Data Collection Methodologies for Health Research Projects

Vicki Lawrence

Health research is commonly based on a priori knowledge about particular conditions or diseases, and the factors or measures that predict the presence of these conditions or diseases. When a researcher decides that he or she wants to investigate relationships of this nature, careful consideration about the variables the researcher will choose to investigate become of substantial importance, as do the methods that an individual chooses to use for collecting and analyzing that information. I wanted to provide some general guidelines for issues a researcher should consider when (1) collecting data and (2) developing an analytic plan.

1. A. Data Collection
2. 1. Data-Do You Need to Collect It?

There is a wealth of data available- much of which has been collected through large surveys and medical studies, but has yet to be analyzed. Before deciding that you must collect your own data, consider:

1. 1. Is the information you want already available in other datasets that have yet to be analyzed?
2. 2. Are you planning to measure the health outcome or predictor measures in a way that has not been done before?
3. 3. Do you have the funds to collect your data, or would it be more reasonable to do secondary data analyses?

Collecting data is an expensive endeavor, particularly for students who typically have limited funds and time limits for the projects they choose to undertake. If the measures you want to assess (including the health outcome and the potential predictors of that outcome) are available in an existing data set, secondary data analyses may be more appropriate for you.

1. 2. Data Collection Decisions

If you decide there is a need to collect your data, carefully consider what can influence the data you can collect. Any type of data collection, whether quantitative, qualitative, or a combination of both, require significant planning and detail. Consider the following:

1. 1. What patients or target population do you want to examine?
2. 2. Do you have a questionnaire that you want to deliver? Have the questions been validated in your target population?
3. 3. How will you establish your sample from your target population? Will this require sophisticated sampling techniques that you need help with?
4. 4. Have other studies examined the reliability and validity of the measures you want to use?
5. 5. Are you proposing a new measure of a health outcome that should be compared to other previously established measures?
6. 6. Will you have multiple interviewers/data collectors? How will you measure inter-rater and intra-rater reliability?
7. 7. Have you received institutional review board approval for the study you want to do?
8. 8. Have you thought about what you will do if the data collection does not go as planned (What will you do if you are unable to get enough participants within your expected time frame? What will you do if many subjects drop out of your study)?

These types of decisions can strongly impact not only whether you are able to collect data, but the quality of any data you are able to collect as well. Although collecting data in itself may not be the most difficult endeavor you may face, having HIGH QUALITY data that allows you to make inferences to the target population of interest with respect to the health outcome you study is very difficult, and requires substantial planning.