amanito Posted May 20, 2012 Hello, I'"m desperately searching for someone who can help out my wife . Sorry for the vague explanation, but I never had statistics and it has been years since my wife had it. She's writing the protocol for her master thesis (fysical therapy) and she's gonna do a questionnaire with two groups and she needs to calculate a sample size. What's she's gonna do is a paired t-test and has to know how much persons are needed and the calculation for this. I don't how the fucking you can do this without having any data yet??? Necessary software: PS POWER AND SIMPLE SIZE PROGRAM Link: http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize She says that she knows the following values already: Power: 0,8 en alpha: 0,05, the rest no clue. (you might wanna doubt this ...) Few pdf's ... http://www.2shared.com/document/F6n41Can/HRQOL_randomized_trial_hypofra.html http://www.2shared.com/document/5kYV5zoE/Kirkwood_PoweraSampleSize.html http://www.2shared.com/document/qvxhltKY/Portney_PoweraSampleSize.html I posted this here cos' there are a lot of smart cookies here and I really hope that someone can help us out cos her deadline is in like two days. THANKS A BILLION. We'll be forever grateful. Share this post Link to post Share on other sites
Halcyon Daze Posted May 20, 2012 I couldn't forget all that stuff fast enough (bad teachers lol). http://www.ats.ucla.edu/stat/R/dae/t_test_power3.htm http://en.wikipedia.org/wiki/Sample_size_determination Anyway, her supervisor would help her with that, or she could phone her old stats lecturer, that's their job. My gf has just given me less than 1 day to apply for a masters for her lol. I'm like WTF! Share this post Link to post Share on other sites
Alice Posted May 20, 2012 (edited) Edit: Probably better as a pm. Edited May 20, 2012 by Alice Share this post Link to post Share on other sites
hebrew Posted May 21, 2012 yeah her supervisors should have already discussed this with her, and she should have already gone through how she was going to be undertaking data collection and analysis at candidature, but i know some quantitative people use Nvivo. I have no idea what that software does, as i am qualitative. but anyway, best bet speak to supervisors, or library staff, or other collegues in her discipline. Share this post Link to post Share on other sites
hebrew Posted May 21, 2012 i guess when you say protocol, you mean proposal? in that case, do some reading, John Cresswell, is a good place to start or Mile and Hubermann, i am sure there are others, you dont want a too small sample size or too large, these things you discuss with your supervisors they will put you her in line as to what to look for. sorry cant really help much more than that. Share this post Link to post Share on other sites
amanito Posted May 21, 2012 Hey, No one wants to help her in her university ... just like that. I'm a teacher and my speciality is foreign languages .... math ... don't even know where to begin. Share this post Link to post Share on other sites