- What is the sociological/criminological rationale for the relationship(s) you have chosen to
analyze? Why have you chosen to look at this/these particular relationship(s)? Why is this a
sensible piece of analysis to attempt? What informed your choice of variables, and what
variables did you choose? You should draw on relevant academic literature/research to support
any claims you make about the issue(s) you are addressing via your analysis.
- Develop and state hypotheses regarding the relationship you will analyse. You should set out
what you expect to find regarding any relationship you have chosen to look into, and it should
also be clear why you expect to find this. (Remember, to interpret figures of ‘statistical
significance’, we need a ‘null hypothesis’.)
- You may need to discuss the nature of the source of the data you are using for your analysis, if
this seems relevant to your analysis. You may need to consider who produced the data, how and
why. Are there any known problems with or limitations of this data that are relevant to your
analysis?
- You should make clear any changes you have made to the dataset. For instance, if you
considered it necessary to recode a variable, or set codes as missing etc. you should explain why
this was done.
- Some univariate analysis of the variables under consideration is likely to be useful. Before we
consider the relationship between variables it often useful to know what characteristics they
have in isolation. For instance, if we wish to account for variation in any variable ‘X’ by
considering its association with another variable ‘Y’, knowing what data we have relating to these
individual variables before we begin is helpful. Results of any univariate analysis, like multivariate
analysis should be appropriate, accurate and clearly presented.
- Conduct and present an analysis of the relationship between two or more variables, using
appropriate tests and techniques to look into the strength and significance of relationships. You
will need to present appropriate findings and offer an accurate interpretation of these findings.
What do the various results tell you about the nature and strength of the relationship you are
analysing? How can these most efficiently be communicated to the reader?
- Draw some conclusions from your analysis and discuss/explain results – why do you think you
have found what you have found? Did you find what you expected? What do the findings mean?
Here you may wish to once again consider relevant literature/existing research. You should also
discuss the strengths and limitations of your analysis and of the data you have used.
