ABSTRACT
This chapter examines internal validity, or the degree of confidence researchers can have that the independent variable, rather than some other factor, affected the dependent variable. The distinction between confounds and artifacts is explained, and the time-related threats to internal validity originally identified by Campbell and Stanley are reviewed: history, maturation, testing, instrumentation change, and statistical regression. The chapter also describes the use of control groups in pretest–posttest research. Selection threats to validity are reviewed, including the volunteer bias, the use of pre-existing groups, and mortality. The importance of ensuring that a study has theoretical validity is discussed. The chapter concludes with a discussion of reactivity, demand characteristics, and experimenter expectancies.
