The research team used the data mining application
“Eureqa”, also known as “Formulize” (creativemachines.
cornell.edu/eureqa), to conduct a symbolic regression
analysis of the MSOSW data. A compact data set was
created with no missing data, and no manipulated or
replacement data, with 157 records of males and 168 records
of females, for a total of n = 325. After a dependent
variable and the predictors were selected, a series of oneminute
searches were conducted using a fitness metric of
maximizing the correlation of the models. The searches
resulted in sets of equations that can be represented as a
network of production relationships among the variables
in the data set. Here we are concentrating on just the production
relationships that explain the CIQ Total score. To
verify the primary result, we then conducted a traditional
linear regression for the dependent variable, CIQ Total (Career
Interest Questionnaire Total Scale Score