Lecture 14

1. Lecture 14 Evaluation techniques Part 2

2. Today’s Topics • We will cover today • Evaluation through user participation • Styles of evaluation • Empirical methods: experimental evaluation • Experimental Design • Analysis of data • Experimental studies on groups

3. Today’s Topics • Observational Methods • Query Techniques • Psychological Methods • Selecting appropriate evaluation technique

4. Evaluating through user Participation User participation in evaluation tends to occur in the later stages of developmentwhen there is at least a working prototype of the system in place. This may range from a simulation of the system’s interactive capabilities, without its underlying functionality.

5. Styles of evaluation • Two distinct styles of evaluation are • those performed under laboratory conditions and • those conducted in the work environment or ‘in the field’

6. Laboratory studies • Advantages: • specialist equipment available • uninterrupted environment • Disadvantages: • lack of context • difficult to observe several users cooperating • Appropriate • if system location is dangerous or impractical for constrained single user systems to allow controlled manipulation of use

7. Field Studies • Advantages: • natural environment • context retained (though observation may alter it) • longitudinal studies possible • Disadvantages: • distractions • noise • Appropriate • where context is crucial for longitudinal studies

8. Empirical methods: experimental evaluation One of the most powerful methods of evaluating a design or an aspect of a design is to use a controlled experiment. This provides empirical evidence to support a particular claim or hypothesis. It can be used to study a wide range of different issues at different levels of detail.

9. Empirical methods: experimental evaluation The evaluator chooses a hypothesis totest, which can be determined by measuring some attribute of participant behavior. A number of experimental conditions are considered which differ only in the valuesof certain controlled variables.

10. Empirical methods: experimental evaluation • Factors to be considered for experiment design mainly includes, • Participants chosen • Variables tested and manipulated • Hypothesis tested • Experimental Design

11. Participants • In evaluation experiments, participants should be chosen to match the expected user population as closely as possible. • This will involve experimental testing with the actual users but this is not always possible. • If participants are not actual users, they should be chosen to be of a similar age and level of education as the intended user group.

12. Participants Their experience with computers in general, and with systems related to that being tested, should be similar, as should their experience or knowledge of the taskdomain.

13. Participants It is no good testing an interface designed to be used by the general public on a participant set made up of computer science undergraduates: they are simply not representative of the intended user population

14. Participants • Choose enough sample size of population to test all aspects of the system. • Nielsen and Landauer suggest that usability testing with a single participantwill find about a third of the usability problems, and that there is little to be gained from testing with more than five. • Always test using large number of participants i.e.; double of five.

15. Variables • Experiments manipulate and measure variables under controlled conditions, in order to test the hypothesis. • There are two main types of variable: • those that are ‘manipulated’ or changed (known as the independent variables) and • those that are measured (the dependent variables).

16. Variables • independent variable (IV) • characteristic changed to produce different conditions • e.g. interface style, number of menu items • dependent variable (DV) • characteristics measured in the experiment • e.g. time taken, number of errors.

17. Hypothesis • A hypothesis is a prediction of the outcome of an experiment. • It is framed in terms of the independent and dependent variables, stating that a variation in the independent variable will cause a difference in the dependent variable. • The aim of the experiment is to show that this prediction is correct.

18. Hypothesis This is done by disproving the null hypothesis, which states that there is no difference in the dependent variable between the levels of the independent variable. If a result is significant at the given level of certainty, that the differences measured would not have occurred by chance (that is, that the null hypothesis is incorrect).

19. Hypothesis • prediction of outcome • framed in terms of IV and DV • For example “error rate will increase as font size decreases” • null hypothesis: • states no difference between conditions • aim is to disprove this • e.g. null hyp. = “no change with font size”

20. Experimental Design • There are two main methods: • between-subjects and • within-subjects.

21. Experimental design • within groups design • each subject performs experiment under each condition. • transfer of learning possible • less costly and less likely to suffer from user variation. • between groups design • each subject performs under only one condition • no transfer of learning • more users required • variation can bias results.

22. Between group design • The advantage of a between-subjects design is that any learning effect resulting from the user performing in one condition and then the other is controlled: • each user performs under only one condition. • The disadvantages are that a greater number of participants are required, and that significant variation between the groups can negate any results.

23. Within-group Design Each user performs under each different condition. This design can suffer from transfer of learning effects, but this can be lessened if the order in which the conditions are tackled is varied between users.

24. Analysis of data • Before you start to do any statistics: • look at data • save original data • Choice of statistical technique depends on • type of data • information required • Type of data • discrete - finite number of values • continuous - any value

25. Analysis - types of test • parametric • assume normal distribution • robust • powerful • non-parametric • do not assume normal distribution • less powerful • more reliable • contingency table • classify data by discrete attributes • count number of data items in each group

26. Analysis of data (cont.) • What information is required? • is there a difference? • how big is the difference? • how accurate is the estimate? • Parametric and non-parametric tests mainly address first of these

27. Experimental studies on groups • Experimental studies of groups and of groupware are more difficult than the corresponding single-user experiments already considered.

28. Experimental studies on groups Problems with: • subject groups • choice of task • data gathering • analysis

29. Experimental studies on groups To organize, say, 10 experiments of a single-user system requires 10 participants. For an experiment involving groups of three, we will, of course, need 30 participants for the same number of experiments.

30. Experimental Studies on groups In addition, experiments in group working are often longer than the single-user equivalents as we must allow time for the group to ‘settle down’ and some rapport to develop. This all means more disruption for participants and possibly more expense payments.

31. Subject groups larger number of subjects more expensive longer time to `settle down’ … even more variation! difficult to timetable so … often only three or four groups

32. The Task • Choosing a suitable task is also difficult. • We may want to test a variety of different task types: creative, structured, information passing, and so on.

33. The task must encourage cooperation perhaps involve multiple channels options: • creative task e.g. ‘write a short report on …’ • decision games e.g. desert survival task • control task e.g. ARKola bottling plant

34. Data gathering several video cameras + direct logging of application problems: • synchronisation • sheer volume! one solution: • record from each perspective

35. Analysis In true experimental tradition, we would like to see statistical differencesbetween experimental conditions. We saw earlier that individual differences made this difficult in single-user experiments. If anything, group variation is more extreme. Given randomly mixed groups, one group will act in a democratic fashion. Thus dominating others.

36. Analysis Vast variation between groups solutions: • within groups experiments • micro-analysis (e.g., gaps in speech) • anecdotal and qualitative analysis look at interactions between group and media controlled experiments may `waste' resources!

37. Field studies Experiments dominated by group formation Field studies more realistic: distributed cognition work studied in context real action is situated action physical and social environment both crucial Contrast: psychology – controlled experiment sociology and anthropology – open study and rich data

38. Observational Methods • A popular way to gather information about actual use of a system is to observe users interacting with it. • Techniques include • Think Aloud • Cooperative evaluation • Protocol analysis • Automated analysis • Post-task walkthroughs

39. Think Aloud • user observed performing task • user asked to describe what he is doing and why, what he thinks is happening etc. • Advantages • simplicity - requires little expertise • can provide useful insight • can show how system is actually use • Disadvantages • subjective • selective • act of describing may alter task performance

40. Cooperative evaluation • variation on think aloud • user collaborates in evaluation • both user and evaluator can ask each other questions throughout • Additional advantages • less constrained and easier to use • user is encouraged to criticize system • clarification possible

41. Protocol analysis • paper and pencil – cheap, limited to writing speed • audio – good for think aloud, difficult to match with other protocols • video – accurate and realistic, needs special equipment, obtrusive • computer logging – automatic and unobtrusive, large amounts of data difficult to analyze • user notebooks – coarse and subjective, useful insights, good for longitudinal studies • Mixed use in practice. • audio/video transcription difficult and requires skill. • Some automatic support tools available

42. Automated Analysis – EVA • Workplace project • Post task walkthrough • user reacts on action after the event • used to fill in intention • Advantages • analyst has time to focus on relevant incidents • avoid excessive interruption of task • Disadvantages • lack of freshness • may be post-hoc interpretation of events

43. post-task walkthroughs • transcript played back to participant for comment • immediately  fresh in mind • delayed  evaluator has time to identify questions • useful to identify reasons for actions and alternatives considered • necessary in cases where think aloud is not possible

44. Query Techniques • Relies on asking the user about the interface directly. • Query techniques can be useful in eliciting detail of the user’s view of a system. • Types of Query Techniques are • Interviews • Questionnaires

45. Interviews • analyst questions user on one-to -one basisusually based on prepared questions • informal, subjective and relatively cheap • Advantages • can be varied to suit context • issues can be explored more fully • can elicit user views and identify unanticipated problems • Disadvantages • very subjective • time consuming

46. Questionnaires • Set of fixed questions given to users • Advantages • quick and reaches large user group • can be analyzed more rigorously • Disadvantages • less flexible • less probing

47. Questionnaires (ctd) • Need careful design • what information is required? • how are answers to be analyzed? • Styles of question • general • open-ended • scalar • multi-choice • ranked

48. Psychological Methods • One of the problems with most evaluation techniques is that we are reliant on observation and the users telling us what they are doing and how they are feeling. • What if we were able to measure these things directly?

49. Psychological Methods Interest has grown recently in the useof what is sometimes called objective usability testing, ways of monitoring physiological aspects of computer use. Potentially this will allow us not only to see more clearly exactly what users do when they interact with computers, but also to measure how they feel.

50. Physiological methods • Areas receiving more attention are, • Eye tracking • Physiological measurement