4 This chapter documents the statistical procedures that PSPP supports so
8 * DESCRIPTIVES:: Descriptive statistics.
9 * FREQUENCIES:: Frequency tables.
10 * EXAMINE:: Testing data for normality.
11 * CORRELATIONS:: Correlation tables.
12 * CROSSTABS:: Crosstabulation tables.
13 * FACTOR:: Factor analysis and Principal Components analysis
14 * NPAR TESTS:: Nonparametric tests.
15 * T-TEST:: Test hypotheses about means.
16 * ONEWAY:: One way analysis of variance.
17 * RANK:: Compute rank scores.
18 * REGRESSION:: Linear regression.
19 * RELIABILITY:: Reliability analysis.
20 * ROC:: Receiver Operating Characteristic.
30 /MISSING=@{VARIABLE,LISTWISE@} @{INCLUDE,NOINCLUDE@}
31 /FORMAT=@{LABELS,NOLABELS@} @{NOINDEX,INDEX@} @{LINE,SERIAL@}
33 /STATISTICS=@{ALL,MEAN,SEMEAN,STDDEV,VARIANCE,KURTOSIS,
34 SKEWNESS,RANGE,MINIMUM,MAXIMUM,SUM,DEFAULT,
35 SESKEWNESS,SEKURTOSIS@}
36 /SORT=@{NONE,MEAN,SEMEAN,STDDEV,VARIANCE,KURTOSIS,SKEWNESS,
37 RANGE,MINIMUM,MAXIMUM,SUM,SESKEWNESS,SEKURTOSIS,NAME@}
41 The @cmd{DESCRIPTIVES} procedure reads the active file and outputs
43 statistics requested by the user. In addition, it can optionally
46 The VARIABLES subcommand, which is required, specifies the list of
47 variables to be analyzed. Keyword VARIABLES is optional.
49 All other subcommands are optional:
51 The MISSING subcommand determines the handling of missing variables. If
52 INCLUDE is set, then user-missing values are included in the
53 calculations. If NOINCLUDE is set, which is the default, user-missing
54 values are excluded. If VARIABLE is set, then missing values are
55 excluded on a variable by variable basis; if LISTWISE is set, then
56 the entire case is excluded whenever any value in that case has a
57 system-missing or, if INCLUDE is set, user-missing value.
59 The FORMAT subcommand affects the output format. Currently the
60 LABELS/NOLABELS and NOINDEX/INDEX settings are not used. When SERIAL is
61 set, both valid and missing number of cases are listed in the output;
62 when NOSERIAL is set, only valid cases are listed.
64 The SAVE subcommand causes @cmd{DESCRIPTIVES} to calculate Z scores for all
65 the specified variables. The Z scores are saved to new variables.
66 Variable names are generated by trying first the original variable name
67 with Z prepended and truncated to a maximum of 8 characters, then the
68 names ZSC000 through ZSC999, STDZ00 through STDZ09, ZZZZ00 through
69 ZZZZ09, ZQZQ00 through ZQZQ09, in that sequence. In addition, Z score
70 variable names can be specified explicitly on VARIABLES in the variable
71 list by enclosing them in parentheses after each variable.
73 The STATISTICS subcommand specifies the statistics to be displayed:
77 All of the statistics below.
81 Standard error of the mean.
87 Kurtosis and standard error of the kurtosis.
89 Skewness and standard error of the skewness.
99 Mean, standard deviation of the mean, minimum, maximum.
101 Standard error of the kurtosis.
103 Standard error of the skewness.
106 The SORT subcommand specifies how the statistics should be sorted. Most
107 of the possible values should be self-explanatory. NAME causes the
108 statistics to be sorted by name. By default, the statistics are listed
109 in the order that they are specified on the VARIABLES subcommand. The A
110 and D settings request an ascending or descending sort order,
120 /FORMAT=@{TABLE,NOTABLE,LIMIT(limit)@}
121 @{STANDARD,CONDENSE,ONEPAGE[(onepage_limit)]@}
123 @{AVALUE,DVALUE,AFREQ,DFREQ@}
126 /MISSING=@{EXCLUDE,INCLUDE@}
127 /STATISTICS=@{DEFAULT,MEAN,SEMEAN,MEDIAN,MODE,STDDEV,VARIANCE,
128 KURTOSIS,SKEWNESS,RANGE,MINIMUM,MAXIMUM,SUM,
129 SESKEWNESS,SEKURTOSIS,ALL,NONE@}
131 /PERCENTILES=percent@dots{}
132 /HISTOGRAM=[MINIMUM(x_min)] [MAXIMUM(x_max)]
133 [@{FREQ,PCNT@}] [@{NONORMAL,NORMAL@}]
134 /PIECHART=[MINIMUM(x_min)] [MAXIMUM(x_max)] @{NOMISSING,MISSING@}
136 (These options are not currently implemented.)
142 The @cmd{FREQUENCIES} procedure outputs frequency tables for specified
144 @cmd{FREQUENCIES} can also calculate and display descriptive statistics
145 (including median and mode) and percentiles.
147 @cmd{FREQUENCIES} also support graphical output in the form of
148 histograms and pie charts. In the future, it will be able to produce
149 bar charts and output percentiles for grouped data.
151 The VARIABLES subcommand is the only required subcommand. Specify the
152 variables to be analyzed.
154 The FORMAT subcommand controls the output format. It has several
159 TABLE, the default, causes a frequency table to be output for every
160 variable specified. NOTABLE prevents them from being output. LIMIT
161 with a numeric argument causes them to be output except when there are
162 more than the specified number of values in the table.
165 STANDARD frequency tables contain more complete information, but also to
166 take up more space on the printed page. CONDENSE frequency tables are
167 less informative but take up less space. ONEPAGE with a numeric
168 argument will output standard frequency tables if there are the
169 specified number of values or less, condensed tables otherwise. ONEPAGE
170 without an argument defaults to a threshold of 50 values.
173 LABELS causes value labels to be displayed in STANDARD frequency
174 tables. NOLABLES prevents this.
177 Normally frequency tables are sorted in ascending order by value. This
178 is AVALUE. DVALUE tables are sorted in descending order by value.
179 AFREQ and DFREQ tables are sorted in ascending and descending order,
180 respectively, by frequency count.
183 SINGLE spaced frequency tables are closely spaced. DOUBLE spaced
184 frequency tables have wider spacing.
187 OLDPAGE and NEWPAGE are not currently used.
190 The MISSING subcommand controls the handling of user-missing values.
191 When EXCLUDE, the default, is set, user-missing values are not included
192 in frequency tables or statistics. When INCLUDE is set, user-missing
193 are included. System-missing values are never included in statistics,
194 but are listed in frequency tables.
196 The available STATISTICS are the same as available in @cmd{DESCRIPTIVES}
197 (@pxref{DESCRIPTIVES}), with the addition of MEDIAN, the data's median
198 value, and MODE, the mode. (If there are multiple modes, the smallest
199 value is reported.) By default, the mean, standard deviation of the
200 mean, minimum, and maximum are reported for each variable.
203 PERCENTILES causes the specified percentiles to be reported.
204 The percentiles should be presented at a list of numbers between 0
206 The NTILES subcommand causes the percentiles to be reported at the
207 boundaries of the data set divided into the specified number of ranges.
208 For instance, @code{/NTILES=4} would cause quartiles to be reported.
210 The HISTOGRAM subcommand causes the output to include a histogram for
211 each specified numeric variable. The X axis by default ranges from the
212 minimum to the maximum value observed in the data, but the MINIMUM and
213 MAXIMUM keywords can set an explicit range. The Y axis by default is
214 labeled in frequencies; use the PERCENT keyword to causes it to be
215 labeled in percent of the total observed count. Specify NORMAL to
216 superimpose a normal curve on the histogram.
217 Histograms are not created for string variables.
219 The PIECHART adds a pie chart for each variable to the data. Each
220 slice represents one value, with the size of the slice proportional to
221 the value's frequency. By default, all non-missing values are given
222 slices. The MINIMUM and MAXIMUM keywords can be used to limit the
223 displayed slices to a given range of values. The MISSING keyword adds
224 slices for missing values.
227 @comment node-name, next, previous, up
231 @cindex Normality, testing for
235 VARIABLES=var_list [BY factor_list ]
236 /STATISTICS=@{DESCRIPTIVES, EXTREME[(n)], ALL, NONE@}
237 /PLOT=@{BOXPLOT, NPPLOT, HISTOGRAM, ALL, NONE@}
239 /COMPARE=@{GROUPS,VARIABLES@}
242 /PERCENTILE=[value_list]=@{HAVERAGE, WAVERAGE, ROUND, AEMPIRICAL, EMPIRICAL @}
243 /MISSING=@{LISTWISE, PAIRWISE@} [@{EXCLUDE, INCLUDE@}]
244 [@{NOREPORT,REPORT@}]
248 The @cmd{EXAMINE} command is used to test how closely a distribution is to a
249 normal distribution. It also shows you outliers and extreme values.
251 The VARIABLES subcommand specifies the dependent variables and the
252 independent variable to use as factors for the analysis. Variables
253 listed before the first BY keyword are the dependent variables.
254 The dependent variables may optionally be followed by a list of
255 factors which tell PSPP how to break down the analysis for each
256 dependent variable. The format for each factor is
262 The STATISTICS subcommand specifies the analysis to be done.
263 DESCRIPTIVES will produce a table showing some parametric and
264 non-parametrics statistics. EXTREME produces a table showing extreme
265 values of the dependent variable. A number in parentheses determines
266 how many upper and lower extremes to show. The default number is 5.
269 The PLOT subcommand specifies which plots are to be produced if any.
271 The COMPARE subcommand is only relevant if producing boxplots, and it is only
272 useful there is more than one dependent variable and at least one factor. If
273 /COMPARE=GROUPS is specified, then one plot per dependent variable is produced,
274 containing boxplots for all the factors.
275 If /COMPARE=VARIABLES is specified, then one plot per factor is produced, each
276 each containing one boxplot per dependent variable.
277 If the /COMPARE subcommand is ommitted, then PSPP uses the default value of
280 The ID subcommand also pertains to boxplots. If given, it must
281 specify a variable name. Outliers and extreme cases plotted in
282 boxplots will be labelled with the case from that variable. Numeric or
283 string variables are permissible. If the ID subcommand is not given,
284 then the casenumber will be used for labelling.
286 The CINTERVAL subcommand specifies the confidence interval to use in
287 calculation of the descriptives command. The default it 95%.
290 The PERCENTILES subcommand specifies which percentiles are to be calculated,
291 and which algorithm to use for calculating them. The default is to
292 calculate the 5, 10, 25, 50, 75, 90, 95 percentiles using the
295 The TOTAL and NOTOTAL subcommands are mutually exclusive. If NOTOTAL
296 is given and factors have been specified in the VARIABLES subcommand,
297 then then statistics for the unfactored dependent variables are
298 produced in addition to the factored variables. If there are no
299 factors specified then TOTAL and NOTOTAL have no effect.
302 If many dependent variable are given, or factors are given for which
303 there are many distinct values, then @cmd{EXAMINE} will produce a very
304 large quantity of output.
307 @section CORRELATIONS
312 /VARIABLES = varlist [ WITH varlist ]
317 /VARIABLES = varlist [ WITH varlist ]
318 /VARIABLES = varlist [ WITH varlist ]
321 [ /PRINT=@{TWOTAIL, ONETAIL@} @{SIG, NOSIG@} ]
322 [ /STATISTICS=DESCRIPTIVES XPROD ALL]
323 [ /MISSING=@{PAIRWISE, LISTWISE@} @{INCLUDE, EXCLUDE@} ]
327 The @cmd{CORRELATIONS} procedure produces tables of the Pearson correlation coefficient
328 for a set of variables. The significance of the coefficients are also given.
330 At least one VARIABLES subcommand is required. If the WITH keyword is used, then a non-square
331 correlation table will be produced.
332 The variables preceding WITH, will be used as the rows of the table, and the variables following
333 will be the columns of the table.
334 If no WITH subcommand is given, then a square, symmetrical table using all variables is produced.
337 The @cmd{MISSING} subcommand determines the handling of missing variables.
338 If INCLUDE is set, then user-missing values are included in the
339 calculations, but system-missing values are not.
340 If EXCLUDE is set, which is the default, user-missing
341 values are excluded as well as system-missing values.
344 If LISTWISE is set, then the entire case is excluded from analysis
345 whenever any variable specified in any @cmd{/VARIABLES} subcommand
346 contains a missing value.
347 If PAIRWISE is set, then a case is considered missing only if either of the
348 values for the particular coefficient are missing.
349 The default is PAIRWISE.
351 The PRINT subcommand is used to control how the reported significance values are printed.
352 If the TWOTAIL option is used, then a two-tailed test of significance is
353 printed. If the ONETAIL option is given, then a one-tailed test is used.
354 The default is TWOTAIL.
356 If the NOSIG option is specified, then correlation coefficients with significance less than
357 0.05 are highlighted.
358 If SIG is specified, then no highlighting is performed. This is the default.
361 The STATISTICS subcommand requests additional statistics to be displayed. The keyword
362 DESCRIPTIVES requests that the mean, number of non-missing cases, and the non-biased
363 estimator of the standard deviation are displayed.
364 These statistics will be displayed in a separated table, for all the variables listed
365 in any /VARIABLES subcommand.
366 The XPROD keyword requests cross-product deviations and covariance estimators to
367 be displayed for each pair of variables.
368 The keyword ALL is the union of DESCRIPTIVES and XPROD.
376 /TABLES=var_list BY var_list [BY var_list]@dots{}
377 /MISSING=@{TABLE,INCLUDE,REPORT@}
378 /WRITE=@{NONE,CELLS,ALL@}
379 /FORMAT=@{TABLES,NOTABLES@}
380 @{LABELS,NOLABELS,NOVALLABS@}
385 /CELLS=@{COUNT,ROW,COLUMN,TOTAL,EXPECTED,RESIDUAL,SRESIDUAL,
386 ASRESIDUAL,ALL,NONE@}
387 /STATISTICS=@{CHISQ,PHI,CC,LAMBDA,UC,BTAU,CTAU,RISK,GAMMA,D,
388 KAPPA,ETA,CORR,ALL,NONE@}
391 /VARIABLES=var_list (low,high)@dots{}
394 The @cmd{CROSSTABS} procedure displays crosstabulation
395 tables requested by the user. It can calculate several statistics for
396 each cell in the crosstabulation tables. In addition, a number of
397 statistics can be calculated for each table itself.
399 The TABLES subcommand is used to specify the tables to be reported. Any
400 number of dimensions is permitted, and any number of variables per
401 dimension is allowed. The TABLES subcommand may be repeated as many
402 times as needed. This is the only required subcommand in @dfn{general
405 Occasionally, one may want to invoke a special mode called @dfn{integer
406 mode}. Normally, in general mode, PSPP automatically determines
407 what values occur in the data. In integer mode, the user specifies the
408 range of values that the data assumes. To invoke this mode, specify the
409 VARIABLES subcommand, giving a range of data values in parentheses for
410 each variable to be used on the TABLES subcommand. Data values inside
411 the range are truncated to the nearest integer, then assigned to that
412 value. If values occur outside this range, they are discarded. When it
413 is present, the VARIABLES subcommand must precede the TABLES
416 In general mode, numeric and string variables may be specified on
417 TABLES. In integer mode, only numeric variables are allowed.
419 The MISSING subcommand determines the handling of user-missing values.
420 When set to TABLE, the default, missing values are dropped on a table by
421 table basis. When set to INCLUDE, user-missing values are included in
422 tables and statistics. When set to REPORT, which is allowed only in
423 integer mode, user-missing values are included in tables but marked with
424 an @samp{M} (for ``missing'') and excluded from statistical
427 Currently the WRITE subcommand is ignored.
429 The FORMAT subcommand controls the characteristics of the
430 crosstabulation tables to be displayed. It has a number of possible
435 TABLES, the default, causes crosstabulation tables to be output.
436 NOTABLES suppresses them.
439 LABELS, the default, allows variable labels and value labels to appear
440 in the output. NOLABELS suppresses them. NOVALLABS displays variable
441 labels but suppresses value labels.
444 PIVOT, the default, causes each TABLES subcommand to be displayed in a
445 pivot table format. NOPIVOT causes the old-style crosstabulation format
449 AVALUE, the default, causes values to be sorted in ascending order.
450 DVALUE asserts a descending sort order.
453 INDEX/NOINDEX is currently ignored.
456 BOX/NOBOX is currently ignored.
459 The CELLS subcommand controls the contents of each cell in the displayed
460 crosstabulation table. The possible settings are:
476 Standardized residual.
478 Adjusted standardized residual.
482 Suppress cells entirely.
485 @samp{/CELLS} without any settings specified requests COUNT, ROW,
486 COLUMN, and TOTAL. If CELLS is not specified at all then only COUNT
489 The STATISTICS subcommand selects statistics for computation:
496 Pearson chi-square, likelihood ratio, Fisher's exact test, continuity
497 correction, linear-by-linear association.
501 Contingency coefficient.
505 Uncertainty coefficient.
521 Spearman correlation, Pearson's r.
528 Selected statistics are only calculated when appropriate for the
529 statistic. Certain statistics require tables of a particular size, and
530 some statistics are calculated only in integer mode.
532 @samp{/STATISTICS} without any settings selects CHISQ. If the
533 STATISTICS subcommand is not given, no statistics are calculated.
535 @strong{Please note:} Currently the implementation of CROSSTABS has the
540 Pearson's R (but not Spearman) is off a little.
542 T values for Spearman's R and Pearson's R are wrong.
544 Significance of symmetric and directional measures is not calculated.
546 Asymmetric ASEs and T values for lambda are wrong.
548 ASE of Goodman and Kruskal's tau is not calculated.
550 ASE of symmetric somers' d is wrong.
552 Approximate T of uncertainty coefficient is wrong.
555 Fixes for any of these deficiencies would be welcomed.
561 @cindex factor analysis
562 @cindex principal components analysis
563 @cindex principal axis factoring
564 @cindex data reduction
567 FACTOR VARIABLES=var_list
569 [ /METHOD = @{CORRELATION, COVARIANCE@} ]
571 [ /EXTRACTION=@{PC, PAF@}]
573 [ /PRINT=[INITIAL] [EXTRACTION] [UNIVARIATE] [CORRELATION] [COVARIANCE] [DET] [SIG] [ALL] [DEFAULT] ]
577 [ /FORMAT=[SORT] [BLANK(@var{n})] [DEFAULT] ]
579 [ /CRITERIA=[FACTORS(@var{n})] [MINEIGEN(@var{l})] [ITERATE(@var{m})] [ECONVERGE (@var{delta})] [DEFAULT] ]
581 [ /MISSING=[@{LISTWISE, PAIRWISE@}] [@{INCLUDE, EXCLUDE@}] ]
584 The FACTOR command performs Factor Analysis or Principal Axis Factoring on a dataset. It may be used to find
585 common factors in the data or for data reduction purposes.
587 The VARIABLES subcommand is required. It lists the variables which are to partake in the analysis.
589 The /EXTRACTION subcommand is used to specify the way in which factors (components) are extracted from the data.
590 If PC is specified, then Principal Components Analysis is used. If PAF is specified, then Principal Axis Factoring is
591 used. By default Principal Components Analysis will be used.
593 The /METHOD subcommand should be used to determine whether the covariance matrix or the correlation matrix of the data is
594 to be analysed. By default, the correlation matrix is analysed.
596 The /PRINT subcommand may be used to select which features of the analysis are reported:
600 A table of mean values, standard deviations and total weights are printed.
602 Initial communalities and eigenvalues are printed.
604 Extracted communalities and eigenvalues are printed.
606 The correlation matrix is printed.
608 The covariance matrix is printed.
610 The determinant of the correlation or covariance matrix is printed.
612 The significance of the elements of correlation matrix is printed.
614 All of the above are printed.
616 Identical to INITIAL and EXTRACTION.
619 If /PLOT=EIGEN is given, then a ``Scree'' plot of the eigenvalues will be printed. This can be useful for visualising
620 which factors (components) should be retained.
622 The /FORMAT subcommand determined how data are to be displayed in loading matrices. If SORT is specified, then the variables
623 are sorted in descending order of significance. If BLANK(@var{n}) is specified, then coefficients whose absolute value is less
624 than @var{n} will not be printed. If the keyword DEFAULT is given, or if no /FORMAT subcommand is given, then no sorting is
625 performed, and all coefficients will be printed.
627 The /CRITERIA subcommand is used to specify how the number of extracted factors (components) are chosen. If FACTORS(@var{n}) is
628 specified, where @var{n} is an integer, then @var{n} factors will be extracted. Otherwise, the MINEIGEN setting will
629 be used. MINEIGEN(@var{l}) requests that all factors whose eigenvalues are greater than or equal to @var{l} are extracted.
630 The default value of @var{l} is 1. The ECONVERGE and ITERATE settings have effect only when iterative algorithms for factor
631 extraction (such as Principal Axis Factoring) are used. ECONVERGE(@var{delta}) specifies that iteration should cease when
632 the maximum absolute value of the communality estimate between one iteration and the previous is less than @var{delta}. The
633 default value of @var{delta} is 0.001.
634 The ITERATE(@var{m}) setting sets the maximum number of iterations to @var{m}. The default value of @var{m} is 25.
636 The @cmd{MISSING} subcommand determines the handling of missing variables.
637 If INCLUDE is set, then user-missing values are included in the
638 calculations, but system-missing values are not.
639 If EXCLUDE is set, which is the default, user-missing
640 values are excluded as well as system-missing values.
642 If LISTWISE is set, then the entire case is excluded from analysis
643 whenever any variable specified in the @cmd{VARIABLES} subcommand
644 contains a missing value.
645 If PAIRWISE is set, then a case is considered missing only if either of the
646 values for the particular coefficient are missing.
647 The default is LISTWISE.
654 @cindex nonparametric tests
659 nonparametric test subcommands
664 [ /STATISTICS=@{DESCRIPTIVES@} ]
666 [ /MISSING=@{ANALYSIS, LISTWISE@} @{INCLUDE, EXCLUDE@} ]
668 [ /METHOD=EXACT [ TIMER [(n)] ] ]
671 NPAR TESTS performs nonparametric tests.
672 Non parametric tests make very few assumptions about the distribution of the
674 One or more tests may be specified by using the corresponding subcommand.
675 If the /STATISTICS subcommand is also specified, then summary statistics are
676 produces for each variable that is the subject of any test.
678 Certain tests may take a long time to execute, if an exact figure is required.
679 Therefore, by default asymptotic approximations are used unless the
680 subcommand /METHOD=EXACT is specified.
681 Exact tests give more accurate results, but may take an unacceptably long
682 time to perform. If the TIMER keyword is used, it sets a maximum time,
683 after which the test will be abandoned, and a warning message printed.
684 The time, in minutes, should be specified in parentheses after the TIMER keyword.
685 If the TIMER keyword is given without this figure, then a default value of 5 minutes
690 * BINOMIAL:: Binomial Test
691 * CHISQUARE:: Chisquare Test
692 * WILCOXON:: Wilcoxon Signed Ranks Test
693 * SIGN:: The Sign Test
698 @subsection Binomial test
700 @cindex binomial test
703 [ /BINOMIAL[(p)]=var_list[(value1[, value2)] ] ]
706 The /BINOMIAL subcommand compares the observed distribution of a dichotomous
707 variable with that of a binomial distribution.
708 The variable @var{p} specifies the test proportion of the binomial
710 The default value of 0.5 is assumed if @var{p} is omitted.
712 If a single value appears after the variable list, then that value is
713 used as the threshold to partition the observed values. Values less
714 than or equal to the threshold value form the first category. Values
715 greater than the threshold form the second category.
717 If two values appear after the variable list, then they will be used
718 as the values which a variable must take to be in the respective
720 Cases for which a variable takes a value equal to neither of the specified
721 values, take no part in the test for that variable.
723 If no values appear, then the variable must assume dichotomous
725 If more than two distinct, non-missing values for a variable
726 under test are encountered then an error occurs.
728 If the test proportion is equal to 0.5, then a two tailed test is
729 reported. For any other test proportion, a one tailed test is
731 For one tailed tests, if the test proportion is less than
732 or equal to the observed proportion, then the significance of
733 observing the observed proportion or more is reported.
734 If the test proportion is more than the observed proportion, then the
735 significance of observing the observed proportion or less is reported.
736 That is to say, the test is always performed in the observed
739 PSPP uses a very precise approximation to the gamma function to
740 compute the binomial significance. Thus, exact results are reported
741 even for very large sample sizes.
746 @subsection Chisquare Test
748 @cindex chisquare test
752 [ /CHISQUARE=var_list[(lo,hi)] [/EXPECTED=@{EQUAL|f1, f2 @dots{} fn@}] ]
756 The /CHISQUARE subcommand produces a chi-square statistic for the differences
757 between the expected and observed frequencies of the categories of a variable.
758 Optionally, a range of values may appear after the variable list.
759 If a range is given, then non integer values are truncated, and values
760 outside the specified range are excluded from the analysis.
762 The /EXPECTED subcommand specifies the expected values of each
764 There must be exactly one non-zero expected value, for each observed
765 category, or the EQUAL keywork must be specified.
766 You may use the notation @var{n}*@var{f} to specify @var{n}
767 consecutive expected categories all taking a frequency of @var{f}.
768 The frequencies given are proportions, not absolute frequencies. The
769 sum of the frequencies need not be 1.
770 If no /EXPECTED subcommand is given, then then equal frequencies
774 @subsection Wilcoxon Matched Pairs Signed Ranks Test
775 @comment node-name, next, previous, up
777 @cindex wilcoxon matched pairs signed ranks test
780 [ /WILCOXON varlist [ WITH varlist [ (PAIRED) ]]]
783 The /WILCOXON subcommand tests for differences between medians of the
785 The test does not make any assumptions about the variances of the samples.
786 It does however assume that the distribution is symetrical.
788 If the @code{WITH} keyword is omitted, then tests for all
789 combinations of the listed variables are performed.
790 If the @code{WITH} keyword is given, and the @code{(PAIRED)} keyword
791 is also given, then the number of variables preceding @code{WITH}
792 must be the same as the number following it.
793 In this case, tests for each respective pair of variables are
795 If the @code{WITH} keyword is given, but the
796 @code{(PAIRED)} keyword is omitted, then tests for each combination
797 of variable preceding @code{WITH} against variable following
798 @code{WITH} are performed.
802 @subsection Sign Test
807 [ /SIGN varlist [ WITH varlist [ (PAIRED) ]]]
810 The /SIGN subcommand tests for differences between medians of the
812 The test does not make any assumptions about the
813 distribution of the data.
815 If the @code{WITH} keyword is omitted, then tests for all
816 combinations of the listed variables are performed.
817 If the @code{WITH} keyword is given, and the @code{(PAIRED)} keyword
818 is also given, then the number of variables preceding @code{WITH}
819 must be the same as the number following it.
820 In this case, tests for each respective pair of variables are
822 If the @code{WITH} keyword is given, but the
823 @code{(PAIRED)} keyword is omitted, then tests for each combination
824 of variable preceding @code{WITH} against variable following
825 @code{WITH} are performed.
828 @comment node-name, next, previous, up
835 /MISSING=@{ANALYSIS,LISTWISE@} @{EXCLUDE,INCLUDE@}
836 /CRITERIA=CIN(confidence)
844 (Independent Samples mode.)
845 GROUPS=var(value1 [, value2])
849 (Paired Samples mode.)
850 PAIRS=var_list [WITH var_list [(PAIRED)] ]
855 The @cmd{T-TEST} procedure outputs tables used in testing hypotheses about
857 It operates in one of three modes:
859 @item One Sample mode.
860 @item Independent Groups mode.
865 Each of these modes are described in more detail below.
866 There are two optional subcommands which are common to all modes.
868 The @cmd{/CRITERIA} subcommand tells PSPP the confidence interval used
869 in the tests. The default value is 0.95.
872 The @cmd{MISSING} subcommand determines the handling of missing
874 If INCLUDE is set, then user-missing values are included in the
875 calculations, but system-missing values are not.
876 If EXCLUDE is set, which is the default, user-missing
877 values are excluded as well as system-missing values.
880 If LISTWISE is set, then the entire case is excluded from analysis
881 whenever any variable specified in the @cmd{/VARIABLES}, @cmd{/PAIRS} or
882 @cmd{/GROUPS} subcommands contains a missing value.
883 If ANALYSIS is set, then missing values are excluded only in the analysis for
884 which they would be needed. This is the default.
888 * One Sample Mode:: Testing against a hypothesised mean
889 * Independent Samples Mode:: Testing two independent groups for equal mean
890 * Paired Samples Mode:: Testing two interdependent groups for equal mean
893 @node One Sample Mode
894 @subsection One Sample Mode
896 The @cmd{TESTVAL} subcommand invokes the One Sample mode.
897 This mode is used to test a population mean against a hypothesised
899 The value given to the @cmd{TESTVAL} subcommand is the value against
900 which you wish to test.
901 In this mode, you must also use the @cmd{/VARIABLES} subcommand to
902 tell PSPP which variables you wish to test.
904 @node Independent Samples Mode
905 @comment node-name, next, previous, up
906 @subsection Independent Samples Mode
908 The @cmd{GROUPS} subcommand invokes Independent Samples mode or
910 This mode is used to test whether two groups of values have the
911 same population mean.
912 In this mode, you must also use the @cmd{/VARIABLES} subcommand to
913 tell PSPP the dependent variables you wish to test.
915 The variable given in the @cmd{GROUPS} subcommand is the independent
916 variable which determines to which group the samples belong.
917 The values in parentheses are the specific values of the independent
918 variable for each group.
919 If the parentheses are omitted and no values are given, the default values
920 of 1.0 and 2.0 are assumed.
922 If the independent variable is numeric,
923 it is acceptable to specify only one value inside the parentheses.
924 If you do this, cases where the independent variable is
925 greater than or equal to this value belong to the first group, and cases
926 less than this value belong to the second group.
927 When using this form of the @cmd{GROUPS} subcommand, missing values in
928 the independent variable are excluded on a listwise basis, regardless
929 of whether @cmd{/MISSING=LISTWISE} was specified.
932 @node Paired Samples Mode
933 @comment node-name, next, previous, up
934 @subsection Paired Samples Mode
936 The @cmd{PAIRS} subcommand introduces Paired Samples mode.
937 Use this mode when repeated measures have been taken from the same
939 If the @code{WITH} keyword is omitted, then tables for all
940 combinations of variables given in the @cmd{PAIRS} subcommand are
942 If the @code{WITH} keyword is given, and the @code{(PAIRED)} keyword
943 is also given, then the number of variables preceding @code{WITH}
944 must be the same as the number following it.
945 In this case, tables for each respective pair of variables are
947 In the event that the @code{WITH} keyword is given, but the
948 @code{(PAIRED)} keyword is omitted, then tables for each combination
949 of variable preceding @code{WITH} against variable following
950 @code{WITH} are generated.
954 @comment node-name, next, previous, up
958 @cindex analysis of variance
963 [/VARIABLES = ] var_list BY var
964 /MISSING=@{ANALYSIS,LISTWISE@} @{EXCLUDE,INCLUDE@}
965 /CONTRAST= value1 [, value2] ... [,valueN]
966 /STATISTICS=@{DESCRIPTIVES,HOMOGENEITY@}
970 The @cmd{ONEWAY} procedure performs a one-way analysis of variance of
971 variables factored by a single independent variable.
972 It is used to compare the means of a population
973 divided into more than two groups.
975 The variables to be analysed should be given in the @code{VARIABLES}
977 The list of variables must be followed by the @code{BY} keyword and
978 the name of the independent (or factor) variable.
980 You can use the @code{STATISTICS} subcommand to tell PSPP to display
981 ancilliary information. The options accepted are:
984 Displays descriptive statistics about the groups factored by the independent
987 Displays the Levene test of Homogeneity of Variance for the
988 variables and their groups.
991 The @code{CONTRAST} subcommand is used when you anticipate certain
992 differences between the groups.
993 The subcommand must be followed by a list of numerals which are the
994 coefficients of the groups to be tested.
995 The number of coefficients must correspond to the number of distinct
996 groups (or values of the independent variable).
997 If the total sum of the coefficients are not zero, then PSPP will
998 display a warning, but will proceed with the analysis.
999 The @code{CONTRAST} subcommand may be given up to 10 times in order
1000 to specify different contrast tests.
1003 @comment node-name, next, previous, up
1009 [VARIABLES=] var_list [@{A,D@}] [BY var_list]
1010 /TIES=@{MEAN,LOW,HIGH,CONDENSE@}
1011 /FRACTION=@{BLOM,TUKEY,VW,RANKIT@}
1013 /MISSING=@{EXCLUDE,INCLUDE@}
1015 /RANK [INTO var_list]
1016 /NTILES(k) [INTO var_list]
1017 /NORMAL [INTO var_list]
1018 /PERCENT [INTO var_list]
1019 /RFRACTION [INTO var_list]
1020 /PROPORTION [INTO var_list]
1022 /SAVAGE [INTO var_list]
1025 The @cmd{RANK} command ranks variables and stores the results into new
1028 The VARIABLES subcommand, which is mandatory, specifies one or
1029 more variables whose values are to be ranked.
1030 After each variable, @samp{A} or @samp{D} may appear, indicating that
1031 the variable is to be ranked in ascending or descending order.
1032 Ascending is the default.
1033 If a BY keyword appears, it should be followed by a list of variables
1034 which are to serve as group variables.
1035 In this case, the cases are gathered into groups, and ranks calculated
1038 The TIES subcommand specifies how tied values are to be treated. The
1039 default is to take the mean value of all the tied cases.
1041 The FRACTION subcommand specifies how proportional ranks are to be
1042 calculated. This only has any effect if NORMAL or PROPORTIONAL rank
1043 functions are requested.
1045 The PRINT subcommand may be used to specify that a summary of the rank
1046 variables created should appear in the output.
1048 The function subcommands are RANK, NTILES, NORMAL, PERCENT, RFRACTION,
1049 PROPORTION and SAVAGE. Any number of function subcommands may appear.
1050 If none are given, then the default is RANK.
1051 The NTILES subcommand must take an integer specifying the number of
1052 partitions into which values should be ranked.
1053 Each subcommand may be followed by the INTO keyword and a list of
1054 variables which are the variables to be created and receive the rank
1055 scores. There may be as many variables specified as there are
1056 variables named on the VARIABLES subcommand. If fewer are specified,
1057 then the variable names are automatically created.
1059 The MISSING subcommand determines how user missing values are to be
1060 treated. A setting of EXCLUDE means that variables whose values are
1061 user-missing are to be excluded from the rank scores. A setting of
1062 INCLUDE means they are to be included. The default is EXCLUDE.
1064 @include regression.texi
1068 @section RELIABILITY
1074 /SCALE (@var{name}) = @{var_list, ALL@}
1075 /MODEL=@{ALPHA, SPLIT[(N)]@}
1076 /SUMMARY=@{TOTAL,ALL@}
1077 /MISSING=@{EXCLUDE,INCLUDE@}
1080 @cindex Cronbach's Alpha
1081 The @cmd{RELIABILTY} command performs reliablity analysis on the data.
1083 The VARIABLES subcommand is required. It determines the set of variables
1084 upon which analysis is to be performed.
1086 The SCALE subcommand determines which variables reliability is to be
1087 calculated for. If it is omitted, then analysis for all variables named
1088 in the VARIABLES subcommand will be used.
1089 Optionally, the @var{name} parameter may be specified to set a string name
1092 The MODEL subcommand determines the type of analysis. If ALPHA is specified,
1093 then Cronbach's Alpha is calculated for the scale. If the model is SPLIT,
1094 then the variables are divided into 2 subsets. An optional parameter
1095 @var{N} may be given, to specify how many variables to be in the first subset.
1096 If @var{N} is omitted, then it defaults to one half of the variables in the
1097 scale, or one half minus one if there are an odd number of variables.
1098 The default model is ALPHA.
1100 By default, any cases with user missing, or system missing values for
1102 in the VARIABLES subcommand will be omitted from analysis.
1103 The MISSING subcommand determines whether user missing values are to
1104 be included or excluded in the analysis.
1106 The SUMMARY subcommand determines the type of summary analysis to be performed.
1107 Currently there is only one type: SUMMARY=TOTAL, which displays per-item
1108 analysis tested against the totals.
1116 @cindex Receiver Operating Characterstic
1117 @cindex Area under curve
1120 ROC @var{var_list} BY @var{state_var} (@var{state_value})
1121 /PLOT = @{ CURVE [(REFERENCE)], NONE @}
1122 /PRINT = [ SE ] [ COORDINATES ]
1123 /CRITERIA = [ CUTOFF(@{INCLUDE,EXCLUDE@}) ]
1124 [ TESTPOS (@{LARGE,SMALL@}) ]
1125 [ CI (@var{confidence}) ]
1126 [ DISTRIBUTION (@{FREE, NEGEXPO @}) ]
1127 /MISSING=@{EXCLUDE,INCLUDE@}
1131 The @cmd{ROC} command is used to plot the receiver operating characteristic curve
1132 of a dataset, and to estimate the area under the curve.
1133 This is useful for analysing the efficacy of a variable as a predictor of a state of nature.
1135 The mandatory @var{var_list} is the list of predictor variables.
1136 The variable @var{state_var} is the variable whose values represent the actual states,
1137 and @var{state_value} is the value of this variable which represents the positive state.
1139 The optional subcommand PLOT is used to determine if and how the ROC curve is drawn.
1140 The keyword CURVE means that the ROC curve should be drawn, and the optional keyword REFERENCE,
1141 which should be enclosed in parentheses, says that the diagonal reference line should be drawn.
1142 If the keyword NONE is given, then no ROC curve is drawn.
1143 By default, the curve is drawn with no reference line.
1145 The optional subcommand PRINT determines which additional tables should be printed.
1146 Two additional tables are available.
1147 The SE keyword says that standard error of the area under the curve should be printed as well as
1149 In addition, a p-value under the null hypothesis that the area under the curve equals 0.5 will be
1151 The COORDINATES keyword says that a table of coordinates of the ROC curve should be printed.
1153 The CRITERIA subcommand has four optional parameters:
1155 @item The TESTPOS parameter may be LARGE or SMALL.
1156 LARGE is the default, and says that larger values in the predictor variables are to be
1157 considered positive. SMALL indicates that smaller values should be considered positive.
1159 @item The CI parameter specifies the confidence interval that should be printed.
1160 It has no effect if the SE keyword in the PRINT subcommand has not been given.
1162 @item The DISTRIBUTION parameter determines the method to be used when estimating the area
1164 There are two possibilities, @i{viz}: FREE and NEGEXPO.
1165 The FREE method uses a non-parametric estimate, and the NEGEXPO method a bi-negative
1166 exponential distribution estimate.
1167 The NEGEXPO method should only be used when the number of positive actual states is
1168 equal to the number of negative actual states.
1169 The default is FREE.
1171 @item The CUTOFF parameter is for compatibility and is ignored.
1174 The MISSING subcommand determines whether user missing values are to
1175 be included or excluded in the analysis. The default behaviour is to
1177 Cases are excluded on a listwise basis; if any of the variables in @var{var_list}
1178 or if the variable @var{state_var} is missing, then the entire case will be