It's nice to get a headline, even just for a blog, with a former Kiwi PM, a former US Treasury Secretary / Harvard President, a (late) billionaire, and a Kiwi education business, Crimson, in the one line. Robertson, the US investor who died recently & owned Kauri Cliffs up north, was also an investor in Crimson Consulting. The National Business Review is running a story today asking, "Crimson without Julian Robertson: (CEO) Beaton mulls what’s next".
Crimson Education states in its radio advertisements and on its website that the company:
"takes your university application to a new level of personalisation & success, increasing your chances of admission to Ivy League, Stanford, MIT, Oxford & other top-ranked universities by up to 700%". The prior line on their website says: "Crimson Students Are Up To 7x More Likely To Gain Admission Into Their Dream University"
Hang on. Maybe Crimson students are 7x more likely to gain admission to a top university but education economists know that correlation is not the same as causation. Why? Because there are typically "selection biases" when it comes to working out the effect of education on outcomes. In Crimson's case, it's mostly super ambitious, hard-working students from better-off families desperate to get into overseas-universities who buy the services of this company. They would've been more likely than the average person to get accepted even without hiring Crimson.
The key question is - to what extent does Crimson cause the chance of being accepted to increase due to the services which it offers? That is an extremely difficult question to answer. It requires sophisticated econometrics, difficult statistics and tons of data to identify the causal relationship - which must be done to determine whether the claim in the ad is correct - or whether it is just the pre-existing characteristics of the student that matters.
Consequently, shouldn't one query whether the ad reflects the full & fair truth about the value which Crimson adds? Usually one may not make such a comment about a business, but since that business has Sir John Key as one of its main "advisers" (and investors) with former Harvard President, Larry Summers, also featured on its website, I'd like to put this question to Sir John Key and Mr Summers:
"Show us the proof that paying Crimson to help with applications to top overseas universities increases admission chances by 7x. How did you control for selection bias? What regressions did you run to work out the size of this effect?"