Invited Speaker: Professor Donald B. Rubin (Harvard)

At the first ERSA Econometrics Workshop, one of the key areas identified for strengthening within Southern African econometric research was that of Bayesian statistics.

In essence, the Bayesian paradigm is concerned with updating prior model beliefs or assumptions with observed data, to generate a posterior view of the model in question. In this way, Bayesians are able to make data-informed probability statements about model parameters (and possible future observations), meaning that this paradigm is well suited for decision-making. In a time where a notable international scientific journal like Basic and Applied Psychology has banned the use of the classical p-value, Bayesian statistics is becoming an invaluable tool for inference based on data, regardless of the research field.

The workshop is intended for Southern African econometricians already familiar with Bayesian concepts, looking to expand their utilisation of Bayes within their own economic research, and incorporate cutting edge Bayesian technologies. The first day of the workshop will include a practical presentation on implementing Bayes using OpenBUGS and intermediate R, followed by paper presentations and discussions. The second day will be dedicated to more complex Bayesian topics, and will include a presentation by world-renowned Bayesian, Prof. Donald Rubin, from Harvard.

The workshop will be a combination of technical presentation and roundtable discussions between academics and policymakers. This is not a training workshop, and only researchers with knowledge of Bayesian theory or who have used Bayesian theory in their research will be considered for participation. Southern African researchers are invited to submit contributions and papers related to the theme for the organising panel to consider.