How is Blakemore's and Cooper's study low in reliability?

- Doesn't state explicitly the sample size, just the 2 conditions - Ideal to see results of more kittens from each environment so we can be certain we didn’t get fluke results - Low external reliability