Assume you have a data set that includes only people with blue, black, green and brown eyes $($each person can have only one eye color$).$ You can create binary $($dummy$)$ variables for each of these eye colors. Assume you want to estimate a multiple linear regression model with binary variables on the right$-$hand side. You can choose to include the following choice of binary variables and an intercept $($a variable including entries of $1$ only$).$ One of these choices would violate the assumption of No Perfect Collinearity $($A$4).$ Which one is it$?$ Each of the colours in the following answers describes a binary variable that turns to $1$ if the individual has that eye colour and is zero otherwise. For example: "Brown, Blue" means that you would include two binary variables: one that is $1$ if the person has brown eyes and one that turns to $1$ if the person has blue eyes. "Black, Brown, Green, intercept" is a set of three variables with the intercept, etc.