Survey Says…
(Ch. 6)
(Ch. 6)
What’s the difference between an open-ended and a forced-choice question?
Open-ended: any response; Forced-choice: limited options.
Define population and sample.
Population = all individuals of interest; sample = subset studied.
A measure of direction and strength between two variables.
What is the correlation coefficient (r)?
Statistically holding one variable constant to isolate another effect.
What does “control for” mean?
A test that compares the means between 2 groups
What is a t-test?
A question asking two things at once (“Do like your dorm and your college football team”).
What is a double-barreled question?
Participants easy to reach but not representative
What is a convenience sample?
______ ______ is when both variables increase; whereas, in a ______ ______ as one variable increases, the other decreases.
Positive correlation is when both variables increase; whereas, in a negative correlation, as one variable increases, the other decreases.
This type of correlation examines how two variables relate when they’re measured at the same time point in a longitudinal design.
What is a cross-sectional correlation?
When participants respond “neutral” to every item on a Likert scale because they don’t want to pick a side, they’re guilty of this response bias.
What is fence-sitting?
When researchers’ expectations influence observations; use blind observers to prevent it.
What is observer bias?
Random assignment improves ____ whereas random sampling improves _____.
Random Assignment = internal validity; Random Sampling = external validity.
Interpret: r = .80, p < .05.
Strong, positive, significant relationship.
This type of correlation checks the consistency or stability of one variable across different time points.
What is an autocorrelation?
A developmental psychologist tests whether age group (children, adolescents, adults) affects reaction time on a task. What analysis should they use?
ANOVA
Give one advantage and one disadvantage of self-report surveys.
Advantage: efficient, subjective data; Disadvantage: social desirability bias.
Divide population into subgroups and randomly sample within each.
What is stratified random sampling?
Just one extreme data point like this can send your correlation from “barely associated” to “deeply committed,” even if most points disagree.
What is an outlier?
Explains the relationship between two variables (why).
What is a mediator?
This correlation tests whether an earlier measure of one variable predicts a later measure of another, helping establish temporal precedence.
What is a cross-lag correlation?
Participants change behavior when observed; minimize by habituation or unobtrusive measures.
What is reactivity, and how can it be minimized?
In a ___ ___, everyone has an equal chance of being selected; where as, in a ___ ___ NOT everyone has an equal chance at being selected
In a probability/random sample, everyone has an equal chance of being selected; where as, in a non-probability/biased sample NOT everyone has an equal chance at being selected
Why can’t bivariate correlation prove causation?
Because directionality and third-variable problems exist.
You find stress predicts GPA, but only for students who sleep <6 hrs. What is “sleep”?
A health psychologist predicts happiness from daily exercise, but also wants to know if this relationship depends on how much social support someone has. What analysis should they run?
moderation regression analysis