|
Size: 310
Comment:
|
← Revision 5 as of 2023-06-11 20:53:05 ⇥
Size: 5131
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 2: | Line 2: |
The '''`CROSSTABS`''' command creates a cross-tabulation. |
|
| Line 11: | Line 13: |
| To create a two-way crosstabulation, try: | {{{ crosstabs /tables=foo by bar. }}} To create a series of tables with `foo` as the row-wise variable in each, try: |
| Line 14: | Line 20: |
| crosstabs /table=FOO by BAR /table=HAM by SPAM EGGS. | crosstabs /tables=foo by ham spam eggs. |
| Line 17: | Line 23: |
| This will create three tables: | To create an N-dimensional table, try: |
| Line 19: | Line 25: |
| 1. `FOO` by `BAR` 2. `HAM` by `SPAM` 3. `HAM` by `EGGS` |
{{{ crosstabs /tables=foo by bar by baz. }}} Multiple `/TABLES` subcommands are allowed. {{{ crosstabs /tables=foo by bar /tables=foo by baz. }}} === Format === The default '''`/FORMAT`''' subcommand is `/FORMAT=TABLES AVALUE`. This causes SPSS to create tables (the alternative being `NOTABLES`) and sort values in ascending order (the alternative being `DVALUE`). === Cells === The default '''`/CELLS`''' subcommand is `/CELLS=COUNT`. This causes table cells to be populated with counts of observations. Options are: ||'''Name''' ||'''Effect''' || ||`COUNT` ||count || ||`ROW` ||row percent || ||`COLUMN` ||column percent || ||`TOTAL` ||table percent || ||`EXPECTED` ||expected value || ||`RESIDUAL` ||residual || ||`SRESIDUAL` ||standardized residual || ||`ASRESIDUAL`||adjusted standardized residual|| ||`ALL` ||all of the above || ||`NONE` ||display nothing || If the `/CELLS` subcommand is specified without any options, the `COUNT`, `ROW`, `COLUMN`, and `TOTAL` options are selected automatically. === Count === By default, SPSS uses raw case weights without rounding for counting. Add the '''`/COUNT=ASIS`''' subcommand to explicitly allow this behavior. To use rounding, first pick between '''`CASE`''' for rounding each case's weight before tabulating, and '''`CELL`''' for rounding the final tabulation. Then pick betwen '''`ROUND`''' for rounding to the nearest integer, and '''`TRUNCATE`''' for rounding towards 0. Add a subcommand with the specifications like `/COUNT=CELL ROUND`. If either `CASE` or `CELL` are specified without `ROUND` or `TRUNCATE`, `ROUND` is selected automatically. === Statistics === The '''`/STATISTICS`''' subcommand causes statistics to be computed for the tabulated variables. ||'''Name'''||'''Effect''' || ||`CHISQ` ||Pearson chi-square, likelihood ratio, Fisher's exact test, continuity correction, linear-by-linear association|| ||`PHI` ||phi || ||`CC` ||contingency coefficient || ||`LAMBDA` ||lambda || ||`UC` ||uncertainty coefficient || ||`BTAU` ||tau-b || ||`CTAU` ||tau-c || ||`RISK` ||risk estimate || ||`GAMMA` ||gamma || ||`D` ||Somers' d || ||`KAPPA` ||Cohen’s kappa || ||`ETA` ||eta || ||`CORR` ||Spearman correlation, Pearson’s r || ||`ALL` ||all of the above || ||`NONE` ||no statistics || Even if selected, a statistic is only calculated when appropriate. If the `/STATISTICS` subcommand is specified without any options, the `CHISQ` option is selected automatically. === Missing === By default, SPSS will exclude observations with a user missing value for a tabulated variable. Add the '''`/MISSING=TABLE`''' subcommand to explicitly allow this behavior, or add '''`/MISSING=INCLUDE`''' subcommand to disable it. ---- == Data Model == The `CROSSTABS` command causes all pending transformations to execute, and reads the active dataset. ---- == See also == [[https://www.gnu.org/software/pspp/manual/html_node/CROSSTABS.html|PSPP manual for CROSSTABS]] |
SPSS Crosstabs
The CROSSTABS command creates a cross-tabulation.
Usage
crosstabs /tables=foo by bar.
To create a series of tables with foo as the row-wise variable in each, try:
crosstabs /tables=foo by ham spam eggs.
To create an N-dimensional table, try:
crosstabs /tables=foo by bar by baz.
Multiple /TABLES subcommands are allowed.
crosstabs /tables=foo by bar /tables=foo by baz.
Format
The default /FORMAT subcommand is /FORMAT=TABLES AVALUE. This causes SPSS to create tables (the alternative being NOTABLES) and sort values in ascending order (the alternative being DVALUE).
Cells
The default /CELLS subcommand is /CELLS=COUNT. This causes table cells to be populated with counts of observations.
Options are:
Name |
Effect |
COUNT |
count |
ROW |
row percent |
COLUMN |
column percent |
TOTAL |
table percent |
EXPECTED |
expected value |
RESIDUAL |
residual |
SRESIDUAL |
standardized residual |
ASRESIDUAL |
adjusted standardized residual |
ALL |
all of the above |
NONE |
display nothing |
If the /CELLS subcommand is specified without any options, the COUNT, ROW, COLUMN, and TOTAL options are selected automatically.
Count
By default, SPSS uses raw case weights without rounding for counting. Add the /COUNT=ASIS subcommand to explicitly allow this behavior.
To use rounding, first pick between CASE for rounding each case's weight before tabulating, and CELL for rounding the final tabulation. Then pick betwen ROUND for rounding to the nearest integer, and TRUNCATE for rounding towards 0. Add a subcommand with the specifications like /COUNT=CELL ROUND.
If either CASE or CELL are specified without ROUND or TRUNCATE, ROUND is selected automatically.
Statistics
The /STATISTICS subcommand causes statistics to be computed for the tabulated variables.
Name |
Effect |
CHISQ |
Pearson chi-square, likelihood ratio, Fisher's exact test, continuity correction, linear-by-linear association |
PHI |
phi |
CC |
contingency coefficient |
LAMBDA |
lambda |
UC |
uncertainty coefficient |
BTAU |
tau-b |
CTAU |
tau-c |
RISK |
risk estimate |
GAMMA |
gamma |
D |
Somers' d |
KAPPA |
Cohen’s kappa |
ETA |
eta |
CORR |
Spearman correlation, Pearson’s r |
ALL |
all of the above |
NONE |
no statistics |
Even if selected, a statistic is only calculated when appropriate.
If the /STATISTICS subcommand is specified without any options, the CHISQ option is selected automatically.
Missing
By default, SPSS will exclude observations with a user missing value for a tabulated variable. Add the /MISSING=TABLE subcommand to explicitly allow this behavior, or add /MISSING=INCLUDE subcommand to disable it.
Data Model
The CROSSTABS command causes all pending transformations to execute, and reads the active dataset.