Important Disclaimer

I have discussed linear regression with you this week for two reasons:

1. I am trying to sketch the methodological development through which the relationship between economic development and democracy has been analysed over time. Lipset (1959) started with a simple correlation analysis, and this quickly developed into linear regression analysis, as it is more powerful and more sophisticated than correlation analysis.
2. It forms the foundation for the binary response models (probit in our case) which we will start with next week. To start with, we will only look at the probability to be democratic in a particular year across different countries. But in Week 7, we will extend this to not only investigate this relationship across countries, but also over time. This dynamic probit, or Markov Transition Model (MTM), is the model you need to apply in the assessment of the module. Please DO NOT stick to linear regression in the assessment, as OLS is not capable to deal with time-series data. The same applies for a cross-sectional probit. The only method permissible int he assessment is an MTM.