Journal of Student Research 2015

75

Contingency Valuation and Interviews in Open-Source Software Market:

and product changes. The strongest characteristics that appeared were past use of software and regression analysis. These two features were discussed at length in all three interviews, which are indicative of their importance to this market. A finding that was unanticipated in the preliminary stage of this study is the impact that previous software experience in graduate school has on the software they currently use in their research. One participant noted that he had not purchased any new software since graduate school. In this case, he was still using Stata 6, which is somewhat shocking because Stata has major releases every two years with minor updates occurring annually. With each major update, their software is discounted to those who purchased the older version. Stata 6 was released in 2006, which means they have had three major releases and four minor updates since this participant last purchased software. This type of analysis was also important to all three participants and was primarily dependent on their research interests. Support, stability, and updates were also important to two of the participants. One participant has used the open-source product R but does not currently use it because of the extensive coding experience required to use the software. The participants showed a positive outlook on the open-source software market’s total quality, flexibility, and potential. One had contributed to the open source project previously but did not anticipate any future involvement. The most com mon piece of open-source software being used is R, as all three participants noted some experience with it. The most common software being used is SPSS because the university provides it to their students. Price has an impact, especially considering the case of Interview T. Future interviews may show a consistent pattern of a single purchase providing adequate utility. size. The graph below displays the design of Contingent Analysis using Dis crete Willingness-To-Pay. All values have a minimum of $0 and a maximum of $1000. A limited selection bias is not an issue because values within single user market research have shown extremely limited markets beyond even the $700 mark. The data currently is far too limited to make any conclusions regarding the statistical significance and/or reliability of models. The research a priori expectations are listed in Table 1. Establishing expectations without considering the data sample is a fundamental tool when working with a limited data set. The specific correlation trends are proceed ing from the qualitative interview research conducted in late 2013. This information, coupled with economic literature, provides a clear anticipation of how we can expect this market to function. The interview research was Survey Methods and Data: The initial survey deployment has been limited due to small sample