What You Need to Know About Sampling in Research

When reading a research study, you should always look at the sample. Samples can range from clinical populations, such as people who are in treatment for depression, to college sophomores. A lot of academic research is conducted using undergraduate students (I'm guilty of this) because it's a convenient way for professors to get their research done without external funding.

The type and size of sample is important to look at particularly if the research is a survey. It's a little less important with experiments because they involve a manipulation and a control group (more on this later).

Why the Sample Is Very Important in Survey Research

If the study is a survey and the sample is college students, the research results might only reflect the population of college students--or even just college students in that geographical area. For example, a survey might look at the attitudes about gender identity.

If the researchers only gave their questionnaire to college students, can you say that their results can be generalized to the U.S. population? Not really. Most people in the United States don't have a college degree, and attitudes when you're in your late teens/early 20s are WAY different from attitudes in your 40s and 50s.

Why Demographics Are Important in Sampling

The researchers should also give the demographics of that sample. How does that affect the population at large? For example, if I have a survey asking whose celebrity workout you'd like to do (ok, that's a pretty stupid survey), I should also report how old and what is the gender makeup of my sample. If 74% of my survey respondents say they'd follow Mark Wahlberg's workout plan, I should also note that 65% of my respondents are male and in their late 20s. If a research report doesn't give a breakdown of the demographics of their survey respondents, throw out that report. It's pseudo-science.

Repeat after me: Discard any survey results that don't report the demographics.

How Researchers Recruited Their Sample is Also Important

This is also an interesting one that is similar to using a convenience sample of college students. Good researchers will not only know the different sampling methods (e.g. random, stratified, cluster, snowball), but they will also report their sampling method.

Clinical researchers will recruit participants for their study mostly by asking volunteers who are already involved in a particular domain. In other words, people who are receiving treatment for cancer or other conditions might be asked to participate in a research study. Sometimes they are paid, sometimes not. Researchers will disclose this.

Other researchers will solicit respondents through an ad or email campaign. This also should be noted. If someone works for a company uses that company's email and/or website to recruit participants, the sample derived will be biased for that company. In other words, if I'm doing a survey on what people's perceptions about Twitter, it wouldn't make much sense for me to advertise for the survey on Twitter.

This also goes for much of the political research. If I'm a political consultant trying to get people's attitudes towards a candidate or an issue, would it make sense for me to only find participants from that politician's email campaign? Sure, you might have very favorable attitudes for that candidate (which might stroke his ego), but it gives an inaccurate view of how the population feels about that candidate.

That's why people are very misled about so much political "research." The results don't very accurately reflect the entirety of the population. More will come later on the survey method and how online and phone surveys really can be wrong.

What Is Sample Size and Why Is It Important?

When conducting research, it is also important to have a sufficient sample size.

A small sample size can produce inconclusive results, while a large sample size produces more accurate results. The precision of a study is also influenced by the sample size. A large sample size yields a more precise estimate than a small sample size.

The power of a study is determined by the ability to detect an effect, if one exists. A large sample size increases the power of a study.

Anytime you come across a poll--particularly a political poll--you should always ask about the sample size. That's why you usually see the plus-minus margin of error. Sampling can really affect the margin of error in research studies and in polling. In other words, can you say that 400 people polled in Nebraska represents the United States population as a whole--particularly when it comes to politics? However, 400 people from different states might give a better picture.

Does Sample Size Matter in Qualitative Research?

You might be wondering whether sample size matters in qualitative research. The short answer is yes, it does, as the size of your sample can affect the data you collect and how accurately it reflects the overall population. A large sample size allows researchers to identify patterns that could not be found with a smaller sample, and it reduces the risk of bias by including a wider range of perspectives and opinions.

However, it's important to remember that there are other factors to consider when assessing the validity of qualitative research studies. While sample size is important, researchers should also assess variables such as diversity, data saturation, and participant engagement for a thorough understanding of their findings.

Conclusion

In research, sample size is the number of participants in a study. The size of a sample affects the validity of research findings. A sample that is too small is not representative of the population and may produce findings that are not generalizable. A sample that is too large is not efficient and may be difficult to manage. The optimal sample size depends on the type of research question being asked and the resources available.