difference between qt and qnorm Can you tell me what is the difference between qt and qnorm?
From my understanding, qt is used for small sample, and qnorm is use for large sample. Am I correct?
If yes, how do I know whether my samples are large or small?
Thank you.
 A: qt returns quantiles (inverse cdf) of the t-distribution if you specify the tail area (left, by default) and degrees of freedom, while qnorm returns quantiles (inverse cdf) of the standard normal distribution.
Each are used for a wide variety of purposes, though you could use it for calculating critical values. However, it would be more typical to use pt or pnorm to compute p-values instead; indeed t.test and summary.lm and so on calculate the p-value for you.

From my understanding, qt is used for small sample, and qnorm is use for large sample. 

Well, no, not really. If you're doing a t-test there's simply no good reason to use normal functions like qnorm instead of t-functions like qt to obtain a critical value even at a very large sample size. What benefit in using something that may or may not be an adequate approximation when you have the t-calculation right there?
There are times when you would use qnorm -- for example, when you have an asymptotic normal approximation but don't have a t-distribution.
A: Thumb rule is for $n$ greater than 30, you can assume normality. But that might not always be useful. For symmetric data, even smaller sample would do. For highly skewed data, you need larger sample. Plot your data. Look at the histogram. Is it symmetric? Fat tail? Based on these, take your call.
