External validity & non-probability sampling

Sir,

Apropos of the article on determinants of Indian physicians satisfaction and dissatisfaction from their job published recently1, the authors deserve credit for their efforts. As the title and the conclusions suggest, the aim was to study the job satisfaction and dissatisfaction levels among Indian physicians and compare it with levels across the world. The authors conclude that “the pattern of high proportion of satisfaction of the Indian physicians reported was similar to the physicians’ satisfaction working particularly in the developed countries”. I have a few concerns here as the methodology used for the purpose of this study may score low on external validity and, therefore, making a conclusion based on non-probability (convenience sampling in this case) sampling may not be correct. This keeping in view the fact that the two institutions chosen for the purpose of this study are not actually representative of institutions across India. One of the institutions chosen is a postgraduate institute only and fully autonomous and the second institution is a central government medical college. Now compare these institutions with other medical institutions across India [private, government (State and central), undergraduate, and undergraduate and postgraduate both, autonomous and non-autonomous, rural and urban] and we can understand the limits in generalizing the findings of this study to medical colleges across India.

The question that needs to be addressed here is to what extent the results of a study conducted in one setting can be generalized to other settings. As pointed above, the situation in this study is not representative of other settings. Further, this representativeness can be expressed in two ways: experimental realism and mundane realism. Experimental realism is the degree to which participants’ psychological experience of a situation is representative of the experience they would have in other situations2. As is obvious, working in the above mentioned institutions is psychologically very different from other institutions across India. Experimental realism asks: Are participants feeling time pressure, or social rejection, or conformity pressure at levels similar to others? Mundane realism is the degree to which the physical setting in a study superficially resembles other physical settings2, again questionable in this study. Though the study has been done wonderfully well in its limited settings but may not be used as a surrogate for Indian physicians.