Study methods that work, ranked by the evidence: ten techniques, three to keep, five to drop

The study methods with the strongest evidence are practice testing (recalling or solving before you look) and spaced practice (short sessions over several days). Interleaving, self-explanation and elaborative questioning have moderate support. Highlighting, rereading, summarizing without training, keyword mnemonics and imagery rank low in the review that rated all ten. Below is the ranking as a table, then each method in a paragraph with a link to its full guide.

How the methods were ranked

In 2013 John Dunlosky and colleagues published a review in Psychological Science in the Public Interest that took ten common study techniques and judged each against the same four questions: does it work across different learning conditions, for different kinds of students, on different materials, and on different kinds of test? Each technique came out rated high, moderate or low utility (Dunlosky et al. 2013). That review is the reason this page can rank rather than list. Later work has not reversed the top two; practice testing and spacing have more evidence now than they had then.

"Which is the best study method?" is the wrong question; nobody studies with one. The right question is which three to keep.

The ranking, as a table

Method	Rating	What it is, in one line	Our guide
Practice testing	High	Recall or solve before you look at the answer, then check	Active recall
Distributed practice	High	Short sessions on several days instead of one long one	Spaced repetition
Interleaved practice	Moderate	Mix problem types in one session instead of blocking them	Interleaving
Elaborative interrogation	Moderate	Ask "why is this true?" and answer it from what you know	Memory techniques, the section on elaboration
Self-explanation	Moderate	Explain each step to yourself as you learn it	The Feynman technique
Summarization	Low	Write a shorter version of the text	A study guide written as questions, in the how-to-study section
Highlighting	Low	Mark text to read later	None; see below for what to do instead
Keyword mnemonic	Low	Link a new word to a familiar one by sound and image	The memory techniques guide
Imagery for text	Low	Picture what a paragraph describes	The memory techniques guide
Rereading	Low	Read the same text again	How to read a textbook

The ratings are the review's (Dunlosky et al. 2013). "Low utility" does not mean useless. It means weak evidence of benefit for the time spent, compared with the methods above it.

The three to keep

Practice testing. In the standard experiment, students who recalled a passage with the text closed remembered more a week later than students who reread it, and the rereaders had been the more confident group (Roediger and Karpicke 2006). Tonight: close the notes, write or solve the answer, then open them and check. Blurting — a blank page, a topic, five minutes of everything you know — is practice testing with a pen.

Spaced practice. A review of hundreds of experiments found that the same study time spread over several days beats the same time in one sitting, in the large majority of comparisons (Cepeda et al. 2006). Tonight: whatever you recall today, recall again in two or three days and again in a week. Date it on paper or let a schedule do it.

Interleaving. Rohrer and Taylor (2007) gave students mixed sets of problem types instead of one type at a time. The mixed group did worse during practice and far better on a test a week later, because they had practised choosing the method (Rohrer and Taylor 2007). Tonight: in a maths or science set, shuffle problems from the last three ideas instead of doing ten of one kind.

The two to use with care

Self-explanation. Explaining each step to yourself as you learn it improved understanding in Chi and colleagues' study, and the students who explained most gained most (Chi et al. 1994). It works when you already know something to connect the new idea to. The Feynman technique is self-explanation with a procedure and a clock: explain on paper, circle the gaps, read the source for the circles only, rewrite shorter.

Elaborative interrogation. Asking "why would this be true?" and answering from what you already know. It depends on prior knowledge too, which is why the review rated it moderate. It suits facts that have reasons behind them (why cold air sinks) and not procedures.

The five to drop (and what to do instead)

• Highlighting. Marks text for later and makes you recall nothing. Instead: write a question in the margin beside each highlight, and answer it tomorrow with the text covered.
• Rereading. Builds familiarity, which feels like learning until the question comes. Instead: read once, then cover and recall; the textbook guide shows a reading method that ends each section with a question.
• Summarizing. Low unless you have been taught to do it well, and most students have not. Instead: a study guide written as questions, not statements, so the guide becomes a test.
• Keyword mnemonics and imagery. Fine for a list you must hold (the cranial nerves), useless for understanding why anything works; the memory techniques guide says where they earn their place.

Once more: "low utility" is about the evidence relative to the time, not a claim that these do nothing. Rereading the night before helps a little. Recalling the night before helps more.

Where notes, flashcards and plans fit

Notes and flashcards are containers, not methods. A Cornell page or a card works when what you do with it is recall and spacing, and fails when you reread it. The guide on Cornell notes shows the cue column as a built-in self-test, and the comparison of five note-taking methods says which layout suits which class. The guide on flashcards that force recall is about a question side that recognition cannot answer. The weekly plan is where the spacing lives: each idea's next recall date on a grid of days. The Pomodoro technique belongs in the same category: a focus ritual, useful for sitting down, not a learning method.

Worked example: one idea, three methods, one week

MHF4U, the idea called the laws of logarithms. Three methods, three evenings.

Monday, practice testing. Blank page. Write the three laws from memory: log(ab) = log a + log b; log(a/b) = log a − log b; log(a^n) = n log a. Open the notes and mark each one; note which you hesitated on.

Wednesday, the spaced return, with interleaving. Five problems, log laws mixed with exponent problems so that each one makes you decide which rule applies. Two of them, with checkable answers: expand log_2(8x^3) = log_2 8 + 3 log_2 x = 3 + 3 log_2 x, since 2^3 = 8. Condense 2 log 5 + log 4 = log 25 + log 4 = log 100 = 2, with log meaning base 10. Mark them, and put any miss back on Thursday's list.

Friday, self-explanation. On a blank page, explain why log(ab) = log a + log b follows from the exponent law b^x × b^y = b^(x+y). The argument: if a = b^x and c = b^y, then ac = b^(x+y), so log(ac) = x + y = log a + log c. Circle any step you hedged on, and read that step in the textbook.

Three evenings, about forty minutes in all: recalled, spaced, mixed with its neighbours and explained.

Summary box

• Keep three: practice testing, spaced practice and, for maths and sciences, interleaving.
• Use two with care: self-explanation and elaborative interrogation, once you know something about the topic.
• Drop five: highlighting, rereading, untrained summarizing, keyword mnemonics and imagery for text.
• Notes, flashcards and plans are containers; recall and spacing are what go in them.
• The one-page version is chapter 10 of the free playbook, the cheat sheet. Every guide in this section is listed below.