Possible model specification error for labelled design
Posted: 08 Mar 2022, 15:36
Hello Professors Hess and Palma,
Hope this finds you doing well. I am working on an analysis for a labelled discrete choice experiment, and I am brand new to this type of experiment and analysis. I have been using support.CEs package to set up my data and apollo for analysis. The study is trying to determine preferences between performing a healthcare service in-house or refer it out to a contractor based on several attributes (time, cost, patient benefit, etc.). We also divided the design into 3 blocks of 6 scenarios each which were randomly assigned to respondents. I am not sure what all you need to assist in troubleshooting, but I will provide as much as possible. After formatting the data using support.CEs, it looks like this:
ID BLOCK QES STR RES.1 ASC.1 X4.hours1.1 X8.hours1.1 X.10.caregiver1.1 X.20.caregiver1.1 X10..decrease1.1 X20..decrease1.1
1 1 1 1 101 FALSE 1 0 1 0 0 1 0
3 1 1 2 102 TRUE 1 1 0 0 1 0 0
5 1 1 3 103 FALSE 1 0 1 1 0 0 0
7 1 1 4 104 FALSE 1 0 0 0 1 0 1
9 1 1 5 105 TRUE 1 1 0 0 0 0 1
11 1 1 6 106 TRUE 1 0 1 0 1 0 1
X5..lower1.1 X10..lower1.1 X5..increase1.1 X10..increase1.1 X4.hours2.1 X8.hours2.1 X.10.caregiver2.1 X.20.caregiver2.1
1 0 1 0 0 0 0 0 0
3 1 0 0 1 0 0 0 0
5 1 0 0 0 0 0 0 0
7 1 0 0 0 0 0 0 0
9 0 0 0 1 0 0 0 0
11 0 1 0 1 0 0 0 0
X10..decrease2.1 X20..decrease2.1 X5..lower2.1 X10..lower2.1 X5..increase2.1 X10..increase2.1 RES.2 ASC.2 X4.hours1.2
1 0 0 0 0 0 0 TRUE 0 0
3 0 0 0 0 0 0 FALSE 0 0
5 0 0 0 0 0 0 TRUE 0 0
7 0 0 0 0 0 0 TRUE 0 0
9 0 0 0 0 0 0 FALSE 0 0
11 0 0 0 0 0 0 FALSE 0 0
X8.hours1.2 X.10.caregiver1.2 X.20.caregiver1.2 X10..decrease1.2 X20..decrease1.2 X5..lower1.2 X10..lower1.2 X5..increase1.2
1 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
5 0 0 0 0 0 0 0 0
7 0 0 0 0 0 0 0 0
9 0 0 0 0 0 0 0 0
11 0 0 0 0 0 0 0 0
X10..increase1.2 X4.hours2.2 X8.hours2.2 X.10.caregiver2.2 X.20.caregiver2.2 X10..decrease2.2 X20..decrease2.2 X5..lower2.2
1 0 0 0 1 0 0 1 0
3 0 0 1 0 0 1 0 0
5 0 0 0 0 1 1 0 0
7 0 1 0 0 0 0 0 0
9 0 0 1 1 0 0 0 1
11 0 0 0 0 0 0 0 0
X10..lower2.2 X5..increase2.2 X10..increase2.2 choice
1 0 1 0 2
3 1 0 0 1
5 1 1 0 2
7 1 1 0 2
9 0 0 0 1
11 0 0 0 1
The following is my apollo code and output. I am brand new to apollo as well so forgive any errors:
apollo_initialise()
apollo_control <- list(
modelName = "CL_apollo",
modelDescr = "DCE for Facebook Support Groups",
indivID = "ID"
)
database <- DCE.data.full.r
apollo_beta <- c(
asc1 = 0,
asc2 = 0,
b_4hours1 = 0,
b_8hours1 = 0,
b_10cg1 = 0,
b_20cg1 = 0,
b_10dec1 = 0,
b_20dec1 = 0,
b_5low1 = 0,
b_10low1 = 0,
b_5inc1 = 0,
b_10inc1 = 0,
b_4hours2 = 0,
b_8hours2 = 0,
b_10cg2 = 0,
b_20cg2 = 0,
b_10dec2 = 0,
b_20dec2 = 0,
b_5low2 = 0,
b_10low2 = 0,
b_5inc2 = 0,
b_10inc2 = 0
)
apollo_fixed <- c()
apollo_inputs <- apollo_validateInputs()
## Several observations per individual detected based on the value of ID.
## Setting panelData in apollo_control set to TRUE.
## All checks on apollo_control completed.
## All checks on database completed.
apollo_probabilities <- function(apollo_beta, apollo_inputs, functionality = "estimate"){
apollo_attach(apollo_beta, apollo_inputs)
on.exit(apollo_detach(apollo_beta, apollo_inputs))
P <- list()
V <- list()
V[['alt1']] = asc1 + b_4hours1*X4.hours1.1 + b_8hours1*X8.hours1.1 +
b_10cg1*X.10.caregiver1.1 +
b_20cg1*X.20.caregiver1.1 + b_10dec1*X10..decrease1.1 + b_20dec1*X20..decrease1.1 +
b_5low1*X5..lower1.1 + b_10low1*X10..lower1.1 + b_5inc1*X5..increase1.1 +
b_10inc1*X10..increase1.1
V[['alt2']] = asc2 +
b_4hours2*X4.hours2.2 + b_8hours2*X8.hours2.2 +
b_10cg2*X.10.caregiver2.2 + b_20cg2*X.20.caregiver2.2 + b_10dec2*X10..decrease2.2 +
b_20dec2*X20..decrease2.2 + b_5low2*X5..lower2.2 + b_10low2*X10..lower2.2 +
b_5inc2*X5..increase2.2 + b_10inc2*X10..increase2.2
mnl_settings <- list(
alternatives = c(alt1 = 1, alt2 = 2),
avail = list(alt1 = 1, alt2 = 1),
choiceVar = choice,
V = V
)
P[['model']] <- apollo_mnl(mnl_settings, functionality)
P <- apollo_panelProd(P, apollo_inputs, functionality)
P <- apollo_prepareProb(P, apollo_inputs, functionality)
return(P)
}
model <- apollo_estimate(apollo_beta, apollo_fixed, apollo_probabilities, apollo_inputs)
##
## Testing likelihood function...
##
## Overview of choices for MNL model component :
## alt1 alt2
## Times available 1038.00 1038.00
## Times chosen 434.00 604.00
## Percentage chosen overall 41.81 58.19
## Percentage chosen when available 41.81 58.19
##
## Pre-processing likelihood function...
##
## Testing influence of parameters......................
## Starting main estimation
## Initial function value: -719.4868
## Initial gradient value:
## asc1 asc2 b_4hours1 b_8hours1 b_10cg1 b_20cg1 b_10dec1 b_20dec1
## -85.0 85.0 -14.0 -44.5 -28.5 -39.5 -85.5 85.5
## b_5low1 b_10low1 b_5inc1 b_10inc1 b_4hours2 b_8hours2 b_10cg2 b_20cg2
## -27.5 -10.0 -30.0 -10.5 26.5 14.0 17.0 28.5
## b_10dec2 b_20dec2 b_5low2 b_10low2 b_5inc2 b_10inc2
## 85.0 85.5 47.5 27.5 44.5 30.0
## initial value 719.486773
## iter 2 value 676.151503
## iter 3 value 581.490273
## iter 4 value 575.289823
## iter 5 value 575.095376
## iter 6 value 574.851886
## iter 7 value 573.647232
## iter 8 value 573.303710
## iter 9 value 573.253082
## iter 10 value 572.958385
## iter 11 value 571.600535
## iter 12 value 571.530286
## iter 13 value 571.522960
## iter 14 value 571.435950
## iter 15 value 571.432525
## iter 16 value 571.428038
## iter 16 value 571.428037
## iter 16 value 571.428037
## final value 571.428037
## converged
## Additional convergence test using scaled estimation. Parameters will be
## scaled by their current estimates and additional iterations will be
## performed.
## initial value 571.428037
## iter 1 value 571.428037
## final value 571.428037
## converged
## Estimated parameters:
## Estimate
## asc1 -0.072715
## asc2 0.072715
## b_4hours1 0.119235
## b_8hours1 -0.244042
## b_10cg1 -0.002335
## b_20cg1 -0.231423
## b_10dec1 -0.453392
## b_20dec1 1.316599
## b_5low1 0.004069
## b_10low1 0.280412
## b_5inc1 -0.042371
## b_10inc1 0.235909
## b_4hours2 -0.052092
## b_8hours2 -0.119235
## b_10cg2 -0.161043
## b_20cg2 0.002335
## b_10dec2 0.935929
## b_20dec2 0.453392
## b_5low2 0.357196
## b_10low2 -0.004069
## b_5inc2 0.266253
## b_10inc2 0.042371
##
## Computing covariance matrix using analytical gradient.
## 0%....25%....50%....75%....100%
## WARNING: Some eigenvalues of the Hessian are complex, indicating that
## the Hessian is not symmetrical.
## Warning in sqrt(diag(varcov)): NaNs produced
## Computing score matrix...
## Warning in sqrt(diag(robvarcov)): NaNs produced
## Calculating LL(0) for applicable models...
## Calculating LL(c) for applicable models...
## Calculating LL of each model component...
apollo_modelOutput(model, list(printPVal = TRUE))
## Model run using Apollo for R, version 0.2.6 on Darwin
## www.ApolloChoiceModelling.com
##
## Model name : CL_apollo
## Model description : DCE for Facebook Support Groups
## Model run at : 2022-03-08 09:05:56
## Estimation method : bfgs
## Model diagnosis : successful convergence
## Number of individuals : 173
## Number of rows in database : 1038
## Number of modelled outcomes : 1038
##
## Number of cores used : 1
## Model without mixing
##
## LL(start) : -719.4868
## LL(0) : -719.4868
## LL(C) : -705.5029
## LL(final) : -571.428
## Rho-square (0) : 0.2058
## Adj.Rho-square (0) : 0.1752
## Rho-square (C) : 0.19
## Adj.Rho-square (C) : 0.1589
## AIC : 1186.86
## BIC : 1295.65
##
## Estimated parameters : 22
## Time taken (hh:mm:ss) : 00:00:1.75
## pre-estimation : 00:00:0.61
## estimation : 00:00:0.28
## post-estimation : 00:00:0.86
## Iterations : 20
##
## Unconstrained optimisation.
##
## Estimates:
## Estimate s.e. t.rat.(0) p(1-sided) Rob.s.e.
## asc1 -0.072715 NaN NaN NaN 150.385
## asc2 0.072715 2931. 2.481e-05 0.5000 20.213
## b_4hours1 0.119235 NaN NaN NaN NaN
## b_8hours1 -0.244042 NaN NaN NaN NaN
## b_10cg1 -0.002335 1589. -1.469e-06 0.5000 178.938
## b_20cg1 -0.231423 1845. -1.2546e-04 0.4999 8.663
## b_10dec1 -0.453392 NaN NaN NaN 242.463
## b_20dec1 1.316599 NaN NaN NaN 26.777
## b_5low1 0.004069 NaN NaN NaN 54.160
## b_10low1 0.280412 NaN NaN NaN 103.734
## b_5inc1 -0.042371 2.609e+04 -1.624e-06 0.5000 NaN
## b_10inc1 0.235909 7424. 3.178e-05 0.5000 42.686
## b_4hours2 -0.052092 NaN NaN NaN NaN
## b_8hours2 -0.119235 NaN NaN NaN NaN
## b_10cg2 -0.161043 1845. -8.730e-05 0.5000 NaN
## b_20cg2 0.002335 NaN NaN NaN 134.929
## b_10dec2 0.935929 NaN NaN NaN NaN
## b_20dec2 0.453392 NaN NaN NaN 280.243
## b_5low2 0.357196 NaN NaN NaN 100.963
## b_10low2 -0.004069 1513. -2.690e-06 0.5000 31.999
## b_5inc2 0.266253 7424. 3.586e-05 0.5000 54.739
## b_10inc2 0.042371 2.965e+04 1.429e-06 0.5000 NaN
## Rob.t.rat.(0) p(1-sided)
## asc1 -4.8352e-04 0.4998
## asc2 0.003598 0.4986
## b_4hours1 NaN NaN
## b_8hours1 NaN NaN
## b_10cg1 -1.305e-05 0.5000
## b_20cg1 -0.026715 0.4893
## b_10dec1 -0.001870 0.4993
## b_20dec1 0.049169 0.4804
## b_5low1 7.514e-05 0.5000
## b_10low1 0.002703 0.4989
## b_5inc1 NaN NaN
## b_10inc1 0.005527 0.4978
## b_4hours2 NaN NaN
## b_8hours2 NaN NaN
## b_10cg2 NaN NaN
## b_20cg2 1.730e-05 0.5000
## b_10dec2 NaN NaN
## b_20dec2 0.001618 0.4994
## b_5low2 0.003538 0.4986
## b_10low2 -1.2717e-04 0.4999
## b_5inc2 0.004864 0.4981
## b_10inc2 NaN NaN
Perhaps it is correct, but I feel as though I have an error given the NA estimates in the output, which look like they come from the Hessian matrix not being symmetrical. I've tried a lot of specification trial and error, and I believe specification is my issue. I had trouble finding details about coding the specification for a labelled design, but if it is readily available in the manual or elsewhere, I'm happy to check if you can direct me, and I apologize for missing it. Otherwise, I was wondering whether you could help me troubleshoot or at least rule out specification as my issue. Thank you so much for all the resources you provide and this forum, and I look forward to hearing from you.
Kyle
Hope this finds you doing well. I am working on an analysis for a labelled discrete choice experiment, and I am brand new to this type of experiment and analysis. I have been using support.CEs package to set up my data and apollo for analysis. The study is trying to determine preferences between performing a healthcare service in-house or refer it out to a contractor based on several attributes (time, cost, patient benefit, etc.). We also divided the design into 3 blocks of 6 scenarios each which were randomly assigned to respondents. I am not sure what all you need to assist in troubleshooting, but I will provide as much as possible. After formatting the data using support.CEs, it looks like this:
ID BLOCK QES STR RES.1 ASC.1 X4.hours1.1 X8.hours1.1 X.10.caregiver1.1 X.20.caregiver1.1 X10..decrease1.1 X20..decrease1.1
1 1 1 1 101 FALSE 1 0 1 0 0 1 0
3 1 1 2 102 TRUE 1 1 0 0 1 0 0
5 1 1 3 103 FALSE 1 0 1 1 0 0 0
7 1 1 4 104 FALSE 1 0 0 0 1 0 1
9 1 1 5 105 TRUE 1 1 0 0 0 0 1
11 1 1 6 106 TRUE 1 0 1 0 1 0 1
X5..lower1.1 X10..lower1.1 X5..increase1.1 X10..increase1.1 X4.hours2.1 X8.hours2.1 X.10.caregiver2.1 X.20.caregiver2.1
1 0 1 0 0 0 0 0 0
3 1 0 0 1 0 0 0 0
5 1 0 0 0 0 0 0 0
7 1 0 0 0 0 0 0 0
9 0 0 0 1 0 0 0 0
11 0 1 0 1 0 0 0 0
X10..decrease2.1 X20..decrease2.1 X5..lower2.1 X10..lower2.1 X5..increase2.1 X10..increase2.1 RES.2 ASC.2 X4.hours1.2
1 0 0 0 0 0 0 TRUE 0 0
3 0 0 0 0 0 0 FALSE 0 0
5 0 0 0 0 0 0 TRUE 0 0
7 0 0 0 0 0 0 TRUE 0 0
9 0 0 0 0 0 0 FALSE 0 0
11 0 0 0 0 0 0 FALSE 0 0
X8.hours1.2 X.10.caregiver1.2 X.20.caregiver1.2 X10..decrease1.2 X20..decrease1.2 X5..lower1.2 X10..lower1.2 X5..increase1.2
1 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0
5 0 0 0 0 0 0 0 0
7 0 0 0 0 0 0 0 0
9 0 0 0 0 0 0 0 0
11 0 0 0 0 0 0 0 0
X10..increase1.2 X4.hours2.2 X8.hours2.2 X.10.caregiver2.2 X.20.caregiver2.2 X10..decrease2.2 X20..decrease2.2 X5..lower2.2
1 0 0 0 1 0 0 1 0
3 0 0 1 0 0 1 0 0
5 0 0 0 0 1 1 0 0
7 0 1 0 0 0 0 0 0
9 0 0 1 1 0 0 0 1
11 0 0 0 0 0 0 0 0
X10..lower2.2 X5..increase2.2 X10..increase2.2 choice
1 0 1 0 2
3 1 0 0 1
5 1 1 0 2
7 1 1 0 2
9 0 0 0 1
11 0 0 0 1
The following is my apollo code and output. I am brand new to apollo as well so forgive any errors:
apollo_initialise()
apollo_control <- list(
modelName = "CL_apollo",
modelDescr = "DCE for Facebook Support Groups",
indivID = "ID"
)
database <- DCE.data.full.r
apollo_beta <- c(
asc1 = 0,
asc2 = 0,
b_4hours1 = 0,
b_8hours1 = 0,
b_10cg1 = 0,
b_20cg1 = 0,
b_10dec1 = 0,
b_20dec1 = 0,
b_5low1 = 0,
b_10low1 = 0,
b_5inc1 = 0,
b_10inc1 = 0,
b_4hours2 = 0,
b_8hours2 = 0,
b_10cg2 = 0,
b_20cg2 = 0,
b_10dec2 = 0,
b_20dec2 = 0,
b_5low2 = 0,
b_10low2 = 0,
b_5inc2 = 0,
b_10inc2 = 0
)
apollo_fixed <- c()
apollo_inputs <- apollo_validateInputs()
## Several observations per individual detected based on the value of ID.
## Setting panelData in apollo_control set to TRUE.
## All checks on apollo_control completed.
## All checks on database completed.
apollo_probabilities <- function(apollo_beta, apollo_inputs, functionality = "estimate"){
apollo_attach(apollo_beta, apollo_inputs)
on.exit(apollo_detach(apollo_beta, apollo_inputs))
P <- list()
V <- list()
V[['alt1']] = asc1 + b_4hours1*X4.hours1.1 + b_8hours1*X8.hours1.1 +
b_10cg1*X.10.caregiver1.1 +
b_20cg1*X.20.caregiver1.1 + b_10dec1*X10..decrease1.1 + b_20dec1*X20..decrease1.1 +
b_5low1*X5..lower1.1 + b_10low1*X10..lower1.1 + b_5inc1*X5..increase1.1 +
b_10inc1*X10..increase1.1
V[['alt2']] = asc2 +
b_4hours2*X4.hours2.2 + b_8hours2*X8.hours2.2 +
b_10cg2*X.10.caregiver2.2 + b_20cg2*X.20.caregiver2.2 + b_10dec2*X10..decrease2.2 +
b_20dec2*X20..decrease2.2 + b_5low2*X5..lower2.2 + b_10low2*X10..lower2.2 +
b_5inc2*X5..increase2.2 + b_10inc2*X10..increase2.2
mnl_settings <- list(
alternatives = c(alt1 = 1, alt2 = 2),
avail = list(alt1 = 1, alt2 = 1),
choiceVar = choice,
V = V
)
P[['model']] <- apollo_mnl(mnl_settings, functionality)
P <- apollo_panelProd(P, apollo_inputs, functionality)
P <- apollo_prepareProb(P, apollo_inputs, functionality)
return(P)
}
model <- apollo_estimate(apollo_beta, apollo_fixed, apollo_probabilities, apollo_inputs)
##
## Testing likelihood function...
##
## Overview of choices for MNL model component :
## alt1 alt2
## Times available 1038.00 1038.00
## Times chosen 434.00 604.00
## Percentage chosen overall 41.81 58.19
## Percentage chosen when available 41.81 58.19
##
## Pre-processing likelihood function...
##
## Testing influence of parameters......................
## Starting main estimation
## Initial function value: -719.4868
## Initial gradient value:
## asc1 asc2 b_4hours1 b_8hours1 b_10cg1 b_20cg1 b_10dec1 b_20dec1
## -85.0 85.0 -14.0 -44.5 -28.5 -39.5 -85.5 85.5
## b_5low1 b_10low1 b_5inc1 b_10inc1 b_4hours2 b_8hours2 b_10cg2 b_20cg2
## -27.5 -10.0 -30.0 -10.5 26.5 14.0 17.0 28.5
## b_10dec2 b_20dec2 b_5low2 b_10low2 b_5inc2 b_10inc2
## 85.0 85.5 47.5 27.5 44.5 30.0
## initial value 719.486773
## iter 2 value 676.151503
## iter 3 value 581.490273
## iter 4 value 575.289823
## iter 5 value 575.095376
## iter 6 value 574.851886
## iter 7 value 573.647232
## iter 8 value 573.303710
## iter 9 value 573.253082
## iter 10 value 572.958385
## iter 11 value 571.600535
## iter 12 value 571.530286
## iter 13 value 571.522960
## iter 14 value 571.435950
## iter 15 value 571.432525
## iter 16 value 571.428038
## iter 16 value 571.428037
## iter 16 value 571.428037
## final value 571.428037
## converged
## Additional convergence test using scaled estimation. Parameters will be
## scaled by their current estimates and additional iterations will be
## performed.
## initial value 571.428037
## iter 1 value 571.428037
## final value 571.428037
## converged
## Estimated parameters:
## Estimate
## asc1 -0.072715
## asc2 0.072715
## b_4hours1 0.119235
## b_8hours1 -0.244042
## b_10cg1 -0.002335
## b_20cg1 -0.231423
## b_10dec1 -0.453392
## b_20dec1 1.316599
## b_5low1 0.004069
## b_10low1 0.280412
## b_5inc1 -0.042371
## b_10inc1 0.235909
## b_4hours2 -0.052092
## b_8hours2 -0.119235
## b_10cg2 -0.161043
## b_20cg2 0.002335
## b_10dec2 0.935929
## b_20dec2 0.453392
## b_5low2 0.357196
## b_10low2 -0.004069
## b_5inc2 0.266253
## b_10inc2 0.042371
##
## Computing covariance matrix using analytical gradient.
## 0%....25%....50%....75%....100%
## WARNING: Some eigenvalues of the Hessian are complex, indicating that
## the Hessian is not symmetrical.
## Warning in sqrt(diag(varcov)): NaNs produced
## Computing score matrix...
## Warning in sqrt(diag(robvarcov)): NaNs produced
## Calculating LL(0) for applicable models...
## Calculating LL(c) for applicable models...
## Calculating LL of each model component...
apollo_modelOutput(model, list(printPVal = TRUE))
## Model run using Apollo for R, version 0.2.6 on Darwin
## www.ApolloChoiceModelling.com
##
## Model name : CL_apollo
## Model description : DCE for Facebook Support Groups
## Model run at : 2022-03-08 09:05:56
## Estimation method : bfgs
## Model diagnosis : successful convergence
## Number of individuals : 173
## Number of rows in database : 1038
## Number of modelled outcomes : 1038
##
## Number of cores used : 1
## Model without mixing
##
## LL(start) : -719.4868
## LL(0) : -719.4868
## LL(C) : -705.5029
## LL(final) : -571.428
## Rho-square (0) : 0.2058
## Adj.Rho-square (0) : 0.1752
## Rho-square (C) : 0.19
## Adj.Rho-square (C) : 0.1589
## AIC : 1186.86
## BIC : 1295.65
##
## Estimated parameters : 22
## Time taken (hh:mm:ss) : 00:00:1.75
## pre-estimation : 00:00:0.61
## estimation : 00:00:0.28
## post-estimation : 00:00:0.86
## Iterations : 20
##
## Unconstrained optimisation.
##
## Estimates:
## Estimate s.e. t.rat.(0) p(1-sided) Rob.s.e.
## asc1 -0.072715 NaN NaN NaN 150.385
## asc2 0.072715 2931. 2.481e-05 0.5000 20.213
## b_4hours1 0.119235 NaN NaN NaN NaN
## b_8hours1 -0.244042 NaN NaN NaN NaN
## b_10cg1 -0.002335 1589. -1.469e-06 0.5000 178.938
## b_20cg1 -0.231423 1845. -1.2546e-04 0.4999 8.663
## b_10dec1 -0.453392 NaN NaN NaN 242.463
## b_20dec1 1.316599 NaN NaN NaN 26.777
## b_5low1 0.004069 NaN NaN NaN 54.160
## b_10low1 0.280412 NaN NaN NaN 103.734
## b_5inc1 -0.042371 2.609e+04 -1.624e-06 0.5000 NaN
## b_10inc1 0.235909 7424. 3.178e-05 0.5000 42.686
## b_4hours2 -0.052092 NaN NaN NaN NaN
## b_8hours2 -0.119235 NaN NaN NaN NaN
## b_10cg2 -0.161043 1845. -8.730e-05 0.5000 NaN
## b_20cg2 0.002335 NaN NaN NaN 134.929
## b_10dec2 0.935929 NaN NaN NaN NaN
## b_20dec2 0.453392 NaN NaN NaN 280.243
## b_5low2 0.357196 NaN NaN NaN 100.963
## b_10low2 -0.004069 1513. -2.690e-06 0.5000 31.999
## b_5inc2 0.266253 7424. 3.586e-05 0.5000 54.739
## b_10inc2 0.042371 2.965e+04 1.429e-06 0.5000 NaN
## Rob.t.rat.(0) p(1-sided)
## asc1 -4.8352e-04 0.4998
## asc2 0.003598 0.4986
## b_4hours1 NaN NaN
## b_8hours1 NaN NaN
## b_10cg1 -1.305e-05 0.5000
## b_20cg1 -0.026715 0.4893
## b_10dec1 -0.001870 0.4993
## b_20dec1 0.049169 0.4804
## b_5low1 7.514e-05 0.5000
## b_10low1 0.002703 0.4989
## b_5inc1 NaN NaN
## b_10inc1 0.005527 0.4978
## b_4hours2 NaN NaN
## b_8hours2 NaN NaN
## b_10cg2 NaN NaN
## b_20cg2 1.730e-05 0.5000
## b_10dec2 NaN NaN
## b_20dec2 0.001618 0.4994
## b_5low2 0.003538 0.4986
## b_10low2 -1.2717e-04 0.4999
## b_5inc2 0.004864 0.4981
## b_10inc2 NaN NaN
Perhaps it is correct, but I feel as though I have an error given the NA estimates in the output, which look like they come from the Hessian matrix not being symmetrical. I've tried a lot of specification trial and error, and I believe specification is my issue. I had trouble finding details about coding the specification for a labelled design, but if it is readily available in the manual or elsewhere, I'm happy to check if you can direct me, and I apologize for missing it. Otherwise, I was wondering whether you could help me troubleshoot or at least rule out specification as my issue. Thank you so much for all the resources you provide and this forum, and I look forward to hearing from you.
Kyle