Evaluating the Dimensionality of Pornography

Abstract

Pornography may be a construct with a single trait or one with many traits. Research in the past was inconsistent in this regard with most researchers assuming that pornography was unidimensional (with one single trait of pornography). However, the considerable amounts of residual variation found in these studies beyond that explained by the single trait hints at what might be a multidimensional construct (with multiple traits such as sensitization and differentiation). Consequently, in this study, we intended to address the question of whether pornography consisted of a single trait or if it was multidimensional. Using MTurk, 2173 participants from the United States and the Commonwealth of Nations (in which pornography is not strictly illegal) were recruited and asked to rate how pornographic they thought a list of different depictions were. The data were analyzed utilizing the cross-validation procedure in which two subsamples were created from the main sample and one was used to establish the model building and the other to validate the model. Various models, including first-order and higher-order exploratory and confirmatory factor models, were tested. Results indicated that a bi-factor (multidimensional) model generated the best model fit, and that it was most appropriate to consider pornography multidimensional. The final model contained two dimensions ("Sensitization" and "Differentiation"). While sensitization revealed the participants' general tendency to rate all items to be more or less pornographic, differentiation revealed the participants' tendency to differentiate highly pornographic items from less pornographic items. Based on the findings of this study, we suggest that future research on the usage and effects of pornography be conducted while taking into consideration the multidimensional nature of pornography.

Untangling the Porn Web: Creating an Organizing Framework for Pornography Research Among Couples

“…Given the increasing focus put on pornography use from a relational lens, the lack of broader theoretical development in this area is a key concern. Research on pornography use has suffered from well-documented limitations and shortcomings, including inadequate and vague measurement (Busby, Chiu, Olsen, & Willoughby, 2017;, a reliance on non-dyadic datasets and disagreement regarding the mechanisms through which pornography might influence individuals and couples (Leonhardt, Spencer, Butler, & Theobald, 2019a). However, of primary importance is the lack of an overarching conceptual framework or model that can give a common language or organizing structure to the relational work in this area.…”

Development and Testing of the Parent-Child Sex Communication Inventory: A Multidimensional Assessment Tool for Parent and Adolescent Informants

“…First, the adolescent and parent samples were analyzed separately using factor analytic techniques to identify best-fitting solutions for the 46 candidate items. Here, we employed a cross-validation strategy in which each of the samples were randomly split into two halves (see Busby et al, 2017). Sample A was used to conduct preliminary exploratory factor analyses (EFA).…”

Measuring the Multidimensional Construct of Pornography: A Long and Short Version of the Pornography Usage Measure

“…Using participants as the content experts also avoids the potential problem with using scholars as content experts as their extensive experience with pornography research may create a skewed perspective that requires the material to be much more extreme before it is considered pornographic. Recently, some researchers have explored what items representing different types of sexual media most individuals will rate as pornographic, thereby providing a foundation for building a measure that has an important type of content validity from the typical user of sexual media (Busby et al, 2017). By asking participants to determine with very specific scenarios what is pornographic or not, this provides a clear reference point and improved measurement validity.…”

“…A second important issue regarding measurement of any construct is whether the construct is unidimensional or multidimensional (Busby et al, 2017;Hartig & Höhler, 2008). Dimensionality has hardly been a focus of the development of Pornography Usage Measurements in the past.…”

“…The measurement of pornography has suffered from the common pitfalls of other areas in that many of the measures had little or no reliability or validity information provided, and they were often single-item measures asking respondents to estimate how often they used the undefined construct of pornography during a specific time frame (Malamuth, Addison, & Koss, 2000). While a single-item measure of simple behaviors can be sufficient, pornography is not a simple construct or behavior as there are differing views of what constitutes pornography (Willoughby & Busby, 2016), substantial social and cultural reasons to underreport pornography (Rasmussen, Grubbs, Pargament, & Exline, 2018), and at least some researchers have found pornography to be multidimensional (Busby, Chiu, Willoughby, & Olsen, 2017).…”