From the article kindly submitted here: "The data comes from an analysis by Payscale, a salary data and software company for individuals and businesses." In other words, the data is correlational, and not based on random assignment to an experimental condition and control condition. So we have no idea what the causation is here. By my observation, the kind of young people who get into private engineering colleges are…
>In other words, the data is correlational, and not based on random assignment to an experimental condition and control condition. If you would like to see studies that use random or quasi-random selection into education, there is a huge literature in labor economics on this topic. For various reasons such literature would never make it onto HN. See a very brief summary here http://www.stlouisfed.org/publications/re/…
Thank you for the link anyway.