Data defines the model by dint of genetic programming, producing the best decile table.

A Pithy History of CHAID and its Offspring
Bruce Ratner, Ph.D.

The introduction of AID – described as an automatic interaction detection program – was in the Journal of the American Statistical Association (Morgan and Sonquist, 1963). "The actual computer program, called the Automatic Interaction Detector, was developed with funds from the National Science Foundation and described in a monograph (Sonquist and Morgan, 1964). AID was later improved and described in another monograph (Sonquist, Baker and Morgan, 1974), and then improved again and renamed Search." [1]

CHAID is extension of AID with the use of the Chi-squared statistic (for the Bonferroni correction for the multiple comparisons carried out in the CHAID algorithm). CHAID was the Ph.D. dissertation topic of South African Gordon V. Kass in 1980. Kass never had an inkling, after his receiving his doctorate, that his original thinking – recursive partition of a population/dataset into mutually exclusive and exhaustive segments – would have the significant contribution it has made to the world of statistics, and subsequently, related quantitative fields. He never revisited his original dissertation, and never got a penny of royalties for his CHAID. Whereas, many practitioners, academics, and venture capitalists, who invested in commercializing CHAID software, have greatly benefitted financially. 

Other programs, mainly by Sonquist and Morgan in 1973, emerged with adjustments for specific quantitative objectives under consideraton: MNA, MCA, and THAID. MAID, a multivariate version of AID, was developed in the late 1980s. Interested readers can pursue the innards of these programs, which will ultimately lead to others, as my list is not comprehensive. 

For more information about this article, call Bruce Ratner at 516.791.3544 or 1 800 DM STAT-1; or e-mail at

Sign-up for a free GenIQ webcast: Click here.