Notes

Independent Variables
- the cause of something
- manipulated variable in one experiment


Dependent Variables
- bears the effect of the independent variables

To distinguish the 2 variables:
Independent Variable Dependent Variable
- manipulator - result
- cause - effect
- influencer - outcome


Extraneous Variables
- extra or unexpected variable cropping up outside the research design

For example,
if you were conducting a taste between two sodas, you should want to control:
* Temperature of the soda so that both are the same
* How long the soda had been on the shelf
* The container, so that both cook the same

Controlling extraneous variables is an important aspect of experimental design. When you control an extraneous variable, you turn it into a control variable.

Types and controls of extraneous variables

Demand characteristics

Demand characteristics are cues that encourage participants to conform to researchers’ behavioral expectations.

Sometimes, participants can infer the intentions behind a research study from the materials or experimental settings, and use these hints to act in ways that are consistent with study hypotheses. These demand characteristics can bias the study outcomes and reduce the external validity, or generalizability, of the results.

Example: Demand characteristics
Research participants in the experimental group easily draw links between the lab setting, being asked to wear lab coats, and the questions on their scientific knowledge.
They work harder to do well on the quiz by paying more attention to the questions.

You can avoid demand characteristics by making it difficult for participants to guess the aim of your study. Ask participants to perform unrelated filler tasks or fill out plausibly relevant surveys to lead them away from the true nature of the study.

Experimenter effects

Experimenter effects are unintentional actions by researchers that can influence study outcomes.

There are two main types of experimenter effects:

Experimenters’ interactions with participants can unintentionally affect their behaviors.
Errors in measurement, observation, analysis, or interpretation may change the study results.

Example: Experimenter effect
You motivate and encourage the participants wearing lab coats to do their best on the quiz. They are more comfortable in the lab environment and feel confident going into the quiz; therefore, they perform well.
Participants wearing the non-lab coats are not encouraged to perform well on the quiz. Therefore, they don’t work as hard on their responses.

To avoid experimenter effects, you can implement masking (blinding) to hide the condition assignment from participants and experimenters. In a double-blind study, researchers won’t be able to bias participants towards acting in expected ways or selectively interpret results to suit their hypotheses.

Situational variables

Situational variables, such as lighting or temperature, can alter participants’ behaviors in study environments. These factors are sources of random error or random variation in your measurements.

To understand the true relationship between independent and dependent variables, you’ll need to reduce or eliminate the effect of situational factors on your study outcomes.

Example: Situational variable
To perform your experiment, you use the lab rooms on campus. They are only available either early in the morning or late in the day. Because the time of day may affect test performance, it’s an extraneous variable.

To avoid situational variables from influencing study outcomes, it’s best to hold variables constant throughout the study or statistically account for them in your analyses.

Participant variables

A participant variable is any characteristic or aspect of a participant’s background that could affect study results, even though it’s not the focus of an experiment.

Participant variables can include sex, gender, age, educational attainment, marital status, religious affiliation, etc.

Since these individual differences between participants may lead to different outcomes, it’s important to measure and analyze these variables.

Example: Participant variables
Educational background and undergraduate majors are important participant variables for your study on scientific reasoning. Participants with strong educational backgrounds in STEM subjects are likely to perform better than others.

To control participant variables, you should aim to use random assignment to divide your sample into control and experimental groups. Random assignment makes your groups comparable by evenly distributing participant characteristics between them.


Confounding Variables
- unstable variable

Confounding variables (a.k.a. confounders or confounding factors) are a type of extraneous variable that is related to a study’s independent and dependent variables. A variable must meet two conditions to be a confounder:

- It must be correlated with the independent variable. This may be a causal relationship, but it does not have to be.
- It must be causally related to the dependent variable.

confounding variable is an unmeasured third variable that influences both the supposed cause and the effect.

It’s important to consider potential confounding variables and account for them in your research design to ensure your results are valid.

Why confounding variables matter

To ensure the internal validity of your research, you must account for confounding variables. If you fail to do so, your results may not reflect the actual relationship between the variables that you are interested in.

For instance, you may find a cause-and-effect relationship that does not actually exist, because the effect you measure is caused by the confounding variable (and not by your independent variable).

Example
You find that more workers are employed in states with higher minimum wages. Does this mean that higher minimum wages lead to higher employment rates?
Not necessarily. Perhaps states with better job markets are more likely to raise their minimum wages, rather than the other way around. You must consider the prior employment trends in your analysis of the impact of the minimum wage on employment, or you might find a causal relationship where none exists.

Even if you correctly identify a cause-and-effect relationship, confounding variables can result in over-or underestimating the impact of your independent variable on your dependent variable.

Example
You find that babies born to mothers who smoked during their pregnancies weigh significantly less than those born to non-smoking mothers. However, if you do not account for the fact that smokers are more likely to engage in other unhealthy behaviors, such as drinking or eating less healthy foods, then you might overestimate the relationship between smoking and low birth weight.

How to reduce the impact of confounding variables

There are several methods of accounting for confounding variables. You can use the following methods when studying any type of subject—humans, animals, plants, chemicals, etc. Each method has its own advantages and disadvantages.

Restriction

In this method, you restrict your treatment group by only including subjects with the same values of potential confounding factors.

Since these values do not differ among the subjects of your study, they cannot correlate with your independent variable and thus cannot confound the cause-and-effect relationship you are studying.

Restriction example
You want to study whether a low-carb diet can cause weight loss. Since you know that age, sex, level of education, and exercise intensity are all factors that may be associated with weight loss, as well as with the diet your subjects choose to follow, you choose to restrict your subject pool to 45-year-old women with bachelor’s degrees who exercise at moderate levels of intensity between 100–150 minutes per week.

✓ Relatively easy to implement
✗ Restricts your sample a great deal
✗ You might fail to consider other potential confounders

Matching
In this method, you select a comparison group that matches the treatment group. Each member of the comparison group should have a counterpart in the treatment group with the same values of potential confounders, but different independent variable values.

This allows you to eliminate the possibility that differences in confounding variables cause variation in outcomes between the treatment and comparison groups. If you have accounted for any potential confounders, you can thus conclude that the difference in the independent variable must be the cause of the variation in the dependent variable.

Matching example
In your study on a low-carb diet and weight loss, you match up your subjects on age, sex, level of education, and exercise intensity. This allows you to include a wider range of subjects: your treatment group includes men and women of different ages with a variety of education levels.
Each subject on a low-carb diet is matched with another subject with the same characteristics who are not on the diet. So for every 40-year-old highly educated man who follows a low-carb diet, you find another 40-year-old highly educated man who does not, to compare the weight loss between the two subjects. You do the same for all the other subjects in your treatment sample.

✓ Allows you to include more subjects than restriction
✗ Can prove difficult to implement since you need pairs of subjects that match every potential confounding variable
✗ Other variables that you cannot match on might also be confounding variables

Statistical control
If you have already collected the data, you can include the possible confounders as control variables in your regression models; in this way, you will control for the impact of the confounding variable.

Any effect that the potential confounding variable has on the dependent variable will show up in the results of the regression and allow you to separate the impact of the independent variable.

Statistical control example
After collecting data about weight loss and low-carb diets from a range of participants, in your regression model, you include exercise levels, education, age, and sex as control variables, along with the type of diet each subject follows as the independent variable. This allows you to separate the impact of diet chosen from the influence of these other four variables on weight loss in your regression.

✓ Easy to implement
✓ Can be performed after data collection
✗ You can only control for variables that you observe directly, but other confounding variables you have not accounted for might remain

Randomization
Another way to minimize the impact of confounding variables is to randomize the values of your independent variable. For instance, if some of your participants are assigned to a treatment group while others are in a control group, you can randomly assign participants to each group.

Randomization ensures that with a sufficiently large sample, all potential confounding variables—even those you cannot directly observe in your study—will have the same average value between different groups. Since these variables do not differ by group assignment, they cannot correlate with your independent variable and thus cannot confound your study.

Since this method allows you to account for all potential confounding variables, which is nearly impossible to do otherwise, it is often considered to be the best way to reduce the impact of confounding variables.

Randomization example
You gather a large group of subjects to participate in your study on weight loss. You randomly select half of them to follow a low-carb diet and the other half to continue their normal eating habits.
Randomization guarantees that both your treatment (the low-carb-diet group), as well as your control group, will have not only the same average age, education, and exercise levels, but also the same average values on other characteristics that you haven’t measured as well.

✓ Allows you to account for all possible confounding variables, including ones that you may not observe directly
✓ Considered the best method for minimizing the impact of confounding variables
✗ Most difficult to carry out
✗ Must be implemented prior to beginning data collection
✗ You must ensure that only those in the treatment (and not control) group receive the treatment