Notes on section 5.3 of the paper "Why do women pursue a PhD in Computer Science?"
Participants' intentions and context
This analysis is based on the following Likert-scale type questions:
Q22. When being a Bachelor student, I plan(ned) to enrol in master studies
Q23. I have thought about doing a PhD.
Q24. I plan to do a PhD.
Q25. I am aware of the aims and contents of PhD studies.
Q26. I am aware of the requirements for starting a PhD.
Q27. I have been encouraged to do a PhD.
Q28. After completing my current studies, I am considering moving abroad for a while.
Q29. My family encourages me to do a PhD.
Q30. I was motivated by industrial contacts to do a PhD.
Table 5 is extended here with three columns making explicit the rank of the hypothesis, by p-value, the Holm-Bonferroni corrected significance level, and the Bonferroni corrected significance level.
Extended Table 5
The detailed data analysis, including the Chi-Square tests can be found in the Excel file SurveyDataAnalysis, on the sheet RQ1. MW-UQ12-Q43Q47-Q54.
To help in the interpretation of the data, the following chart summarises the results obtained for these questions. The data for this chart can be found in the Excel file SurveyDataAnalysis, on the sheet RQ1. Q22-Q30Diagram. This chart only presents results for questions where there was a statistically significant difference between the two groups, in descending order of effect size.
The distribution of responses to questions related to participants’ in- tent and context factors for those questions with statistically significant differ- ences, with medium effect sizes, from those who chose to do a PhD and those who didn’t.
Participants encouraged to do a PhD were much more likely to enrol in one. Having such an incentive from one’s family, or by an industrial contact is also relevant, but the latter is very uncommon. Our data also shows how many participants were unaware of the requirements for starting a PhD, and how being aware of such requirements can make a difference.
