Reflections on Theme 3 - What I've learnt

I think this week's seminar/exercise and lecture was quite interesting and fruitful. As I hoped last week, I’ve learnt quite a lot more about quantitative methods. Or not, really depends on your epistemology definitions. Anyway, of the lecture and the seminar/exercise , I found the exercise much better. It was a good introduction and quite fun to tinker with SPSS. Even though the GUI was quite simplistic, there were some serious capabilities underneath. I think it would’ve been better if we have had more than one hour to get acquainted to the program. Sure, you could sit yourself and play around with it, but then you wouldn’t get any help from a teacher/PhD. I’ll see if I’ll sit down with SPSS further down the road.

So far, my reflection has been quite general, so from here on, I’ll go through step by step (kind of) on what I did together with Niklas during the exercise.

As pictured above, the first thing we did was to make a table for question 32 “Did you listen to radio yesterday? IF yes; for how long in total?” It was a single choice question, so there weren’t really any hassle to analyze it with the program. Pretty straightforward.

The second task was to do the same but with a multiple response question. This was also pretty straightforward, although it was kind of hard to make out the responses in the window since you couldn’t resize it. We forgot to screenshot it, but we got a screenshot of a multivariable cross table, with when you listened and with what, as seen above.

The next task was to make a visualization with excel. You can see two of the above; the staple diagram and a linear one. Of them both, the staple one is far superior and really makes it easy to identify the times in which people listen and with what. The distribution seems to be the same for all the mediums, which could point out that there may be some problems with the questions. However, as Ester pointed out, the study from which we got our data is rigid, so the patterns are correlating to reality.

The last task was to tinker with the explore option. We chose to check out the statistical page. As you can see above, you get the statistical information presented neatly, which is good, since doing it manually is quite a hassle. Although it doesn’t say much now, but if you compare the same answers over several years you can see if the deviation is a coincidence or a “permanent” change in behaviour.