SOCI832: Lesson 2.2: Exercise

Lesson 2.2: An example

Practice Exercise


  • Can you identify the following concepts in our model/demonstration article:
    • theory & hypotheses
    • units of analysis and variable
    • dependent variable, independent variables, control variables
    • conceptualisation of a variable
    • operationalisation of a variable
    • one of each type of variable: categorical, binary, ordinal, continuous (or interval)
    • a scale or index
    • the source for the scale or index
    • evidence the scale, index, or other variables are reliable and valid

Write your answers in the Google Doc.

This is a link to the Google Doc for this week’s class.

Background: About the article

This week and next we are focusing on dissecting the following article:

Ian McAllister (2016) Internet use, political knowledge and youth electoral participation in Australia, Journal of Youth Studies, 19:9, 1220-1236, DOI: 10.1080/13676261.2016.1154936

Ian McAllister is a Profession at the ANU, and one of the leading experts on electoral behaviour in Australia. You can read more about him at these websites. Note that Google Scholar says his articles have been cited over 18,000 times.

You can download this article via this link: (it will ask for your MQ login, and then click on the “PDF” button.

The paper deals with an important problem of democratic societies: the low level of political knowledge and political disengagement of young people. It asks whether internet use – something which has been heralded as way to reinvigorate political engagement amongst the young – actually does have a positive impact on political knowledge and engagement.

The Full Dataset from the AES (don’t use this yet!)

I picked this article because the dataset for this paper is publicly available. You can download any of the Australian Election Study datasets from this page:

If you have a look around this website, you will see there are other interesting datasets too, particularly the Australian Candidate Study (a survey of candidates for the House of Representatives and Senate).

The dataset for the McAllister 2016 paper is actually the AES 2013 data.

If you go to the website for the 2013 data you will see something like Screenshot 1:

Screenshot 1: The Australian Electoral Study website for downloading the 2013 AEC data Notice there are three files to download:

  1. Data (SPSS): This is the data file in SPSS format
  2. Codebook: This is the codebook for the dataset (the codebook describes all necessary information about the variables in the dataset, including the values of the variables, and their meanings.
  3. Questionnaire: This is the actual questionnaire given to the participants in the study. Feel free to download and have a look at these three files, particularly the codebook and questionnaire. 5.3.2 A simplified dataset for this class (use this when starting out!)

When preparing for this class, I’ve been reviewing the SPSS data file provided for this dataset, and honestly I think it is a bit too complex and burdensome for the purposes of teaching this class – at least for simple demonstrations. The main issues are that (1) each variable is labelled by a question number (e.g Gender has the code ‘h1’) – which makes it hard to intuitively use (you need to keep referring back to the codebook); (2) there are some complex missing data codes; (3) SPSS format – with labels – can be a bit complex in R when using them for the first time (you need to use the package ‘sjlabelled’ with SPSS data). For these reasons, I’ve prepared a simplified an easier to use version of the AES 2013 data for us to use for the purposes of this week’s exercises. This simplified dataset, and the codebook for it, can be downloaded here:

Dataset for Weeks 1 + 2 – McAllister 2016 example: (Link 1)

Codebook for Weeks 1 +2 – McAllister 2016 example: < aes 2013.html> (Link 2)