Deconstructing Prospect Theory

Deconstructing Prospect Theory

Keywords: 
Risk preferences
Decision making
Psychology
Economics
Prospect Theory
multidisciplinary research
Status: 
Active Analysis
Reference: 
Article under preparation
Summary: 

Coming soon.

SPSS Syntax: 
DATASET ACTIVATE DataSet1.
USE ALL.
do if $casenum=1.
compute #s_$_1=140.
compute #s_$_2=443.
end if.
do if  #s_$_2 > 0.
compute Holdout=uniform(1)* #s_$_2 < #s_$_1.
compute #s_$_1=#s_$_1 - Holdout.
compute #s_$_2=#s_$_2 - 1.
else.
compute Holdout=0.
end if.
VARIABLE LABELS Holdout '140 cases from the whole file (Holdout)'.
FORMATS Holdout (f1.0).
FILTER  BY Holdout.
EXECUTE.
 
USE ALL.
compute main = 0.
IF (Holdout =  0) main=1.
VARIABLE LABELS  main 'Main Sample (n=303)'.
FORMATS main (f1.0).
EXECUTE.
 
 
43. Now use the main sample for reliability
----------------------------------------------------------
PROSPECT THEORY ------------------------------------------
----------------------------------------------------------
 
USE ALL.
FILTER BY main.
EXECUTE.
 
RELIABILITY
  /VARIABLES=PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08
  /SCALE('PT-Certainty effect') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
---------
RELIABILITY
  /VARIABLES=PT_PTMR01 PT_PTMR02 PT_PTMR03 PT_PTMR04
  /SCALE('PT-Mirror effect') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
-----------
RELIABILITY
  /VARIABLES=BIS01_PTIS01 BIS02_PTIS02 BIS03_PTIS03
  /SCALE('PT-Isolation effect') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
--------------
ON A FULL SAMPLE I get exactly the same reliability results
 
FILTER OFF.
USE ALL.
EXECUTE.
 
----------------------------------------------------------
DOSPERT-R  -----------------------------------------------
----------------------------------------------------------
 
RELIABILITY
  /VARIABLES=DSPRT_01F DSPRT_02F DSPRT_03F DSPRT_04F DSPRT_05F DSPRT_06F
  /SCALE('DOSPERT-Financial') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
  
RELIABILITY
  /VARIABLES=DSPRT_01F DSPRT_03F DSPRT_05F
  /SCALE('DOSPERT-Financial-betting') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
 
RELIABILITY
  /VARIABLES= DSPRT_02F DSPRT_04F DSPRT_05F 
  /SCALE('DOSPERT-Financial-investing') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
 
----------------------------------------------------------
SIRI  ----------------------------------------------------
----------------------------------------------------------
 
 
RELIABILITY
  /VARIABLES=SIR_ZS01 SIR_ZS02 SIR_ZS03 SIR_ZS04 SIR_ZS05 SIR_ZS06 SIR_ZS07 SIR_ZS08 SIR_ZS09 
    SIR_ZS10 SIR_ZI01 SIR_ZI02 SIR_ZI03 SIR_ZI04 SIR_ZI05 SIR_ZI06 SIR_ZI07
  /SCALE('SIRI - Full scale') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
  
Subscales
SOCIAL RISK
 
RELIABILITY
  /VARIABLES=SIR_ZS01 SIR_ZS02 SIR_ZS03 SIR_ZS04 SIR_ZS05 SIR_ZS06 SIR_ZS07 SIR_ZS08 SIR_ZS09 
    SIR_ZS10
  /SCALE('SIRI - SOCIAL RISK') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
 
RELIABILITY
  /VARIABLES=SIR_ZS01 SIR_ZS02 SIR_ZS04 SIR_ZS05 SIR_ZS07 SIR_ZS09 SIR_ZS10
  /SCALE('SIRI - SOCIAL RISK') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
 
INSTRUMENTAL RISK
  
RELIABILITY
  /VARIABLES=SIR_ZI01 SIR_ZI02 SIR_ZI03 SIR_ZI04 SIR_ZI05 SIR_ZI06 SIR_ZI07
  /SCALE('SIRI - INSTRUMENTAL RISK') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.  
 
Now combined scale
 
RELIABILITY
  /VARIABLES=SIR_ZS01 SIR_ZS02 SIR_ZS04 SIR_ZS05 SIR_ZS07 SIR_ZS09 SIR_ZS10 SIR_ZI01 SIR_ZI02 SIR_ZI03 SIR_ZI04 SIR_ZI05 SIR_ZI06 SIR_ZI07
  /SCALE('SIRI - Combined') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE
  /SUMMARY=TOTAL.
 
alpha = .869 (14 items, 2 x 7)
n = 124 (out of 303)
 
----------------------------------------------------------
Holt-Laury  ----------------------------------------------
----------------------------------------------------------  
MAkes no sense to prune down.
--------------------------------------------------------------------------------------------
  
- Preliminary transformation of data were done. H-L scale results transformed into three variables:
  - HLCut: [MEAN/HL] Cutoff (mean Last A option, First B). If all A, last B = 10.5, if all B, last A = 0.5
  - HLCutPon: [MEAN/PONDERED/HL] Cutoff with ponders - not only means of rows, but also means of values pondered by location. Essentially ((ValueA1*position + ValueA2*position +... / nA) + (ValueB1*position + ValueB2*position +... / nB)) / 2
  - NonLinearC [nr of oscilations between options] Is this a non-monotonic, non-linear case? 0 - keeps to one option, 1 - no (common) 2,3,4 - yes.
    0 "NO. Linear A or B only (monotonic)"
    1 "NO. Linear between A or B or missing (in that case monotonic)" 
    2 "YES. Non-monotonic (2 oscilations between A and B and missing)"
    3 "YES. Non-monotonic (3 oscilations between A and B and missing)"
    4 "YES. Non-monotonic (4 oscilations between A and B and missing)"
    5 "YES. Non-monotonic (5 oscilations between A and B and missing)"
    6 "YES. Non-monotonic (6 oscilations between A and B and missing)"
    7 "YES. Non-monotonic (7 oscilations between A and B and missing)"
    8 "YES. Non-monotonic (8 oscilations between A and B and missing)"
    9 "YES. Non-monotonic (9 oscilations between A and B and missing)"
 
----------------------------------------------------------
EFA  ----------------------------------------------
----------------------------------------------------------  
 
FACTOR
  /VARIABLES PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08 PT_PTMR01 PT_PTMR02 PT_PTMR03 PT_PTMR04 BIS01_PTIS01 BIS02_PTIS02 BIS03_PTIS03
  /MISSING LISTWISE 
  /ANALYSIS PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08 PT_PTMR01 PT_PTMR02 PT_PTMR03 PT_PTMR04 BIS01_PTIS01 BIS02_PTIS02 BIS03_PTIS03
  /PRINT INITIAL KMO EXTRACTION ROTATION
  /FORMAT SORT BLANK(.3)
  /PLOT EIGEN
  /CRITERIA FACTORS(2) ITERATE(25)
  /EXTRACTION PAF
  /CRITERIA ITERATE(25) DELTA(0)
  /ROTATION OBLIMIN
  /METHOD=CORRELATION.
 
 
RELIABILITY
  /VARIABLES=DSPRT_01F DSPRT_02F DSPRT_03F DSPRT_04F DSPRT_05F DSPRT_06F
  /SCALE('DOSPERT-R') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
  /SUMMARY=TOTAL MEANS.
 
 
Certainty Effect
  RELIABILITY
    /VARIABLES=PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08
    /SCALE('Cert Effect') ALL
    /MODEL=ALPHA
    /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
    /SUMMARY=TOTAL MEANS.
 
Mirror Effect
  RELIABILITY
  /VARIABLES=PT_PTMR01 PT_PTMR02 PT_PTMR03 PT_PTMR04
  /SCALE('Mirror Effect') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
  /SUMMARY=TOTAL MEANS.
 
Isolation Effect
 
RELIABILITY
  /VARIABLES=BIS01_PTIS01 BIS02_PTIS02 BIS03_PTIS03
  /SCALE('ISOL Effect') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
  /SUMMARY=TOTAL MEANS.
 
 
RELIABILITY
  /VARIABLES=BIS01_PTIS01 BIS02_PTIS02
  /SCALE('ISOL Effect') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
  /SUMMARY=TOTAL MEANS.
 
  What about the reliability of the whole PT (with ISOL03 removed)?
 
  RELIABILITY
  /VARIABLES=BIS01_PTIS01 BIS02_PTIS02 BIS03_PTIS03 PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 
    PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08 PT_PTMR01 PT_PTMR02 PT_PTMR03 PT_PTMR04
  /SCALE('PT Full') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
  /SUMMARY=TOTAL MEANS.
 
 
FREQUENCIES VARIABLES=PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 
    PT_PTCR08 BIS01_PTIS01 BIS02_PTIS02
  /STATISTICS=STDDEV SEMEAN MEAN SKEWNESS SESKEW KURTOSIS SEKURT
  /HISTOGRAM NORMAL
  /ORDER=ANALYSIS.
 
  ... and they are skewed. So PAF it is.
 
  FACTOR
  /VARIABLES PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08 
    BIS01_PTIS01 BIS02_PTIS02
  /MISSING LISTWISE 
  /ANALYSIS PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08 
    BIS01_PTIS01 BIS02_PTIS02
  /PRINT INITIAL CORRELATION DET KMO EXTRACTION ROTATION
  /FORMAT BLANK(.30)
  /PLOT EIGEN
  /CRITERIA MINEIGEN(1) ITERATE(35)
  /EXTRACTION PAF
  /CRITERIA ITERATE(35) DELTA(0)
  /ROTATION OBLIMIN
  /METHOD=CORRELATION.
 
 
----
Creating tables: 
RELIABILITY
  /VARIABLES=DSPRT_01F DSPRT_02F DSPRT_03F DSPRT_04F DSPRT_05F DSPRT_06F
  /SCALE('DOSPERT-R') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
  /SUMMARY=TOTAL MEANS.
 
RELIABILITY
  /VARIABLES=DSPRT_01F DSPRT_03F DSPRT_05F
  /SCALE('DOSPERT-R - Betting') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
  /SUMMARY=TOTAL MEANS.
 
RELIABILITY
  /VARIABLES=DSPRT_02F DSPRT_04F DSPRT_06F
  /SCALE('DOSPERT-R - Investing') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
  /SUMMARY=TOTAL MEANS.
 
  RELIABILITY
  /VARIABLES=BIS01_PTIS01 BIS02_PTIS02 BIS03_PTIS03 PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 
    PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08 PT_PTMR01 PT_PTMR02 PT_PTMR03 PT_PTMR04
  /SCALE('PT Full') ALL
  /MODEL=ALPHA
  /STATISTICS=DESCRIPTIVE SCALE CORR ANOVA
  /SUMMARY=TOTAL MEANS.
 
FACTOR
  /VARIABLES PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08 
    BIS01_PTIS01 BIS02_PTIS02
  /MISSING LISTWISE 
  /ANALYSIS PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 PT_PTCR08 
    BIS01_PTIS01 BIS02_PTIS02
  /PRINT INITIAL KMO EXTRACTION ROTATION
  /FORMAT SORT BLANK(.1)
  /PLOT EIGEN
  /CRITERIA FACTORS(1) ITERATE(25)
  /EXTRACTION PAF
  /CRITERIA ITERATE(25) DELTA(0)
  /ROTATION OBLIMIN
  /METHOD=CORRELATION.
 
FREQUENCIES VARIABLES=PT_PTCR01 PT_PTCR02 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07 
    PT_PTCR08 BIS01_PTIS01 BIS02_PTIS02
  /STATISTICS=STDDEV MEAN SKEWNESS SESKEW KURTOSIS SEKURT
  /HISTOGRAM NORMAL
  /ORDER=ANALYSIS.
 
FACTOR
  /VARIABLES PT_PTCR01 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07
    BIS01_PTIS01 BIS02_PTIS02
  /MISSING LISTWISE
  /ANALYSIS PT_PTCR01 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07
    BIS01_PTIS01 BIS02_PTIS02
  /PRINT INITIAL KMO EXTRACTION ROTATION
  /FORMAT SORT BLANK(.1)
  /PLOT EIGEN
  /CRITERIA FACTORS(1) ITERATE(25)
  /EXTRACTION PAF
  /CRITERIA ITERATE(25) DELTA(0)
  /ROTATION OBLIMIN
  /METHOD=CORRELATION.
 
FACTOR
  /VARIABLES PT_PTCR01 PT_PTCR03 PT_PTCR06 PT_PTCR07 BIS01_PTIS01 BIS02_PTIS02
  /MISSING LISTWISE
  /ANALYSIS PT_PTCR01 PT_PTCR03 PT_PTCR06 PT_PTCR07 BIS01_PTIS01 BIS02_PTIS02
  /PRINT INITIAL KMO EXTRACTION ROTATION
  /FORMAT SORT BLANK(.1)
  /PLOT EIGEN
  /CRITERIA FACTORS(1) ITERATE(25)
  /EXTRACTION PAF
  /CRITERIA ITERATE(25) DELTA(0)
  /ROTATION OBLIMIN
  /METHOD=CORRELATION.
 
RELIABILITY
  /VARIABLES=PT_PTCR01 PT_PTCR03 PT_PTCR04 PT_PTCR05 PT_PTCR06 PT_PTCR07
    BIS01_PTIS01 BIS02_PTIS02
  /SCALE('KT after EFA') ALL
  /MODEL=ALPHA
  /STATISTICS=SCALE
  /SUMMARY=TOTAL.
 
RELIABILITY
  /VARIABLES=PT_PTCR01 PT_PTCR03 PT_PTCR06 PT_PTCR07 BIS01_PTIS01 BIS02_PTIS02
  /SCALE('KT after EFA') ALL
  /MODEL=ALPHA
  /STATISTICS=SCALE
  /SUMMARY=TOTAL.