1. The app outputs projected salary based on two different models.
2. Follwing data were used to build salary models:
   • For my app I used the follwoing data
   • Available in R package car.
   • Ranks: assistant professor, associate professor, professor.
   • Discipline: theoretical” departments or “applied” departments.
   • Sex: male or female.
   • Years since PhD varies between 1 and 56.
   • Salary: 2008-09 nine-month academic salary in a college in the U.S.

Two linear regression models are created to model professor salary based on the rank, discipline, sex and years since the completion PhD.

• Model 1 is created based only on two factors: rank and discipline.

 model1 <- lm(salary ~ rank + discipline, data = Salaries)

• Model 2 is based on four factors: rank, discipline, sex and years since PhD.

 model2 <- lm(salary ~ rank + discipline + sex + yrs.since.phd, 
              data = Salaries)

Let's consider the following example:

  newdata <- data.frame(rank = "Prof", discipline = "A",sex = "Female", 
                        yrs.since.phd = 20)
  predict(model1,newdata)

##        1 
## 119788.2

  predict(model2,newdata)

##        1 
## 115137.1

Models produced different salary as they were build using different factors.