Evaluation Toolkit for Breastfeeding Programs and Projects

June 2012

4.5 Analysing and interpreting data

Page last updated: 04 November 2013

The analysis approach will depend on the kind of data you collect.

Some data may be extracted from clinical software systems and gathered together into spreadsheet software, where simple calculations of rates and proportions, means and medians can be done. More complex cross-tabulation of results or multivariate analysis may accompany complex numeric evaluations, but these exercises are likely to be undertaken by specialist evaluators and are not usually required as part of a general analysis as to whether the outcomes of a program is moving toward or away from its objectives.

If you are collecting data on an audit sheet, or other template in which activities or interventions are being counted, this data can be entered into spreadsheet software and analysed. Alternatively, in a small service, it might be as easy to do this with pen, paper and a pocket calculator – it depends on your own needs and your resources.

For a qualitative analysis there are a variety of approaches one can take. The most direct is simply to read through interview notes or survey responses in a structured way, writing down themes or phrases which occur again and again, noting responses which are unique (it may be that one person is very unhappy but everyone else is satisfied), and building up a picture of the collective experience of the respondents. There are rigorous approaches to qualitative coding and analysis which ensure that the interpretation of words is clear, however on a small scale it can be as efficient to simply read and consider the responses (if there aren’t too many of them). It is also possible to quantify this data by counting numbers of times people say certain things, but generally this is not so important since the experiences are personal and can’t be taken to be representative of all service users.

Once you have the data in a collective form (either statistics or experiences) it can be useful to sit down (with others, as possible) and consider what the data are telling you. Here’s a diagram which depicts an analysis process:

Image showing the steps involved in the analysis process. For detailed description of this image please refer to the descriptive link next to the image. D

It is important to check your analysis of the results and to ensure that possible factors that have led to a particular finding are not overlooked. This can be done by discussing the conclusions with a few of the people who provided the data.