Due to uncertainty in economic growth concepts, existence of various proxies for effective factors on economic growth, and lack of indication of the most appropriate econometric model for investigating effective variables on economic growth, the empirical growth regressions have always faced uncertainty. To investigate this phenomenon, we should move from the classical econometrics towards the approaches which can appropriately deal with uncertainty. One of these approaches is the ‘bayesian averaging of classical estimates approach’ which has been used in this article to investigate the efficacy of effective factors on economic growth through 7 different economic growth theories, for 79 countries across the world which are divided into two groups of 'all countries' and 'undeveloped countries' (52 countries) during 1970 and 2006. The research findings indicate that the effective factors on economic growth of different groups of countries can be different from one another. On the same basis, it is not possible to conclude that various economic growth theories for different groups of countries obtain similar results, regarding the effectiveness of proxies on economic growth.