I am trying to caluclate mean hours worked by post graduates (defined as educ >=120) and get a weird blip around 92 - see graph below. Any ideas what is causing this discontinuity/how to solve?
To a lesser extent, I also get this for mean hours worked for graduates (defined as educ >=100)
.
What you are observing is likely due to the fact that EDUC is actually a constructed variable that is a combination of two other variables, HIGRADE and EDUC99, which measure educational attainment in different ways. HIGRADE is available for years prior to 1992 and gives the respondent’s highest grade of school or year of college completed. EDUC99 is available beginning in 1992 and classifies high school graduates according to their highest degree or diploma attained.
As a result, while EDUC was constructed to maximize comparability over time, it is not perfect.