Advances in heuristic usability evaluation method
Heuristic evaluation method is one of the most used usability evaluation method in both industry and academia (Rosenbaum, 2000). Based on the characteristics of heuristic evaluation method and E-Commerce websites, a conceptual model with three parts which address the evaluator's cognitive style, heuristic evaluation process, and the heuristics' impacts respectively is proposed. The first part examines effect of evaluator's cognitive style on the evaluation results. The second part incorporates the Taguchi quality control method into the heuristic evaluation process to find the optimal combination of factors including task type, heuristic set, and evaluation mode which is least sensitive to variations in evaluator's cognitive style. The third part relates website usability heuristics with user's purchase intention on E-Commerce websites. Three experiments and a survey were conducted to test the proposed four hypothesis associated with testing the validity of the proposed conceptual model. The results of the three experiments and survey study indicated the following: (1) The optimized evaluation process which combines the Taguchi method with the traditional heuristic evolution is to have evaluators carry out evaluation in pairs with the help of domain specific heuristic set. (2) Using the developed optimized evaluation process significantly improves the evaluation effectiveness by over 17.6% and reduced the variance among the evaluation effectiveness by 3.5 times in relation to using the traditional heuristic evaluation method. (3) Field independent (FI) evaluators find problem sets with higher thoroughness and validity than field dependent (FD) evaluators. (4) FI evaluators use more analytical approaches than FD evaluators during heuristic evaluation. (5) Using the newly developed E-Commerce heuristic set results in finding larger number of real usability problems than using Nielsen's (1994d) heuristic set. (6) Users' purchase intention on E-Commerce website is affected, in descending importance by the following five factors: prevention from error, information access quality, shopping support, ease of comprehension, and hedonic quality. ^
Gavriel Salvendy, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our