Normalising, stationarising and Creating ARIMA models (very urgent please)

Hi any and all,

I need help desperately with

1) Normalising a set of data which proved to be non-normal after Shapiro-wilks and kolmogorov-smirnov tests

2) Sationarising this data so I can use it for ARIMA modelling.

3) Creating the right ARIMA time series model to be able to extract the unpredicted component from the variables

My Statistics package is SPSS

Please, do help. I'm in dire straits. I'm no expert in statistics so technical language will only make the issue a bigger mystery.

Thanks!