Table 1.2 on page 6 lists statistical procedures according to task. Table A1.1 on page 1355 lists the most common statistics and the procedures that compute them.
To produce | Use this procedure | Which |
---|---|---|
Descriptive statistics | CORR | computes simple descriptive statistics. |
MEANS or SUMMARY | computes descriptive statistics; can produce printed output and output data sets. By default, PROC MEANS produces printed output and PROC SUMMARY creates an output data set. | |
REPORT | computes most of the same statistics as PROC TABULATE; allows customization of format. | |
SQL | computes descriptive statistics for data in one or more DBMS tables; can produce a printed report or create a SAS data set. | |
TABULATE | produces tabular reports for descriptive statistics; can create an output data set. | |
UNIVARIATE | computes the broadest set of descriptive statistics; can create an output data set. | |
Frequency and cross-tabulation tables | FREQ | produces one-way to n -way tables; reports frequency counts; computes chi-square tests; computes tests and measures of association and agreement for two-way to n -way cross-tabulation tables; can compute exact tests and asymptotic tests; can create output data sets. |
TABULATE | produces one-way and two-way cross-tabulation tables; can create an output data set. | |
UNIVARIATE | produces one-way frequency tables. | |
Correlation analysis | CORR | computes Pearson s, Spearman s, and Kendall s correlations and partial correlations ; also computes Hoeffding s D and Cronbach s coefficient alpha. |
Distribution analysis | UNIVARIATE | computes tests for location and tests for normality. |
FREQ | computes a test for the binomial proportion for one-way tables; computes a goodness-of-fit test for one-way tables; computes a chi-square test of equal distribution for two-way tables. | |
Robust estimation | UNIVARIATE | computes robust estimates of scale, trimmed means, and Winsorized means. |
Data transformation | ||
Computing ranks | RANK | computes ranks for one or more numeric variables across the observations of a SAS data set and creates an output data set; can produce normal scores or other rank scores. |
Standardizing data | STANDARD | creates an output data set that contains variables that are standardized to a given mean and standard deviation. |
Low-resolution graphics [*] | ||
CHART | produces a graphical report that can show one of the following statistics for the chart variable: frequency counts, percentages, cumulative frequencies, cumulative percentages, totals, or averages. | |
UNIVARIATE | produces descriptive plots such as stem and leaf, box plot, and normal probability plot. | |
[*] To produce high-resolution graphical reports, use SAS/GRAPH software. |
For a large sample size n , the calculation of quantiles, including the median, requires computing time proportional to n log( n ). Therefore, a procedure, such as UNIVARIATE, that automatically calculates quantiles may require more time than other data summarization procedures. Furthermore, because data is held in memory, the procedure also requires more storage space to perform the computations . By default, the report procedures PROC MEANS, PROC SUMMARY, and PROC TABULATE require less memory because they do not automatically compute quantiles. These procedures also provide an option to use a new fixed-memory quantiles estimation method that is usually less memory intense . See Quantiles on page 555 for more information.
To compute statistics for several groups of observations, you can use any of the previous procedures with a BY statement to specify BY-group variables. However, BY- group processing requires that you previously sort or index the data set, which for very large data sets may require substantial computer resources. A more efficient way to compute statistics within groups without sorting is to use a CLASS statement with one of the following procedures: MEANS, SUMMARY, or TABULATE.
Appendix 1, SAS Elementary Statistics Procedures, on page 1353 lists standard keywords, statistical notation, and formulas for the statistics that base SAS procedures compute frequently. The individual statistical procedures discuss the statistical concepts that are useful to interpret the output of a procedure.