How to study a whole textbook with quizzes, chapter by chapter

A thick book cannot be held in view. A bookmark at page 300 says where your eyes have been, not whether chapter 4, read five weeks ago, is still there.

To study a whole textbook with quizzes, give every chapter a short quiz of its own, take it more than once, and write each score on a map of the book. The map shows which chapters are held, which have faded and which never went in.

EdProcto's page for students is built round that: it splits a book at its headings, makes a study quiz for each chapter and keeps the map.

Why does a whole book need a map and not a bookmark?

A whole book needs a map because readers judge their own understanding of a text badly, and because chapters behind the bookmark fade as it moves on.

Glenberg et al. (1982) called the first problem the illusion of knowing. Their readers hunted for contradictions planted between neighbouring sentences. An illusion was counted when a reader missed one and still rated their comprehension as high, and it was more frequent when the clash closed a three-paragraph text than a single paragraph.

Yang et al. (2023) measured the gap across 115 studies and 15,889 participants. With nothing done to help, the weighted mean correlation showing how well people tell the texts they learned well from the ones they learned poorly was .178, where 1 would be perfect.

The feeling matters because it steers the studying. Thiede et al. (2003) asked readers to write five keywords for each text. Those who wrote them after a delay judged their own comprehension more accurately than those who wrote them at once or wrote none, chose what to restudy more effectively, and scored higher on the tests that followed. A quiz after a gap does the same work.

How fast does an unrehearsed chapter fade?

Quickly at first, though we found no study that timed a textbook chapter week by week. Rawson and Kintsch (2005) had 423 college students read a text once, twice in a row, or twice a week apart. Tested at once, two readings in a row beat one. Two days later they no longer did, while readings a week apart did.

Over years, a review by Custers (2010) of studies in general and medical education found roughly two thirds to three quarters of the knowledge left after one year and slightly under half after the next. So the map needs dates as well as scores.

Does a quiz after each chapter beat reading it again?

Yes, in the experiments we read, one of them on a real textbook chapter.

Uner and Roediger (2018) gave students a 40-page chapter of a biology textbook. Some questions, each followed by its answer, came after a section, some after the whole chapter, some at both points. One comparison group read those answers without being asked, and another read the chapter once. Two days later, answering had produced more recall than either. One round of questions did equally well in either place, and two rounds did best.

McDaniel et al. (2009) tested a routine a reader can run alone: read, recite, review. In two experiments it gave better free recall of fact-based passages than re-reading or note-taking, at once and a week later, and on longer engineering passages it beat re-reading on multiple-choice and problem-solving questions.

Should old chapters be mixed into each sitting?

Yes: bring earlier chapters back at widening gaps, out of book order.

Rawson and Dunlosky (2011) had 533 students learn concepts by recalling them and tested them one to four months later. Relearning on later days had a pronounced effect on what lasted, at little cost in extra practice. They advise three correct recalls at first learning, then three rounds of relearning set far apart, hence the four sittings a chapter in the timetable below.

Lindsey et al. (2014) built review software into a semester-long foreign-language course in a middle school and compared strategies that took equal time. On a cumulative exam after the semester, review personalised to each student gave a 16.5% boost in retention over massed study, which the authors call current practice, and 10.0% over spacing on one schedule for everybody.

Mixing has its best evidence in mathematics. Rohrer et al. (2015) gave 126 seventh-grade students the same practice problems over three months, blocked by skill or interleaved. An unannounced test a day after the closing review favoured interleaving (Cohen's d of 0.42), and one 30 days after favoured it more (0.79). Interleaving chapters of a prose textbook is our extension; we read no trial of it.

How do you study a whole textbook with quizzes, chapter by chapter?

Run one loop on every chapter and let the map decide what comes back.

1. Split the book and make a quiz for each chapter. EdProcto's Books page does both; a single chapter can also be pasted into the quiz maker on the home page.
2. Read a chapter, then quiz it the same day, book shut, before opening the next. Check each answer and retry the misses.
3. Mark the map: the date, the score, the ideas missed.
4. Bring each chapter back twice, about one week and about three weeks later, when newer chapters lie on top of it.
5. Mix the old ones. A sitting is the newest chapter's quiz, then one or two old chapters out of book order.
6. Believe the later score. A same-day score includes whatever is merely fresh.

What goes on the map

One line a chapter, four facts: the date it was last quizzed, that day's score, the best score so far and a verdict. Our rule of thumb: held when two sittings at least a week apart both reach 80%, fading when the latest score is below the one before. The 80% line is EdProcto's, a working threshold and not a research finding.

What does a 20-chapter book over ten weeks look like?

It comes to 75 chapter quizzes in 32 sittings and about 15 hours of quizzing, on the assumptions below.

Twenty chapters, eight questions a chapter, the exam at the end of week 10. Read in weeks 1 to 8, three chapters one week and two the next. Quiz each chapter the day it is read, a week on, three weeks on, and once more in the final fortnight, with no more than three quizzes a sitting. We allow 12 minutes a quiz, a minute and a half a question for the answer, the explanation and the retry. That allowance is ours, not a measured figure, so time your first week and correct it. Reading time is extra.

Weeks 9 and 10 are the last pass: chapters 1 to 10 in week 9 and 11 to 20 in week 10, plus the four sittings still owed to chapters 14, 15, 19 and 20. Chapters 16 to 20 get three sittings, not four, because the calendar runs out.

Where should the last fortnight go?

Where the map points. Sort the chapters by latest score, lowest first. A chapter under 80% on its last pass gets one more sitting two or three days later, once you have re-read the passages behind its misses. Chapters held twice are left alone.

What do you do with a chapter that will not stick?

Change the treatment: a chapter still under half at its third sitting needs more than the same quiz again.

• Look behind it. A chapter often fails because an earlier one it leans on has faded. Quiz that one first.
• Recite the misses. Re-read only the passages behind them, then shut the book and recite each idea, the middle step of the routine McDaniel et al. (2009) tested.
• Change the questions. If you remember which option was right, you know the quiz and not the chapter. Swap in questions that make you write the answer. Our guide to the questions you got wrong goes further.

What can you skip?

Chapters the syllabus leaves out, appendices, and re-reading what the map shows as held. Never the same-day quiz or the old chapters, the first things dropped in a busy week.

How does EdProcto turn a book into chapter quizzes and a map?

EdProcto is the best tool for this method when the questions must come from your own book, because it does the three chores that stop most people. It splits the book, writes each chapter's quiz from that chapter's text alone, and keeps the scores in one table. In Uner and Roediger (2018), answering questions on a textbook chapter beat reading the same answers.

Its Books page does exactly this:

• Reads a PDF or a Word file (.docx) of up to 15 MB, a plain text file, or pasted text.
• Finds chapters by rule, with no AI, at heading lines: a line beginning Chapter, Unit, Part, Section, Lesson, Module or Topic with a number or Roman numeral; a numbered heading such as "3.1 Photosynthesis"; or a short line standing alone without a full stop. A piece under 250 words joins the one before, the list stops at 60 pieces, and a text with no headings comes back as one piece, "The whole text".
• Lets you choose which pieces become quizzes, not where they break. Each is listed, ticked, with its word count and most-used terms. Sub-headings can count as headings, so check the list against the contents page; if it is wrong, paste the text with only the chapter titles on lines of their own.
• Makes a study quiz for each ticked piece. Where Procto AI's writing is available, that is eight multiple-choice questions written from the piece alone, each with an explanation, fewer if one comes back without exactly one right option. Where it is not, the quiz keeps the chapter's text and no questions, and the map offers Write the questions, which opens the builder. The first 120,000 characters of a piece are used.
• Keeps the map, one table a book. Each chapter shows its number of questions and its best score as a word: not yet (nothing handed in), go back (under 50%), getting there (50% to 79%) or known (80% and over). Above it: how many chapters are tested and the average of their best scores. Best is the highest percentage among the attempts you have handed in.

Each quiz runs on the study screen in our guide to study mode, and the dashboard lists it as due two days after an attempt with misses and a week after a clean one.

What the Books page does not do

• Read a scanned book. A PDF with no text layer returns a notice that no text could be read; have the text recognised first, as in our guide to a quiz from a PDF.
• Mix chapters into one quiz. Interleave by taking two or three chapters' quizzes in a sitting.
• Show dates or latest scores. A best score never falls, so a chapter marked known in week 2 still reads known in week 9. Fading shows only on your own map.
• Lengthen the gaps. The due list knows only two days and a week.

The plans list a whole book with its map under Learner, with self-tests without limit and 100 questions a month from Procto AI. At eight a chapter that is twelve chapters a month (96 questions), so a twenty-chapter book (160) needs two months' allowance: chapters 1 to 12 first, then 13 to 20 the month after, which the map lists as a second book. Free carries ten quizzes and 20 written questions a month, two chapters at eight each, enough to try the loop on one chapter. See the pricing page, and the kits for a certification or entrance exam.

Upload only a copy you are entitled to use, such as your own notes, an openly licensed textbook or a file your course provides for study; this is not legal advice.

What are the limits of studying a book by chapter quizzes?

The limits are sampling, shape and subject.

• A quiz samples a chapter. Eight questions cannot reach every idea: known means those eight were answered.
• Choosing is not producing. If the exam wants sentences, add questions you answer in writing; the templates show the shapes.
• Problem subjects need problems. In mathematics, physics, accounting or programming, a quiz on the text shows the method was read. Working the problems, mixed across chapters as in Rohrer et al. (2015), builds the skill.
• The evidence has edges. The studies used single texts, one chapter, school courses or lists of concepts, and none tested this timetable on a whole book. We read them at their published abstracts, so every figure above is one printed there.

Self-testing shows what you know while there is time to act on it. It does not promise a grade.

Quick questions

Should I take a chapter quiz straight after reading or the next day?

The same day, as a first check, and again after a gap. Believe the later score: in Rawson and Kintsch (2005) the advantage of an immediate second reading was gone within two days.

Can I make chapter quizzes from a PDF of my textbook?

Yes, if the PDF has a text layer and is no larger than 15 MB. EdProcto, described on its page for students, splits it at its headings and makes a study quiz for each chapter you tick.

Is it better to quiz chapters in order or mixed?

Read them in order and review them mixed: the new chapter's quiz first, then one or two older chapters out of sequence.

What score means I know a chapter?

EdProcto's map marks a chapter known at a best score of 80%. Our stricter rule of thumb is 80% at two sittings a week or more apart, because a same-day score flatters.

References

• Custers, E. J. F. M. (2010). Long-term retention of basic science knowledge: A review study. Advances in Health Sciences Education, 15(1), 109–128. https://doi.org/10.1007/s10459-008-9101-y
• Glenberg, A. M., Wilkinson, A. C., & Epstein, W. (1982). The illusion of knowing: Failure in the self-assessment of comprehension. Memory & Cognition, 10(6), 597–602. https://doi.org/10.3758/BF03202442
• Lindsey, R. V., Shroyer, J. D., Pashler, H., & Mozer, M. C. (2014). Improving students' long-term knowledge retention through personalized review. Psychological Science, 25(3), 639–647. https://doi.org/10.1177/0956797613504302
• McDaniel, M. A., Howard, D. C., & Einstein, G. O. (2009). The read-recite-review study strategy: Effective and portable. Psychological Science, 20(4), 516–522. https://doi.org/10.1111/j.1467-9280.2009.02325.x
• Rawson, K. A., & Dunlosky, J. (2011). Optimizing schedules of retrieval practice for durable and efficient learning: How much is enough? Journal of Experimental Psychology: General, 140(3), 283–302. https://doi.org/10.1037/a0023956
• Rawson, K. A., & Kintsch, W. (2005). Rereading effects depend on time of test. Journal of Educational Psychology, 97(1), 70–80. https://doi.org/10.1037/0022-0663.97.1.70
• Rohrer, D., Dedrick, R. F., & Stershic, S. (2015). Interleaved practice improves mathematics learning. Journal of Educational Psychology, 107(3), 900–908. https://doi.org/10.1037/edu0000001
• Thiede, K. W., Anderson, M. C. M., & Therriault, D. (2003). Accuracy of metacognitive monitoring affects learning of texts. Journal of Educational Psychology, 95(1), 66–73. https://doi.org/10.1037/0022-0663.95.1.66
• Uner, O., & Roediger, H. L. (2018). The effect of question placement on learning from textbook chapters. Journal of Applied Research in Memory and Cognition, 7(1), 116–122. https://doi.org/10.1016/j.jarmac.2017.09.002
• Yang, C., Zhao, W., Yuan, B., Luo, L., & Shanks, D. R. (2023). Mind the gap between comprehension and metacomprehension: Meta-analysis of metacomprehension accuracy and intervention effectiveness. Review of Educational Research, 93(2), 143–194. https://doi.org/10.3102/00346543221094083