A Person-Centered Perspective of Death Awareness During the COVID-19 Pandemic

Supportive, resistant, or both? A person-centric view on employee responses to diversity initiatives.

“…We began the profile enumeration step of latent profile analysis (Step 1; Asparouhov & Muthén, 2014) by extracting two profiles and then increasing the number of profiles extracted until the fit statistics and interpretations of new profiles did not improve enough to warrant the inclusion of another latent profile (Nylund et al, 2007). Similar to others (e.g., Chawla et al, 2021;Zhong et al, 2021), we followed recommendations to improve model convergence (Bauer & Curran, 2003;Diallo & Lu, 2017;Mäkikangas et al, 2018) and allowed the mean of each indicator to vary across profiles, but modeled variances and residual variances as fixed. We then continued with profile membership classification (Step 2; Asparouhov & Muthén, 2014) and estimated the counts and proportions for each profile from the posterior distribution, which accounts for error associated with profile classification (Wang & Hanges, 2011).…”

“…Perhaps the most important criterion for determining the number of profiles is that the profiles are not redundant (i.e., they must be distinct from each other) and parsimonious (Zhong et al, 2021). If profiles have a similar theoretical meaning, it could suggest that a smaller profile solution is more appropriate especially when fit statistics also provide support for a more parsimonious solution (Gabriel et al, 2018).…”