Common Data Types in Public Health Research

Data Types

Quantitative data is measurable, often used for comparisons, and involves counting of people, behaviors, conditions, or other discrete events (Wang, 2013).
Quantitative data uses numbers to determine the what, who, when, and where of health-related events (Wang, 2013).
Examples of quantitative data include: age, weight, temperature, or the number of people suffering from diabetes.

Qualitative data is a broad category of data that can include almost any non-numerical data.
Qualitative data uses words to describe a particular health-related event (Romano).
This data can be observed, but not measured.
Involves observing people in selected places and listening to discover how they feel and why they might feel that way (Wang, 2013).
Examples of qualitative data include: male/female, smoker/non-smoker, or questionnaire response (agree, disagree, neutral).
Example of qualitative data from a health care setting includes (Curry, Nembhard, & Bradley, 2009):
- Measuring organizational change.
- Measures of clinical leadership in implementing evidence-based guidelines.
- Patient perceptions of quality of care.

Primary data analysis in which the same individual or team of researchers designs, collects, and analyzes the data, for the purpose of answering a research question (Koziol & Arthur, nd).

You collect exactly the data elements that you need to answer your research question (Romano).
You can test an intervention, such as an experimental drug or an educational program, in the purest way (a double-blind randomized controlled trial (Romano).
You control the data collection process, so you can ensure data quality, minimize the number of missing values, and assess the reliability of your instruments (Romano).

Existing data collected for another purposes, that you use to answer your research question (Romano).

Large samples
Can provide population estimates : for example state data can be combined across states to get national estimates (Shaheen, Pan, & Mukherjee).
Less expensive to collect than primary data (Romano)
It takes less time to collect secondary data (Romano).
You may not need to worry about informed consent, human subjects restriction (Romano).

Study design and data collection already completed (Koziol & Arthur, nd).
Data may not facilitate particular research question o Information regarding study design and data collection procedures may be scarce.
Data may potentially lack depth (the greater the breadth the harder it is to measure any one construct in depth) (Koziol & Arthur, nd).
Certain fields or departments (e.g., experimental programs) may place less value on secondary data analysis (Koziol & Arthur, nd).
Often requires special techniques to analyze statistically the data.