Hypothesis Tests and Confidence Intervals
The position of the graphically represented keys can be found by moving your mouse on top of the graphic. The top row of function keys is considered to be row 1.
On this page, I will
describe how to do the following functions:
A one-sample t-test
A one-sample z-test
A z-confidence interval
A t-confidence interval
T-tests and Z-tests | |||
Note: a z-test and z-interval are used when the population standard deviation is known. If it not known, the t-test and t-interval are used. The sample standard deviation is computed in lieu of the population standard deviation. Technically, you should only use z-tests with data sets of more than 30 numbers. The examples shown here are thankfully smaller. | |||
T-tests: one variable | |||
The Problem: Given a list of numbers, is the mean of that list significantly different than m0? | |||
The Solution: Press
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
|||
Example: For the following
list of numbers: 7 6 5 7 4 3 4 4 6 5:
Solution:
You should see the following for the first question: Since p<0.05, we reject the null hypothesis and affirm that the mean is not 4. For the second question, we get Since p>0.05, we cannot reject the null. The mean could be 5. |
|||
Z-tests: one variable | |||
The Problem: Given a list of numbers, is the mean of that list significantly different than µ0? | |||
The Solution: Press ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
|||
Examples: The following
list of numbers: 1 9 6 5 3 8 come from a
distribution with s = 2.75.
Solutions:
For question 1, you should get
z=.03449757 and p=.9724804. |
Confidence intervals | ||||
Using the t-distribution | ||||
The Problem: Find an interval for which you can be p% confident that it contains the population mean. | ||||
The Solution: Press ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
||||
Example: For the following
list of numbers: 8 4 3 8 1 2 2 0 0 6, construct a 95% confidence interval. Solution: Compute
the mean and standard deviation. The mean is 3.4. The standard deviation is
3.03. Now press
You should see the following: Critical T=±2.262157 Hence the confidence interval is (1.232, 5.568) |
||||
Using the z-distribution | ||||
The Problem: Find an interval for which you can be p% confident that it contains the population mean. The standard deviation is known to be s. | ||||
The Solution: Press ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
||||
Example: For the following
list of numbers: 8 4 3 8 1 2 2 0 0 6, construct a 95% confidence interval.
Take s to be 3. Solution: Compute
the mean. The mean is 3.4. Now press
You should see the following: Critical Z=±1.959964 |
Other Hp50g Pages:
* Technically, this is the manual for the 49g+, but the commands are the same for both calculators.