Normally a regression application involving more than one DV is analyzed using canonical correlation (sometimes called multivariate regression), but SPSS requires multiple predictor and multiple dependent variables:

http://www.statisticssolutions.com/canonical-correlation/

Thank you Paul Yarnold for sharing your knowledge

You are welcome, and it is my pleasure, Amin Rahman.

Perhaps an answer to your question lies in (heuristic) "causal" approaches, I am not current in these methods, and I don't know if this is supported by SPSS.

Hello,

I advise you to use R, which is more efficient in statistical analysis in the case of multiple dependant variables (interactions).

In R, you find more Predefined commands for such cases of data analysis.

Regards

Canonical correlation analysis is a methodology that identifies the association between composites of sets of multiple dependent and independent variables. For ordered measures:

• One DV, one IV --- correlation analysis
• One DV, two or more IVs --- multiple regression analysis
• Two or more DVs, two or more IVs --- multivariate regression (canonical correlation) analysis

Here are two more introductions to CC:

• http://www.mvstats.com/Downloads/Supplements/Canonical_Correlation_6e.pdf
• Chapter 9 in: http://www.apa.org/pubs/books/431643A.aspx

"Canned" ("predefined") linear dimension-reducing algorithms can identify confounded solutions. Here is an example of this problem (and the solution) for PCA:

• https://odajournal.com/2013/09/19/the-use-of-unconfounded-climatic-data-improves-atmospheric-prediction/

Hello , You may try to use step wise regression.  

step wise regression is also a good option

How do you conduct regression analysis using SPSS when there is more than one dependent variable?

For more than one dependent variable i.e. multivariate analysis you might want to consider Structural Equation Modeling (SEM) like AMOS (additional module of SPSS) or SmartPLS etc.