Behavior Genetic Research Methods: Testing Quasi-Causal Hypotheses Using Multivariate Twin Data

Citation

Turkheimer, Eric & Harden, K. Paige (2014). Behavior Genetic Research Methods: Testing Quasi-Causal Hypotheses Using Multivariate Twin Data. In Judd, Charles M. & Reis, Harry T. (Eds.), Handbook of Research Methods in Social and Personality Psychology (pp. 159-187). Cambridge: Cambridge University Press.

Abstract

Research design is inextricably linked to data analysis. Research designs can be divided into three fundamental categories: experimental design, quasi-experimental design, and nonexperimental or passive observational design. This chapter focuses almost exclusively on experimental design, for two main reasons. In an experimental design there are always at least two factors such as fixed and random factors or independent variables, variables that are considered to influence the dependent variable response. The chapter discusses a series of questions that must be answered regarding each individual factor in a design. It considers the possible ways that multiple factors included in a single experiment may be interrelated. Multiple factors in a design can be crossed, nested, or confounded. Issues involving the dependent measures in a study are also part of experimental design, and these choices can influence statistical conclusion validity (power), internal validity, construct validity, and external validity.

URL

https://doi.org/10.1017/CBO9780511996481.012

Keyword(s)

construct validity

Reference Type

Book Chapter

Book Title

Handbook of Research Methods in Social and Personality Psychology

Author(s)

Turkheimer, Eric
Harden, K. Paige

Editor(s)

Judd, Charles M.
Reis, Harry T.

Year Published

2014

Pages

159-187

Edition

2

Publisher

Cambridge University Press

City of Publication

Cambridge

ISSN/ISBN

9781107011779

DOI

10.1017/CBO9780511996481.012

Reference ID

5868