DISCUSSION
The aim of this study was to investigate whether the data supported the initial goals of the museum's kiosk software design.
A combination of two of the museum's goals for the program was to introduce the visitor to the museum, its collection, and its exhibits in a graphically interesting manner. The ability for the software program to introduce the visitor to the museum was assessed by determining whether images encouraged exploration of the exhibits. Seventy-seven percent of the subjects did not feel that any image sparked their interest enough to try to find that piece of artwork in the museum (Appendix 6 & 7, Question 1a & 1b). However, most people (73%) did recognize images from the software while touring the galleries (Appendix 6 & 7, Question 2a).
Subjects indicated they would like more images of artwork in the program (Appendix 6 & 7, Question 2b), even though the observation data more frequent image errors. This may indicate the subjects used images as prompting cues for movement between screens. The software did contain some images that were links as well as some that were not. Clicking on an image that was not intended to be a link was counted as an image error - this may explain the high frequency of image errors observed. Consistent use of images as links may help reduce click on image errors.
Appendix 4 summarizes ten user responses related to software preferences. Only 30% of the participants liked using the software and only 33% of the subjects reported they liked the design of the software. Some of the changes that could be implemented come from the other preference responses within the data: Users generally felt that the graphic display of text was not too long to read, but 70% felt that the text itself was too small to see. A standard practice within the field of interface design is to use high contrast of colors between the text and its background (for example, black-on-white, yellow-on-blue). Within the software program important information was displayed using red text on a blue background. This is not a recommended color scheme because of two reasons: 1) chromatic aberration (color blindness) and 2) the contrast between blue and red is too difficult to detect.
Forty-three percent of the subjects felt that they would like to see a graphic representation of the museum's floor plan or a print-out of one of the screens. Only 18% of the subjects felt that the computer helped them find their way around the museum (Appendix 6 &7). As well, the lack of signage was indicated as the least liked aspect of the museum. These results, coupled with the fact that half of the subjects were from out of town and first-time visitors, indicate that clearly there is a need for better way-finding tools within the museum.
Another goal of the museum was to create a program that was easy to use for all user groups. The most obvious problem with the program concerns the "click & hold" function of the software. Those errors represent 59% of the total errors. This function was created in order to reduce the amount of time that the users would spend on each screen, unfortunately it became the most reoccurring complaint about the software. From the data collected, its appears that both the youngest and the oldest age groups erred the most often when navigating through the program (Appendix 8 & 9).
Ergonomic guidelines for user interface design states that "certain aspects of an interface should behave in consistent ways at all times for all screens (http://www.glover.com)." Mouse operations (i.e. number and manner of clicking) should be consistent throughout the software. For instance, if a single click opened the first layers, a single click should be used to open further layers (as opposed to click and hold).
The consistent use of single click could alleviate the click and hold problems. Another solution may include adding images and screens that include the words "open" and "close" to indicate function.
Results may have been biased due to the following issues:
Subjects were not randomly selected. Researchers asked visitors for their participation in the study. The only subjects that approached the kiosk without persuasion were children. A customized kiosk may be better suited to the task of getting a visitorís attention and inviting them to use the program. One possible solution other than a new kiosk design would be to install a rug that would lead to the computer.
During the observations, all of the subjects knew that they were being observed, and this may have influenced time spent using the program. Perhaps users felt obligated to look throughout the entire program and ended up spending a grater amount of time than they normally would have. On the other hand, subjects may have felt pressured by the observers and cut their time short. A better study would allow the subjects to use the computer kiosk by themselves, without observers looking over their shoulder.
The effects of the presence of other people on computer use patterns and learning patterns is a potential area for further study. A crowded museum lobby may inhibit the ability of a program to effectively "teach" the user. And if two or more people are using the program at one time (i.e. a parent-child combination or a married couple) each user may not get as much benefit from the program as if they had used it alone.
Sample size and attrition may have effected statistical results. Specifically, a small sample size of 30 may have increased the statistical error, and attrition further reduced the sample size to 22 at the time of the exit interview.
There are several ways to improve upon this study. The most obvious is to increase the sample size. It would also help to have a control group of subjects who did not use the computer to answer questions on way-finding and anticipated amount of time they will spend in the museum.
It would be more accurate to test those subjects who approach the computer unprompted, as they would do in a non-experimental setting. Observers should record data from a more distant, non-invasive position. For instance, a second monitor/screen could be used in an adjacent room, or the interaction could be video taped and evaluated at a later time.