Frequently Asked Questions

Will the code in abcd-rstools work if I don’t have the ABCD Study data?

No, you need the ABCD Study data for the code to work.

You can only get the ABCD data from the NIH via the NBDC Data Hub.

Is the code available in another programming language besides R?

No, the code is only available in R.

What is the “item metadata” and how did you create it?

The item metadata (ABCD-item-metadata.csv) is a table that identifies which variables in the ABCD Study data could be survey items on a risk score. The table also has additional information that might be helpful when creating a risk score (e.g., which topic that question asked about, who the question was asked of).

We made the item metadata by looking at every single variable in the ABCD 6.0 release. For each variable, we checked whether it met the following criteria:

(a) it could be measured via survey item (vs., e.g., required brain imaging)

(b) it was administered in the core study, to the full ABCD sample (vs. a subsample or the eligibility screener)

(c) if skip/display logic was present, it could be appropriately handled by recoding

Guided by this table, we then did all the recoding that would be necessary to create a clean dataframe of items that are ready to enter into risk scores and test as predictors.

The script worked-example.R in abcd-rstools shows how the finished item metadata can be used to create and evaluate a risk score. If you want to know more detail about how we created the metadata, and handled edge cases, view some further documentation here.

Are there any problems in the item metadata I should be aware of?

1. Items the measured the youth’s substance use are currently excluded from the pool of items. This is because the ABCD Study’s assessment of substance use entailed some complicated skip logic that differed over time. We are still finishing the recoding of the substance use variables into items and expect to finish in January 2025.

I think there is a bug in your code or error in your item metadata — what should I do?

[pending: check back later]

How do I acknowledge / cite this tool?

Cite this paper: Pelham III, W. E., Dontha, M., Aks, I. R., Nooney, A., Chitty, T. C., Gonzalez, O., Patel, H., Kemp, E. C., Blyth, S. H., Tapert, S. F., & Brown, S. A. (2025). New risk indexes for identifying children at risk of high-risk substance use during adolescence. Preprint: https://osf.io/preprints/psyarxiv/4rfve_v1

Where can I read more about creating and evaluating predictive modeling tools?

Here are some references!

Steyerberg, E. W., & Harrell, F. E. (2016). Prediction models need appropriate internal, internal–external, and external validation. Journal of Clinical Epidemiology, 69, 245–247. https://doi.org/10.1016/j.jclinepi.2015.04.005

Japkowicz, N., & Shah, M. (2011). Evaluating Learning Algorithms: A Classification Perspective. Cambridge University Press.

Hemingway, H., Croft, P., Perel, P., Hayden, J. A., Abrams, K., Timmis, A., Briggs, A., Udumyan, R., Moons, K. G. M., Steyerberg, E. W., Roberts, I., Schroter, S., Altman, D. G., & Riley, R. D. (2013). Prognosis research strategy (PROGRESS) 1: A framework for researching clinical outcomes. BMJ, 346, e5595. https://doi.org/10.1136/bmj.e5595 

Riley, R. D., Hayden, J. A., Steyerberg, E. W., Moons, K. G. M., Abrams, K., Kyzas, P. A., Malats, N., Briggs, A., Schroter, S., Altman, D. G., Hemingway, H., & Group,  for the P. (2013). Prognosis research strategy (PROGRESS) 2: Prognostic factor research. PLOS Medicine, 10(2), e1001380. https://doi.org/10.1371/journal.pmed.1001380 

Steyerberg, E. W., Moons, K. G. M., Windt, D. A. van der, Hayden, J. A., Perel, P., Schroter, S., 

Riley, R. D., Hemingway, H., Altman, D. G., & Group,  for the P. (2013). Prognosis research strategy (PROGRESS) 3: Prognostic model research. PLOS Medicine, 10(2), e1001381. https://doi.org/10.1371/journal.pmed.1001381

Hingorani, A. D., Windt, D. A. van der, Riley, R. D., Abrams, K., Moons, K. G. M., Steyerberg, E. W., Schroter, S., Sauerbrei, W., Altman, D. G., & Hemingway, H. (2013). Prognosis research strategy (PROGRESS) 4: Stratified medicine research. BMJ, 346, e5793. https://doi.org/10.1136/bmj.e5793

My question isn’t answered here - whom should I contact?

Email <wpelham at ucsd dot edu>.