Exam 2-Psych-Spring 2013-Lourdes University

Front 1

Between-subjects design

Back 1

Experimental design in which the contrast between differently treated groups measures the treatment effect

Front 2

Blocking

Back 2

Dividing subjects into groups according to a measured variable

Front 3

Control Group

Back 3

Condition in which the experimental treatment is withheld to provide a comparison with the treated group

Front 4

Correlation design

Back 4

Research approach in which the independent variable is measured rather than fixed by an intervention

Front 5

Experimental design

Back 5

Research approach in which the independent variable is fixed by a manipulation or natural occurrence

Front 6

Group Threats

Back 6

Internal validity threats to between-subjects designs not protected by random assignment to groups, including selection, regression toward the mean and selection-by-time threat interactions

Front 7

Internal Validity

Back 7

Truthfulness of the assertion that the observed effect is due to the independent variable(s) in the study

Front 8

Matching

Back 8

Assigning subjects to experimental and control conditions to equalize the groups on selected characteristics; can be combined with random assignment but when used alone, cannot guarantee group equivalence on variables not used for matching

Front 9

Mixed design

Back 9

Experimental design that includes both within-subjects and between-subjects features

Front 10

Mortality

Back 10

Subject attrition from pretest to post-test, which casts doubt on validity of the study; here conceptualized as a threat to measurement construct validity because protection against this threat is not provided by a control group or random assignment

Front 11

O

Back 11

Symbol used in design diagrams to represent one or more observations collected at some time point

Front 12

On-line data collection

Back 12

Using a computer to receive and store data directly from an experiment in progress

Front 13

Panel design

Back 13

Correlation design in which a group of subjects is surveyed or measured at more than one time point; the group itself is called a panel

Front 14

Preexperiments

Back 14

Class of very primitive quasi-experimenal designs that are highly vulnerable to internal validity

Front 15

Prerandomization

Back 15

Random assignment of subjects to treatment conditions before obtaining informed consent

Front 16

Quasi-experimental design

Back 16

Experimental approach in which the researcher does not assign subjects randomly to treatment and control conditions

Front 17

R

Back 17

Symbol used in design diagrams to represent random assignment to differently treated groups

Front 18

Random assignment

Back 18

Method of placing subjects in different conditions so that each subject has an equal chance of being in any group to avoid systematic subject differences between the groups

Front 19

Reverse causation

Back 19

Threat to internal validity that an observed relationship results from causation of opposite direction from that hypothesized; protected against by fixing the independent variable to occur before the dependent variable

Front 20

Time Threats

Back 20

Internal validity threats to within-subjects designs not protected by control groups, including history, maturation, measurement decay, and test reactivity

Front 21

True Experiment

Back 21

Experimental design in which subjects are randomly assigned to two or more differently treated conditions

Front 22

Within-subject design

Back 22

Experimental design in which the change in subjects from before to after the manipulation measures the treatment effect

Front 23

X

Back 23

Symbol used in design diagrams to represent presence of an experimental manipulation

Front 24

Blind

Back 24

Technique of avoiding experimenter expectancy by concealing assignment of subject from researcher or of avoiding demand characteristic by concealing assignment of subject from subject. When both subject and experimenter are blind to the assignment, the study is called "double blind"