Statsmodels P>|t|

Question

in the statsmodels summary, what does the P>|t| and t mean in relation to the variables when it says something like:

    coef    std err    t    P>|t|    [95.0% Conf. Int.]
Intercept   7.0326  0.458   15.360  0.000   6.130   7.935
TV          0.0475  0.003   17.668  0.000   0.042   0.053

Ragi

Answer 1

t is the value of the t-statistic for testing if the corresponding coefficient is different from 0. (Hypothesis H0: coef==0, H1: coef!=0)

Pr>|t| is the p-value for this hypothesis test. A low p-value means, that you can reject the null-hypothesis and accept the alternative hypothesis (coef!=0).

Bad

can u plz clarify what qualifies as low and acceptable p-values.. – Serialize

In social science, Pr>|t| under .05 is generally considered significant, .10 mildly significant, and .01 very significant. But note that this only applies if one plans in advance what variables to run. If one instead keeps running regressions with different variables until significance is found, it defeats the purpose of the statistic, and you will have a falsely significant result. – Kweisui

From a scientific perspective the previous comment is false. One doesn't say "mildly significant", "very significant" or similar because significance levels have no declination. Instead, to be correct, one can say something is significant at P < 0.05. Or something is significant at a 5% probability level (or a different level). – Rattlesnake