Getting Down to Stats

This last week has been one of facing up to the volume of data that this research is generating.

Completion of the pre-course questionnaire continues at a trickle with 561 students having now completed it yielding 481 usable sets of data. So far 70 students have returned their completed diaries. Interestingly some appear to have chosen to give someone else's ID for the diary!

How to analyse the diary content has been high in my thoughts. During the teaching week in London last week I had a one-to-one session with one of the lecturers who knows about stats and SPSS. It was really helpful to 'voice' the research and data issues, and the result is I am much clearer about the form in which the data needs to be recorded .

One issue that stats cannot resolve is that of sequencing ie the order and times that students spend on different activities during a day. My thinking at the moment is that this will have to be looked at visually. I have therefore created a spreadsheet that shows the time students spend on the four main activities (resting, working, studying and other) as colour coded 10 minute blocks. Already this serves to show whether an individual is blocking longer periods of time or grabbing short sessions, and whether this is happening during the week or at weekends.

Another tool I have created is a proportional 2-D chart that summarises the student's week, and is based on one illustrated by Gershuny (see previous blog). Created in Excel, it shows the proportion of time spent on the 4 main activities as a grid with columns of width equal to the percentage of time spent on that group of activities. Each column then sub-divides into the time activities within that group with the height of each block proportional to the time spent on it. One idea for using this is to print each chart on acetate and to overlay them. This should help to identify common patterns of time usage (and may be simpler than trying to do multi-variate analysis!!). I'll let you know how I get on.

I'm one of those people who is more than happy crunching number but still get lost with the more advanced stats - mainly due to not using the techniques often enough. For anyone in a similar position, or even for those who are unsure about statistics, can I recommend a book by Reva Berman Brown and Mark Saunders called Dealing with Statistics. The great thing is that there are no formulae or numbers included at all!! What they do is to talk you through the basics of statistics and what techniques can / should be used for different purposes. It is only a short book so doesn't take long to read but quickly gives the sorts of reminders that a rusty mind like mine needs.