R help - Jan 2018 - MCMC Estimation for Four Parametric Logistic (4PL) Item Response Model

Good day Sir/Ma'am! This is Alyssa Fatmah S. Mastura taking up Master of
Science in Statistics at Mindanao State University-Iligan Institute
Technology (MSU-IIT), Philippines. I am currently working on my master's
thesis titled "Comparing the Three Estimation Methods for the Four
Parametric Logistic (4PL) Item Response Model". While I am looking for a
package about Markov chain Monte Carlo (MCMC) method for 4PL model in this
R forum, I found an sirt package of an MCMC method but only for the three
parametric normal ogive (3PNO) item response model. However, my study focus
on 4PL model. In line with this, I would like to know if there exist a
function of MCMC method for 4PL model in R language. I am asking for your
help to inform me if such function exist. I highly appreciate your response
on this matter. Thank you so much. Have a great day ahead!



Alyssa

<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
Virus-free.
www.avast.com
<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

-- 
---
*DISCLAIMER AND CONFIDENTIALITY NOTICE* The Mindanao Sta...{{dropped:24}}

I know of no existing functions for estimating the parameters of this model
using MCMC or MML. Many years ago, I wrote code to estimate this model using
marginal maximum likelihood. I wrote this based on the using nlminb and
gauss-hermite quadrature points from statmod.

I could not find that code to share with you, but I do have code for estimating
the 3PL in this way and you could modify the likelihood for the upper asymptote
yourself.



-----Original Message-----
From: R-help [mailto:r-help-bounces at r-project.org] On Behalf Of ALYSSA FATMAH
MASTURA
Sent: Thursday, January 18, 2018 10:15 AM
To: r-help at r-project.org
Subject: [R] MCMC Estimation for Four Parametric Logistic (4PL) Item Response
Model

Good day Sir/Ma'am! This is Alyssa Fatmah S. Mastura taking up Master of
Science in Statistics at Mindanao State University-Iligan Institute Technology
(MSU-IIT), Philippines. I am currently working on my master's thesis titled
"Comparing the Three Estimation Methods for the Four Parametric Logistic
(4PL) Item Response Model". While I am looking for a package about Markov
chain Monte Carlo (MCMC) method for 4PL model in this R forum, I found an sirt
package of an MCMC method but only for the three parametric normal ogive (3PNO)
item response model. However, my study focus on 4PL model. In line with this, I
would like to know if there exist a function of MCMC method for 4PL model in R
language. I am asking for your help to inform me if such function exist. I
highly appreciate your response on this matter. Thank you so much. Have a great
day ahead!



Alyssa

<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
Virus-free.
www.avast.com
<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

--
---
*DISCLAIMER AND CONFIDENTIALITY NOTICE* The Mindanao Sta...{{dropped:9}}

If you have the expression of the model, package nlsr should be able to
form the Jacobian analytically for nonlinear least squares. Likelihood
approaches allow for more sophisticated loss functions, but the
optimization is generally much less reliable because one is working
with essentially squared quantities and possibly multiple minima where
some are not the ones you want.

JN

On 2018-01-18 12:32 PM, Doran, Harold wrote:
> I know of no existing functions for estimating the parameters of this model
using MCMC or MML. Many years ago, I wrote code to estimate this model using
marginal maximum likelihood. I wrote this based on the using nlminb and
gauss-hermite quadrature points from statmod.
> 
> I could not find that code to share with you, but I do have code for
estimating the 3PL in this way and you could modify the likelihood for the upper
asymptote yourself.
> 
> 
> 
> -----Original Message-----
> From: R-help [mailto:r-help-bounces at r-project.org] On Behalf Of ALYSSA
FATMAH MASTURA
> Sent: Thursday, January 18, 2018 10:15 AM
> To: r-help at r-project.org
> Subject: [R] MCMC Estimation for Four Parametric Logistic (4PL) Item
Response Model
> 
> Good day Sir/Ma'am! This is Alyssa Fatmah S. Mastura taking up Master
of Science in Statistics at Mindanao State University-Iligan Institute
Technology (MSU-IIT), Philippines. I am currently working on my master's
thesis titled "Comparing the Three Estimation Methods for the Four
Parametric Logistic (4PL) Item Response Model". While I am looking for a
package about Markov chain Monte Carlo (MCMC) method for 4PL model in this R
forum, I found an sirt package of an MCMC method but only for the three
parametric normal ogive (3PNO) item response model. However, my study focus on
4PL model. In line with this, I would like to know if there exist a function of
MCMC method for 4PL model in R language. I am asking for your help to inform me
if such function exist. I highly appreciate your response on this matter. Thank
you so much. Have a great day ahead!
> 
> 
> 
> Alyssa
> 
>
<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
> Virus-free.
> www.avast.com
>
<https://www.avast.com/sig-email?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=webmail>
> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
> 
> --
> ---
> *DISCLAIMER AND CONFIDENTIALITY NOTICE* The Mindanao Sta...{{dropped:9}}
> 
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>