You should investigate the trendline before and after the $8/15$ initiative. What impact did the changes have? If they had not made any changes, how many defects would you have predicted in $2017$ and $2018($use analysis, do not guess$)?$ How does this compare to the actual number of defects?