Hi Xinyang, Mattis, and Kristina,
Item Response Theory (IRT) was used to score the HCP Toolbox data, but this was
done completely internally by the NIH Toolbox processing team and we don't have
any other details beyond what is described in the NIH Toolbox Scoring and
Interpretation Guide from August 2012 that we provided you. From what that
Guide says, I believe the IRT statistical model was constructed from the NIH
Toolbox norming sample, but we do not know the details about the type of IRT
model.
You can ask the NIH Toolbox helpdesk for more clarification by writing them. We
correspond with Odessa Castro at Toolbox through the
h...@nihtoolbox.org<mailto:h...@nihtoolbox.org> email address.
Best,
Jenn
Jennifer Elam, Ph.D.
Scientific Outreach, Human Connectome Project
Washington University School of Medicine
Department of Neuroscience, Box 8108
660 South Euclid Avenue
St. Louis, MO 63110
314-362-9387
e...@wustl.edu<mailto:e...@wustl.edu>
www.humanconnectome.org<http://www.humanconnectome.org/>
From: hcp-users-boun...@humanconnectome.org
on behalf of Xinyang Liu
Sent: Thursday, February 21, 2019 7:19:24 AM
To: HCP 讨论组
Subject: [HCP-Users] A few questions about HCP behavioral data
Dear HCP experts,
We're currently analyzing some of the behavioral data from your dataset and
have three short questions about the IRT-modelling and theta-score derivation
that is also mentioned in the "NIH Toolbox Scoring and Interpretation Guide".
To be able to build comparable scores between NIH-task data and Penn- and
Working Memory data we would like to know more about the NIH IRT-scores. We
hope you can help us out.
1. In the "NIH Toolbox Scoring and Interpretation Guide" it is mentioned
that IRT-scores were calculated before the consecutive scores, such as
Age-adjusted or Unadjusted. Was that also true for the HCP data?
2. What was the database to estimate the IRT model for the HCP data? Was the
model estimated based on the HCP data only or was it somehow combined with the
norming sample of the NIH Toolbox?
3. What type of IRT model was run? A simple Rasch model or did you add other
free parameters, such as guessing probability (2PL)?
Thank you very much for your help.
Kind regards,
Xinyang
___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users
___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users