One Sample Test
Performs a one sample t-test and/or normality test
Move all test variables from the left hand list to the variable list.
One Sample t-test
The one sample t-test tests whether the mean of a population is not equal to a specified value. For example, one might be interested to see if the mean weight change in a group of study participants is equal to zero. This test requires that either the sample size be sufficiently large, or the variable be normally distributed. 'Sufficiently large' in this case is fairly small for most distributions), as convergence of the test statistic to normality occurs rapidly. If outliers are present, this can lead to misleading inferences, so pre-screening of the data is recommended. The options button (the gear) allows for the specification of the alternate hypothesis and the null hypothesis mean.
Shapiro-Wilk Normality Test
This option performs the Shapiro-Wilk test against normality. This test determines if there is enough evidence to conclude that a variable in not normally distributed.
A better alternative to normality testing is to examine the histogram. Is it skewed or roughly symmetrical? Are its tails fat? If your sample is small and your variable is skewed you can try switching to non-parametric methods or transforming the variable so it looks more normal. Alternatively if you can program, you can run simulations to assess the procedure's sensitivity to departures from normality.
Two types of plots can be obtained through this dialog. If 'box plot is unchecked, a histogram is produced, otherwise, box and jitted plots are given. If 'Scale variables' is checked, all variables are standardized.