Skip to content
Prev 3008 / 15274 Next

Statistically significant in linear and non-linear model

Thank you for your comments.

Maybe I could specify my question more clearly. For example, a linear model is y = alpha + beta*x + epsilon, and this beta is not statistically significant. Then a non-linear model, e.g., y = eta + gamma*exp(x) + omega, since exp(x) = 1 + x + (x^2)/(2!) + (x^3)/(3!) + ..., then y = eta + gamma*(1+x+x^2/2+...) + omega. But beta in the former formula is not statistically significant, is there probability that gamma is statistically significant?

And I will post this to r-help also, thanks.

HC