**Mood:** quizzical

**Now Playing:** Time Marches on (Tracey Lawrence)

**Topic:** Data Analysis

I’m looking at my analysis of my data for my LP section. So, what I did, I got Biglan (1973) paper (both of them) and decided to categorized my disciplines according to his methodology that is into hard vs soft; pure vs applied; and life systems vs non-life system. Well, I had some problems in decided which category the disciplines fitted into. For example computer science I felt it should be hard, applied and non-life system. But is computer science considered a pure or applied subject. I wasn’t certain about that. Anyway, he had some categories of disciplines in a table so follow that to some extent. What I did too, was that I went to the course websites and see what it is about and what department they were in to fit it in better if I was unsure – such as things that had built and natural environments (just to tell you – I decided those were life systems).

Well, after I did … decided to look at how the responses were distributed for the coverage and delivery for formulation, solution and sensitivity analysis. Well, I decided to do ANOVAs but something just wasn’t going so right with it – because you know this is ordinal and nominal data rather than continuous. So, looked at a bit how to do analysis with that – at first tried some loglinear analysis – wasn’t sure what results I got … well, finally decided to do logistic regression – which seems to be the way to go. I’m doing ordinal logistic regression using the PLUM module in SPSS and it seems to be working out to some extent … except I’ve discovered one problem with my logic … my ordinal variables (i.e. my responses to delivery and formulation) may not be truly ordinal since I have the ‘not sure’ variable at ‘5’ … so, I was wondering what to do with that!

Well … I thought maybe I should make it a missing variable (I’ve given the value of 999 for my missing variables) – well, I am seriously considering doing that – well, I meet John tomorrow so we’ll see what he says.

### Like this:

Like Loading...

*Related*