Premium
Using nonequivalent dependent variables to reduce internal validity threats in quasi‐experiments: Rationale, history, and examples from practice
Author(s) -
Coryn Chris L.S.,
Hobson Kristin A.
Publication year - 2011
Publication title -
new directions for evaluation
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.374
H-Index - 40
eISSN - 1534-875X
pISSN - 1097-6736
DOI - 10.1002/ev.375
Subject(s) - internal validity , external validity , dilemma , psychological intervention , mechanism (biology) , inclusion (mineral) , variables , psychology , computer science , risk analysis (engineering) , social psychology , management science , medicine , economics , statistics , epistemology , mathematics , philosophy , psychiatry , machine learning
Threats to the validity of inferences and conclusions regarding the effects of applied interventions have been a major dilemma for social scientists and evaluators for several decades. One mechanism for reducing threats to internal validity and improving warrants for cause‐and‐effect conclusions in nonrandomized investigations and evaluations is the inclusion of nonequivalent dependent variables as an element of structural design. In this chapter, the rationale for, history of, and examples from practice for using nonequivalent dependent variables to reduce internal validity threats, as well as some warrants supporting their increased use, are described. © Wiley Periodicals, Inc., and the American Evaluation Association