Statistics

The materials below come from an eight week course. Each week students had a 2-hour lecture and a 1-hour tutorial in which homework exercises were discussed. Students also took two 2-hour Stata workshops. At the end of the course students needed to write a 3,000 word statistics paper.

Many students coming to the course either do not like statistics, do not see the relevance, or have an outright fear of statistics. I try to make statistics interesting by using examples that are relevant and based on real data. I try to make the course ‘do-able’ by employing a variety of teaching tools that accommodate to different learning styles, and by providing a lot of support including in-class exercises, office hours, small group tutorials and extensive answer keys to homework exercises.

In my lectures I use clips from Hans Rosling’s energetic plea on The Joy of Stats , XKCD cartoons, examples based on real data, and in-class exercises. Some of these exercises come from Andrew Gelman and Deborah Nolan’s brilliant book ‘Teaching Statistics: a bag of tricks’. I am also a big fan of The Cartoon Introduction to Statistics which includes one of the most accessible explanations of the central limit theorem I have come across.

The big advantages of doing in-class exercises is that students are actively thinking about statistics in class.  This not only encourages their learning  I also found this allows me to better understand what students find difficult and where their reasoning takes a wrong turn.

Key to a successful statistics course is a good team of teaching assistants who are both knowledgeable and patient.  If you, like me, are lucky enough to have a good team, use them to not only help students, but also help you by commenting on homework exercises and answer keys and providing feedback on what students are failing to grasp.

Syllabus

Stata guide
Covers the codes for basic descriptive and inferential statistics, generating new variables and making custom tables. All examples are based on round 5 of the European Social Survey. I have used the rename file to change the variables names to letter-number combinations.

Example of class exercise: Central limit theorem

Examples of homework exercises

Most of the homework exercises focus on problems that are relevant to students in migration studies and use real data (including the European Social Survey). Some exercises are based on published social scientific articles. Students are encouraged to critically think about sampling and measurement, elements too often ignored in many statistics courses and textbooks.

  • Week 1: sample vs population, measurement level and two-way tables
  • Week 2: normal distribution and central limit theorem
  • Week 3: confidence intervals
  • Week 5: chi-square and independent t-test
  • Week 6: correlation and regression