First of all I have to apologize because my english is horrible.
I am working in a basic and practical statistic course.
I will speak about confidence intervals, of course.
But when I was preparing this chapter I had a doubt: I wonder if it has sence
today to explain the confidence intervals for a mean in terms of the standard
normal distribution.
If we show that the mean follows a normal distribution with variance/n then it
is easy to explain (and to understand) that the confidence interval is due to
two percentiles, the percentile alpha/2 and the percentile 1-alpha/2 from a
normal distribution with the sample mean and the sample variance/n.
I think that the standard gaussian distribution was usefull when we had to
calculate CI in the past.
That is, it was a simple way to have all possibilities in a single table (in
paper).
But now, we have R functions that let us to obtain the exact values for any
normal distribution.
I think that is not necessary to explain how to work with transformations if we
can explain that a CI is simply to identify percentiles.
What do you think?
Thanks
PD I know that the normal distribution for CI need some hipothesis. But I have
focused the question
_________________________
David Monterde
Lo difícil se hace.
Lo imposible se intenta.
_________________________
[[alternative HTML version deleted]]
_______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-teaching
