Data can be analysed quantitatively or qualitatively – depending on the type of data and needs of the evaluation.  In some cases, qualitative data can be collated, categorised and counted within categories to also provide quantitative data.

QualDATA can assist in planning or undertaking analysis of evaluation data.

Quantitative data

A critical distinction with quantitative data is whether it is ‘statistically representative’ of a population or whether it is a not necessarily representative.  This makes a difference in whether statistical tests are used and whether statements can be made about the broader population from the results.  There is a difference between “the survey showed that 10% (plus or minus 2%) of a population is using the new technology” and “10% of survey respondents are using the new technology”.  The former is based on a randomised statistical sample and the latter is based on a particular sub-set (not random nor necessarily fully representative) of the community who were selected to be surveyed – or who self selected to respond to a survey.

There are excellent statistical packages that can be used to undertake statistical analysis of quantitative results – and Excel has a great capacity for such tests.  Likewise, Excel and other spreadsheet programs allow results to be collated, converted to tables and a range of graphs for presenting numbers, figures and percentages – whether statistically representative or not.

Qualitative data

Qualitative data can be analysed relatively easily if you are systematic about it.  Answers to open questions in a survey or interviews can be collated, categorised, grouped and counted using a spreadsheet program.  Categories (groupings of similar topics or types of responses) are either pre-determined because of your evaluation questions and context – or they emerge because they don’t quite ‘fit’ into your predetermined categories.  As pointed out above, these can then be shown as tables or graphs alongside quantitative data – supported by quotes taken from the responses.

There are a number of software programs that allow you to code your qualitative data (and hence collate comments/sentences/ideas that are coded the same) or which can analyse the data with some initial guidance from you – looking for patterns in the data.

Using small narratives or vignettes as illustrations in an evaluation report also adds to the understanding of results.