Think aloud protocol

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Think-aloud protocol (or thinking aloud; also talk-aloud protocol) is a used to gather data in usability testing in product design and development, in psychology and a range of social sciences (e.g., reading, writing, translation research, decision making, and process tracing). The think-aloud method was introduced in the usability field by Clayton Lewis [1] while he was at IBM, and is explained in Task-Centered User Interface Design: A Practical Introduction by C. Lewis and J. Rieman.[2] The method was developed based on the techniques of protocol analysis by Ericsson and Simon.[3][4][5] However, there are some significant differences between the way Ericsson and Simon propose that protocols be conducted and how they are actually conducted by usability practitioners, as noted by Boren and Ramey.[6] These differences arise from the specific needs and context of usability testing; practitioners should be aware of these differences and adjust their method to meet their needs while still collecting valid data. For example, they may need to prompt for additional information more often than Ericsson and Simon would allow, but should take care not to influence what participants say and do.

Think-aloud protocols involve participants thinking aloud as they are performing a set of specified tasks. Participants are asked to say whatever comes into their mind as they complete the task. This might include what they are looking at, thinking, doing, and feeling. This gives observers insight into the participant's cognitive processes (rather than only their final product), to make thought processes as explicit as possible during task performance. In a formal research protocol, all verbalizations are transcribed and then analyzed. In a usability testing context, observers are asked to take notes of what participants say and do, without attempting to interpret their actions and words, and especially noting places where they encounter difficulty. Test sessions are often audio- and video-recorded so that developers can go back and refer to what participants did and how they reacted.

A related but slightly different data-gathering method is the talk-aloud protocol. This involves participants only describing their actions but not other thoughts. This method is thought to be more objective in that participants merely report how they go about completing a task rather than interpreting or justifying their actions (see the standard works by Ericsson & Simon).

As Kuusela and Paul [7] state, the think-aloud protocol can be distinguished into two different types of experimental procedures. The first is the concurrent think-aloud protocol, collected during the task. The second is the retrospective think-aloud protocol, gathered after the task as the participant walks back through the steps they took previously, often prompted by a video recording of themselves. There are benefits and drawbacks to each approach, but in general a concurrent protocol may be more complete, while a retrospective protocol has less chance to interfere with task performance.

See also[edit]


  1. ^ Lewis, C. H. (1982). Using the "Thinking Aloud" Method In Cognitive Interface Design (Technical report). IBM. RC-9265. 
  2. ^ Task-Centered User Interface Design: A Practical Introduction, by Clayton Lewis and John Rieman.
  3. ^ Ericsson, K., & Simon, H. (May 1980). "Verbal reports as data". Psychological Review 87 (3): 215–251. doi:10.1037/0033-295X.87.3.215. 
  4. ^ Ericsson, K., & Simon, H. (1987). "Verbal reports on thinking". In C. Faerch & G. Kasper (eds.). Introspection in Second Language Research. Clevedon, Avon: Multilingual Matters. pp. 24–54. 
  5. ^ Ericsson, K., & Simon, H. (1993). Protocol Analysis: Verbal Reports as Data (2nd ed.). Boston: MIT Press. ISBN 0-262-05029-3. 
  6. ^ Thinking Aloud: Reconciling Theory and Practice, by Ted Boren and Judith Ramey.
  7. ^ Kuusela, H., & Paul, P. (2000). "A comparison of concurrent and retrospective verbal protocol analysis". American Journal of Psychology (University of Illinois Press) 113 (3): 387–404. doi:10.2307/1423365. JSTOR 1423365. PMID 10997234.