About

I require some simple descriptive statistics and analyses of pre/post course survey data from a medical education course. I would have done this myself but am very pressed for time right now, so would appreciate your help. Thanks for your consideration! Tasks: - Match pre and post surveys using participant ID codes - Create comprehensive descriptive statistics table for a set of questions for all pre and post responses to the surveys, and format these results in publication-ready tables - Create a summary variable based on 4 individual questions - Calculate a proportion + 95% CI of students with a certain value of the summary variable - Test for changes in reported level of knowledge across a number of different topics, with FDR correction - Provide clean, commented R or Python code ready to share on OSF in association with a paper - (Optionally, instead of being listed in the acknowledgements of the paper, also become a co-author on the paper if you want to contribute to the results and/or discussion sections) Data is already cleaned and in CSV format, which I can share with you via Google Sheets or CSV depending on your preference. Budget: 5-15 hours. Timeline: 3 days. Ideally, if you could complete this in the next 2 days, that would be amazing. The survey responses are already de-identified, but I require you to treat all the information related to this task as confidential. I plan to hire at least 2 people to do the same task so that the two outputs can be compared to each other. If you use AI to assist with this task, I have a strong preference for you to use Claude please. That's because I prefer Anthropic's approach to data privacy. (Perplexity offers access to Claude, and as far as I know it's possible to send some free messages directly to Claude each day too.) To apply, please message me, briefly listing the things you'd want to keep in mind when conducting multiple statistical tests on a small dataset. Please also state how you would approach matching pre- and post-course responses based on identification codes, including scenarios where there may be ~single-character mismatches between codes and where manual checks may be required in addition to code-based matching. This task involves working with a small dataset containing information from less than 40 participants in total. Thank you in advance!