ABSTRACT
Analysing the data and interpreting the results is one of the most exciting stages in a research study because this provides the answers to the study questions. In general, any data analyses should progress through the logical steps of first conducting univariate analyses before progressing to the bivariate and then the multivariate analyses. Missing data must be treated very carefully. Missed data that occur in a random pattern reduce statistical power but rarely affect the results. Every stage in the process of collecting data, coding responses, and making decisions about data management must be documented clearly in a data management file. For categorical data, histograms and frequency counts will show whether any groups with small numbers need combining with other categories. The categories that are used in the data analyses must be meaningful and must preferably be in concordance with any prior evidence of effect.
