Chapter 2, part 2 Human Aspects: The Capacity of Short-term Memory......must be taken into account, i.e. important to limit the number of items a user.

Chapter 2, part 2Human Aspects:

The Capacity of Short-term Memory...

...must be taken into account, i.e. important to limit the number of items a user has to deal with at any particular moment.

...make sure that the user does not have to remember or write down items from one screen to the next.

Capacity of Short-term Memory continued

Large tasks should be structured into smaller tasks, so that the user does not have to mentally transfer information from one sequence of actions to another.

Design user interface in a way that user can make use of recognition rather than recall.

It is often useful to apply memory aids and visual cues (use of pictures, metaphors, etc.)

Feedback

When users perform actions, they typically want to know what happened, e.g. “Installation successful”.

3 types of feedback:

1. Future Feedback (“what will happen if I carry out command”)

2. Present Feedback (continuing feedback during application)

3. Past Feedback (“Installation successful”)

Feedback continuedWhen the system processes a long job, the user should receive feedback about the fact that the system is working. This is particularly important if the user cannot interact with the system whilst it is in progress.

System should provide user-centered wording, i.e. simple terms in non-technical language. Exceptions apply to expert users where an error code that can be looked up is sometimes preferred.

Methods to get theuser’s attention

Many ways, e.g. blinking, use of colour, flashing animations etc.

Because there are so many ways to enhance attention, one must be careful not to misuse them, e.g. it is easy to create a scenario where the user gets confused.

Methods to get theuser’s attention continued

Some common guidelines:

1. Text

2. Colour

3. Blinking

4. Audio

Methods to get theuser’s attention continued

1 Text:

Use no more than 2 levels of intensity on a single screen.

Make use of underlining, bold or italic text, inverse video.

Upper and lower case letters as in normal sentences, because if using only upper-case letters, this slows down the reading speed by over 10 percent (Hix and Hartson, 1993)

Methods to get theuser’s attention continued

2 Colour:

Consider that there might be people with visual difficulties, e.g. colour-blind people.

Consider that different people might have different screens (in the past it was often considered a good idea to design for monochrome screens first, but now most people have colour screens)

The layout should make sense independently of the use of colour.

Methods to get theuser’s attention continued

2 Colour (continued):

generally, it is considered best not to have more than 4 different colours on a single screen.

No more than 7 different colours should be applied in one application.

Will people print screen display? (if yes, consider that not so many people have colour printers)

Methods to get theuser’s attention continued

2 Colour (continued):

Use colour conservatively and do not switch colours between screens if you repeat showing the same display on some screen locations.

Depending on application, it is often a good idea to let the user customise her/his preferred colours to the application.

Methods to get theuser’s attention continued

3 Blinking:

It is generally recommended to use blinking only for the most important items.

4 Audio:

Audio information can enhance very important events, but consider that not every user may have their speakers turned on. The information should also be highlighted in other ways, as deaf users would otherwise have a serious disadvantage.

Knowledge andMental Models

1. Knowledge

• Analogical representations

• Propositional representations

• Distributed representations

2. Mental Models

“the models people have of themselves, others, the environment, and the things with which they interact.”

1. Knowledge

Questions to be analysed:

What do users know about the systems they use?

How do they reason about the system?

Why? ...because then it might be possible to predict what users find simple and what users find difficult (and to design interfaces appropriately)

1. Knowledge continued

Analogical representations = picture-like, e.g. a lock and a key could stand for encryption.

Propositional representations = abstract and language like, e.g. the laptop is on the desk (rule-based, as in classical artificial intelligence).

Distributed representations = interconnected network of nodes where knowledge is stored in the connections between nodes (e.g. a neural network)

1. Knowledge continuedKnowledge is typically very organised, but it is still unclear how knowledge is exactly organised in long-term memory (apart from the episodic/semantic distinction, there might be a range of other possibilities to represent knowledge in long-term memory).

It is generally accepted that knowledge is represented in some sort of network, which is analogous to the way information is stored in the brain.

1. Knowledge continuedFor example: the brain is a network of approximately 1011 nerve cells and each nerve cell receives on average 1000 connections from other nerve cells.

However, it is still unclear what content the bundles of nerve cells and connections represent, so cognitive scientists have mainly focussed on how the mind represents knowledge.

1. Knowledge continuedThe idea is that highly organised information is stored in a way that things that belong together are stored near each other rather than widely distributed.

Examples: the names of countries and their capitals, the names of football clubs and their stadiums, the concept “surf” and Hawaii, but also the concept “surf” and the internet etc.

1. Knowledge continued

Theory 1: Semantic Network Theory

It is assumed that information is often stored in the form of semantic networks.

A semantic network consists of nodes and links between these nodes.

Semantic networks are meant to structure information (often hierarchically).

1. Knowledge continued

Theory 1: Semantic Network Theory continued

for example: nodes correspond to objects (Java, C++, Visual Basic etc.) or classes of objects (Programming Languages) and links are meant to represent the relationship between these objects (C++ is a programming language).

1. Knowledge continued

Theory 2: Schemata and Scipts

A schema is a network of general knowledge based on previous experience (this does not have to be based on personal experience, i.e. this schema can also be acquired through TV, stories from other people etc.)

Classical example: how a typical hospital looks like.

1. Knowledge continued

Theory 2: Schemata and Scipts continued

Schemata have the function to facilitate people’s understanding of commonplace events, e.g. if you hear that a friend has been admitted to hospital, you can deduce that your friend may need medical treatment.

1. Knowledge continued

Theory 2: Schemata and Scipts continued

A Script is a special case of a Schema. A script describes a sequence of events that occur in a particular setting, e.g. when you go to a restaurant, you will typically be offered a table, you receive the menu, order drinks and meals, drinks arrive, meals arrive a bit later, you ask for the bill, pay, leave a tip and leave the restaurant. This sequence of action is a script.

1. Knowledge continued

Theory 2: Schemata and Scipts continued

How can we apply this to User Interface Design?

People often develop scripts when they repeatedly carry out sequences of actions, e.g. when dealing with computers, mobile phones, etc.

1. Knowledge continued

Theory 2: Schemata and Scipts continued

Because schemata and scripts are often viewed as guiding user’s behaviour, user interface design tries to figure out what schemata and scripts people have already developed.

New interfaces based on existing schemata may help the user to adapt better, because these interfaces will seem more meaningful.

1. Knowledge continued

Theory 2: Schemata and Scipts continued

Criticism: Schema-based theories are often considered too inflexible, because they cannot explain how people make inferences in complex situations, predict future states accurately, etc.

People however often succeed in these situations.

2. Mental models

Mental models have been developed to account for human flexibility in these situations.

They are based on Schema-Theory, but they are constructed dynamically, i.e. depending on the moment and situation, where they activate already existing schemata.

2. Mental models continued

Mental models include Meta-knowledge, i.e. knowledge that people have about themselves, as well as other knowledge such as knowledge about other people, the environment, and things they interact with, e.g. knowledge about computers and application programs (“this old web Browser does not support frames”).

2. Mental models continued

at least 4 ways people can construct mental models:

1. Own experience

2. Training

3. Instruction

4. Models (not mentioned in study guide, but observational learning by watching others deal with specific situations has a large impact).

2. Mental models continued

According to Johnson-Laird, mental models can either be based on analogical representations (e.g. learning to deal with a new mobile phone by making analogies to other mobile phones one has already used).

Mental models are often constructed when making inferences, predictions etc. It is often said that a mental “simulation” run is carried out before one performs an action in reality.

2. Mental models continued

So mental models are not like images, more like imagery videos.

2 types of mental models are often referred to:

1. Structural model

2. Functional model

2. Mental models continued

1. Structural model

User knows the structure of the device or how the system works and has stored this in memory, e.g. having thorough knowledge about both software and hardware of mobile phones, e.g. what chips they typically have, what assembly language is used to program the processors etc.

2. Mental models continued

1. Structural model continued

Structural knowledge is typically most useful when a device breaks down and one has to find out what the reason for the break-down is.

They can answer unexpected questions and make predictions.

2. Mental models continued

1. Structural model continued

Structural models are context free.

This is an advantage, as this makes it possible to enhance knowledge and to integrate other knowledge.

However, structural knowledge is often more difficult to gain.

2. Mental models continued

2. Functional model

The user has stored procedural knowledge, i.e. the user knows how to use the device (In a mobile phone, this could be sending text messages, taking pictures and sending them, playing games, listening to MP3s and finally: making a phone call).

2. Mental models continued

2. Functional model continued

Functional models often develop from past knowledge of a similar device, i.e. they develop from already existing functional models.

Functional models are centered around specific tasks, i.e. context dependent.

2. Mental models continued

2. Functional model continued

The advantage of the context dependence in functional models is that this makes the model easier to use (by making analogies to already existing, similar models).

2. Mental models continued

Utility of Mental Models

It is often assumed that people have pretty vague mental models.

Preece et al. (1994) argue, however, that the user’s mental models are nevertheless important for user interface design.

2. Mental models continued

Utility of Mental Models

Why?

...because knowing mental models of people can be very productive when constructing an appropriate model of the system.

Many HCI experts suggest that a user interface has to enable the user to form a mental model of the system.

2. Mental models continued

Utility of Mental Models

The mental model will ultimately have the effect that it governs how well the user understands the system and how well the user will interact with it.

If the user is able to form a consistent mental models based on the task s/he performs, the user will be more likely to succeed on similar tasks in the future.

2. Mental models continued

Utility of Mental Models

Graphical user interfaces and visual cues that are incorporated into them are typically considered very helpful by most users to formulate a good mental model.