Trying to do a subjective analysis of a regression

[acemcmac]

Regression Statistics
Multiple R 0.998549946
R Square 0.997101995
Adjusted R Square 0.995411493
Standard Error 1.553187767
Observations 20

ANOVA
df SS MS F Significance F
Regression 7 9960.236793 1422.89097 589.8257121 2.93494E-14
Residual 12 28.94870687 2.41239224
Total 19 9989.1855

Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Lower 95.0% Upper 95.0%
Intercept 16.54794708 9.453493918 1.75045832 0.105537828 -4.04944673 37.14534089 -4.04944673 37.14534089
col2 0.358128136 0.262328459 1.3651898 0.197240936 -0.213436475 0.929692747 -0.213436475 0.929692747
col3 0.277825795 0.161904067 1.715990219 0.111842711 -0.074932863 0.630584453 -0.074932863 0.630584453
col4 -0.201387607 0.271910508 -0.740639295 0.473156004 -0.793829709 0.391054495 -0.793829709 0.391054495
col5 0.093030658 0.039491549 2.355710565 0.036333587 0.006985966 0.179075351 0.006985966 0.179075351
col6 0.404645413 0.280207411 1.444092473 0.174310969 -0.205874088 1.015164913 -0.205874088 1.015164913
col7 -0.243703564 0.17985855 -1.354973474 0.20038499 -0.63558168 0.148174552 -0.63558168 0.148174552
col8 0.155194397 0.099103288 1.565986353 0.143329223 -0.060733119 0.371121912 -0.060733119 0.371121912

So.... the higher the T-stat, the less fits or the more it fits into the regression? P value? Should I pull any of the columns? I'm trying to establish a relationship.... Fvcking econ....

 

[Kalbi]

does anyone know what the play is called when the ball hits the ground on the 3rd strike and you have to tag or throw the batter out?

 

[silverpig]

Chi-Squared baby... minimize it and you're golden

 

[acemcmac]

Originally posted by: Kalbi
does anyone know what the play is called when the ball hits the ground on the 3rd strike and you have to tag or throw the batter out?

Originally posted by: silverpig
Chi-Squared baby... minimize it and you're golden

WTF are you guys talking about?

 

[silverpig]

Originally posted by: acemcmac

Originally posted by: Kalbi
does anyone know what the play is called when the ball hits the ground on the 3rd strike and you have to tag or throw the batter out?

Originally posted by: silverpig
Chi-Squared baby... minimize it and you're golden

WTF are you guys talking about?

Chi-Squared is the parameter that tells you how good your fit is. Basically you take the value of the function of your fit at a certain point, and find the difference between that function and the data point it passes closeset to. You square it, do the same for all of the points, average them, and then square root it (I think).

 

[UTmtnbiker]

That's also the R-squared value in this paritcular case. Also....look for large p-values. Generally, anything over 10% (.1) you'd like to throw out as they don't correlate well to the regression. Throw out that particular column and then re-run the regression.

In general, when I'm looking for fit, I look at the R-squared for trend line fit and the p-value to identify which parameter doesn't fit the regression. Throw it out, re-run, re-evaluate R-squared and p-value.

Originally posted by: silverpig

Originally posted by: acemcmac

Originally posted by: Kalbi
does anyone know what the play is called when the ball hits the ground on the 3rd strike and you have to tag or throw the batter out?

Originally posted by: silverpig
Chi-Squared baby... minimize it and you're golden

WTF are you guys talking about?

Chi-Squared is the parameter that tells you how good your fit is. Basically you take the value of the function of your fit at a certain point, and find the difference between that function and the data point it passes closeset to. You square it, do the same for all of the points, average them, and then square root it (I think).