Latent profile analysis of human values: What is the optimal number of clusters?

Abstract

Latent Profile Analysis (LPA) is a method to extract homogeneous clusters characterized by a common response profile. Previous works employing LPA to human value segmentation tend to select a small number of moderately homogeneous clusters based on model selection criteria such as Akaike information criterion, Bayesian information criterion and Entropy. The question is whether a small number of clusters is all that can be gleaned from the data. While some studies have carefully compared different statistical model selection criteria, there is currently no established criteria to assess if an increased number of clusters generates meaningful theoretical insights. This article examines the content and meaningfulness of the clusters extracted using two algorithms: Variational Bayesian LPA and Maximum Likelihood LPA. For both methods, our results point towards eight as the optimal number of clusters for characterizing distinctive Schwartz value typologies that generate meaningful insights and predict several external variables.

Relationship of self-determined motivation with time-related academic behavior in Korean primary school students: A person-centered approach

“…Based on the assertion of Tein et al. (2013) and Schmidt et al. (2021) that the actual number of classes was three to five as a result of analysis of previous studies related to the latent profile, in this study, the number of classes was identified to be from two to five.…”

Processed Dietary Patterns during Pregnancy Are Associated with Low Birth Weight at Term among Women of Advanced and Non-Advanced Age

“…The study tools included: (1) a self-designed structured questionnaire, which comprised demographic information (maternal age and the family's socioeconomic status), pre-pregnancy BMI, and categories of weight gain during pregnancy. Family socioeconomic status was classified into high (41-55), middle (30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40), and low (11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29) groups based on the Hollingshead [20] 2-factor index of social position, which was adapted to Taiwan by Lindsay [21]. Maternal body weight and height were self-reported, and participants' pre-pregnancy body mass index (BMI) was calculated by the researcher.…”

“…The LPA explored the dietary frequencies of each food that create mutually exclusive groups of people with relatively homogeneous food consumption patterns; individuals could only belong to one identified pattern. A combination of the lowest Bayesian information criterion (BIC) and the lowest parametric bootstrapped likelihood ratio test (BLRT) [31] helped to determine which model provided the best fit. Based on the fit index for BIC and BLRT derived from the LPA results, two group SCs (BIC = 34574.87, p = 0.001) were identified.…”

A within-study cross-validation of the values-as-ideals measure: levels of value orientation explain variability in well-being

“…Could latent value profiles provide more meaningful and systematic information regarding the association between values and well-being than value priorities? A recent study makes a step toward this direction ( Schmidt et al., 2021 ) by using Latent Profile Analysis (LPA) to identify clusters that characterize distinctive typologies based on responses on the PVQ. However, the scope of the study is to demonstrate that LPA can extract more than the recommended 3 to 5 number of profiles ( Tein et al., 2013 ).…”