* do-file for Supplementary Exercises 6.13 and 6.14 of IPS7e
version 17 /* works also with versions 14-16, see below for version 13 and earlier */

import delimited "R:\Chapter 6\ex06_013.csv"
summarize yield

* 6.13
ztest yield=0, sd(10) level(90)
ztest yield=0, sd(10) level(95)
ztest yield=0, sd(10) level(99)
* the ztest command offers inference for a normal distribution with known sigma
* the hypothesized mean (for a test) is required, but we if we're not interested
* in a test, we can set it at any value and ignore the result
* note: the default confidence level is 95, so the option "level(95)" is not needed

* 6.14
ztesti 50 123.8 10 0, level(95)
* note: we don't have data for this situation, so we use the ztesti command instead
* where the relevant statistics are typed directly 


**************************************************************************
* Stata 13 (and earlier) code

* Stata 13 does not offer confidence intervals with KNOWN sigma.
* One way to trick the program is to pretend the dataset to be huge;
* below we assume 10000 times as many observations, and a standard deviation
* 100 (=sqrt(10000)) times as large, so that the standard deviation of
* the mean is the same.
cii 150000 123.8 1000, level(90)
cii 150000 123.8 1000, level(95)
cii 150000 123.8 1000, level(99)
* As you'll have guessed by now, the cii command with parameters 
* n, mu and sigma, gives a confidence interval for a sample of size n
* from a normal distribution with mean mu and standard deviation sigma.
* Other variants of the command give confidence intervals for other
* situations (later in course).

* 6.14
cii 500000 123.8 1000, level(95)