Slide 1: „Keeping it real” while thinking aloud Ensuring credibility of eye-tracking usability studies employing think aloud method. Wojtek Chojnacki Tobii User meeting, 2007-04-18
Slide 2: Agenda • • • • • • About Symetria Think aloud method Misconceptions about eyetracking Problems with eyetracking in think aloud studies Context of situations as an important analysis factor Concept of checkpoints Keeping it real while thinking aloud
Slide 3: About Symetria Design & planning Production & promotion Broadcasting Usability Usability evaluation User-Centered Design Eyetracking analysis (layouts, ads) Usability consulting E-marketing Marketing communication on the Web Web campaigns E-mail marketing Market analysis Media buying Media planning Ad production Media buying Monitoring Keeping it real while thinking aloud
Slide 4: Think aloud as conducted by Symetria • Trunk test: – – – What is this site? What is it for? What is your first impression? • Tasks: – – – – The more users talk, the better User can stop and explain No heavy time pressure Task fulfillment levels as usability measures Keeping it real while thinking aloud
Slide 5: Eyetracking misconceptions • Eyetracking results can be obtained by simply summarizing data for each page • Every piece of data is important • The more data gathered, the more accurate the results Keeping it real while thinking aloud
Slide 6: Eyetracking & think aloud convergence – problems Result analysis errors User specific problems Speech and gaze correlation Keeping it real while thinking aloud
Slide 7: Eyetracking & thinking aloud convergence – problems Speech and gaze correlation – Users tend to look at the elements they are describing, influencing the results – Questions asked influence reactions – this is not always intended Keeping it real while thinking aloud
Slide 8: Eyetracking & thinking aloud convergence – problems User specific problems – Overeager users – concentrate too much on talking – Distractible users – forget about the task, – Tendency to describe the page element by element Keeping it real while thinking aloud
Slide 9: Eyetracking & thinking aloud convergence – problems Result analysis errors • Summarizing data without inspecting the context • Mixing gaze data for different types of activities User describes screen actions reactions User operates page task 1 task 2 task n Keeping it real while thinking aloud
Slide 10: Eyetracking misconceptions corrected • Eyetracking is only a tool enhancing the study • Gaze data cannot be summarized without analyzing the context • It is not important to track the whole study, but specific moments • Users that concentrate on talking make results unreliable Keeping it real while thinking aloud
Slide 11: „Keeping it real” during the study • Don’t push users to talk • Ask users to describe their reactions, not their actions – The recording is enough to understand what happened – It is not important for user to know why he/she did something • In case of overeager users, suggest concentrating on the tasks • Plan ahead what questions to ask on pages which will be analyzed Keeping it real while thinking aloud
Slide 12: Planning eyetracking analysis for thinking aloud studies • Typical situations can be distinguished for each task • Online store – user looks for a 2GB mp3 player: Home page Recommended buyings Product description Product list Another Product description Shipment info Product description Shopping cart Checkout Keeping it real while thinking aloud
Slide 13: Concept of checkpoints • An instance of user interaction with a page/type of pages that happens in a specific context defined before the study • Example: „Product list page” checkpoint  Type of checkpoint: task fulfillment  Purpose of interaction: user is looking for a mp3 player  Condition: User is seeing a product list for the first time  Hypothesis: user looks at text labels rather than icons Keeping it real while thinking aloud
Slide 14: Types of checkpoints • Free interaction checkpoints: – Home page – Products page • Task fulfillment checkpoints: – Registration form – Product description • Question-response checkpoints: – What page is currently displayed? – How would you go around contacting the site’s owner? Keeping it real while thinking aloud
Slide 15: Conclusion • Plan situations in which eyetracking data will be comparable between the users • Plan questions to ask • During the study, look for situations identified before • Don’t summarize results automatically Keeping it real while thinking aloud
Slide 16: Thank you for your attention Wojtek Chojnacki wojciech.chojnacki@symetria.pl Keeping it real while thinking aloud