The NC State All-campus Data Science and AI Project-based Teaching and Learning (ADAPT) Model: A mechanism for interdisciplinary engagement in workforce-relevant learning

(2) College of William; David Stokes (1) ((1) NC State University; James B. Harr III (2); Mary); Rachel Levy (1)

arxiv: 2604.02597 · v1 · submitted 2026-04-03 · 💻 cs.CY

The NC State All-campus Data Science and AI Project-based Teaching and Learning (ADAPT) Model: A mechanism for interdisciplinary engagement in workforce-relevant learning

Rachel Levy (1) , James B. Harr III (2) , David Stokes (1) ((1) NC State University , (2) College of William , Mary) This is my paper

Pith reviewed 2026-05-13 19:13 UTC · model grok-4.3

classification 💻 cs.CY

keywords data science educationAI in higher educationinterdisciplinary learningproject-based learningacademy modelworkforce developmentuniversity structure

0 comments

The pith

NC State University established an Academy to weave data science and AI into teaching, research, and engagement across every academic unit.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper describes NC State's creation of an Academy as a new structure to advance data science and AI throughout the university. Unlike traditional methods that isolate these fields in new departments or colleges, the Academy works across existing units to support project-based learning. The goal is to prepare students for data-rich careers by embedding relevant skills in diverse disciplines. Readers might care because it proposes a solution to the challenge of adapting higher education quickly to emerging technologies without adding more silos.

Core claim

The authors describe how NC State formed an Academy to catalyze data science and AI work in all aspects of the university mission by connecting disciplines, departments, colleges, centers, and institutes, with a focus on project-based teaching and learning through the ADAPT model.

What carries the argument

The Academy, a cross-cutting entity that organizes project-based data science and AI learning across the campus.

Load-bearing premise

That establishing a central Academy will reduce departmental silos and lead to better interdisciplinary and workforce-relevant outcomes than conventional departmental approaches.

What would settle it

A study comparing rates of interdisciplinary data science projects, student skill assessments, or post-graduation employment in data-related roles at NC State before and after implementing the Academy.

read the original abstract

Academic institutions have been challenged to adapt as data science and AI have rapidly evolved into disciplines, degrees and careers. Efforts to provide students with learning experiences have led to the development of novel credentials, renamed departments, new schools and even additional colleges within universities. Generally, these approaches are siloed in some way, perhaps separating STEM students from those in the humanities or separating faculty assigned to these courses from their colleagues in their home departments. NC State University decided to take a novel approach by creating a new type of entity called an Academy that would reach across all disciplines, departments, colleges, centers and institutes to catalyze work in data science and AI in all points of the university's mission: teaching, research and engagement.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

NC State's ADAPT Academy is a descriptive account of a cross-cutting structure for data science and AI, but it supplies no metrics or outcomes to show whether the approach reduces siloing or improves results.

read the letter

This paper describes NC State University's creation of the ADAPT Academy as a university-wide entity meant to integrate data science and AI across teaching, research, and engagement without the usual departmental silos. The authors explain the governance setup and how it reaches every college and institute rather than carving out a separate school or department. That part is clear and gives a concrete picture of one institution's structural choice. The description itself is the main contribution here, as it spells out an alternative to the renamed departments or new colleges mentioned in the abstract. It does a reasonable job showing the intent behind the design. The soft spot is straightforward: there is no data at all. No enrollment numbers across colleges, no counts of joint projects or publications, no pre/post measures of collaboration, and no student outcome assessments. The claim that this setup catalyzes better interdisciplinary, workforce-relevant learning rests entirely on the structure being in place, with nothing to verify it works better than prior approaches. This leaves the central assertion as an untested hypothesis. The paper is aimed at higher-education administrators or faculty committees thinking about scaling data science programs. It has little for technical researchers in data science or AI itself. I would not bring it to a reading group focused on methods or results, and I would not cite it. It does not deserve peer review in its current form because the effectiveness claims cannot be evaluated without evidence; adding even basic metrics would change that.

Referee Report

2 major / 2 minor

Summary. The manuscript describes NC State University's creation of the ADAPT Academy as a novel cross-cutting entity intended to catalyze data science and AI activities across teaching, research, and engagement missions by spanning all disciplines, departments, colleges, centers, and institutes, in contrast to traditional siloed structures such as new departments or colleges.

Significance. If the ADAPT model can be shown to deliver measurable reductions in departmental siloing and improvements in interdisciplinary, workforce-relevant outcomes, it would provide a replicable structural alternative for universities adapting to rapidly evolving fields. The paper's value lies in its detailed account of governance and program design, but the absence of supporting evidence limits its contribution to the literature on educational innovation.

major comments (2)

[Abstract and sections describing programs/governance] Abstract and main descriptive sections: The central claim that the Academy successfully catalyzes interdisciplinary work and avoids siloing is presented as a structural solution but is unsupported by any quantitative metrics, student outcome data, cross-college enrollment figures, joint publication rates, or pre/post collaboration indicators. This absence leaves the effectiveness assertion as an untested hypothesis rather than an evidenced result.
[Sections on implementation and intent] Sections on implementation and intent: No comparison is provided against traditional departmental or college-based approaches, nor are any evaluation methods, assessment instruments, or longitudinal tracking plans described that would allow verification of improved learning outcomes or reduced fragmentation.

minor comments (2)

[Title] The title is lengthy and could be streamlined for readability while retaining key terms.
[Overall structure] Consider adding a dedicated evaluation or outcomes section in a revision to strengthen the manuscript.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for their constructive feedback on our manuscript describing NC State's ADAPT Academy. We appreciate the recognition of the model's novelty as a cross-cutting entity and agree that stronger framing around evidence and evaluation is needed. We will revise to clarify the descriptive intent of the paper while adding details on planned assessments and comparisons.

read point-by-point responses

Referee: Abstract and main descriptive sections: The central claim that the Academy successfully catalyzes interdisciplinary work and avoids siloing is presented as a structural solution but is unsupported by any quantitative metrics, student outcome data, cross-college enrollment figures, joint publication rates, or pre/post collaboration indicators. This absence leaves the effectiveness assertion as an untested hypothesis rather than an evidenced result.

Authors: We agree that the manuscript should not overstate outcomes. The paper is primarily a descriptive account of the Academy's governance and program design as a structural alternative to siloed approaches. We will revise the abstract and relevant sections to emphasize intended mechanisms rather than demonstrated success, and we will include any available preliminary indicators from the initial launch phase. Comprehensive longitudinal data are not yet available given the recency of the initiative. revision: partial
Referee: Sections on implementation and intent: No comparison is provided against traditional departmental or college-based approaches, nor are any evaluation methods, assessment instruments, or longitudinal tracking plans described that would allow verification of improved learning outcomes or reduced fragmentation.

Authors: We will add a dedicated subsection comparing the Academy model to traditional departmental and college-based structures, drawing on existing literature regarding interdisciplinary education and workforce preparation. We will also describe the evaluation framework under development, including specific assessment instruments, cross-college enrollment tracking, and longitudinal plans for measuring collaboration and learning outcomes. revision: yes

standing simulated objections not resolved

Provision of comprehensive quantitative metrics (e.g., student outcomes, joint publications, pre/post collaboration data) because the ADAPT Academy is newly established and such data do not yet exist.

Circularity Check

0 steps flagged

Descriptive account of ADAPT Academy shows no circularity

full rationale

The manuscript is a purely descriptive case study of the NC State ADAPT Academy without equations, fitted parameters, predictions, derivations, or load-bearing self-citations. The central description of creating a cross-cutting Academy to catalyze data science and AI work across teaching, research, and engagement is presented as an institutional structural choice rather than derived from or equivalent to its own inputs by construction. No steps reduce to prior results or fits; the account is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The description rests on the untested premise that cross-cutting structures outperform traditional departments for data and AI education; no empirical support or external validation is supplied.

axioms (1)

domain assumption Siloed departmental structures hinder effective data science and AI education across disciplines
Explicitly stated in the abstract as the motivation for creating the Academy.

invented entities (1)

Academy no independent evidence
purpose: Cross-cutting entity to integrate data science and AI into teaching, research, and engagement without removing faculty from home departments
New organizational form introduced to solve the stated siloing problem; no prior evidence of effectiveness is cited.

pith-pipeline@v0.9.0 · 5452 in / 1260 out tokens · 49919 ms · 2026-05-13T19:13:12.414968+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

[1]

Oh, I don't know how to wire this machine. How do I put a potentiometer so that my bubble machine can change and vary speed?

you will ﬁnd links to the R/Python book, a network analysis book and ADAPT related research articles. DSA has a number of ways that we provide workshops to external groups through Data and AI at work . We have an annual ADAPT ShareFair in which instructors and students share their experiences with the ADAPT mode and provide examples of student work. You a...

work page 2025
[2]

professional learning community

Acknowledgements The authors of this article are listed alphabetically as authorship was a collaboration with both contributing to the inception and execution of this paper. Dr. Rachel Levy, Executive Director of the Data Science Academy and Professor of Mathematics at North Carolina State University, initiated this report and was the lead on the developm...

work page doi:10.1002/tea.20237 2001

[1] [1]

Oh, I don't know how to wire this machine. How do I put a potentiometer so that my bubble machine can change and vary speed?

you will ﬁnd links to the R/Python book, a network analysis book and ADAPT related research articles. DSA has a number of ways that we provide workshops to external groups through Data and AI at work . We have an annual ADAPT ShareFair in which instructors and students share their experiences with the ADAPT mode and provide examples of student work. You a...

work page 2025

[2] [2]

professional learning community

Acknowledgements The authors of this article are listed alphabetically as authorship was a collaboration with both contributing to the inception and execution of this paper. Dr. Rachel Levy, Executive Director of the Data Science Academy and Professor of Mathematics at North Carolina State University, initiated this report and was the lead on the developm...

work page doi:10.1002/tea.20237 2001