This is not scientific research, but many folks quote Joel's article on Smart, Get's Stuff Done. http://www.joelonsoftware.com/articles/fog0000000073.html http

This is not scientific research, but many folks quote Joel's article on Smart, Get's Stuff Done. http://www.joelonsoftware.com/articles/fog0000000073.html http://www.joelonsoftware.com/articles/fog0000000073.html Steve McConnell reported on the 10x phenomenon, but again it's not so rigorous that it could pass a Physics journal. http://www.construx.com/10x_Software_Development/Measuring_Productivity_of_Individual_Programmers/ http://www.construx.com/10x_Software_Development/Measuring_P... Now venturing into the realm of survey design... If you were to ask 100 technical programming manager, and 100 banking vice presidents, "Would you be more likely to hire and promote your intern if they had 10 more IQ points, or put in 10 more hours per week" you would get a statistically significant different answer. This is based on a small sample size of personal observations across both industries. In finance and banking there gets to be a point where people are "smart enough". There is enough complexity in software that you can always extract more value from more intelligence.