Public Health 3100: Quantitative Data Collection

Self Reports

An individual’s own account of factors such as attitudes, symptoms, beliefs, or behaviors. This type of data is relatively easy to gather through surveys or interviews, and is commonly used in public health research.

Cognitive Response Testing

Another way to gain insight into how respondents may perceive questions. Involves an in-depth interview with a small number of respondents who are similar to the target survey population. Researchers using this method of pretesting ask the participants for details on how they interpret the questions, whether the concepts make sense, and what factors determined their answers. Data can inform revisions needed, thus reducing the chance of measurement error. The content of the responses is less relevant than the participant’s perception of the question.

Open-ended Questions

Worded in such a way that the respondent can answer in phrases or sentences. These questions elicit a broad and unique set of responses. Is also helpful when researching new topics or ideas. However, these can be more time-consuming to use for the researcher. Information from responses has to be categorized or coded in order to analyze it quantitatively.

Closed-ended Questions

Those for which response options are provided. Data is typically easier to analyze because the response categories can easily be converted into numeric codes for quantitative analysis. However, the response options may not be exhaustive or limited to unique responses. Allows respondents to choose more than one answer or add in their response if it is not an option (such as “Other”—fill in the blank).

How can the percieved benefits for responding be increased?

•  Provide information about the survey and how results will be used to benefit participants or the population.   

•  Asking for help or assistance. People may feel a sense of reward for helping.   

•  Support group values by making the connection from the survey to the target respondent group.  

•  Give tangible rewards. Appropriate and timely incentives given in exchange for survey burden may increase perceived benefit. Look to past research, pilot testing, or stakeholder input to determine the most relevant incentive.   

•  Inform people that opportunities to respond are limited. This may increase the perceived value of responding.

How can the percieved costs for responding be decreased?

•  Make it convenient to respond. Many web-based survey platforms now offer mobile-friendly options so people can conveniently respond to a survey using their cell phones.   

•  Make the questionnaire short. Consider including only questions essential to your research focus.   

•  Minimize requests for sensitive information, and, if needed, emphasize confidentiality or anonymity in responses.  

•  Include a way to save survey responses, in case they want to return to the survey at a later time.

How can trust be established so that people believe the rewards will outweigh the costs of responding?

•  Obtain sponsorship from a legitimate authority. Carefully consider history and perceived trust between academic institutions and community when determining sponsorship.   

•  Ensure the confidentiality of the information being collected. Institutional review board certification with contact information can help boost faith in confidentiality.   

•  Identify the request for information as important to some outcome relevant to the respondent or population group.

Response Rate

The number of people who completed your survey divided by the number in the sample population.

What is The Behavioral Risk Factor Surveillance System (BRFSS)

One of the largest telephone survey systems in the United States.

Panel Research

Online survey companies merge commercial lists of potential respondents to develop a sample frame, which can then be narrowed to match characteristics of the target research population (e.g., women age 50–65 years). These frames can be useful if the research study includes multiple data collection from the same people.

The Health and Retirement Study

An example of panel research. The sample consists of 20,000 Americans over the age of 50, and the study collects data every 2 years on topics such as income, work, health, and healthcare.

Objective Measurements

Recording empirical data, i.e., weight on a scale. Reduces the bias from the subjectivity inherent in self-reported data.

Observations

Another way of collecting objective data. Researchers can document people, places, or situations directly. For example, rather than relying on self-reports of neighborhood resources available, a researcher can visit a community and record the quantity and quality of those resources.

Systematic Observation

A way to ensure accuracy in reporting what is observed. Developing a structure for methodical observation reduces the chance of observer bias.

Crowdsourcing

Uses input and data from a large number of people with the help of technology. This method is growing in popularity in public health surveillance and has been used in the areas of tobacco control, physical activity, and environmental health. Because it is an emerging way to collect data, more evidence is needed on the rigor and validity of its use.

Secondary Data

Data already collected for surveillance or from a previous study, is a viable option for exploring many public health topics. The U.S. Census and the CDC’s BRFSS are two common public health data sources.