Paired t

Graphs - Histogram of Differences

  

The histogram illustrates the following properties of your data:

·    Shape. As a requirement for the t-procedures, the differences should be normally distributed, therefore you should ensure that the histogram is roughly symmetric and bell-shaped. You can also conduct a normality test to determine if the differences are normally distributed.

·    Mean. The point below the histogram labeled image represents the mean difference between paired observations, which serves as your estimate of m D.

·    Confidence Interval. The line and brackets on either side of the mean represent the confidence interval.

·    Reference Value. The point labeled H0 represents the value for m D specified in the null hypothesis.

Example Output

image\pair_1n.gif

Interpretation

The histogram of the heart rate differences seems to indicate that they are roughly normally distributed. (It is often difficult to evaluate normality when the sample size is small.)

The graph also illustrates that m D is likely to be greater than the reference value specified in H0, since this value (0) is smaller than those within the confidence interval.