Exercise

# Collinearity and inflation (2)

In the last exercise, you experimented with removing each of three covariates (`sector`

, `exper`

, and `age`

) from the following model of wages:

```
model_1 <- lm(wage ~ educ + sector + exper + age, data = CPS85)
```

For each variation, you measured the effect on inflation and found that leaving out either `age`

or `exper`

brought a signifcant reduction in the variance inflation factor.

In this exercise, you'll build a new model excluding `exper`

, then find the new confidence interval on the effect of `educ`

on `wage`

.

Instructions

**100 XP**

- Build
`model_2`

the same as`model_1`

, but leaving out the worst offending covariate,`exper`

. - Find the new confidence interval on the effect of
`educ`

on`wage`

for`model_2`

.