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 * CROSSTABS:: Crosstabulation tables.
12 * NPAR TESTS:: Nonparametric tests.
13 * T-TEST:: Test hypotheses about means.
14 * ONEWAY:: One way analysis of variance.
15 * RANK:: Compute rank scores.
16 * REGRESSION:: Linear regression.
17 * RELIABILITY:: Reliability analysis.
27 /MISSING=@{VARIABLE,LISTWISE@} @{INCLUDE,NOINCLUDE@}
28 /FORMAT=@{LABELS,NOLABELS@} @{NOINDEX,INDEX@} @{LINE,SERIAL@}
30 /STATISTICS=@{ALL,MEAN,SEMEAN,STDDEV,VARIANCE,KURTOSIS,
31 SKEWNESS,RANGE,MINIMUM,MAXIMUM,SUM,DEFAULT,
32 SESKEWNESS,SEKURTOSIS@}
33 /SORT=@{NONE,MEAN,SEMEAN,STDDEV,VARIANCE,KURTOSIS,SKEWNESS,
34 RANGE,MINIMUM,MAXIMUM,SUM,SESKEWNESS,SEKURTOSIS,NAME@}
38 The @cmd{DESCRIPTIVES} procedure reads the active file and outputs
40 statistics requested by the user. In addition, it can optionally
43 The VARIABLES subcommand, which is required, specifies the list of
44 variables to be analyzed. Keyword VARIABLES is optional.
46 All other subcommands are optional:
48 The MISSING subcommand determines the handling of missing variables. If
49 INCLUDE is set, then user-missing values are included in the
50 calculations. If NOINCLUDE is set, which is the default, user-missing
51 values are excluded. If VARIABLE is set, then missing values are
52 excluded on a variable by variable basis; if LISTWISE is set, then
53 the entire case is excluded whenever any value in that case has a
54 system-missing or, if INCLUDE is set, user-missing value.
56 The FORMAT subcommand affects the output format. Currently the
57 LABELS/NOLABELS and NOINDEX/INDEX settings are not used. When SERIAL is
58 set, both valid and missing number of cases are listed in the output;
59 when NOSERIAL is set, only valid cases are listed.
61 The SAVE subcommand causes @cmd{DESCRIPTIVES} to calculate Z scores for all
62 the specified variables. The Z scores are saved to new variables.
63 Variable names are generated by trying first the original variable name
64 with Z prepended and truncated to a maximum of 8 characters, then the
65 names ZSC000 through ZSC999, STDZ00 through STDZ09, ZZZZ00 through
66 ZZZZ09, ZQZQ00 through ZQZQ09, in that sequence. In addition, Z score
67 variable names can be specified explicitly on VARIABLES in the variable
68 list by enclosing them in parentheses after each variable.
70 The STATISTICS subcommand specifies the statistics to be displayed:
74 All of the statistics below.
78 Standard error of the mean.
84 Kurtosis and standard error of the kurtosis.
86 Skewness and standard error of the skewness.
96 Mean, standard deviation of the mean, minimum, maximum.
98 Standard error of the kurtosis.
100 Standard error of the skewness.
103 The SORT subcommand specifies how the statistics should be sorted. Most
104 of the possible values should be self-explanatory. NAME causes the
105 statistics to be sorted by name. By default, the statistics are listed
106 in the order that they are specified on the VARIABLES subcommand. The A
107 and D settings request an ascending or descending sort order,
117 /FORMAT=@{TABLE,NOTABLE,LIMIT(limit)@}
118 @{STANDARD,CONDENSE,ONEPAGE[(onepage_limit)]@}
120 @{AVALUE,DVALUE,AFREQ,DFREQ@}
123 /MISSING=@{EXCLUDE,INCLUDE@}
124 /STATISTICS=@{DEFAULT,MEAN,SEMEAN,MEDIAN,MODE,STDDEV,VARIANCE,
125 KURTOSIS,SKEWNESS,RANGE,MINIMUM,MAXIMUM,SUM,
126 SESKEWNESS,SEKURTOSIS,ALL,NONE@}
128 /PERCENTILES=percent@dots{}
129 /HISTOGRAM=[MINIMUM(x_min)] [MAXIMUM(x_max)]
130 [@{FREQ,PCNT@}] [@{NONORMAL,NORMAL@}]
131 /PIECHART=[MINIMUM(x_min)] [MAXIMUM(x_max)] @{NOMISSING,MISSING@}
133 (These options are not currently implemented.)
139 The @cmd{FREQUENCIES} procedure outputs frequency tables for specified
141 @cmd{FREQUENCIES} can also calculate and display descriptive statistics
142 (including median and mode) and percentiles.
144 @cmd{FREQUENCIES} also support graphical output in the form of
145 histograms and pie charts. In the future, it will be able to produce
146 bar charts and output percentiles for grouped data.
148 The VARIABLES subcommand is the only required subcommand. Specify the
149 variables to be analyzed.
151 The FORMAT subcommand controls the output format. It has several
156 TABLE, the default, causes a frequency table to be output for every
157 variable specified. NOTABLE prevents them from being output. LIMIT
158 with a numeric argument causes them to be output except when there are
159 more than the specified number of values in the table.
162 STANDARD frequency tables contain more complete information, but also to
163 take up more space on the printed page. CONDENSE frequency tables are
164 less informative but take up less space. ONEPAGE with a numeric
165 argument will output standard frequency tables if there are the
166 specified number of values or less, condensed tables otherwise. ONEPAGE
167 without an argument defaults to a threshold of 50 values.
170 LABELS causes value labels to be displayed in STANDARD frequency
171 tables. NOLABLES prevents this.
174 Normally frequency tables are sorted in ascending order by value. This
175 is AVALUE. DVALUE tables are sorted in descending order by value.
176 AFREQ and DFREQ tables are sorted in ascending and descending order,
177 respectively, by frequency count.
180 SINGLE spaced frequency tables are closely spaced. DOUBLE spaced
181 frequency tables have wider spacing.
184 OLDPAGE and NEWPAGE are not currently used.
187 The MISSING subcommand controls the handling of user-missing values.
188 When EXCLUDE, the default, is set, user-missing values are not included
189 in frequency tables or statistics. When INCLUDE is set, user-missing
190 are included. System-missing values are never included in statistics,
191 but are listed in frequency tables.
193 The available STATISTICS are the same as available in @cmd{DESCRIPTIVES}
194 (@pxref{DESCRIPTIVES}), with the addition of MEDIAN, the data's median
195 value, and MODE, the mode. (If there are multiple modes, the smallest
196 value is reported.) By default, the mean, standard deviation of the
197 mean, minimum, and maximum are reported for each variable.
200 PERCENTILES causes the specified percentiles to be reported.
201 The percentiles should be presented at a list of numbers between 0
203 The NTILES subcommand causes the percentiles to be reported at the
204 boundaries of the data set divided into the specified number of ranges.
205 For instance, @code{/NTILES=4} would cause quartiles to be reported.
207 The HISTOGRAM subcommand causes the output to include a histogram for
208 each specified numeric variable. The X axis by default ranges from the
209 minimum to the maximum value observed in the data, but the MINIMUM and
210 MAXIMUM keywords can set an explicit range. The Y axis by default is
211 labeled in frequencies; use the PERCENT keyword to causes it to be
212 labeled in percent of the total observed count. Specify NORMAL to
213 superimpose a normal curve on the histogram.
214 Histograms are not created for string variables.
216 The PIECHART adds a pie chart for each variable to the data. Each
217 slice represents one value, with the size of the slice proportional to
218 the value's frequency. By default, all non-missing values are given
219 slices. The MINIMUM and MAXIMUM keywords can be used to limit the
220 displayed slices to a given range of values. The MISSING keyword adds
221 slices for missing values.
224 @comment node-name, next, previous, up
228 @cindex Normality, testing for
232 VARIABLES=var_list [BY factor_list ]
233 /STATISTICS=@{DESCRIPTIVES, EXTREME[(n)], ALL, NONE@}
234 /PLOT=@{BOXPLOT, NPPLOT, HISTOGRAM, ALL, NONE@}
236 /COMPARE=@{GROUPS,VARIABLES@}
239 /PERCENTILE=[value_list]=@{HAVERAGE, WAVERAGE, ROUND, AEMPIRICAL, EMPIRICAL @}
240 /MISSING=@{LISTWISE, PAIRWISE@} [@{EXCLUDE, INCLUDE@}]
241 [@{NOREPORT,REPORT@}]
245 The @cmd{EXAMINE} command is used to test how closely a distribution is to a
246 normal distribution. It also shows you outliers and extreme values.
248 The VARIABLES subcommand specifies the dependent variables and the
249 independent variable to use as factors for the analysis. Variables
250 listed before the first BY keyword are the dependent variables.
251 The dependent variables may optionally be followed by a list of
252 factors which tell PSPP how to break down the analysis for each
253 dependent variable. The format for each factor is
259 The STATISTICS subcommand specifies the analysis to be done.
260 DESCRIPTIVES will produce a table showing some parametric and
261 non-parametrics statistics. EXTREME produces a table showing extreme
262 values of the dependent variable. A number in parentheses determines
263 how many upper and lower extremes to show. The default number is 5.
266 The PLOT subcommand specifies which plots are to be produced if any.
268 The COMPARE subcommand is only relevant if producing boxplots, and it is only
269 useful there is more than one dependent variable and at least one factor. If
270 /COMPARE=GROUPS is specified, then one plot per dependent variable is produced,
271 containing boxplots for all the factors.
272 If /COMPARE=VARIABLES is specified, then one plot per factor is produced, each
273 each containing one boxplot per dependent variable.
274 If the /COMPARE subcommand is ommitted, then PSPP uses the default value of
277 The ID subcommand also pertains to boxplots. If given, it must
278 specify a variable name. Outliers and extreme cases plotted in
279 boxplots will be labelled with the case from that variable. Numeric or
280 string variables are permissible. If the ID subcommand is not given,
281 then the casenumber will be used for labelling.
283 The CINTERVAL subcommand specifies the confidence interval to use in
284 calculation of the descriptives command. The default it 95%.
287 The PERCENTILES subcommand specifies which percentiles are to be calculated,
288 and which algorithm to use for calculating them. The default is to
289 calculate the 5, 10, 25, 50, 75, 90, 95 percentiles using the
292 The TOTAL and NOTOTAL subcommands are mutually exclusive. If NOTOTAL
293 is given and factors have been specified in the VARIABLES subcommand,
294 then then statistics for the unfactored dependent variables are
295 produced in addition to the factored variables. If there are no
296 factors specified then TOTAL and NOTOTAL have no effect.
299 If many dependent variable are given, or factors are given for which
300 there are many distinct values, then @cmd{EXAMINE} will produce a very
301 large quantity of output.
310 /TABLES=var_list BY var_list [BY var_list]@dots{}
311 /MISSING=@{TABLE,INCLUDE,REPORT@}
312 /WRITE=@{NONE,CELLS,ALL@}
313 /FORMAT=@{TABLES,NOTABLES@}
314 @{LABELS,NOLABELS,NOVALLABS@}
319 /CELLS=@{COUNT,ROW,COLUMN,TOTAL,EXPECTED,RESIDUAL,SRESIDUAL,
320 ASRESIDUAL,ALL,NONE@}
321 /STATISTICS=@{CHISQ,PHI,CC,LAMBDA,UC,BTAU,CTAU,RISK,GAMMA,D,
322 KAPPA,ETA,CORR,ALL,NONE@}
325 /VARIABLES=var_list (low,high)@dots{}
328 The @cmd{CROSSTABS} procedure displays crosstabulation
329 tables requested by the user. It can calculate several statistics for
330 each cell in the crosstabulation tables. In addition, a number of
331 statistics can be calculated for each table itself.
333 The TABLES subcommand is used to specify the tables to be reported. Any
334 number of dimensions is permitted, and any number of variables per
335 dimension is allowed. The TABLES subcommand may be repeated as many
336 times as needed. This is the only required subcommand in @dfn{general
339 Occasionally, one may want to invoke a special mode called @dfn{integer
340 mode}. Normally, in general mode, PSPP automatically determines
341 what values occur in the data. In integer mode, the user specifies the
342 range of values that the data assumes. To invoke this mode, specify the
343 VARIABLES subcommand, giving a range of data values in parentheses for
344 each variable to be used on the TABLES subcommand. Data values inside
345 the range are truncated to the nearest integer, then assigned to that
346 value. If values occur outside this range, they are discarded. When it
347 is present, the VARIABLES subcommand must precede the TABLES
350 In general mode, numeric and string variables may be specified on
351 TABLES. Although long string variables are allowed, only their
352 initial short-string parts are used. In integer mode, only numeric
353 variables are allowed.
355 The MISSING subcommand determines the handling of user-missing values.
356 When set to TABLE, the default, missing values are dropped on a table by
357 table basis. When set to INCLUDE, user-missing values are included in
358 tables and statistics. When set to REPORT, which is allowed only in
359 integer mode, user-missing values are included in tables but marked with
360 an @samp{M} (for ``missing'') and excluded from statistical
363 Currently the WRITE subcommand is ignored.
365 The FORMAT subcommand controls the characteristics of the
366 crosstabulation tables to be displayed. It has a number of possible
371 TABLES, the default, causes crosstabulation tables to be output.
372 NOTABLES suppresses them.
375 LABELS, the default, allows variable labels and value labels to appear
376 in the output. NOLABELS suppresses them. NOVALLABS displays variable
377 labels but suppresses value labels.
380 PIVOT, the default, causes each TABLES subcommand to be displayed in a
381 pivot table format. NOPIVOT causes the old-style crosstabulation format
385 AVALUE, the default, causes values to be sorted in ascending order.
386 DVALUE asserts a descending sort order.
389 INDEX/NOINDEX is currently ignored.
392 BOX/NOBOX is currently ignored.
395 The CELLS subcommand controls the contents of each cell in the displayed
396 crosstabulation table. The possible settings are:
412 Standardized residual.
414 Adjusted standardized residual.
418 Suppress cells entirely.
421 @samp{/CELLS} without any settings specified requests COUNT, ROW,
422 COLUMN, and TOTAL. If CELLS is not specified at all then only COUNT
425 The STATISTICS subcommand selects statistics for computation:
432 Pearson chi-square, likelihood ratio, Fisher's exact test, continuity
433 correction, linear-by-linear association.
437 Contingency coefficient.
441 Uncertainty coefficient.
457 Spearman correlation, Pearson's r.
464 Selected statistics are only calculated when appropriate for the
465 statistic. Certain statistics require tables of a particular size, and
466 some statistics are calculated only in integer mode.
468 @samp{/STATISTICS} without any settings selects CHISQ. If the
469 STATISTICS subcommand is not given, no statistics are calculated.
471 @strong{Please note:} Currently the implementation of CROSSTABS has the
476 Pearson's R (but not Spearman) is off a little.
478 T values for Spearman's R and Pearson's R are wrong.
480 Significance of symmetric and directional measures is not calculated.
482 Asymmetric ASEs and T values for lambda are wrong.
484 ASE of Goodman and Kruskal's tau is not calculated.
486 ASE of symmetric somers' d is wrong.
488 Approximate T of uncertainty coefficient is wrong.
491 Fixes for any of these deficiencies would be welcomed.
497 @cindex nonparametric tests
502 nonparametric test subcommands
507 [ /STATISTICS=@{DESCRIPTIVES@} ]
509 [ /MISSING=@{ANALYSIS, LISTWISE@} @{INCLUDE, EXCLUDE@} ]
511 [ /METHOD=EXACT [ TIMER [(n)] ] ]
514 NPAR TESTS performs nonparametric tests.
515 Non parametric tests make very few assumptions about the distribution of the
517 One or more tests may be specified by using the corresponding subcommand.
518 If the /STATISTICS subcommand is also specified, then summary statistics are
519 produces for each variable that is the subject of any test.
521 Certain tests may take a long time to execute, if an exact figure is required.
522 Therefore, by default asymptotic approximations are used unless the
523 subcommand /METHOD=EXACT is specified.
524 Exact tests give more accurate results, but may take an unacceptably long
525 time to perform. If the TIMER keyword is used, it sets a maximum time,
526 after which the test will be abandoned, and a warning message printed.
527 The time, in minutes, should be specified in parentheses after the TIMER keyword.
528 If the TIMER keyword is given without this figure, then a default value of 5 minutes
533 * BINOMIAL:: Binomial Test
534 * CHISQUARE:: Chisquare Test
535 * WILCOXON:: Wilcoxon Signed Ranks Test
536 * SIGN:: The Sign Test
541 @subsection Binomial test
543 @cindex binomial test
546 [ /BINOMIAL[(p)]=var_list[(value1[, value2)] ] ]
549 The /BINOMIAL subcommand compares the observed distribution of a dichotomous
550 variable with that of a binomial distribution.
551 The variable @var{p} specifies the test proportion of the binomial
553 The default value of 0.5 is assumed if @var{p} is omitted.
555 If a single value appears after the variable list, then that value is
556 used as the threshold to partition the observed values. Values less
557 than or equal to the threshold value form the first category. Values
558 greater than the threshold form the second category.
560 If two values appear after the variable list, then they will be used
561 as the values which a variable must take to be in the respective
563 Cases for which a variable takes a value equal to neither of the specified
564 values, take no part in the test for that variable.
566 If no values appear, then the variable must assume dichotomous
568 If more than two distinct, non-missing values for a variable
569 under test are encountered then an error occurs.
571 If the test proportion is equal to 0.5, then a two tailed test is
572 reported. For any other test proportion, a one tailed test is
574 For one tailed tests, if the test proportion is less than
575 or equal to the observed proportion, then the significance of
576 observing the observed proportion or more is reported.
577 If the test proportion is more than the observed proportion, then the
578 significance of observing the observed proportion or less is reported.
579 That is to say, the test is always performed in the observed
582 PSPP uses a very precise approximation to the gamma function to
583 compute the binomial significance. Thus, exact results are reported
584 even for very large sample sizes.
589 @subsection Chisquare Test
591 @cindex chisquare test
595 [ /CHISQUARE=var_list[(lo,hi)] [/EXPECTED=@{EQUAL|f1, f2 @dots{} fn@}] ]
599 The /CHISQUARE subcommand produces a chi-square statistic for the differences
600 between the expected and observed frequencies of the categories of a variable.
601 Optionally, a range of values may appear after the variable list.
602 If a range is given, then non integer values are truncated, and values
603 outside the specified range are excluded from the analysis.
605 The /EXPECTED subcommand specifies the expected values of each
607 There must be exactly one non-zero expected value, for each observed
608 category, or the EQUAL keywork must be specified.
609 You may use the notation @var{n}*@var{f} to specify @var{n}
610 consecutive expected categories all taking a frequency of @var{f}.
611 The frequencies given are proportions, not absolute frequencies. The
612 sum of the frequencies need not be 1.
613 If no /EXPECTED subcommand is given, then then equal frequencies
617 @subsection Wilcoxon Matched Pairs Signed Ranks Test
618 @comment node-name, next, previous, up
620 @cindex wilcoxon matched pairs signed ranks test
623 [ /WILCOXON varlist [ WITH varlist [ (PAIRED) ]]]
626 The /WILCOXON subcommand tests for differences between medians of the
628 The test does not make any assumptions about the variances of the samples.
629 It does however assume that the distribution is symetrical.
631 If the @code{WITH} keyword is omitted, then tests for all
632 combinations of the listed variables are performed.
633 If the @code{WITH} keyword is given, and the @code{(PAIRED)} keyword
634 is also given, then the number of variables preceding @code{WITH}
635 must be the same as the number following it.
636 In this case, tests for each respective pair of variables are
638 If the @code{WITH} keyword is given, but the
639 @code{(PAIRED)} keyword is omitted, then tests for each combination
640 of variable preceding @code{WITH} against variable following
641 @code{WITH} are performed.
645 @subsection Sign Test
650 [ /SIGN varlist [ WITH varlist [ (PAIRED) ]]]
653 The /SIGN subcommand tests for differences between medians of the
655 The test does not make any assumptions about the
656 distribution of the data.
658 If the @code{WITH} keyword is omitted, then tests for all
659 combinations of the listed variables are performed.
660 If the @code{WITH} keyword is given, and the @code{(PAIRED)} keyword
661 is also given, then the number of variables preceding @code{WITH}
662 must be the same as the number following it.
663 In this case, tests for each respective pair of variables are
665 If the @code{WITH} keyword is given, but the
666 @code{(PAIRED)} keyword is omitted, then tests for each combination
667 of variable preceding @code{WITH} against variable following
668 @code{WITH} are performed.
671 @comment node-name, next, previous, up
678 /MISSING=@{ANALYSIS,LISTWISE@} @{EXCLUDE,INCLUDE@}
679 /CRITERIA=CIN(confidence)
687 (Independent Samples mode.)
688 GROUPS=var(value1 [, value2])
692 (Paired Samples mode.)
693 PAIRS=var_list [WITH var_list [(PAIRED)] ]
698 The @cmd{T-TEST} procedure outputs tables used in testing hypotheses about
700 It operates in one of three modes:
702 @item One Sample mode.
703 @item Independent Groups mode.
708 Each of these modes are described in more detail below.
709 There are two optional subcommands which are common to all modes.
711 The @cmd{/CRITERIA} subcommand tells PSPP the confidence interval used
712 in the tests. The default value is 0.95.
715 The @cmd{MISSING} subcommand determines the handling of missing
717 If INCLUDE is set, then user-missing values are included in the
718 calculations, but system-missing values are not.
719 If EXCLUDE is set, which is the default, user-missing
720 values are excluded as well as system-missing values.
723 If LISTWISE is set, then the entire case is excluded from analysis
724 whenever any variable specified in the @cmd{/VARIABLES}, @cmd{/PAIRS} or
725 @cmd{/GROUPS} subcommands contains a missing value.
726 If ANALYSIS is set, then missing values are excluded only in the analysis for
727 which they would be needed. This is the default.
731 * One Sample Mode:: Testing against a hypothesised mean
732 * Independent Samples Mode:: Testing two independent groups for equal mean
733 * Paired Samples Mode:: Testing two interdependent groups for equal mean
736 @node One Sample Mode
737 @subsection One Sample Mode
739 The @cmd{TESTVAL} subcommand invokes the One Sample mode.
740 This mode is used to test a population mean against a hypothesised
742 The value given to the @cmd{TESTVAL} subcommand is the value against
743 which you wish to test.
744 In this mode, you must also use the @cmd{/VARIABLES} subcommand to
745 tell PSPP which variables you wish to test.
747 @node Independent Samples Mode
748 @comment node-name, next, previous, up
749 @subsection Independent Samples Mode
751 The @cmd{GROUPS} subcommand invokes Independent Samples mode or
753 This mode is used to test whether two groups of values have the
754 same population mean.
755 In this mode, you must also use the @cmd{/VARIABLES} subcommand to
756 tell PSPP the dependent variables you wish to test.
758 The variable given in the @cmd{GROUPS} subcommand is the independent
759 variable which determines to which group the samples belong.
760 The values in parentheses are the specific values of the independent
761 variable for each group.
762 If the parentheses are omitted and no values are given, the default values
763 of 1.0 and 2.0 are assumed.
765 If the independent variable is numeric,
766 it is acceptable to specify only one value inside the parentheses.
767 If you do this, cases where the independent variable is
768 greater than or equal to this value belong to the first group, and cases
769 less than this value belong to the second group.
770 When using this form of the @cmd{GROUPS} subcommand, missing values in
771 the independent variable are excluded on a listwise basis, regardless
772 of whether @cmd{/MISSING=LISTWISE} was specified.
775 @node Paired Samples Mode
776 @comment node-name, next, previous, up
777 @subsection Paired Samples Mode
779 The @cmd{PAIRS} subcommand introduces Paired Samples mode.
780 Use this mode when repeated measures have been taken from the same
782 If the @code{WITH} keyword is omitted, then tables for all
783 combinations of variables given in the @cmd{PAIRS} subcommand are
785 If the @code{WITH} keyword is given, and the @code{(PAIRED)} keyword
786 is also given, then the number of variables preceding @code{WITH}
787 must be the same as the number following it.
788 In this case, tables for each respective pair of variables are
790 In the event that the @code{WITH} keyword is given, but the
791 @code{(PAIRED)} keyword is omitted, then tables for each combination
792 of variable preceding @code{WITH} against variable following
793 @code{WITH} are generated.
797 @comment node-name, next, previous, up
801 @cindex analysis of variance
806 [/VARIABLES = ] var_list BY var
807 /MISSING=@{ANALYSIS,LISTWISE@} @{EXCLUDE,INCLUDE@}
808 /CONTRAST= value1 [, value2] ... [,valueN]
809 /STATISTICS=@{DESCRIPTIVES,HOMOGENEITY@}
813 The @cmd{ONEWAY} procedure performs a one-way analysis of variance of
814 variables factored by a single independent variable.
815 It is used to compare the means of a population
816 divided into more than two groups.
818 The variables to be analysed should be given in the @code{VARIABLES}
820 The list of variables must be followed by the @code{BY} keyword and
821 the name of the independent (or factor) variable.
823 You can use the @code{STATISTICS} subcommand to tell PSPP to display
824 ancilliary information. The options accepted are:
827 Displays descriptive statistics about the groups factored by the independent
830 Displays the Levene test of Homogeneity of Variance for the
831 variables and their groups.
834 The @code{CONTRAST} subcommand is used when you anticipate certain
835 differences between the groups.
836 The subcommand must be followed by a list of numerals which are the
837 coefficients of the groups to be tested.
838 The number of coefficients must correspond to the number of distinct
839 groups (or values of the independent variable).
840 If the total sum of the coefficients are not zero, then PSPP will
841 display a warning, but will proceed with the analysis.
842 The @code{CONTRAST} subcommand may be given up to 10 times in order
843 to specify different contrast tests.
846 @comment node-name, next, previous, up
852 [VARIABLES=] var_list [@{A,D@}] [BY var_list]
853 /TIES=@{MEAN,LOW,HIGH,CONDENSE@}
854 /FRACTION=@{BLOM,TUKEY,VW,RANKIT@}
856 /MISSING=@{EXCLUDE,INCLUDE@}
858 /RANK [INTO var_list]
859 /NTILES(k) [INTO var_list]
860 /NORMAL [INTO var_list]
861 /PERCENT [INTO var_list]
862 /RFRACTION [INTO var_list]
863 /PROPORTION [INTO var_list]
865 /SAVAGE [INTO var_list]
868 The @cmd{RANK} command ranks variables and stores the results into new
871 The VARIABLES subcommand, which is mandatory, specifies one or
872 more variables whose values are to be ranked.
873 After each variable, @samp{A} or @samp{D} may appear, indicating that
874 the variable is to be ranked in ascending or descending order.
875 Ascending is the default.
876 If a BY keyword appears, it should be followed by a list of variables
877 which are to serve as group variables.
878 In this case, the cases are gathered into groups, and ranks calculated
881 The TIES subcommand specifies how tied values are to be treated. The
882 default is to take the mean value of all the tied cases.
884 The FRACTION subcommand specifies how proportional ranks are to be
885 calculated. This only has any effect if NORMAL or PROPORTIONAL rank
886 functions are requested.
888 The PRINT subcommand may be used to specify that a summary of the rank
889 variables created should appear in the output.
891 The function subcommands are RANK, NTILES, NORMAL, PERCENT, RFRACTION,
892 PROPORTION and SAVAGE. Any number of function subcommands may appear.
893 If none are given, then the default is RANK.
894 The NTILES subcommand must take an integer specifying the number of
895 partitions into which values should be ranked.
896 Each subcommand may be followed by the INTO keyword and a list of
897 variables which are the variables to be created and receive the rank
898 scores. There may be as many variables specified as there are
899 variables named on the VARIABLES subcommand. If fewer are specified,
900 then the variable names are automatically created.
902 The MISSING subcommand determines how user missing values are to be
903 treated. A setting of EXCLUDE means that variables whose values are
904 user-missing are to be excluded from the rank scores. A setting of
905 INCLUDE means they are to be included. The default is EXCLUDE.
907 @include regression.texi
917 /SCALE (@var{name}) = @{var_list, ALL@}
918 /MODEL=@{ALPHA, SPLIT[(N)]@}
919 /SUMMARY=@{TOTAL,ALL@}
920 /MISSING=@{EXCLUDE,INCLUDE@}
923 @cindex Cronbach's Alpha
924 The @cmd{RELIABILTY} command performs reliablity analysis on the data.
926 The VARIABLES subcommand is required. It determines the set of variables
927 upon which analysis is to be performed.
929 The SCALE subcommand determines which variables reliability is to be
930 calculated for. If it is omitted, then analysis for all variables named
931 in the VARIABLES subcommand will be used.
932 Optionally, the @var{name} parameter may be specified to set a string name
935 The MODEL subcommand determines the type of analysis. If ALPHA is specified,
936 then Cronbach's Alpha is calculated for the scale. If the model is SPLIT,
937 then the variables are divided into 2 subsets. An optional parameter
938 @var{N} may be given, to specify how many variables to be in the first subset.
939 If @var{N} is omitted, then it defaults to one half of the variables in the
940 scale, or one half minus one if there are an odd number of variables.
941 The default model is ALPHA.
943 By default, any cases with user missing, or system missing values for
945 in the VARIABLES subcommand will be omitted from analysis.
946 The MISSING subcommand determines whether user missing values are to
947 be included or excluded in the analysis.
949 The SUMMARY subcommand determines the type of summary analysis to be performed.
950 Currently there is only one type: SUMMARY=TOTAL, which displays per-item
951 analysis tested against the totals.