BSc Global Development with International Study / Course details

The further statistics course unit builds on the intermediate statistics course unit. Its main aim is to ensure that students have (i) all of the generic skills necessary for an undergraduate dissertation which is rich in quantitative methods, and (ii) the ability to find information about additional statistical tools and to apply these tools in appropriate way.

The unit builds on Intermediate Statistical Methods (MGDI20251), providing students with a grounding in techniques that can be used for the statistical analysis of time series and longitudinal data. The unit finishes with a section designed to equip students for lifelong statistical learning

Weeks 1-2: recap on binary outcomes and discrete choices (Greene chapter 17)

Weeks 3-4: multinomial choices and event counts (Greene chapter 18)

Weeks 5-6: introduction to time-series analysis (Greene chapter 20)

Weeks 7-8: cointegration and multivariate times-series modelling (Greene chapter 21)

Weeks 9-10: lifelong statistical learning

Teaching and learning will be based on lectures (twice per week) and hands-on classes (once per week). Lectures will combine presentation of the learning material with some interactive discussion. Reading lists, lecture slides and tutorial questions will be posted on Blackboard. Classes will provide students with the opportunity to grapple with data under with guidance from a tutor. Advice on how to prepare for each class will be posted on Blackboard.

• Select longitudinal statistical methods that are appropriate, given the characteristics of their data
• Identify the main properties of time series relevant to statistical analysis

• Evaluate choices about statistical model selection in the context of longitudinal and time-series data
• Critically analyse results from the estimation of time-series and panel data models

• Use Stata to fit a longitudinal model appropriate to the data
• Analyse the characteristics of a time series using Stata
• Use Stata to fit a time-series model appropriate to the data

• Search effectively for literature on a statistical problem novel to them
• Manage data using Stata

1. First draft of assignment 1, a report based on an analysis of a longitudinal dataset. (1,000 words) (Formative - no weighting)
2. Second draft of assignment 1. (1,000 words) (weighting 35%)
3. Assignment 2, a report based on an analysis of a time-series dataset. (1,000 words) (weighting 35%)
4. Assignment 3, a short essay evaluating the statistical methods appropriate for a dataset with properties new to the student. (1,000 words) (weighting 30%)

Written comments via Blackboard.

William H. Greene (2017) Econometric Analysis (8th edition), Pearson: New York, NY