Question

Provide three different examples of Y and X (do not use income and education) in which we can expect the following situations to be true. These examples should be in some way connected to economics. Provide at least one sentence justifying each example. (a) In a simple linear regression model: Y = Bo + B1X+U the conditional variance of U for any given value of X is most likely very large. (b) In a simple linear regression model Y = Bo + B1X +U, the conditional expectation of U for any given value of X is most likely not constant.

Answer 1

a,

Regression of instances of late arrival to college on distance from home to college couldhave very large
conditional variance of 'U' for any given valueof 'X'. because there are many significant factors which affect 'U' like traffic , weather etc. which may vary greatly from day to day.

b,

Regression of revenue made by firms on size of the firm,it  might not have constant error term. as 'U' becomes
larger the size becomes large, there will be more factors while making variation in the revenue. whereas smaller firms would not face as many factors of variation.