Verifiable Random Function (VRF): Research Randomisation

A Verifiable Random Function (VRF) is a public-key cryptographic primitive that generates a deterministic, pseudorandom output from an input and a secret key, alongside a non-interactive proof. This proof allows third parties holding the corresponding public key to verify that the output was correctly computed without revealing the secret key.

While cryptographic VRFs secure distributed systems and consensus networks, randomisation in research serves a distinct methodological purpose: assigning participants or stimuli to experimental conditions without bias. Understanding the distinction between a cryptographic VRF and an auditable randomisation workflow is essential when selecting methodology for academic studies, trial allocations, and reproducible science.

What Is a Verifiable Random Function?

Formalised in IETF RFC 9381, a VRF can be conceptualised as the public-key equivalent of a keyed cryptographic hash function or Message Authentication Code (MAC).

A standard VRF scheme consists of three core components:

1. Input (x): The data seed supplied to the evaluation function.
2. Output (y): The deterministic pseudorandom value derived from the input and secret key.
3. Proof (π): Cryptographic evidence enabling any verifier possessing the public key (pk) to validate the output.

For example, a blockchain protocol might use a VRF to select a participant for a particular role. The selected participant can publish the output and its proof, allowing other network participants to verify the result without gaining access to the secret key.

The same principle of independent verification is relevant to research methodology: a reader should be able to examine the evidence supporting a reported randomisation outcome rather than having to rely entirely on the researcher's description.

However, a cryptographic VRF and an auditable random selection tool are not necessarily the same technology. Each must be evaluated according to the specific evidence and security guarantees it provides.

How Does a Verifiable Random Function Work?

How a VRF Operates

1. Key Generation: A key pair (sk, pk) is produced. The secret key (sk) remains private; the public key (pk) is distributed publicly.
2. Evaluation: The secret key and input are processed by the algorithm: (y, π) ← Eval (sk, x).
3. Determinism: For any given (sk, x), the output y is fixed and invariant.
4. Public Verification: Third parties run Verify (pk, x, y, π), returning 1 if the output is valid or 0 if tampered with.

Verifiable Random Function vs. Random Number Generator (RNG)

Although the terms are related, a verifiable random function and a random number generator have different purposes.

A random number generator may be entirely suitable for a research task, but the existence of a random result does not automatically establish that the process is independently verifiable.

Using a standard PRNG generates random data, but it does not produce independent proof that the reported result was the one generated. Conversely, implementing a cryptographic VRF provides proof of generation but does not inherently validate the experimental design or recruitment process.

What Is Verifiable Randomisation in Research?

Verifiable randomisation is a process in which a random selection is accompanied by evidence that allows another person to inspect or check how the result was produced.

In empirical research, verifiable randomisation requires providing an audit trail demonstrating that experimental conditions, treatments, or stimulus presentation orders were selected without researcher bias.

Randomisation, reproducibility and verification are different

These concepts are related, but they should not be treated as interchangeable.

Randomisation is the procedure used to make a selection or assign a condition.

Reproducibility concerns whether a researcher or another investigator can repeat an analysis or research procedure using the documented materials and methods.

Verification concerns checking whether a particular result or process satisfies specified requirements.

A saved randomisation record can support research transparency, but it does not independently prove that the study design was appropriate, the participants were recruited fairly or the statistical analysis was correct.

For this reason, researchers should describe exactly what their evidence establishes.

Common Pitfall: Binary Choice Allocation

A simple 50/50 binary selection (e.g., assigning participants to Treatment A or Control B via a fair coin flip) does not guarantee equal group sizes.

• Simple Random Allocation: Independent 50/50 probability per participant. Out of 100 participants, variance often yields unbalanced distributions (e.g., 53 vs. 47).
• Blocked Randomisation: Groups assignments into predefined blocks (e.g., 4 or 6) to maintain equal sample sizes across study arms throughout recruitment.
• Stratified Randomisation: Categorises participants by covariates (e.g., age, disease severity) prior to allocation to ensure balanced subgroups.

Protocol for Documenting Randomisation Methods

To ensure methodological transparency in academic publications, document the randomization workflow using the following checklist:

1. Define Alternatives & Mapping: Explicitly state the conditions (e.g., Option A = Experimental, Option B = Control) prior to executing the selection.
2. Specify Tooling & Seed Management: State the exact software, version, pseudo-random generator, or cryptographic tool used. Avoid describing standard PRNG draws as cryptographic VRFs.
3. Detail the Probability Model: Report whether allocations were simple (unweighted), weighted, blocked, or stratified.
4. Preserve Audit Artifacts: Store raw output logs, seeds, timestamped receipts, or verification proofs alongside open-data repositories.
5. Acknowledge Limitations: Clearly delineate what the verification artifact proves (e.g., selection integrity) versus what it does not (e.g., recruitment fairness or statistical power).

Sample Methods Template

"Participant allocation was conducted via a simple binary randomization procedure. Mappings between random outputs and experimental arms (Condition A vs. Condition B) were established prior to selection. Randomization seeds, output logs, and software configuration details were archived in the study's OSF repository to enable independent verification."

Using Spin the Wheel for Randomisation in Research

For researchers, students and educators who need a straightforward way to randomise between predefined alternatives, Spin the Wheel's research resources provide a starting point for exploring randomisation workflows.

A two-option wheel can represent a binary choice such as Yes or No, Condition A or Condition B, or one of two presentation orders. When equal weighting is configured, the two alternatives can represent a 50/50 selection for a single draw.

For research tasks that need a record of the selection, Spin the Wheel also describes a separate provably fair workflow. Its documentation explains how users can generate a record intended to support later checking of the selection process.

This distinction matters: an ordinary spin provides a random selection, while a spin performed through a supported verification workflow may provide additional evidence for research documentation. Researchers should check the available records and verification mechanism before describing the result as independently verifiable.

To explore the tool and its research applications, visit Spin the Wheel Research.

For academic use, document the options, weighting, procedure and retained evidence. A verified selection can support transparency, but it does not replace a suitable study design, statistical analysis or the additional safeguards required for high-stakes research.

Frequently Asked Questions About Verifiable Random Functions

What is a verifiable random function in cryptography?

A verifiable random function is a public-key cryptographic primitive that produces a deterministic pseudorandom-looking output for a given secret key and input, together with a proof that can be checked using the corresponding public key. It enables independent verification of the output without revealing the secret key.

Is a verifiable random function the same as a random number generator?

No. A random number generator produces random or pseudorandom values, whereas a VRF is designed to provide a verifiable cryptographic output. A conventional generator does not necessarily provide a public verification proof.

What is verifiable randomisation in research?

Verifiable randomisation is a selection procedure supported by evidence that allows the process or recorded outcome to be checked. Depending on the tool, this evidence may consist of a cryptographic proof, a replayable record or an audit trail. The precise guarantees depend on the implementation.

Can I use a Yes or No wheel for binary choice randomisation?

Yes. A two-option wheel can be used for a simple binary selection, provided the options and intended probabilities are configured appropriately. For formal research, document the procedure and distinguish an ordinary random result from a result supported by a verification mechanism.

Does a 50/50 randomiser guarantee equal research groups?

No. A 50/50 probability gives each outcome an equal chance on each independent selection, but the final group sizes can differ. If a study requires balanced allocation, a method such as blocked randomisation may be more appropriate.

Can a randomisation result be cited in a research paper?

A randomisation procedure can be described in a paper's methods section, and a retained verification record can be referenced or archived when the tool supports it. Researchers should cite the relevant technical specification or tool documentation, describe what was actually done and preserve the evidence needed to substantiate the claim.

A citation to a tool or technical standard does not, by itself, prove that a particular randomisation was performed correctly.

Is Spin the Wheel a verifiable random function?

The terms should not be used interchangeably. A cryptographic VRF has a specific public-key construction and verification proof. Spin the Wheel describes a provably fair randomisation workflow with its own verification records. Researchers should consult the tool's technical documentation to identify the mechanism used and describe its guarantees accurately.