Confirmatory Factor Analysis for CSR perception on stakeholders in SMEs

Abstract

This is the study done on small and medium sized enterprises (SMEs) who are engaged in the corporate social responsibility (CSR) activities. India is a country of SMEs. It is imperative that India works towards making the smaller enterprises CSR compliant. Employing close to 40% of India’s workforce and contributing 45% to India’s manufacturing output, SMEs play a critical role in economic development and especially generating millions of jobs. We can extrapolate and comfortably say that the geographical reach through SMEs is vastly higher than through the larger enterprises. In India, prior to the Companies Act, 2013, the CSR was more-or-less voluntary activities and philanthropy. (indiacsr.in/csr-an-equal-responsibility-of-smes/)

The objectives of this study are: (a) to assess the perception of CSR initiatives in relation to stakeholders, adopted by the small and medium sized enterprises, leading to an impact on society and (b) build a hypothesized measurement model by using confirmatory factor analysis (CFA) to test statistical significance on CSR perceptions.

AMOS, abbreviated for Analysis of MOment Structures, is a visual program for structural equation modeling and is applied for this study. SEM is a set of statistical techniques that is used to analyze structural relationships. SEM is employed to test a model on observed data from CSR practicing small and medium sized companies of cross-sectors in the vicinity of Bengaluru City, India. For all constrained parameters in the model, AMOS has calculated a modification index. Using modification indices, the model is improved for a better good-fit.

The distinctive feature of CFA is that the factors are from theory, not from statistical results unlike exploratory factor analysis. CFA is conducted with the knowledge about how many factors really exist or which variables belong to a particular construct. The number of factors and the observed variables (indicators) that load on each construct are specified in advance of the analysis. It is helpful in validating a measurement theory.