Isn’t this a bad thing? Hence if a T-stat is too high, it is statistically significant and thus isn’t reliable for forecasting? I ran into a question, in which the correct answer stated “the main weakness in this model is the lack of significan”. The t-star for this independet variable was -.48

Am I missing something?

Thanks for your help

No, statistical significance is actually a good thing. When a variable is statistically significant (different from zero), it helps to explain the dependent variable. When the t-statistics (coefficient/standard error of variable) is greater than the critical value (in case of an two-sided test), you can say the variable is statistically significant.

A low level of statistical significance means that it’s difficult to say that the coefficient didn’t arise from chance. A high level means that it’s likely that the relationship really exists.

Oh man, I have been interpreting it wrong for months. But this does make sense now.

Thanks!!!