1. Introduction and background.
Discuss what previous academic studies have shown with respect to your research question. Include at least three academic references. Discuss any factors that others (in the readings) have found to have an effect on your dependent variable [note: where possible these should be factors that you also examine in your coursework]. For full credit answers must:
• Provide at least three reasonable and relevant references, integrated into the text and described clearly.
• Identify the methods used in reported studies.
Provide background information on the data set (Understanding Society Teaching Data 2011-12) and identify the benefits and limitations of using the Understanding Society data to answer your particular research question(s). Here you may want to provide information on the following:
• Coverage
• Purpose
• Sample size
• Methodology (in particular sampling strategy)
• Questionnaire wording of some of the key variables relating to your RQs.
• Benefits
• Limitations
(1) There is useful info on the teaching dataset here:
https://beta.ukdataservice.ac.uk/datacatalogue/studies/study?id=7549
You can also see how to cite the dataset on this webpage.
(2) Information on the precise questionnaire wording, which sometimes diverge a little from the brief description in the dataset, can be found in the Understanding Society Mainstage Questionnaire http://doc.ukdataservice.ac.uk/doc/6614/mrdoc/pdf/6614_wave3_questionnaires.pdf I have provided a copy of this for you on Moodle (under data and data analysis). Please also note the teaching dataset contains fewer variables than the full questionnaire contains.
[20 marks]
2. Select and describe the variables (both dependent and explanatory) that you will use in your analysis.
I. Describe and defend your choice of dependent variable (Y variable). You can elect to have more than one dependent variable if you prefer. Your choice of dependent variable(s) will be informed by the operationalisation you decide on for your research question. Your dependent variable can be an interval-ratio variable or a categorical variable. Explain why the variable(s) you have selected (or created) are a suitable way of measuring the concepts in your research question. Note: It may be that you will need to recode this variable to do the different analyses described below.
II. Describe and defend your choice of independent variables (X variables). You must include between four and seven explanatory/independent variables in your assignment. These must include at least one interval-ratio and at least one categorical variable. For each of the explanatory variables you have selected explain why you think it may be related to the dependent variable. Ideally, you would link this discussion back to the readings reviewed in part 1.
At least one of your variables (dependent or explanatory) must be recoded. You must explain why you are recoding the variable you are recoding and how you did it. The new variable must be named and have suitable variable labels.
Please refer to worksheet 3a for info on how to recode. This can be found under week 3’s teaching materials.
[10 marks]
3. Produce appropriate descriptive statistics for each of your dependent and explanatory variables. These might include, for example: mean, median, minimum, maximum, standard deviation, frequencies or proportions. You will need to decide which measures are appropriate and useful to report for each variable. You should discuss in the text those measures that are most relevant. You may also include any charts that you think are helpful. It is important to only include appropriate statistics. Marks will be lost for irrelevant or inappropriate statistics, tables or charts.
[10 marks]
4. Produce bivariate analyses, showing the relationships between your dependent variable and each of your explanatory variables. (It will be necessary to recode one or more variables to produce these analyses – you should describe all recoding as part of your response to Q2, above). Choose either one or both of the following types of bivariate analysis:
Cross-tabulation: examine how your dependent variable (recoded if necessary) varies across categories of any categorical explanatory variable(s). Provide a brief commentary on any similarities or differences. Use a chi-squared test to evaluate the statistical significance of any relationship that appears to exist. Provide a substantive commentary on the association (or lack of association) between the variables, quoting appropriate percentages.
Correlation: explore the relationship between an interval-ratio dependent variable and an interval ratio explanatory variable. Discuss the direction and strength of any relationship that appears to exist.
[15 marks]
5. Produce a series of (at least two) charts. Choose either one or both of the following ways of displaying data:
Clustered or stacked bar-charts show how your categorical dependent variable is distributed across categories of one or more of your explanatory variables. Discuss any patterns you see in the bar charts and any differences across categories of your explanatory variable(s).
Histograms show how your categorical dependent variable is distributed across one or more of your interval-ratio variables (e.g. how hours of work is distributed among males versus among females). Discuss any patterns you see in the histograms and any differences across your different histograms.
[5 marks]
6. Carry out a logistic regression analysis (recode your response variable if need be) in which you include between four and six explanatory variables. Explain your methodology and write up the substantive implications of your results. You should also identify possible weaknesses in your analysis.
[25 marks]
7. What would you conclude on the basis of the results of the analyses you have carried out in this project? How do your findings relate to the existing body of academic work that you discussed in section 1?
Which are your most interesting findings? What, if anything, might you have done differently? How could you develop this analysis in the future?
[15 marks]
Include your syntax in full as an appendix. Include information to clarify to which part of your analysis each part of the syntax refers.
