It is one thing to run usability tests…..and another to provide useful recommendations. How many times do you see a usability report that seems like it is pressed from the same mould over and over again? You see recommendations that look like not much thought has gone into it. Or you see recommendations that are based upon one or a small number of comments from users, without much analysis or thought about the overall effect or problem that is occurring.
As a usability professional, it is important that we provide recommendations that are meaningful, and are not based purely upon inferences. Take the name fields on forms (i.e. do we put the first name before the surname, or vice versa?). A user may say, “I always put my first name first”. It’s easy to then infer that the first name should be before the surname. But is this true in all cases? How many users are having problems with this ordering? How about in Asian countries where the surname usually comes first?
Jared M. Spool takes a look at this issue of strong recommendations in The Road to Recommendation that looks at jumping to conclusions, validating our inferences, making strong recommendations, and (my personal favourite) that there are alternate means of analysis other than usability studies, such as analysis of web site logs. It is well worth a read!