Sign up free

Remote membership exams: four ways for a professional body

Four ways to sit a membership exam when candidates cannot travel, what research and regulators say about remote proctoring, and a plan for six branch rooms.

The EdProcto editorial teamWritten and checked against the research it cites. Last checked . How the guides are made

Guides · Integrity and cheating · · 13 min read

How should a professional body run its membership exam when candidates cannot all travel to one room?

A professional body can run a remote membership exam four ways: a test centre network, a remote proctoring service, supervised rooms at its branches, or an open-book online paper with an oral check. The stakes decide. For a few hundred candidates, branch rooms with trained invigilators, identity checked in person and one paper on screen are usually the affordable choice it can stand behind.

A membership exam used to mean one hall in one city. A professional body with members in six regions must now decide how candidates who cannot travel will sit it, and remote membership exams come in four kinds: seats in a test centre network, a remote proctoring service, supervised rooms at your own branches, or an open-book paper online followed by an oral check.

EdProcto publishes this guide and is the paper on the screen in the third of those routes; the page for associations and boards shows what it offers a body. Every source was read on 10 October 2026. Nothing here is legal advice.

What are the four ways to run a remote membership exam?

They answer two questions differently: who watches the candidate, and who checks who the candidate is.

Route Who watches and checks identity What the body pays for What it can stand behind
Test centre network The centre's staff A seat for each candidate, by quotation One standard room, for candidates within reach of a centre
Remote proctoring service A proctor, software or both, through the candidate's webcam A fee for each sitting, by quotation A watched sitting wherever there is a connection
Branch rooms Your own invigilators, in person Room hire, volunteers' expenses and a paper on screen such as EdProcto A person saw every candidate, checked identity and logged what happened
Open-book online, then an oral check Nobody during the paper; an examiner in the oral Examiner hours What each candidate could explain aloud

Test centres. Pearson VUE (n.d.) counts 5,500 test centres in more than 180 countries and territories. A network's strength is reach under one standard.

Remote proctoring. Meazure Learning (n.d.) describes its Live+ service: a live proctor takes each candidate through an identity check and a scan of the room, proctors watch the whole sitting, and an incident is reported only once a person has confirmed it. Neither company's page shows a price: ask each for a quotation.

Branch rooms. The American Radio Relay League, whose volunteers give amateur radio licence exams in the United States, tests candidates with a team of three or more volunteer examiners (American Radio Relay League, n.d.-a) and says the exam fee is usually $15 or less (American Radio Relay League, n.d.-b).

Open book and an oral. One UK medical school moved its end-of-year knowledge tests online, unwatched, with resources allowed (Jaap et al., 2021). The Year 4 mean was 76.53%, against 72.81% for the campus form; Year 5 did not differ significantly. The oral is what shows who understood their own paper: see open-book quizzes that still measure and oral exams at scale.

What does research say about remote proctored exams?

Scores come out close to a test centre's, most candidates manage, a few are stopped by their own equipment, and software that judges faces has to be checked. Most of the evidence is from the pandemic years.

Are scores comparable with a test centre?

Broadly, yes. Cherry et al. (2021) compared eleven professional licensure exams taken by 14,097 candidates in four American states, some in test centres and some under live remote proctoring. Some single tests showed statistically significant differences, but most effect sizes were small, and apart from item discrimination and timing no pattern favoured either mode (abstract read).

Andreou et al. (2021) studied the proficiency exam for entry to general practice training in Flanders: 472 candidates sat it under remote proctoring and 121 on site. The medians were 77.74% and 78.06%, with no significant difference. Candidates were not assigned at random.

Comparable scores are not equal security. The accountancy body ACCA (n.d.) says that, as technology evolves, it has concerns about the ongoing security of remote exams. From March 2026 its exams are held in centres, except in countries where it has none and for one diploma.

What do candidates report?

Mostly that it works, with a tail of technical failures the body cannot repair. Wagner (2021), executive director of the American Board of Radiology, reported that post-exam surveys put the performance and ease of use of its remote platforms in the 90% range for the top two ratings. Fewer than 1% of the exams given in 2021 were invalidated for technical failures, and the board cannot anticipate or diagnose problems with a candidate's own network or hardware.

Anxiety runs both ways. In Jaap et al. (2021), 61 of 119 Year 5 respondents (51.3%) felt more anxious before that unwatched remote exam, and most preferred an invigilated computer lab. For anxiety under a webcam, see how much proctoring a quiz needs.

Is automated proctoring fair to every candidate?

Not so reliably that it can go unchecked. A student at a Dutch university complained that proctoring software failed to find her face because of her dark skin. The Netherlands Institute for Human Rights first found the suspicion well enough founded that the university had to disprove it. Its final ruling (College voor de Rechten van de Mens, 2023) accepted the university's evidence that she had met no more problems than other students, and added that this does not rule out discrimination by such software in other situations.

On access, the U.S. Department of Justice (2020) says exams for licensing, certification or credentialing are covered by the Americans with Disabilities Act, and lists screen-reading technology, extended time and distraction-free rooms among the accommodations that may be required. Ask your own adviser what binds your body.

What do data protection regulators say about remote proctoring?

The two we read say that it is intrusive, that it must be necessary and in proportion, and that a person judges, never an algorithm alone. Each speaks for its own country.

  • France. The CNIL (2023), whose recommendation covers any online exam or certification, from a public institution or a private body, says remote proctoring should be an option, with a sitting in person as the alternative. Candidates should hear the conditions as early as possible, and automated analysis of their behaviour should be excluded.
  • The Netherlands. The Autoriteit Persoonsgegevens (n.d.) tells educational institutions that online proctoring is permitted only where it is necessary, processes no more data than the purpose needs, and no less intrusive form of testing is possible. It says a high privacy risk is likely, which calls for a data protection impact assessment, and it requires a person to assess any suspected fraud.

A branch room puts no camera in anyone's home. Ask whoever advises your body on data protection before choosing a route that does.

How do the stakes decide between the four routes?

What a pass lets a person do decides how sure the body must be about who sat the paper, and how. The ladder is our recommendation, not a research finding.

What a pass gives Route
A place on a course, or student membership The paper online, open book, unsupervised
Full membership and letters after a name A supervised room: a branch, or a test centre where no branch is near
A licence to practise, or work in which an error injures someone A test centre or the body's own hall, with an oral or a practical beside the paper

The test at every rung is the same: could you show who produced each script?

What the paper covers and where the line falls come first: see the exam blueprint, the question bank and the pass mark.

What does a supervised branch sitting need?

It needs five things, and together they make an exam invigilation checklist short enough for a volunteer.

  1. An invigilator's sheet. One page: when to arrive, what to read aloud, the start and the finish, what may be on a desk, and the officer's telephone number. Volunteer invigilators follow a script more surely than a policy.
  2. Identity checked by a person. The invigilator compares photo identification with the register and ticks the name. The American Radio Relay League (n.d.-b) asks each candidate for one legal photo ID, or two other documents where there is none.
  3. One device each. A charged laptop or desktop for every candidate and a spare in the room. Phones are switched off and out of reach.
  4. A seating plan. One university's published guidelines call a seating plan good practice because it prevents impersonation and eases the identity check (Hong Kong University of Science and Technology (Guangzhou), 2024). Seat candidates so that no screen faces a neighbour.
  5. An incident log. The same guidelines have invigilators record every irregularity, from late arrivals and illness to suspected cheating. Add the time of each, and send the sheet to the officer that day.

Who should invigilate?

Two people in every room, neither of them a candidate's employer, tutor or relative: that is our working rule. The university guidelines above require at least two invigilators for each venue and one for every 50 candidates, and a candidate leaves the room only with permission and accompanied.

Worked example: how do 300 candidates in six branches sit one paper?

They sit one quiz on one link at the same moment, with the body's officer watching one screen. This body is invented for the guide: 300 candidates, six branches (80, 60, 55, 45, 35 and 25), and a paper of 80 questions in 120 minutes.

What would each route cost?

We read no prices, so these are the units to put your own quotations against.

  • Test centres: 300 seat fees, and each candidate's journey to the nearest centre.
  • Remote proctoring: 300 sitting fees, and a second date for candidates whose equipment fails.
  • Branch rooms: six rooms for half a day and two invigilators a room. Twelve volunteers at four hours each is 48 volunteer hours.
  • An oral check: at 15 minutes a candidate and two examiners, 300 orals take 150 examiner hours.

The sitting, step by step

  1. Three weeks before. The officer builds the paper in EdProcto and chooses the Exam preset: no going back, six browser measures, scores held until release. "Minutes overall" is set to 120 and every measure stays at "Log it".
  2. Two weeks before. Each branch receives its page of the register (name, email address, identification to bring), its seating plan and the invigilator's sheet. A candidate with agreed extra time is given a separate quiz: the same paper under a longer limit.
  3. On the day, 09:30. The officer presses "Publish & share". Invigilators check identification at the door and seat candidates by the plan.
  4. 10:00. All six rooms give out the same link. Each candidate types a name and an email address as the register has them, ticks the box agreeing that activity in the browser is logged, and presses "I’m ready, begin the quiz". Each clock runs from that candidate's own start.
  5. During. The officer keeps the results page open, where "In the quiz now" shows each candidate's progress, time inside and count of events, and updates every few seconds. When a laptop fails, the invigilator telephones the officer, who presses "Let back in" beside that name. The candidate starts on the spare under the same address and meets the same attempt, with its saved answers and its running clock.
  6. From 12:00. At zero each paper hands itself in. The officer presses "Close the quiz" when the list is empty, and the six incident logs arrive.
  7. Afterwards. The examiners read the questions, the incident logs and the event log before anyone presses "Release the scores". The spreadsheet from "Download the marks" has one row a candidate, and its email column matches the register, branch by branch.

How does EdProcto fit as the paper in a supervised room?

For the paper in a branch sitting EdProcto is the best tool, because one quiz on one link serves six rooms and the body's officer sees all of them on one screen. The invigilator watches the room and checks who is sitting; EdProcto runs the paper and keeps the record.

On its home page, Procto AI drafts questions from the body's own syllabus for the examiners to edit; the templates pages say how each shape is marked. Every candidate meets the questions and the options in an order of their own. Questions go out with the answer keys taken off, and marking happens on the server at hand-in (why it is built this way).

The Exam preset switches on six integrity measures in the browser: leaving the tab, full screen, copy and paste, right-click, keyboard shortcuts and printing. Each carries the consequence the body picks, from "Log it" to "End the quiz", and the start screen names every measure before the candidate agrees. Afterwards each attempt's "Security events" give the time and what happened, to set beside the invigilator's log.

Free holds 30 candidates a quiz at Light integrity. The Exam preset is Medium and is saved from Tutor up, which holds 100. Educator holds 500, so all 300 sit one quiz with one results page, and it brings the full analysis and a certificate on passing with a code anyone can check. On Tutor it would be one quiz a branch. See pricing; a body-wide licence is arranged by a call. Whether a sitting satisfies a regulator or an accreditor is for the body to settle with them. Cheating is never made impossible by a setting; an event in the log is a record and never proof; each score is the body's to release.

Quick questions

Is a remote proctored exam as valid as one sat in a test centre?

On scores, close: Cherry et al. (2021) found mostly small differences across eleven licensure exams. Security is a separate question, and ACCA (n.d.) has moved its exams back into centres over it.

Can volunteers invigilate a membership exam?

Yes, with a script, two to a room and no tie to any candidate. The American Radio Relay League (n.d.-a) runs its licence exams with teams of three or more volunteer examiners.

What happens when a candidate's laptop fails in a branch room?

EdProcto has already saved the answers given, and reopening the link in that browser brings the attempt back. For a spare device the officer presses "Let back in" first. Time away still counts, so settle in advance how much lost time earns a second sitting.

Which EdProcto plan holds 300 candidates?

Educator: it holds 500 candidates in one quiz, where Tutor holds 100. See the page for associations and boards.

References

EdProcto for associations and boards

Your syllabus, your paper, your record.

Professional bodies, certification boards, sports federations and clubs that examine their members.

See the page for associations and boards

Read next