Sunday, November 20, 2011

multiple regressions

after spending last weekend re-reading Adrian's ESM study to remind me of my hypotheses and actually writing out what i want to test with my data, i discovered that while i knew what i wanted to test, i didnt know what test to actually carry out in spss. so while i had hoped to present data analysis (at least something) to adrian on thursday, on wednesday i told him that it'd be more of me asking what tests to do. (he was fine with that luckily).

i did manage to make a few charts. just simple percentages stuff about the participants and about when (day/hours) music was vs. was not heard. but at least it was something to show adrian.

(thursday, 17 november)

so adrian and i sat down with the data and questionnaire questions and we determined i should use multiple regressions for a number of analyses (and some chi squares for some down the line). i didnt know how to compute those, so we spent some time looking at that. but it required a number of additional re-coding to happen first.
glad to have a direction to go with it. and it felt good to clearly be learning something - felt like being in a class at school which i haven't felt in awhile.

No comments:

Post a Comment