Social Psychology
10th EditionElliot Aronson, Robin M. Akert, Samuel R. Sommers, Timothy D. Wilson
525 solutions
Consumer Behavior: Buying, Having, Being
13th EditionMichael R Solomon
449 solutions
Myers' Psychology for the AP Course
3rd EditionC. Nathan DeWall, David G Myers
955 solutions
Psychology
4th EditionLaura Namy, Scott Lilienfeld, Steven Lynn
256 solutions
In a quasi-experiment by Minke (2011), the effectiveness of a Danish halfway house (a place for people with criminal backgrounds to learn skills for reintegrating into society) that mixes offenders with nonoffenders was compared with that of the control condition of halfway houses housing only offenders. The measurement of effectiveness was based on rate of recidivism (relapse in criminal behavior) in former inmates.
Identify the threat to internal validity in each scenario
that might have occurred in this study.
Inmates in the mixed halfway house were provided with job training opportunities, while inmates in the only-offenders halfway house were not.selection-history threat
Correct label: selection-history threat
Inmates in the mixed halfway house showed lower rates of recidivism only because they understood the purpose of the study.demand characteristics
Correct label: demand characteristics
Inmates were allowed to choose their own halfway
house.selection effects
Correct label: selection effects
In their measurement of recidivism, researchers did not include inmates who were sent directly to prison from their halfway house for violations.attrition threat
Correct label: attrition threat
HDEV5
6th EditionSpencer A. Rathus
380 solutions
Consumer Behavior: Buying, Having, Being
13th EditionMichael R Solomon
449 solutions
Social Psychology
10th EditionElliot Aronson, Robin M. Akert, Timothy D. Wilson
525 solutions
Social Psychology
10th EditionElliot Aronson, Robin M. Akert, Samuel R. Sommers, Timothy D. Wilson
525 solutions
Experiments also can be classified by setting. Field experiments and laboratory experiments share the properties of control, randomization, and manipulation and involve the same design characteristics but are conducted in different environments. Laboratory experiments take place in an artificial setting created specifically for the purpose of research. In the laboratory, the researcher has almost total control over the features of the environment, such as temperature, humidity, noise level, and participant conditions. Conversely, field experiments are exactly what the name implies: experiments that take place in a real, pre-existing social setting, such as a hospital or clinic, where the phenomenon of interest usually occurs.
Pre-experimental designs follow similar experimental steps but do not include a control or comparison group. There is only a single group, with no comparison with an equivalent or nonequivalent nontreatment group. Examples are the one-group pretest-posttest (O1 → X → O2) and the one-group posttest-only (X → O1) designs, where X is the treatment or intervention, and O is the data-collection points.
In the one-group pretest-posttest design, data are collected before and after an experimental treatment on this one group of participants. In this type of design, the participants act as their own controls, and no randomization occurs. Because controls and randomization are important characteristics that enhance the internal validity of the study, the evidence generated by the findings of this type of pre-experimental design needs to be interpreted with careful consideration of the design limitations.
The advantage of these designs is that they can be used to evaluate treatments, ruling out ineffective treatments before large-scale experimental or quasiexperimental studies are initiated. The disadvantage of this design is that without a control or comparison group, it is difficult to make any conclusions as to whether the treatment, (X) really caused the outcomes or changes.
various designs for research studies differ in the amount of control the researcher has over the antecedent and intervening variables that may affect the results of the study. True experimental designs, which yield level II evidence, offer the most possibility for control, whereas nonexperimental
designs, which
yield level IV, V, or VI evidence, offer the least. Quasiexperimental designs, which yield level III evidence, offer evidence that lies somewhere in between. Research designs must balance the needs for internal validity and external validity in order to produce useful results. In addition, judicious use of design requires that the chosen design be appropriate to the problem, free of bias, and capable of answering the research question.