Chapter 10: The Nonexperimental and Quasi-Experimental Strategies: Nonequivalent Group, Pre–Post, and Developmental Designs

The core distinction centers on the absence of random assignment and direct manipulation of independent variables in these approaches, which prevents researchers from establishing definitive causal relationships between variables. The chapter systematically covers nonequivalent group designs, including differential research examining existing groups, posttest-only designs that measure outcomes without baseline data, and pretest-posttest control group designs that attempt to establish equivalence through measurement. Pre-post designs are explored through one-group pretest-posttest formats and time-series approaches, which track changes across multiple time points. A central theme involves understanding internal validity threats specific to these designs, such as individual differences between groups, historical confounds, maturation effects, and order effects that can rival alternative explanations for observed outcomes. Quasi-experimental designs attempt to minimize these threats through strategic design choices, whereas nonexperimental designs typically remain vulnerable to multiple confounds. The chapter addresses developmental research designs, contrasting cross-sectional approaches that sample different age cohorts simultaneously with longitudinal designs that follow the same participants over extended periods. Cohort effects and generation effects emerge as important considerations when interpreting developmental data, and participant attrition in longitudinal studies presents practical challenges. The chapter integrates relevant statistical techniques for analyzing data from these designs, including t-tests for group comparisons, chi-square tests for categorical data, repeated-measures analysis of variance for within-subject changes, and mixed-design analysis of variance when both between-group and within-subject factors are present. The concept of quasi-independent variables—variables that cannot be manipulated but function as independent variables in the analysis—receives particular attention as essential to understanding the limitations and possibilities of these research strategies.