Undertaking Experiments in Social Sciences: Sequential, Multiple Time Series Designs for Consideration

Abstract

In social sciences, the use of stringent methodological approaches is gaining increasing emphasis. Researchers have recognized the limitations of cross-sectional, nonmanipulative data in the study of causality. True experimental designs, in contrast, are preferred as they represent rigorous standards for achieving causal flows between variables. The Solomon four-group design, for example, is ideal for its positioning to account for, and factor out, confounded influences of predictors on outcomes. However, in daily life settings, it is often difficult to emulate true experimental conditions. Identified limitations include financial resources, logistic difficulties, time constraint, and small sample sizes in social science research settings. There are, of course, other experimental designs that are noteworthy for consideration. Time series and single-case designs, quasi in nature, are effective alternatives for educators and researchers to consider in their research foci. This article examines the different experimental designs that may be implemented in naturalistic classroom settings. In particular, one important inquiry of our theoretical discussion pertains closely to the conceptualization of two innovative designs that we have made, consequently as a result of our research development and examination of the literature: a 'sequential, multiple time series multi-group design and a multi time series, multi-group single-case design'. These experimental designs are innovative and enable comparisons for within and between differences under different experimental conditions.