Precision of displayed values in data mining results

In this section Hide

This topic covers the precision with which various computed numeric columns in the data mining results table appear. Note that the full precision of numeric values is present in the underlying database.

There are three types of computed numeric columns to consider: relative reporting ratio measures, expected values, and probability values.

Relative reporting ratio measures

MGPS columns of this type: EB05, EB95, EBGM, ER05, ER95, ERAM, INTSS, PRR, ROR05, ROR95, ROR, RR, EBMAX, RR_IND, EB05_IND, EB95_IND, EBGM_IND, EBGMDIF_IND, E2D_DIV_E_IND, E2D_DIV_F_IND.

Logistic regression columns of this type: RR, LROR, LR05, LR95 in the tables for Results and Covariates, INT_REGR, INT_LOF, INT_TOT, INT05, INT95 in the Interactions table

Expected values

Columns of this type: E, F, EXCESS, EXCESS2, PRR_CHISQ, E_IND, EXCESS_IND, EXCESS2_IND, E2D_IND.

Probability values

(In the event that a value is 1 or greater, it will display with two digits to the right of the decimal point.)

Columns of this type: P_VALUE, Q.