In a recent analysis of salaries (in $1,000) of financial analysts, a regression of salaries on education, experience, and gender is run. Gender equals one for men and zero for women. The regression results from a sample of 230 financial analysts are presented below, with t-statistics in parenthesis. Salaries = 34.98 + 1.2 Education + 0.5 Experience + 6.3 Gender (29.11) (8.93) (2.98) (1.58) Holding everything else constant, do men get paid more than women? Use a 5% level of significance. No, since the t-value: A) does not exceed the critical value of 1.96. B) does not exceed the critical value of 1.65. C) exceeds the critical value of 1.96. ------------------------------------ I thought the answer would be A but it is B. Why? Thanks.
Because it is a one sided hypothesis test. Null: M>W Alternative: M<=W.
that’s right. You’re the man.
> Because it is a one sided hypothesis test. Null: M>W Alternative: M<=W. Not correct. Null hypothesis has to have an equal sign.
yes dreary is right, the Null hypothesis must have an equality sign in it. the Null would have to be M<=W, alternative M>W
How did you find Salaries of men and women here? I am totally forgetting everything. Plz help here to have me recall this dummy variable concept.
dummy variable is something that has either a 0 or a 1 value. So say belongs to Q1=1, does not belong=0 and you run a regression after doing the above conversion. Subsequent to running the regression - you would get the coefficients pertaining to that assumption, and you can subsequently perform that regression calculation. Is there anything else that I am forgetting?
There was a little reason why we used n-1 dummy vars, when we had n classifiable units. Something related to intercept being the avg salary when gender = 0 and interset+dummyCoeff being the avg sal if gender = 1 …Or something like that.
Dummy variables exist so that you can include qualitative factors into the equation I think. hence the 0 or 1. The reason why you can use up to n-1 variables is as follows: say if you had an equation y = a + b1x1 + b2x2 + b3x3 + b4x4 +d1q1 +d2q2 + d3q3 + d4q4 The four dummy variables relate to which seasonal quarter it is (ie 1q2008 and so forth). if its the fourth quarter then only d4q4 is equal to 1. if its the 1st quarter than the first dummy variable =1. Now lets say thats dummy variable 1,2 and 3 are all equal to 0. This in itself forces the 4th dummy variable to be equal to 1. Which would be ok except that then the dummy variables are not independent from eachother. to get around this you would have the equation above but with one less dummy variable, hence the n-1. I think this is right, but my understand is a little shady.
yea - winner! This ‘correlation of the independent variables’ is what I was looking for.
so you were testing us?
Not at all. The concept was just not comin’ back to me and I was not able to focus on corp finance (dividend policy) which was the POA today. ThnX!
multicollinearity or whatever is it