Id wonder if anyone could help me. Im having one of those brain freeze moments when you think you may know the answer but really would appreciate a second opinion…

Im looking at change between two points of time (from two cross sectional surveys conducted in 2006 and 2008) for 2 areas (say area X and area Y). Am I right in thinking area X change would be significant compared to area Y change if the minimum change for area X from CI’s is greater than the max change by area Y from CI’s. Im using confidence intervals for proportions

e.g. For a hypothetical question
Area X Area Y
2006 2008 2006 2008
Percentage stating 'Z' 50.0 51.4 60.0 60.2
Upper CI 50.8 52.2 61.8 61.9
Lower CI 49.2 50.6 58.2 58.5

So from the table above:

Area X min change = 50.6 - 50.8 = -0.2 i.e lower ci 2008 – upper ci 2002
Area Y max change = 61.9 - 58.2 = 3.7 i.e upper ci 2008 – lower ci 2002

Min change for area X is not greater than max change for area Y = difference in change is not significant

Or am I completely of the mark with how I am trying to do this. In the back of my head I think I may be terming some differences not significant when in fact they are significant.

Thanks for any help they you can provide


My head is hurting from going round in circles!