Hi,

Wondered if anyone could help.

I've created an Interrupted Time Series analysis approach to interpret the effect that can be associated with the implementation of a revised planning policy framework within the UK.

I only have 12 annual observations relating to 2007 to 2018. The policy can be considered to have been implemented from 2012. In a traditional ITS approach the pre-policy trend would therefore be calculated based solely upon 2007 to 2011.

However, it is known that the effect of the policy would not be evident until around 2014. Therefore, years 2012 and 2013 are almost certain to merely be reflections of the pre-2011 policy. In such circumstances is it appropriate to include the 2012 and 2013 observations as part of the pre-policy trend?

There doesn't appear to be any research precedence for such, but it would seem logical to do so......

Any advice would be hugely appreciated

John