RESEARCH PSYCHOLOGY

Research Psychology

Table of Contents

Introduction2

Post-test-Only and the Pretest-Posttest Experimental Designs2

Matched Pair Design3

Selection of Research Participants4

Measuring the Dependent Variable6

Factorial Designs8

Statistical Analysis Employed With a Factorial Experimental Design9

Ethical Considerations of Animal Research10

Aspects of the Experimental Method12

Relationship between Variables14

Types of Reliability16

Types of Validity17

Scales of Measurement18

Example of a Research Project19

Quasi-Experimental Designs20

Descriptive Statistics22

Use of Correlation as a Research Design23

Type I and Type II Errors24

Generalizing the Results of a Research Project26

Goals of Science27

Research Psychology

Introduction

The essay answers the questions related to the research, design, measurements, and scales of the research keeping in view the psychological and social research.

Post-test-Only and the Pretest-Posttest Experimental Designs

The basic randomized experimental design is the randomized post-test- only control group design because of randomly assigned participants to the intervention and control groups, and data being collected only once during the Post-intervention phase. Random assignments produce intervention and control groups that do not differ with respect to characteristics related to the outcome variables of interest. This design can be expanded to more than two groups namely the randomized and non randomized control groups. The randomized pre-test-post-test control group design builds on the randomized post-test control group design by adding pre-intervention measures that allow for explicit evaluation of pre to-post changes, and the use of certain statistical analysis that can take advantage of available baseline data (e.g. analysis of covariance) (Shadish, 2002).

This traditional design is the most widely used an experimental design in drug trials and other clinical research. As with the randomized post-test-only control group design. This design typically controls for threats to internal validity because any existing threats are expected to affect both groups equally. For example, the testing effect may exist which should be present in both the control groups. Therefore, any changes in the intervention group beyond those in the control group would be attributable to the intervention. However, the design does not control for external validity including the testing of variables which can arise because the pretest may sensitize participants to the intervention, producing results that would not be seen in a non-pre-tested situation (Shadish, 2002).

Marczyk (2005) noted that this design commonly controls for threats to internal validity. If testing is likely to be a threat to internal validity or to external validity through the interaction of testing and x, this design should be considered because its lack the pre-test measures that eliminate the testing effect. However, the design does not allow for the explicit evaluation of pre-to-post changes in the dependent variables (although the control group post-test measure may serve as a proxy for the pre-test measures).

Matched Pair Design

Matching can be combined with random assignment. Matched-pairs design refers to which provides a better bet that the groups are equivalent before the introduction of the independent variable manipulation. In a matched-pairs design, researchers create pairs of participants that are matched on a variable that is strongly related to the dependent variable. Researchers then randomly assign one of the pairs to the experimental condition and the other to the control condition (Marczyk, 2005). That is, each member of the pair is assigned ...