Please answers all parts of the qestion and write clearly and I will give you a
ID: 1162715 • Letter: P
Question
Please answers all parts of the qestion and write clearly and I will give you a good rating!
4. You are hired by an NBA team to determine what factors determine a player's points per game "Points". You decide that the following regression would be appropriate, where "Minutes" is minutes played per game, "Height" is the player's height in inches, and "Talent" is a measure of the player's basketball skill. Points Bo +Bi Minutes +B2ln(Heighti) +BsTalent+u (a) What signs do you expect for Bi, B2, and Bs? Unfortunately, the variable for "Talent" is omitted. You wish to figure out how omitting this variable is going to bias the other coefficients in your regression. (b) What correlation do you expect between "Talent" and "Minutes"? Explain. In light of this, is your estimated Bi likely to be biased upwards or downwards? (c) Suppose that taller people have an easier time making it to the NBA, even if they are not quite as talented as their shorter peers. In light of this, explain how omitting talent bias B2. (d) List two reasons why researcher may omit an important variable in a regression.Explanation / Answer
Answer
a)
The more time a player played more should will be his scoring points Hence Positive
As height plays huge positive role in basketball, hence it should also be positive.
Talent should also have positive impact on scoring points
b)
If a player is talented then he should be played more time in a game. Hence there is some sought of positive relation between time played and skill. Hence correlation should be positive
We have to calculate VIF to check multicollinearity. Without that it is not possible to check biasness of an estimate. If there is no multicollinearity then there should be no biasness.
c)
Lets take into consideration that situation mentioned in the statement is generally true.
Then there is negative relation between talent and height and there is positive relation between talent and points scored (as mentioned in part a)
Hence ommision of talent leads to downward biasness.
d)
There can be many reasons for omission of important variables in regression:
1) Lack of data for important variable
2) Chances for having multicollinearity between explainatory Variables
3) They saw decrease in Adjusted R2 when they include that important variable.