Psychological Research Methods: Variables, Correlations, and Experimental Design

Psychologists employ specific methodologies to conduct experiments and research studies.
A course-specific note mentions an extra credit quiz, which is described as very easy and worth approximately $5$ bonus points.

A variable is defined as anything that varies.
More specifically, a variable is anything that can be measured.
When discussing correlations, the distinction between dependent and independent variables is initially disregarded as all variables are treated the same for the purpose of the relationship analysis.

Correlation involves identifying relationships between variables to see if one can be used to predict the other.
The Height and Arrival Time Example:
- If one were to measure the heights of students in a class (the variable), it could be correlated with the time the students showed up for class.
- In this scenario, there would likely be no correlation found between the two variables.
The Lecture Attendance and GPA Example:
- A study of data from the end of a quarter might compare the amount of time a student spends in lectures with their overall GPA.
- This relationship often shows a strong positive correlation, specifically cited as a coefficient of $0.75$ .
- Implication: While a strong correlation exists, it does not mean that one variable causes the other; however, it does allow for prediction. Specifically, higher lecture attendance can predict a better GPA.
- Bidirectionality: Variables in a correlation can move in both directions.

Causality is distinct from correlation in that causality only flows in one direction.
A major limitation of correlation is that it merely reflects an observed relationship, not a causal one.
Confounding Variables (Third Variables):
- There could always be an external third factor causing both of the observed variables to move together.
- Example: Students who attend lectures regularly might also study significantly more during lunch. In this case, the extra studying (a causal factor) might be the actual reason for higher grades, not the attendance itself.
Spurious Correlations Example (Nicolas Cage and Drownings):
- Data from 1999 through 2009 shows a correlation of approximately $0.7$ between the number of movies Nicolas Cage appeared in per year and the number of people who drowned in pools per year.
- These variables move together remarkably well, with a notable peak around the year 2007.
- Despite the statistical movement, there is no causal relation. No action should be taken based on such data (e.g., stopping Nicolas Cage from acting to lower drowning rates).

In experimental research, researchers manipulate specific elements to observe outcomes.
Independent Variable (IV): This is the variable that the researcher adds, changes, or manipulates. It is the hypothesized cause.
Dependent Variable (DV): This is the variable that the researcher measures. It is the outcome that changes or responds to the manipulation of the Independent Variable.
Neuroscience Supplement Example:
- The researcher creates a brain supplement hypothesized to increase memory and test scores.
- Sample: The current class participants.
- Independent Variable: The brain supplement being added.
- Dependent Variable: The test scores being measured.

Experimental Group: This group of participants receives the actual treatment or manipulation (e.g., the brain supplement).
Control Group: This group provides a baseline for comparison. To be effective, the control group must be well-set.
Potential Pitfalls in Control Groups:
- Comparing a psychology class receiving a supplement to a math class that does not would be a poor comparison.
- This is because differences in test scores could be explained by the subject matter, the professor, or other class-specific factors rather than the supplement.

Placebos: A more effective experimental design involves giving half the participants a fake supplement (a placebo) that contains nothing.
The placebo ensure that everyone thinks they took the supplement, accounting for the fact that placebos have been proven to "work" via psychological expectation even when they contain no active ingredients.
Falsifiability: The goal of scientific experiment design is often to test a theory by trying to prove it wrong (disproving/falsifying the theory) rather than just looking for confirmation.

Student Question: Wouldn't a better way to do it be to give half the group a fake supplement that has nothing so that everyone thinks they took it because it's been proven that placebos will work even if there's nothing in them?
Response: Yes, that is a perfect methodology and aligns with the concept of placebo-controlled studies.
Student Question regarding Little Albert: Isn't that for Little Albert? He was, like, exposed to a rat or something with fur, and then they would play, like, a sound to scare him, and then he associated the scary sound with seeing a furry thing?
Context of Little Albert: This case study involves classical conditioning where a child was conditioned to fear a white rat by pairing the sight of the animal with a loud, frightening noise, eventually resulting in the child fearing any furry objects.