Abstract
This empirical research conducted four studies in distinct stages to conceptualize, construct and validate a multidimensional reflective-reflective scale of resonant leadership (RL). Study 1, comprising three phases, the researchers mapped dimensions and the nomological network of RL through a systematic literature review applying the ADO (Antecedents-Dimensions-Outcomes) framework. The dimensions were further validated through a focus group comprising experts from industry and academia. In Phase two, items representative of the construct domain were generated and subjected to another focus group discussion to establish face validity. In the third phase, content validity indices at the item-level and scale-level (ICVI and SCVI) and modified kappa were estimated to refine the item pool. Study 2, pilot tested the items on responses from 100 professionals of the Indian IT/ITeS sector. Following data factorability assessment, principal component analysis with varimax rotation was applied for exploratory factor analysis. A 19-item, five-dimensional solution was extracted. Study 3 validated the latent structure through confirmatory factor analysis on another sample of 564 IT/ITeS professionals and empirically corroborated the hierarchical component model. Reliability analyses using Cronbach’s alpha, composite reliability and average variance extracted were established with values well in the recommended range. Tests for common method bias, discriminant, convergent and nomological validities, and tetrad analyses confirmed the robustness of the scale. Study 4 tested the generalizability of the RL scale by cross-validating the psychometric fitness of the measure on 325 professionals working in the Indian BFSI sector. The study confirms teamwork, mindfulness, inclusive decision-making, compassion, and emotional management as the distinct first-order reflective dimensions of RL. The study contributes to literature and practice a unique five-dimensional, 19-item, second-order reflective-reflective scale of WLF, psychometrically fit for use on samples from diverse contexts.