Abstract
We review several methodological practices that are integral to enhancing developmental research in the social sciences. Our primary theme is shifting the dominant statistical and methodological paradigm in the developmental sciences to embrace modern modeling practices. Although null-hypothesis significance testing is unavoidable in some cases, techniques such as latent variable modeling and planned missing data designs can improve research. We also address instances in which only quasi-experiment designs are possible, highlighting three alternative designs: regression discontinuity, propensity score matching, and regression point displacement. Ill-guided and outdated methodological paradigms in research lead to biased results which can lead to biased conclusions.
Acknowledgment
Special thanks to Steven Chesnut, Eugene Wang, and Elizabeth Plowman for their assistance in the initial planning phases.
Funding
This work was supported by the Institute for Measurement, Methodology, Analysis and Policy (Todd D. Little, director) at Texas Tech University, and by a grant from the National Science Foundation (NSF) to Wei Wu and Todd. D. Little.