Conclusions, limitations and scope for future research
This study has certain limitations which open the avenues for future research. First, it uses cross-sectional research design which puts restriction on the limits to which causality can be established among the variables investigated in the study. Therefore, we encourage future researchers to conduct longitudinal studies to gain better insights into the dynamics of relationship among predictor and outcome variables. Second, the study utilizes experimental research design where prospective employees, who were presented with hypothetical company scenarios, were in the best position to provide ratings on their organizational reputation, environmental values and JPIs. However, the use of self-report measures and single source data collection could artificially inflate or deflate the relationship among the study variables due to common method bias (Podsakoff et al., 2012). Yet, common method variance is less likely to have affected the study results. The common latent factor test performed using confirmatory factor analysis in AMOS 24 also showed lower probability of common method variance as the common latent factor model showed poor fit to the data ( χ 2 /df¼ 3.46, GFI¼ 0.625, NFI¼ 0.631, TLI¼ 0.641, CFI¼ 0.657) and explained only 24 percent of the variance. Additionally, we took a number of precautionary measures such as keeping the survey anonymous and introducing proximal separation between the measures via inter-mixing the items of each of the study constructs (Podsakoff et al., 2012). Though this study was conducted in an economic environment full of job opportunities for the engineering graduates, we recommend future work to examine the relationship among study variables in different job conditions characterized by varying combination of demand and supply in the labor market. The dynamics of the investigated set of relationships may change in adverse job market where there are few job opportunities for the fresh engineering graduates. Although final year undergraduate students may soon be in the labor market for jobs, future studies may replicate the above studies with actual labor pool available in the market for better generalization of the study findings. Future studies may also look for alternative mechanisms which may throw additional light on the linkage among study variables. Although we examined gender and value that individuals place on environment as moderators, future studies should look for other individual difference variables such as age, education, personality factors as moderators to delve deeper into the nature of relationships.
