[Limdep Nlogit List] Interpretation of the alternative specific constants (alt1, alt2)

Brett Smith brett.smith at uwa.edu.au
Thu Nov 24 18:24:40 AEDT 2022


Harold 

My guess is that your variable None=1 and that has led to the model being unidentified. Or you are choices are actually labelled and caged and acdoors  are your alternate specific constants. 

One final comment the code has the conditional means of caged(n) ,acdoors(n), 
As a function of male and female? Unidentified
All age groups? Unidentified
All incomes? Unidentified.

Include male only
Convert your age and incomes to a scale (i.e. use the mid points of the brackets)

All the best,

Brett

-----Original Message-----
From: Limdep <limdep-bounces at mailman.sydney.edu.au> On Behalf Of Harold Mayaba via Limdep
Sent: Thursday, 24 November 2022 1:53 PM
To: limdep at mailman.sydney.edu.au
Cc: Harold Mayaba <mayabach2009 at yahoo.com>
Subject: [Limdep Nlogit List] Interpretation of the alternative specific constants (alt1, alt2)

Dear Professor Greene and Nlogit users,
Please help me understand how to interpret the results for alt1, alt2, and none in the results below. Alt1 and alt2 are used as alternative-specific constants. Also, how to interpret the coefficients for  Distns. of RPs. Std.Devs in a model. Thank you for your continued help. 


Regards
Harold. 


NLOGIT

; lhs = choice

; choices = alt1, alt2, none

;rpl = female, male, age1, age2,age3, age4, age5,inco1, inco2, inco3,inco4, inco5

;fcn= caged(n) ,acdoors(n), nprs(n|#000000000000),cert1(n|#000000000000),cert2(n|#000000000000),mmort(n|#000000000000),hmort(n|#000000000000)

;pds =10

; pts=100

;maxit=500

;describe

;pwt

;effects:caged(*)/acdoors(*)/price(*)

 ;full

;parameters

; halton

;effects: caged (*) / acdoors(*)/price(*)

; wtp=caged/price,acdoors/price,nprs/price ,cert2/price

,hmort/price

; model:

U(alt1) = alt1 +caged*caged+acdoors*acdoors+nprs*nprs+cert1*cert1+cert2*cert2+mmort*mmort+hmort*hmort +price*price/

U(alt2) = alt2 +caged*caged+acdoors*acdoors+nprs*nprs+cert1*cert1+cert2*cert2+mmort*mmort+hmort*hmort  +price*price/
U(None) = None$

--------+--------------------------------------------------------------------
        |                  Standard            Prob.      95% Confidence

  CHOICE|  Coefficient       Error       z   |z|>Z*         Interval

--------+--------------------------------------------------------------------

        |Randomparameters in utility functions..............................

   CAGED|   -3.23814***      .49400   -6.55  .0000    -4.20636 -2.26993

 ACDOORS|    4.76545***      .56989    8.36  .0000     3.64849  5.88242

    NPRS|    -.63055**       .27633   -2.28  .0225    -1.17214  -.08896

   CERT1|     .98566***      .26379    3.74  .0002      .46864  1.50268

   CERT2|     .18598         .19201      .97 .3328     -.19036    .56232

   MMORT|    -.41961***      .14551   -2.88  .0039     -.70480  -.13442

   HMORT|   -1.01844***      .18335   -5.55  .0000    -1.37780  -.65909

        |Nonrandomparameters in utility functions...........................

    ALT1|     .45298     .8389D+07      .00 1.0000***********  ***********

   PRICE|    -.20525***      .01812  -11.32  .0000     -.24077  -.16972

    ALT2|     .37546     .8389D+07      .00 1.0000***********  ***********

    NONE|    -.82843     .8389D+07      .00 1.0000***********  ***********

 

        |Distns. ofRPs. Std.Devs or limits of triangular....................

 NsCAGED|    2.33247***      .17476   13.35  .0000     1.98995  2.67500

NsACDOOR|   4.41155***      .32553    13.55 .0000     3.77352  5.04958

  NsNPRS|    2.00750***      .14234   14.10  .0000     1.72852  2.28648

 NsCERT1|     .12418         .27668      .45 .6536     -.41809    .66646

 NsCERT2|     .58488***      .11404    5.13  .0000      .36137   .80839

 NsMMORT|     .35905         .22849     1.57 .1161     -.08878    .80688

 NsHMORT|     .98402***      .19336    5.09  .0000      .60504  1.36301

_______________________________________________
Limdep site list
Limdep at mailman.sydney.edu.au
https://protect-au.mimecast.com/s/2lzvC2xMQzik3JmPxtnjbuL?domain=aus01.safelinks.protection.outlook.com



More information about the Limdep mailing list