Chapter Summary

Data analysis is an iterative process and the first move in this iteration for variable-based analyses is usually to undertake univariate analysis. This means inspecting the data one variable at a time, and will include displaying categorical and metric variables in tables, charts and graphs, calculating summary measures that will pinpoint key characteristics of each distribution, and perhaps, where the data are derived from a random sample, evaluating the likely accuracy of estimates made from it, or the statistical significance of univariate hypotheses that have been put forward. At this point, the researcher may, in addition, reflect upon the implications of this analysis for the client, for the research objectives and for the next steps to be undertaken in the analysis of the data.

 

The most commonly used summary measures for metric variables (discrete or continuous), or for variables derived from summated ratings and assumed to be metric, are measures of central tendency, dispersion and distribution shape. Summary measures for categorical variables are somewhat limited to percentages, proportions and in some cases the modal category. Once univariate analysis is complete, researchers will usually proceed to the next step, which is looking for patterns of relationship between variables, initially two at a time.