8.2: Initial Equivalence

Learning Objectives

1. Recognize examples of studies that are experiments and studies that are not experiments.
2. Explain how random assignment and initial equivlance of groups leads to the ability to show cause-and-effect relationships.

What Is an Experiment?

A true experiment is a type of study designed specifically to answer the question of whether there is a causal relationship between two variables. In other words, whether changes in one variable (the independent variable) cause a change in another variable (the dependent variable). Experiments have two fundamental features: Initial equivalence and ongoing equivalence.

Initial Equivalence

As you might guess, initial equivalence means that groups start (are initially) similar (equivalent). What this means in practical terms is usually that the researcher has control over the IV, and randomly assigns participants into the different IV groups; these different groups, or levels of the independent variable, are called conditions. For example, in Darley and Latané’s (1968) experiment, the independent variable was the number of witnesses that participants believed to be present. Darley and Latané (1968) manipulated this independent variable by telling participants that there were either one, two, or five other students involved in the discussion, thereby creating three conditions. For a new researcher, it is easy to confuse these terms by believing there are three independent variables in this situation: one, two, or five students involved in the discussion, but there is actually only one independent variable (number of witnesses) with three different levels or conditions (one, two or five students).

While Darley and Latané (1968) did not use the phrase "random assignment," the data provided and their interpretation of the results suggest that suggest that they used random assignment to put participants into the IV conditions. Random assignment is putting participants in the different IV conditions randomly. This can be done through any computerized randomizer, or flipping a coin, rolling a die, picking a condition out of a hat, or any method that randomly distributes participants into the conditions. If an assignment method probably leads to higher proportions of certain kinds of people in any one condition, then it is not random. For example, you could alternate assigning participants into conditions as they enter such that the first person to arrive for the study is assigned into the one-person group, the second person is assigned into the two-person group, the third person is assigned into the five-person group, the fourth person to arrive is assigned to the one-person group, and so on. What wouldn't be random is assigning the first 10 people to arrive for the study are assigned to the one-person group, the next 10 to arrive to the two-person group,a nd the last 10 people to arrive to the five-person group because there may be something different about the type of people who arrive early compared to the type of people to arrive late.

There are two things to understand about random assignment and initial equivalence. First, random assignment does not ensure that all relevant characteristics are randomly distributed across the IV conditions. Randomization relies on probability, which is, by definition, not a sure thing. While it's a great method to try to control for variables that are not the IV and might affect the DV, we could always somehow still end up with a bunch of really helpful people in the one-person condition. That's why replication of research findings is so important.

The second important idea to note is that there may be other ways to reach initial equivalence other than random assignment. You could use the same participants in all IV conditions or use pretest scores; this would results in initial equivalence between the two groups if you assume that the participants are similar to themselves at the beginning of each part of the study. For example, you could have participants complete a survey on digital stress at the beginning of the study, then limit participants' access to social media, then have the participants re-take the survey on digital stress. Analysis would compare each individuals' first score (pretest) to their second score (posttest) to see if there were any changes in digital stress. If nothing else changed in the participants' lives, we could be fairly confident that it was the social media restrictions and nothing else that could have reduced their digital stress.

The thing to remember is that we want initial equivalence so that we can argue that any changes in the DV must be due to the IV since everything else started out similar between the IV conditions. Random assignment is the most trusted and most commonly used method to try to attain initially equivalence IV groups.

Random Assignment ^[1]

The primary way that researchers attempt to reach initial equivalence is through random assignment, which means using a random process to decide which participants are tested in which conditions. Do not confuse random assignment with random sampling (sometimes called random selection). Random sampling, discussed in the chapter on the scientific method, is a method for selecting a sample from a population based on probability to represent the population. Random assignment is a method for assigning participants in a sample to the different IV conditions.

In its strictest sense, random assignment should meet two criteria. One is that each participant has an equal chance of being assigned to each condition (e.g., a 50% chance of being assigned to each of two conditions). The second is that each participant is assigned to a condition independently of other participants. Thus one way to assign participants to two conditions would be to flip a coin for each one. If the coin lands heads, the participant is assigned to Condition A, and if it lands tails, the participant is assigned to Condition B. For three conditions, one could use a computer to generate a random integer from 1 to 3 for each participant. If the integer is 1, the participant is assigned to Condition A; if it is 2, the participant is assigned to Condition B; and if it is 3, the participant is assigned to Condition C. When the procedure is computerized, the computer program often handles the random assignment.

One problem with coin flipping and other strict procedures for random assignment is that they are likely to result in unequal sample sizes in the different conditions. Unequal sample sizes are generally not a serious problem, and you should never throw away data you have already collected to achieve equal sample sizes. However, for a fixed number of participants, it is statistically most efficient to divide them into equal-sized groups. It is standard practice, therefore, to use a kind of modified random assignment that keeps the number of participants in each group as similar as possible. One approach is block randomization.

Random assignment is not guarantee of initial equivalence. The process is random, so it is always possible that just by chance, the participants in one condition might turn out to be substantially older, less tired, more motivated, or less depressed on average than the participants in another condition. However, there are some reasons that this possibility is not a major concern. One is that random assignment works better than one might expect, especially for large samples. Another is that the inferential statistics that researchers use to decide whether a difference between groups reflects a difference in the population takes the “fallibility” of random assignment into account. Yet another reason is that even if random assignment does result in a confounding variable and therefore produces misleading results, this confound is likely to be detected when the experiment is replicated. The upshot is that random assignment to conditions, although not infallible, is usually considered a strength of a research designs that attempt to show cause-and-effect relationships.

Manipulation of the Independent Variable

Again, to manipulate an independent variable means to change its level systematically so that different groups of participants are exposed to different levels of that variable, or the same group of participants is exposed to different levels at different times. For example, to see whether limiting social media affect digital stress, a researcher might instruct some participants to limit their social media use to less than 1 hour a day and others to limit their social media use to 1-2 hours a day. The different levels of the independent variable are referred to as conditions, and researchers often give the conditions short descriptive names to make it easy to talk and write about them. In this case, the conditions might be called the “1 hour a day” and the “2 hours a day.”

Notice that the manipulation of an independent variable must involve the active intervention of the researcher. This is why (random) assignment to reach initial equivalence is so connected to the researcher being the one who is responsible for experiencing the IV conditions.

Comparing groups of people who differ on the independent variable before the study begins is not the same as manipulating that variable. For example, a researcher who compares the digital stress of people who already spend less than an hour a day of social media with the digital stress of people who spend 1-2 hours on social media has not manipulated this variable and therefore has not conducted an experiment; this is a natural groups design, which will be discussed in more detail next. This distinction is important because groups that already differ in one way at the beginning of a study are likely to differ in other ways too. For example, people who choose to already spend less than an hour a day on social media might not respond as strongly to fear of missing out or other indicators of digital stress compared to people who spend 1-2 hours a day on social media. Therefore, any observed difference between the two groups in terms of their digital stress might have been caused by the time spend on social media, or it might have been caused by any of the other differences between people who spend different amounts of time on social media. Thus the active manipulation of the independent variable is crucial for eliminating potential alternative explanations for the results. Random assignment is the method that researchers most often use to reach initial equivalence; with initial equivalence between the groups, there are no other alternative explanations for why the DV changed other than the IV condition.

Of course, there are many situations in which the independent variable cannot be manipulated for practical or ethical reasons and therefore an experiment is not possible. For example, whether or not people have a significant early illness experience cannot be manipulated, making it impossible to conduct an experiment on the effect of early illness experiences on the development of hypochondriasis. This caveat does not mean it is impossible to study the relationship between early illness experiences and hypochondriasis—only that it must be done using nonexperimental approaches. We will discuss this type of methodology in detail later in the book.

The "Other" IV Condition

In these kinds of experiments, there are often at least two groups to compare. The group that the researcher is most interested in is the one that they think causes changes in the DV. This group is often called the treatment group or an intervention group. To determine whether an intervention works, though, there needs to be at least one other group that is initially equivalent to the intervention group as a comparison. This is the "other" IV condition, the one that's not the one being tested. Students new to research first think about the control group (or control condition), in which participants do not receive the treatment. If participants in the intervention condition end up with statistically significantly different scores on the DV than participants in the control condition, then the researcher can conclude that the intervention works. An alternative approach is to use a wait-list control condition, in which participants are told that they will receive the treatment but must wait until the participants in the treatment condition have already received it. This disclosure allows researchers to compare participants who have received the treatment with similar participants who are not currently receiving it but who still expect to improve (eventually). If participants were randomly assigned to the intervention condition or the wait-list control condition, then the two groups would (probably) have initial equivalence.

But does it really make sense to compare the intervention group with a group who received nothing at all? If a psychiatrist thinks that a new drug would work better for their clients, would it really make sense to compare that to clients who aren't taking any medication? No! We would want to compare the mental health outcome (DV) with the new IV condition (new drug) and an established treatment (commonly-used drug). In these studies, participants are randomly assigned into the intervention condition and to another treatment condition. The other treatment conditions could be a placebo condition (which is really not a treatment, but the participants think that it is) or a different treatment (comparison group). In research on the effectiveness of psychotherapies and medical treatments, this type of experiment is often called a randomized clinical trial (or RCT), and is often considered the gold standard of research.

Placebo

Let's discuss the comparison with placebos first. A placebo is a simulated treatment that lacks any active ingredient or element that should make it effective, and a placebo effect is changes in the DV based on such a treatment (compared to no real or fake treatment at all). In medical studies, participants would receive a placebo that looks much like the treatment but lacks the active ingredient or element thought to be responsible for the treatment’s effectiveness. When participants in a treatment condition take a pill, for example, then those in a placebo control condition would take an identical-looking pill that lacks the active ingredient in the treatment (a “sugar pill”). In research on psychotherapy effectiveness, the placebo might involve going to a psychotherapist and talking in an unstructured way about one’s problems.

Figure \(\PageIndex{1}\): Hypothetical Results From a Study Including Treatment, No-Treatment, and Placebo Conditions

Although placebo effects are not well understood, they are probably driven primarily by people’s expectations that they will improve. Having the expectation to improve can result in reduced stress, anxiety, and depression, which can alter perceptions and even improve immune system functioning (Price et al., 2008). Placebo effects are interesting in their own right (see “The Powerful Placebo” box below or Psychology Today's page on placebos), but they also pose a serious problem for researchers who want to determine whether a treatment works. Figure \(\PageIndex{1}\) shows some hypothetical results in which participants in a treatment condition improved more on average than participants in a control condition. If these conditions (the two leftmost bars in Figure \(\PageIndex{1}\)) were the only conditions in this experiment, however, one could not conclude that the treatment worked. It could be instead that participants in the treatment group improved more because they expected to improve, while those in the control condition did not. If participants in both the treatment and the placebo control groups expect to improve, then any improvement in the treatment group over and above that in the placebo control group must have been caused by the treatment and not by participants’ expectations. This difference is what is shown by a comparison of the two outer bars in Figure \(\PageIndex{1}\)).

The Powerful Placebo

Many people are not surprised that placebos can have a positive effect on disorders that seem fundamentally psychological, including depression, anxiety, and insomnia. However, placebos can also have a positive effect on disorders that most people think of as fundamentally physiological. These include asthma, ulcers, and warts (Shapiro & Shapiro, 1999). There is even evidence that placebo surgery—also called “sham surgery”—can be as effective as actual surgery.

Medical researcher J. Bruce Moseley and his colleagues conducted a study on the effectiveness of two arthroscopic surgery procedures for osteoarthritis of the knee (Moseley et al., 2002). The control participants in this study were prepped for surgery, received a tranquilizer, and even received three small incisions in their knees. But they did not receive the actual arthroscopic surgical procedure. Note that the IRB would have carefully considered the use of deception in this case and judged that the benefits of using it outweighed the risks and that there was no other way to answer the research question (about the effectiveness of a placebo procedure) without it. The surprising result was that all participants improved in terms of both knee pain and function, and the sham surgery group improved just as much as the treatment groups. According to the researchers, “This study provides strong evidence that arthroscopic lavage with or without débridement [the surgical procedures used] is not better than and appears to be equivalent to a placebo procedure in improving knee pain and self-reported function” (p. 85).

To learn more about this and other studies on the placebo effect, check out this Hidden Brain podcast episode (Vedantam, 2019) on your favorite podcast app.

Comparison Groups

An additional option to compare the target IV condition with is a comparison group. These IV conditions are usually created by randomly assigning the participants into the currently used treatment. For a psychiatrist, this could be continuing to provide the same drug that you've been prescribing to the comparison group. Some call these control groups, which is correct in the sense that the IV condition is not receiving the target intervention. However, it is more clear to identify whether the group is not receiving anything (control group) or is receiving something but not what we're testing (comparison group).

The use of comparison groups is helpful to test whether the cost and effort of any intervention is worth it. Let's go back to our psychiatrist. Imagine that the cost of the current drug is pretty low, but the cost of the new, comparison treatment is pretty high. We want clients to receive the most effective treatment, but a comparison of the effectiveness of the new drug with the old drug may show that the new drug is equally effective. In this case, there's no benefit to switching to the new, higher cost treatment. If the psychiatrist compared the new drug to a wait-list control group, they would not have this important information about the comparative effectiveness of the new treatment.

Final Considerations of Comparisons

You may have figured it out, but it is good to make it clear: You can combine all of these control and comparison groups in one study! A psychiatrist could compare participants who:

• Receive a new drug treatment (intervention)
• Receive no drug treatment at all, but will receive medication once the study is completed (wait-list control)
• Receive a sugar pill (placebo)
• Receive the commonly-used drug (comparison group)

While that would require more participants, you may decide that the information gained in the comparisons is worth it.

A last consideration about all of these comparison group options is the principle of informed consent. Participants must be informed of what condition they may be assigned to, so they should be told that they will be assigned to either a treatment or control group or a wait-list control group or a placebo condition or a comparison group. Researchers have decided that it is both necessary and ethical to not tell the participants which of these groups each individual participant is in until the experiment ends.

Ongoing Equivalence

The second fundamental feature of an experiment is ongoing equivalence. As you might guess, ongoing equivalence means that the IV groups continue to be similar, other than what they experience with the IV. Note that you can't have ongoing equivalence if you didn't start out with initial equivalence; two groups can't stay similar if you didn't start out similar. For ongoing equivalence, the researcher exerts control over, or minimizes the variability in, variables other than the independent and dependent variable, so that the only difference between the groups is the IV. These other variables are called extraneous variables. Darley and Latané (1968) tested all their participants in the same room, exposed them to the same emergency situation, and so on. Notice that although the words manipulation and control have similar meanings in everyday language, researchers make a clear distinction between them. They manipulate the independent variable by systematically changing its levels and control other variables by holding them constant.

Extraneous Variables as “Noise”

Extraneous variables make it difficult to detect the effect of the independent variable in two ways. One is by adding variability or “noise” to the data. Imagine a simple experiment on the effect of mood (happy vs. angry) on the number of happy childhood events people are able to recall. Participants are put into a negative or positive mood (by showing them a video clip designed to elicit happiness or anger) and then asked to recall as many happy childhood events as they can. The two leftmost columns of Table \(\PageIndex{1}\) show what the data might look like if there were no extraneous variables and the number of happy childhood events participants recalled was affected only by their moods. Every participant in the happy mood condition recalled exactly four happy childhood events, and every participant in the angry mood condition recalled exactly three. The effect of mood here is quite obvious. In reality, however, the data would probably look more like those in the two rightmost columns of Table \(\PageIndex{1}\). Even in the happy mood condition, some participants would recall fewer happy memories because they have fewer to draw on, use less effective recall strategies, or are less motivated. And even in the angry mood condition, some participants would recall more happy childhood memories because they have more happy memories to draw on, they use more effective recall strategies, or they are more motivated. Although the mean difference between the two groups is the same as in the idealized data, this difference is much less obvious in the context of the greater variability in the data. Thus one reason researchers try to control extraneous variables is so their data look more like the idealized data in Table \(\PageIndex{1}\), which makes the effect of the independent variable easier to detect (although real data never look quite that good).

Table \(\PageIndex{1}\): Hypothetical Number of Happiness Memories Recalled (Noiseless Data and Realistic Noisy Data)

Idealized “noiseless” data		Realistic “noisy” data
Happy Video	Angry Video	Happy Video	Angry Video
4	3	3	1
4	3	6	3
4	3	2	4
4	3	4	0
4	3	5	5
4	3	2	7
4	3	3	2
4	3	1	5
4	3	6	1
4	3	8	2
M = 4	M = 3	M = 4	M = 3

One way to avoid extraneous variables is to hold them constant. This technique can mean holding situation or task variables constant by testing all participants in the same location, giving them identical instructions, treating them in the same way, and so on. It can also mean holding participant variables constant. For example, many studies of language limit participants to right-handed people, who generally have their language areas isolated in their left cerebral hemispheres (Knecht et al., 2000). Left-handed people are more likely to have their language areas isolated in their right cerebral hemispheres or distributed across both hemispheres, which can change the way they process language and thereby add noise to the data.

In principle, researchers can control extraneous variables by limiting participants to one very specific category of person, such as 20-year-old, white, bisexual, right-handed psychology majors who are cis-gender women. The obvious downside to this approach is that it would lower the external validity of the study—in particular, the extent to which the results can be generalized beyond the people actually studied. For example, it might be unclear whether results obtained with a sample of younger Black women would apply to older Latino men. In many situations, the advantages of a diverse sample (increased external validity) outweigh the reduction in noise achieved by a homogeneous one.

Confounding Variables Can Affect Initial or Ongoing Equivalence

As discussed in the Empirical Study section of the introduction to the scientific method, a confounding variable is an extraneous variable that differs on average across levels of the independent variable (i.e., it is an extraneous variable that varies systematically with the independent variable). A confound can affect initial equivalence if there's no random assignment, or it can affect ongoing equivalence if the groups started out similar but something affected only one IV group during the study. The example of mood and memories could show confounds related to initial equivalence for any differences that random assignment didn't effectively distribute across the two IV conditions. There could also be a confound related to ongoing equivalence if all of the happy video clips were shown in Classroom A and all of the angry video clips were shown in Classroom B, Classroom B had a really noisy heater. While it might seem like the happy video clips caused higher recall of happy memories, it might be that the noisy classroom reduced the total number of total memories that could be recalled. The classroom noise was a confound, affecting only one IV group. As you can see, confounds make it so that you can't state that the IV caused changes in the DV (even if you used random assignment to reach initial equivalence) because you did not have ongoing equivalence. To confound means to confuse, and this effect is exactly why confounding variables are undesirable. Because they differ systematically across conditions—just like the independent variable—they provide an alternative explanation for any observed difference in the dependent variable.

---

References

Darley, J. M., & Latané, B. (1968). Bystander intervention in emergencies: Diffusion of responsibility. Journal of Personality and Social Psychology, 4, 377–383.

Knecht, S., Dräger, B., Deppe, M., Bobe, L., Lohmann, H., Flöel, A., . . . Henningsen, H. (2000). Handedness and hemispheric language dominance in healthy humans. Brain: A Journal of Neurology, 123(12), 2512-2518. http://dx.doi.org/10.1093/brain/123.12.2512

Moseley, J. B., O’Malley, K., Petersen, N. J., Menke, T. J., Brody, B. A., Kuykendall, D. H., … Wray, N. P. (2002). A controlled trial of arthroscopic surgery for osteoarthritis of the knee. The New England Journal of Medicine, 347, 81–88.

Price, D. D., Finniss, D. G., & Benedetti, F. (2008). A comprehensive review of the placebo effect: Recent advances and current thought. Annual Review of Psychology, 59, 565–590.

Shapiro, A. K., & Shapiro, E. (1999). The powerful placebo: From ancient priest to modern physician. Baltimore, MD: Johns Hopkins University Press.

Footnote

1. This section was originally in Jhangiani et al.'s (2019) section 5.3, while the rest of the information from Jhangiani et al. (2019) in this page is from section 5.2. ↵