The Bureau of Labor Statistics is an office within the US Department of Labor. Every three months the bureau releases a report containing jobs and salary data of US workers. One statistic reported is the median weekly salary for full-time workers. The scatterplot shows the growth of median weekly salary, starting in January 2010. An equation of the least-squares for the data in the scatterplot is Median salary hat = 17.327 (year) minus 729.327, where 2010 represents year zero. The value of r for the scatterplot is 0.977. A graph titled median salary versus years since 2010 has years since 2010 on the x-axis, and median salary (dollars) on the y-axis. Points form a line with positive slope.

A graph titled residuals versus years since 2010 has years since 2010 on the x-axis, and residual on the y-axis. Points decrease, and then increase.

Based on the least-squares regression and residual plot, is a linear model suitable for this data set?

A linear model is suitable because the value of r is close to 1.
A linear model is suitable because the residual plot shows a curved pattern.
A linear model is not suitable because the residual plot shows a clear pattern.
A linear model is not suitable because there are more positive residuals than negative residuals.