 | |
 |
 |
 | |
|
Experiment
2.1.1 Independent and dependent variablesThe independent variable, the style of presentations, was defined as either Graphical or Tabular displays. We call our graphical representation Flight-Lines, which shows each schedule of flight by a line segment. In the Flight-Lines, the length of the line segment represents the flight time, and the thickness represents the cost of the flight. The thicker the line segment is, the more expensive the flight is.
FIGURE2.1 Flight-Lines The Tabular representation is Fare-Finder in Travelocity.com, which displays the information by a tabular format. The dependent variables we examined were the time to complete tasks, the number of errors made by the users and the subjective ratings to the two display methods. 
FIGURE2.2 Fare-Finder 2.2 Pilot study resultsFor our experiment, we designed two display formats: Flight-Lines (graphical representation) and Fare-Finder (tabular representation). We also designed three simple tasks, three complex tasks, three recall test questions and eight background and subjective questions.For our pilot study, we used all the above designs. The primary goal of this was to get a sense of whether our format for running the experiment was going to work or not. Three pilot tests were done for Fight-Lines and two pilot tests were done for Fare Finder. According to the time used in the pilot tests, the planned times to complete each phase are estimated. These planned times give us a guide to run the final experiment. However, we gave the subjects unlimited time to finish all the questions when we did the real experiment for better performance. The more important thing is that the pilot study gave us a feel for what needed to be changed in our final implementation of the project. We made some modifications in our designs for the real experiment. We refined the questions to be more specific and more understandable. This helped to cut down on the overall time for each subject to complete the tasks, and it provided more accurate, and less varied data. Three questions in the Background Survey were modified. The question "Are you familiar with graphical user interface?" was changed to "How often do you use graphical user interface?", the question "Have you ever taken airplane?" was changed to "How often do you take airplanes?" and the question "Have you ever booked ticket on the Internet?" was changed to "How often do you book air ticket on the Internet?" We found that two questions in the Subjective Questionnaire were not very appropriate in our experiment. We deleted them finally. A very important factor for evaluating the flight time is the time difference between different places. The pilot study showed us that there was some inconsistency in our design. First of all, we included the time differences between airports at the beginning of the interface in Fare-Finder. Second, in any questions, when a time is shown, it was clearly stated that which airport the time was relative to. The final thing we should mention is the third question in complex tasks. When doing the pilot study, there was no answer for this question actually. This fact confused our subjects. Also, we found that it was quite easy to do this question base on the experience of doing the first question in the simple tasks. We corrected this mistake. In the current version, this task is quite difficult and has a definite answer. 2.3 SubjectsTwenty students of the University of Maryland served as subjects. Ten students were male and the other ten students were female. They are from different departments, such as Computer Science and Telecommunication, etc..2.4 Materials2.4.1 Instructional MaterialsStep by step instructions had been set up for the subjects. Six steps were given to finish the whole experiment. In our instructions, we told the subjects the environment of our experiment. We also designed three simple questions as warm-up exercises. By this, we could provide some training of our designs.Since our team has four members, to guarantee the consistency of our experiment, we designed and formalized the step by step instructions for consistent use individually. 2.4.2 Tasks and QuestionnairesAs we mentioned before, we designed three simple tasks, three complex tasks, three recall test questions and eight background and subjective questions. We have listed them in the appendices.All the materials can be found in the Appendices. 2.5 Procedures and problems2.5.1 ProcedureOur experiment was a between-group design, which was used to look at subject's performance in terms of the format used for data representation.First of all, the subject was asked to read and sign consent forms, then fill out the background questionnaire. Second, we briefly explained the search function and information expression method of our designs to the subject. We helped the subject to do the warm-up exercises and answered their questions. Third, the subject did the simple and complex tasks by following the instructions. For each task, we went through the question with the subject until we were sure that this question was well understood. Then we recorded the time and the answer. Fourth, the subject did the recall tests and subjective questionnaire. 2.5.2 ProblemsThe first problem is we should design the questions to be more specific since some subjects asked us the meaning of some questions during the experiment.We should improve our method for recording the time since we recorded the time manually during the experiment. The improvement would reduce the varied data in our experiment. |