statable Statistics for study and research

The "Commands" page

In brief

Commands run here as in PSPP or SPSS, such as .sps files from a course.

Syntax that SPSS writes when pasting from the dialogs contains parts PSPP does not know: /ES DISPLAY(TRUE) with every t test (since SPSS 27), /ES and CILEVEL with ONEWAY, /MISSING LISTWISE and OUTS with REGRESSION, LOGISTIC REGRESSION VARIABLES … /METHOD=ENTER …, UNIANOVA. Statable leaves such parts out or rewrites them when this does not change the other numbers, and names every place with its reason below the output ("Adjusted for PSPP"). Anything that would change numbers (for example PAIRWISE, stepwise regression, another reference category) stays; then the message explains the error.

Statable runs PSPP commands, not SPSS scripts: an .sps file runs on the "Commands" page as far as PSPP knows the commands; Python programs and SPSS extensions do not run. When importing a questionnaire with a codebook (dialog "Questionnaire recognised", "Codebook as well"), Statable takes only labels, missing values and data from an .sps and skips all other commands. PSPP cannot do GLM with WITH (analysis of covariance) or NONPAR CORR (Spearman); for these Statable has the questions "Do groups differ?" (with a control variable under "More options") and "Are two things related?".

QUICK CLUSTER on the commands page is computed by PSPP. PSPP picks different initial centres than SPSS and shows them; the final centres are the same.

Without open data only commands that need no data will run.

Related

Open in Statable