@guzdial It wasn't published, but we did an internal analysis and found that our first year discrete math (largely proof based) course was a better predictor of upper year GPA than CS1, CS2, Calculus or Linear Algebra
Does anyone have any empirical evidence (not anecdote or personal experience) that being able to construct proofs is predictive of future success in computing courses or career? Or that failing to be able to construct proofs is detrimental to later success? #icer2018