Using NIH Cognitive data

[Hannah Alves Hagy]

Hello -

I am wanting to use the raw data from the NIH Toolbox Cognitive battery, specifically the following tests:

1. Pattern Comparison Processing Speed
2. List Sorting Working Memory Test
3. Flanker Inhibitory Control and Attention Test
4. Dimensional Change Card Sort (DCCS) test
5. Picture Sequence Memory test (PSMT)
6. Picture Vocabulary Task (TPVT) 

I have read and re-read the data manual and am not sure which variables I should be using in the analysis as some seem to require some combination of reaction time and accuracy while others are pretty straightforward (just raw score). Does anyone have any resources as to which variables to use? I have been looking at the LS2.0 crosswalk spreadsheet but still cannot figure it out.

Thanks in advance for any help you can provide!

Hannah Alves Hagy 

Graduate Student

Loyola University Chicago

Department of Psychology

1032 W. Sheridan Road

Chicago, IL  60660

[Elam, Jennifer]

Hi Hannah,

Are you just making comparisons between HCA subjects for your analyses? From the LS 2.0 Crosswalk it looks like all but the Picture Vocabulary Task (TPVT) has a Raw score you can use, Raw Score for X Instrument (Inst) specified, e.g. nih_picseq_raw

For the Picture Vocabulary Task, I would use the Uncorrected Standard score (tpvt_uss).

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
el...@wustl.edu
www.humanconnectome.org

---

From: Hannah Alves Hagy <hannah....@gmail.com>
Sent: Monday, January 8, 2024 10:17 AM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: [hcp-users] Using NIH Cognitive data

 

* External Email - Caution *

--
You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.
To view this discussion on the web visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAGWN3GUtEZzzA%3Dr0HyXBX%2BF_w6jrsNTY1u_4tXrjvR82BECkzQ%40mail.gmail.com.

 

---

The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail.

[Toshihiko Aso]

Dear all,

I want to merge the HCP-YA and HCP-Aging dataset to see the effect of crystalized intelligence, but, for example,  read_uss values are missing in orrt01.txt from HCP-Aging, while ReadEng_Unadj is found in HCP-YA "unrestricted" table. Is there a way to calculate the uncorrected values for HCP-A?

Best regards,

Toshihiko ASO

2024年1月9日火曜日 5:40:21 UTC+9 el...@wustl.edu:

[Toshihiko Aso]

Dear all,

I want to merge the intelligence measures, Gc/Gf, from HCP-YA and HCP-Aging datasets, but some items such as 
"Reading test Uncorrected Standard Score" is missing in all HCP-A subjects (orrt01.txt). In addition, a subset of subjects have not "uncorrected standard score" while they have "adjusted" TPVT values.

I also found this message: 

https://nihtoolbox.zendesk.com/hc/en-us/articles/18799834304788-Raw-Scores-for-Picture-Vocabulary-Test-and-Oral-Reading-Recognition-Test

So is it impossible to compare uncorrected values for these Gc measures with HCP-YA dataset for which those are provided?

Best regards,

Toshihiko ASO

2024年1月9日火曜日 5:40:21 UTC+9 el...@wustl.edu:

Hi Hannah,

[Elam, Jennifer]

Hi Toshihiko,

As we announced with the LS2.0 release, you should be using our Behavioral data "Crosswalk" Data Dictionary Excel spreadsheet to find out what data is included in the release and translate NDA data structures/elements to HCP behavioral variables and instruments. The NDA Data Dictionary has duplicative and sometimes misleading variable names and descriptions. The NDA also did not map our variables to the most obvious data elements for a particular structure every time. We recommend removing empty columns when you download the data and only referring to the crosswalk to find your variables of interest.

 

The “tbx_reading_score” variable in the “orrt01” data structure has the “Uncorrected Standard score for the Oral Reading Recognition Test Instrument (Inst) specified”.  There are some subjects that are missing that score, but the vast majority have it.

This score is equivalent to “ReadEng_Unadj” in HCP-YA.

 

For Picture Vocabulary you should use “tpvt_uss” in the “tpvt01” data structure: “Uncorrected Standard score for the Picture Vocabulary Test Instrument (Inst) specified”.

This score is equivalent to “PicVocab_Unadj” in HCP-YA

 

You may also be interested in the composite cognition scores in cogcomp01 and the analogous variables for HCP-YA for the analysis you are doing.

 

Best,

Jenn

 

________________________________________

Jennifer Elam, Ph.D.
Scientific Outreach, Human Connectome Project
Washington University School of Medicine

el...@wustl.edu
www.humanconnectome.org

 

To view this discussion visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/2f1f3c1e-be11-42cb-80d9-62d09d740f0bn%40humanconnectome.org.

[Toshihiko Aso@Gmail]

Dear Jenn,

Thanks for the prompt reply! (And sorry for the duplicate post) I should have dug more before asking.

Now I see that tbs_reading_score is supposed to replace ReadEng in S1200, but I still find some inconsistency in the distribution of values as in the attached plots. Actually this was my first attempt. 

Am I still missing something? And is it safe to use the age-adjusted version for adults >18 despite this possible change in battery? 

Toshihiko

2025年8月27日(水) 0:03 'Elam, Jennifer' via HCP-Users <hcp-...@humanconnectome.org>:

You received this message because you are subscribed to a topic in the Google Groups "HCP-Users" group.
To unsubscribe from this topic, visit https://groups.google.com/a/humanconnectome.org/d/topic/hcp-users/iLBNIeM014Y/unsubscribe.
To unsubscribe from this group and all its topics, send an email to hcp-users+...@humanconnectome.org.
To view this discussion visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CH3PR02MB10309B7A1ADC68A7DF83E9938CC39A%40CH3PR02MB10309.namprd02.prod.outlook.com.

[Toshihiko Aso@Gmail]

Dear Jenn,

OK, maybe the entry "Unclear" in "harmonized" is indicating the known inconsistency of ORRT test between YA and LS and we are not supposed to combine these. Thanks for your help!

Best regards,

Toshihiko

[Toshihiko Aso@Gmail]

For those who thought there might be a plotting error, I changed the color and marker to easily view the difference in distribution. 

Another Gc component, PicVocab, also has a similar shift in the "raw" values. It looks fine after age-adjustment.

Interesting trend in HCP-A (like crystalized intelligence) goes away after combining the YA results.

2025年8月27日(水) 16:22 Toshihiko Aso@Gmail <aso.to...@gmail.com>:

[Elam, Jennifer]

Hi Toshihiko,

There are different version numbers for the Oral Reading Recognition Test assessment available in 2012 (NIH TB Oral Reading Recognition ENG Age 7+, web based, used in HCP-YA) and in 2016 (NIH Toolbox Oral Reading Recognition Test Age 3+ v2.0, iPad app, part of NIH Toolbox v2 used in Lifespan). Scoring guides for each are attached but the descriptions of the assessment in each are not substantively different. I can't find anything that documents what changes were made between the two versions or how the scoring was calculated. However, NIH Toolbox has evolved over time, see the comparisons between v2 and v3 https://nihtoolbox.zendesk.com/hc/en-us, this about differences in scoring between versions (although it doesn’t mention the ORRT specifically): NIH Toolbox V2 2016 Rescoring Summary, and this about the v2 language scoring.

 

So, all this to say that there is likely a scoring difference over the two versions and that is why you see “unclear” in the “harmonized” column.  One way to possibly deal with that would be to use theta scores. Theta scores for Lifespan were directly released in the orrt01 data structure. For HCP-YA we were only provided computed scores, which are derived from theta scores, but I don’t have the formula used for v1 of the NIH Toolbox (it might be similar to that given for v2 language scoring, but likely not the same). The HCP-YA computed scores are available on request as they were output from NIH Toolbox and I will send you that file off list.

 

I would try reaching out to the helpdesk at NIH Toolbox and see if they have documentation of v1/v2 differences for ORRT and its scoring: he...@nihtoolbox.org

NIH Toolbox development was centered at Northwestern University, the staff that I had contact with have since left, but the faculty are still there and might know where to find information if the Toolbox helpdesk doesn’t provide what you need:

Richard Gershon and Cindy Nowinski

 

Best,

Jenn

 

---

From: Toshihiko Aso@Gmail <aso.to...@gmail.com>
Sent: Wednesday, August 27, 2025 5:01 PM
To: hcp-...@humanconnectome.org <hcp-...@humanconnectome.org>
Subject: Re: [hcp-users] Using NIH Cognitive data

 

For those who thought there might be a plotting error, I changed the color and marker to easily view the difference in distribution. 

 

 

Another Gc component, PicVocab, also has a similar shift in the "raw" values. It looks fine after age-adjustment.

Interesting trend in HCP-A (like crystalized intelligence) goes away after combining the YA results.

 

2025年8月27日(水) 16:22 Toshihiko Aso@Gmail <aso.to...@gmail.com>:

Dear Jenn,

 

OK, maybe the entry "Unclear" in "harmonized" is indicating the known inconsistency of ORRT test between YA and LS and we are not supposed to combine these. Thanks for your help!

 

Best regards,

Toshihiko

--

You received this message because you are subscribed to the Google Groups "HCP-Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hcp-users+...@humanconnectome.org.

To view this discussion visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CAAciWi_gJKPQH0K9RojLK04AhSzgZt3natAhmuf9rJ%3DJPKj6vA%40mail.gmail.com.

[Toshihiko Aso@Gmail]

Thank you for the clear and helpful information!!! I’ll work on it.
Best regards,

Toshihiko ASO, MD, PhD
麻生 俊彦

2025年8月29日(金) 1:21 'Elam, Jennifer' via HCP-Users <hcp-...@humanconnectome.org>:

To view this discussion visit https://groups.google.com/a/humanconnectome.org/d/msgid/hcp-users/CH3PR02MB10309082F6E4EFFEE202DA3DBCC3BA%40CH3PR02MB10309.namprd02.prod.outlook.com.