▼ In Pair Programming (PP) two programmers are working together side by side on the same task using a single computer. PP is one of the twelve practices in Extreme Programming which promises to produce reliable, efficient, and well-factored software products. The benefits of PP are well documented in experimental studies. Some of the benefits include improving design quality, reducing defects, reducing staffing risks, enhancing technical skills, and improving team communications. Pair programming has become increasingly popular in the field; however, some programmers are skeptical of pairing. Many believe that pair programming success highly depends on the compatibility of the pairs, which is determined by the personality traits of the partners. In this research, personality traits’ effects on pair programming success were studied empirically. The hypothesis is that certain personality traits are crucial for the success of PP, and PP partners should be chosen based on these personality traits. Programmers in the software development industry were surveyed to identify the perceived important personality traits for successful PP. A series of experiments were conducted among undergraduate students in computer science classes to determine the significance of these personality traits in successful PP. The successfulness in PP has been evaluated by measuring the quality of the produced code and the pairs’ compatibility. Then, the collected data were analyzed through ANOVA and multiple comparison tests. The study found that there are some indications that personality traits of the pairs influence success in pair programming in terms of producing high quality codes and pairs’ compatibility. Advisors/Committee Members: Chao, Joseph.

▼ In recent years, one of the more popular ways of presenting information is as a three-dimensional graphical model on a computer. With 3-D graphics becoming more popular, they are starting to be used for instructional purposes. Instructions presented in 3-D should make performing a task easier than a task presented with two-dimensional instructions. BGSU researchers in computer science and psychology have conducted a number of studies that involved presenting instructions with three-dimensional graphics. The current study represents a convergence of the BGSU work involving the presentation of instructions for construction tasks and the use of stereo graphics. The task itself used a true three-dimensional artifact, a LEGO block, rather than the paper artifacts of the prior BGSU studies. The purpose of this study was to determine the effects of stereoscopic technologies on participants’ ability to perform a LEGO block construction task. This was a qualitative study that consisted of six participants who were all computer science graduate students. The task for this study was to construct a 111 piece LEGO tower for a pre-built castle. The tower construction task was divided into two parts with the participants building one part with the stereo glasses on and the other part with them off. The results of the study found that two of the six participants thought that the stereo glasses were a benefit in constructing the object. Most of the participants felt that the glasses did not help in constructing the object. The stereoscopic glasses for the most part do not appear to affect the results of the study. One reason the glasses did not appear to help was because the software provided enough features that the glasses were not necessary. There was also some indication that while performing the higher complexity steps the participants were slower when wearing the glasses. Advisors/Committee Members: Leventhal, Laura Marie.

▼ Acceptance tests are formal testing conducted to determine whether a system satisfies its acceptance criteria or not and whether the acquirer should accept the system or not. A suite of acceptance tests for large projects might include a large number of test cases; therefore, automation of acceptance test is in great demand. Framework for Integrated Tests (FIT) is a popular tool employed in Agile Software Development to automate acceptance tests. Its most attractive feature is that it uses customer readable tables as test cases so that customers can write test cases. Refactoring is the process of restructuring or rewriting code without changing its interface and functionality. Refactoring make the code easier to read, understand and maintain, and sometime helps to improve the performance of the system. In a typical project that uses FIT as an acceptance test tool, the size of FIT acceptance tests grows as the size of system code grows, and the acceptance design may go far away from the original design (this may happen in any project, not restricted in a project using FIT). At this stage, it would be difficult to read and maintain the FIT acceptance test, and it is time to improve the quality of the acceptance test. In this research, we introduce the concept and reveal the importance of FIT Refactoring. Several FIT Refactoring methods are introduced as examples to show the needs of FIT Refactoring and the methods how it can be accomplished. Of course, the methods given here are some obvious refactoring methods, and new methods can be discovered in further research. We also discuss the proper time to do FIT refactoring and proper efforts that should be devoted in it. The similarities and differences between system code refactoring and FIT acceptance test refactoring are also one part of the research. During the research, there are some unexpected findings. One of them is that sometimes, the bad code in FIT acceptance test indicates bad code in the system code. Advisors/Committee Members: Chao, Joseph.