2.1 Experiment Introduction/Hypotheses

The experiment explores three user controlimplementations and their benefits/ drawbacks with respect to varying degrees of browsing a number of images at atime. Subjects were given a unique description of the image to search for. The time to locate the image, aswell as the number of errors made and the degree of satisfaction were measured.

Independent Variables
1. User Control Type
Treatments:
1. Button-based
'Next' and 'previous' buttons which navigate through the database a screenful at a time.
2. Scrollbar
A scrollbar to quickly scroll through the database.
3. Velocity Slider
A widget that controls the velocity at which images scroll by.
2. Thumbnail Type (area occupied by images is 512x512 pixels)
Treatments:
1. 3x3 grid of thumbnails per screen.
2. 5x5 grid of thumbnails per screen.
Dependent Variables
1. Mean time for subject to locate specified image.
2. Mean number of errors.
3. Degree of subjective satisfaction.
Hypotheses
User satisfaction will be greatest for the scrollbar implementation, but lowest for the Velocity Slider. Error rates generally will be very low, however the button-based browser will give the least error rate of the three browser types. The fastest recorded times will be for the velocity slider. However, the variance of times for this treatment will also be the greatest since the widget is not as familiar to most users as the other two, and some may have more difficulty manipulating it than others. These results are expected to be consistant across all thumbnail types.