In Statistica workspace node browser, there are 3 different SVB nodes for I-Tree CHAID regression analysis:  “CHAID Standard Regression (SVB)”, “Advanced Regression CHAID (SVB)”, and “Exhaustive Regression CAHID (SVB)”.

In-DB CHAID does the regular/non-exhaustive type of I-Tree CHAID analysis. The Advanced CHAID node in above screenshot is the WPF node version of the Standard CHAID algorithm.

Workspace In-DB CHAID node uses similar algorithm as in Advanced CHAID node, except some minor differences observed when comparing the parameters for the two nodes:

1. The intervals for continuous variables are based on percentiles (value range 2-50 allowed) for In-DB CHAID node, while ITree CHAID (Advanced CHAID) uses General Chi-square statistics. 

In-DB CHAID Interval Options for continuous variables:

Advanced I-Tree CHAID Chi-Square statistics for continuous variables:

2. The stopping parameter: In-DB CHAID node uses "minimum n of cases" while Advanced ITree CHAID node uses "minimum n (%) of cases".

In-DB CHAID stopping parameter: minimum n of cases:

Advanced I-Tree CHAID stopping parameter: minimum n % of cases:

Note:

1. Users may not get similar results for In-DB CHAID and Advanced I-Tree CHAID due to above two differences.

2. If user only has categorical predictors and no continuous predictors in the analysis, he may can get similar results by adjusting the stopping parameter of minimum number or % of cases in the two nodes. In such situation, user can compute the minimum % of cases such that it will match the minimum number of case as stopping parameter.