> It’s a bit like how Gallup can summarise Indonesians’ smartphone habits by calling 1,500 of them; it works fine if you’re looking for a general pattern, but i

> It’s a bit like how Gallup can summarise Indonesians’ smartphone habits by calling 1,500 of them; it works fine if you’re looking for a general pattern, but it might skew the data if you’re looking for data about a tiny niche of smartphone users or if Gallup happened to call up relatively too many Nokia users that day. This is an incorrect statement and interpretation of how statistics work. Is there a chance that the 1,500 Indonesians that they call that day not representative of the over population? Of course, but the probability of that is very low. This concept specifically is known as statistical significance[0]. The conclusion is: sampling error can lead to incorrect conclusions, but if you can eliminate any biases to your sampling, then it can indeed be representative. Personally, the more important take away is this: before you start deriving conclusions from your metrics, it's necessary to fully grok the concept of statistics. [0] - http://en.wikipedia.org/wiki/Statistical_significance http://en.wikipedia.org/wiki/Statistical_significance