Investigating the Multidimensionality of Need Fulfillment: A Bifactor Exploratory Structural Equation Modeling Representation

Abstract

This article assesses the multidimensionality of the Basic Psychological Need Satisfaction and Frustration Scale (BPNSFS) using bifactor exploratory structural equation modeling (bifactor ESEM). The first study relies on a sample of community adults (N = 2,301), and revealed the superiority of a bifactor ESEM representation, supporting the 6-factor structure of BPNSFS ratings, and the presence of a single continuum of need fulfillment relative to 2 distinct dimensions reflecting need satisfaction and frustration. These results were replicated in a second representative sample of the Hungarian adult population (N = 504), as well as across gender, and found no evidence of differential item functioning as a function of age. Relative to males, females presented higher levels of relatedness satisfaction and lower levels of competence satisfaction. Finally, autonomy frustration decreased with age, whereas competence satisfaction and frustration presented opposite curvilinear tendencies, showing that the fulfillment of this need increased sharply for younger participants, a tendency that became less pronounced with age.

Keywords:

• age
• Basic Psychological Need Satisfaction and Frustration Scale (BPNSFS)
• bifactor
• exploratory structural equation modeling (ESEM)
• gender
• measurement invarianceneed frustration
• need satisfaction

FUNDING

István Tóth-Király was supported by the ÚNKP-16-3 New National Excellence Program of the Ministry of Human Capacities. Gábor Orosz was supported by the Hungarian Research Fund (NKFI PD 116686, FK 124225) and the Hungarian Academy of Sciences (Lendület Project LP2012-36); Emberi Eroforrások Minisztériuma [ÚNKP-16-3 New National Excellence Program].

Supplemental data

Supplemental data for this article can be accessed on the www.tandfonline.com/HSEM.

Notes

¹ We report omega coefficients of composite reliability, calculated as (McDonald, 1970): ω = (Σ|λ_i|)² / ([Σ|λ_i|]² + Σδ_ii) where λ_i are the factor loadings and δ_ii the error variances. Compared to classical estimates of scale score reliability (e.g., α), ω has the advantage of taking into account the strength of association between the items and the latent factors (λ_i), as well as item-specific measurement errors (δ_ii; e.g., Dunn, Baguley, & Brunsden, 2014; Sijtsma, 2009).

² Given that partial measurement invariance is sufficient to support comparisons of results across groups (Byrne, Shavelson, & Muthén, 1989; Steenkamp & Baumgartner, 1998), and the similar gender and age composition of the samples being combined, all items (including Item 24) were retained for these analyses.

Additional information

Funding

István Tóth-Király was supported by the ÚNKP-16-3 New National Excellence Program of the Ministry of Human Capacities. Gábor Orosz was supported by the Hungarian Research Fund (NKFI PD 116686, FK 124225) and the Hungarian Academy of Sciences (Lendület Project LP2012-36); Emberi Eroforrások Minisztériuma [ÚNKP-16-3 New National Excellence Program].