Evaluation Toolkit for Breastfeeding Programs and Projects

June 2012

3.7 What information should be collected?

Page last updated: 04 November 2013

Data can be quantitative (that is, something that can be counted) or qualitative (that is, something that describes a quality or experience). Typically, quantitative data refers to numbers such as clinical measurements or population statistics, while qualitative data refers to word-based data such as people’s interview responses or conversations in group meetings.

It is important that data is collected because it will answer the evaluation question, and not just because it is available. There may be a range of data available to answer the question you are asking, for instance exclusive breastfeeding duration may be recorded in the clinical notes (quantitative), or you could ask the mother whether she has any concerns about feeding (qualitative).

Whether it is quantitative or qualitative, there are another two categories of data, primary and secondary.

Primary data is collected specifically for the evaluation itself, such as surveys of clients or staff, extracting data from clinical records, clinical audit data, interview notes, or data recorded by staff while undertaking their usual tasks (for instance, recording waiting times in a reception area, or mapping the journeys of clients through the service).

Two of the biggest challenges for a service in conducting an evaluation are 1) ensuring that data collection is not a burden on service staff, and 2) ensuring that the right data is collected to answer the evaluation questions. Nothing will discourage people from doing evaluation like finding that the data they have collected doesn’t actually tell them what they want to know.

Secondary data is already collected and available through external sources, such as data published in peer-reviewed literature, national data sets such as birth statistics, the Australian Childhood Immunisation Register, census data, survey results such as the 2010 Australian National Infant Feeding Survey Indicator Results or state/territory level surveys, or routine data collected through states and territories, such as perinatal data collections. This can include both published data, as well as additional unpublished data or access to datasets that may be available by contacting the organisation that collected the data or a repository such as the Australian Data Archive which holds the full data from the 2010 Australian National Infant Feeding Survey.

Other questions to consider when thinking about data are: who are the likely contributors to your data? Are these recipients of breastfeeding services, program staff or stakeholders in other organisations? It may be that a number of stakeholder groups are involved in contributing to the evaluation data.

Note that primary and secondary data may be quantitative (for example, numbers or rates of women breastfeeding) or qualitative (for example, levels of satisfaction with a service, or perceptions of the quality of training received).

Consider whether the data sought will be informing a process evaluation (was the program effectively delivered?) or an outcomes evaluation (how did the program change breastfeeding initiation or duration?), as this will assist in deciding the data sources and the methods of data collection. Either qualitative or quantitative methods may be chosen, or both, depending upon the aims of the evaluation and the available resources.