Comparing predictive validity of additive versus multiplicative AMO models to explain attitudinal outcomes of job satisfaction

Abstract

Purpose: This study examines the impact of human resources management (HRM) systems on attitudinal outcomes in Spain, comparing the predictive capacity of two alternative individual-level models (additive and multiplicative) with a focus on job satisfaction.

Design/methodology/approach: The analysis draws on data from the Spanish sample of the OECD's Programme for the International Assessment of Adult Competencies (PIAAC). Partial least squares regression (PLS) is applied, along with complementary tests of observed and unobserved heterogeneity and out-of-sample predictive power.

Findings: The models account for up to 10% of the variance in job satisfaction across the whole sample. Ability-enhancing practices have the greatest effect, followed by opportunity-related practices, while motivation contributes minimally. The multiplicative model performs slightly better than the additive one, revealing significant interaction effects among ability, motivation, and opportunity, whose practical relevance future research should further examine. The explanatory power increases slightly when focusing on employees in managerial positions. Accounting for unobserved heterogeneity identifies five employee segments, for which the multiplicative model achieves R² values ranging from 0.42 to 0.76.

Research limitations/implications: PIAAC is a secondary, cross-sectional dataset, limiting analytical flexibility and preventing causal inference.

Practical implications: Understanding how AMO dimensions interact may help HR managers design more effective practice configurations to foster job satisfaction.

Originality/value: Empirical comparisons between additive and multiplicative applications of the AMO model remain scarce. Most prior research focuses on extrinsic motivation practices (such as pay-for-performance) and shows weak links with job satisfaction, while intrinsic motivation is rarely addressed. This study contributes to closing that gap and identifies five employee groups that differ in how they weigh the factors driving job satisfaction.