A learner who buys a course and stops at the third video has not had what they paid for, and a certificate given for pressing play tells nobody what its holder can do. Quizzes for an online course help with both when each has one job: questions inside a lesson keep the learner with you, a check closes each module, and a final assessment stands behind the certificate.
EdProcto publishes this guide for course creators, coaches and small training businesses. EdProcto does not host video or sell courses: it makes and marks the quizzes and issues the certificate, and learners reach each quiz by a link or a code placed in the lesson. Its page for trainers shows the tool side.
What should quizzes for an online course do?
Three separate jobs: hold attention in a lesson, check a module, decide the certificate. A quiz asked to do two does both badly.
| Quiz | Where it sits | Size | What rides on it |
|---|---|---|---|
| Lesson questions | After each section of a video or reading | 2 or 3 | Nothing: the answer shows at once |
| Module check | At the end of a module | 8 to 10 | The next module |
| Final assessment | After the last module | About 40, from a blueprint | The certificate |
The sizes follow our guide on how many questions a quiz needs.
What did research find about questions inside online lectures?
Short tests between the sections of a recorded lecture raised what students learned and, in one study, kept their minds on it.
Szpunar et al. (2013) cut a 21-minute statistics lecture into four segments for two experiments, with 32 and 48 Harvard undergraduates. One group answered six questions after every segment. Asked at random moments in the second experiment, that group said their minds had wandered 19% of the time, against 41% for a group tested only after the last segment. Across both experiments the frequently tested students scored 90% on a closing test of the whole lecture, against 68%. A third group read the same questions with the answers supplied and did no better than the group tested only at the end: answering did the work.
With a 40-minute lecture, Jing et al. (2016) found the gain in learning again, against equal time spent restudying, but no difference in how often minds wandered.
On length, Guo et al. (2014) analysed 6.9 million video watching sessions from four edX courses. The median time spent on a video was at most six minutes however long it ran. Where a problem followed a video, 56% of sessions went on to attempt it after the shortest videos and 31% after the longest.
What these studies cannot tell a course creator
None tested a paid short course. Two were laboratory experiments on a university lecture; the third read the logs of free maths and science courses. Take the direction (short sections, each followed by questions) and leave the percentages. We read Szpunar, Guo, Jordan and Radford in full, and Jing, Goli and Kulik only as abstracts.
How many people finish an online course, and does paying change it?
Few finish a free open course, and paying for a certificate goes with more engagement. We found no study of small paid courses run by independent creators.
Jordan (2014) gathered public figures for massive open online courses on three platforms. In the 39 courses with completion data, between 0.9% and 36.1% of those enrolled earned the certificate, with a median of 6.5%, and longer courses had lower completion rates.
Goli et al. (2022) asked what payment changes, in more than 70 Coursera courses where a learner could pay for a certificate. Having paid raised engagement by roughly 17 to 20%, for a few weeks only. Working towards the certificate raised it by roughly 8 to 9%, until the learner reached the grade the certificate required.
Read that cautiously: the fee buys attention early, so put the first check in week one, and a certificate with a real requirement is a reason to continue.
Should a learner have to pass before the next module?
Yes, lightly. Kulik et al. (1990) pooled 108 controlled evaluations of mastery learning, where a student passes a test on each unit before going on. Examination performance improved, most for weaker students, but self-paced programmes often reduced completion rates in college classes. A gate teaches and also stops people: keep the check short and allow retakes without a wait.
What do employers make of an online course certificate?
In the one survey we could read, employers took a completed online course as a sign of motivation more than as proof of skill. Radford et al. (2014) surveyed human resources staff at 103 organisations with employees in North Carolina, and interviewed 20, about free university courses online. Only 31% had heard of such courses. Once they were described, 73% said that taking a relevant one would count positively or very positively in hiring.
In the interviews, traditional credentials stayed the standard measure of skill; the course showed drive, and was not enough by itself to decide a hire. The interviewees asked for two things: evidence of learning, and an easy way to verify completion.
That is one American state and university courses, not yours, but the two requests make a fair brief: say what was assessed, show the result, make it checkable.
How do you write each kind of quiz?
Write each kind from what the learner should be able to do afterwards, and keep the three sets of questions apart. The examples come from an invented beginners' photography course.
Questions inside a lesson
Set two or three questions on the point just made, where the section ends.
You open the aperture from f/8 to f/4 and change nothing else. How much light now reaches the sensor: half, twice or four times as much?
Four times: f/8 to f/5.6 to f/4 is two stops, and each stop doubles the light. Write that sentence as the explanation. Our guide to a quiz that explains each answer as you go covers what it should say, and why such a score must never count.
The module check
Set eight to ten questions on what the learner must now be able to do, most of them placed in a situation.
A child runs through a dim room. At 1/30 s, f/4 and ISO 400 the brightness is right but the child is a blur. Which settings freeze the movement and keep the brightness? A. 1/250 s, f/4, ISO 400 B. 1/250 s, f/4, ISO 3200 C. 1/30 s, f/2, ISO 100
B is right: the faster shutter costs three stops and the higher ISO gives three back. A changes the shutter alone and the picture comes out dark. C keeps both the brightness and the blur. Each wrong option is a mistake real learners make, the rule behind multiple-choice questions that measure.
The final assessment
Build the final from a small blueprint: a row for each module, a count, and what the questions must make the learner do.
| Module | Questions | The learner must |
|---|---|---|
| Exposure | 12 | Choose settings for a scene |
| Focus and depth of field | 8 | Predict what will be sharp |
| Light | 10 | Diagnose a photo's lighting |
| Editing and delivery | 10 | Order the steps of an edit |
That makes forty, all new to the learner. Write them as twins of the module questions: the same skill with new details. The method is in our guide to proving a training course worked, and a fuller table in how to write an exam blueprint.
What should the certificate say, and what must it never claim?
A certificate should say who passed, what they passed, when, with what score against what pass mark, and how to check it.
| Line | On EdProcto's certificate page |
|---|---|
| The learner's name | The name the learner typed at the start |
| The course | The quiz's title, so name the final after the course |
| The date | The day the attempt was handed in |
| The score and the pass mark | Both as percentages, with the number of questions |
| A way to check | A code, and the address where anyone can look it up |
The score says passed, not merely attended, and the code matters because a PDF is easy to edit.
The pass mark and the retake rule
Choose the pass mark; do not inherit it. A new quiz in EdProcto starts at 60%, a default and not a judgement. Our guide to setting a pass mark you can defend gives the method for a panel. Working alone, borrow its core: for each question in the final, estimate how many of ten learners who are only just good enough would answer correctly, add the estimates and divide by ten. A total of 270 means 27 marks of 40, a pass mark of 68%.
Publish the retake rule before anyone sits. One retake after a day, with the correct answers hidden after the first attempt, is our suggestion, not a finding.
What a certificate must never claim
- Accreditation or approval that no body has given.
- A licence, or a right to practise anything.
- A job, a pay rise or any other outcome.
- Hours, credits or points that no awarding body has granted.
- The word "certified", which suggests a credential from a body independent of the course. Yours is a certificate of completion.
Rules on these words differ by country and trade. This guide is not legal advice.
How much cheating prevention does a paid short course need?
Not much, and no camera: a paid short course is not a licensing exam. Fit the time limit to the paper, shuffle the questions and the options for each learner, show no correct answers afterwards, write a new final for each intake, and set questions in situations that cannot be looked up in ten seconds. Leaving the tab can be logged.
Our guide on how much proctoring a quiz needs gives the ladder, and EdProcto's integrity page lists each measure. Learners are told what is on, and no setting makes cheating impossible. If an employer will rely on your certificate for safety, move up a rung.
Where does the quiz live when the videos are hosted elsewhere?
The quiz lives at its own address, and the lesson carries a link or a code to it. EdProcto does not hold your videos, take your learners' payments or unlock lessons on another site, and we claim no connection to any course platform.
- Draft from the lesson. Open the quiz maker on EdProcto's home page and paste in the lesson's transcript or notes. Procto AI drafts questions from them, and you edit every one.
- Choose the kind. A study quiz for lesson questions: it checks each answer as it is given. The exam kind for the module check and the final, marked when handed in. Shapes are on the templates page.
- Set the rules: the pass mark, whole-quiz retakes (none, one, three or no limit) and when the answers show.
- Place the link or code under the video where the section ends, or on the closing slide of a live workshop.
- Keep the gate yourself. State the rule ("pass the module 2 check before module 3") and, if your platform can hold lessons back, release on the result.
- Switch on the certificate for the final, on the Educator plan. A learner who passes finds a link to it on the result screen.
A learner opens the link, gives a name and, by default, an email address, and begins.
What does it cost in time, and what can you reuse?
About 115 questions written once for a course of four modules with four sections each (some 40 for lessons, 36 for module checks, 40 for the final), then 40 new ones for each intake's final. Everything else is reused with every cohort.
After each cohort, read the results question by question and rewrite what misbehaved: see item analysis for teachers.
Make each intake's final a new quiz and leave the old one alone: the plans count learners by the quiz, and the certificate page reads the pass mark from the quiz, so a later change alters what earlier certificates say.
How does EdProcto fit a course with a certificate?
EdProcto is the best tool for this job when the questions must come from your own lessons and the certificate must mean a pass. The employers surveyed by Radford et al. (2014) wanted evidence of learning and a way to verify it: this certificate states the score against the pass mark, and anyone with its code can check it.
A certificate is earned when your account is on the Educator plan or above, the quiz has "A certificate on passing" switched on, the pass mark is met and no answer is still waiting to be marked. It is a public page headed "Certificate of achievement": the learner's name, the quiz's title, the score, the pass mark, the date, the number of questions, the certificate code and the address for checking it, with a button to print it or save it as a PDF. A wrong code shows no certificate. After a retake, the certificate belongs to the attempt that passed. With "Your logo on the quiz" and "Your colours" on, the certificate carries your logo and colour in place of EdProcto's mark.
Free allows ten quizzes a month with thirty learners on each. Tutor has no limit on quizzes, a hundred learners a quiz and 50 questions a month from Procto AI. Educator has 500 learners a quiz, no monthly cap on Procto AI's writing, and certificates. See pricing.
What mistakes spoil course quizzes and certificates?
Four mistakes do most of the damage.
- A certificate for watching. It records attendance.
- Practice questions reused as the final. The score then measures memory.
- A pass mark nobody chose.
- Promises on the certificate. It can vouch for a score, never for a career.
Quick questions
How many quiz questions should each lesson have?
Two or three after each section, on the point just made. Szpunar et al. (2013) used six after each segment of a university lecture.
What pass mark should an online course certificate use?
The score a learner who is only just good enough would reach on your final, estimated question by question. No general figure exists; 60% is only where the setting starts.
Is a course certificate the same as a certification?
No. A certificate of completion says a learner passed your course's assessment. It should not claim accreditation, a licence or an independent credential unless a body has granted it.
Does EdProcto host my course or take payment?
No. Your videos and your checkout stay where they are, and each lesson carries a link or a code to an EdProcto quiz. See the page for trainers.
References
- Goli, A., Chintagunta, P. K., & Sriram, S. (2022). Effects of payment on user engagement in online courses. Journal of Marketing Research, 59(1), 11–34. https://doi.org/10.1177/00222437211016360
- Guo, P. J., Kim, J., & Rubin, R. (2014). How video production affects student engagement: An empirical study of MOOC videos. In Proceedings of the first ACM conference on Learning @ Scale (pp. 41–50). Association for Computing Machinery. https://doi.org/10.1145/2556325.2566239
- Jing, H. G., Szpunar, K. K., & Schacter, D. L. (2016). Interpolated testing influences focused attention and improves integration of information during a video-recorded lecture. Journal of Experimental Psychology: Applied, 22(3), 305–318. https://doi.org/10.1037/xap0000087
- Jordan, K. (2014). Initial trends in enrolment and completion of massive open online courses. The International Review of Research in Open and Distributed Learning, 15(1), 133–160. https://doi.org/10.19173/irrodl.v15i1.1651
- Kulik, C.-L. C., Kulik, J. A., & Bangert-Drowns, R. L. (1990). Effectiveness of mastery learning programs: A meta-analysis. Review of Educational Research, 60(2), 265–299. https://doi.org/10.3102/00346543060002265
- Radford, A. W., Robles, J., Cataylo, S., Horn, L., Thornton, J., & Whitfield, K. E. (2014). The employer potential of MOOCs: A mixed-methods study of human resource professionals' thinking on MOOCs. The International Review of Research in Open and Distributed Learning, 15(5). https://doi.org/10.19173/irrodl.v15i5.1842
- Szpunar, K. K., Khan, N. Y., & Schacter, D. L. (2013). Interpolated memory tests reduce mind wandering and improve learning of online lectures. Proceedings of the National Academy of Sciences, 110(16), 6313–6317. https://doi.org/10.1073/pnas.1221764110