Suggested layout of the research report

This assumes you will use the GHS data. You can use any other dataset supplied. Alternate datasets for your report are also fine as long as they are cleared with your tutor.

This is quite a conventional model for the structure of a research report. The aim is to cover all this in around 1500 words so you will need to make each word count.

Introduction –
• What the research question is (or what hypotheses you were testing);
• Why these are of interest with brief reference to some appropriate literature.

Data and methods –
• the nature of the secondary data and the sample (GHS Social Capital dataset or other data supplied);
• how the concepts (eg social participation) in the research are operationalised in this research (what questions were asked in the survey; what the dependent and independent variables are in your analysis based on these questions); what are the possible limitations if any of this data (or to put it another way is this a valid way of measuring whatever it is you think it is measuring);
• some descriptive statistics of the variables.

Results – of the analysis. What does your analysis show? The introduction will state what question or hypotheses you were addressing in this piece of secondary analysis so your analysis will provide the evidence for your conclusions and answer to the research question (or tests of the hypotheses). Ignore any analysis which did not contribute to answering your research question (or testing your hypotheses) and any odd results not relevant to your question. You therefore have to be selective about the SPSS you include. We are not looking for masses of output but very selective use of output to support your findings from the analysis. In particular you don’t need to have SPSS output for all tests of statistical significance. This can be set out in the text using p values. It might be helpful to look at a paper using quantitative analysis.

Conclusions – your answer to the research question; what further issues were raised by the work (if any) and whether any generalisations might be made from these results (NB the survey once weighted is said to be representative of the population of GB so it might be a good basis for generalisation).