Data classification problem

[ecognom]

Hi to all!

Your help is highly appreciated in advance.

I have a problem with data. I have a longitude data, which is allocated in different classes e.g. education. But due to a reform in education evaluation,I have a spike in my data. More specifically, spike occurs due to individuals in one education-level being nominated to a higher education level. There was no concept change in education, only that education level , which prior was lower, now became higher one and this causes a spike (seems like more high educated individuals but in reality not) which has no real terms.

My idea is to fix the years prior the change to correspond the current classification.

Would you have any thoughts of this?

Thank you! :D