Results Summary
What was the research about?
For research studies to be meaningful, people from many kinds of communities need to take part in deciding what types of research are important to them. Some communities, such as rural areas or areas where people of color live, are underrepresented in research. This means these people are often not invited to create ideas for research. Researchers want to find better ways to learn what types of research people from these groups think are important.
In this study, the research team tested two different ways to recruit people with long-term pain to come up with research ideas to improve living with pain. The team recruited people from two mostly Hispanic, rural counties in Texas. In one county, the team used snowball sampling. For snowball sampling, several people recruit other people they know to participate in the study. In the second county, the team used purposive sampling and convenience sampling. For purposive sampling, the team asks people who know a community well, such as business leaders or members of community groups, to help recruit. Convenience sampling is when the research team recruits people who are easy to reach.
The research team wanted to see which method was better at recruiting people to participate in meetings about research.
What were the results?
Compared with purposive plus convenience sampling, snowball sampling got more people to come to all the meetings. Snowball sampling also recruited more people who were Hispanic and more people who were disabled by long-term pain. In addition, the people recruited by snowball sampling came up with more ideas for research on long-term pain.
There was no difference between the two groups of people in their ratings of the importance of the ideas. There was also no difference in the number of people who agreed to attend the meetings.
Who was in the study?
The team recruited people ages 35 to 75 who lived in two rural counties in south Texas. Most of these people had long-term pain not caused by cancer. Others were taking care of people with long-term pain. Among the 55 people recruited by snowball sampling, 69 percent were women, and 87 percent were Hispanic. In addition, 16 percent spoke Spanish as their main language. The average age was 58. Among the 62 people recruited by purposive plus convenience sampling, 65 percent were women, and 73 percent were Hispanic. In addition, 11 percent spoke Spanish as their main language. The average age was 57.
What did the research team do?
The research team spent four months recruiting people for a series of three meetings. The team used snowball sampling in the first county and purposive sampling in the second county. When large businesses were not interested in helping in the second county, the team added convenience sampling.
People spoke either English or Spanish in the meetings. They learned about long-term pain and watched videos of people talking about their pain. They then shared their ideas for future research on services and programs to help people with long-term pain. Finally, they rated the importance of those ideas.
A group of community members with an interest in long-term pain led the study in each county.
What were the limits of the study?
The study focused on recruiting people from two rural communities with a large Hispanic population. Results may be different for other types of communities.
Future studies could test these methods in other underrepresented communities.
How can people use the results?
Snowball sampling may work better than purposive plus convenience sampling for recruiting Hispanic people living in rural areas to find out what types of research are important. Researchers can use these results when planning how to learn from people from communities that are underrepresented in research.
Professional Abstract
Objective
To compare the effectiveness of two sampling methods to recruit participants from underrepresented communities and engage them in identifying research priorities
Study Design
| Design Element | Description |
|---|---|
| Design | Empirical analysis |
| Data Sources and Data Sets | 55 research participants recruited through snowball sampling in Frio County, Texas; 62 research participants recruited through purposive and convenience sampling in Karnes County, Texas; |
| Analytic Approach | Quantitative and qualitative analysis |
| Outcomes | Number of participants consenting to meetings with research team, number of participants attending all meetings, demographics of research participants, productivity of participant feedback (number of ideas generated by research participants), quality of participant feedback (ratings of ideas for importance and feasibility) |
The research team compared two sampling methods to recruit participants from two predominantly Hispanic, rural counties in south Texas to identify and prioritize research ideas. In one county, the research team used snowball sampling, for which newly recruited participants invite others from their social network to participate. In the second county, the team used purposive sampling. With this method, people with in-depth knowledge about a population, such as key community groups and businesses, recruit participants. However, because larger employers did not wish to grant access to their employees, the team added convenience sampling, using a local businesswoman to help recruit people available to participate in the study.
The research team asked participants to attend three meetings. During these meetings, participants learned about long-term pain and watched videos of people talking about their pain. Participants also brainstormed and prioritized ideas for research to improve the lives of persons with chronic pain in their community. Participants completed a Likert-type scale ranking of each idea on importance and feasibility to implement.
The research team compared the number of participants who consented to attend the meetings, the number who attended all three meetings, and the percentage of Hispanic participants and participants with disabilities who attended meetings in each county. The team evaluated the productivity of participants recruited by each method by counting the number of unique ideas generated and assessed the quality of these ideas based on their relative ranking on importance and feasibility.
Community members were eligible to participate if they were ages 35–75 and had chronic pain that was not cancer related or were a caregiver for an individual with chronic pain. In the county using snowball sampling, 69% of participants were women, 87% were Hispanic, and 16% spoke Spanish as their primary language. The average age was 58. In the county using purposive and convenience sampling, 65% of participants were women, 73% were Hispanic, and 11% spoke Spanish as their primary language. The average age was 57.
Two community advisory boards that comprised residents from each county with an interest in chronic pain served leading roles in conducting and evaluating the study.
Results
Compared with the county using purposive plus convenience sampling, the county that used snowball sampling
- Recruited more participants (65% versus 42%, p<0.05) who attended all three meetings
- Recruited more Hispanic participants (87% versus 73%, p<0.05) and more participants who had disabilities (47% versus 10%, p<0.001)
- Generated more unique ideas (115 versus 101, p<0.05)
The counties did not differ in the number of participants who consented to attend the meetings or in the average rating of ideas generated at these meetings on importance and feasibility.
Limitations
The study recruited Hispanic participants from rural areas. These methods may produce different results in other communities that are underrepresented in research.
Conclusions and Relevance
The research team found that snowball sampling may perform better than purposive plus convenience sampling for recruiting participants from predominantly Hispanic, rural areas to generate ideas for future research. However, people recruited by the two sampling methods produced generally similar ideas for research in terms of importance and feasibility.
Future Research Needs
Future studies could test these methods in other underrepresented communities.
Final Research Report
View this project's final research report.
Journal Citations
Related Journal Citations
Peer-Review Summary
Peer review of PCORI-funded research helps make sure the report presents complete, balanced, and useful information about the research. It also assesses how the project addressed PCORI’s Methodology Standards. During peer review, experts read a draft report of the research and provide comments about the report. These experts may include a scientist focused on the research topic, a specialist in research methods, a patient or caregiver, and a healthcare professional. These reviewers cannot have conflicts of interest with the study.
The peer reviewers point out where the draft report may need revision. For example, they may suggest ways to improve descriptions of the conduct of the study or to clarify the connection between results and conclusions. Sometimes, awardees revise their draft reports twice or more to address all of the reviewers’ comments.
Peer reviewers commented and the researchers made changes or provided responses. Those comments and responses included the following:
- Reviewers asked the researchers to qualify their findings, as the study was more exploratory than hypothesis testing. The researchers did so, noting in their report the need for future research to confirm their findings.
- The reviewers asked for additional justification for conducting statistical comparisons on the two groups of research participants. The researchers responded that although they had not randomly assigned members to the groups, they still had an interest in exploring the differences between the groups. Thus, the investigators conducted the appropriate statistical tests.
- The reviewers asked for additional discussion of potential reasons why Spanish-speaking participants were more likely to drop out of the study than English-speaking participants. The researchers expanded their discussion and noted that the research team had engaged with the Spanish-speaking community for many years. They believed that the main reasons for this difference in study completion were employment and logistical barriers.
- Reviewers questioned the apparent lack of focus on known cultural and community factors in understanding participant engagement. The researchers explained that rather than assuming cultural and community barriers to engagement based on past literature on rural and Spanish-speaking populations, they gathered this information from their research participants and community advisors. The investigators incorporated what they learned into their understanding of participant behaviors.