5.2.3 Hypothesis Tests (Means and Sigmas) - tests concerning the difference between two means and standard deviations


This statlet tests hypotheses about the means and standard deviations of two normal distributions. The tabs are:

Input

t test

Power curve

F test

Power curve


Input

To use this statlet, enter information about your data:

Sample labels - names to be displayed on the output.

Sample sizes - the number of observations in the samples.

Sample means - the means of the samples.

Sample standard deviations - the standard deviations of the samples.

Back to Top


t test

This tab shows the result of a hypothesis test concerning the difference between the means:

It shows:

Summary statistics.

A confidence interval for the difference between the means.

The result of a t-test run to test a hypothesis about the difference between the population means.

Use the Options button to specify the hypotheses to be tested. If the P-value is less than the alpha risk which you specify, you should reject the null hypothesis at the corresponding significance level.

Options button

Enter:

Null hypothesis - the value of the mean difference specified by the null hypothesis.

Alt. Hypothesis - select a two-sided test (~=) or a one-sided test.

Alpha - the probability of a Type I error, which is a situation where a true null hypothesis is incorrectly rejected. Typical values for alpha are 10%, 5%, and 1%. The confidence interval is also affected by this setting and uses a confidence level equal to (100-alpha)%.

Assume equal sigmas - whether to assume that the samples come from populations with the same standard deviation.

Back to Top


Power curve

This tab shows the power curve for the t test:

The power curve shows the probability of rejecting the null hypothesis as a function of the true difference between the population means.

Options button

Same as previous tab.

Back to Top


F test

This tab shows the result of a hypothesis test concerning the ratio of the variances:

It shows:

Summary statistics.

A confidence interval for the ratio of the variances.

The result of an F test run to test a hypothesis about the ratio of the population variances.

Use the Options button to specify the hypotheses to be tested. If the P-value is less than the alpha risk which you specify, you should reject the null hypothesis at the corresponding significance level.

Options button

Enter:

Null hypothesis - the value of the ratio specified by the null hypothesis (usually 1.0).

Alt. Hypothesis - select a two-sided test (~=) or a one-sided test.

Alpha - the probability of a Type I error, which is a situation where a true null hypothesis is incorrectly rejected. Typical values for alpha are 10%, 5%, and 1%. The confidence interval is also affected by this setting and uses a confidence level equal to (100-alpha)%.

Back to Top


Power curve

This tab shows the power curve for the F test:

The power curve shows the probability of rejecting the null hypothesis as a function of the true ratio of the population variances.

Options button

Same as previous tab.

Back to Top