Question

The following model is estimated to find the effect of experience
(?????), age (???), and years played in college (????) of the player on points per
game (??????):
?1????? = 33.21 − 1.07 ??? + 2.36 ????? − 1.28 ???? − 0.07?????!
(6.986) (0.295) (0.450) (0.405) (0.023)
? = 269, ?! = 0.1412, ? A ! = 0.1282
a. (8 pts) What is the causal effect of experience on points per game? Write
mathematically and explain verbally. Test whether the quadratic term ?????! is
statistically significant at 5% significance level.

Answer 1

The points per game initially increase with experience but will eventually decline after a certain level of experience.

dpoints/dexper = 2.36 - 2*0.07 exper

setting this to 0 we get,

exper = 2.36/0.14 = 16.85

This means that points per game are increasing as we increase experience to roughly 17 years but will decrease after that.

Testing the statistical significance:

Test statistic = coefficient/standard error = -0.07/0.023 = -3.04

Z critical value at 95% = -1.96

since the absolute value of the test statistic = 3.04 is greater than the absolue value of z critical value at the 5% level, the null hypotheisis of no sigbificance can be rejected.