People in tech aren’t necessarily computer science grads

For this reason I wanted to see how important formal education is to break into the tech sector and what common majors people have there.Data setThere were 48788 respondents in this survey in 2018..These are some of the 129 questions in the survey.SurveyHow well can Formal Education predict Salary?To answer this question, I wanted to calculate how much variation in salaries can be explained by Formal Education..The statistical tool for that is know as r squared.There was some data cleaning and data engineering that I had to do first..I dropped the rows with missing salaries from the data set; and I encoded the Formal Education variable..When it comes to categorical variables (i.e. not numbers like 15, or 789, but rather groups, or categories, like bachelor degree, doctoral degree etc.), they must be encoded to be treated quantitatively, as either 1 or 0.Once the data is ready to be processed, I trained the regression on 70% of the data and I tested it on the other 30% of the data..Whenever it comes to models, even as simple as linear regressions, we want to have training and testing sets..The reason is that we do not want the model to fit extremely well on the training set, but fail to predict on other sets..We want the model to capture the overall trend.To my surprise, the r squared value for this model was extremely low, close to 0..This tells us that Formal Education is useless in predicting Salary..In the context of this data set at least, formal education was not a good predictor of salaries at all..This might not sound so strange, when we think that the people in tech break in using non-traditional ways..Many rely on online education, others attend bootcamps, or they get experience via internships.What is the most common undergrad major?I wanted to see what people studied in university to work as developers, or software engineers, or data scientists etc.. More details

Leave a Reply