1. There are three categories of quantitative research design: experimental forms, quasiexperimental forms, and descriptive forms.
2. Experimental research is used to establish cause-effect relationships between or among variables, and is most often conducted in a laboratory.
3. In an experiment, the researcher controls the manipulation of the independent variable by randomly assigning participants to treatment or control groups; this ensures that the treatment and control groups are equivalent before any treatment is applied or withheld.
4. Manipulation checks should be conducted to ensure that participants perceived variation in the independent variable as the researcher intended.
5. In the posttest only design, only individuals in the treatment group are exposed to the stimulus.
6. In the pretest-posttest design, only individuals in the treatment group are exposed to the stimulus; the dependent variable is measured for all participants prior to and after the treatment group receives the stimulus.
7. In the factorial design, treatment groups are based on two or more independent variables, with random assignment occurring on one of the variables.
8. The time between the multiple measurements of the dependent variable in a longitudinal design is based on the communication phenomena under study and the theoretical foundation of the study.
9. In quasi-experiments, the researcher uses the natural variation that exists on the independent variable to assign participants to treatment and control conditions.
10. Field experiments are a form of quasiexperimental research design conducted in a naturalistic setting.
11. Descriptive designs are those studies that do not use random assignment of participants or researcher manipulation of the independent variable; as a result of lacking these controls, these research designs cannot demonstrate causation.
12. Communication researchers often use descriptive designs when communication phenomena do not lend themselves to experimental or quasi-experimental designs.
13. All research designs can suffer from bias from researcher effects or procedural bias.
