Summary: Lavery et al., Comparison of Evaluation Methods Using Structured Usability Problem Reports

Because heuristics inspection and user testing are two complementary evaluation methods, people usually conduct both of them to get a system thoroughly evaluated. However, there is a problem: the nature of the usability problems found through analytic and empirical methods are not exactly the same, and the way by which evaluators document these two types of usability problems are also different; therefore, people find it difficult to consolidate the evaluation results from heuristics inspection and user testing. The aim of this paper is to propose a “usability problem report format,” which provides clearly defined components for describing a usability problem. After all the problems are documented using this format, the evaluators could match the usability problems from the above two types of evaluation methods by analyzing each component of the usability report. However, the rules for matching usability problems with different natures are yet to be developed in the future studies. Here are the basic components of the proposed format:

· Context: “user context,” “interaction context,” and the “work context;”

· Cause: the “design fault” or a “knowledge requirement on the user” that lead to the breakdown;

· Breakdown: the point at which “user does not take an appropriate step in the interaction;”

· Outcome: the results of a breakdown, including behaviors, performance, or preference;

· Solution: proposed design change for addressing the problem.

Summary: Hornbæk & Frøkjær, Making Use of Business Goals in Usability Evaluation: An Experiment with Novice Evaluators

The paper conducted an experiment aiming at examining the effect for using business goals in usability evaluation. The results indicate that, if the evaluators are asked to consider business goals in usability evaluation, less usability problems are found than people who are not required to do so. However, these usability problems are deemed as of “higher business relevance and overall utility to the company that commissioned the evaluation.”

This experiment is a between group experimental design. Two groups of evaluators are asked to conduct “think-aloud tests” to a website. Before the test, both of the groups are given a background material to read regarding the business goals of the company who owns the website. The only difference is that one group is required to consider the relevance of the usability problems they found to the business goals, but the other is not. After the test, both of the groups’ evaluation reports are given back to the company’s representatives for assessment. The assessment is based on a four-dimension criterion: clarity, interaction, business, and overall usefulness.

The implication of this experiment is that, besides usability knowledge, other kinds of knowledge, such as business goals, could play an important role in usability evaluation. They could help evaluators eliminate the problems that are not relevant to the business goals and focus on the problems which are more useful for the company.

Summary: Jared Spool et al., Designing for the Scent of Information

The main theme of this paper is based on the analogy that users’ information searching behavior on a website is analogous to “animals hunting their prey”: After users arrive at a website, if they could smell the scent of the information they need, they would stay on the website and approach the information step by step guided by the scent. However, if the scent disappears in the middle, they need to “backtrack” to re-pick up the scent. Here, the term “scent” refers to “the placement of critical information, the length of links and pages, and a few other perceptible elements.” Whether users are satisfied with the scent is reflected by their confidence level.

The most important element on a website that could create scent is the “trigger words.” The first thing that users do when they open a web page is scanning for the words that represent the information they need. To create good scent, make these trigger words prominent both on the links that guide users to the needed information and the web page that contains the information.

We should be careful the things that might block scent:

· Iceberg Syndrome: Irrelevant information that contains the trigger word might prevent users from scrolling down the webpage to see the information that they really need;

· Camouflaged Links: The link doesn’t look like a link;

· Banner Blindness: Links that appear in the top 60 pixels of the home page would be ignored, because top 60 pixels is the place where banner ads usually appear and people are trained during years of using websites that banner ads are of little use;

· Links that Lie: The inconsistency between a link an its content would impair users’ confidence;

· Missing Words: Users open a link and find that the new page doesn’t contain the trigger words;

· Cute Link Names: Using cute but obscure link names would weaken the scent;

· Misplaced Links: “Designers see entire section, while users see only the pages they visit.” Do not place a link that would confuse people.

The other guidelines for creating scent are: Longer links are clearer in meaning than the short ones, but do not make them too long, which may prevent users in spotting the trigger word; navigation and content graphics help creating scent, but decorative graphics do not; to create scent for a website, “start with the most important content and determine the most effective trigger words.”