A register of who joined a webinar shows that members were there. It is silent on what any of them took away.
A CPD quiz for members closes that gap from the body's side. Each webinar, journal article or branch evening ends in five to eight questions written from the session itself and one line of reflection, and every sitting leaves a dated record.
EdProcto, which publishes this guide, runs such a quiz from one link: see its page for associations and boards. The member's side is in CPD evidence: how to show what you learned.
What do regulators and professional bodies ask of a CPD record?
They ask for a dated record of activities with evidence behind it; most count hours or units, and most ask what was learned or applied. We read three bodies' own pages on 10 October 2026 and report only what each says. None of the three asks for a test at the end of an activity.
| Body and page | Counted in | What the record holds | Learning and reflection |
|---|---|---|---|
| Health and Care Professions Council (2018) | No hours or points on the page; at least two types of learning | A record that is continuous, current and accurate, in any format | At audit: what was learnt, how it improved practice and benefited service users |
| Association of Chartered Certified Accountants (n.d.-b; n.d.-a) | 40 units a year, one unit an hour, 21 verifiable | Evidence the activity was undertaken, kept three years (longer for some) | Why it was chosen, what was learned, how it was or will be applied |
| General Dental Council (n.d.), on a provider's certificate | Total hours, stated on the certificate | Subject, content, aims and objectives, expected development outcomes, date, participant's name, who quality assured it | The provider confirms the details are complete and accurate |
What this means for a body that provides CPD
After each session your members need three things from you: dated proof that they took part, the hours, and something true to write under what they learned. The quiz supplies the third.
The checking falls on the record. Karas et al. (2020) reviewed the CPD schemes of the nine UK health and care regulators and found that they verify participation chiefly by sampling the records practitioners keep. In all, 81% of registrants must reflect on their learning in some form. The General Dental Council (n.d.) adds that it does not quality assure CPD itself: a certificate rests on the provider's own checks.
Does CPD change practice, and what does a quiz at the end add?
On average a session changes practice a little, a test shows what a member knows and not what they do, and a quiz with feedback can help members keep what the session taught. Nearly all of this research is on doctors.
Attendance, knowledge and practice
The reviews of courses and practice are in our guide for members, linked above: the average effect on what professionals do is small.
One old trial shows why a quiz score is not practice. Sibley et al. (1982) randomly allocated 16 Ontario family doctors to receive education packages or not, then judged more than 4,500 episodes of care against preset criteria. Tests confirmed that the doctors had learned, yet the overall quality of care barely moved.
Tests after CPD
Three trials tried quizzes after real CPD events, and they do not agree.
- Larsen et al. (2015) followed 35 people from four courses at a neurology meeting. Each point taught was assigned to repeated short-answer quizzing, repeated study or nothing more. Five and a half months later they scored 55%, 46% and 44% on the three, from 36% before the course.
- Feldman et al. (2018) randomised 186 doctors at a conference; 126 gave complete data. Those tested before their workshop and again, with feedback, two weeks after it did better four weeks after the conference (Hedges g 0.46, a moderate effect).
- McConnell et al. (2018) found no difference. Of 75 doctors randomised at a three-day course, 49 finished: 26 had taken two multiple-choice quizzes with no feedback and 23 had read a handout. Eight weeks on they scored 75% and 77%.
The authors of the null trial point to its missing feedback. Our reading, not a finding: show every answer with its explanation.
The place of reflection
A written intention goes with reported change. Domino et al. (2011) asked 80 clinicians to choose up to three commitments straight after a lecture, and 64 controls to write theirs a week later. At 30 days, 58% of the first group and 22% of the controls reported a change in practice. The change was self-reported: a reason to ask, not proof.
We read Karas and McConnell in full, and Sibley, Larsen, Feldman and Domino as abstracts only.
How do you write a five-to-eight question check from the session itself?
Write one question for each thing a member should be able to do after the session, five to eight in all, then one reflection that is not marked.
- Start from the session's outcomes. Three or four outcomes give six or seven questions.
- Draft from the session's own text. Paste the slides' text, the handout or the transcript into the quiz maker on EdProcto's home page. Procto AI drafts the questions, and you read each one against the source.
- Ask for decisions. Make at least half the questions a situation from members' work. Multiple choice carries most of them, and writing multiple-choice questions that measure covers the wrong options.
- Use the mistake the speaker warned about as a wrong option.
- Write an explanation for every answer. It is the feedback the trials turn on.
- Add the reflection last. Add a short answer question such as "What will you do differently, and by when?" and switch on "Not marked: a reflection, not a question". It then carries no points and holds up neither the score nor a certificate. EdProcto's exit ticket is the same structure ready made.
- Have the speaker sit it first, so a wrong key is corrected before any member's marks are made.
Five to eight is our working convention, not a research finding: enough to touch each outcome once or twice, short enough for the last ten minutes. See how many questions a quiz needs.
What does a pass mark mean on a CPD quiz?
On six questions a pass mark is a threshold for credit under your own rule, not a standard of competence. Our own arithmetic shows why, treating each question as an independent draw.
| A member who | Reaches 4 of 6 | Reaches 5 of 6 |
|---|---|---|
| Would answer half of all such questions correctly | 34 in 100 | 11 in 100 |
| Would answer 70% of them correctly | 74 in 100 | 42 in 100 |
A threshold of four turns away a quarter of the members who know 70% of the material and lets through a third of those who know half.
- Set it before the session, in whole questions. Four of six shows as 66.7%, so type 66 in "Pass mark": 67 would fail it.
- Allow one retake. "Whole-quiz retakes" at 1 and "Hours between retakes" at 24 give a second sitting the next day. It repeats questions whose answers the member has seen, so read it as study finished. Both sittings stay in the record.
- Say what a pass earns. Your scheme decides whether the hour is credited for attending or for passing.
Never describe the score as a grade, or the pass as proof of competence. For an examination that decides something, see setting a pass mark you can defend.
What should the record hold for an audit, and for how long?
Hold six things for every sitting: who sat, which activity, the date, the questions, the score and the member's reflection. Keep them for the period your own retention policy sets.
| The record holds | Where it comes from |
|---|---|
| Who | "Name" and "Email" in EdProcto's marks file, as the member typed them at the start |
| Which activity | The quiz's title, which names the marks file: put the activity, its date and its hours in it |
| The date | "Handed in (your time)", to the minute, in your own time zone |
| The questions | "Paper" in the builder: the quiz, with its answer key on a page of its own |
| The score | "Score %", "Points" and "Out of", then the points on each marked question |
| The reflection | The sitting's own page, opened from the results page: copy it into your register |
Add what only you know: the hours credited, the learning outcomes and how you confirmed who attended.
How long to keep it
How long is your decision, and names and email addresses are personal data. Two public sources frame it.
- The members' side. The accountants' body asks members to keep CPD records for three years (Association of Chartered Certified Accountants, n.d.-a), and the Health and Care Professions Council (2018) looks at the previous two years at audit.
- The data side. The Information Commissioner's Office (n.d.), the UK's data protection regulator, says personal data must not be kept longer than needed, that the UK GDPR fixes no particular periods, and that an organisation should justify its own and set them in a policy.
Other countries have their own rules; this guide is not legal advice.
In EdProcto a quiz and its sittings stay under your account for as long as you keep the quiz. "Close the quiz" ends new sittings and keeps the record; "Delete this quiz" removes the quiz and every submission for good. So press "Download the marks" after each activity, file it with your register, and delete the quiz when your retention period ends.
What does this look like for one webinar?
For one webinar it comes to six questions and a reflection, one link in the chat and a register backed by three files. The example is invented: an association of bookkeepers and a one-hour evening webinar on reading a cash flow forecast.
The quiz. Six questions, one point each, from the speaker's slides: two on terms, three situations from a client's books and one calculation.
A client opens the month with £12,000 in the bank and expects receipts of £30,000 and payments of £37,500. What is the closing balance?
A. £4,500. B. £19,500. C. An overdraft of £7,500. D. £79,500.
A is right. B swaps receipts and payments, C is the month's movement alone, D adds all three. The seventh item is the unmarked reflection: "What will you change in your next forecast, and by when?"
The settings. The title reads "Reading a cash flow forecast: webinar, 6 October 2026, 1 hour CPD", then the association's name. "Pass mark" 66, "Whole-quiz retakes" 1, "Hours between retakes" 24, and "Show the correct answer" and "Show the explanation" at "After submitting". "A certificate on passing" is on, which puts the example on the Educator plan.
The evening. As the session ends the host presses "Publish & share" and puts the quiz's link in the chat. The webinar platform lists 84 members present; 76 hand in a sitting and 68 reach four of six. Five of the other eight sit again a day later and four pass.
The record. "Download the marks" gives 81 rows: 76 first sittings and five retakes. One row reads: the name and the address as typed, "Handed in (your time)" 2026-10-06 19:12, "Score %" 83.3, "Points" 5, "Out of" 6, "To grade" 0, "Minutes" 6.4, "Security events" 0, then 1, 1, 1, 0, 1, 1 under "Q1 (1)" to "Q6 (1)".
What is filed. The marks file, the paper with its answer key, the platform's attendance list and a register with a line for each member: 72 passed, each with a certificate code; four sat and did not pass; eight attended without sitting. Under this association's rule all 84 are credited the hour and the 72 are marked as assessed.
Each of the 72 opens a certificate showing the name as typed, the title, the score against the pass mark of 66%, the date, "6 questions" and a code: see quiz certificates that can be checked by code. The association's own CPD statement carries the hours and the outcomes.
The record shows that these members answered these six questions that evening. Whether their forecasts improve is for a refresher test six months on.
How does EdProcto run a quiz for each CPD activity?
EdProcto is the best tool for a CPD quiz when the questions must come from the session itself and every sitting must leave a record. In Larsen et al. (2015), quizzing on a course's own points kept almost twice as much of what the course had added; EdProcto writes its questions from the session's own text and shows each member the answers and explanations at hand-in.
- One link for the activity. "Publish & share" opens the quiz at one link and one code, shared by every member, on any device.
- A name and an email address. With "An account is needed to take it" on, each member types both. Ask for the address your membership register holds.
- Every sitting listed on the results page with its score and time; "Download the marks" saves them as a spreadsheet file.
- A member across activities. The Students page puts each person's sittings on one line, by the email address typed.
A CPD check needs little security. The integrity page lists each measure, and no setting makes cheating impossible.
On Free a quiz takes thirty members, with ten quizzes a month. Tutor takes a hundred, with no monthly limit. Educator takes five hundred and adds certificates and the full analysis. See pricing.
EdProcto produces the record. Whether that record meets your regulator's rules, or your own scheme's, is for the body to check.
Quick questions
Do regulators require a quiz at the end of a CPD activity?
None of the three bodies' pages we read asks for one. They ask for a dated record, evidence and, mostly, what was learned or applied. Check your own regulator's rules.
How many questions should a CPD quiz have?
Five to eight, written from the session's outcomes, with one unmarked reflection. That is our convention, not a research finding.
Does passing a CPD quiz show that a member is competent?
No. It shows that the member answered most of the session's questions correctly on the day: knowledge, not practice.
Can EdProcto run a quiz after every webinar for our members?
Yes. Each activity gets its own quiz, link and marks file, and a certificate on passing on the Educator plan. See EdProcto's page for associations and boards.
References
- Association of Chartered Certified Accountants. (n.d.-a). Managing your CPD. https://www.accaglobal.com/gb/en/member/cpd/managing-your-cpd.html
- Association of Chartered Certified Accountants. (n.d.-b). Your guide to CPD. https://www.accaglobal.com/gb/en/member/cpd/your-guide-to-cpd/cpd-what-you-need-to-do.html
- Domino, F. J., Chopra, S., Seligman, M., Sullivan, K., & Quirk, M. E. (2011). The impact on medical practice of commitments to change following CME lectures: A randomized controlled trial. Medical Teacher, 33(9), e495–e500. https://doi.org/10.3109/0142159X.2011.599452
- Feldman, M., Fernando, O., Wan, M., Martimianakis, M. A., & Kulasegaram, K. (2018). Testing test-enhanced continuing medical education: A randomized controlled trial. Academic Medicine, 93(11S), S30–S36. https://doi.org/10.1097/ACM.0000000000002377
- General Dental Council. (n.d.). Information for CPD providers. https://www.gdc-uk.org/education-cpd/cpd/information-for-cpd-providers
- Health and Care Professions Council. (2018, June 14). Standards of continuing professional development. https://www.hcpc-uk.org/standards/standards-of-continuing-professional-development/
- Information Commissioner's Office. (n.d.). Principle (e): Storage limitation. https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/data-protection-principles/a-guide-to-the-data-protection-principles/storage-limitation/
- Karas, M., Sheen, N. J. L., North, R. V., Ryan, B., & Bullock, A. (2020). Continuing professional development requirements for UK health professionals: A scoping review. BMJ Open, 10(3), Article e032781. https://doi.org/10.1136/bmjopen-2019-032781
- Larsen, D. P., Butler, A. C., Aung, W. Y., Corboy, J. R., Friedman, D. I., & Sperling, M. R. (2015). The effects of test-enhanced learning on long-term retention in AAN annual meeting courses. Neurology, 84(7), 748–754. https://doi.org/10.1212/WNL.0000000000001264
- McConnell, M., Hou, C., Panju, M., Panju, A., & Azzam, K. (2018). Does testing enhance learning in continuing medical education? Canadian Medical Education Journal, 9(3), e83–e88. https://doi.org/10.36834/cmej.42236
- Sibley, J. C., Sackett, D. L., Neufeld, V., Gerrard, B., Rudnick, K. V., & Fraser, W. (1982). A randomized trial of continuing medical education. New England Journal of Medicine, 306(9), 511–515. https://doi.org/10.1056/NEJM198203043060904