True or false is the fastest question to write, to answer and to mark, and it has the worst reputation of any shape. Both facts come from the same place: there are two options, so a student who knows nothing is right half the time.
That weakness is real, but it is arithmetic, and arithmetic can be planned for. This guide covers what the research says the shape does well, the number of items it takes for luck to stop mattering, the rules for writing a statement, and the marking that separates a guess from knowledge. EdProcto carries the shape as its true or false template, along with the two repairs for guessing described below.
Are true or false questions any good?
True or false questions are good at one thing: covering a lot of content in very little time. A student answers them faster than any other shape, so a quiz can touch more ideas in the same sitting. They are poor wherever the truth is a matter of degree.
The objections are old. Ebel (1970) listed them as teachers put them then: the items are said to be trivial, to encourage rote learning, to be ambiguous and, worst, to be passed by blind guessing. His reply was that these charges tend to exaggerate the defects, and that most of them describe badly written items, not the shape.
The textbooks do not settle it. Frisbie and Becker (1991) read seventeen educational measurement textbooks for their advice on true or false tests and found no general agreement on the advantages, the disadvantages or the rules, and concluded that many of the disadvantages the textbooks listed were unfounded. The standard taxonomy of item-writing rules still lists true or false among the formats it recommends for classroom use (Haladyna et al., 2002).
What the evidence does support is narrower. Frisbie (1973) converted multiple-choice questions into true or false statements and gave both forms to 1,018 high school students, three true or false items for every two multiple-choice items attempted. The true or false tests were significantly less reliable, but they tended to measure the same thing. So the shape measures the right knowledge, less precisely, question for question.
How much is a guess worth on a true or false quiz?
A guess is worth half a mark on average, and on a short quiz that is enough to pass. The chance that a student who guesses every item reaches a given score depends almost entirely on how many items there are.
| Items on the quiz | Chance of 70% or more by guessing every item |
|---|---|
| 10 true or false | about 17 in 100 |
| 20 true or false | about 6 in 100 |
| 40 true or false | fewer than 1 in 100 |
| 10 multiple choice, three options | about 2 in 100 |
| 10 multiple choice, four options | fewer than 1 in 100 |
These are binomial probabilities for pure guessing, worked out for this guide. The same sums explain Frisbie's result: each true or false item carries less information than a multiple-choice item, so it takes more of them to reach the same confidence in the score.
Burton (2001) worked through this for both shapes and showed how the unreliability from guessing, and from the luck of which questions happen to be asked, shrinks as the test gets longer and as the number of options rises. His conclusion was that many such tests, as usually run, are too unreliable for the decisions made from them. He was not arguing for dropping the shape. He was arguing for knowing the size of the error.
The student who knows half
Pure guessers are rare. The common case is a student who knows half the material and guesses the rest. On twenty true or false items that student expects 10 marks from knowledge and 5 from luck: 75%, a comfortable pass, for someone who holds half the content. A raw true or false score flatters everyone, and it flatters the weakest most.
How do you write a good true or false statement?
Write one idea per statement, make it entirely true or entirely false, and make the false version say what a confused student actually believes. The rules below follow from what the shape is, and where the standard taxonomy speaks to one, it is named.
- One idea only. "Evaporation happens at any temperature and is faster in still air" is two claims, one true and one false. A student cannot answer it honestly.
- No absolutes as bait. Statements with always, never, all and none are usually false, and students learn that quickly. The taxonomy lists these words among the clues that give answers away.
- Test the idea, not the sentence. A line lifted from the textbook with one word changed rewards memory of the page. Paraphrase.
- Write false statements from real misconceptions. "Heavier objects fall faster in a vacuum" is worth asking. "The Moon is made of iron filings" is not.
- Keep true and false statements the same length. True statements tend to grow qualifiers, and length becomes a clue.
- Balance the key, roughly. About half true and half false, in no pattern.
- Avoid negatives. "It is not true that plants do not respire" tests logic, not biology.
An example, fixed
Weak: "Water always boils at 100 degrees." The word always marks it false to a student who knows nothing about pressure.
Better: "At the top of a high mountain, water boils at a lower temperature than at sea level." It is one idea, has no bait, and a student has to know the relationship to answer it.
Can a false statement teach the wrong thing?
It can, if the student never sees the correction. Toppino and Brochin (1989) had 64 college students read passages, take a true or false exam, and a week later rate how true each of a new set of statements was. Statements they had seen on the first exam were rated as more true than new ones, whether the statements were in fact true or false. Reading a false statement once had made it feel familiar, and familiar felt true. The authors called it a negative suggestion effect, a name that goes back to work in the 1920s.
The remedy is feedback. For multiple choice, where the same risk comes from reading wrong options, Butler and Roediger (2008) found that showing the right answers, immediately or after a delay, raised later correct recall and reduced the number of wrong options students later gave as answers. That study did not test true or false items, so the transfer is a reasonable inference, not a finding. It is also cheap: show the corrected statement after the quiz, with one line saying why.
How should you mark true or false questions?
Mark them so that a guess earns nothing, or so that the student can say when they are guessing. There are four workable ways, and they suit different quizzes.
| Method | What it does | Use it when |
|---|---|---|
| More items | Luck averages out with length | The quiz is low stakes and short items are the point |
| Negative marking | A wrong answer costs a mark, so blind guessing averages zero | The score will be used for a decision |
| An unsure option | The student can say they do not know, for zero where wrong answers cost nothing | You want an honest map of the class |
| A confidence rating | Sure and right earns more, sure and wrong costs | You want to find misconceptions, not just gaps |
Negative marking on a two-option question means one mark off for a wrong answer, because there is one wrong option. Tell students before the quiz, and tell them that leaving a blank is safe. Some students leave blanks more readily than others, so a negatively marked score measures a little caution along with the knowledge. That is a known limit of the method.
The confidence rating is the most informative of the four. A wrong answer given with certainty is a misconception. A right answer given with doubt is luck or fragile knowledge. Neither shows up in a plain score.
How does EdProcto run a true or false quiz?
EdProcto is the best place to run true or false when the score has to mean something, because both repairs for the fifty-fifty guess are built into the shape and neither needs a workaround.
The true or false template takes a statement and a key and is marked by rule the moment the student answers. Negative marking is a setting on the quiz: a wrong answer costs one mark divided by the number of wrong options, which for true or false is the full mark, so blind guessing averages zero. For a confidence rating, set the statement in the multiple choice with confidence shape with True and False as its two options and switch the confidence bonus on: the student says how sure they are, a sure and right answer earns more, and a sure and wrong answer loses more. Where you would sooner let a student say they do not know, the yes, no or maybe shape gives the third option, which scores zero as long as negative marking is off.
The statements can come from your own material. Paste a chapter into the quiz maker on the home page, choose true or false, and Procto AI writes statements from that text for you to edit, including false ones built on the confusions the text itself warns about. Nothing reaches a student until you have read it.
After the quiz, the feedback settings decide what students see and when: the correct answer and your explanation at once, after submitting, or not at all. The results show each statement's difficulty, and the full analysis on the Educator plan shows how well it separated stronger students from weaker ones, which is how a statement that everyone guessed on gives itself away.
True or false is quick to try on the Free plan, which carries the core templates, marking by rule and the results question by question. The Tutor plan opens all 111 templates. The pricing page lists what each plan carries.
When should you use a different shape?
Use a different shape when the idea has more than two possible states, or when you need the student to produce the answer.
- More than two plausible answers: write a multiple-choice question, where each wrong option names a different misconception.
- A term the student should recall unaided: a fill in the blank has no options to guess between.
- A judgement that needs a reason: follow the statement with a short answer that asks why.
True or false earns its place as the quick pulse at the start or end of a lesson, as a pre-test of what a class already believes, and as the check on a list of facts that are simply right or wrong, such as safety rules.
Quick questions
How many true or false questions should a quiz have?
At least twenty if the score matters, and forty is better. On ten items a student who guesses everything reaches 70% about 17 times in 100. On forty items it is fewer than 1 in 100.
Should I use negative marking on true or false questions?
Use it when the score feeds a decision, and say so before the quiz starts. One mark off for a wrong answer makes blind guessing worth zero on average. EdProcto applies it as a setting on the quiz, and a blank costs nothing.
Should there be the same number of true and false statements?
Roughly the same, in no pattern. If most statements are true, a student who answers true throughout scores well without reading. Keep the true and false statements the same length too.
Is true or false easier than multiple choice?
Each item is easier to get right by chance, and Frisbie (1973) found true or false tests less reliable than multiple-choice tests on the same content. They measured the same knowledge, so the fix is more items or better marking, not a different subject.
Can students cheat more easily on true or false?
A one-letter answer is easy to pass along. EdProcto can shuffle the order of the questions for each student and keeps the answer key on the server, and the integrity settings add what the stakes need. Each student seeing a different order removes most of the value of a whispered answer.
References
- Burton, R. F. (2001). Quantifying the effects of chance in multiple choice and true/false tests: Question selection and guessing of answers. Assessment & Evaluation in Higher Education, 26(1), 41–50. https://doi.org/10.1080/02602930020022273
- Butler, A. C., & Roediger, H. L. (2008). Feedback enhances the positive effects and reduces the negative effects of multiple-choice testing. Memory & Cognition, 36(3), 604–616. https://doi.org/10.3758/MC.36.3.604
- Ebel, R. L. (1970). The case for true-false test items. The School Review, 78(3), 373–389. https://doi.org/10.1086/442915
- Frisbie, D. A. (1973). Multiple choice versus true-false: A comparison of reliabilities and concurrent validities. Journal of Educational Measurement, 10(4), 297–304. https://doi.org/10.1111/j.1745-3984.1973.tb00807.x
- Frisbie, D. A., & Becker, D. F. (1991). An analysis of textbook advice about true-false tests. Applied Measurement in Education, 4(1), 67–83. https://doi.org/10.1207/s15324818ame0401_6
- Haladyna, T. M., Downing, S. M., & Rodriguez, M. C. (2002). A review of multiple-choice item-writing guidelines for classroom assessment. Applied Measurement in Education, 15(3), 309–333. https://doi.org/10.1207/S15324818AME1503_5
- Toppino, T. C., & Brochin, H. A. (1989). Learning from tests: The case of true-false examinations. The Journal of Educational Research, 83(2), 119–124. https://doi.org/10.1080/00220671.1989.10885940