Today I spent some time exploring the data that I’ll be using for my thesis. I initially set a goal of an hour of work at around midday, but ended up doing about 3 hours. Instead of forcing the exploration, I gently meandered through, and once my set time was up I ended up continuing my work anyway.

The data I’m using comes from a big Randomised Control Trial (RCT) done on the Activate! leadership program between 2013 and 2015. The data was collected as an investigation into the effectiveness of the program and the results were really positive. The paper hasn’t been published yet, but if you’d like to read about the study you can find info here.

The CES-D Score

Part of the data collection included questions from the CES-D scale, which is used as an indicator of depression. The CES-D score has been shown to be a good predictor of depression and is used frequently in literature on mental health. A high score indicates a greater likelihood and severity of clinical depression.

I’ll be using this as my main variable in my thesis, and will begin exploring how the program affected depression and how depression affected the program outcomes. Depending on the results, this could inform how to improve this program or others (e.g. by including a psychologist as part of the program team).

Some interesting relationships

So today I looked at the relationships between the CES-D score and the descriptive variables. I ran a bunch of regressions and found a number of really interesting results, with some variables showing strong relationships with depression. Note, that these correlations do not imply causation, just a relationship. It is difficult to tell the direction of this relationship without further analysis. The relationships were as follows:

  • Exercise: there was a small, positive relationship between frequency of exercise and depression. In other words, more exercise was linked to lower depression
  • Unemployment: an unemployed individual was predicted to have a higher level of depression. The magnitude was quite significant, with an increase of around 0.5 points on a 16 point scale.
  • Present health: a very strong predictor of depression was physical health, with those reporting better health expected to have much lower CES-D scores.
  • Smoking: This one was fascinating to me. It was by far the highest effect size of the variables I explored. If a person was a smoker they were predicted to be more depressed than a non smoker by around 1 point on the scale. Again, we can’t be sure of the causality but it’s still really interesting.
  • Wanting to move neighbourhood: finally, I found a strong connection between depression and wanting to move. People who wanted to change the neighbourhood they lived in were predicted to have higher CES-D scores.

I think this is a great start, and has really made me excited about what else I might undercover in the data. I can’t wait to get back into it tomorrow

Image was taken on the promenade during my run this afternoon. 



Blog 58/365. Read more about my #365of25 journey here


Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.