# The reflective measurement model of adherence to non-pharmaceutical interventions (NPIs) in accordance with normalization process theory (NPT) in coherent and convenient social subgroups: PLS-SEM analysis

## TL;DR

This study evaluates how well people follow non-pharmaceutical interventions during the pandemic, using a model based on normalization process theory in different social groups.

## Contribution

The paper introduces a novel reflective measurement model for adherence to NPIs using PLS-SEM in distinct social subgroups.

## Key findings

- Coherent subgroups showed better integration of NPIs into daily life.
- Higher education and younger age correlated with external locus of control and adherence to NPIs.
- The reflective model demonstrated excellent goodness-of-fit and inverse relationships between key factors.

## Abstract

Non-pharmaceutical interventions (NPIs) decrease COVID-19 transmission. Reliability and validity of adherence to NPIs in accordance with normalization process theory (NPT) in coherent and convenient social subgroups using reflective measurement model assessment has not been evaluated.

In February 2021, a sample of medical students and people with substance use disorders in treatment as coherent (based on continuous probability distribution) vs. convenient groups (based on convenience, not equal probability) composed of travellers and COVID-19 suspected persons from Split-Dalmatia County (SDC) (n = 656) in the Mediterranean completed self-administered surveys. Partial least squares structural equation modelling (PLS-SEM) was used to measure reflective model assessment of adherence to NPIs according to NPT.

PLS-SEM reflective model assessment provided two-group specific factors in inverse relationships which determined adherence to NPIs with excellent goodness-of-fit [χ2 = 1.292, df = 1; P = 0.297, CFI = 1, TLI = 0.997, RMSEA = 0.011 (90% CI 0–0.105), RMSEA P = 0.604, SRMR = 0.008, Hoelter CN (α = 0.05) = 2322.757]. Significant negative factors covariance estimate (−0.716) revealed an inverse relationship between first (adherence to NPIs and internal locus of control (LoC) (0.640)) and second factor; young adulthood age (≤25) and highest level of education (1362). As the first factor increased the second tended to decrease. LoC is expected potential mechanism by which sex (MLsex = −0.017, SE = 0.007, P < 0.016) and belonging to coherent subgroups (MLgroup = −0.008, SE = 0.003, P = 0.015) can produce indirect effect of adherence to NPIs.

Coherent subgroups had a more pronounced tendency toward integration of NPIs in everyday life. Group factors that facilitate the normalization were higher educated younger adults with a tendency toward external LoC.