Please enable JavaScript.
Coggle requires JavaScript to display documents.
EXPERIMENTAL RESEARCH, Chapter 10, Field experiments: Performed in a field…
EXPERIMENTAL RESEARCH
Two-group experimental designs: Involves one treatment and one control group. Ideal for testing the effects of a single independent variable that can be manipulated as a treatment.
Pretest-posttest control group design: Participants that are randomly assigned to treatment and controls groups take a pre-test measuring the dependent variable of interest. The treatment groups is administered a treatment and afterward the dependent variable is assessed through a post test..
Posttest-only control group design: Similar to the pretest-post test, but without the initial pretest measurement. with omission of pretest.
Covariance designs=Concomitant variable design: When dependent variables are influenced by covariates rather than dependent variables
-
Key Concepts
Treatment and control groups: When one group of participants receives the experimental treatment, while the control group does not receive any treatment.
-
-
Factorial designs: Manipulating four or higher group designs. Allows researcher to evaluate the separate effects of each treatment on the dependent variables, as well as their interaction.
Main effect: When a dependent variable shows a significant difference between multiple levels of one factor, at all levels of other factors
Quasi-experimental designs: Inferior to true experimental designs because it lacks random assignment and due to selection related threats.
Non-equivalent group design: a quasi-equivalent version of pretest-posttest control group
non-equivalent switched replication design: quasi-experimental version of switched replication design
Regression discontinuity design: A non-equivalent pretest-posttest design where participants are assigned to the treatment or control group based on a cut-off score on a preprogram measure.
What is experimental research?: The most rigorous of all research designs. Independent variables are manipulated, participants are randomly assigned to different treatments, and results of the dependent variables are observed. Links cause and effect through changing treatments and controlling for the effects of extraneous variables.Uses of experimental research: Used in explanatory research because the goal is to examine cause-effect relationshipsPerils of experimental research: Experimental research should be used with a multitude of methodological problems. 1. Requires theories for framing hypotheses current research lacks this.
- Instruments not testing for reliability and validity. 3. A lot of the research uses inappropriate designs 4. treatment may be diverse, incomparable, and inconsistent across studies.
NOTE: Internal validity=causality
External validity=generalizability
-
Field experiments: Performed in a field setting and are high in both internal and external validity.
Treatment manipulation: Treatments design that is unique from all other research methods. This controls for cause-effect relationships. Must have pre-test measures or post-test measures prior to start of study.
Random selection and assignment: When a participant is randomly drawn from the population or sampling frame. This concept ensures treatment groups are equivalent to each other and the control group prior to treatment.
History threat: Dependent variables are influenced by historical events instead of he experimental treatment.
Maturation threat: Dependent variables are influenced by maturation of subjects rather than experimental treatment. (e.g., improvement of intellectual ability to understand complex subjects
Instrumentation threat: The difference between pre-and post test designs were manipulated such that the degree of difficulty is not the same as the pre-test.
Testing threat: When post-test responses are influenced by pretest responses. This can be avoided by eliminating the pre-test
-
Regression threat: The group's overall scores tend to move closer to the average on the post-test instead of changing as expected.
Interaction effect: Occurs when the impact of one factor various depending on the level of another factor
Proxy pretest design: When a researcher tests the efficacy of a program after the program has started and pretest data isn't available.
Separate pretest-posttest samples designs: Not a strong design because its used when pretest and posttest data is not available from the same participant.
Non-equivalent dependent variable design: Single-group pre-post quasi-experimental design with tow outcome measures, where one is theoretically expected to be influenced by the treatment and the other is not.
-