Term
|
Definition
| results are technically sound and accurate |
|
|
Term
|
Definition
| possible effects of extraneous variables on the DEPENDENT variable are controlled for (more difficult in field research) |
|
|
Term
|
Definition
| the ability to generalize results beyond the current experiment (more difficult in laboratory research) |
|
|
Term
|
Definition
events that happen between pre and post tests that may effect the dependent variable.
avoided by: using a control group (people not exposed to the measure being studied) |
|
|
Term
|
Definition
changes that occur because of the passage of time rather than experimental measures.
avoided by: using a control group (people not exposed to the measure being studied) |
|
|
Term
|
Definition
in doing a pre and post test, participants learn how to do better on the test or become "test-wise"
avoided by:using a control group (people not exposed to pre- testing) |
|
|
Term
|
Definition
implements used to collect data are not accurate/procedures not standardized
avoided by: well designed instruments, careful calibration, and thorough training of individuals implementing the measures |
|
|
Term
|
Definition
non-random selection of participants leading to potential differences between groups BEFORE the study even begins (however, selection cannot always be random)
avoided by: recruiting volunteers THEN randomly assigning them to groups, matching participants based on selected characteristics THEN randomly assigning them to groups, or pre-testing to discover pre-study differences |
|
|
Term
| selection-maturation effect |
|
Definition
intact groups also vary in maturation level (combination of selection bias and maturation effect)
avoided by: pre-testing and pre- screening participants for maturation level |
|
|
Term
|
Definition
when participants are selected based on high or low scores-- when retested students with high scores tend to score lower and students with low scores tend to score higher (regression towards the mean)
avoided by: using random samples that encompass a full range of scores |
|
|
Term
|
Definition
if a study is done over the course of time, when participants do not continue to participate for any reason
avoided by: oversampling and incentives to continue participation |
|
|
Term
|
Definition
people behave differently when they know they are being watched
avoided by: give control group treatment that will make them think they are NOT in the control group but that will not effect the dependent variable |
|
|
Term
|
Definition
people behave differently when they expect an effect
avoided by: use of placebos, blind studies (those involved do not know if they are in the experimental or control group), double blind studies (neither the researcher nor the participants know who is in the experimental group and who is control) |
|
|
Term
|
Definition
the treatment of one group contaminates another via discussion of the experimental measures between groups
avoided by: somehow having groups who do not come into contact |
|
|
Term
|
Definition
when the physical location of the study has an effect on the outcome
avoided by: making sure the location is the same for all participants (if this is not possible, minimizing differences as much as possible) |
|
|
Term
|
Definition
when the person implementing the study has an effect on the outcome (whether because the level of skill is different between researchers or the researcher favors one group over the other)
avoided by: making sure everyone is properly trained and (when possible) one who is neutral toward the two groups should be responsible for implementation |
|
|
Term
| selection treatment interaction |
|
Definition
whether results from the study can be generalized beyond the study population. Results can only be generalized to those who had a CHANCE of being chosen for participation
avoided by: random samples of study participants from as broad a pool as possible |
|
|
Term
| setting treatment interaction |
|
Definition
the extent to which the physical setting has an effect on the outcome of the study/the ability to generalize results to other settings
avoided by: sampling as broadly as possible across as many settings as possible |
|
|
Term
| history treatment interaction |
|
Definition
when the researcher tries to generalize findings to situations other than the present
avoided by: longitudinal studies |
|
|
