The Reliability Paradox of the Parent-Child Conflict Tactics Corporal Punishment Subscale

In the present investigation we consider and explain an apparent paradox in the measurement of corporal punishment with the Parent-Child Conflict Tactics Scale (CTS-PC): How can it have poor internal consistency and still be reliable? The CTS-PC was administered to a community sample of 453 opposite...

Full description

Saved in:
Bibliographic Details
Published in:Journal of family psychology 2018-02, Vol.32 (1), p.145-150
Main Authors: Lorber, Michael F, Slep, Amy M. Smith
Format: Article
Language:English
Subjects:
Adult
Child
Child, Preschool
Children
Corporal punishment
Family Conflict
Fathers
Female
Human
Humans
Item Response Theory
Male
Measurement
Measures
Mothers
Parent Child Relations
Parenting - psychology
Parents
Parents & parenting
Psychometrics
Psychometrics - instrumentation
Psychometrics - standards
Punishment
Punishment - psychology
Quantitative psychology
Reliability
Reproducibility of Results
Tactics
Test Reliability
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In the present investigation we consider and explain an apparent paradox in the measurement of corporal punishment with the Parent-Child Conflict Tactics Scale (CTS-PC): How can it have poor internal consistency and still be reliable? The CTS-PC was administered to a community sample of 453 opposite sex couples who were parents of 3- to 7-year-old children. Internal consistency was marginal, yet item response theory analyses revealed that reliability rose sharply with increasing corporal punishment, exceeding .80 in the upper ranges of the construct. The results suggest that the CTS-PC Corporal Punishment subscale reliably discriminates among parents who report average to high corporal punishment (64% of mothers and 56% of fathers in the present sample), despite low overall internal consistency. These results have straightforward implications for the use and reporting of the scale.
ISSN:0893-3200
1939-1293
DOI:10.1037/fam0000307