I am currently doing some coursework which requires me to run a multivariate regression analysis to determine changes in Hong Kong Rental Growth (Y-variable) from various X-variables. The X-variables I chose are Hong Kong GDP, Y-o-Y Change in Vacancy, Y-o-Y change in unemployment rate, best lending Y-o-Y change, Change in CPI & Private Office completions. I am currently trying to transform the data so that I can see a significant regression. Please can you give any advice on which transformations to perform to get a good regression? What I would like to know is, if I perform transformations, do I need to perform the same type of transformations on all variables eg. only log my variables or can I apply different transformations on my model eg. log transformations and year-over-year transformations at the same time