PAPP102 - S06:  Internal evaluation of demographic data

Diagnostics (cont.)

Evaluating internal consistency of the data

In addition to evaluation of the data by reference to external sources of information described on the previous pages, it is possible (and necessary, and desirable) to evaluate the internal consistency of the data collected. Quite what is investigated depends fundamentally on the nature of the research question posed, and the data used to answer it: if the analyst wants to examine sex differentials in income of adults, for example, it follows that these two variables should receive the greatest attention.

In general, the approach to evaluating the internal consistency of the data relies on tabular and graphical analysis of the variables of interest, and then seeking to identify (and, where possible, to offer plausible explanations of) any inconsistencies in the results. To this end, it is sometimes asserted that demographers need to be detectives; seeking out errors and flaws in the data so that the overall quality of the data can be understood better.

Since most demographic analysis hinges centrally on the age and sex structure of a population (age and sex jointly are the most important determinants of fertility, mortality and migration in a population), we set out below the important investigations that should be conducted.