AutoCWW Research Program

Institute of Cognitive Science (ICS), University of Colorado at Boulder

Marilyn Hughes Blackmon, Ph.D.
Research Associate, Institute of Cognitive Science. University of Colorado, Boulder, CO 80309-0344, phone 303-859-5060, fax 303-492-7177

AutoCWW Research Program Home > CWW Tutorial June, 2004

Cognitive Walkthrough for the Web (CWW)
Tutorial (June, 2004)

The Cognitive Walkthrough for the Web (CWW) transforms the original Cognitive Walkthrough (CW)

  • CWW bypasses the
  • CWW uses a slight adaptation of the four questions of CW.
  • Good fit with website design processes. The web developer applies the CWW to one web page at a time, the page under construction. The developer starts with the home page and then moves to lower and lower levels of the hierarchy (for example, home page, subsite main pages, intermediate level pages, terminal node pages)
  • Capacity to model representative goals for diverse user groups, including English-speaking users with various levels of reading comprehension and background knowledge. CWW has the potential to add new LSA semantic spaces, and CWW will consequently gain the capacity to simulate users from different cultures/subcultures and users who speak various languages.
  • Takes as input long (100-word to 200-word) narrative goals that richly describe users' understanding of their tasks and their underlying motivation
  • Built on a computer simulation model of goal-driven exploration of websites— the CoLiDeS model. According to CoLiDeS, the user first parses the web page into regions and focuses on the region that is semantically most similar to the user goal. Then the user comprehends the actions available in the focused-on region and selects the one that is semantically most similar to the user goal. As a result, the heading labels for regions play a crucial role in directing/misdirecting user attention, and an important contribution of CWW is its ability to accurately identify and repair problematic headings.
Subgoal 1. Compile a representative set of realistic user goals for the website 
  Subgoal 2. Create initial design for web page under construction: Example of original home page for

Subgoal 3. Overview of how to do 3 types of LSA analyses

  1. LSA term vector length, the procedure needed to identify unfamiliar headings/links
  2. LSA Matrix Comparison, the procedure needed to identify confusable headings/links
  3. LSA One-to-Many Comparison, the procedure needed to identify goal-specific competing headings/links

Subgoal 4. Inspect headings/links for problematic heading/link texts liable to interfere with all user goals

  1. Identify and repair unfamiliar heading/link labels
  2. Identify and repair confusable heading/link labels

Subgoal 5. Identify and repair goal-specific competing headings/links, problems that affect only some goals.


Subgoal 6. Further evaluate repaired web page

  1. If not already completed for home page and/or subsite main page, perform CWW of navigation and layout common to all pages in the site/subsite.
  2. Check repaired home page: Example of repaired home page for
  Subgoal 7. Understand the CoLiDeS model underlying the CWW (external link)
This material is based upon work supported by the National Science Foundation under Grant No. 0137759. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.