Academic journal article Fathering

The Influence of Survey Format on Reports about Fathers and Mothers

Academic journal article Fathering

The Influence of Survey Format on Reports about Fathers and Mothers

Article excerpt

Similarities or differences in parenting between reports about mothers and fathers may be associated with the format of the questionnaire. The purpose of this study was to examine three questionnaire formats for collecting the same data about fathers and mothers from adolescents and emerging adults. Study 1 used secondary data analyses from three cross-sectional studies with 820 Latino adolescents using the three survey formats. Study 2 used data from a blind experimental design with 472 emerging adults where the three survey formats were randomly assigned to participants. The analyses from studies with adolescents and emerging adults demonstrated that when the items are asked in the Top Bottom and Left Right formats there are significantly higher correlations between responses about mothers and fathers than when the items are asked in the Separate Pages format.

Keywords: instrumentation threat, method error, mother, father, parenting behaviors

Despite compelling evidence that aspects of parenting are associated with a wide range of youth outcomes, the examination of both fathers' and mothers' parenting is an emerging science (Henry & Hubbs-Tait, in press). Although earlier parenting research focused on mothers as parents, since the mid 1970s scholars increasingly recognized that fathers, as well as mothers, engage in parenting that is integrally involved with child well-being (Palkovitz, 2007; Pleck, 2010). Methodological issues emerge as researchers use assessments of the same parenting variables about mothers and fathers within the same models. Adamsons and Buehler (2007), for example, advocated examination of measurement equivalence before using the same instruments with fathers and mothers to reduce the likelihood of finding exaggerated differences in parenting by mothers and fathers. Another methodological concern affecting the examination of both mothers and fathers using the same instrument is the degree of influence the questionnaire format has on how similarly mothers and fathers are rated by their children. Hence, the purpose of this study is to examine three questionnaire formats for collecting the same data about mothers and fathers from adolescents and emerging adults. This information holds potential to provide important insights into the similarity or differences between perceptions of fathers and mothers, and is of particular relevance to scholars who seek to understand the role of perceived parenting by each parent.


Researchers utilize a range of approaches to investigating youth responses about both mothers and fathers. The early study of fathering focused on variables such as engagement, availability, and responsibility utilizing diaries or observation (Pleck, 2010). Currently, self-report measures (parents' and or children's) are widely used including parental support and control variables that emerged from earlier studies of mothers (Pleck). Given the calls for including both fathers' and mothers' variables within the same research models (e.g., Barber, Stolz, & Olsen, 2005) or controlling for mothers' variables in fathering research (Pleck), a notable challenge is that high correlations may be present between fathers and mothers on the same variables (e.g., Tein, Roosa, & Michaels, 1994). It is likely that mothers and fathers engage in some similar parenting behaviors (i.e., a "true" correlation between reports of mothers and fathers), yet it is also possible that some "spurious" correlation exists due to shared-method variance (a type of error variance).

These correlations run the risk of spurious results due to multicollinearity if reports about mothers and fathers are entered into the same analyses (e.g., ordinary least squares [OLS] regression; Barber, Stolz, & Olsen, 2005). Specifically, researchers may erroneously conclude that some behaviors by mothers or fathers are not important if they are not significant in the OLS regression. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.