How to do your own experiments in psychology and education
5.2 Plan how to work with your Participants
As a psychologist, you will usually study people. Who will participate in your experiment? How many people will you need? What characteristics do they need to have? Why did you choose these people?
Can you do anything you want with and to your participants, in the name of science? Obviously not. Your participants have rights that you need to respect. Another very important goal of experimental design, then, is this:
Ethical conduct on the part of the researcher is fundamental for successful research. A summary of guidelines for Ethical research is included below. Read it very carefully.
At this point, you have to do three things:
There are two main components of ethical experimentation that refer to participants: minimizing risk and maximizing informed consent.
Risk management. Under no circumstances should you expose participants to any kind of risk. This includes risk of physical, cognitive or emotional harm, even risk of embarrassment.
NOTE: For Research Methods, you can only choose a research problem that will NOT CREATE ANY POSSIBLE RISKS AT ALL for either you, the experimenter, or the participants who you recruit.
That means no experimental surgery in Research Methods; no drugs; no use of dangerous, vulnerable, or at-risk populations; no dangerous tasks or settings. Especially with your first experiments, you should not do anything risky!
Some kinds of research, such as medical research, involve unavoidable risks to the participants. In these cases, the researchers need to take especially meticulous care that the participants understand all of the risks and benefits very clearly and that they agree to participate in the experiment even so.
Risk management is why you need permission to carry out any research that involves human participants. There are two kinds of permission here: permission to do the study (from your Research Methods professor and from the University) and permission from the sources (institution and/or family) of the participants as well as the participants themselves.
Permission from the University
Finally, the main reason for all of this is that those are the rules: ethical research demands risk management and informed consent.
Anonymity. When you take a look at the consent form in Appendix A, you’ll see that participants are guaranteed anonymity.
Never, ever, ever mention the name of a participant to anyone, much less in print or in public.
There are at least two reasons for this: one is that participants may feel embarrassed or upset by other people seeing their performance. The other is that if the data is anonymous, then the data analysis won’t be influenced by other information about the participant. For example, “This is Stevie’s data. He’s smart; it must all be correct.” This kind of thinking leads to analysis errors. It’s harder to be influenced by “participant #20” . Finally, we know that someone’s name doesn’t systematically affect their behavior on most psychological tasks. It’s irrelevant information that we simply don’t need.
The easiest solution is simply: Don’t ask for the participants’ names.
Informed consent. For this Research Methods course, you must work with normal adults, usually university students. For normal adults, you should get them to sign a consent form. This is called “informed consent” to participate: participants have to be informed enough to give their consent with a clear idea of the risks and benefits of the experiment.
There is a sample consent form in Appendix A. This is the version to use if you do NOT have IRB permission. Note that the material in brackets [like this] is what has to be changed for each experiment. You need two copies for each of your participants (you keep one; they keep one), plus a few extras just in case.
Informed consent means that the participants have to understand clearly:
Notice that there is no mention of explaining the specific goals of the experiment. In fact, it’s very important NOT to tell the participants exactly what you are looking for until after you have finished collecting data. You can give them some more information right after they participate but remember that they might communicate with your next batch of participants and that will often be enough to invalidate your results. General answers will usually do. You can offer to answer any detailed questions at all about the study “next week” (after you’ve finished collecting all of your data), for example.
Also, you cannot deceive your participants, i.e., lie to them. Deception is unethical. Vagueness about the specific research goals and questions is often necessary, though. If your participants feel the need for more details, you can tell them more about the tasks, but avoid giving them specific information about your goals until after you collect your data.
For schools and other institutions, rather than just answer the individual participants’ questions, it’s much better to return later and present your study and results (in non-technical terms!). That is the best way of showing your appreciation for their collaboration. It also establishes a better relationship with them in case you need to collect more data later on (because most researchers disappear and never come back!).
Read this topic next: Define and find your participants
Can you do anything you want with and to your participants, in the name of science? Obviously not. Your participants have rights that you need to respect. Another very important goal of experimental design, then, is this:
| Use maximally ethical procedures. Ensure that everything possible has been done to protect all of your participants from risk and to guarantee them both informed consent and full anonymity. |
At this point, you have to do three things:
| a Define and find your participants |
b Choose a sample of the possible participants |
c Assign the participants to your experimental conditions |
| You need to decide who you want to study, how many people you need, and then you need to start to find sources of participants. Plan ahead to be sure to deal with your participants ethically. | When you do an experiment, because of practical limitations you will work with a sample of the population of interest. For example, you might study 40 college students (the sample) because you’re interesting in how adults in general (the population) perform some task. | To make generalizations more reliable, researchers use random assignment (of participants to experimental conditions). The use a table of random numbers to decide which people will participate in which experimental conditions. |
Ethics in dealing with Participants
The American Psychological Association has detailed norms on ethical behavior of both clinical and research psychologists. Review the Association’s Ethical Principles of Psychologists and Code of Conduct (APA, 2002) so that your behavior is as professional as possible.There are two main components of ethical experimentation that refer to participants: minimizing risk and maximizing informed consent.
Risk management. Under no circumstances should you expose participants to any kind of risk. This includes risk of physical, cognitive or emotional harm, even risk of embarrassment.
NOTE: For Research Methods, you can only choose a research problem that will NOT CREATE ANY POSSIBLE RISKS AT ALL for either you, the experimenter, or the participants who you recruit.
That means no experimental surgery in Research Methods; no drugs; no use of dangerous, vulnerable, or at-risk populations; no dangerous tasks or settings. Especially with your first experiments, you should not do anything risky!
Some kinds of research, such as medical research, involve unavoidable risks to the participants. In these cases, the researchers need to take especially meticulous care that the participants understand all of the risks and benefits very clearly and that they agree to participate in the experiment even so.
Risk management is why you need permission to carry out any research that involves human participants. There are two kinds of permission here: permission to do the study (from your Research Methods professor and from the University) and permission from the sources (institution and/or family) of the participants as well as the participants themselves.
Permission from the University
- University policy dictates that research for publication which uses human participants must first be reviewed and approved by the Institutional Review Board. Remember that if anything goes wrong in an experiment, then the University is in trouble, too.
- In the case of Research Methods courses, the professor plays the role of the Review Board for research that will not be published.
- For this Research Methods course, you will be working with normal adults, usually university students. For normal adults, you should get them to sign a consent form. [See Appendix A for a sample consent form.] So, you won’t be dealing with sources of participants like schools and clinics.
- If you ever do work with special populations of any kind (i.e., not normal adults), get permission from everyone possible: the participant’s family, the teacher, the school / clinic / hospital, the participant, etc. Note that other institutions like schools, clinics, and hospitals also have Institutional Review Boards, so this process can take a long time. Policies on permission for research vary widely, so start your discussions early.
- Do NOT accept your favorite teacher’s friendly invitation to test on her elementary or high-school class without getting permission from the school, too. All it takes is one talkative kid and one ill-informed parent for accusations to start flying and then everyone is in trouble. Even with permissions in place, all sorts of strange accusations happen sometimes.
Finally, the main reason for all of this is that those are the rules: ethical research demands risk management and informed consent.
Anonymity. When you take a look at the consent form in Appendix A, you’ll see that participants are guaranteed anonymity.
Never, ever, ever mention the name of a participant to anyone, much less in print or in public.
There are at least two reasons for this: one is that participants may feel embarrassed or upset by other people seeing their performance. The other is that if the data is anonymous, then the data analysis won’t be influenced by other information about the participant. For example, “This is Stevie’s data. He’s smart; it must all be correct.” This kind of thinking leads to analysis errors. It’s harder to be influenced by “participant #20” . Finally, we know that someone’s name doesn’t systematically affect their behavior on most psychological tasks. It’s irrelevant information that we simply don’t need.
The easiest solution is simply: Don’t ask for the participants’ names.
Informed consent. For this Research Methods course, you must work with normal adults, usually university students. For normal adults, you should get them to sign a consent form. This is called “informed consent” to participate: participants have to be informed enough to give their consent with a clear idea of the risks and benefits of the experiment.
There is a sample consent form in Appendix A. This is the version to use if you do NOT have IRB permission. Note that the material in brackets [like this] is what has to be changed for each experiment. You need two copies for each of your participants (you keep one; they keep one), plus a few extras just in case.
Informed consent means that the participants have to understand clearly:
- Potential benefits of the research
- Potential risks of the research for them individually
- How the researcher will protect their individual privacy
- That participation is voluntary and they can opt out of the experiment at any time
Notice that there is no mention of explaining the specific goals of the experiment. In fact, it’s very important NOT to tell the participants exactly what you are looking for until after you have finished collecting data. You can give them some more information right after they participate but remember that they might communicate with your next batch of participants and that will often be enough to invalidate your results. General answers will usually do. You can offer to answer any detailed questions at all about the study “next week” (after you’ve finished collecting all of your data), for example.
Also, you cannot deceive your participants, i.e., lie to them. Deception is unethical. Vagueness about the specific research goals and questions is often necessary, though. If your participants feel the need for more details, you can tell them more about the tasks, but avoid giving them specific information about your goals until after you collect your data.
For schools and other institutions, rather than just answer the individual participants’ questions, it’s much better to return later and present your study and results (in non-technical terms!). That is the best way of showing your appreciation for their collaboration. It also establishes a better relationship with them in case you need to collect more data later on (because most researchers disappear and never come back!).
Read this topic next: Define and find your participants
Replies to This Discussion
-
Permalink Reply by Benjamin Su on -
Please attach a sample consent form
-
Permalink Reply by Mike Dillinger, PhD on -
See below.
-
Permalink Reply by Mike Dillinger, PhD on
-
I've attached a sample Consent form for use when (as in our class) your did NOT get approval from the Institutional Review Board to do your experiment.
- Attachments:
-
-
ConsentForm_noIRB.doc, 30 KB
-
About
© 2009 Created by Mike Dillinger, PhD on Ning. Create a Ning Network!
