Test Yourself
True or False
Feedback
False
A CI tells us something about sampl-ING variability, i.e, how much we'd expect the point estimate (eg; sample mean) to vary from sample to sample. It does not tell us about the variability of values in the sample, the sample standard deviation tells us this.
Feedback
False
The 97.5th centile refers to individuals in the observed sample or the observed distribution of values, in contrast the 95% CI for the mean refers to the theoretical sampling distribution of the point estimate (eg; sample mean).
Feedback
True
Feedback
False
This is an approximate 95% reference range (calculated as mean+/-1.96*SD). The 95% CI for a mean is calculated as mean +/- 1.96*standard error. The standard error for the mean in this example is equal to SD/√n = 0.1. So a 95% CI would be 9.8 to 10.2. Completely different!
State whether the following interpretations of a 95% CI are true or false.
Feedback
False
We are 100% confident of this; confidence intervals are built around the sample statistic or point estimate.
Feedback
False
Confidence intervals do not make inference about other samples; they make inference about populations.
Feedback
False
The population parameter is fixed, it is either in the CI or it is not, but we do not know. We’d expect to capture it 95% of the time if we did independent repeat samples and calculated a CI each time.
Feedback
False
The population parameter is fixed; so it is either 100% in the interval or 100% not.
Feedback
True