This should be an option under Regression Modeling so as not to require R programming experience in the use of the poly() function. Have seen similar problem from other users under TIBCO Community, I posted this same fix below after I researched the R function poly(). Quote from a Support Case:
A Spotfire user of the Linear Regression feature who wants 'poly()' coefficients that correspond directly to the x, x^2, x^3, etc. model components needs to specify 'poly(.., raw = TRUE )' in the formula the user provides in the Regression dialog's Formula field.
Using syntax like this:
poly(`X1`, degree = 2, raw = TRUE)
in place of this:
poly(`X1`, degree = 2)
Here is a link to the problem under TIBCO community: https://community.tibco.com/questions/how-do-i-interprete-estimates-given-quadratic-formula-linear-regression-tool