-
Validity
Does it measure what it’s supposed to measure?
-
Reliability
Consistency and reproducibility
-
Nominal
Measurement involves observations in qualitatively different categories
-
Ordinal
Data according to size
-
Interval
Measurement data same distance apart
-
Quota sampling
Nonrandom sampling technique when subgroups from each group are chosen
-
Stratified sampling
Process of sampling in which groups of interest are identified and participants selected at random from the groups
-
Convenience sampling
Picking population that’s easy or available
-
Subject variable
Independent variable resembling true IV but created on basis of pre-existing characteristics of participant
-
Manipulated variable
True IV changed by the experimenter
-
Measured variable
Variable used to create groups to be measured, but participants are assigned based on their characteristics rather than researchers’ system
-
Quasi-experiment
Research project resembling experiment that compares groups but no random or systematic assignment (assignment based on participant characteristics)
-
Ex post facto study
Resembles experiment but uses existing grouped data that didn’t involve random assignment
-
Qualitative variable
Variable whose different values are based on qualitative rather than numeric differences
-
Quantitative variable
Variable whose differences are based on numerical differences like size or duration
-
Repeated Measures design
Design in which a single participant is observed & measured on more than one level of independent variable rather than individuals on each level of individual variable
-
ANOVA (Analysis of variance)
Family of statistical tests that compares group means to assess whether differences across means are reliable
-
Planned comparison
When you know in advance which groups you’re going to compare
-
Post hoc comparison
When you compare groups after looking at results
-
Factorial design
Research design when investigator manipulates more than one variable, each level crossed with each level of all other IVs
-
Main effect
In factorial design, differences among groups for single IV that are significant, temporarily ignoring all other IVs
-
Interaction effect
In a factorial design, differences across groups of a single IV that are predictable only by knowing the level of another IV
-
Higher order interaction
Interaction in factorial design that involves joint effect of more than 2 IVs
-
Within-subjects design
Repeated measures – collecting data from same people more than once
-
Between groups design
Test participants in different groups to see if there are differences between groups
-
Sequence effects
The result of multiple or repeated measurements of individuals in different experiment conditions such that they behave differently on later measurements as a result of having undergone the earlier measurements
-
Order effects
The result of multiple or repeated measurements of individuals in different experimental conditions such that a particular behavior changed depending on which condition it follows
-
Transfer
A change in behavior in a repeated measures design that results from learning that takes place in an earlier condition
-
Symmetric transfer
A change in behavior in a repeated measures design that results from learning in an earlier condition, with the same degree of change in later behaviors, regardless of the order of conditions
-
Asymmetric transfer
A change in behavior in a repeated measures design that results from learning in an earlier condition, with differences in the amount of transfer in a later condition depending on which conditions occur first
-
Counterbalancing
In a repeated measures design, the changing of the order of conditions to avoid contamination of data
-
Complete counterbalancing
In a repeated measures design, the use of all possible orders of experimental conditions to avoid contamination of data because of systematic sequence
-
Partial counterbalancing
In a repeated measures design the use of a subset of all possible orders of experimental conditions to avoid contamination of data because of systematic sequence, order, or transfer effects
-
Statistical regression
Participants categorized or selected for research participation on basis of initial observation that involves significant measurement error that is not likely to repeat itself on later measurements, giving the false impression that change is due to a treatment when it is really due to the difference in measurement error.
-
Attrition threat
Participants dropping out
-
Maturation threat
Short or long term changes in participant
-
Selection threat
Groups seem to differ – could be due to initial differences rather than the IV
-
Testing threat
Participants’ previous tests changes their behavior
-
Nonequivalent control group
A quasi-experimental research design in which 2 groups differ on pre-existing dimension measured on a pre-test, exposed to treatment, and measured on a posttest
-
Snowball sampling
An individual from a hidden population is likely to know others in the group
-
Observational research
Investigators are trained to record human or nonhuman behavior exactly as it occurs, avoiding interpretation
-
Cohort effects
Differences across age groups having to do with characteristics of the era in which the person grew up rather than age specifically
-
Cohort study
Longitudinal research in which investigator samples randomly from population because of specific characteristic, often age
-
Panel study
Longitudinal research in which investigator studies the same individuals over time
-
Withdrawal design
Investigator observes baseline, applies treatment, watches change, and then removes or withdraws treatment and observes
|
|