Have you ever wondered why and how to collect data? You are not alone. MPact partners increasingly request technical assistance onData Collection Guidelines data collection and research, which we provide whenever possible as a part of collaborative projects we have with community partners around the world. Based on these collaborations, we have developed this guide as an outline for how to begin planning for data collection.
We at MPact collect and analyze data for the purpose of action. We provide evidence in order to influence:
Advocacy: amplify community voices
Programming: document best approaches
Policy: guide decisions | make it difficult to deny a problem
Funding: show funders where the need is greatest
What is data collection?
Are you or your organization interested in building an evidence base that can inform your advocacy, programming or policy work? It’s important to understand the fundamental idea behind data collection in order to begin this process.
Data are bits of information that represent experiences (e.g., knowledge, attitudes or behaviors) that are important for understanding, trends, patterns, or how things are related to each other.
Data collection in research is the practice of systematically (or consistently) collecting and measuring bits of information that are relevant for a topic of interest.
Data Analysis involves methodically examining patterns that can reveal answers to questions or clarify next steps in a broader category of investigation.
How to Begin Data Collection?
The more systematic and consistent the data collection is, the more confidently it can inform.
Data collection begins with an exploration of what it is you really want to know and why you want to know it. For example, let’s assume you want to improve health care access. The first step is to frame this as a question and then, specify, specify, specify. Below is an example of steps you would take toward refining your area of interest in improving health care access.
Area of interest: Improving health care access.
Frame it as question: How can we improve health care access?
Specify the focus (Selecting aspects of the question that you believe may contribute to answering the question): What are the barriers to and facilitators of health care access?
Specify the population or group/community of people with whom you want to learn about these barriers. For example: What are the barriers to and the facilitators of health care access among gay, bisexual and other men who sex with men?
Specify the location or where it is you want to learn about these barriers. For example: What are the barriers to and the facilitators of health care access among gay, bisexual and other men who sex with men in Kenya?
As you refine your question, the boundaries of your data collection become clearer. The more specificity you can provide in your question, the easier it will be to conduct your data collection. You would continue to specify based on what it is you really want to know as well as the feasibility of the data collection. For example, if you have limited time or resources, you might further specify your location.
Further specify and refine based on feasibility: What are the barriers to and the facilitators of health care access among gay, bisexual and other men who sex with men in Nairobi, Kenya?
What Kind of Data Collection is Best?
There are multiple kinds of data collection, such as observational (documenting what one sees), qualitative interviewing (recording narratives from representative members of a group), and quantitative or survey data (repeated questions across multiple members of a group whose answers can be counted).
Just as in construction a hammer is not superior to a screw driver, there is no superior kind of data collection. In construction what determines whether to use a hammer or screwdriver is the task at hand, i.e., whether you need to inset a nail or a screw. For data collection, your question and what is known about it help determine the kind of data collection you should use.
Both methods can be used in the same study with the same population. This is called multi-methods.
What is already known and what is not yet known?
Based on your experience, the literature, and existing data, you would review what is known about your question. Using the example on health care access, you might break down the question as shown in Table 1. If you know a lot already, you would begin with quantitative. If you know very little, you would want a deeper explanation and start with qualitative.
|Goal||Gain depth & breadth of participants’ subjective experiences
Understand the underlying reasons for behavior, thoughts and feelings
|Generalize findings from a sample of respondents to larger group|
|Assumptions||Complexity, dynamic, & little is known about the specific context, event, people or relationship||Fixed, measurable, & enough is known about the phenomenon as to be able to reduce the range of possible answers|
|Data Collection||Observation, focus groups, individual interviews||Computer or investigator-led survey, existing data|
|Analysis||Themes based on participant descriptions||Numerical frequencies, comparisons, tests of statistical significance|
|Dissemination||Findings presented using language of participant||Findings presented as statistical analyses|
|Strengths||Tells story from the participant’s viewpoint
Descriptive detail that explains a broader human context
|Quickly summarizes, organizes & compares a lot of information
Reveals general conclusions
Can measure progress & success
Identifies trends & magnitude
|Limitations||Potential to assume findings are true for larger group
Analyses and interpretation are labor-intensive
|Potential to simplify or lead to an inaccurate understanding of complex situations and realities
How to share findings?
Decide who is your audience, what is your goal, what story do you want to tell.
|Community Engagement||Share findings with community of interest; create space for community to respond to findings||Community events, presentations, discussions regarding findings|
|Advocacy||Amplify community voices & needs||Use evidence/findings to support recommendations that advance community needs in local, national and international meetings; reports; conferences|
|Programming||Document best approaches & inform development||Internal organizational briefings, strategic planning meetings|
|Policy||Guide decisions | make it difficult for policy makers to deny a need||Brief reports, policy meetings, conferences, national planning meetings|
|Funding||Show funders where the need is greatest||Graphs, Reports, PPT slides|
|Knowledge production||Contribute to a growing body of knowledge||Conference posters or presentations, peer review journal articles, all of the above|
For possible next steps and more resources on data collection, stay tuned for more from MPact!
This was developed by Sonya Arreola, Johnny Tohme, and Greg Tartaglione at MPact.