A good question in a seminar is short, open and tied to a line of the reading, and it gives the room something to work on: what a term means, what holds a claim up, what the author takes for granted, or where a number came from. Ask one thing at a time. An MA module at the University of Warwick (2025) wants seminar questions that are "detailed and focussed", concise, and lead beyond a yes or a no.
EdCitation publishes this guide. Every page and study in it was read on 2 October 2026, and the examples marked "ours" are our own wording. One kind of question needs evidence in hand: "where does that figure come from?" EdCitation's free Find sources takes the sentence the figure sits in and searches the published record for it; Cite a source then turns the paper into a reference. EdCitation never writes any part of anyone's work: the question is yours, which is what makes it worth asking.
If speaking at all is the hard part, start with how to speak up in seminars when you are nervous; for one-to-one help, see how to use office hours. This guide is about the question itself.
What makes a good question in a seminar?
A good seminar question is specific, open and anchored in the text, so several people can answer it from what they read. Warwick asks students to start from what genuinely interests them in the reading. The University of Pittsburgh's teaching centre, writing for lecturers, says open questions with many credible answers draw in more students than questions with one right answer, and "how" and "why" get a livelier response than "what" and "which" (Wright and Onufer, n.d.). That holds from the student's chair too.
Where do real questions come from? Graesser and Person (1994), who recorded tutoring sessions, found that students' questions about gaps in their own understanding (29% of their questions) were set off by a contradiction, a fact that did not fit, or an unfamiliar word. Mark those as you read.
Closed and open questions
A closed question can be answered with yes, no or a single word; an open one needs an explanation. The Right Question Institute (n.d.), whose Question Formulation Technique is used in classrooms, has people sort their questions into the two kinds and practise turning one into the other. An example, ours: "Was the survey only of British students?" becomes "How might the results change with students from outside Britain?" Closed questions are not wrong; sometimes a fact is what the room needs.
What types of question can you ask in a seminar?
Three published schemes help: Bloom's taxonomy, Richard Paul's Socratic question types, and Graesser and Person's shallow and deep questions. None was designed as a seminar script.
Bloom's taxonomy as question levels
Bloom's taxonomy was built to classify learning objectives and exam items. Krathwohl (2002), one of its original authors, explains that the 1956 version had six categories (Knowledge, Comprehension, Application, Analysis, Synthesis, Evaluation) and the 2001 revision renamed them as verbs (Remember, Understand, Apply, Analyze, Evaluate, Create), with Create replacing Synthesis at the top. He adds that analyses of courses and tests kept finding a heavy emphasis on recall.
Teaching pages use both versions. The University of Waterloo Centre for Teaching Excellence (n.d.) gives question stems under the 1956 names, up to "What criteria would you use to assess …?" for evaluation. The label matters less than the direction: one question above recall is worth more than five that check who did the reading.
Socratic questions: Paul's six types
The six types usually called Socratic questions come from Richard Paul. Kwantlen Polytechnic University's learning centres (n.d.) list them with stems: clarification, assumptions, reasons and evidence, viewpoints and perspectives, implications and consequences, and questions about the question. Paul and Elder (2006) add four directions in which any belief can be followed: back to its origin or source, down to its reasons and evidence, forward to its implications, and outward to opposing views. An assumption question often targets what how to argue a position in writing calls the warrant.
The name carries baggage. Wright and Onufer argue that "the Socratic method" is a loose label, often tied to hostile cross-examination in law schools, and that questioning works as shared discovery, not attack. Kwantlen's learning aid itself calls Paul's six "deep" questions. In a seminar you ask alongside classmates; you do not cross-examine them.
Shallow and deep questions
Graesser and Person sorted questions into 18 categories by the information sought, six of them deep-reasoning questions: why, why not, how and what if, about causes, consequences, goals, procedures, enabling conditions and expectations. Rated against Bloom's levels, 70% of students' questions sat at the lowest level and 1% at the highest. In the second half of the course, students who asked a larger share of deep questions scored higher on the exam (a correlation of .47); the sheer number of questions did not. The limits: 27 undergraduates on one psychology research methods course, one-to-one tutoring rather than seminars, and correlation, not cause.
Good seminar questions and Socratic questions: examples by type
The examples are ours, written for an imagined reading on remote work; each type and its purpose come from the source named.
| Type | What it does | Example (ours) | Source |
|---|---|---|---|
| Clarification | Pins down a term the argument depends on | "On page 4, does 'productivity' mean output per hour or per worker?" | Paul, via Kwantlen |
| Assumption | Brings an unstated premise into view | "The argument assumes people can choose where they work. Does it hold for those who cannot?" | Paul, via Kwantlen |
| Reasons and evidence | Asks what holds a claim up | "Is the claim on page 7 based on one firm or several?" | Paul, via Kwantlen; Waterloo |
| Origin of a figure | Traces a number to the study behind it | "Where does the figure on page 9 come from, and what exactly was measured?" | Paul and Elder (2006); Caulfield (2019); Find sources |
| Viewpoint | Brings in another perspective | "How would a manager and an employee each read this result?" | Paul, via Kwantlen |
| Implication | Follows a claim forward | "If this is right, what should our own university do?" | Paul, via Kwantlen |
| Connection | Links the reading to earlier work | "How does this tie in with last week's reading on office design?" | King's stems, via Rosenshine et al. (1996) |
| Method | Questions how material was chosen or grouped | "How did the authors decide which jobs count as remote-capable?" | Corfield and Hitchcock (2022) |
| Evaluation | Asks for a judgement against criteria | "What criteria would show that the policy worked?" | Waterloo |
| The question itself | Checks the group is asking the right thing | "Are we asking whether remote work works, or for whom?" | Paul, via Kwantlen |
How do you ask about a reading or a source?
Ask where a claim or a number came from, then whether the source says what the reading says it does. Corfield and Hitchcock (2022), two historians writing about research seminars, note that requests for information or clarification, including about a source's provenance and reliability, often make up most questions, and that a simple question is not a lesser one. Their example: a talk claiming Britain imposed strict Anglicanism across its eighteenth-century territories was undone by one listener asking about the Quebec Act of 1774, which allowed Catholic worship in Quebec.
Caulfield (2019) names the habit in his SIFT method: trace a claim, quote or image back to its original context and see whether it was fairly represented. In a seminar, open the reading's reference for the figure, or search the claim when there is none. A question built that way (ours): "The chapter says home workers are more productive and cites one study, whose abstract describes call-centre staff in one firm. Does the chapter stretch it?"
What Find sources returned for one claim
On 2 October 2026 we pasted one claim, word for word, into Find sources: "Teaching students to generate their own questions improves their reading comprehension." The search ran on OpenAlex and reported 102,956 matching works. The first result was a 1983 article on teaching reading comprehension by P. David Pearson and Margaret C. Gallagher. The fourth was Chin and Osborne (2008), a review of research on students' questions, cited 742 times, whose abstract lists the effects of teaching questioning skills among its four areas. The first dozen also held a 2016 dataset paper on machine reading, so read each abstract before relying on a result. The review that tests the claim most directly (Rosenshine et al., 1996) was not in the first 100 results: a claim search is where checking starts, not where it ends. The method is in how to search by a claim, not a topic.
What Cite a source returned
We gave Cite a source the review's DOI, 10.1080/03057260701828101. It read the record at Crossref, found no retraction, and returned this APA 7 reference:
Chin, C., & Osborne, J. (2008). Students’ questions: A potential resource for teaching and learning science. Studies in Science Education, 44(1), 1–39. https://doi.org/10.1080/03057260701828101
It capitalised the word after the colon, as APA 7 asks, and gave the in-text form "(Chin & Osborne, 2008)": the names and year to write on your card and say aloud.
How to ask good questions in class: a step-by-step method
This method is ours, built from the sources above; no study has tested it as a whole.
- Read for three triggers. Mark a contradiction, a surprising fact and a word you could not define (Graesser and Person, 1994).
- Write questions fast. Follow the Question Formulation Technique's rules: ask as many as you can, do not stop to discuss, judge or answer them, write each down exactly, and turn any statement into a question.
- Open two closed questions so they need an explanation.
- Push one up a level with a stem such as "What are the strengths and weaknesses of …?", one of King's generic stems as listed by Rosenshine et al. (1996).
- Check the source behind one claim. Search it in Find sources, read the abstract, and put the DOI into Cite a source for the author and year.
- Choose three. Corfield and Hitchcock suggest a general question to open, a definitional one to go deeper, and a tough critique in reserve.
- Ask it short. Name the page, ask one question, and stop.
- Write down the answer and one question it raised.
Does practice help? Rosenshine et al. (1996) reviewed 26 studies in which students were taught to generate questions about what they read. Comprehension of new material improved: the median effect size was 0.36 on standardised tests and 0.86 on tests the researchers wrote. Generic question stems did best on the researchers' tests (1.12 across four studies), and all five studies with college students found significant gains. The limits: the outcome was reading comprehension, most studies were in schools, and none measured seminar talk. We found no controlled test of the Question Formulation Technique itself in what we read.
How do you ask a question at a conference?
At a conference, ask one short question about the talk and do not make a speech. An editorial in Nature Energy (2019) says the aim is not to show how clever you are, that simple questions of clarification are often the most valuable because others were probably confused too, and it names the unloved "more-of-a-comment-than-a-question". Corfield and Hitchcock warn against the counter-lecture: a newcomer giving a first paper is usually treated gently, while experienced speakers expect a hard question.
A pattern that keeps you short (ours): one sentence to locate the point, one question, then stop. "On the slide with the regional results, how did you handle firms that moved region during the study?" If you freeze, the coffee break counts too; see how to network at a conference.
Why do students ask so few questions in class?
Students ask few questions in class, mostly for social reasons. Graesser and Person summarise earlier classroom studies from several countries: a whole class asked 1.3 to 4 questions an hour (median 3), about 0.11 per student, while teachers asked 96% of the questions. In their tutoring sessions a student asked about 240 times as many. They name three barriers: students find it hard to spot gaps in their own understanding, they fear looking ignorant (social editing), and teachers seldom model good questions; only 4% of teacher questions in earlier studies were high-level. Chin and Osborne (2008) review the wider research on students' questions, including the effects of teaching questioning skills.
Corfield and Hitchcock note that asking is easier once the decision to speak has been made. Our reading: a question written down before the seminar has already passed the social edit.
How EdCitation helps you follow a question to its evidence
EdCitation is the best tool for the evidence step of a seminar question, because it takes every reference from the publisher's record and writes none of them: the Chin and Osborne reference above came from Crossref, screened for retraction, not from a chatbot's memory.
- Find sources, free with no account, covers about 300 million published works. Enter a topic, or paste the sentence you want evidence for, then narrow by kind of source, year range or open access.
- Cite a source, free: works from a DOI, title, PubMed ID, ISBN or web address to build the reference in six styles (APA 7, MLA 9, Harvard, Chicago 18 author-date, Vancouver and IEEE), and screens the source for retraction.
- Verify references, free: marks each entry in a pasted list verified, doubtful or not found, and keeps "could not check" apart from "not found".
- A free account keeps a reference list and a reading list, and shares a list with classmates.
- Pro is $8 a month and Max $24 a month (pricing); Max adds the Library, with more than 5 million books to read or borrow, useful when the set text is a book.
EdCitation never writes any part of anyone's work, so what you say in the seminar is entirely your own.
Quick questions
Which questions work best in a seminar?
Open questions anchored in the text: what a term means, what supports a claim, what the author takes for granted, or where a figure comes from. Contradictions, surprising facts and unfamiliar words in the reading are where real questions start.
What are some Socratic questions examples?
Paul's six types each have a stock form, such as "What could we assume instead?" for assumptions or "What is another way to look at it?" for viewpoints. Kwantlen's learning centres list stems for all six.
How do I ask a question at a conference without making a comment?
Give one sentence of context, ask one question, and stop. A Nature Energy editorial singles out the question that is really a comment.
How many questions should I ask in a seminar?
No rule fits every course. Corfield and Hitchcock suggest preparing at least one possible question for every presentation, even if you do not get to ask it.
Can EdCitation help me check a figure before I ask about it?
Yes. EdCitation's free Find sources looks for the paper behind a figure when you paste the claim it supports, and Cite a source gives the paper's reference from its DOI; the question stays yours.
References
- Caulfield, M. (2019, June 19). SIFT (the four moves). Hapgood. https://hapgood.us/2019/06/19/sift-the-four-moves/
- Chin, C., & Osborne, J. (2008). Students' questions: A potential resource for teaching and learning science. Studies in Science Education, 44(1), 1–39. https://doi.org/10.1080/03057260701828101
- Conference considerations [Editorial]. (2019). Nature Energy, 4(7), 527. https://doi.org/10.1038/s41560-019-0442-y
- Corfield, P. J., & Hitchcock, T. (2022). Asking and answering seminar questions. In Becoming a historian: An informal guide (chap. 14). University of London Press. https://read.uolpress.co.uk/read/becoming-a-historian/section/c0cf218f-d301-47c9-a1d9-d0eafea188c1
- Graesser, A. C., & Person, N. K. (1994). Question asking during tutoring. American Educational Research Journal, 31(1), 104–137. https://doi.org/10.3102/00028312031001104
- Krathwohl, D. R. (2002). A revision of Bloom's taxonomy: An overview. Theory Into Practice, 41(4), 212–218. https://doi.org/10.1207/s15430421tip4104_2
- Kwantlen Polytechnic University Learning Centres. (n.d.). Critical thinking through deep questioning [Learning aid]. https://www.kpu.ca/sites/default/files/Learning%20Centres/Think_Critical_LA.pdf
- Paul, R., & Elder, L. (2006). The thinker's guide to the art of Socratic questioning [Excerpt]. Foundation for Critical Thinking. https://www.criticalthinking.org/files/SocraticQuestioning2006.pdf
- Right Question Institute. (n.d.). What is the QFT? https://rightquestion.org/what-is-the-qft/
- Rosenshine, B., Meister, C., & Chapman, S. (1996). Teaching students to generate questions: A review of the intervention studies. Review of Educational Research, 66(2), 181–221. https://doi.org/10.3102/00346543066002181
- University of Warwick Department of English and Comparative Literary Studies. (2025, October 6). Seminar questions. EN942 Fundamentals of World Literature. https://warwick.ac.uk/fac/arts/english/currentstudents/postgraduate/masters/modules/worldlitworldsystems/seminarquestions
- University of Waterloo Centre for Teaching Excellence. (n.d.). Asking questions: Six types. https://uwaterloo.ca/centre-for-teaching-excellence/catalogs/tip-sheets/asking-questions-six-types
- Wright, J. D., & Onufer, L. (n.d.). Teaching through questioning: "Socratic" no more. University Center for Teaching and Learning, University of Pittsburgh. https://teaching.pitt.edu/resources/teaching-through-questioning-socratic-no-more/