Highly Interactive Visual Interfaces
Measuring Interaction Extracts

These video clips are extracted from programme 2 of a series of video tapes entitled Highly Interactive Visual Interfaces. These were produced by Stephen Draper and Keith Oatley and are available by contacting University of Glasgow Media Services.

The descriptions on this page were written by Mark Dunlop.

The videos are all in wmv format (using a fairly standard codec) and about 700kbps. They can be saved locally if you have trouble playing over your network connection at 700kbps.


Questionnaires

Questionnaires are a simple evaluation tool and can be administered to large groups of users. Questionnaires can be used at many stages of a project from initial requirements capture through to final system quality evaluation. Questionnaires contain two types of question: open and closed. Closed questions ask specific questions and typically have a series of checkboxes as possible answers - these tend to lead to statistically testable quantitative results. Open questions, on the other hand tend to lead to qualitative results but are more likely to raise issues the question setters hadn't considered.

12M / 2:29

Focus Groups

Focus groups provide a form of interaction with users to find out their needs from a system/product. Often used in the early stages of design focus groups are normally used as a form of requirements capture and are based around a semi-structured discussion between one of the system team and a group of users. They can also be used in the very early stages of a project to develop some directions for the project group.

13M / 8:29

Interviews

Interviews are an easy way to find the opinions of users on a system or working environment. Interviews range from completely formal where the interviewer asks a fixed series of questions to completely informal where the system developer and the user simply have a discussion about the system. Usually the most helpful model is the semi-structured interview. Here the interviewer has a list of topics to discuss but lets the conversation flow naturally in the hope of raising the issues which are really important to the user and some issues which the interviewer had not considered.

12M / 2:28

Think Alouds

Think Alouds are the most common form of cooperative evaluation and are usually based on the evaluator monitoring a single user working through a set of tasks with the system. The user is asked to talk out loud about their thought process and this often leads to insights about how the user's model of the system differs from the developer's and about places where users find the interaction hardest. As with interviews, a large amount of skill lies in the person conducting a think aloud - they have to prompt the user the right amount to keep them talking but without interfering with the results.

23M / 4:38

Experiments

The Interaction Experiment is an evaluation approach based on getting large groups of users to carry out a fixed set of tasks with an application. Often the group is split in two and comparison is made between those using system A and system B. Normally designed to reach to answer a specific hypothesis, the experiment is typically designed to achieve statistically significant results. The results are usually quantitative and measure interaction affects such as time to complete tasks, number of problems encountered, number of "task fatal" problems or quality of final results.

18M / 3:38

Incident Diaries

When software is released into a user's environment it is hard to keep track of problems that develop. Incident Diaries are designed to help users keep track of problems that develop. The diaries are based on a small form which the user completes each time (s)he discovers a problem (either interaction difficulty or simple bug) with the system. The form typically includes contextual information to help the developers reproduce the problem.

5M / 1:01

Checklists

Checklists are a form of questionnaire. The sheet contains all the functions of a programme, perhaps organised by the menu structure (e.g. New, Open, Close, Save...). For each function the person completing the checklists states whether they use the function regularly, infrequently, know what it does, or don't know the function. These help the developers of a system model which aspects of the application users use to achieve their tasks. The designers can then either focus their design on these features or increase the visibility of ones they consider important which the users are missing.

9M / 1:48

Original Video Copyrighted University of Glasgow - clips digitally reproduced with permission


UoG              Back