# How to use the t-statistic

- Determine the level of significance which is 100% minus the confidence level.
- Determine if it is a one-tail or a two-tail test.
*Confidence Intervals are always two-tailed.* - Determine degrees of freedom which is usually one less than the sample size or (n-k-1).
- Look up the critical t-value
- A
**Standard Confidence Interval**is: sample mean +/- t-critical * standard deviation

**Confidence Interval Around the Mean** requires the use of standard error instead of standard deviation, standard error is calculated as follows:

standard error = standard deviation / √n

Advertisements

## Trackbacks & Pingbacks