Experimental Psychology Midterms Flashcards
The degree of relationship between 2 traits, behaviors, or events
Correlation
What are the purposes when measuring correlation?
- To explore behaviors that are not yet understood
- To study observable characteristics that can take on different values
- To make predictions about behaviors
Which levels of measurement can correlation be applied to?
Interval to Ratio level
This is the coefficient that results from a statistical measure of correlation
Pearson Product Moment Correlation
The statistical formulas for this type of correlation use a General Linear Model
Simple Correlation
What are the 4 things correlation can determine?
- Direction of relationship
- Strength of relationship
- Coefficient of determination
- Scatterplot
What is the difference between a negative and positive relationship?
Positive means as variable 1 increases, variable 2 also increases. Negative means as variable 1 increases, variable 2 decreases or vice versa
True or False: the nearer the r is to 1, the weaker the relationship
False
This is determined by r squared, it is the proportion of shared common variance between 2 variables
Coefficient of Determination
This is a graphic representation of the relationship between 2 variables
Scatterplot
True or False: a perfect correlation also indicates a causal relationship
False
True or False: the regression line goes through most points on the graph
True
This is the estimate of a score on one of the measured behaviors based on the score from the other
Linear Regression
This is the mathematical equation that best describes the linear relationship between 2 scores
Regression Line
This tests the relationship between 3 or more predictor variables with a criterion variable
Multiple Correlation
This predicts the scores of criterion variables based on one variable from scores on sets of other predictor variables
Multiple Regression
True or False: multiple regression only requires 1 predictor variable
False, it requires at least 2
This shows the weight or degree of influence of each predictor variable
Beta Weights
This is a multiple regression design where subjects are measured on several related behaviors and causal sequences for these behaviors are established
Path Analysis
This multiple regression design measures several related characteristics on two separate points in time. It is based on correlation
Cross-lagged Panel Design
These are study designs that resemble but are not classified as experimental designs
Quasi-experimental Designs
True or False: random assignment is recommended for quasi-experimental studies
False, it is not possible to do random assignment
What is the level of internal validity of quasi-experimental designs?
Low Internal Validity
If experimental designs are based on controlled treatments, quasi designs are based on?
Pre-existing conditions/naturally occurring events