an experiment in which the researcher has complete control over the manipulation of the IV.
holding constant as many extraneous variables as possible
a process used to produce random, and hopefully equivalent, groups by ensuring that any extraneous influence is as likely to affect one group as it is the other and thereby hold it constant.
when the experimenter cannot use random assignment to make groups equivalent because they select levels of the IV, not manipulate it. This happens when the IV's not manipulable, like race, gender or age.
an experiment with one IV.
an experiment with more than one IV.
the particular way in which participants are treated.
experimental group/treatment group
group that is given the level of IV of interest
provides a means to determine changes that might occur naturally in the absence of treatment
a measure of the outcome in both groups after the experimental group has received treatment. All true experiments have this.
a measure of the DV before experimental intervention. All true experiments have this.
independent groups receive different levels of the IV
all research participants receiver all levels of the IV
randomized comparative change design/ pretest-post-test control group design
a randomized experimental design with a pretest and posttest.
a procedure used to equate groups, but is a poor substitute for randomization. Pair people in control and treatment group on basis of similarity in gender, age, etc. Because unmatched characteristics can skew results and there are only so many characteristics you can match in two individuals, if matching is used instead of random assignment it becomes quasi-experimental; not a true experiment.
matching combined with random assignment. Used in between-subject studies, create pairs of participants matched on a variable that's strongly related to the DV. Each pair is split up and assigned to either control or treatment group.
technique to control for order of treatments. More effective than randomization because ensures e/ treatment/level of IV occurs in all combinations.
the extent to which the systematic manipulation of one or more IV's produces the predicted/hypothesized effect on the DV.
the extent to which a researcher has constructed the best possible study to evaluate the topic under investigation.
occurs when an IV fails to produce any effect on the DV.
include subject roles/biases, experimenter effects. Threat to internal validity.
both experimenter and participant are deliberately kept blind to the conditions of study.
to quantify the effects of variables that cannot be randomized but are known to be strongly related to the DV.
statistical technique which allows the researcher to test whether a difference on the dependent measure exists, even after controlling or co-varying.
the likelihood that the observed effect revealed by data analysis could be due to chance.
the strength of the relationship of the IV on the DV.
the hypothesis of no difference or no relationship which the researcher tries to reject, nullify or disprove.
research hypothesis/alternative hypothesis
predicts a certain relationship typically on the basis of theory.
Type 1 error
when the null hypothesis is wrongly rejected. Type 1 error= stat. significance/alpha.
Type 2 error
when the null hypothesis is wrongly accepted. Type 2 error= beta/ power. Usually occurs when sample size is small and relationship between IV and DV is not revealed.
the extent to which results of a particular study that involves a sample of subjects will generalize to the wider population.
tool for maximizing generalizability.
form of external validity concerned with the real-life impact of results generated from studies.
when a correlation between two variables might be caused by something other than an effect of the presumed IV on the DV.
when you make conclusions about an individual based only on analyses of group data.
when you make conclusions about a group based only on analyses of individual data.
any influence on the DV that isn't solely to the changes in the IV.