/AAS-ongoing-tutorials

The Jupyter notebooks for the applied analytic statistics course (MSc Social Data Science) in 2023

Primary LanguageJupyter NotebookMIT LicenseMIT

Applied Analytic Statistics 2023

The notebooks in this repository are being written for use in the course Applied Analytic Statistics for MSc Social Data Science 2023. In each of the directories, the questions notebook contains the questions for the practical sessions, and the answers one has partial answers which will be uploaded after each session. Please give the questions a solid attempt before looking at the answers as this is the best way to check that you really understand them.

Each notebook will be made available on Canvas and GitHub the Wednesday afternoon prior to the TA session. You have a full week to work through the notebook and must submit it to Canvas by the following Wednesday at 5pm. It is expected that the notebook is started during the TA session and completed in your own time. The partial solutions will be uploaded Fridays at 5pm. Please attempt and persevere with questions before turning to the solutions.

Contents

Notebooks

  • The Week 1 Notebook contains an overview of the software packages we will be using and basic Python skills (primarily plotting with matplotlib).
  • The Week 2 Notebook covers probability, distributions and the central limit theorem.
  • The Week 3 Notebook is an introduction to simple linear regression.
  • The Week 4 Notebook extends linear regression to multiple regressors.
  • The Week 5 Notebook covers contingency tables and hypothesis testing with the chi-squared distribution.
  • The Week 6 Notebook looks into binary and multinomial logistic regression.
  • The Week 7 Folder contains model answers from past students.
  • The Week 8 Notebook considers multilevel modelling (under week 7 on Canvas).

Miscellaneous

  • The setup page gives some instructions for setting up the software and notebooks.
  • The additional resources page gives some pointers to additional resources including a list of good places to find datasets.