I do not know the specific SPSS routine but in general the standardization is based upon the normalization on the basis of mean and variance of each single variable (so to eliminate the effects due to the order of magnitude of each regressor) and not on the 'relative weight' of each descriptor (indpendent variable) in the final model. Thus it is perfectly sound you can have a standardized regression coefficient greater than one

There is some relevant discussion here:

• https://stats.stackexchange.com/questions/120201/magnitude-of-standardized-coefficients-beta-in-multiple-linear-regression

I speculate that one reason for confusion about this issue is that in a simple linear regression model (i.e., an OLS model with one explanatory variable), the standardized regression coefficient = Pearson r, and |r|

I would say, a standardized beta weight greater than one is a sign of suppression, especially cooperative suppression. Such suppression increases the predictive value of the predictors and thus is of potentially great value.

Generally, a common way to standardize you can do that, divide both the predictor and response variable by their standard deviations and fit the regression, then it is mathematically impossible to see a coefficient outside of the -1 to 1 range (since the slope will be the same as the correlation). otherwise, if you didn't standardize the response variable then it would be easy to see an estimated coefficient outside of -1 to 1 depending on the scale of the response variable.

1 in which unit? Your question makes no sense to me.