Task analysis. Readings:

Transcription

1 Task analysis Readings: Dix et al 5.5; 15.1, 15.2, 15.3, 15.7 The 1984 Olympic Message System: a test of behavioral principles of system design by John Gould et al. CACM, v30 n9,

2 What is Task Analysis? The process of analyzing and improving the way people perform their jobs The process of modeling and manipulating user tasks so as to optimize the human computer interface Methods to analyze people's jobs: what people do what things they work with what they must know 2

3 A comparison of analysis techniques 3

4 Tasks and goals Goals and tasks are sometimes equated; we ll separate them. What is a task? A procedure Performed by one or more classes of users May have a supertask Broken down into subtasks, with possible loops and alternation (if-then) Involves actions (at the bottom level) A task that involves no problem solving and cannot be successfully broken down into components 4

5 Tasks and goals (cont d) What is a goal? A state that a user wants to be in Achieved by performing a task Subgoals are the goals of subtasks The user has to figure out what task to perform in order to achieve the goal May or may not be explicitly known in advance of task execution E.g. of implicit goal: writing a letter E.g. of explicit goal: adding an item to the inventory 5

6 Internal vs external tasks and actions Internal tasks Mental tasks Thoughts E.g. figuring out which command to use External tasks E.g. interacting with the computer (typing a key ) Manual tasks e.g. getting a disk from the cabinet Task analysis can focus on one or the other (preferably both) Cognitive task analysis focuses on internal tasks Q: can you find some major differences and complementary aspects of internal and external tasks? 6

7 Why task analysis? A task is a high-level activity that motivates the user to operate their computer system in the first place. Identifying the structure of a task is not trivial. It also depends on the application e.g. design of web-based applications versus design of a content-based web site 7

8 Task analysis for content-based web sites What information is needed/desired? This becomes What information is the most important? This becomes What are the main topics and which are subtopics? These become the... Within a topic, what are the related topics? These become... 8

9 Task analysis for web-based applications What are the main tasks performed by the intended users? Which tasks are parts of other tasks? Which tasks are performed linearly? Which tasks are performed repeatedly in a single session? What user assistance (online Help) is needed each step of the way? Which tasks are performed frequently and which infrequently? Would users prefer to have the tasks organized by frequency of use or by the order in which they re performed? 9

10 Techniques for task analysis Task decomposition (HTA): Looks at the way a task is split into subtasks Knowledge based techniques: look at what the users need to know about the objects and actions involved in a task, and how that knowledge is organized Entity-relationship-based analysis: object-based approach, identifies actors and objects and actors, relationships between them 10

11 How to do a task analysis (cont d) General method: First iteration from requirements specification Subsequent iterations: observe the users in the process of interaction; gather data about the interaction E.g. Thinking aloud method: ask users to talk while performing tasks imagine asking the user the question: what are you doing now? for the same action the answer may be: typing ctrl-b making a word bold emphasising a word editing a document writing a letter etc. 11

12 In-class exercise Task analysis from a thinking aloud protocol. Write down: The main goal of the user A decomposition of this goal into subgoals The tasks (actions) necessary to accomplish these goals Redesign suggestions? 12

13 13

14 How to do a task analysis? Cognitive walkthrough: Specify representative task(s) and the actions needed to complete the task(s) Walk through the action sequence, answering and documenting the analysis questions for each action Possible levels of granularity for task analysis Micro level (e.g. mouse clicks, key presses) Intermediate level (e.g. entering a string, selecting a link) Abstract level (e.g. logging in, spell checking a document) Streamlined analysis questions Will the user know what to do at this step? If the user does the right thing, will the user know that he or she did the right thing, and is making progress towards the goal? 14

15 How to do a task analysis? Scenarios: stories of typical use of your UI offer context, but may fail to capture the big picture Informal analysis, but A careful identification of a meaningful scenario allows designers to build a description of most activities to be included in the task model. Scenarios can be represented by text, diagrams, videos etc. 15

16 Scenario as a Diagram simple notation clear, easy to read single or multiple threads 16

17 4. Task Model 17

18 Verifying Requirements scenarios can be animated from task models users can see what systems would do, if built, and correct any mistakes or omissions allows systems to be 'tested' before they are designed 18

19 Hierarchical task analysis Starts with the overall goal of the interaction and decompose it into tasks Example: Self-service grocery checkout system: Overall goal: Tasks (1 level deep hierarchy) 19

20 Textual HTA description Hierarchy description in order to clean the house 1. get the vacuum cleaner out 2. get the appropriate attachment 3. clean the rooms 3.1. clean the hall 3.2. clean the living rooms 3.3. clean the bedrooms 4. empty the dust bag 5. put vacuum cleaner and attachments away... and plans Plan 0: do in that order. when the dust bag gets full do 4 Plan 3: do any of 3.1, 3.2 or 3.3 in any order depending on which rooms need cleaning N.B. only the plans denote order 20

21 Diagrammatic HTA 21

22 Refining the description Given initial HTA (textual or diagram) How to check / improve it? Some heuristics: paired actions restructure balance generalise e.g., where is `turn on gas' e.g., generate task make pot' e.g., is `pour tea' simpler than making pot? e.g., make one cup.. or more 22

23 Refined HTA for making tea 23

24 Types of plan fixed sequence then 1.2 then 1.3 optional tasks - if the pot is full 2 wait for events - when kettle boils 1.4 cycles - do while there are still empty cups time-sharing - do 1; at the same time... discretionary mixtures - do any of 3.1, 3.2 or 3.3 in any order - most plans involve several of the above 24

25 How do I conduct a hierarchical task analysis? Starting point: build a comprehensive list of tasks that make up the main task. 1. Cluster or group the tasks. For inclusion in a group, select tasks that bear close resemblance to each other (alternatives) or that are related in a linear sequence. Each task must be included in at least one of the groups, but a task may also be common to several groups. Label groups with meaningful terms. Initial clustering or grouping of tasks may be tentative. The composition of the groups may change as a result of decisions you make later on. Do not hesitate to regroup tasks when it seems appropriate. 2. Confer with a domain expert to determine the hierarchy s accuracy. 25

26 Combining plans and task analysis What needs to be done? Task for goal accomplishment What must be done first to make it possible Preconditions Tasks on which this task depends Information that must be known by user What steps are involved in doing the task? Subtasks Subtasks may be decomposed recursively 26

27 Example: self-service grocery checkout Goal: Enter groceries into register Preconditions All the groceries you want are in your cart Subtasks Enter prepackaged item Enter loose produce 27

28 Veggie vision: a system for checking out vegetables developed at IBM TJ Watson Research Center Bolle, Connel and others VeggieVision: A produce recognition system WACV Currently under field test in several supermarkets in US and Germany. 28

29 Veggie vision media coverage BBC News Online about the "Future Store" in Rheinberg, Germany, part of the METRO retail group The intelligent scale doesn't like our bunch of bananas. When we had put tomatoes on the scale, its digital camera took just a split second to recognize the produce, weigh it and print a bar-coded price tag. [..] No need to find "tomatoes" on a 50-button display. But now the scale is baffled, and offers four choices: Are we weighing bananas, chicory salad, long beans or avocados? Touching the banana logo on the screen solves the slip-up. Story from BBC NEWS: Published: 2004/05/18 BBC MMV 29

30 Dangers of task analysis from observation Duplicating a bad existing procedure in software Failing to capture good aspects in existing procedure 30