Correlations vs. Controlled Between-groups Designs

This set of flashcards covers key concepts related to correlations and controlled between-groups designs, including measures of relationships, interpretations of correlation coefficients, and experimental methodology.

What does Karl Pearson’s Correlation Coefficient (r) measure?

It describes a linear relationship between pairs of quantitative variables, indicating the direction and strength of that relationship.

What is the range of values for Pearson's r?

Between -1.00 and 1.00.

What does an r value of +1 indicate?

A perfect positive linear relationship between two variables.

What r value indicates a weak positive correlation?

An r value of 0.15 shows a weak positive correlation.

What is the difference between interpolation and extrapolation?

Interpolation predicts data points within the range of known data, while extrapolation predicts outside of that range.

What does a negative r value signify?

It indicates a negative direction of the relationship between two variables.

What is Simpson’s paradox?

The reversal of apparent relationships when data is separated by another variable; combining different populations may show opposite trends.

What type of studies are correlational studies?

Observational/descriptive studies that analyze the relationships between variables without manipulating them.

What are extraneous variables?

Variables that can influence the dependent variable and potentially compromise the interpretation of a study.

What is the main disadvantage of observational studies?

They cannot make causal claims, as correlation does not imply causation.

What is a spurious relationship?

A relationship that appears to exist between two variables but is actually coincidental or due to a third variable.

What does the phrase 'Cum hoc ergo propter hoc' mean?

'With this, therefore because of this', indicating that correlation does not imply causation.

What are the two basic requirements for a controlled between-groups experiment?

The groups must be equivalent before manipulation and free of confounds.

How do we operationalize 'performance' in an experiment?

By defining it through measurable outputs, such as speed and accuracy in a task.

What is the independent variable in the caffeine experiment?

Caffeine or no caffeine.

What does a correlation coefficient of +0.67 indicate?

A strong positive relationship.

Why are observational studies useful?

They are useful for collecting data, finding relationships, and generating hypotheses.

What does the range of the multiple correlation coefficient represent?

The degree of relationship strength between predictors and a predicted variable, ranging from 0.0 to 1.00.

Why may observational studies be ethical?

They can study variables or relationships that cannot be manipulated or are unethical to manipulate.

What is the key concept behind controlled experiments?

The ability to manipulate an independent variable to infer causal relationships with the dependent variable.

What are demand characteristics?

Cues in an experiment that may influence participants' behavior or responses.

In the controlled experiment about caffeine, what are the dependent measures?

Response time (words per minute) and accuracy (number of errors).

What is the error term in controlled experiments?

Variability within groups due to sampling error and other uncontrolled factors.

What type of sampling minimizes variability in populations?

Random selection from the population.

What is a potential outcome of sampling bias?

It can affect the external validity of the experiment.

What needs to be controlled to ensure equivalent groups in experiments?

Natural variability and potential confounding variables.

What does a weak correlation of r = 0.1 indicate?

A weak positive relationship between the two variables.

How does participant attrition affect studies?

It may lead to biased results as the remaining participants may not represent the original sample.

What factors can influence the dependent variable in typing speed experiments?

Medications, sleep, experience, age, etc.

Why might interpolation be safer than extrapolation?

Because interpolation relies on known data ranges, while extrapolation relies on assumptions about unknown data.

What example illustrates a third variable problem?

The positive correlation between murder rates and ice cream sales being driven by summer temperatures.