EXAMINE VARIABLES=AgeAtLos DepressT PVLoss SuppTotl
  /PLOT HISTOGRAM NPPLOT
  /MESTIMATORS HUBER(1.339) ANDREW(1.34) HAMPEL(1.7,3.4,8.5) TUKEY(4.685)
  /STATISTICS DESCRIPTIVES
  /CINTERVAL 95
  /MISSING LISTWISE
  /NOTOTAL.

*Correlation and Robust correlations based on ranking the data

CORRELATIONS
  /VARIABLES=AgeAtLos DepressT PVLoss SuppTotl
  /PRINT=TWOTAIL SIG
  /MISSING=LISTWISE.
NONPAR CORR
  /VARIABLES=AgeAtLos DepressT PVLoss SuppTotl
  /PRINT=SPEARMAN TWOTAIL SIG
  /MISSING=LISTWISE.

*Scatterplot

GRAPH
  /SCATTERPLOT(MATRIX)=DepressT AgeAtLos PVLoss SuppTotl
  /MISSING=LISTWISE.

REGRESSION
  /DESCRIPTIVES MEAN STDDEV CORR SIG N
  /MISSING LISTWISE
  /STATISTICS COEFF OUTS CI(95) R ANOVA COLLIN TOL
  /CRITERIA=PIN(.05) POUT(.10)
  /NOORIGIN 
  /DEPENDENT DepressT
  /METHOD=ENTER AgeAtLos PVLoss SuppTotl
  /SCATTERPLOT=(*ZRESID ,*ZPRED)  
  /RESIDUALS HIST(ZRESID) NORM(ZRESID)
  /SAVE PRED MAHAL COOK LEVER RESID.

*Statistical Test of normality of residuals

EXAMINE VARIABLES=RES_1
  /PLOT NPPLOT
  /STATISTICS NONE
  /CINTERVAL 95
  /MISSING LISTWISE
  /NOTOTAL.

*Influence plot to detect outliers.

IGRAPH
  /VIEWNAME='Scatterplot'
  /X1=VAR(COO_1) TYPE=SCALE
  /Y=VAR(LEV_1) TYPE=SCALE
  /COLOR=VAR(MAH_1) TYPE=SCALE
  /COORDINATE=VERTICAL
  /POINTLABEL=$case ALL
  /YLENGTH=5.2
  /X1LENGTH=6.5
  /CHARTLOOK='NONE'
  /SCATTER COINCIDENT=NONE.

* NonLinear Regression. The only way to do the bootstrap for a regular regression is to first run the regression then input the parameters from there into the nonlinear regression procedure.
*There is no control apparently of how many bootstrapped samples one can take, and while 100 is fine for validation purposes, it is low for interval estimation.
*The outfile has to be changed to a location on your computer

Set Printback on.
MODEL PROGRAM  b1=-0.16437 b2=-0.10648 b3=.61466 b0=59.96181.
COMPUTE  PRED_=b0+b1*SuppTotl+b2*AgeAtLos+b3*PVLoss.
_set Printback off.
CNLR  DepressT
  /OUTFILE='C:\Documents and Settings\mjc0016\Desktop\SPSSFNLR.TMP'   
  /PRED PRED_
  /BOOTSTRAP
  /CRITERIA ITER 500 STEPLIMIT 2 ISTEP 1E+20.

*What's left out: Statistical Test for heteroscedasticity and nonlinearity, Robust regression, Validation/bias-adjustment (you could do it the old fashioned way by randomly splitting the data into training
 and tests sets and going from there), comparison measures of fit.  Gist: more code to do a lot less.