On page 201 of Schweser book 1: In the example of detecting multicollinarity, part of the answer said “none of the independently variables individually is statistically significant to any reasonable degree, since the p-values are larger than 10%, and we never test at significant levels larger than 10%” where is the 10% thing coming from? and do we never test at significant levels >10% as they said? thanks! vitamin
Because testing at levels > 10% would be pointless. The statement you can make if you are testing at the 20% level is “one in 5 times, This is nonsense”. That’s not desperately useful. It’s not a hard and fast rule, it’s just a bit of common sense.
We usually test at 0.05 because one time RA Fisher said “5% is no less arbitrary than any other number”
10% is an arbitrary number. You can test at the 1% level of significance, the 5% level of significance or the 10% level of significance.
folks, some how can’t get the best part of study session 11 to stick in ma mind especially readings 42-45. any body help: are they crucial parts of equity and how best cani remember them?
sorry this was supposed to be a new post…pls ignore