What is one piece of information does on obtain about a relationship between two variables from a correlation coefficient?
(1) Degree/Magnitude of relationship (very strong to nonexistent)
(2) Direction of relationship (positive, negative)
(3) Form of relationship (typically used for linear relationships)
100
What is one of the two ways to avoid mislead findings that will not generalize to the population of interest?
(1) Create a sampling plan that allows you to estimate the potential amount of error in your estimates (margin of error) and the percentage of time that the estimate will be within that margin (confidence level)
(2) Include diverse population elements AND take them into account in the proportion to which they exist in the population
100
What is the primary difference between experimental and individual-difference variables?
Experimental variables: properties or characteristics that an experimenter can manipulate or expose people to
Individual difference variables: qualities or characteristics that subject in a study brings with them; not easily manipulated, or impossible to manipulate
100
What are two of the three goals served by using a laboratory setting.
(1) Control
(2) Manipulation
(3) Ability to construct a setting
100
What is one reason why matching is a bad strategy to deal with lack of random assignment?
(1) Doesn’t control for regression toward the mean
(2) Doesn't match on all variables
200
What are the characteristics of an IV that would motivate you to use an Independent Samples t-test?
Categorical, with two levels (e.g., an experiment with two treatment groups, an individual-difference variable with two groups [gender])
200
What is the primary distinction between nonprobability and probability sampling techniques?
Nonprobability: no way to estimate the probability that each element has of being included in the sample
Probability: Can specify the probability that each element will have of being included in the sample
200
What is the name of the design where you have subjects randomly assigned to two levels of your IV and where subjects are administered both a pretest and posttest DV measure?
Pretest-Posttest Two-Group Design
200
What is a judgment study?
A study where some sort of stimulus material provided and the person makes responds in some way (recognition, recollection, classification, evaluation); participants are more passive than in impact studies
200
What is a sampling distribution?
The distribution of statistics (e.g., mean, correlation, proportion) for all possible samples of a given size
300
What is the difference between between-subjects and within-subjects (repeated measures) analyses, in terms of the comparisons that one is making?
Between-subjects: Making comparisons between subjects or groups
Within-subjects: Making comparisons within subjects or groups
300
What is quota sampling?
Select elements nonrandomly according to some fixed quota
300
What is one of the two strengths of randomized experimental designs?
What is the difference between experimental realism and mundane realism?
Experimental realism: extent to which the manipulations/measures perceived in the intended ways by the participants (construct validity)
Mundane realism: degree of resemblance between the lab operational definitions and some target object/event outside the lab (similar to face validity)
300
What statistical method would you use if you had a randomized multiple-group design?
Between-subjects ANOVA
400
What is the shared reason for why one would use correlation or regression (i.e., what characteristic of an IV would motivate one to use CORR or REG)?
In both cases, you have a continuous IV.
400
When would you use snowball sampling?
When the research question involves a special sample that is hard to identify or reach.
400
What is one reason why we might use a nonrandomized design as opposed to a randomized design?
(1) Exploring distribution of a variable in some population
(2) Discovering/documenting association between variables in some population
(3) Examining effect of IV when random assignment is impossible or unethical
400
We stated the external validity can be a problem for lab settings. Name one of the other two problems associated with laboratory research and define it.
(1) Experimenter Expectancy Effects: Experimenter expectations influence how participants behave
(2) Demand Characteristics: Participants search for and pick up subtle cues in the experiment to use as guidance for their behavior
400
What statistical method would you use if you had a pretest-posttest nonequivalent control group design?
Mixed-Model ANOVA
500
Why would you use correlation instead of regression AND why would you use regression instead of correlation?
Regression: Clear distinction between IV and DV; want to predict values on the DV
Correlation: Not a clear distinction between IV and DV; want to assess degree of relationship
500
Fill in the following blanks pertaining to the selection of participants in simple random sampling: Select n units out of N elements such that each element AND each possible combination has an ______ and _________________ probability of being selected
Equal; independent
500
What is the major assumption about the subjects in the various groups that must be made in nonrandomized designs?
Nonequivalence: Must assume that the groups may differ in regards to the characteristics of the participants within them
500
What information is a manipulation check capable of telling us?
(1) Construct validity of the manipulation
(2) What may have gone wrong, if the results didn't turn out as expected
500
What sampling method entails dividing the population of interest into two or more groups or segments and randomly sampling participants from each segment?