Statistics Breakdown (Non-Factorial) - Descriptives tab
Select the Descriptives tab of the Statistics Breakdown (non-factorial) dialog box to access options to compute various statistics for each table (all possible combinations of grouping variables) in batch; that is, without requiring any further user input.
Summary
Click this button to display various statistics for the analysis variables on a by-group basis in spreadsheet format.
Among the specific types of statistics is the mean, standard deviation, confidence levels, etc, which you can include in the output spreadsheet by selecting the desired options on this tab.
Compute Statistics
The selections in the group boxes determine the statistics that are reported in the Breakdown Table of Descriptive Statistics that is produced when you click the Summary button (also located on the Statistics Breakdown (non-factorial) dialog box). Note that the inclusion of these statistics is optional. If none is selected, the spreadsheet contains estimates of the mean by default.
Variation, moments
Select the check box next to the statistics you want computed.
The statistics in this group box are used to estimate the variation for the dependent variable for each group as well as a variety of its moments and their standard errors. You can estimate the Standard Deviation, Confidence Interval for Sample Standard Deviation, Coefficient of variation, Variance, Standard error of the mean, and Confidence limits for the mean. When computing confidence levels, you must specify the confidence Interval via the corresponding edit box. Refer to the Introductory Overview for a basic discussion of confidence limits. You can also calculate the Skewness and its Standard error as well as the Kurtosis and its Standard error.
Location, valid N
Select the check box next to the statistics you want computed.
The Mean check box is selected by default. When it is cleared, the mean is not included in the output. The other statistics in this group box are used to estimate the location of the distribution as well as to determine the number of valid cases for the dependent variable for each group (Valid N, % valid N, or Percent of valid observation). You can estimate a variety of location statistics including Sums, % valid sums, Mode, Medians and quartiles, Geometric mean, and Harmonic mean.
Medians and quartiles are only computed for the Summary table of means option.
Percentiles, ranges
Select the check box next to the statistics you want computed.
The options in this group box are used to estimate a variety of percentiles and ranges for the dependent variables for each group. You can select Percentile boundaries (see note below), Range, and Minima and Maxima values. Note that you can specify what percentile boundaries to compute using the First and Second fields.
Note: Percentile boundaries
Use the Boundaries option button (located under the Percentile boundaries check box) to enter which percentiles you want calculated by editing the percentages in the adjacent edit field (the default boundaries are the deciles).
Use the Intervals option button to divide the percentiles evenly among the interval number in the adjacent edit field (the default interval is 10, which would also give the deciles of the data).
Note: computation of median and quartiles (25th, 50th, and 75th percentiles)
When the distribution of values cannot be exactly divided into halves and quartiles, there are different ways to compute the respective median and quartile values (and percentiles).
The specific method of computation for those values can be configured "system-wide" via the Computations of percentiles option in the Options dialog box - Analyses/Graphs: Limits tab.