ABSTRACT
A true experiment includes a manipulated independent variable with experimental and control conditions; the characteristics of a good manipulation include validity, reliability, strength, and salience. The chapter describes the importance of controlling extraneous variance by holding potential sources of variance constant and by using appropriate methods of assigning research participants to conditions of the independent variable, such as simple and matched random assignment. The chapter compares and contrasts between- and within-subjects designs, focusing on the strengths and limitations of each and strategies for controlling for order effects in within-subjects designs are described. Multiple group designs—those that use more than two conditions of the independent variable—are explained, as are the distinctions between qualitative and quantitative independent variables. Designs with multiple independent variables, or factorial designs, are reviewed, with a focus on interpreting the results of these designs. The chapter explains how factorial designs can be used to test moderator hypotheses or to detect order effects. The chapter also discusses the role of parametric statistics in experimental design.
