Hacker News new | past | comments | ask | show | jobs | submit login

Agreed regarding the correlation being mistaken for causation in the article. Although the correlations are still worth noticing.

Where it really impacts the articles credibility is when the author claims the ability to predict relationships:

> By observing these types of interactions, Gottman can > predict with up to 94 percent certainty whether > couples—straight or gay, rich or poor, childless or not— > will be broken up, together and unhappy, or together and > happy several years later. Much of it comes down to the > spirit couples bring to the relationship. Do they bring > kindness and generosity; or contempt, criticism, and > hostility?

Really? 94 percent certainty? I'm willing to say that if you can do that, people will pay you to tell them whether they're going to make it. I suspect the "certainty" is only found when predicting the future of couples who have been together for a few years and have developed these responsive behaviors toward their partners. It's much less useful to be able to "predict" an outcome after you're already committed to a course of action.

The value of being able to "predict" that a car is going to be a lemon diminishes greatly if you can't make the prediction until the car has been bought and driven 10,000 miles.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: