Six Sigma Part 1

wdp consulting six sigma

Six Sigma

Part 1

Hill Farmhouse 20 Chishill Road Heydon Royston Herts SG8 8PW Tel: 01763 837111 Fax: 01763 837223

www.wdpconsulting.com


Contents – Part 1

Introduction Six Sigma Model Project Management Process Mapping Measurement Statistical Thinking Basic Statistics Lean Operation Leading Teams


Introduction to Six Sigma So What Exactly Is… Six Sigma? In 2000, Six Sigma was a phrase that was relatively new to the majority of businesses in Britain. Only 16%, according to an exit poll at the Best Factories conference in Birmingham, thought that it was a current manufacturing initiative (reported in Works Management July 2000). However, in the USA it was heralded as the next step on the road to quality improvement and business excellence. Since then Six Sigma has received a higher profile in the UK with many large multi-national companies adopting the methodology. However, Six Sigma is still seen as a methodology for large companies and has struggled to make headway into small medium sized companies. What’s it about? Six Sigma is about understanding the variation within products or processes. Reducing this variation produces a more consistent functionality in a product, or more consistent output from a process – leading to better processes, more reliable products, lower costs and ultimately happier customers. Where did it come from? The origin of Six Sigma goes back to Motorola’s quality initiatives of the eighties. In 1985, Bill Smith of Motorola presented a paper showing that if a product were assembled error free, the product rarely failed during early use by the customer. One of the members of Motorola at that time was Mikel Harry. He set up his own consultancy (Six Sigma Academy) and modified the principles. Harry says, ‘I was just the godfather. I put the math to it and dressed it up for school’. In the mid-nineties, Jack Welch proclaimed that great savings were being made at GE due to Six Sigma. From there, other major American companies joined the Six Sigma devotees. The ideas have since crossed the Atlantic. What is this Sigma and why Six? The term ‘sigma’ is used to describe the spread of the output from a process. As sigma increases the process spread reduces, product reliabilty improves, the need for testing and inspection diminishes, work in progress declines, costs go down, cycle time goes down and customer satisfaction goes up. The benefits can be understood if we consider a car which just fits into a garage. One small shift to either side and the car will be damaged. This depicts the situation where a processes output can just meet the required specification. (This would be ‘Three’ sigmas) However, if we halved the width of the car, getting it into the same


garage becomes far easier. The aim of Six Sigma is to reduce the width of the car such that two car widths would fit into the garage. This would represent ‘Six’ sigmas. Achieving this would, in process terms, reduce the number of defects to as low as 3.4 defects per million opportunities (DPMO).

6σTarget

Six Sigma

Three Sigma

MeasuredCharacteristic

Lower Spec Upper Spec

Variation around target

Understanding variation

Six Sigma Model The generic Six Sigma process follows the accronym DMAIC. Define the goals of the existing system: - choose a process or product that is critical to business success Measure the current system: - select characteristics critical to quality, map and measure the process. Analyse the system: - use benchmarking to identify potential areas for improvement. The simple quality tools and basic statistics are used at this stage. Improve the system: - determine how to significantly change the process and reduce the defect levels. More advanced statistical tools such as Experimental Design may be used at this stage. Control the new system: - once the new process is up and running, standardise and monitor to ensuring that the gains are realised. This may involve the use of Statistical Process Control (SPC).


Six Sigma Training. Typical Six Sigma training programmes rely strongly on statistical methods. These are taught to proposed project leaders (Black Belts) usually in four one week sessions with one session taking place each month. This is complimented by a project applying the techniques learnt. Green Belts (project team members) would receive two, one week training sessions.

6σ

Six Sigma tools

FMEADesign of ExperimentsStatistical Process ControlQFDMeasurement Systems AnalysisBasic StatisticsReliability EngineeringRegression Analysis

Seven Quality ToolsMistake ProofingTaguchi MethodsBenchmarkingHypothesis TestingT-TestsANOVAResponse Surfaces

Subjects covered

Six Sigma Standards Unfortunately there is no standard describing Six Sigma and each provider offers a different view and curriculum. Some companies say a Black Belt qualifies having successfully completed two projects, some say three, others are happy to take your money, giving accredition to those who have simply completed their training course. Motorola, itself, has additional Yellow belts and one consultancy even has the position of Brown belt. One company has combined Six Sigma and Lean Manufacturing to create Lean Sigma! The American Society for Quality (ASQ) has produced its ‘Body of Knowledge for Black Belts’ although this is by no means accepted as standard. Nothing New under the Sun? Those who have lived through quality initiatives during the last twenty years will recognise many of the concepts. So why is this different? In simple terms, there is little difference except many of the successful quality techniques have been brought together. In addition, there is focus on projects with measured improvement, something which has for many years been advocated by Juran.


So what is the secret? Advocates of Six Sigma would probably say that it is the correct application of the steps and tools. However, it is interesting to note that GE along with many others had past initiatives using tools similar to those of Six Sigma that gradually faded out. The key to Six Sigma is that the leadership at the highest level of an organisation drives it. As with any successful initiative, Six Sigma requires top management commitment, provision of appropriate resources and appropriate training but certainly not four weeks of statistical methodology. A positive aspect is that, in most cases, Black Belts are released from their day-to-day work to be full time on projects. In GE, 40% of executives’ incentives are tied to key Six Sigma achievements. In Motorola, Black Belts receive bonuses for project success. Team skills? Knowledge of teamworking and facilitation is essential, and basic project management skills will be needed for any projects that are developed. More advanced tools can be picked up as and when they are needed. Although individuals are trained, it is highly unlikely that they will be able to achieve their project aims without the help of others.

Benefits of Six Sigma As with any improvement methodology for which the appropriate resources and commitment are applied the benefits can be enormous. For many of the large companies small savings can become huge when applied worldwide. On a smaller scale these improvements may not reach billions of dollars but can still be significant. For example; a manufacturer of plastic packaging had tended to make the bags overlarge to avoid the customer receiving a bag that was too short for the application. By using experimental design the factors affecting bag length were understood and the production was able to reduce the average bag length without creating risk for the customer. In fact, the process was more stable and meant that the customer was even less likely to get an undersized bag. The savings on material amounted to about £0.005 per bag. Not a huge saving, except that the company produced 40 million bags per year giving a saving of £200,000 per year. The benefits of Six Sigma are often talked of in terms of savings to the company. However, it should always be remembered that the original aim of Six Sigma was to improve customer satisfaction, reduce errors, improve products and processes and as a result save money.


6σ

Benefits of Six Sigma

Produces significant bottom line resultsRequires and gets top management supportProvides better products and services for customersDevelops and empowers employeesWorks in all parts of the businessSustains gains over timeUses previous tools

Six Sigma Pitfalls Six Sigma cannot be introduced as something that only impacts in one area eg. manufacturing. If it is to be successful it has to be applied throughout an organisation. The key to achieving this is to ensure that Six Sigma is aligned and woven into the organisational strategy, objectives and plans. Although the tools dominate the training it is not simply enough to have the tools without management commitment to changing the way the organisation operates. Six Sigma is in essence a platform for change.

6σ

Six Sigma Success

Advert for Satellite Broadcasts Oct 2001

‘Dr Harry will show why successful installation of Six Sigma has little to do with statistics but, rather, with leadership setting performance goals, establishing valid deployment plans and defining execution tactics…..’


Six Sigma Model The DMAIC model is intended to be cyclical i.e. the improvement cycle never ends. Define, Measure, Analyse, Improve, Control, (Transfer)

Define the Customer, their Critical to Quality (CTQ) issues, and the Core Business Process involved.

• Define who customers are, what their requirements are for products and services, and what their expectations are

• Define project boundaries the stop and start of the process

Define the process to be improved by mapping the process flow Measure the performance of the Core Business Process involved.

• Develop a data collection plan for the process

• Collect data from many sources to determine types of defects and metrics

• Compare to customer survey results to determine shortfall

Analyse the data collected and process map to determine root causes of defects and opportunities for improvement.

• Identify gaps between current performance and goal performance

• Prioritise opportunities to improve

• Identify sources of variation

Improve the target process by designing creative solutions.

• Create innovate solutions using technology and discipline

• Develop and deploy implementation plan

Define

Measure

Analyse Improve

Control


Control the improvements to keep the process on the new course.

• Prevent reverting back to the "old way"

• Require the development, documentation and implementation of an ongoing monitoring plan

Institutionalise the improvements through the modification of systems and structures. Some companies add Transfer after Control. That is; transfer the learning to other relevant areas of the company.


Six Sigma Checklist (adapted from P.S. Pande “The Six Sigma Way”)

Define Has the project team ….

1. Confirmed that the project is a worthwhile improvement priority for the organisation and is supported by the business leaders?

2. Been given a brief project rationale explaining the potential impact of the project on customers, profits and its relationship to the company’s business strategies?

3. Composed and agreed a two to three sentence description of the situation as currently seen?

4. Prepared a goal statement defining the results expected from the project, with a measurable target. (note: results not solution)?

5. Completed a Team charter including a list of constraints, team membership, a preliminary plan and schedule and process scope?

6. Reviewed the charter with the Champion / Sponsor and confirmed their support?

7. Identified the primary customer and key requirements of the process being improved and created a SIPOC diagram of the areas of concern?

8. Prepared a detailed process map of the areas of the process where the initial measurement will be focussed?

Measure Has the project team….

1. Determined what needs to be understood about the process and where to go to get the answer?

2. Identified the types of measures to be collected and have a balance between effectiveness / efficiency and input / process / output?

3. Developed clear, unambiguous operational definitions of the things to be measured?

4. Tested the operational definitions with others to ensure their clarity and consistent interpretation?

5. Made a clear, reasonable choice between gathering new data or taking advantage of existing data?

6. Clarified the factors needed to facilitate the analysis of the data? 7. Developed and tested data collection forms to ensure the collection of

consistent and complete data?


8. Identified an appropriate sample size and sampling frequency to ensure a valid representation of the process under study?

9. Prepared and tested the measurement system, including training of collectors and assessment of data collection stability/

10. Used data to prepare baseline process performance measures, including proportion defective and yield?

Analyse Has the project team …

1. Examined the process and identified potential bottlenecks, disconnects and redundancies that could contribute to poor performance?

2. Conducted a value and cycle time analysis, locating areas where time and resources are devoted to tasks that are not critical to the customer?

3. Analysed data about the process and its performance, identified reasons for variation in the process and identified possible areas for improvement?

4. Evaluated whether the project should focus on process design or redesign, as opposed to process improvement and confirmed the decision with the Champion / Sponsor?

For Process Design / Redesign

5. Ensured that the key areas of the process are understood so that a new process can be created to meet the needs of the customer efficiently and effectively?

For Process Improvement

6. Investigated and understood whether they are dealing with Special Cause or Common Cause variation?

7. If Special Cause, identified and verified the root cause? 8. If Common Cause, identified the significant factors influencing the variation in

the process?

Improve Has the Project team …

1. Created a list of innovative ideas for potential solutions? 2. Identified clear criteria for a successful solution? 3. Made a final choice based on the above criteria?


4. Verified that the solution will give the required improvements? 5. Agreed with the Champion / Sponsor and received the go-ahead? 6. Developed a plan for piloting and implementing the solution including a pilot

strategy, action plan, results assessment, schedule etc. 7. Evaluated the results and confirmed that the results in the Goal statement

can be achieved? 8. Identified and implemented refinements to the solution based on lessons

learnt from the Pilot? 9. Created and put in place a plan to completely implement the solution? 10. Considered potential problems and unintended consequences of the solution

and developed preventive and contingent actions to address them?

Control Has the Project team …

1. Compiled results data confirming that our improvement has achieved the Goal defined in the Team charter?

2. Selected on-going measures to monitor performance of the process and continued effectiveness of the solution?

3. Determined key charts / graphs for a process ‘Scorecard’ on this process? 4. Prepared all essential documentation of the revised process, including key

procedures and process maps? 5. Identified an ‘owner’ of the process who will take over responsibility for the

improvement and for managing continuing operations? 6. Documented the team’s work and data collected during the project? 7. Forwarded other issues / opportunities, that were identified during the

project, to senior management? 8. Celebrated the work and efforts of the team?


Project Management Introduction These notes have been compiled to give the reader an introduction into project management skills. It should be used as a reference point and as a guide. It cannot tell you exactly how to manage projects; that is down to an individual’s style and aptitude. What it does do is give some key pointers, and things to think about. The picture below shows a juggler. Project management is often like juggling. A good juggler will have an awareness of their own ability, the mood they are in when juggling, the people who may be juggling with them and those who may just be watching. The balls they have in the air relate to different things they are trying to keep in balance at any one time. At all times the juggler must be aware of where the balls are, and will be catching and throwing them up in the air again so they do not fall in the ground. As a project manager, you will have to do very similar interventions, albeit rarely with balls!! What you should not do is expect the impossible, how many jugglers have you seen spinning 20 plates at the same time as juggling 5 chainsaws?


Definition of Project Management Projects come in different forms. Many people think of a project as being the Channel Tunnel or British Library or other large engineering projects that hit the news headlines, usually for being grossly overspent!! Projects can be both large and small. There are a variety of definitions of a project, which include: -

“A human activity that achieves a clear objective against a timescale. Projects nearly always have the following characteristics - one clear objective, a fixed timescale, a team of people, no practice or rehearsal, and change”

“A human endeavour that creates change; is limited in time and scope; has mixed goals and objectives; involves a variety of resources; and is unique”

All projects have to be managed. Project management has been described as, “the process of planning, controlling, and managing people as a temporary team. This is no different in Six Sigma. Different Types of Project “Many traditional definitions of projects assume that the objectives of a project, and the methods of achieving them, are well understood throughout the project.” Projects can be judged against two parameters: “whether the goals are well defined, and whether the methods of achieving them are well defined”. This concept leads to the definition of four types of project, see diagram below: -

Type one projects: goals and methods are well-defined - typically large engineering projects. Type two projects: goals are well defined but the methods of achieving them are not e.g. product development projects.

Greater chance of failure

Goals Well Defined

Methods Well Defined

No

Yes

Yes NoGreater chance of success

Type 2 Project

Type 1 Project Type 3 Project

Type 4 Project

Product developmentt

Research & Organisational

change

Applications software

developmentEngineering


Type three projects: goals are not well defined but the methods are - e.g. software development projects where the user requirements are hard to specify. Type four projects: neither the goals nor methods are well-defined - e.g. typically organisational development projects.

Some people argue that the best way of achieving these projects is to move the project into the type three or two category (preferably type two), and then into the type one category as quickly as possible, thereby fixing the goals and methods. Milestones become key decision points and configuration management is suggested as a way of managing this. The conventional view of projects as complex, but well defined, sequences of activities to deliver clearly defined goals and objectives may be invalid for a large number, if not the majority, of projects. Whatever the type of project, one of the first things the project manager must do is to identify the external environment, particularly who the stakeholders are, in order to build appropriate relationships and satisfy competing demands. The Development of Projects Projects can come from a variety of sources. They may be part of the logical strategy of an organisation and be created on an annual basis, programmed into the ongoing efficient running of an organisation, and many large projects will often fall into this category. There are also other sources that may drive projects. For example, an operation manager may see the scope to introduce technology to make the operational process much more efficient. This will inevitably create a project to implement the technology. External influences can also create the need for projects. If there are plans to change legislation then organisations often have to react to this, and create projects to deal with the external changes. Above all else customer demands may create the need for projects. Some of these demands will be focused in an organisation’s annual strategy, but others will be much more immediate. For example, if service provision falls down to an extent that a great deal of bad publicity is created, then customers will place demands upon the organisation to make changes in the way the service is delivered. There are four key stages in the life of a project.

WDP

Proposal and initiation

Design and approval

PROJECT LIFE SPAN

BIRTHImplementation and control

Finalisation & control & evaluation

GROWTH

MATURITY

DEATH


The first is the Birth stage. This will be where the seeds for the project are sown, and the outline proposal documentation is drawn up. This may be the end of the project as it quickly becomes apparent that it is a not feasible. However, if the project passes successfully through the first stage then it moves into the Growth stage. During this phase the project design is made clear, the scope of the project agreed and people understand the likely impact on resources. This is followed by the Maturity stage of the project wherein the plans will be implemented, monitored and controlled. Even at this stage of the project, progress may and should be aborted if things change so radically as to make the project no longer feasible. One of the worst crimes of the project manager is to ‘press on regardless’. Analogies can be drawn with the Titanic, where despite iceberg warning the ship continued on at high speed, only to suffer drastic consequences. The Death stage of the project marks the end of the project, hopefully completed, and will be the time that project evaluation and learning takes place.

Defining The Project As already described, the definition of the project will be paramount to a successful completion. There are four key elements to this: - Scope, Stakeholders, Self and Team. Charter / Scope Writing a scope document that incorporates the necessary information will mean that there will be a formal agreement between Project Manager and the Project Champion as to what should be delivered. This document will become ever increasingly useful to the Project Manager as the project progresses. It should be a document that is constantly referred to and ensures that all work is focused and that only the right amount of work is undertaken, thereby ensuring an effectively managed project. Within Six Sigma this known as the Project Charter. Areas to be considered: Project Title: this should give others a quick view of the project and use key words or phrases. Project Leader: who is leading the efforts? Champion / Sponsor: who is responsible for “creating the environment” for the team leader and team to undertake the project. Project Start date: the date the project formally started.


Project End date: the Champion will probably set the anticipated project end date. The time allowed for the project should give enough time to complete the project given its complexity and current business conditions. Cost of Poor Quality: quantifying the total cost of quality can be difficult but there may be information on such things as scrap produced, excess hours spent etc. This will give an idea of the potential savings and size of the project. Process: which process is to be improved? Why does it need improving? How will it affect the customer? I.e. what is the need for this improvement? Project boundaries: the team cannot solve everything. The team needs clear boundaries to help it focus on the extent of the project. Project objectives: what is the project expected to achieve? Are the objectives SMART? Process Measurements: what are the measures that will determine the success of the project? Team members: who should be part of the team? Subject matter experts? Project Milestones: what are the major milestones within the projects? Once the scope is defined, the process of scope management begins and this is covered later.

Stakeholders The management of a project’s “stakeholders” means that the project is explicitly described in terms of the individuals and institutions that share a stake or an interest in the project. Stakeholders can include a variety of people, often outside the

WDP

OBJECTIVES

Specific

Measurable

Achievable

Relevant

Time based

wdp consulting

control of the project manager and therefore presenting serious management problems. It is only by identifying what is important to each of the stakeholders in terms of both hard and soft expectations, that the project manager can develop options for managing any interference or conflicting expectations and for managing the project to ensure the stakeholders are satisfied at its conclusion. There are few prescribed tools and techniques for the management of stakeholders, and what does exist is often based on simple forms identifying the stakeholders and their level of power and interest, as demonstrated in the model below: It is critical thaway will all extregarding stakeand dovetail thproject.

Self Only by knowimake an effectpossibly using able to see themap how this ensure the proplanning. The following

What exper

What are y

Key

Low

Level of interest Low High

Keep informed

Keep

Minimal effort

Power

six sigma

t all stakeholders are identified and none overlooked. Only in this ernal influences be considered. While deciding the approach to take holder management, it is necessary to look at the project life cycle, e two accordingly since stakeholders may appear during the life of the

ng him or herself / herself will the Project Manager ensure that they ive contribution to the project. By being realistic about this, and some assessment tools to aid the process, the Project Manager will be ir own strengths and weaknesses. This information can be used to might affect the project and then the necessary action can be taken to ject does not falter due to a weakness in a key skill area such as

questions will be useful in assessing this: -

iences have you had?

ou good at?

players High satisfied


What are you not so good at?

What do you want to develop?

How will you measure this?

The strengths and weaknesses identified should then be matched against expected role and performance. Strengths are often easier to build on and identify whereas weaknesses are often hard to admit to, accept and then either develop or resolve. The development of a weakness may well take some time, and so the project manager should seek to overcome a weakness by using the skills of others where necessary. Admitting to the weakness does not have to be a part of this, although doing so may well allow the designated person to act as a mentor for the development of the skill in question. Teams Teams are an essential feature of the work pattern of almost any organisation. Once in a team, the people within it influence each other in many different ways. The team may even develop its own hierarchy and its own leaders. Both may be different to that of the organisation from which it originates. Group pressures can have a major influence on the behaviour of individual members within that team and their performance to the team’s overall aims. The activities of the team can also be closely associated with the process and style of leadership adopted by the leader / facilitator of the team.


There are two different types of group: formal and informal. Formal groups are created to achieve specific organisational objectives and concerned with the co-ordination of work activities. People are brought together on the basis of defined roles and the structure of the organisation. The nature of the task to be undertaken is an important feature of the formal group with goals being identified by management and certain ground rules, relationships and perhaps norms of behaviour established. Informal groups are based more on personal relationships and agreement of group members than on defined role relationships. These groups tend to satisfy psychological and social needs, not necessarily related to the task being undertaken. These groups tend to devise ways of attempting to satisfy member’s affiliation and other social motivations that are perhaps lacking in the work situation. The membership and roles of informal groups can cut across formal structure. It is very unusual for an informal group to be exactly the same as a formal group. Understanding the informal group network can be critical in determining how to facilitate formal teams. The performance of a team is essentially its effectiveness. This is influenced by: (1) the givens - the team, the task, the environment (2) the intervening factors - leadership / facilitation style, motivation, processes and

procedures

leading to (3) the outcomes - productivity and member satisfaction. The Givens The size of the team is an important factor in its performance. As team size increases, problems arise with communication and co-ordination. If a team becomes too large it may split in two sub-teams with friction developing between them. The optimum performance of team size is probably five to seven people. Teams with more than seven or eight permanent members start to split into sub-groups, and can under perform. For larger projects it therefore makes sense to have teams and sub-teams for different work packages, with the occasionally ‘whole project’ team briefings to manage communication. Compatibility of members of a team is also important. The more homogenous the team in terms of background, interests, attitudes and values, the easier it is to promote cohesiveness. Differences in personality or skills in the team may actually complement each other and help the overall performance. However, they can also cause difficulties and conflict may arise when members are in competition with each other. If teams are too cohesive, they can suffer the phenomenon of ‘group think’, where people are so in tune with what the others are thinking they do not always see the broader picture and may overlook something incredibly obvious.


The individual roles in a group can therefore be important. Belbin has done some study into this and defines a variety of roles within a team. For example, the Shaper will be the task leader and is needed to spur action; the Plant will be introverted but have many ideas; the Finisher will check details and deadlines. When setting up a team it is useful to examine individual strengths and weaknesses and recognise the most beneficial role each team member can play. For example, if someone is particularly good at planning, but not so good at communication, they will be more suited to the co-ordination of the project plan, rather than primary liaison with stakeholders. It has been found that the more complex the task is, the more important it is for the team to work well together and the need for consensus on a focal person or leader. Good performance therefore requires that a leader / facilitator should be picked because of ability to direct and influence rather than status. Hidden agendas can give rise to conflict, and will prevent trust from building. Hidden agendas may be protecting the interest of one sub-group, impressing the boss, making a particular alliance, or even covering up past errors. For as long as hidden agendas exist, team performance is likely to be hindered. It is important for the team leader / facilitator to reveal hidden agendas, either by talking to individuals separately, or preferably encouraging members to admit their own issues to the team. Once recognised, hidden agendas can be overcome by careful allocation of task The type of task will naturally affect the kind of team that is formed. The benefits of task allocation are that it allows a structured approach. The team that is formed to formulate ideas about the detail of the task needs a more supportive management style and may not initially have a structured approach to its work. The timescale of the task also reflects how the team will work together. If the timescales are very tight, then the team may not develop to deal with individual needs. This can result in work of a lower quality than if the timescales were broadened. However, performance in terms of quantity may be a lot higher with team members motivating each other. Another important factor affecting team performance is the ‘salience of the task’, the more important the task is to the individual within the team, the more committed they are to achieving it. This enables greater demands to be placed upon the team. It is important that the task is clear and unambiguous. This will assist the team through the stages of team development. However, if the task requires ambiguity, then this can lead to the formation of different roles within the team and also increase stress, therefore having implications for the leadership of that team. The timescale set should therefore be realistic so that good performance can be achieved, but not too long as procrastination may occur.


Team Development When examining the components of team performance, it is also important to consider team development. Teams, like individuals, have a growth cycle. The team will function differently at various stages of development, and performance expectations must be altered accordingly. A model that is frequently used splits the development into four stages:- Forming this is the initial stage of development. Participants can be anxious, testing to discover the nature of the situation, trying to discover what help is available from the leader or facilitator, and what behaviour is or is not appropriate. Participants try to discover the nature of the task, and the ground rules. Storming conflict begins to emerge between individuals or sub-groups; the authority of the leader, the make up of the group, indeed its purpose for existing is challenged. Discussion causes differences of opinion to emerge. Norming the group begins to harmonise; cohesion and unity begin to be experienced by participants; norms emerge for handling conflict; mutual support develops. Plans are made and work standards laid down. Communication of views and feelings begins to develop. Performing the group structures itself or accepts a structure that fits most appropriately a common task. Roles are established and accepted in relation to the task. The group begins to apply more of its energy to being effective in the area of its common task. This results in progress and achievement. Performance will improve if teams can ‘get through’ the first stages as quickly as possible. Teambuilding exercises may help with this, and should not be overlooked even if the team already know each other. It is also important to recognise that teams need to be maintained, and team-building exercises should not just be saved

WDP

BUILDING THE TEAM

4 stages of team developmentformingstormingnormingperforming

motivation is key

Balancing needs

Task needs

Individualneeds

Groupmaintenance

needs


for the beginning of a team’s task, but undertaken at various intervals, as appropriate, throughout the life of the project. The Forming Stage During the initial stage of development, the primary concern of new team members is to be included. They typically wonder: ‘How do I fit here?’ ‘what’s expected of me?’ They may think that if they speak out they’re misbehaving, so they don’t express their negative feelings – they don’t want to rock the boat. They look to the facilitator for guidance and assistance. That can be the beginning of a long-term dependency. Here are some things you can do to encourage team responsibility from the start. Help the team develop ground rules, and explain their use. Encourage team members to assume responsibility not only for adhering to the ground rules, but also for enforcing them. Provide a mini lesson on how to intervene regarding breaking the rules: • make a statement or ask a question that describes the inappropriate behaviour,

and correlate it to the ground rules. Example: Joe, one of our ground rules if no cheap shots. When Sarah offered her idea, you said ‘You need to get out of the office more, Sarah.’

• invite comment. This step isn’t always necessary, but it can be used to help someone recognise his or her behaviour and its effect on the team

• request that the person change his or her behaviour to be more consistent with the ground rules, or suggest the desired behaviour. Example: What did you intend by that? Or, I felt that comment was a cheap shot. Would you be willing to hear how Sarah’s idea might work?

Clarify roles Working with the team, define and spell out in specific terms (preferably written) the roles of facilitator, team member, and team leader. That not only establishes immediate ownership for the success of the team, but also individual responsibility for fulfilling roles and expectations. Check progress Team progress or process checks are an easy way to encourage team ownership of results and how the team works together to achieve them. Progress checks are led by the facilitator and are usually the last item on a meeting agenda. Each member is asked to respond to a few questions – such as; What did we do well? What did we not do well? What do we need to do differently the next time? The facilitator asks the questions one at a time. Team members take turns answering, and then the facilitator offers his or her view. The Storming Stage Once team members are oriented, they should focus on being heard and having control. At this stage, conflict is often evident. As facilitator, your focus should be on group dynamics and positive confrontation to help team members find workable ways to address their differences. This is the stage at which many teams (and facilitators) either make it or break it.


Get the team to own responsibility for recognising and addressing difficulties. A key role of the facilitator during this stage is to play back the specifics of what’s happening. Example: Paul, you just said to Eileen ‘There’s no way I’ll go for that’ and suggested another alternative. Before you talk about how your idea will work, would you be willing to paraphrase what Eileen proposed and state your concerns? Use graduated interventions. Make observations aloud to get team members to see the effect of their actions or behaviours on others personally and on the team and its tasks. Example: Patti, that’s the third time you rolled your eyes as John has responded to that? (Patti’s response: ‘Yes, he said, ‘If you cared about anything other than your promotion, you wouldn't be so blind.’ How do you think these exchanges between the two of you are affecting the team?). The Norming Stage By now, team members have worked their way through the struggles in the storming stage. They have become comfortable giving and receiving feedback, taking risks, and taking steps to balance everyone’s participation. Now, in the Norming stage, they should focus their energies on being open and encouraging others. Your primary focus as facilitator is to uncover unspoken issues and promote the full exploration of ideas. You should make such statements as: • Why do you think you were so effective today? Or, Why do you suppose you

had such difficulty? • I sense that you’re not comfortable with what has been proposed, Ron. You

haven’t said anything in the past 10 minutes. What’s going on? • Janet, each time the group mentions the benchmarking visit, you drop your eyes

and start to doodle, Were you aware you were doing that? What are your concerns?

This stage is also a great time to provide idea-generating tools that stretch people’s thinking or encourage differences of opinion. Devil’s Advocate is an exercise in which one team member is appointed the devil’s advocate and has to argue why certain ideas or proposals won’t work. The Performing Stage This stage leads to the final handoff. The definition of final handoff, however, depends on the type of team. But in general, a handoff means transferring most or all of a facilitator’s typical duties to the team. Your key focus is to encourage the team to become self-facilitating. It’s here that the ties that bind are cut. The process of letting go, although simple, is difficult for many facilitators. There are two reasons why some don’t let go and the dependency deepens. One it feels good to be the expert or advisor. Two, the more skilled a facilitator is, the more second nature his or her responses are. So when an issue crops up that the team could handle on its own, the facilitator steps in automatically.


The process for empowering a team to become self-facilitating is Socratic. By asking the right questions, you can get a team to recognise what’s happening (observe its own process) and determine what to do. It’s here that you need to draw on your questioning abilities. These six steps will help you hone them. 1. Ask a question that causes the team to recognise what’s happening.

Example: What’s happening right now? (Team response: ‘We’re foundering. We probably need to use some kind of process to hash this out.’)

2. Check Members’ understanding. Rather than give the answer, ask a question that will help you determine whether team has the skills and knowledge to address what’s happening. Example: You say some kind of process will help. What process? What do you think you need to do?

3. Check your understanding. This means paraphrasing what was said. Example: You said you could probably look at the pros and cons of each idea, one at a time.

4. Ask outcome questions. This step involves helping the team see the link between the proposed approach and the desired results. Rather than correct team members or elaborate on their ideas, ask questions that encourage them to test the idea. Example: How will that help you overcome the difficulties you were having? (Team response: ‘It will force us to look at one alternative at a time and thoroughly explore it.’) Will that provide you with all of the information you need to make an informed decision? (Team response: ‘Probably not. We need some data from X, and we should probably see how our customers would feel about it.’)

5. Offer relevant observations and experience, if needed. Wait until this step to offer your suggestions, assistance, or expertise. If the team can’t resolve barriers to success using steps 1 through 4, ask leading questions that will determine its direction. Example: I think those are both good approaches. Is there any value in checking those alternatives with some others outside the group?

6. Ask ‘What will you do’. This final step gives you an opportunity to check the end result and action plan. If those are OK the team then puts its plan to work.

Knowing that a team depends on you can be satisfying. Ironically, the skills for putting yourself in the background are greater than for staying at the forefront. Team independence is the best measure of a facilitator’s success. A mediocre facilitator tells a team what to do, a good one demonstrates and encourages, a superlative one inspires and motivates. But the absolute best facilitators are able to step aside and allow a team to take charge.


Project Planning & Control Project planning is a form of management, evolved and established to co-ordinate complex activities - some man-made ones still survive today, For example, the Egyptian pyramids. However it has not been until the later part of this century that many of the techniques have been developed, partly as a result of nations’ competition for supremacy in weapons and defence systems. Although variations exist, there is a core set of techniques that are commonly used. Before looking at the various ways in which plans can be developed it is worth understanding the sequence of planning. It is also worth noting that plans should be developed with the project team, and other relevant stakeholders. The plan should be targeted at its audience, which will often mean that different levels of a plan will exist. For example, senior people in the organisation will not be interested in the detail of the tasks to be undertaken, but they will be interested in the overall picture and some of the key milestones to be achieved. The methods and tools for planning projects can be as follows: - Work breakdown structure (WBS) this involves the division of an activity into sub activities and sub-sub activities in order to control the project effectively. It is something that provides better control of work definition; it allows work to be delegated in coherent and often independent packages; it allows work to be defined at an appropriate level for estimating and control; and it allows risk to be contained more easily. It simply follows the principle that when faced with a mountain, turn it into a series of molehills. Milestone plan this is a tool that allows the project manager to show how the intermediate deliverables build towards the final objectives of the project. It

WDP

SEQUENCE OF PLANNING

Consider the ‘big picture’

Break the project into sub projects

Determine the relevant milestones

Break each sub project down

Clarify resources


therefore allows different members of a project team to see the interfaces between their activity and others. Milestones are useful to show the logical sequence of events, highlighting dependencies. Where these are not always clear, the milestone plan is likely to be much more broadly focused. Defining milestones can be a tricky task. They do not show how something is to be done, but what is to be achieved. An example of a milestone would be ‘arrival at work’, but not ‘when I have driven to work’; the reason for this is that you could arrive at work by bus should your car breakdown. This can get quite complicated, but the key issue is ‘what will have been achieved’. Once milestones have been defined a result path can be drawn to show the interdependencies between tasks, as we see illustrated here. Responsibility charts can be described as a matrix with tasks shown as rows and organisational resources as columns. Symbols are placed in the body of the matrix to represent the involvement of each resource type in each work element. This method can be invaluable in managing and allocating known tasks. If tasks are predictable they can be developed in advance, if not, as is often the case in open projects, they can be used over shorter time frames, but still providing a useful tool for the project manager.

WDP

activities (subWBS3)

WBS1 WBS2 WBS3

tasks

interdependencies

milestone

result path

MILESTONE PLAN

WDP

RESPONSIBILITY CHART

\\\\

Companies / Departments / Functions

Resources / People

Milestones

Activity / TaskRoles / Responsibilities


Time is often seen as a critical element of projects, and a classic tool for managing time is the critical path method (also known as programme evaluation and review technique). The critical path shows the overall duration of the job. It is the longest path of the project and therefore the shortest time in which the job can be done. Any delay in the critical path activities will cause the project to overrun. Gantt charts, or bar charts also show time. They depict the projects schedule by showing the duration of an activity, its planned start and finish dates and the logic of events. See example below: - If plans also calculate the early start and finish, the late start and finish and the float of elements in a project, ‘what-if’ analysis can be more readily conducted. Some projects are time critical others are not. Managing to achieve deadlines is most critical, even for open projects where the stakeholders will often want to see something by a certain date - they may not be sure what or how, but something!!! The use of time charts as named above allows people to see what is going to happen and when. Clearly if activities, start dates, logic etc, is unknown then it can be difficult to draw. Another traditional technique is the Cost specification methodology that focuses on three objectives, managing cost, scope and organisation. Cost can be measured through a cost breakdown structure or project financial plan. Estimating costs at the start of a project clarifies the total budget required and predicts cash flow requirements. Estimating project costs can be achieved through careful planning, assessing historic projects and group brainstorming, or preferably a combination of all three. The costs that occur can be split into a variety of categories - Labour; Materials; Plant and Equipment; Subcontractors; Management; Overheads and Administration; Fees and Taxation etc. Controlling and assessing variance enables plans to be made to manage some of the inevitable overruns!! It is a simple technique really, but needs to be linked to the WBS, and to control costs you must compare the actual expenditure not to the schedule of expenditure, but to some measure of the value of the value of work done. The WBS provides the means to so this. Some projects will be very capital

GANTT CHART

ID Task Name Duration Start Finish Work1 Task A 1 day Wed 10/7/98 Wed 10/7/98 1 day

2 Task B 2 days Thu 10/8/98 Fri 10/9/98 1 day

3 Task C 8 days Wed 10/7/98 Fri 10/16/98 9 days

4 Task C1 2 days Wed 10/7/98 Thu 10/8/98 3 days

5 Task C2 2 days Fri 10/9/98 Mon 10/12/98 1 day

6 Task C3 2 days Tue 10/13/98 Wed 10/14/98 2 days

7 Task C4 2 days Thu 10/15/98 Fri 10/16/98 3 days

8 Task D 1 day Thu 10/15/98 Thu 10/15/98 0 days

9 Half Way Party !! 0 days Thu 10/15/98 Thu 10/15/98 0 days

10 Task E 3 days Fri 10/16/98 Tue 10/20/98 8 days

11 Task E1 1 day Fri 10/16/98 Fri 10/16/98 4 days

12 Task E2 1 day Mon 10/19/98 Mon 10/19/98 1 day

13 Task E3 1 day Tue 10/20/98 Tue 10/20/98 3 days

14 Task F 1 day Mon 10/26/98 Mon 10/26/98 1 day

15 Task G 1 day Tue 10/27/98 Tue 10/27/98 2 days

16 Task H 1 day Wed 10/28/98 Wed 10/28/98 1 day

10/15

M T W T F S S M T W T F S S M T W T F S S M T W TOct 4, '98 Oct 11, '98 Oct 18, '98 Oct 25, '


intensive whilst others will not be. OD projects by their very nature are less capital intensive, and the costs can often be, and are often, hidden - For example, people’s time (it is often just fitted in!!). The costs will frequently arise in the expertise bought in to assist with certain interventions. Complete cost predictability is therefore impossible until the tasks are known! Quality can be managed through a variety of means. At the start of the project the manager should draw up a quality plan to define how quality will be achieved, how the company’s procedure will work on this project, and how the manager intends to assure and control quality. Quality assurance systems, such as ISO9000 are not a product standard, but a management standard system, relating to how quality is provided. It provides a framework of procedures with which to achieve quality, but it is a means to an end and should not be seen as anything more. It is geared to the achievement of quality of production, not the production of quality. In order to assure the quality of the project there needs to be a clear specification with defined standards, appropriately qualified people, periodic reviews and reasonable change control. A project handbook may well be used, either as part of the organisations approach to projects generally or specifically produced for major projects. Quality control is the process of diagnosis and cure. As the facility is erected and commissioned it is checked against the specification to ensure it is of the required standards, and any variances are eliminated. The activities by which this is done must be planned, tested, recorded and analysed. Techniques are employed to produce the best product and service possible - through constant innovation. Three tools that can be adapted from regular initiatives include the Cause and Effect diagram, Pareto's 80/20 rule and ‘Just-in-Time’ methodologies.

WDP

QUALITY

Give the customers what they want!Achieve by:

Carefully developed & managed processes

An inbuilt attitude of mind

‘We never have time to do anything, but wealways find time to do it twice!’


Risk management – There are probably five sources of risk, external-unpredictable; external-predictable; internal - technical; internal - non-technical; and legal. The techniques that can be used to identify these risks include - expert judgement; plan decomposition; assumption analysis; decision drivers’ analysis; and group brainstorming. These can then be reduced by avoidance, deflection or contingency planning. Whilst risks exist in every project, the above categories may not be appropriate to OD projects; For example, internal-technical risks may not be relevant to a quality improvement programme. Risk can be identified by using the simple table shown. Project Management Information Systems a PMIS is used for collecting data from across various functions, and analysing and presenting those data in a form suitable for all the parties involved in a project. The systems are used for planning and controlling projects, throughout the project management lifecycle, providing timely control information. The systems are also used for recording historical data for estimating future projects. Microsoft Project is an example of a computer software package that can help in the management of a project. It should not however be assumed that by purchasing such a system, a project will automatically be well managed and planned. A typical system has two key purposes, to act as a planning system and also as a control system - as illustrated below.

THE PLANNING SYSTEM

• scope • organisation • time • cost • quality

THE CONTROL SYSTEM

• purpose • resources • management

data structured information

management decision and direction

WDP

RISK

Affected by:technology - new and oldtimepeople - skills andteamworkuncertainty

Overcome by:avoidance - identify andreplandeflection - pass the buckcontingency

Outcome Simple Some NewProcess innovation

Simple low risk high

MediumHigh high high riskcomplexity


This sort of system builds, and relies upon much of the data set-up in the ‘scoping tools and techniques’ (as illustrated in the previous diagram), and therefore the same arguments apply when considering their usefulness with open projects. Summary A key part of Six Sigma is the running of projects to achieve the improvements within a process or product. Thus, it is necessary to use the knowledge developed from running projects to help ensure that the Six Sigma project is successful. The key parts of the project are : Defining the Charter Selecting the most appropriate team Planning and monitoring the project steps using tools described Communication with all stakeholders

Process Mapping


A Process Map is a graphical representation of a process, showing the sequence of tasks or events using a number of symbols. (see appendix ). Mapping can be applied to anything from the movement of an invoice to the flow of materials in manufacturing or the steps involved in making a sale. You will discover that individuals will have a different understanding of what the map looks like and how the process works. Hence maps should be created by a team rather than by an individual. The objective of creating a process map is to help understand the process, highlight where problems exist and ultimately improve the process. A good map is the foundation for continuous improvement. What is a Process? A process is a set of steps/tasks/activities that have a beginning and an end. Another definition could be; a group of logically related tasks, decisions or activities which together produce a result or output. In its simplest format it can be seen as: INPUT PROCESS OUTPUT The process transforms the inputs into the outputs. This simple diagram can be expanded to include Suppliers and Customers. SUPPLIERS INPUTS PROCESS OUTPUT CUSTOMERS This diagram is sometimes referred to as SIPOC. Customers and Suppliers may be either internal or external to the organisation. A further development of the diagram brings in reference to Controls and Resources. CONTROLS SUPPLIERS INPUTS PROCESS OUTPUT CUSTOMERS RESOURCES Definitions:


Process – any activity that transforms inputs into outputs Inputs – material and/or data transformed by the process to create the output Outputs – the results of the transformation of the inputs Controls – regulators or influences on the process Resources – people, equipment or material not transformed by the process Supplier – individual, department or organisation that supplies the inputs Customer – individual, department or organisation that receives the outputs In looking at a process for the first time it can be useful to list the elements under each of the S.I.P.O.C. This helps remind the team of aspects of the process that need to be considered. E.g. have they considered the requirements of all the customers listed?

Suppliers Inputs Process Customers Requirements

Step 1:

Outputs

Step 2: Step 3: Step 4:

See Below

Step 5:


Suppliers Inputs Process Outputs CustomersManufacturer Copier Copies You

Office SupplyCompany Paper File

Toner Others

Yourself Original

Making aphotocopy

Power Company Electricity

Put original on glass

Put original on glass Close LidClose Lid Adjust

SettingsAdjust Settings

Press STARTPress START

Remove originals

and copies

Remove originals

and copies

Process Steps

SIPOC Example

Levels of a Process Your position within an organisation will affect how you see a process. A process can be shown at a number of levels. High Level Process (Core Process) E.g. introduce new products to the market Within this there are Key Processes e.g. Launch Product. Key Process (Major Process) E.g. Launch Product Each of theses processes can be further broken down into Activities or Tasks.

Identify Market Need

Develop Product Design

Develop Production Capability

Launch New Product

Prepare Sales Literature

Train Sales Team

Pilot Sales ofNew Product

Sell New Product


Sub-Process E.g. Prepare Sales literature

Activity / Task E.g. Design brochure If needed even each of these could be broken down further. It is important to agree from the start at which level the process is being considered. For most improvement situations, the more detailed activity / task level will be used. However, this may not always be the case. For example it may be decided that a radical overhaul of the method of identifying a market need is required. Whatever level is being considered, it is important that an accurate map of what currently happens is created. It is very tempting for teams to produce a map of what should happen according to procedures or what they would like to happen. Honesty is a key requirement before any real improvement can be made. Example A chemical process was showing high and at times irregular variability. When the Supervisors were consulted over this they had no answer even though some of the changes appeared to be shift related. It was only during conversations with operators that it became known that at the start of a shift, once the previous Supervisor had left, the new Supervisor got out his little black book and changed the process settings. Initially, the Supervisors denied they did this and it was only under pressure from Operators saying they did, that it was acknowledged. Once the Supervisors stopped changing the process that they were able to move forward and improve the process which ultimately led them to running a line for several days without any breaks or changes; a situation unheard of before.

Design Brochure

Agree Design

Print Brochure

Distribute Brochure

Write Text Agree Layout

Prepare Artwork

Produce Draft Brochure


Link to Key Business Objectives There are many processes within an organisation, thus if significant improvements are to be made, then there needs to be an understanding of which processes are linked to which key business objectives. Amongst the key business objectives should be those that directly relate to serving the external customer. Make sure that the objectives are not purely internal. A decision making process (see Tools) can then be used to identify the most critical process i.e. the process having the greatest impact on the business objectives. Process Mapping Steps Step 1: Decide on the process to be mapped Step 2 Decide where the process begins and ends (boundaries) Step 3 List the outputs and identify the customers for these outputs Step 4 Determine the requirements for these customers Step 5 List the inputs and their suppliers Step 6 List the main activities and decision points in the process Step 7 Arrange these activities in their proper order, using arrows to show the direction of flow Step 8 Agree this represents the current ‘as is’ process

Customers and Suppliers (Steps 3 – 5) In drawing the Process map it is important to recognise those who may have some interest in this process, i.e. customers, suppliers and other stakeholders. Think of a process that you are involved in: Who are your main customers?


Do you know what their requirements are? Are your customers aware of what you can or cannot do for them? Who are your main suppliers? Do you know what their capabilities are? Are any of your customers also your suppliers? Do any other groups have an interest in your process? Too often organisations produce Process Maps without any reference to customer needs and expectations.

Map the Process (Steps 6 – 7) In order to display the process than different methods may be used. Systems Flowcharting Operations Flowchart (See appendix for description)

10

20

30

40

50

wdp consulting

Example: Washing dishes after a meal

Clear dishes fromtable

Food on dishes?

Stack dishes in dishwasher

Add detergent

Set dishwasher controls and turn on

o

yes
Dispose of food?
Store food

o

yes
Put food in bin
N
N
six sigma


Improving the Process Map Simply getting a team to draw a Process Map can lead to the identification of areas for improvement. For example discovering output that goes nowhere or realising some things could be done earlier in the process. However, on studying the final Process Map it is necessary to ask whether or not this represents the ideal process flow. If the answer is no, then improvements must be made. In many cases it may not be possible to answer this question immediately and it is necessary to collect data from the process in order to identify areas for improvement.

Measurement The measures chosen should relate to customer requirements or necessary controls required to ensure that the process works correctly. Typical areas of measurement might be: Effectiveness: output measures of your customer’s requirements Efficiency: measures of the amount of time, cost or labour Supplier’s effectiveness: how good is the input material or information? Process controls: how stable is the process? In deciding on the measures to be used, take care to ensure that their definitions are agreed. For example in a delivery performance project, the key measure was on-time delivery. It was not until part way through the project that it was discovered that half of the team thought that on-time meant as measured against the factory scheduled delivery date whilst the remainder thought that it was measured against the customer’s required delivery date; sometimes there were weeks between them. A typical measure of a process is that of Cycle Time. This is the total amount of time to complete the process from boundary to boundary.

Value Analysis There may be many steps within a process. Each step will take time, material and add cost to the process. Thus, it is important that each step adds value to the process. A step that contributes to customer satisfaction adds value. If this step were eliminated then the customer would notice. The definition of whether a step adds value or not needs to meet three criteria (Eckes):


1. The customer must be willing to pay for that step in the process 2. The step must physically change or transform the product or service 3. The step must be done right the first time

It is often easy to identify non-value added work as the process step begins with the prefix “re” e.g.: rework, review, reclaim, rewrite, reject, retest, recall. Eckes identifies six major categories of non-value adding work: Internal failures – steps related to correcting errors/ rejects in the process External failures – steps relating to failures at the customers Control / Inspection – does this add value? Why can’t the process get it right first time? Delays – often seen as waiting for the next stage of the process Preparation/set up – the time this activity takes up has led to the demand for long runs Movement – excess movement is often created due to poor factory layout Compare this list to the Seven Wastes (see Lean Operations) identified by Taiichi Ohno:

1. Overproducing 2. Waiting 3. Transporting 4. Inappropriate processing 5. Unnecessary inventory 6. Unnecessary motion 7. Defects

Having identified non-value adding steps it is useful to calculate the impact they are having on the total cycle time of the process. It will also enable priority to be given to the step that has the largest detriment on the process. Given that most processes have been in operation for many years then it is quite possible for steps and procedures to have been added for someone or purpose many years ago but are no longer of value. At each step consider key customers’ needs. If the step does not satisfy a customers’ needs then look to eliminate it. For each step ask: Does this step add value in our customers’ eyes? What would happen to the customer if this step were eliminated? Are others in the organisation affected by this step? If a value-adding step adds considerably to the cycle time, try to brainstorm ways of reducing the time taken without losing the value.


Inspection / Control Points A point of contention is the identification of inspection or control points. Inspection does not add value and appears when a step within a process cannot produce its output right-first-time. In the interests of process improvement it would be better to not have such inspection. However, within critical processes some inspection may be necessary in order to monitor the process and to check for any unforeseen changes to the process. Needless to say, such points should be kept to a minimum. Where an inspection point is deemed necessary then it is important that the standards to which the inspection is made are clear. Inspection/ Control points should be positioned close to where an error may be first detected.

Suppliers The quality of the inputs will influence the quality of the process. Often there is little communication between Customers and Suppliers, whether internal or external, such that there is often little knowledge of exactly what the customer requires and from the customer’s point of view exactly what the Supplier can provide. Too often Suppliers take on contracts knowing that they will find it difficult to meet them. An example in the electronics industry found a Supplier receiving returns from a customer for oversized mounting holes in a circuit board. After several rejected batches and varied excuses from the Supplier, they eventually admitted that they could not maintain the hole-tolerance of +/-1mm. Further investigation at the customer’s revealed that the tolerance was a default tolerance and the process could easily cope with a tolerance of +/-2mm, a figure the Supplier could achieve. A lot of unnecessary hassle and frustration could easily have been avoided.


Measurement “What gets measured gets done.” Measurement is an important part of process / product improvement. If you don’t measure, how do you know whether things have improved? Within an organisation there are many things that could be measured but for historical reasons the things that are currently measured may have little to do with the areas that need to be improved. By measuring we can: Ensure customer requirements have been met Monitor progress against objectives Create standards for comparisons – Benchmarking Provide visibility for everyone to understand their performance level Estimate the costs of poor quality Identify areas for improvement However, beware what gets measured affects behaviour. “Tell me how you will measure me and I will tell you how I will behave. If you measure me in an illogical way, do not complain about illogical behaviour.” (E. Goldratt: The Haystack Syndrome) For example: in a chemical company each shift was measured by its yield. A poor yield resulted in aggressive messages on the notice board and strong words from the departmental manager. In most cases the operators were powerless to affect the yield yet they still were blamed. So to avoid the hassle, the Supervisors under-reported the quantities of raw material used. This led to an improvement in the yields and a quiet Management. The only problem that Management had to confront was to account for the thousands of pounds of missing raw material at the annual stock check. The choice of measurement is important. There needs to be a clear understanding of why something is being measured and whether it is the most appropriate measure. Three criteria for measurement (Eckes) are: 1. Measure only what is important to the customer. 2. Only measure those outputs of a process that you can improve. 3. Measures of effectiveness in which you have no history of customer dissatisfaction should be avoided.


In Six Sigma a term that is commonly used is that of Critical to Quality or CTQ. (This can also be referred to at Critical to Customer or CTC).This refers to any features of a product or aspects of a process that a customer would view as being important in the performance of the product or process. Thus, if they are critical to the customer then they should be measured. A word of warning; there are many things that can be measured. Concentrate on the critical few.

Measurement Systems Measurement Systems are the means by which data is obtained. For measuring dimensions this will include gauges, scales etc. For process related measures it may be the point at which the data is collected or the data itself. It is often assumed, incorrectly, that the tools used to create the data are suitable for the job. It is important that several aspects of the measurement system are known. Accuracy – the measured value reflects the true value Repeatability – the same person repeating the measurement gets the same

result Reproducibility – a different person repeating the measure gets the same result Stability – the same measurement repeated over time does not vary Adequate Resolution – the measure has enough different values to pick up differences. For measuring equipment then a Gauge R & R study should be carried out before data is collected. Just because a piece of equipment has been calibrated does not guarantee its suitability. This process is more difficult for qualitative data i.e. yes/no, opinions etc and if possible quantitative data should be collected. Qualitative data can be made to look more like quantitative data. For example in the measuring of “pain” in drug trials a 20cm line is used marked “no pain” at one end and “severe pain” at the other. The individual is then asked to mark on the line the appropriate level of pain. This point is then measured in either centimetres or millimetres. This is better than a qualitative five-point scale: “no pain”, “a little pain”, “moderate pain”, “a lot of pain”, “severe pain”


Measurement Variability When collecting data it is recognised that there will be variation from the process, but there will also be variation introduced by the measurement system. It is important to know how much variation is being introduced by the measurement system hence the need for an assessment of the system. Questions to be answered. How big is the measurement system variability? What are the sources of the measurement variability? Is the measurement system stable over time? Is the measurement system capable of making the required measurements of the process? What needs to be done to improve the measurement process? The purpose of measurement systems analysis is to quantify the variation due to the measurement system itself. Too much variation from the measurement system and - changes in the process may be hidden

- the measurement system may give rise to a false belief that a change in the

process has taken place.

Measurement System requirements

- the measurement system must be in a state of statistical control (stability) - the variability of the measurement system must be small compared with the

manufacturing process variability and any specification limits - the increments of measure must be small relative to either the process

variability or the specification limits - the measurement system should give the true or reference value, within

acceptable limits (accuracy and precision)

Accuracy and Precision Confusion often exists between these two terms. Accuracy is were the data on average reflects the true value. Precision is where the data shows very little variation. Ideally a measurement system should be both accurate and precise. Sometimes accuracy is called bias – the difference between the observed average and a reference value. This should be close to zero.


Precision / Accuracy

Accurate not precise Precise not accurate

Neither precise nor accurate Precise and accurate

Repeatability – this is the variation observed when a gauge is used by one person repeatedly to measure the same part. Theoretically the observed value should always be the same, however, in practice this is not the case. Repeatability is usually much smaller than the precision of the system. Reproducibility – This is the variation present when the same part is measured by different people. Variation of Measurement System (precision) = Repeatability plus Reproducibility σ2 ms = σ2 rpt + σ2 rpd Stability – this is the variation present when the same part is measured by the same person over time. This can be assessed using a Control Chart and must be stable. Measurement Discrimination – this sis the capability of the measurement tool to detect small changes in a measured characteristic. The higher the discrimination, the smaller the changes that can be detected. As a rule of thumb the tool should be able to detect a change of a tenth of the total process spread.

Gauge R&R This stands for Gauge Repeatability and Reproducibility. This is a set of trials conducted to assess the Repeatability and Reproducibility of the measurement system. In simple terms – several operators measure several items a number of times. For example: 3 operators measure ten parts three times.


The calculations associated with this study will be covered in the training.

Sample Size One of the difficulties, in collecting data, is knowing how much to collect. There are statistical methods of calculating the minimum sample size but the result of the calculation is often a size that is beyond the practicalities of the study. Formula: Sample size > (2 x process sd / difference to be detected)^2

(95% confidence level) Example: Process has an sd of 0.25. What sample size is required to detect a difference of 0.1 mm? Sample size = ((2 x 0.25) / (0.1))^2 = 25 units Basic principles to follow are: Collect enough data to represent the total variation of the usual process. Try to collect quantitative data Ensure that the data represents “common cause” variation and not “special

cause” variation

Sigma Calculation Six Sigma has introduced the measure known as Process Sigma or the Sigma metric. This is simply a metric that reflects the ability of the process to meet the specifications of the process. It is important to ensure that the specifications used represent the customer requirements and do not as often happens represent the organisations view. Sigma is derived from DPMO (Defects per Million Opportunities). DPU = number of defects / number of units DPMO = (DPU / Opportunities for error)*1,000,000


DPMO is then converted to Sigma from tables (see Appendix)

DPMO Process Sigma 308,537 2 66,807 3 6,210 4 233 5 3.4 6

Where Does Sigma come from? The Sigma value is based on mathematical calculations of the area under the Normal distribution curve.

Normal curve Under the standardised Normal curve it is known that 68.27% of the data lie within one standard deviation of the mean, 95.45% of the data lie within two standard deviations of the mean. Thus if we consider what lies outside of two standard deviations from the mean we can calculate 100-95.45 = 4.55. If we convert to a proportion (divide by 100) and multiply this by 1,000,000 we have the number of defects we would expect in every million items. (45,000 defects). So at the level of two sigma we would expect 45,000 defects. Notice that this is not the same value as in the tables. This is because the table values are adjusted to allow for a long-term shift of the process (estimated at 1.5 sigma). Thus two sigma is really 0.5 sigma. The Sigma metric has both benefits and problems Benefits Allows different processes to be compared. Accounts for the fact that one process may have many steps whilst another only has a few. Moves the mindset away from percentages (many managers would stop improvement if they had a process that was producing 95% good product) Provides a baseline to compare improvement Disadvantages


Is based on the theoretical Normal distribution Is used as an exact measure when at the edges of any distribution the ability to predict accurately declines The 1.5 sigma allowance may be inappropriate Sigma can be inflated by increasing the number of opportunities for error Some organisations have avoided the problem of defining an “opportunity for error” by simply using defects per unit. Thus if a process has a failure rate of 5% then DPU = 0.05, DPMO = 50,000. From tables Sigma = 3.1.


Statistical Thinking

Understanding Variation Variation appears in all processes. Look around you; no one is exactly the same – even identical twins have differences. Write your signature several times – they are different. Measure the output from a machine – they will never be exactly the same. In essence statistics is the study of variation. Without understanding variation, process improvement cannot be achieved. In order to describe the amount of variation then a number of measures are used: mean, mode, median, standard deviation and range. Together these will describe the variation. These values will be discussed in basic statistics. Understanding the variation within a process will help determine what needs to be done to reduce it.

Why reduce Variation? Life becomes so much easier if every time you do something the outcome is exactly the same. You can predict what is going to happen and therefore be prepared. This is especially true for the output from a manufacturing process or service provision. If you knew that your journey to work would take exactly twenty minutes every time then you know what you have to do to arrive at work on time. However, most things do vary and so we have to allow for the variation. Variation presents problems in both manufacturing and services. For instance, a screw has to be put into a block. Variation can mean that the screw and block do not match. Or a bearing is produced slightly large creating increased friction causing a poor performance of the product. A service call takes longer than expected thus impacting on planned calls for the rest of the day. Understanding the variation will allow us to reduce the main sources and thus incur less problems and failures. This has been the basic premise of Quality improvement.


Special Cause and Common Cause Variation Variation can be of two types: Common Cause - there all the time - part of the natural variation inherent in the process Special Cause - exceptional

- unpredictable

It is important in starting process improvement to be able to distinguish between Common Cause and Special Cause variation. Six Sigma looks at reducing Common cause variation and in doing so may change the process such that Special Causes are eliminated. However, if Special Causes are present in the process then they need to be identified and removed. To identify and remove Special Causes for which the root cause is unknown, a rigorous and logical problem solving methodology, such as RISE, is required. Special Causes must be removed to allow a clearer understanding of the Common Cause variation. They must also be removed because they may fail to appear during the data collection and give a false sense of security to the improved process only to appear later and produce unexpected rejects. In terms of responsibilities: Common causes – management must change the system. Use process improvement, Six Sigma methodologies Special cause – operators/administrators to identify cause and eliminate. Deming estimated that 94% of processes exhibit Common Cause variation and 6% exhibit Special Cause variation

Sources of Variation Variation can be found in a number of sources, some of which may be easy to affect; others may be difficult. For a process the sources of variation may be understood by use of MEPEM or the M’s.


MEPEM M’s Materials Materials Equipment Machines People Men Environment Mother Nature Methods Methods Measurement Maintenance For a product then the above categories can be used or the following:

1. Piece to piece variation 2. Changes over time (wear) 3. Customer usage 4. External environment 5. Interaction with neighbouring sub-system

This is known as a Noise Tree. Consider a fountain pen. Why might a pen not function as intended? Piece to piece – variation due to size of nib, size of casing, size of ink cartridge/ holder, ink viscosity Wear – nib wear, amount of ink Customer Usage – user style, angle of writing, pressure applied, storage, hand size Environment – temperature System to system – type of paper In looking for improvement then there needs to be an understanding of what factors are influencing these sources in order to make changes to reduce the effect.

Taguchi’s Loss Function Six Sigma tends to have the philosophy that the aim is to reduce variation such that it is well within the customer specification. This focuses on the specification limits. However, Taguchi points out that the philosophy should be to focus on the centre point (mean) and minimise variation around that point irrespective of the specification limits.


6σ

Understanding Variation

Variation around target (Loss function)

Target Value

Loss FunctionLoss

Although both will give potentially similar improvement, the philosophies are a little different. By using specification limits then the danger is the process can appear to have been improved by simply widening the limits, the failure rate goes down and the Sigma value goes up. Taguchi’s Loss Function can be a little too hypothetical for many people who need to have cut-off points to determine whether a product or service is acceptable or not.

The Normal Distribution The Normal distribution is a theoretical distribution created by Gauss. Equation:

Its importance is for several reasons: It is a good approximation to the data from many processes It is symmetrical It can be used to estimate the probability of events occurring outside of defined parameters


6σ

Six Sigma MetricsArea under Normal Curve

99.73%99.99%

95.45%

µ−4σ µ−3σ µ−2σ µ−1σ µ µ+1σ µ+2σ µ+3σ68.27%

µ+4σ

Areas under the Normal distribution 68.27% of the data lies within one sd. of the mean 95.45% of the data lies within two sd of the mean 99.73% of the data lies within three sd of the mean Calculating areas under the Normal Curve. Z = (x – mean) / sd Look up the z value in tables. Example A process has a mean of 10cm with an sd of 2. Calculate the probability that a value lies beyond 14 cm Z = (14 – 10) / 2 =2.0 From tables a Z= 2.0 gives 0.02275 i.e. there is a probability of 0.023 that a value will lie outside of 14cm Or we would expect 2.3% of the data to lie outside of 14 cm


Displaying Data A graphical representation of data can have far more impact than simply a set of numbers. It can also help in the understanding of the variation. Here we will look at: Check Sheet Histograms Dot Plots Pareto Chart Run Charts Scatter Diagram Definitions: Frequency – is the number of times a value or values within a defined category occurs. Frequency distribution – is a display of the pattern created by plotting data frequencies versus quality characteristics. Class Width – the width of the interval into which data is grouped. Class Boundaries – the extremes of the class widths. These boundaries should be carefully defined to make it clear in which class each result should fall. For example: if data is to two decimal places then a boundary of 5.10 –5.19cm is clearer than 5.1 – 5.2 cm. Check sheet (Tally Chart) This is an easy to understand method of grouping data. 1.Decide on the data to be collected 2.Decide on the period of data collection and from what source 3.Create a Check Sheet on which to record the occurrence of each event or data value. For count data the groups will be simple categories. For numerical data the groups may be in the form of class widths. 4.Collect data


This type of chart is often the forerunner of a graphical method e.g.: Histogram

6σ

Check Sheets

44Total

1I5.58

3III 5.57

5IIIII5.56

7IIIII II5.55

12IIIII IIIII II5.54

9IIIII IIII5.53

4IIII5.52

2II5.51

1I5.50

TotalTallyDiameter

Histogram This graphical method displays the number of data points in each class width as a bar. The height of which represents the number of data points in that class width. In its purest state the area of the bars of the histogram are proportional to the frequency. Thus it is possible to have bars of different widths. However, for convenience the bars are kept the same width so that the height represents the frequency of occurrence. Creating a Histogram 1.Decide on the process measure – data measured on a continuous scale is best 2.Gather the data – collect enough data to give a good representation of the process (some recommend 50- 100 points) 3.Decide on the class width – be practical Note: too few class widths will produce a tight, high pattern; too many class widths will produce a flat, spread out pattern. Guidance:

Number of data points Number of class widths <50 5 – 7

50 - 100 6 – 10 100 -250 7 –12

>250 10 - 20


As a rule of thumb the number of classes equals the square root of the number of data points. 4.Calculate the class boundaries 5.Use a check sheet or tally chart to sort the data into the appropriate class widths. 6.Draw the Histogram 7.Interpret the Histogram Interpretation Centre – compare with customer requirements Spread – is variation well within the customer requirements? Shape – Normal distribution? Bi-modal distribution? Skewed? Any strange shapes? If the number of measurements taken were very large and if the class widths were very small then the Histogram would look more like a curve than a series of steps. This curve is known as the Frequency Curve and will be used to represent the shape of the distributions from where the data comes.

6σ

Histogram

Histogram

0

2

4

6

8

10

12

14

5.5 5.51 5.52 5.53 5.54 5.55 5.56 5.57 5.58

Tube diameter (cms)

Fre

quen

cy

Dot plot A graphical method, similar to the Histogram, where each data value is represented as a dot. Each data point within the class width is then represented by a dot. The Dot Plot is created in the same way as the Histogram except that the frequency of each class width is represented by a number of dots rather than a bar.


6σ

Dot Plot

5.50 5.51 5.52 5.53 5.54 5.55 5.56 5.57 5.58

Fr e

quen

c y

Diameter of tube (cms)

Pareto Chart This was first created by Vifredo Pareto in the 19 Century. He was looking at wealth distribution. He discovered that 20% of the people held 80% of wealth. This led to the 80:20 rule. That is 80% of the problems are caused by 20% of the sources. In considering the 80:20 rule, Juran talks about the vital few and the useful many. Creating a Pareto Chart 1.Decide on the problem area to be considered 2.Identify the causes of the problem 3.Quantify the number of times each cause creates the defined problem 4.Order the causes according to their magnitude 5.Draw a bar chart to represent the frequency of occurrence of each cause 6.Interpret the results The largest occurring cause should be the one first dealt with as this contributes most to the problem.

6σ

Pareto Charts

Pareto Chart

0 5 10 15 20 25 30

too late

poor seals

thin material

discoloured material

wrong product

wrong quantity

invoice incorrect

Rea

son

for

Co

mpl

aint

Number of Complaints


Run Chart This is a graphical representation of data over time. Creating a Run Chart 1.Decide on the process measure 2.Gather data over time 3.Create a graph with a vertical line (y axis) and a horizontal line (x axis) 4.Determine the scale to be used on the y-axis

Use a scale such that the data collected uses approximately 2/3rds of the y-axis scale

5.Plot the data 6.Calculate the mean of the data 7.Draw the mean as a horizontal line on the graph 8.Interpret the chart (This is the basis for creating an SPC chart)

6σ

Run Charts

Run Chart

0

0.5

1

1.5

2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Week Number

% R

ejec

ts

Scatter Diagram This is a graphical method of observing the relationship between two variables. It is used to assess whether the variation in one variable could be related to the variation in another. This is represented by the degree of Correlation. A correlation coefficient of +1 or –1 shows a high degree of association. A correlation coefficient of around zero suggests no association. It is important to note that a correlation does not prove that variation in one variable causes the variation in the other. Creating a Scatter diagram 1.Decide on the variables to be studied 2.Collect paired samples relating to the chosen variables 3.Create a graph with one variable on the x-axis and the other as the y-axis 4.Plot the paired data values on the graph 5.Interpret the data Look for either positive / negative relationships or simply no relationship.


A lot can be gained from the plot rather than simply calculating the correlation coefficient. Note: beware of false relationships. It is often seen that apparent relationships exist between two variables. E.g.: number of storks and the number of babies born in an area (which proves that storks really do deliver babies). There is often an unseen variable to which both are related. – the degree of urbanisation in the case of the storks – the storks nest in the chimneys.

6σ

Scatter Diagrams

Scatter Diagram

0

100

200

300

400

500

600

700

800

0 50 100 150 200 250 300

Number of storks in area

Num

ber

of b

abie

s bo

rn

per

yea

r


Basic Statistics Recommended book; Statistics without Tears; Derek Rowntree 0-14-013632-0 There are a number of parameters that typical in any statistical calculation. These parameters tell use where the centre of the data is (location) and how far the data spreads.

Measures of Location : mean, mode, median Measures of Spread : range, standard deviation, inter-quartile range If we know these parameters then we have a good idea of what the distribution of the data looks like.

Measures of Location Three main measures : Arithmetic mean (Mean) Median Mode Mean (arithmetic mean)

= ( x1 + x2 + x3 + ……….xn ) / n where n is the number of data points

Example: 3,4,5,5,6,9,10 = 42 / 7 = 6

Median (middle value) The middle value when placed in numerical order. Ex: 3,4,5,5,6,9,10

3, 4, 5, 5, 6, 9, 10 therefore = 5 If there are an even number of data items then the Median is the mean of the two middle values when the data is placed in order. Ex: 3,4,5,5,6,9,10,11 3,4,5,5,6,9,10,11 : median = mean of 5 + 6 = 5.5

X

X


Mode This is the most frequently occurring value. There may be one or more points with this value. Hence there are bi-modal, multi-modal sets of data. Ex: 3,4,5,5,6,9,10 Five occurs twice and is therefore the most frequently occurring value.

= 5

Relationship between Mean, Mode and Median

Symmetrical data Mean = Mode = Median

Mean, Median, Mode

Measured Characteristic


Positive skew data Mode<Median<Mean

Negative skew data Mean<Median<Mode Generally: Mode = Mean –3 (Mean – Median)

Mode Median Mean

ModeMedianMean

wdp consult

Measures of Spread Look at the data points for A, B and C. They have the same mean but there spread is very different. Hence knowledge of the spread of data is also important.

A

B

0 1 2 3 4 5 6 7 8 911

1213 14 15 16 17 18

0 2 4 6 8 10 12 14 16 18 20

0 2 4 6 8 10 12 14 16 18 20

ing

C
0 2 4 6 8 10 12 14 16 18 20
six sigma


Range Range = maximum value – minimum value A: 0 B: 8 C: 20 The Range is easy to calculate but is influenced by extreme values

Standard Deviation (sd) Sometimes called root mean square deviation. Two measures; Population and Sample Population sd (only used when you have all the data possible) Sample sd (most commonly used when calculating based on a sample of the data) From the dot plots above: A: 0 B: 2.83 C: 7.48 It is rare to have all the data so it is usually the sample standard deviation that is calculated. Why two different calculations? It can be shown that when taking samples of data then the second equation on average gives a better estimate of the true population sd than the first. In statistical jargon the second equation is said to be BLUE (Best Linear Unbiased Estimate).

( ) X X

2 −

n∑

σ = √

( )X Xn 1

2−

−∑

s = √


The standard deviation is not so easy to calculate by hand compared to the Range but this is not a problem with calculators and computers. Remember to use the “n-1” calculation. The standard deviation is less influenced by extreme values compared to the Range.

Interquartile Range This shows the range of the middle 50% of the data. Interquartile range = Q3 – Q1 Where Q1 = 25% value and Q3 = 75% value (can be found from a cumulative frequency) A: 0 B: 12 – 8 = 4 C: 16 – 4 = 12 Calculators do not calculate this value and hence is not used very often. Its advantage is that it is uninfluenced by extreme values.


Statistical Distributions As seen in Histograms, if we take enough data from a population and break it into small enough class widths then the resulting shape will approximate to a curve. Bi-modal The description of the data is known as the frequency distribution and the curve can be represented by a mathematical function.

wdp consulting

In order to use the frequency distribution for calculations, the data is standardised such that the area under the curve equals one. It is now possible to calculate areas under the curve. This distribution is called the Probability Density Function (p.d.f.)

Specific Distributions

Normal Distribution This is the most commonly used p.d.f.

- it is a good representation of much naturally occurring data - it is symmetrical - other distributions look like a Normal distribution in certain

circumstances Because of its properties there is almost an obsession with it. The area under the Normal curve can be calculdistribution tables. Z = (X – ) / σ Where Z is in fact the number of standard devimean. This is used in conjunction with the stand X is the value being considered, x-bar is the mestandard deviation of the population and may bdeviation.

68.27%

µ−4σ µ−3σ µ−2σ µ−1σ µ µ+2σ µ+3σ µ+4σ

X
µ±1σ
six sigma

ated from standardised Normal

ations that the value is away from the ard tables.

an value of the data, sigma is the e estimated by the sample standard

wdp consulting

Example: The mean diameter of a steel ball is 50mm with a s.d. of 10mm. Calculate the % of balls that will have a diameter greater than 57mm. Z = (X – ) / σ Z = (57 – 50) / 10 = 0.7 From tables Z = 0.7 gives a value of 0.2420 Thus we expect that 24.2% of the balls will have a diameter of > 57mm.

Standardised Normal Tables Take care if using these tables as the areas are sometimes tabulated differently. The area being tabulated is usually shown along with the title. Example: The mean diameter of a pipe is 50.2 mm withWhat percentage of the pipes will have:

a. a diameter greater than 50.6mm? b. a diameter less than 49.6mm c. diameters between 49.6mm and 50

a. Z = (50.6 – 50.2) / 0.5 = +0.8

From tables 0.2119, = 21.2% > 50

b. Z = (49.6 – 50.2) / 0.5 = -1.2

µ−4σ µ−3σ µ−2σ µ−1σ µ µ+2σ µ+3σ µ+4σ

X
µ+1σ
six sigma

a sd of 0.5mm.

.6mm?

.6


From tables 0.1151, = 11.5% < 49.6

c. if 21.2%>50.6 and 11.5% < 49.6 then the remainder must be between these values ie: (100 – (21.2 + 11.5)) = 67.3%

Normal Distribution and Link to Six Sigma Graphical representation At three sigma the data from the distribution just fits within the specification limits. At six sigma the variation of the data has been reduced such that the distribution of data would fit exactly twice into the specification limits.

Binomial Distribution The Binomial distribution occurs where there are only two possible outcomes: Pass / Fail; Good / No Good; Yes / No; Present / Absent This distribution is often used when looking at numbers of defectives. That is the part or service is either defective or it is not.

Target

Six Sigma

Three Sigma

Measured Characteristic

Lower Spec Upper Spec


nCx px q n-x

where nCx = (n!) / ((x!)(n-x)!) and n! = 1 x 2 x 3 x 4 x 5 x …………n p = probability of success, q = probability of failure, n= number of data points The mean of a Binomial distribution is n .p The sd of a Binomial distribution is √n.p.q When n is large and p and q are not small then the Binomial distribution can be represented by the Normal distribution. Example: Tossing a coin Probability of a head = 0.5, probability of a tail = 0.5 Probability get three heads in three tosses Pr (3 Heads) = ((3!) / (3!)(0)!) (p3)(q0)

= ((3 x 2 x 1) / (3 x 2 x 1))(0.5)3(0.5)0 = (0.5)3 = 0.125 Probability of getting two heads in three throws Pr (2 H) = ((3!) / (2!)(1!)) (0.5)2(0.5)1 = ((3 x 2 x 1) / (2 x 1)(1)) (0.5)2(0.5)1 = (3) (0.5)3 = 0.375

Note: Pr (3H) + Pr (2H) + Pr (1H) + Pr (0H) = 1


Poisson Distribution Events whose probability of occurrence is low tends to give a Poisson distribution. Time between events (e.g. Time between failures) can also be modelled by a Poisson distribution.

Pr ( X) = ((e-µ)( µx)) / x! Where e = 2.7183, µ= mean = n.p, x is the number of events For a Poisson distribution : mean = µ, variance = µ For large n and a small probability of success (occurrence) then the Binomial distribution tends towards a Poisson distribution. (n > 20, p< 0.1). Example: Probability that a product fails in its first year is 0.01. What is the probability of four failures from 500 products sold. n = 500 p = 0.01 µ = n.p = 500 x 0.01 = 5 Pr (4 events) = (e-5)(54) / 4! = 0.1775 Pr (0 failures) = (e-5)(50) / 0! = 0.0067 Pr (1 failure) = (e-5)(51) / 1! = 0.0337 Pr (2 failures) = (e-5)(52) / 2! = 0.0842 Pr (3 failures) = (e-5)(53) / 3! = 0.1404 Etc


Central Limit Theorem This is important in understanding the relationship between a sample taken at random and its underlying distribution. The Central Limit Theorem states that the distribution of means from samples of a population tends to be Normally distributed with a sd of the sample means equal to (� / �n) where n is the number in each sample taken. The value (� / �n) is known as the Standard Error of the Mean. This will be used in Statistical Process Control. The Central Limit Theorem can be demonstrated by a simple dice experiment.

a. roll three dice 20 times record each individual value and the mean of the three dice. Plot a histogram of the individual values. Plot a histogram of the mean values Calculate range of the individuals. Calculate range of the means

b. roll 5 die 20 times record each individual value and the mean of the five dice. Plot a histogram of the individual values. Plot a histogram of the mean values Calculate range of the individuals.

Calculate range of the means Compare the histograms and ranges.


Hypothesis Testing Too often people collect data without thought about what they are testing. In statistics, the decision about tests to be applied should be taken before the experiment has been run or the data collected. It is all too common for people to collect data organise it and then test the top result against the bottom result, often totally ignoring all the data in between. For example, if we wish to compare a set of data with a previous known mean. i.e. has the mean of the process changed then we need to decide how large a difference are willing to accept before declaring the process has changed and what risk are willing to take on making the wrong decision. Jargon H0 : Null hypothesis – what is to be tested Eg. The means of two samples are equal H0 : µ1 = µ2 For any Null hypothesis there must always be an alternative (H1). That is if H0 is not accepted then what is the decision. Eg: the means of the two samples are not equal H1 : µ1 ≠ µ2 or H1 : µ1 > µ2

Note: to be correct, statistics only ever rejects the Null hypothesis. It never ‘accepts’ the Null hypothesis but rather states that ‘there is no evidence to reject the Null hypothesis’. Collect more data and we might change our minds. The use of a hypothesis test also leads to a decision of at what point is the Null hypothesis rejected. This is where statistical significance is used. In defining a significance test then a decision must be made beforehand what the Alternative hypothesis will be. Single-sided Hypothesis test H0 : µ1 = µ2 H1 : µ1 > µ2 or H1 : µ1 < µ2 ie. we are only interested in a difference in one direction. Double-sided (two-sided) Hypothesis test H0 : µ1 = µ2 H1 : µ1 ≠ µ2 ie. we are interested in whether or not there is a difference and not the direction. The decision of a single-sided test or a double-sided test will affect the probability of ‘accepting’ or rejecting the Null hypothesis.


Significance Testing In comparing any sets of data then there are a number of options that may arise. This is best represented as a table.

Actual Situation No

difference Difference

No difference √ Type II

Decision Made

Difference Type I √ If we make the right decision compared to the actual situation then all is well. However, we need to be aware of the risks that we make an incorrect decision. i.e. we conclude that there is a difference between the groups when they are from the same distribution. This is known as a Type I error (α). We could also conclude, in error, that there is no difference when the two groups come from different populations. This is a Type II error (β). In significance testing it is necessary to decide on the risks we are prepared to take in making an incorrect decision. In most cases it is the Type I error (α) that we set. That is; what is the probability we can accept that we will conclude that there is a difference when there is no difference? i.e. we will reject the Null hypothesis. The Type I error is often set at 5% although in risk averse situations value of 1% may be set. In theory any value could be used. By choosing the Type I error we automatically set the Type II error. If the Type I error is set at 5% then the Type II error will be around 10%. That is that there is approximately a 10% risk that we will say that there is no difference when the samples are in fact different. As the Type I error goes down the Type II error goes up and vice versa. (1 - β) is known as the Power of the test to discriminate between groups. As the number of data points in the sample increases then so does the power. (100 - α) is referred to as the confidence level. Knowing whether a test is one-sided or double-sided and the risk that we are prepared to take in making an incorrect conclusion we are in a position to use statistical methods to compare sets of data.


t- Distribution A distribution can be standardised by the use of the equation: Z = (X - µ) / σ According to the Central Limit Theorem the distribution of sample means with have a standard deviation of (σ / √n). Replacing this in the Z calculation we have: Z = (X - µ) / (s / √n) To show that the estimate of the standard deviation has come from a sample the symbol s is used rather than σ. This new formula has a value that follows the Student’s t distribution. (So named because the creator of this distribution did not want to be known and called it “Student’s t” to hide his identity). Thus when comparing data from samples we use t tables rather than Normal tables. As the number of data points in a sample increases then the t distribution tends towards a Normal distribution. If we have collected a sample of data, with mean x-bar, from a process and wish to benchmark it against a previously recorded value(�) to determine if the process has changed and wish to take a risk of 5% that we will conclude that there is no difference when there is, then the test would look like: H0 : x-bar = µ H1 : x-bar ≠ µ (ie. two-sided test) α = 0.05 (5%) Pr ( -t.975 < (x-bar – µ) / ( s / √n) < +t.975) If x-bar lies in either of the shaded areas then we would reject H0 and conclude that the process has changed. Example: An efficiency expert claims that a new machine must reduce the time by eight hours to be acceptable. Six experiments gave an average decrease of 8.4 hours with a sd of 0.32 hours. Should the new machine be bought? H0 : x-bar = µ H1 : x-bar > µ (one-tailed test) where µ = 8 hours N = 6 , x-bar = 8.4, s = 0.32 one-sided test use α = 0.01 (1% level) Pr ( (x-bar – µ) / ( s / √n) > t.99)


From t tables α = 0.01, degrees of freedom (ν) = n-1 = 5 t value = 3.365 ( (x-bar – µ) / ( s / √n) > t.99) (8.4 – 8.0) /(0.32/ √6) >3.325) 3.062 > 3.325 since this is false, have no reason to reject the Null hypothesis. That is the new machine has not reduced the time by more than eight hours. However, had we used α = 0.05 (5% level) the answer would have been: From t tables α = 0.05, degrees of freedom (ν) = n-1 = 5 t value = 2.015 3.062 > 2.015 is true therefore we would reject the Null hypothesis with 95% confidence. This reinforces the need to decide on the parameters of the test before seeing the data.

Confidence Limits When we take a sample, the mean of the sample is an estimate of the mean of the underlying population. However, if we were to take a different sample from the same population the estimate of the mean would be slightly different. By using confidence limits we can predict what the range of estimates of the population mean are likely to be. Example; What are the confidence limits for a sample of 12 measurements with an average breaking strength of 7.38g and an sd of 1.24g. n=12, x-bar = 7.38, s = 1.24 assume α = 0.05 from t tables degrees of freedom (ν) = n-1 = 11, α = 0.05 t value = 2.201 Pr ( -t < (x-bar – µ) / (s / √n) < +t ) = 0.95 Pr( -2.201 < ( 7.38 – µ) /( 1.24 / √12) < 2.201) = 0.95 Pr( 7.38 –2.201 (1.24 / √12) < µ < 7.38 +2.201 (1.24 / √12)) = 0.95 Pr (6.59 < µ < 8.17) = 0.95


Thus the 95% confidence limit for the estimate of the population mean is 6.56 to 8.20. This may be written as 7.38 ± 0.79


Comparisons Statistical tests are often used to compare data sets e.g. a before and after improvement. (Most of the analyses will be shown using Excel). t-test Can be used for : Comparison of the mean of a sample with a population mean Comparison between the means of two samples

Comparison of a Sample Mean with a population mean Example A company claims that the average life of a battery is 21.5 hours. Tests of six batteries gave the following results: 19,18, 22, 20, 16, 25 hours Do these results confirm the company’s claims? H0 : x-bar = µ H1 : x-bar ≠ µ µ = 21.5, x-bar = 20, s = 3.16 2 tailed test. df = 5, t (0.025) = 2.571 x-bar ± t. (s / √n) 20 ± 2.571 (3.16 / √6) 20 ± 3.32 16.68 < µ < 23.32 batteries could have come from a population with 21.5 hours. Note the effect of increasing the number in the sample. N = 120, x-bar = 20, sd = 3.16 t (0.025) = 1.98 ( 2 tail)


20 ± 1.98 (3.16 / √120) 20 ± 0.57 19.43 < µ < 20.57 In this case we would reject the Null hypothesis. i.e. the batteries did not come from a population with a mean of 21.5 hours. Analysis using Excel Many of the simpler statistical tests can be run in Excel. In order to do this Excel needs an add-in for Excel called ‘data analysis’. This is usually held on the original Excel CD or may be held on the computer but not activated.

Comparison of Two Groups There are two situations one where the data from the two groups is linked (paired) and the other where the data is not linked. The mathematical calculations will be shown at the end of this chapter for those who are interested. Otherwise the data will be analysed through Excel.

Un-paired t-Test

Group A Group B 1.38 1.03 1.41 1.22 1.07 0.98 1.75 1.62 1.77 1.69 1.20 0.67 0.78 0.84 1.04 0.84 1.22 1.25 1.63 1.22

In this example we wish to compare two sets of data; group A and group B.

Select Tools Select Data Analysis


Select t-test, two sample assuming unequal variance. (selecting this option rather than assuming equal variance avoids having to make the decision whether the variances of the two groups are similar).

Identify the data for variable 1 (group A) and variable 2 (group B). Select your alpha level (0.05 in this case). Identify an area on the spreadsheet to display the results. (Output range). Ensure that the output range does not go over your data.


If the calculated value (t-stat) is greater than ‘t critical’ then the difference is said to be significant. If the t-stat is less than the ‘t critical’ then the difference is said to be not significant.

Note: that the results for a one-tail test and a two-tailed test are calculated. Hence it is important to have made the decision before hand as to the Hypothesis i.e. is this a one-tailed or two –tailed test. The results also show the exact probability that the difference could have occurred by chance.

Calculated Value

Table Value

Exact probability

df = degrees of freedomIf the Calculated value is less than the Table value (t critical) then the difference is not significant

If the Calculated value is greater than the Table value (t critical) then the difference is statistically significant


Paired t-Test This is a special case as we have additional information in the pairing of the results.

Machine Group A Group B 1 1.38 1.03 2 1.41 1.22 3 1.07 0.98 4 1.75 1.62 5 1.77 1.69 6 1.20 0.67 7 0.78 0.84 8 1.04 0.84 9 1.22 1.25 10 1.63 1.22

Here the groups’ data values are paired in that each group produces output on each machine. As described in the unpaired t-test; Select tools Select data analysis Select t-test: paired two sample

Identify data set 1, data set 2 and the output range. Set the alpha value.


In this case the degrees of freedom are 9 since there are ten pairs. (d.f. = n-1). The t-statistic is 3.1389 which is greater than the t-critical two-tail of 2.262. Therefore there is a significant difference between the two groups.


Analysis of Variance (ANOVA) This is a method of analysing the variation due to different groups or factors. The t-test is a special case of ANOVA for one or two groups. ANOVA can handle more than two groups. One-way ANOVA

Group A Group B 1.38 1.03 1.41 1.22 1.07 0.98 1.75 1.62 1.77 1.69 1.20 0.67 0.78 0.84 1.04 0.84 1.22 1.25 1.63 1.22

In this case we only have one factor i.e. machines, sites, one type of product. We could have several groups within this one factor. Select Tools Select Data Analysis Select Anova: single factor Note: Excel can only handle two factors at a time,. However it can handle several groups of data for any one factor.

Identify the data as one group. Select your alpha level (0.05 in this case). Identify an area on the spreadsheet to display the results. (Output range). Ensure that the output range does not go over your data.


If the calculated value (F) is greater than ‘critical value’ then the difference is said to be significant. If the F-value is less than the ‘F critical’ then the difference is said to be not significant. The results also show the exact probability (P-value) that the difference could have occurred by chance.

Calculated value = 1.65

Critical value = 4.41

Since the calculated value is less than the critical value we conclude that there is no difference between the groups


Two-way ANOVA Here we are considering two factors. As for the one-factor ANOVA we can have several groups of data for each factor. The number of groups for each factor does not have to be identical. e.g. output from three machines on five different days. The factors here are machines and days. Method From the Data Analysis select ANOVA two factor without replication, if there is only one data point per combination of factor settings. Otherwise select, with replication.

Identify the data as one group. Select your alpha level (0.05 in this case). Identify an area on the spreadsheet to display the results. (Output range). Ensure that the output range does not go over your data.

Critical values

Calculated values

Since both calculated values are greater than their corresponding critical values, we conclude that there is a difference between machines and a difference between groups.

If the calculated value (F) is greater than ‘critical value’ then the difference is said to be significant. If the F-value is less than the ‘F critical’ then the difference is said to be not significant. Since there are two factors there are two comparisons shown as rows and columns. In the example rows are machines and groups are columns. The results also show the exact probability (P-value) that the difference could have occurred by chance.


ANOVA becomes more useful in the case where is more than one category being assessed. For example comparing several cars driven by several drivers. We want to know if there is a difference between the cars and / or the drivers. Time Driver Σ A B C

1 7.15 7.20 7.18 21.53 2 7.13 7.17 7.19 21.49

Car

3 7.14 7.19 7.18 21.51 Σ 21.42 21.56 21.55 64.53 Select Tools Select Data Analysis Select Anova :two factor (in this case without replication) Identify data Select alpha Identify output range

In this case Rows = cars Columns = drivers From the Anova table the F-value for cars is 1.1428 which smaller than the F crit 6.944. thus there is no evidence of a difference between cars. However, the F-value for drivers is 17.42 which is greater than the F-crit of 6.944. thus there is evidence of a difference between the drivers. However, Excel does not


show where the difference lies, other than it is between drivers. This can be identified using mathematics as shown in the mathematical analysis.


Chi – Squared ( χ2 )

- used to calculate confidence limits of sd’s - used to test goodness of fit of models to data - used in contingency tables

Goodness of Fit Example throw a die 50 times and record the results. If the die is unbiased then we would expect a similar number of each face, theoretically from 50 throws, we would expect each face to appear 8.3 times (50 /6) Face 1 2 3 4 5 6 Actual : observed (O) 12 6 6 9 7 10 Theoretical : expected (E) 8.3 8.3 8.3 8.3 8.3 8.3 χ2 = ((observed – expected)2 / (expected) df = 6 – 1 = 5 (ie number of possibilities) χ2 = (12 – 8.3)2 + (6-8.3)2 + (6 – 8.3)2 + (9 – 8.3)2 + (7 – 8.3)2 + ( 10 – 8.3)2 / 8.3 = 29.34 / 8.3 = 3.53 from tables χ2 (0.05) on 5 df = 11.1 Since 3.53 < 11.1 the die is considered to be unbiased. We have no evidence to reject the Null hypothesis.


Regression This is a methodology to analyse the relationship between two quantitative variables.

Simple Linear Regression

Distance travelled over Tim e

0

10

20

30

40

50

60

70

0 5 10 15 20 25 30 35 40 45 50 55 60

Tim e (minutes)

Dis

tanc

e (m

iles)

60 mph

30mph

In its simplest case the method fits a straight line to the data in such a way to give the best fit. This is achieved by minimising the variation from the data points to the line. The equation of a straight line relationship between variables is y = a + bx where a and b are constants a: intercept – the point where the line crosses the y-axis b: slope – the gradient or slope of the fitted line. The mathematical equations needed to calculate the regression line can be quite involved and so it is suggested that any linear regression is carried out using Excel or statistical software. Excel only performs simple linear regression. For more complex regression statistical software should be used. Method From the Data Analysis select Regression. Identify the independent data (X) and the dependent data (Y). i.e. the data that will create the equation Y= mX + constant The screen shows a confidence limit of 95%. This is equivalent of alpha = 0.05.

Identify an area on the spreadsheet to display the results. (Output range). Ensure that the output range does not go over your data.


The output for the regression shows: The correlation coefficient R squared The significance of the slope (from an ANOVA table) The coefficients for the equation i.e. the slope gradient and the constant.

There are other options available in Excel but at this stage it is probably best to ignore them.


Mathematical Analysis Unpaired t-test In this example we wish to compare two sets of data; group A and group B.

Group A Group B 1.38 1.03 1.41 1.22 1.07 0.98 1.75 1.62 1.77 1.69 1.20 0.67 0.78 0.84 1.04 0.84 1.22 1.25 1.63 1.22

H0 : x-bar = µ H1 : x-bar ≠ µ Group A : n =10, x-bar = 1.325, sd = 0.325 Group B : n =10, x-bar = 1.136, sd = 0.333 Combined sd: sdcom = √(((na – 1) s a

2 + (nb – 1)s b 2) / (na +nb –2))

= √ (((9)0.3252 + (9)0.3332) / 18 = 0.329 Also s.e. (difference) = sdcom √((1/na) + (1/nb)) = 0.329 √ (2/10) = 0.147 Comparison of means Pr ( difference) ± t (se(difference)) from tables t (0.05 two tailed test) df = na+nb-2 = 18 : t = 2.101 Pr( 1.325 – 1.136) ± 2.101 x 0.147 0.189 ± 0.987


-0.798 < difference < 1.176 since this range includes zero we would not reject the Null hypothesis i.e. there is no difference between the groups. Paired t-test This is a special case as we have additional information in the pairing of the results.

Machine Group A Group B 1 1.38 1.03 2 1.41 1.22 3 1.07 0.98 4 1.75 1.62 5 1.77 1.69 6 1.20 0.67 7 0.78 0.84 8 1.04 0.84 9 1.22 1.25 10 1.63 1.22

Here the groups’ data values are paired in that each group produces output on each machine. Calculate the difference between the pairs.

Machine Group A Group B difference 1 1.38 1.03 0.35 2 1.41 1.22 0.19 3 1.07 0.98 0.09 4 1.75 1.62 0.13 5 1.77 1.69 0.08 6 1.20 0.67 0.53 7 0.78 0.84 -0.06 8 1.04 0.84 0.2 9 1.22 1.25 -0.03 10 1.63 1.22 0.41

Mean difference = 0.189, s = 0.190, n=10 s.e. = s / √n = 0.190 / √10 = 0.0601 t = 0.189 / 0.0601 = 3.145 on 9 d.f. from tables (α = 0.05), t (9 d.f.) = 2.262 since 3.145 > table value , we reject the Null hypothesis that there is no difference between the groups.


Analysis of Variance (ANOVA) This is a method of analysing the variation due to different groups or factors. The t-test is a special case of ANOVA for one or two groups. ANOVA can handle more than two groups.

Group A Group B 1.38 1.03 1.41 1.22 1.07 0.98 1.75 1.62 1.77 1.69 1.20 0.67 0.78 0.84 1.04 0.84 1.22 1.25 1.63 1.22

∑ 13.25 11.36 Total Sum of Squares = ∑(x – xbar)2 = ∑ x2 – (∑ x)2 / n = 1.382 + 1.412 + ……………..1.222 – ((1.38 + 1.41 + ……..1.22)2 / 20) = 32.4093 – ((24.61)2 / 20) Total Sum of Squares (TSS) = 2.1267 Between Groups Sum of Squares = (∑xa)

2 / n + (∑xb)2 / n - (∑x)a

2/ n = ((13.25)2 / 10) + ((11.36)2 / 10) - ((24.61)2 / 20) = 0.1786 Total Sum of Squares = Between Groups Sum of Squares + Within Groups Sum of Squares 2.1267 = 0.1786 + Within Group SS ∴Within Group SS = 1.9481 for the two sample case this is also known as the Residual Sum of Squares


ANOVA Table Source df Sum of Squares Mean Square F Between groups (na –1) 1 0.1786 0.1786 1.65 Residual 18 1.9481 0.1082 Total (n –1) 19 2.1267 Using F tables : F 1,18 (0.05) = 4.41 F (from calculation) < F from tables ∴no difference between the groups. Note: The √ 0.1082 = 0.3289 ………… the sd com from the t-test Two-way ANOVA Here we are considering two factors. As for the one-factor ANOVA we can have several groups of data for each factor. The number of groups for each factor does not have to be identical. e.g. output from three machines on five different days. The factors here are cars and drivers. Time Driver ∑ A B C

1 7.15 7.20 7.18 21.53 2 7.13 7.17 7.19 21.49

Car

3 7.14 7.19 7.18 21.51 ∑ 21.42 21.56 21.55 64.53 Total sum of squares = ∑ x2 – (∑ x)2 / n = 462.6849 – (64.53)2 / 9 = 0.0048 Between Drivers SS = (21.42)2 / 3 + (21.56)2 / 3 + 21.55)2 / 3 - (64.53)2 / 9

= 0.0041 Between Cars SS = (21.53)2 / 3 + (21.49)2 / 3 + (21.51)2 / 3 - (64.53)2 / 9 = 0.0003


Two-way ANOVA Table Source df SS MS F Drivers 2 0.0041 0.0021 21.0 Cars 2 0.0003 0.0002 2.0 Residual 4 0.0004 0.0001 Total 8 0.0048 F 2,4 (0.05) = 6.94 Since the F value of 21.0 is > 6.94 there is a difference between drivers. But since the f value of 2.0 is < 6.94, there is no difference between cars. Note: S1

2 / S22 ∼ F distribution

Since there is a difference between the drivers we need to compare the means. Driver A B C mean 7.14 7.18 7.19

If we simply compared two groups H0 : B – A = 0 H1 : B – A ≠ 0 From the t test B – A ± t . s( 1/n1 + 1/n2) s comes from the ANOVA table s = √ (0.0001) for any comparison we can calculate the critical difference required for significance. t from tables df = 4 (na + nb –2) (0.05) = 2.776 critical difference = 2.776 (√(0.0001) (1/3 + 1/3)) = 0.0185 ∴B = C > A In words. Drivers B and C are similar but A is faster than both of them


Regression This is a methodology to analyse the relationship between two or more quantitative variables. A simple example would be the relationship of distance travelled with time it travelling at a constant speed. This can be shown as a linear relationship signified by: Y = a + bX Where a is the point at which the line crosses the y-axis (the intercept) and b is the slope of the line. (other common forms are ; Y = mX + c ; Y = b0 + b1X). Thus if we know a and b, we can solve for any value of X.

Distance travelled over Time

0

10

20

30

40

50

60

70

0 5 10 15 20 25 30 35 40 45 50 55 60

Time (minutes)

Dis

tanc

e (m

iles)

60 mph30mph

However, all data does not give rise to a perfect straight line and we have to ‘fit’ the best line to describe the data. To do this we use the ‘method of least squares’. This fits the line that has the smallest squared distance between the observed values and the fitted line.


The calculation of the line of best fit is derived from two ‘Normal’ equations; ∑ Yi = n.a + b ∑Xi ∑ Xi Yi = a ∑Xi + b ∑Xi

2

By re-arranging these equations we have: Slope b = Sxy / Sx2 And Sxy = ∑XY – (∑X ∑Y / n) and Sx2 = ∑ X2 - ((∑X)2 / n) Intercept

X

X

X

XX

X

d3

d2

d1

d4

d5 d6

Minimise ∑di2

Y = a + bX

a = Y - b. X

wdp consulting

Example: Individual Exam scores (%) French (X) German (Y)

1 10 11 2 10 22 3 18 22 4 25 19 5 28 35 6 33 27 7 34 33 8 39 40 9 42 42 10 43 47

Can the German scores be predicted from the French scores? X = 28.2 ∑X = 282 ∑X2 = 9312 Y = 29.8 ∑Y = 298 ∑Y2 = 10066 ∑XY = 9539 n = 10 b = (9539 – ((282 x 298) / 10)) / (9312 – ((282 x 282 b = 1135.4 / 1359.6 b = 0.84 a = 29.8 – 0.84 x 28.2 a = 6.25 Equation: German score = 6.25 + 0.84 x French score We can estimate the German score if the French score German score = 6.25 + 0.84 x 30 = 31.5 Regression can be taken further by looking at how gooddata and how precise any estimate from the equation is

six sigma

) / 10))

is 30 :

the fitted line explains the in predicting the Y values.

wdp consulting

The deviation of each observed value from the fitted line is known as the ‘residual’. These values can be plotted to test various assumptions that lie behind regression analysis. Assumptions: Residuals are : from a Normal distribution, stable (do not change over time), are not related to the X values, occur at random Any patterns in the plotted residuals indicate that the fitted line is not explaining the data. ANOVA The significance of a regression line can be tested via the use of Analysis of Variance. Source df Sum of Squares Mean Square F Regression 1 SSreg Residual n-2 SSresidual Total n-1 SStotal Where: SStotal = ∑ ( Y- Y )2 = ∑Y2 - ((∑Y)2 / n) SSreg = ∑ (Y – Y )2 where Y = predicted value from equation a + bX SSresidual = ∑ (Y

∧

∧

–Y )2

∧

six sigma


Using the previous example

X Y = 6.25 + 0.84(X) French German Predicted Y - predicted

10 11 14.65 -3.65 10 22 14.65 7.35 18 22 21.37 0.63 25 19 27.25 -8.25 28 35 29.77 5.23 33 27 33.97 -6.97 34 33 34.81 -1.81 39 40 39.01 0.99 42 42 41.53 0.47 43 47 42.37 4.63

SS residual = (-3.652 + 7.352 + ……….4.632) = 237.6522 SStotal = ∑ ( Y- Y )2 = ∑Y2 - ((∑Y)2 / n) = 10066 – ((298)2 / 10) = 1185.6 Source df Sum of Squares Mean Square F Regression 1 947.948 947.948 31.91 Residual 8 237.652 29.707 Total 9 1185.600 F 1, 8 (0.05) = 5.32 The regression is significant. German results can be predicted from French scores.

wdp consulting

Correlation This is related to regression and provides a dimensionless measure of the relationship between two variables. It is also known as Pearson’s Correlation coefficient. The value(r) ranges from –1 to +1 r = -1 : perfect negative relationship r = 0 : no linear relationship r = +1: perfect positive relationship The squared value of r measures the percent of variation in the Y-values that is explained by the linear relationship with the X-values. Warning: a high correlation does not mean causation. Mathematically :

r = ∑ ( X –X) (Y – Y ) / √(∑ ( X – X)2 ∑ (Y – Y )2) r = (∑XY - n. X. Y ) / √( (∑X2 – ((∑X)2 / n) (∑Y2 – ((∑Y)2 / n) ) From French – German data X = 28.2 ∑X = 282 ∑X2 = 9312 Y = 29.8 ∑Y = 298 ∑Y2 = 10066 ∑XY = 9539 n = 10 r = (9539 - 10 x 28.2 x 29.8) / √ (9312 – (282)2 / 10) (1 r = 1135.4 / √ (1359.6) (1185.6) r = 0.8943 r2 = SSreg / SStotal r2 = 947.9478 / 1185.6 = 0.7996 compare with 0.8943 x

six sigma

0066 – (298)2 / 10)

0.8943 = 0.7997


Exercise (Part 1) Calculate the regression equation and correlation for the following data. X Y 1 10 8.04 2 8 6.95 3 13 7.58 4 9 8.81 5 11 8.33 6 14 9.96 7 6 7.24 8 4 4.26 9 12 10.84 10 7 4.82 11 5 5.68 X-bar 9.0 Y-bar 7.5 X = 9.0 ∑X = 99 ∑X2 = 1001.0 Y = 7.5 ∑Y = 82.51 ∑Y2 = 660.173 ∑XY = 797.6 n = 11 Exercise (Part 2) The following data has the same regression equation as the previous data Y = 3.0 + 0.5X And the same correlation coefficient (r) B C D X Y X Y X Y 1 10 9.14 10 7.46 8 6.58 2 8 8.14 8 6.77 8 5.76 3 13 8.74 13 12.74 8 7.71 4 9 8.77 9 7.11 8 8.84 5 11 9.26 11 7.81 8 8.47 6 14 8.10 14 8.84 8 7.04 7 6 6.13 6 6.08 8 5.25 8 4 3.10 4 5.39 19 12.50 9 12 9.13 12 8.15 8 5.56 10 7 7.26 7 6.42 8 7.91 11 5 4.74 5 5.73 8 6.89 average 9.0 7.50 9.0 7.50 9.0 7.50

r 0.82 0.82 0.82 What is different? How do you know?


Lean Operation Lean Thinking has developed from the Toyota Production System (TPS). It is interesting that the emphasis of Lean, like Six Sigma, has been on the teaching of the tools. However, TPS is essentially about the culture and way of working. For example TPS focuses on 3 Ms: Muda (waste) Mura (unevenness – levelling) Muri (overburdening) All Lean teaching emphasises the removal of waste, some talk of levelling flow, but few ever mention the concept of not overburdening people or machines. In the West Lean Thinking has become synonymous with being ‘lean and mean’ i.e. doing more with less to such an extent that the organisation becomes ‘anorexic’. Such action does not fit in with the TPS concept of Muri. “experienced leaders within Toyota kept telling me that these (lean) tools and techniques were not the key to TPS. Rather the power behind TPS is a company’s management commitment to continuously invest in its people and promote a culture of continuous improvement” (The Toyota Way; Jeffrey Liker). Despite the above teaching from Toyota, Lean promoters still emphasise the tools. Lean Thinking has five main principles:

- specify value from the point of view of the customer - identify the value stream

o physical flow o information flow o new product introduction

improve value-adding activities eliminate non-value adding activities

- make value flow (remove obstacles / queues) - pull at the customer’s rate of demand (use one : make one) - seek perfection through continual improvement

There are two stages to the introduction of lean operation: Stage 1: Prepare the organisation for flow, flexibility, short lead time and high quality. Areas involved:

- Demand management (stability) - Standards - Maintenance (TPM)


- Quality (Prevention, 5S) - Layout (cell concept, workstations) - Change-over time reduction (SMED) - Team preparation (Policy, Rewards and Incentives, Training)

This stage is about preparing for the creation of a lean operation. Stage 1 does not need to be completed before starting on Stage 2, but work should have been started. Stage 2: Produce with agility and no waste. Areas involved:

- Team involvement (actions, opportunities) - Visibility and Simplicity (management by walking about) - Process data collection (inventory, machines) - Improvement (process mapping) - Master scheduling (stability, capacity, control) - Inventory management (Batch size, Buffers, Flow control) - Pull and synchronisation (Kanban) - Measurement (wastes and improvement)

Stage 2 are the techniques for operating in a lean way. They need some of the Stage 1 ideas to be in place. Some of the Stage 2 techniques may not quite as applicable for low volume manufacturing. It is not the purpose to cover every aspect of lean operations. The areas focussed on will be:

- 5S - Waste - On-going improvement


5 S This is basic housekeeping. Sort – classify everything by frequency by use. Red tag – label with date – if not used within a time period, throw away. Simplify – “a place for everything and everything in its place.” Use shadowboards. Sweep – physical tidy-up of anything out of place. Five minute clean up each week. Standardise – set standards and audit Self-discipline – participation and improvement The S’s are sometimes written as : Simplify, Straighten, Scrub, Stabilise, Sustain Or CANDO : Cleanup, Arrange, Neatness, Discipline, Ongoing improvement


Seven Wastes : Muda (Taiichi Ohno) Waste is defined as: “… Anything other than the minimum amount of money, equipment, materials, parts, space and workers’ time, which are ‘absolutely essential’” (To add value to the product) Taiichi Ohno defined seven wastes Waste : Muda Examples of waste

Overproducing too early, too much, just-in-case

Waiting materials queuing, not moving, people not productively employed, expediting

Transporting all material movement, double handling

Inappropriate processing too fast, too big, too variable, one big machine rather than several smaller

Unnecessary Inventory Stores, buffers, batch sizes, control systems

Unnecessary Motions Reaching, bending, exertion, excess walking, excessive turns to loosen

Defects Rework, rejects, unnecessary inspection, consequences of not doing the right thing first time

Taiichi Ohno said to the workers “may I ask you to do at least one hour’s worth of work each day?” Believing themselves to have been working hard all day long, the workers resented this remark. What Ohno actually meant, however was “Will you do your value-adding work for a least one hour a day?” So what exactly is meant by ‘value-adding’ work? Value-added Work

A primary operation that transforms, converts or changes a product towards that which is sold to a customer.

Non-value added work

Those operations or activities that take time and resources but do not add to the value of the product which is sold to the customer.


If the ‘value-added’ definition is applied to any process then it will be discovered that most of what goes on within the process is in fact non-value adding. Think; how much of your time is spent moving around or waiting for something to happen or arrive. How much of the work that you do adds value to the process or the product? You may even be employed in a department responsible for putting things right. At each step of the process value should be added to the product, information or service and then passed onto the next step. However, in many processes much that we do does not add value to the product. Overproduction: extra work put into making more items than are needed, usually just in-case there is a problem; rejects, failures. This leads to additional storage and may eventually result in this material being scrapped. How much of your material in the warehouse has been there for more than 12 months? Will it ever be used? Attitudes that create overproduction are:

Produce as many as possible because the cost per unit drops Encourage each line to focus on productivity alone Produce more to increase machine utilisation Such attitudes do not focus on the overall process and usually end up with having to deal with excess production. Waiting: in a production environment this is when the operator can do no value-adding work because they are waiting for additional material or a machine breakdown to be repaired. This waste can also be seen when someone is simply monitoring a machine that is doing the value adding work itself. In a administration this can be waiting for information or waiting for a decision. We can often fill the time with other things and therefore do not notice how much time is spent waiting. Transporting: this can often be overlooked as we see forklift trucks, conveyor belts as part of the production process. However the need for such transporting is due to a process being built around existing structures. Often when someone follows a product or piece of paper around a factory they are astonished at the distance it travels. The question should always be can this transportation be eliminated. In a chemical company the output from an extrusion process was transported via a tractor from one building to another for further processing. This involved delays time to load, unload and logistical problems. Inappropriate processing: this can be seen in a number of areas such as running machines knowing that they will produce rejects or simply adding additional steps in the process that add no value. For example a small circuit board arrives in one company inside a plastic bag separated in the box by dividers. Other boards arrive


simply separated by dividers without problems. The use of the bag adds steps into the process that appear to add nothing and leaves a company with lots of small bags to dispose of, or return to the supplier of the boards. Unnecessary Inventory: raw materials, semi-finished products and finished products do not add value. They have taken resources to produce and now occupy space, requiring a method and people to record and manage the storage. High inventory will often hide problems within a process. Rather than fix a process we stock more. Unnecessary Motion: this aimed primarily at an individual. How far does an operator have to walk to get a tool or how far does an administrator have to walk to get to a filing cabinet or a phone? This movement delivers no added value to the process. Defects: rejects interrupt production and require resources for either rework or disposal. In one company it was prepared to live with a defect rate of 15% and employing 10 people to repair the rejects rather than fix the process. Defects do not only occur in products; how many errors occur on an invoice leading to delayed payment and additional handling. Taiichi Ohno used the term ‘rejects’ for one of his wastes. This has in some companies been reworded as ‘correction’, i.e. the additional cost of dealing with the defect e.g. scrap, sort, rework. Eighth Waste In addition an eighth waste has been added to the original seven. This is the waste of not fully utilising human resources. This is probably the largest waste of all. Elimination of waste: many of the areas of waste appear to be simply part of the way we do business and few people think about removing them because they are not aware of the cost. It is estimated that in manufacturing 15-20% of turnover is wasted but for Service industries this can be as high as 40 – 50% of turnover. Often when a process is mapped these wastes do not appear. It is important to identify how much waiting time is involved how far things are moved. It is here that significant improvements can be made. Waste elimination and good housekeeping (5 S’s) go hand in hand.


Improvement Make improvement a way of life.

- Be aware of the seven/eight wastes - Question why things are done - Kaizen (on-going improvement) - Mapping (identify problems / opportunities) - Materials – reduce inventory - Methods – go to gemba (work place) to collect data - set standards and audit (adopt best practice not police)

Summary Creating a Lean Operation requires more than simply introducing 5S or Kaizen. Books such as ‘The Toyota Way’ by Jeffrey Liker will give more information. “these are just technical tools and they can be effective only with the right management and philosophy – the basic way of thinking. At the centre of TPS is people.” (The Toyota Way; J. Liker)


Leading Teams Part 1 – Self Team leader: Competencies and Behaviours Self-Assessment Questionnaire The Team Leadership Role The Facilitator’s Role Presentation Skills

Part 2 – Individual Giving and Receiving Constructive Feedback The Intervention Descriptive versus Evaluative Feedback Handling Feelings and Conflict Part 3 – Group P.R.E.P. Team Performance Review Team Behaviours


Introduction These notes have been compiled to give the reader an introduction into team leader and facilitation skills. It should be used as a reference point and as a guide. It cannot tell you exactly how to lead a team; that is down to an individual’s style. What it does do is give some key pointers under the main aspects, notably Self, Individual, Group and Task. Team Leadership is like juggling. A good juggler will have an awareness of their own ability and the mood they are in when juggling. The balls they have in the air relate to the task, the group and the individual. At all times the juggler must be aware of where the balls are, and will be catching the balls and throwing them up in the air again so they do not fall in the ground. As a team leader, you will make various interventions to the group you are facilitating to ensure the meeting is a success, as they juggler hopes his / her juggling act will be a success.


SELF

Team leader Competencies And Behaviours

Question No Individual

Score Total Score

Average Score

7 Spot opportunities for individual or departmental improvement

20 Turn good ideas into action

31 Try to change things rather than just live with them

14 Take prompt action when opportunities arise

Initiative Average

26 Change current thinking for the “way we've always done it”

1 Adapt to new ways of doing things

41 Engage in idea generating or "what if" discussions

19 Generate innovative ways to deal with situations

Creativity Average

30 Exercise confidentiality when dealing with sensitive issues

3 Deal with people consistently

40 Keep promises once they are made

22 Inspire the trust of others

Personal Integrity Average



Score Total Score

Average Score

2 'Reads' and interprets people's behaviour accurately

42 Has a flexible style in dealings with others

29 Appear to be approachable and easy to talk with

32 Ask questions to learn more about (or to clarify) what you are saying

17 Give praise / recognition where appropriate

Skills With People Average

12 Identify the process being used to deal with a situation

21 Focus on the key issues in a mass of detail

10 Think of practical ways of implementing ideas

39 Demonstrate an ability to think on your feet

Process and Logical Thinking Average

4 Describe what took place when giving feedback

36 Remain uninvolved when observing situations

27 Make insightful comments on behaviour in a group situation

6 Identify trends and patterns in situations

38 Actively listen and note what others say

Observation Average



Score Total Score

Average Score

13 Keep appointments and arrive at meetings on time

33 Plan meetings in advance as appropriate

5 Organise activities and timescales to achieve objectives on schedule

23 Establish contingency actions in case things go wrong

Planning and Organising Average

9 Choose an appropriate time to interrupt a meeting

37 Keep the overall picture in mind when looking at details

25 Show a willingness to change their mind

16 Assess the possible consequences of decisions

Judgement and Decision Making Average

28 Make effective formal training presentations

11 Express ideas or information clearly when speaking

35 Presents a good first impression

15 Answer questions specifically and to the point

Presentation

18 Act as an effective facilitator between people who have different points of view

34 Persuade people to change their minds

24 Anticipate objections or counter-arguments and have solutions ready

8 Influence people not under your direct control

Influencing Average


The Team Leadership Model

All teams have leaders. Sometimes these are elected or pre-chosen, and sometimes the leader will emerge as the team begins to work together. Good leaders will have a certain set of behaviours; people who possess these naturally are likely to be those who emerge as leaders of the group, without the formality of being chosen. However, most people will need to develop their team leadership skills, and the following model is designed so that team leaders can assess what they are strong at and which skills require some development. These behaviours are shown as a set of competencies that fall into certain categories. Each competency is defined in the table and two other columns are also shown. The first of these highlights the typical behaviours that one would expect the leader to exhibit. The last column shows what the outcome or impact of that particular behaviour should be on the team. The model is split into two parts. The first summarises the competencies involved, showing them in 4 distinct categories. The second is the table, which shows everything in more detail. Both of these are shown below.

The Team Leadership

Model

LookingAhead

• initiative and risk taking• vision• planning and goal setting• creativity / innovation

PersonalValues

ManagingPeople

ManagingOutputs

• personal integrity• performance under pressure• diversity

• empowerment• communicating• delegating• motivating• coaching• team building• mentoring

• quality• problem solving and

decision making

Competency Definition Typical Behaviours Outcome/Impact On Team

Initiative And Risk Taking

The ability to demonstrate proactive thinking, drive and accept responsibility for their actions.

Spotting and acting on opportunities; taking personal responsibility; persevering; displaying energy; standing up for what they believe in; showing a willingness to take the lead; showing a willingness to take risks; making things happen and continuously looking forward.

Team has a sense of urgency. Team members feel they have encouragement to take measured risks and initiatives, but within the context of the wide picture. Team never happy with the status quo.

Personal Integrity The ability to gain the trust and confidence of others by acting fairly and honestly.

Keeping their promises; dealing with sensitive issues in a confidential way; sharing information openly; dealing with people fairly and consistently; representing their team to the organisation and the organisation to their team.

Members feel that they will be supported and the team defended where necessary. People feel that they can speak up without fear of reprisal. Team has confidence that the leader will deliver on promises. Team sees leader as its champion. Team are comfortable in their working environment, and have a sense of equality.

Vision The ability to create and describe an ideal state or condition and focus people’s attention on it.

Taking a step back from day to day activities; describing and discussing the purpose and strategies of the team; reviewing the purpose periodically; keeping people focused on the future rather than the past; inspiring others; displaying charismatic behaviours which support the vision; acting in a way that is consistent with the vision.

People understand why the team exists, where the team is going, and their own part in the scheme of things. Team feels a sense of purpose. The team has its own identity.

Quality The commitment to produce high quality results via high quality processes.

Producing what is required first time round; operating in an evidence based manner; demonstrating a consistent personal commitment to quality; balancing attention to results with attention to work and interpersonal processes; performing work that exceeds expectations.

People are committed to continuous improvement; people are encouraged to regularly improving their performance; people are encouraged to view mistakes and problems as learning experiences. Team takes a pride in its output.


Empowerment The willingness to create a work environment in which people are encouraged to develop their full potential.

Allowing people to use their judgement when making decisions; allowing people to initiate actions; involving people on key projects; creating a learning organisation.

Team members feel that they have worthwhile work to do and worthwhile responsibility; people feel that they have the freedom to develop and express themselves through their work; people are willing to take responsibility for their actions and are given authority for actions.

Communicating The ability to express oneself clearly and to listen to others.

Expressing ideas clearly; presenting information in a well-organised way; answering questions specifically and to the point; allowing people to finish what they have to say; asking pertinent questions; showing an interest in what others are saying; aware of others feelings and concerns.

Team members understand what is communicated to them. They feel listened to. Misunderstandings about issues are uncommon. Everybody understands the same thing. Team as well as individual performance is known and understood. Team communication with others is clear.

Delegating The ability to identify and assign work, clarifies expectations, and defines how individual performance will be measured.

Explaining new assignments clearly and patiently; using delegation as a development tool; ensuring people receive training for tasks; letting people know what is expected of them; assigning responsibilities that match individual capabilities.

Team members receive interesting and challenging work to do. They understand what is expected of them and how they will be measured. Skill levels and coverage are raised within the team. Tasks are perceived to be allocated fairly.

Motivating The ability to create a satisfying work environment that encourages others towards achieving team goals.

Setting an appropriate example; providing a sense of belonging to the team; praising individuals or the team for good work; recognising contributions to team performance.

Team members feel committed to individual and team performance; there is an identifiable team spirit; people enjoy working in the team. Team output is high.

Coaching The ability to provide feedback and offer support when people are confronted with performance problems and/or development requirements.

Giving constructive performance feedback to individuals or the team; taking time to develop people’s effectiveness; looking for opportunities to create learning; encouraging team members to learn from each other.

People feel that they can approach team colleagues for help; performance problems are dealt with quickly and effectively; team members are prepared to admit their mistakes. People have a positive attitude towards development.


Planning And Goal Setting

The ability to plan and organise work in a structured and visible manner, and set realistic, measurable objectives for individuals and the team.

Developing realistic plans and timetables; setting clear, measurable objectives; using time effectively; establishing priorities.

There is congruence between individual and team goals. Objectives and schedules are clearly communicated. The team creates a disciplined environment that does not inhibit the freedom of its members. The team is involved in the planning process. Plans are changed as a result of environmental influences and tasks completed on schedule.

Problem Solving And Decision Making

The ability to analyse a situation, identify root causes and alternative solutions, and make appropriate decisions.

Making sense of complex situations; identifying problems at their early stages; involving team members in problem solving; dealing with the cause of problems, not just the symptoms; encouraging democratic decision making; demonstrating a willingness to take tough decisions.

There is an appropriate balance between team and individual decision-making; decisions are made in a timely fashion based on facts as well as opinions. Team members feel that they have been involved in solutions. Difficult decisions are not put off. Problems are shared amongst the whole team.

Creativity And Innovation

The ability to develop and apply new and innovative ideas and practices.

Encouraging team members to be creative in their work; making sure that good ideas are shared within the team; adapting to new ways of doing things; listening to the ideas of others even when these do not coincide with their own; engaging in idea generating or ‘what if’ discussions within the team.

Many ideas are generated within the team. Many ideas are adopted. People feel valued for their creativity. Team members share ideas with each other.

Team Building The ability to ensure that the value and effectiveness of the team is greater than the sum its individual members.

Paying attention to the team’s internal processes such as purpose and commitment, harmonious group working, contributions from individuals, interactive processes and team focus; ensuring that the team and its members enjoy excellent relationships with other teams and with its customers.

The priorities of the team are in line with those of the organisation. The team functions effectively as a unit. Team members feel that the team is good to belong to. The team is seen as adding value to the organisation. Customers feel able to give feedback that will be listened to and result in action.


Performance Under Pressure

The ability to maintain high levels of performance and personal conduct in challenging or unpredictable situations.

Remaining calm when confronted with problems or crises; demonstrating strength of character in adversity; acting with professional detachment and objectivity; being able to adapt to changing circumstances.

Team members feel that external pressures are being managed. Team is aware of steadiness of purpose. Team’s output is not unduly affected to external pressure. Team can take changes on board.

Diversity The willingness to work with individuals and integrate the differences that exist among others.

Making the effort to deal with others in a co-operative fashion, including those in other teams; negotiating differences of opinion openly and fairly; demonstrating a willingness to be flexible and open minded; bringing together people with varied talents and perspectives to resolve work group problems.

Team members work well with each other and with the team leader. The team are compatible with and able to encompass and synergise the many different ideas and approaches of its members.

Mentoring The willingness to support others in their personal and professional development.

Identifying career and development opportunities for team members; increasing levels of responsibilities to prepare people for a higher level position; sharing relevant personal insights or experiences; providing suggestions for personal and professional development.

Team members feel that positive thought is being given to their future and feel able to request only support needed. Team members develop during the life of the team and receive feedback on their performance at regular intervals.


Presentation Skills Structure

Opening impact Tell them what you are going to tell them Main body of presentation Tell them what you have told them Close

Techniques

Start with the listener Appearance Eye contact Use examples Distracting mannerisms Visual aids

Presenting yourself Talk from experience

When a team member illustrates a proposal with examples of someone else doing something else somewhere else, attention becomes short. When the same team member gives a personal example, something that happened to them, eyes swivel back and attention is complete. Speak to, not about people present

It may seem inconceivable that team member A will speak about team member B or something that team member B has done by addressing everyone else but team member A, but watch for it as this frequently happens. A statement like: “I thought Joe’s proposal was excellent, we should take this further” may sound positive and supportive, but the speaker has not actually made any personal contact with Joe. The message may have been delivered but the relationship or contact between the team members has not been enhanced.

Address the person by name

Not always, but from time to time. The use of someone’s name suggests that the speaker is focused on the person spoken to (rather than rambling in his or her own head) and therefore can help to establish contact.


Look at the person you are talking to

Many people attempting to express a complicated idea will look up to the ceiling or out of the window. In such cases contact will be maintained if the speaker catches the eye of the person spoken to before and after launching into an exposition. Team members who habitually talk whilst looking out of the window may discover they are not heard very well. Say ‘I’ not ‘we’, ‘you’, ‘one’, ‘people’, etc.

Again, members who take responsibility for their own opinions make strong contact with the rest of the team. Confusion, if not lack of interest, often results from the use of ‘we’, ‘you’, etc. Opinions are often then disguised as established facts. Make statements before questions

Behind every question lies a statement. The statement will always clarify the question and make a response much simpler. When a team member repeatedly challenges an idea or asks question after question, it is frequently because they already have an opinion and are searching for information that will confirm their opinion. This process often prolongs unnecessary discussion, frustrating everyone except the individual who is holding back his or her ‘ace’. Flush hidden positions out with the question: “Is there a statement behind your question?” Trace opinions back to observations

When giving feedback to a team member on performance. Only specific instances, described as if by a video camera, can avoid an argument or further discussion. Describe, don’t judge

Judgement implies evaluation. Evaluation involves comparison. Comparison established difference. Differences establish how people are separate from each other. While recognition and acceptance of differences is fundamental for individuals to make contact, difference should not be used as a means to compare and judge the relative world of others. Feelings contribute

Even apparently negative feelings can help the team become more aware of what it is doing. Feelings result from our interpretation of what we see happening. If your interpretation is correct some other team members will usually be experiencing the same feeling. Recognition of this can instigate an immediate positive change of focus.


Say ‘I would like’ rather than ‘You should’ or ‘We should’

‘I would like’, is followed by a statement. ‘You should’, is followed by an opinion, which will frequently lead to a lengthy argument. Saying ‘I would like’ is clear and can drastically shorten meeting time.

Summary If we frame information clearly and speak appropriately to our team-mates and colleagues, we understand each other better and generate more genuine contact. Right relationships are enhanced. If we foster good contact amongst team members and in our relationships with others, we create an environment in which the content of our communication is given the meaning we intend by the context we create. The circular causality of content and context in communication should be apparent: good communication fosters good contact amongst colleagues and vice versa.


INDIVIDUAL Giving And Receiving Constructive Feedback For most of us, feedback is an essential ingredient of learning and developing. You will have many opportunities to give and receive feedback from others. It is important to make your exchange constructive, so that each person’s development is supported. Guidelines for Giving Constructive Feedback

1. Be timely Give your feedback as soon as possible after the event, in an appropriate setting.

2. Be specific Describe specific behaviours and reactions; particularly choose those they should keep and

those they should change.

3. Be descriptive not judgmental

Describe what you see, hear and feel, do not make assumptions about why

4. Describe behaviour not personality

Focus on what someone does, not what you think they are. Behaviour is easier to change than

personality.

5. Be constructive not destructive

Why are you giving the feedback? Is it to make yourself feel better or to help the receiver?

6. Own your own feedback

Speak for yourself, not for others.

EXAMPLE 1: Judgement, speaking for others, personality focused: “You should not

be so aggressive; it is rude, and it got everyone upset”. EXAMPLE 2: Descriptive, speaking for self, behaviour focused: “When you

interrupted me several times, I felt as though there was no point in explaining my idea”.


Guidelines for Receiving Constructive Feedback

1. Ask for it Be selective, find people whose views you will value. Ask them to be specific and descriptive.

2. Direct it If you need information, ask for it. Tell people what you are trying to find out, and explain the kind of feedback that will be especially helpful to you.

3. React positively Accept criticism by acknowledging that there may be some truth in what they say, or how they see it, without having to apologise.

4. Accept it Do not defend or justify your behaviour. Listen to what people have to say. Use what is helpful; quietly discard the rest.

EXAMPLE 1: Defensive, not listening: “No, you are completely wrong there. I did

it like that because I had to and I know I am right, and I am not prepared to listen to anything else”.

EXAMPLE 2: Positive, accepting: “Yes, I can see from your point of view that it

might look like that. Thank you for telling me, perhaps I need to explain it in more detail”.


Intervention 1. WHEN TO INTERVENE The facilitator should intervene (i.e. interrupt) the meeting when he/she has something important to which to draw the group’s attention. For guidance on how to do this read again the notes on Descriptive Feedback and Ways of Speaking. Balance. You should be constantly aware of the balance of meeting behaviours between Initiating, Clarifying, and Reacting. If one of these is lacking for any length of time, then the meeting will become ineffective. Typical Problems Caused By Dominant or Missing Behaviours

When a meeting is unusually Characteristic Problems are

High on INITIATING

• too many ideas and alternatives to handle • lack of attention to detail • “up in the clouds” feeling

Low on INITIATING

• meeting becomes backward-looking • lack of enthusiasm and excitement • undue attention to detailed analysis

High on REACTING

• meeting becomes emotional • misunderstandings become more frequent • people take sides and issues become entrenched

Low on REACTING

• tendency for repetition • people withhold important information • meeting is awkward and forced

High on CLARIFYING

• very time-consuming • obsession with minor issues • “swimming in syrup” impression

Low on CLARIFYING

• meeting becomes disorganised • hasty decisions are made • people cannot agree afterwards on what has

been decided

For example, in a meeting that was low on clarifying, the facilitator might try to rectify matters by supplying the missing behaviours himself. Thus he/she might spend time on seeking information bringing out the details of proposals under discussion and encouraging group members to follow his example


He/she could also test understanding and summarise to make sure that these important activities are not neglected and the correct degree of understanding is achieved. He/she could also bring in other group members encouraging them to give information on the reasons for their initiating and reacting behaviour. He/she could shut out group members who attempted to overload the meeting with further initiating and reacting behaviour. 2. BRINGING IN AND SHUTTING OUT Bringing in involves identifying those members who are low contributors (either because they are naturally so or because they are being shut out by high initiators). Use expressions like: What do you think, Harry? You seem to have gone a bit quiet lately, Jean, is there something you want to say? Do you agree with this, Angela? Tom, this subject is one in which you have some experience. What do you think about it? The important thing is to give people the OPPORTUNITY to speak, although whether they take it is up to them. Shutting out can be done deliberately by the Facilitator in order to give others a chance to speak. Shutting out is basically an interruption and needs to be done very carefully. Here are some examples:- Before you continue, Jim, perhaps you might like to ask the others what they think. I think you should get people’s reaction to what you have just said. Are you sure people understand what you have been saying? Let’s leave that for now and move on. We can always come back to it later when people have had a chance to digest it. 3. DEALING WITH CONFLICT AND DISAGREEMENT Encourage people to concentrate on the arguments rather than personalities. Encourage people to look for the good points even if they disagree with other parts of the proposal (“3 likes and a wish”). If the disagreement is between two people check to see what the other team members think. Test for consensus; ask each team member to give their view individually.


“Park” the item for tabling later after a cooling off period. Try to make the protagonists see that merely re-stating their positions will convince nobody. 4. “BEHIND THE SCENES” Sometimes the meeting is not the most appropriate setting for the facilitator to carry out his/her full role. Some things are best dealt with on an individual or more private basis before or after the meeting, or during a break. The facilitator can use these occasions to take soundings about what people feel about the meeting, and can act as a “go-between” between individuals or sub-groups. The facilitator can also use this time to talk to members privately about their performances as team members. For example, if someone is not contributing, or is over-contributing, this can be pointed out to them in private rather than in front of the whole group. Descriptive Versus Evaluative Feedback Descriptive feedback describes what you see or hear the other person is doing or saying. Thus “I notice that you look out of the window when I’m talking to you” or “Bill, you’ve been talking non-stop for the last three minutes” are both forms of descriptive feedback. Based on evidence ‘I imagine this is .........’ Evaluative feedback, on the other hand, involves giving a judgement on what you are seeing or hearing. “I think your performance is bad” and “Your man management skills are weak” are both forms of evaluative feedback. Descriptive feedback is more likely to result in change because:

it is more impartial

it is not emotional

it is difficult to argue with. Evaluative feedback is more likely to produce a negative or defensive reaction because the person themselves feel criticised. Descriptive feedback can also be used to describe your internal feelings: thus “I feel happy when you do that”


Exercise on Descriptive Feedback Look at the list of statements below. Which of them are descriptive and which are evaluative? How might you convert the evaluative ones into more descriptive language? Your presentation skills are poor Your presentation skills need developing In your speech I did not notice any mention of costs I feel uncomfortable when you discuss my performance as a facilitator I feel that you should look at Peter when you talk to him I notice that you always disagree with what Tom says and never with what Harry says Your methods of man management are very good Your methods of people management need improving I like the way you get to the heart of a problem People come to you for advice You are very reliable


Handling Feelings and Conflict Handling feelings and conflict is an important part of a team leader’s role; challenging the status quo often causes difficulties and skills are needed to deal with the situations that arise. These skills often involve separating:-

issues from solutions facts from opinions feelings from logical reasoning.

Whilst sometimes uncomfortable, conflict can be positive when it improves the quality of decisions, encourages creativity and interest. It can provide a medium for problems to be shared and tensions released. There can of course also be destructive outcomes of conflict such as discontent, diminished communication, and reduced group effectiveness. A framework for tackling difficult situations can be helpful, particularly if people find dealing with feelings and conflict unpleasant. This is shown below as a series of steps. 1. Face up to the issues and acknowledge that these exist. Encourage people to be open,

communicating that you feel issues are present and affecting the performance of the group.

2. Ensure there is common understanding of the different positions. Encourage people to express their wants, wishes and preferences, and restate the positions expressed.

3. Define the issues, trying to reach a mutually acceptable definition of the problem. Avoid a combative approach and continually reiterating the positives of each position.

4. Search for and evaluate alternative solutions - maximise collaboration on common ground and separate creating alternatives from evaluating them.

5. Gain agreement on the best solution - ensure everyone knows who will do what by when.

The solutions generated may happen in a variety of ways that are as follows:


• Competition (assertive and uncooperative) - when one party seeks to achieve certain goals, without regard to the impact on others in the process.

• Collaboration (assertive and co-operative) - when each of the parties in the conflict

wants to satisfy fully the concern of all parties, and want to work towards a mutually beneficial outcome.

• Avoidance (unassertive and uncooperative) - a party may recognise that a conflict exists

but reacts by ignoring it, withdrawing or suppressing the conflict. • Accommodation (unassertive and co-operative) - when parties seek to appease their

opponents so much that they may be willing to place their opponents’ interests above their own.

• Compromise (midway on both assertiveness and co-cooperativeness) - when each party

has to give up something important to them, and therefore the outcome is compromised. However much you desire to handle feelings and conflict, inevitably blocks to this will arise such as:- Wrong language - combative rather than collaborative Not confronting - people failing to acknowledge differences Excessive talking - not listening to others point of view, being too ready to criticise Misuse of language - ‘people - blaming’ rather than ‘I’ statements; high level of emotion, and repetition of views Poor anger control - people's frustrations very evident In addition blocks to resolving conflict situations may also stem directly from people asking poorly thought through questions which can lead to greater anxiety, rather than reducing it. For example if your boss asks you ‘where were you yesterday?’ it could either be followed by: ‘because I wanted to invite you to lunch’, or ‘because I could not find you and had to deal with a problem which was your responsibility. Questions thus spring from a variety of directions or motives and if those on the receiving end are unsure about these they may be defensive, guarded or aggressive. All of these blocks can be reduced by the use of facilitation skills.


P. R. E. P.

PLACE RELATIONSHIPS END RESULT

PLAN P.R.E.P. How many meetings have you been to where one of the following has happened? • Some people do not know why the meeting is taking place, or have different views • Some people do not know each other


• Attendees do not know why they are there • There are not enough chairs to go round • The room is too dark • People are unsure about the agenda • The meeting never really seems to ‘get going’ • Little, if anything, is achieved. Most people have experienced some of these problems, yet they are quite easy to avoid. They are allowed to happen by everyone getting immediately involved with the content of the meeting or their own agenda. A little time invested at the start of the meeting can pay big dividends when it comes to achieving meeting effectiveness. The process outlined below is a means of ensuring that important matters are attended to before the agenda starts in earnest. The process is called P.R.E.P and should be carried out at the beginning of every meeting. Usually, the shorter the meeting the less time is needed for P.R.E.P., but it should always be done. It should be conducted by the Leader of the meeting, and all members should be actively involved. Introducing P.R.E.P. may be done very overtly by saying that ‘P.R.E.P.’ will be used, or more subtly by simply ensuring all four steps are covered. This will depend on the mood of the meeting. Place Pay some attention to the physical environment in which the meeting is taking place. Is the room appropriate? Is the arrangement of furniture as it should be? If visual aids are to be used can everybody see them? Are the temperature, lighting etc. appropriate? Will everyone be able to hear what is going on? Is the venue ok? Or should an alternative be found next time? Relationships


Consider the meeting attendees. Does everyone know everyone else? Has this group worked together before or are they new to the meeting and each other? Are there any special roles involved? For example, who is the Chair; should there be a Facilitator, a Scribe, a Time Manager or a Recorder? Does everyone know what is expected of them? For teams that have been meeting for a while, there is less of a need for introductions, but often a need to establish outstanding issues. When did you last meet? What has happened in between? How do people feel about what is going on? How is the team relating with people external to it? End Result Everybody must be clear about the purpose of the meeting. If the group is to meet several times, the purpose of each specific occasion needs to be clarified at the beginning of each meeting. Where specifically do you want to be at the end of the allotted meeting time? What should have been achieved? Plan / Process This involves agreeing a means of getting the team to the end results specified. The agenda, which should have been sent out beforehand, should be clarified by the Team leader, who should ask if everyone understands it and thinks it is feasible. Time slots should be allocated, including time for warm up and warm down.

Team Performance Review All teams go through a process of development whereby they move through four stages from forming to storming to norming and then performing. This process does not happen automatically, and requires time for the team members to review what is happening within a team and assess the performance being achieved. As far as possible this should be based upon facts, with limited emphasis placed upon opinions, although these do of course have their place. For example, if someone dominates the discussion, others perceptions and opinions may be that the person does not want to listen to what others have to say, whereas the opposite may be true. By stating the fact that other participants of the meeting observed the person talking more than others, it becomes easier to establish different norms of behaviour, conversely, telling someone they are dominating is more likely to cause resentment and ill-feeling. To review performance the team needs to break away from what it is doing and undertake the following three key stages:- 1. Each individual should spend time reflecting on what is going on within the team - i.e.

outside themselves. There is a series of issues that could be thought about:-


Achievements and Assessment of processes used Roles and relationships Other people - attitudes, actions and body language Making decisions Ambience - the place and environment 2. Each individual should then assess how they feel about the meeting in terms of their own

level of comfort and emotion i.e. inside themselves. 3. Based on assessments made by the individuals, the team should then share some of their

thoughts; assess what is working well in the meetings and what could improve. Once this has happened the team should look for any patterns that may emerge - for example if the team recognises that it consistently generates a lot of ideas but does not actually manage to do anything with them, then some change of process is necessary. As far as possible the developments should be based on facts and descriptions, with limited emphasis placed on perception and gut reaction, although recognising the value this can have. Each time the team meets they should focus on one thing about their behaviour they are going to develop.

By going through this process the team will develop closer and more productive working relationships more quickly, and will help the team and the individuals within it get a greater sense of satisfaction.


Team Performance Review A

Achiev

emen

ts &

Awaren

ess of

Proce

ssR

Roles &

Relationships

OOther People M

MakingDecisions

AAmbience

Outside

Insideof

Self

of Self

How well is the team working?What can we change?


Team Behaviours When teams are working together there are three types of behaviour that may be observed, especially when ideas are being brainstormed and developed. These are initiating, reacting, and clarifying. To ensure all members of a team understand a new concept, the facilitator should ensure that all three behaviours are brought out. Initiating, Reacting, Clarifying Initiating This behaviour involves making suggestions or proposals, building on or modifying someone else’s proposal, and inputting new ideas. Reacting This behaviour involves supporting (either verbally or non-verbally giving support for another person or their concepts), disagreeing (stating disagreement or raising obstacles or counter arguments), and defending/attacking (attacking another person or behaving defensively). Clarifying This behaviour involves testing understanding (establishing whether an earlier statement was understood), summarising (re-stating previous discussion in a more succinct form) seeking information (seeking facts, opinions or clarification from others) and giving information (offering clarification to others).

Initiating

ClarifyingReacting


GROUP PERFORMANCE

By Simon Haben


INTRODUCTION & DEFINITIONS In this document I explain some of the theory relating to groups and performance and then follow this by my own experiences, outlining improvements that could be made. The first part of this document defines groups and performance; the second illustrates different types of groups and their uses; the third examines the influences on group performance. Before examining the factors affecting the performance of a group it is worth defining the two key words, group and performance. According to Mullins (1989) "a group consists of a number of people who have:

a common objective or task; an awareness of group identity and 'boundary'; a minimum set of agreed values and norms, which regulates their relatively mutual interaction."

This definition of a group is rather a 'clinical' one and I feel Handy (1993) expresses the definition of a group more practically:- "Groups vary from the formal - a work group, a project team, a committee, a board - to the informal - the ad hoc meeting or discussion, the luncheon group, the clique, the cabal. They are permanent or temporary. They are liked by their members or regarded as a waste of time. They can be a most effective device for blocking and obstructing new ideas, or the best way of putting them into practice.......They have a common objective." The definition of `performance' according to the Oxford Dictionary is `the execution, carrying out, doing, notable feat, performing of play, etc.’ Once a group has formed, it is not too hard to expect some level of performance according to that definition. However, what I am explaining in this paper is how effective the performance of a group is. Performance is something that can be measured tangibly, both in terms of quantity and quality.


GROUPS Groups are an essential feature of the work pattern of almost any organisation. Once in a group, the people within it influence each other in many different ways. The group may even develop its own hierarchy and its own leaders. Both may be different to that of the organisation from which it originates. Group pressures can have a major influence on the behaviour of individual members within that group and their performance to the group's overall aims. The activities of the group can also be closely associated with the process and style of leadership adopted by the leader / facilitator of the group. There are two different types of groups: formal and informal. Formal groups are created to achieve specific organisational objectives and concerned with the co-ordination of work activities. People are brought together on the basis of defined roles and the structure of the organisation. The nature of the task to be undertaken is an important feature of the formal group with goals being identified by management and certain ground rules, relationships and perhaps norms of behaviour established. Informal groups are based more on personal relationships and agreement of group members than on defined role relationships. These groups tend to satisfy psychological and social needs, not necessarily related to the task being undertaken. These groups tend to devise ways of attempting to satisfy member’s affiliation and other social motivations that are perhaps lacking in the work situation. The membership and roles of informal groups can cut across formal structure. It is very unusual for an informal group to be exactly the same as a formal group. Understanding the informal group network can be criticised in determining how to facilitate formal groups.


GROUP PERFORMANCE The performance of a group is essentially its effectiveness. This is influenced by:- (1) the givens i.e. the group, the task, the environment; (2) the intervening factors i.e. leadership / facilitation style, motivation, processes and

procedures; leading to:- (3) the outcomes i.e. productivity and member satisfaction. Below I have examined the givens and the intervening factors that lead to the outcome. The Givens The size of the group is an important factor in its performance. As group size increases, problems arise with communication and co-ordination. If a group becomes too large it may split in two sub-groups with friction developing between them. I believe that the optimum performance of group size is of five to seven people. In my experience groups with more than 7 or 8 permanent members starts to split into sub groups, and under perform. While this may be beneficial if both groups are focusing on different aspects of the task, this is often not the case. Compatibility of members of a group is also important. The more homogenous the group in terms of background, interests, attitudes and values, the easier it is to promote cohesiveness. Differences in personality or skills in the group may actually complement each other and help the overall performance. However, they can also cause difficulties and conflict may arise when members are in competition with each other. If groups are too cohesive, they can suffer the phenomenon of `group think', where people are so in tune with what the others are thinking they do not always see the broader picture and may overlook something incredibly obvious. The individual roles in a group can therefore be important. Belbin (1981) has done some study into this and defines a variety of roles within a group. For example the Shaper will be the task leader and is needed to spur action; the plant will be introverted but have many ideas; the finisher will check details and deadlines. When setting up a group it is not often possible or practical to test prospective members to find out which role they are likely to play. However, an understanding of Belbin's roles and people’s background is useful. An assessment can then be made of what different members the group is likely to need (ensuring a good mix thereby overcoming any problems of 'group think') and of people most likely to meet those characteristics when recruited to the group. Where a group has been pre-selected, the leader / facilitator of the group should try to maximise performance by encouraging people to adopt a role slightly different to the one they would normally play.


It has been found that the more complex the task is, the more important it is for the group to work well together and the need for consensus on a focal person or leader. On a project I was working on, the group leader, although very senior in the organisation, was very weak, offering almost no leadership, this resulted in fragmented group, no focus, much time wasting and duplication. Good performance therefore requires that a leader / facilitator should be picked because of ability to direct and influence rather than status. Hidden agendas can give rise to conflict, and will prevent trust from building. Hidden agendas may be protecting the interest of one sub-group, impressing the boss, making a particular alliance, or even covering up past errors. For as long as hidden agendas exist, group performance is likely to be hindered. It is important for the group leader / facilitator to reveal hidden agendas, either by talking to individuals separately, or preferably encouraging members to admit their own issues to the group. Once recognised, hidden agendas can be overcome by careful allocation of task. It is my experience that this can be a painful process, however the long-term benefits are worthwhile. The type of task will naturally affect the kind of group that is formed. The benefits of task allocation are that it allows a structured approach. The group that is formed to formulate ideas about the detail of the task needs a more supportive management style and may not initially have a structured approach to its work. The timescale of the task also reflects how the group will work together. If the timescales are very tight, then the group may not develop to deal with individual needs. This can result in work of a lower quality than if the timescales were broadened. However, performance in terms of quantity may be a lot higher with group members motivating each other. Another important factor is the 'salience of the task'. The more important the task is to the individual within the group, the more committed they are to achieving it. This enables greater demands to be placed upon the group. It is important that the task is clear and unambiguous. This will assist the group through the stages of group development. However, if the task requires ambiguity, then this can lead to the formation of different roles within the group and also increase stress, therefore having implications for the leadership of that group. The timescale set should be realistic so that good performance can be achieved, but not too long as procrastination may occur. In my experience, groups at a very minimum need a vision, and if this is not specific the group must make it so as soon as possible. I once worked with a group that had no plan of what it was trying to achieve, motivation was low, and the outcome was poor. Performance will be improved the task is made clear and a plan formulated with the groups involvement. Group cohesiveness will often be enhanced if members are situated in the same location or within close proximity of each other. Physical barriers can hinder cohesiveness and often act as communication barriers.


If a group is formed to work on a task that is not part of there everyday work, it can be useful to have a 'project room', so that members can work in the room when working on the project to avoid interruptions. I have worked on a project where this has not been the case, and performance has suffered due to the level of interruptions and distractions on everyday matters. However the needs of the project would dictate the need for this facility. Communication is important in developing group cohesiveness. The more easily groups communicate with each other the more cohesive they become. People have varying levels of belonging needs. Groups of people that have higher belonging needs are likely to require greater access to communicate with each other. Regular meetings ensure a certain level of communication and make people feel part of the group, leading to better performance. However it is important the meetings allow discussion. One group I worked with held regular meetings, but it was essentially a briefing, and discussion was discouraged by the leader of the group. The result was that people did not feel as though they belonged to the group. Technology can affect the way in which work groups function. The nature of technology can be unfavourable for the creation of work groups and a source of alienation, especially for manual workers. The impact of information technology is likely to lead to new patterns of work organisation, and affect the structure of groups. For example, physical location may cease to be a problem in forming a group if information technology links can easily be made. However this does not allow much interpersonal contact, which may negatively affect the performance of the group. If technology is likely to lead to alienation, mechanisms should be put in place to overcome this. External threat can serve to dramatically increase the cohesiveness of a group, with members uniting against the common threat. This may continue once the threat has been removed. If this occurs, the group should be managed to make best use of the increased cohesiveness, ensuring effort is focused in the right direction. However threats should not be over-exaggerated and the group leader must be honest about the threat. The Intervening Factors These factors can be changed in the short term to help the performance of a group. Leadership / facilitator style can affect a group substantially. A change in style can have tremendous rewards for group performance, and therefore the style has to be 'right' for the groups and the situation. Conversely, poor leadership / facilitation is likely to result in poor performance. Guidance, support and encouragement must be tackled according to the group needs, ensuring good employee relations. To improve performance the leader / facilitator must be constantly aware of group and individual needs and try to satisfy these. In my experience the more successful groups have leaders who are


willing to listen, observe and react according to the situation, and thereby maximising participation. Poor performing groups often have leaders who 'bulldoze' on regardless of the group. Motivation is a key influence to group performance. If individual motivational needs are not recognised and met, individual satisfaction and performance will diminish which will inevitably lead to poorer group performance. Performance will improve if individual’s needs are clearly recognised and met. Motivation theory suggests, and my own experience supports this, that individuals have differing needs. Some are motivated by money, but many by recognition and the sense of achievement. It is therefore important to reward people satisfactorily, give praise where it is due - both privately and publicly, and ensure individuals have from which they will get a sense of achievement from completing.


Group motivation will be enhanced if individual motivation needs are satisfied and if the members like each other; if they approve of the task; if they wish to be associated with the standing of the group in the organisation; and if members know the expected results and are given feedback. Motivation will show itself in the actual behaviours of an individual that will be related to the role they play in the group. It is therefore important to regularly clarify these. The diagram below shows the factors affecting positive and negative performance. Various processes and procedures have to be undertaken in order for any group to be effective. These can be categorised into two functions - task & maintenance. The task function is often focused around problem solving, and the way in which a group tackles this may affect the solution. The maintenance functions are important to ensure the general well being of the group and they include - encouraging, arbitrating, peacekeeping, clarifying, summarizing and standard setting. The methodology used for undertaking the processes and procedures will affect performance. For example it is important to involve the group in standard setting; otherwise people will not accept and adopt them. I once led a pre-selected project team to a specific goal. The team had the same line manager who was not on the team. In this case it was not just my responsibility to undertake some of the maintenance functions, but also the line manager. He declined to do this, and as a result I believe performance was adversely affected. It is therefore important to recognise and ensure that the maintenance functions are dealt with by the key stakeholders, and the group leader / facilitator.

Expected patterns of behaviour

Role perception Role motivation Role capability

Role sanctions

ROLE CONFLICT Role ambiguity Role incompatibility Role overload Role underload

Role Stress

ACTUAL PATTERNS OF BEHAVIOUR

Negative Performance

Positive Performance


Good management of the given and intervening factors will enhance group performance, leading to the desired outcome. Many of these factors can be seen in the following group behaviour model:-

The core of the model is the group structure. Every group has roles, norms, and a status hierarchy that defines its structure. This core is moderated by the personality characteristics of the group’s members, the size of the group and the stage in its development. While leadership provide direction, group members interact through communication and collective decision-making. Group Development When examining the components of group performance, it is also important to consider group development. Groups, like individuals, have a growth cycle. The group will function differently at various stages of development, and performance expectations must be altered accordingly. One model identifies four distinct stages in group development.

1. Developing mutual acceptance and membership. 2. Communication and decision-making. 3. Motivation and productivity. 4. Control and organisation.

Group decision making

Leadership

CONTINGENCY VARIABLES

Communication

GROUP STRUCTURE

- Roles - Norms - Status

Member personalities

Development Stage Size

Power & Politics Conflict

Group Performance

Member satisfaction

Group cohesiveness


This model shows that a group build up to achieving a high level of performance by the third stage by understanding and communicating with each other. The fourth stage allows for a more flexible group, with members working more independently. Another model is also split into four stages:-

1. Forming - bringing the group together. 2. Storming - individual views become more open and conflict may occur. 3. Norming - group behaviour and standards are developed. 4. Performing - the group effectively performs the task set.

This model assumes the group does not perform well until the last stage in development, and does not recognise more independent working in the later stages. Neither of the two models illustrates the later stage in group development, which inevitably must occur at some time or other, and that is the destruction of a group, leading to the 'mourning' stage, where performance may fall when members leave the group. Performance will improve if groups can 'get through' the first stages as quickly as possible. Teambuilding exercises may help with this, and should not be overlooked even if the team already know each other. It is also important to recognise that teams need to be maintained, and team-building exercises should not just be saved for the beginning of a group’s task, but undertaken at various intervals, as appropriate, throughout the life of the task. This will also help overcome the 'mourning' phase of a group, when performance will naturally diminish.


CONCLUSIONS There are many differing factors that can affect the performance of a group. Some of these factors are given, while others can be changed. The given factors may include the members of the group, the objective of the group the resources available to the group and any external influences. The factors that can be altered include leadership style, motivation and some of the processes adopted. There are differences between formal and informal groups. As informal groups tend to be voluntary, communication and openness may be easier to obtain, but formal groups are more likely to be more task orientated. The group development cycle is an important consideration when examining the level of performance to expect. In my experience the most influential factors are:- • the interest level of the task to the team members, and the potential

tangible outcomes; • the members of the group and their willingness to share ideas and

communicate; • the leadership / facilitator of the group which should be a positive role

model but contribute to the task and the group. It is these factors that I recommend are initially examined when seeking performance improvements.

Sigma Table

wdp consulting

Yield % Sigma Defects Per Million Opportunities

99.9997 6.00 3.4

99.9995 5.92 5

99.9992 5.81 8

99.9990 5.76 10

99.9980 5.61 20

99.9970 5.51 30

99.9960 5.44 40

99.9930 5.31 70

99.9900 5.22 100

99.9850 5.12 150

99.9770 5.00 230

99.9670 4.91 330

99.9520 4.80 480

99.9320 4.70 680

99.9040 4.60 960

99.8650 4.50 1350

99.8140 4.40 1860

99.7450 4.30 2550

99.6540 4.20 3460

99.5340 4.10 4660

99.3790 4.00 6210

99.1810 3.90 8190

98.9300 3.80 10700

98.6100 3.70 13900

98.2200 3.60 17800

97.7300 3.50 22700

97.1300 3.40 28700

96.4100 3.30 35900

95.5400 3.20 44600
six sigma
Yield % Sigma Defects Per Million Opportunities

94.5200 3.10 54800

93.3200 3.00 66800

91.9200 2.90 80800

90.3200 2.80 96800

88.5000 2.70 115000

86.5000 2.60 135000

84.2000 2.50 158000

81.6000 2.40 184000

78.8000 2.30 212000

75.8000 2.20 242000

72.6000 2.10 274000

69.2000 2.00 308000

65.6000 1.90 344000

61.8000 1.80 382000

58.0000 1.70 420000

54.0000 1.60 460000

50.0000 1.50 500000

46.0000 1.40 540000

43.0000 1.32 570000

39.0000 1.22 610000

35.0000 1.11 650000

31.0000 1.00 690000

28.0000 0.92 720000

25.0000 0.83 750000

22.0000 0.73 780000

19.0000 0.62 810000

16.0000 0.51 840000

14.0000 0.42 860000

12.0000 0.33 880000


Symbols for Process Mapping / Flowcharting System Approach

Process step (activity)

Decision

End of process

Start of process or link

Information block (inputs or outputs)

Direction of flow


Process Approach Symbol

Name

Description

Operation

Inputs are transformed / changed

Multiple operations

Operation 20 makes 6 products at a time

Operation with inspection

Eg: grinding operation with automatic gauging

Inpection

Manual or automatic inspection

Primary flow

Movement in the process

Transportation

Planned movement where potential control is required

Process path - decision

Different products may take different routes

Incoming material

Raw materials / sub-assemblies

Planned process delay

Eg: curing time for rubber

Storage

Product is intentionally placed in a bin / rack/ pallet

Operator

Used to show where people are found in the process

10

20

6

30

40

?

Batch 001

50

60


Process Flow Approach Symbol

Name

Description

Manual information flow

Movement in the process

Electronic information flow

Information

Manufacturing process description

A process in which material or information is flowing

Push movement arrow

Inventory

Where information or material accumulates

Data box

Shows information such as cycle time, number of people, reject level

Six Sigma Part 1

Documents

central limit

champion sponsor

taguchis loss

reveal hidden

satisfy members

enables greater

preferably

hidden agendas