Presentation2003

TopicsTopicsData – SpreadsheetManipulating data

- Pivot tables - Visualisation(static and dynamic)

Comparing spreadsheet & databaseSupporting a hypothesis using dataSpreadsheet & database – which is

more appropriate for supporting hypothesis?

Part of Excel SpreadsheetPart of Excel SpreadsheetData regarding 2 weeks of group

members activities

Week DateUQ student

No. First Name Last Name Category Activity Time Duration

1 14/09/2008 41201396 Andrew McMillen Recreation Drinking 12:00:00 AM 6

1 14/09/2008 41201396 Andrew McMillen Recreation Drinking 3:00:00 PM 3

1 14/09/2008 41201396 Andrew McMillen Socialising Video Games 7:00:00 PM 2

1 14/09/2008 41201396 Andrew McMillen Education Internet 10:00:00 AM 4

1 14/09/2008 41201396 Andrew McMillen Travel Train 2:00:00 PM 1

1 14/09/2008 41201396 Andrew McMillen Rest Sleeping 11:00:00 PM 1

1 14/09/2008 41613298 Harry Kim Rest Sleeping 12:00:00 AM 91 14/09/2008 41613298 Harry Kim Religion Church 10:30:00 AM 2

1 14/09/2008 41613298 Harry Kim Recreation Videos 4:00:00 PM 3

1 14/09/2008 41613298 Harry Kim Education Study 8:00:00 PM 3

1 14/09/2008 41613298 Harry Kim Recreation Reading 12:00:00 AM 0.5

1 14/09/2008 41788468 Samuel Ninness Rest Sleeping 12:00:00 AM 9

1 14/09/2008 41788468 Samuel Ninness Recreation Videos 10:00:00 AM 8

1 14/09/2008 41788468 Samuel Ninness Recreation Videos 8:00:00 PM 2

1 14/09/2008 41788468 Samuel Ninness Rest Sleeping 10:00:00 PM 2

Pivot TablesPivot Tables

Weekly Duration

Week 1 Week 2

Andrew Harry Samuel Andrew Harry Samuel

Education 18 48 24.5 11.5 41 28.5

Exercise 2 1 1 3.5

Housework 3 2.5 2 2.5

Recreation 21 14 28.5 16 12.5 21

Religion 2 2

Rest 54 55.5 70 52.2 57.5 69

Socialising 6 10.5 4.5 16 9.65 4.5

Travel 6 10 8.3 5 9 8

Work 36 8.5 38 16

Manipulating data to for specific goalsE.g. Comparing Weekly Totals of

Category per personData much more useful

In PercentageIn Percentage

Week 1 Week 2

Weekly Duration Weekly Duration

Andrew Harry Samuel Andrew Harry Samuel

Education 20% 53% 27% Education 14% 51% 35%

Exercise 67% 33% 0% Exercise 22% 78% 0%

Housework 55% 0% 45% Housework 44% 0% 56%

Recreation 33% 22% 45% Recreation 32% 25% 42%

Religion 0% 100% 0% Religion 0% 100% 0%

Rest 30% 31% 39% Rest 29% 32% 39%

Socialising 29% 50% 21% Socialising 53% 32% 15%

Travel 25% 41% 34% Travel 23% 41% 36%

Work 81% 0% 19% Work 70% 0% 30%

Visualization Visualization Another way of manipulating dataLike pivot tables, allows data to be

represented in a useful wayDisplays data graphically e.g. Graphs2 types: Static and Dynamic

visualization

Static representationStatic representationWeekly Category total per person

Static continued…Static continued…

But what if too many graphs are needed?

Dynamic RepresentationDynamic Representation“Dynamic” – non-static

visualisationE.g. Daily Total Category per

person - Over 2 weeks, 14 graphs are

needed!So static visualisation is

inappropriate in certain cases

Dynamic Continued…Dynamic Continued…Hence we resort to dynamic

representationHere is one about Daily Total

Category per Person produced using Google docs

StructureStructure

SpreadsheetsTable made up of individual cells

DatabasesCollection of tables storing related dataEach table contains columns/fieldsAlso Queries, reports, forms

Database Structure Database Structure ExampleExample

Activity Log

Student Info

UQ student

No.First

NameLast

Name41201396 Andrew McMillen41613298 Harry Kim41788468 Samuel Ninness

Date UQ student

No. Activity Time Duratio

n 14/09/2008 41201396 Train 2:00:00 PM 1

14/09/2008 41201396 Sleeping11:00:00

PM 114/09/200

8 41613298 Sleeping12:00:00

AM 9

14/09/2008 41613298 Church10:30:00

AM 2

Category Activity Travel TrainRest Sleeping

Religion ChurchRecreation Videos

Activity Types

Relationships between similar data in tables

Field

Additional ConstraintsAdditional ConstraintsSpreadsheetsEnforces data format constraintsNumerical, currency, date/time, text formats

DatabasesSame formats as spreadsheetsAlso minimum and maximum field size,

required field, default values assigned and validation rules

Spreadsheet constraints Spreadsheet constraints exampleexample

Numbers Text Date Currency

1.00 Where 19/09/2009 $13.00

5423.00 Is 19/09/2009 $56.00

234.00 My 19/09/2009 $47.00

52.00 Cow 19/09/2009 $85.00

76.00 ? 19/09/2009 $99.00

Each Column is formatted to display

the specified information only

Data ManipulationData ManipulationSpreadsheetsStatic and dynamic visualizationsPivot TablesExtensive mathematical calculations

DatabasesFew graphical visualizationsQueries, reportsLimited calculation functions in reports

Reports exampleReports example

This report based the this query

Calculations ExampleCalculations Example

Num A Num B Total

20 2 22

10 6 16

5 5 10

7 3 10

9 8 17

Num A + Num B = Total

LimitationsLimitationsSpreadsheetsData in large spreadsheet systems

redundant and unreliableMultiple copy complicationsOne user at a time on centrally stored

spreadsheets

DatabasesEliminates spreadsheet problemsChanging user requirements necessitates a

new database

DevelopmentDevelopmentSpreadsheetsSimple to create.Requires considerable user maintenanceMultiple spreadsheets -> inconsistencies

occur

Databasesconsiderable time and energy to create. Little maintenance neededneed to be replaced when they become

outdated.

Supporting Hypothesis using Supporting Hypothesis using datadataHypothesis: It is argued that

students who have no more than 10 hours of paid work a week are more effective than students who do not work or work longer hours

Using our group dataUsing our group data

Excessive work correlates with lower time into education

However, one non-working person put a large amount of time into education

Working a lot over 10 hr correlated with low GPA

However, non-working person achieved high results

SummarySummaryBoth cases show mixed resultsHence data does not (fully)

support the hypothesis Non-working person had higher

Education hours and grade then someone close to 10 hr of working

Working vs Non-working students continued (data from another research)Dr Kerri-Lee Krause, Sept 200521st century undergraduate

student engaged, inert, or otherwise occupied?

Engagement = time, energy and resources devoted to uni activities

Working vs Non-working studentsHypothesis:No more than 10 hours of paid work a week = more effective than students who do not work, or work longer hours?

Working vs Non-working studentsKrause et al 2004: The First Year

Experience in Australian Universities: Findings from a decade of national studies

‘Effective’ student = more ‘engaged’ student

This = more time, energy and resources devoted to uni activities. (in theory)

Working vs Non-working studentsPaid students study less (10.5 hours)

than non-employed (11.8 hours per week)

Average uni contact hours per week for full-time first year students has declined to 16 per week in 2004 - was 17.6 in 1995

Paid part-time workers = fewer weekly contact hours (15.5) compared to their non-employed peers (16.8 hours per week)

Working vs Non-working studentsHypothesis unsupported by our

data

Working vs Non-working students

Spreadsheet vs DatabaseSpreadsheet good for the

purposes of this small-scale project

Easily create visualisations using graphs and pivot tables

If project was larger, recommend DBMS for stability, versatility and relational capabilities

Project LimitationsProject LimitationsCategorisation of activities sometimes

confusingUni grades vs work experience?Vague task descriptionsSmall sample size – not indicative of

habits across entire semesterHence unrepresentative of the whole

population of first year studentsOur group was unable to find statistics

regarding work and education

In Conclusion:In Conclusion:Comparison between parts of the

assessment were enjoyableOnline collaboration is highly

recommendedThis opens the door to further

research – are you interested?Cheers