Document type
Preprint
Version
1.0
Date
21 September 2026
Peer review
Not yet reviewed
DOI
Not yet assigned
Licence
CC BY 4.0
Preprint — not peer reviewed
Findings may change after review

The Substitution Paradox

Digital capacity for virtual science laboratories is lowest in the Indian schools that lack physical laboratories — a national analysis of 1,471,473 schools

Sreeram Venugopal
Independent researcher · Puducherry, India
ORCID 0009-0009-2916-7633

Abstract

Virtual laboratories are widely proposed as a remedy for the shortage of physical science laboratories in Indian schools. The argument contains an untested empirical premise: that schools lacking laboratories possess the electricity, computing devices and connectivity needed to run a digital substitute.

This study tests that premise against the complete 2024–25 UDISE+ administrative census of Indian school education. Because running a cloud-connected virtual laboratory requires electricity and a device and internet simultaneously, while published statistics report only the three marginal distributions, the governing quantity is not point-identified; sharp Fréchet–Hoeffding bounds are derived rather than independence assumed.

Capacity and need are found to be inversely aligned, and the association survives adjustment for state income, exclusion of small territories, regional demeaning and every alternative definition of capacity examined. The conclusion is not that virtual laboratories fail to teach — the evidence that they support conceptual learning is substantial — but that their reach is bounded, and that aggregate claims about democratising laboratory science through cloud delivery are not supported by the infrastructure distribution.

Principal findings

63.47%Upper bound on schools able to run a cloud-connected virtual laboratoryFréchet upper bound · assumption-free
537,486Schools that cannot, under any configuration consistent with the data≥ 36.53% of all schools
−0.553Correlation between laboratory deficit and digital capacity ceiling95% CI [−0.762, −0.278] · ρ = −0.523
59.2%Of all Indian schools sit in states below median on both physical and digital provision13 states · 871,398 schools
124,741Secondary sections with no functional integrated science laboratory42.9% of 290,950
36States and union territories analysed — a census, not a sampleUDISE+ 2024–25 · ref. 31 Mar 2025

Scope and status

Study 1 — complete. The national infrastructure analysis above is finished, reproducible and verified. Every reported figure regenerates from the published raw tables, and 28 automated checks confirm the manuscript matches the computed values.

Study 2 — protocol only. A preregistered cluster-randomised trial of conceptual learning is specified in full, with instruments and analysis code deposited in advance. No participant has been recruited, no datum collected and no ethics approval obtained. Nothing in that section may be cited as a finding.

Declarations

Competing interests
The author is associated with SciPhyLabs, an interactive physics-learning platform. Study 1 uses only government administrative data and does not involve that platform. In the Study 2 protocol the platform appears solely as a delivery and logging environment, with simulation content specified to come from an independently developed third-party suite. No finding here is offered as evidence for any platform's effectiveness.
Ethics
Study 1 used only publicly available aggregate data concerning institutions, not individuals, and required no ethical review. Study 2 has not received ethical approval and has not begun.
Data and code
All data are public. Raw transcribed tables, the derived analysis panel, full provenance and every analysis script are openly available. The pipeline rebuilds all reported figures in under one minute.
AI assistance
Generative AI tools were used for literature search and screening support, drafting and editing assistance, and writing and checking analysis code. All statistical results were computed by the deposited scripts. All cited works were verified against publisher records; verification corrected seven author or year errors and removed four sources that could not be verified.

How to cite

Venugopal, S. (2026). The Substitution Paradox: Digital Capacity for Virtual Science Laboratories Is Lowest in the Indian Schools That Lack Physical Laboratories (Version 1.0) [Preprint].

Access