How to Recruit a Sample

Previously, you learned about the experiment in Psychology - how randomly assigning participants to different groups allows you to investigate the causes of behavior by manipulating an independent variable.  But how do you recruit these participants - these people who will take part in your experiment? 
Think Critically

Jared, Ali and Brent are three students in a U.S. high school, and they decide to carry out a research project on attitudes towards teen drinking at their school.  They want to find out if students at the school regard teen drinking as a dangerous activity, or harmless partying.  There are a total of 680 students in the school, but they decide to give out surveys to just 40 students in order to save time.  But how to choose which 40 students to survey?

  • Ali is on the football and basketball team, so suggests handing out the surveys to 40 of his teammates

  • Brent thinks they should hand out the surveys to students in the library after school

  • Jared suggests obtaining a list of student e-mails, and randomly selecting students to receive the survey by e-mail, until 40 surveys are returned

Can you think about a potential problem with each of these sampling methods?  Which idea do you think is best?



Target Population and Sample

To discuss which student has the best sampling idea, we need to first define some key terms.  The target population is the entire category of people you are interested in studying.  It could be as broad as all humans, but could also focus on a more specific group - for instance, children of divorced parents, teenagers who have been victims of cyberbullying, or elderly people living in nursing homes.  In the example above, the target population is simply the 680 students at the particular American high school that Jared, Ali and Brent attend.

In most cases, studying every person in the target population is too time consuming and impractical.  So, researchers select a smaller number of people to study - called the sample.  In the example above, Jared, Ali and Brent have chosen to select a sample of 40 students.

The key challenge with sampling is to make sure that your sample is representative of the target population.  In other words, the characteristics of the sample should be as similar as possible to the characteristics of the target population.  If, for example, the target population is about half male and half female, evenly divided between grades 9 to 12, and have a mix of different ethnicities, then the sample should reflect all of these characteristics.

If the sample is representative of the target population, then its reasonable to generalize the findings from the sample to the target population.  This means that the results from the sample can be applied to the entire target population.  So, for example, if most of the 40 students in a representative sample think that teen drinking is a bad idea, then that is probably the view of the school as a whole.
Try it Out

For each study, identify the target population

A.  An experiment to see if mindfulness exercises can increase the performance of long distance runners

B. An experiment to test whether playing memory games helps delay memory loss in people with Alzheimer's disease

C. An experiment to investigate the effects of violent cartoons on kindergarten students

D.  A study on whether wearing a perfume can influence how female job applicants are rated by males

E.  An investigation on the effects of sleep on memory

Scroll to the bottom of the page for the answers
Now that we've defined some key terms, let's go back and compare how Ali, Brent and Jared want to recruit a sample of high school students, and which method will give the most representative sample.

  • Ali wants to obtain a sample by handing out surveys to his teammates on the football and basketball teams.  However, this will result in a sample consisting entirely of high school athletes, who are not representative of the entire student population.  It's possible that high school athletes have different attitudes to drinking than non-athletes.  If U.S. teen stereotypes are to be believed, maybe high school atheletes are typical "jocks" who really like partying and binge drinking.  Or perhaps the opposite is true - maybe athletes would be more concerned about the effects of alcohol on their physical health.

  • Brent wants to obtain a sample by handing out surveys in the library after school.  But maybe the students who spend lots of time in the library are different from students who rarely venture in the library, and thus may have different attitudes to drinking.  Again, if popular stereotypes are to be believed, perhaps students who frequent the library are "nerds" who rarely go to parties or engage in risky behavior, and hence their views do not represent everyone in the school.

  • Jared suggests obtaining a sample by sending out e-mails to random students.  This is most likely to produce a  representative sample, as any student has an equal chance to be chosen.  Jared's sample will likely consist of a good cross-section of the student population, and hence can be generalized to the school as a whole.  But even Jared's sampling method is not perfect.  Suppose, for example, that many students don't reply to the e-mail that he sends out.  Its conceivable, therefore, that the students who do reply to the survey might be different in some respects from the students who don't reply, and so the attitudes of the students who do reply might not be representative of everyone.
Sampling Methods in Psych

Four common sampling methods used in Psychology are convenience sampling, self-selected sampling, snowball sampling, and random sampling.  Some sampling methods are relatively easy and simple to carry out, but may not result in a very representative sample, while other sampling methods are more complex, but produce a better sample.

Convenience sample (or opportunity sample) involves simply asking anyone who happens to be there if they would like to participate.  For instance, asking students in the library if they would be willing to complete a survey.

  • Advantages:  The quickest and easiest way to obtain participants

  • Limitations: Sample may not be representative, since the people who happen to be in that one location may not have all the characteristics of the target population.  Furthermore, convenience sampling can only be feasibly done for studies that can be completed in just a few minutes.

Self-selected sample (or volunteer sample) involves recruiting people who volunteer to participate in a study, often for payment.  Typically, posters will be placed on a university campus or in a public location, or classified ads will be placed online or in a newspaper. 

  • Advantages: An easy way to obtain participants.  Furthermore, participants have freely chosen to participate, so they will usually be motivated and co-operative.  There are no ethical issues since no pressure is being put on people to participate who may not want to.

  • Disadvantages: The people who choose to volunteer may be different from the people who don't volunteer, so the sample may not be representative of the target population.  For instance, in a study on teen drinking, perhaps many heavy drinkers would fear being judged or reported to school authorities, and hence would not volunteer and be included in the sample.

Snowball sampling is when participants are asked to refer other people they know (ex. family or friends) who would be willing to participate in the study.

  • Advantages: An easy way to otain participants, especially in studies which involve hard-to-reach populations.  For instance, if a researcher wants to study teens who belong to street gangs, it can be hard to earn the trust of gang members.  However, if a researcher succeeds in finding one willing participant in a street gang, that can help open the door to other participants.

  • Disadvantages: Participants will all know each other and come from the same social circle, which suggests that the sample may not represent everybody in the target population.  For instance, in a study on street gangs, if all the participants come from the same gang, they might not represent the views and behavior of teens in different gangs.  Also, since participants know each other, anonymity is not strictly maintained (the names of people who participate is not kept secret)

Random sampling is when every person in the target population has an equal chance of being sampled.  Typically, researchers will have the names of every person in the target population, and will select names at random.  This method is often used in telephone surveys before an election - every telephone number has an equal chance of being selected.

  • Advantages: This should result in a very representative sample, and so the findings from the sample can be generalized to the entire target population.

  • Disadvantages: The most expensive and time-consuming sampling method, and often not feasible to carry out.  Furthermore, there is no guarantee that the people you randomly choose will agree to participate.  If many people decline to participate, this can cause response bias - the people who agree to participate may be different in some way from the people who refuse.







Think Critically

Much of the research carried out in Psychology occurs on university campuses, with samples made up primarily of university students.  This raises a very important question:  are university students really a representative sample?  Would the results be different if other groups of people were studied?


  • According to the article, what sort of people tend to be studied in psychology experiments?  What is the target population for which the results are applied?

  • What does WEIRD stand for?  

  • What are some ways in which WEIRD samples think and behave differently from most other people in the world?

Scroll to the bottom of this page for answers!

 
Checklist

  • I can describe the sample and target population of a research study

  • I can evaluate whether the sample is representative of the target population

  • I can describe a variety of sampling methods, including opportunity, self-selected, random and snowball sampling

  • I can discuss strengths and limitations of sampling methods
Quiz Yourself

1.  The entire category of people you are interested in studying is the ___

(a) Sample

(b) Sampling method

(c) Target population

(d) Sample characteristics


2.  The key objective in sampling is to obtain a sample that is _____

(a) Randomly selected

(b) Representative of the target population

(c) Equally made up of males and females

(d) Consists of different nationalities


3.  Geno hands out surveys on consumer spending to passerby at a local shopping mall.  What sampling method is this?

(a) Random

(b) Opportunity

(c) Self-selected

(d) Snowball


4.  Which sampling method is most likely to be representative of the target population?

(a) Random

(b) Opportunity

(c) Self-selected

(d) Snowball


5.  A limitation of this sampling method is that volunteers may be different from non-volunteers.  Which sampling method is this?

(a) Random

(b) Oppurtunity

(c) Self-selected

(d) Snowball


6.  This sampling method is particularly useful for reaching hard-to-reach populations.  

(a) Random

(b) Opportunity

(c) Self-selected

(d) Snowball


7.  Which is NOT a way that WEIRD samples are different from most people in the world?

(a) WEIRD samples are mostly from highly individualistic countries

(b) WEIRD samples tend to be highly educated

(c) WEIRD samples tend to be aged 18-22

(d) WEIRD samples tend to be define themselves by their relationships






Answers


Try it Out

A. Target population = Long distance runners
B. Target population = People with Alzheimer's disease
C. Target population = Kindergarten students
D. Target population = Males
E. Target population = Humans

Think Critically

Nearly all of the people who are studied in psychology experiments are from Western countries, with a majority from the United States.  Of the American participants, most are undergraduate university students studying Psychology.  The results from these participants are then assumed to apply for all people.  The target population is all of humanity but the sample does not represent the diversity of humankind.

WEIRD stands for Western, Educated, Industrialized, Rich, Democratic

WEIRD participants are different from other people in a multitude of ways.  For instance, WEIRD participants tend to be prize individualism, independence, and personal choice.  They tend to define themselves in terms of personal traits (ex. "I'm outgoing and talkatative") rather than in terms of their relationships.