“The methods that feel the most productive while you're doing them tend to be the ones that work the least.”
In 2013, a group of researchers went through the actual evidence behind ten common study techniques and rated each one on how well it's actually supported by controlled studies, not by how popular it is or how good it feels. The results split cleanly into two groups, and the split is almost exactly backwards from how often people use each method.
Here's what they found, and what to actually do with it.
Low utility: the methods everyone uses
Highlighting and underlining. Almost universal, almost useless on its own. The problem is that deciding what to highlight requires so little thought that it barely counts as engaging with the material, and studies comparing highlighted text to plain text found next to no difference in what people remembered afterward. It feels productive because you're doing something with a pen. Doing something isn't the same as learning something.
Rereading. The most common study strategy there is, and one of the weakest. We wrote a longer piece on why rereading feels like understanding without being it, but the short version is that recognising a sentence you've already read feels like knowing it, and those are two different things that happen to feel identical from the inside.
Summarising, done as a passive exercise of condensing text, sits in the middle. It helps more than rereading because it forces some processing, but it depends heavily on whether you're good at identifying what matters, which is a skill on its own. Done well, with the goal of finding the core idea rather than just making the text shorter, it's genuinely useful. We wrote a full guide to doing it properly because the difference between a good and bad summary is enormous.
High utility: the methods almost nobody uses enough
Practice testing came out at the top of nearly every comparison. Not practice as in "read the material and then feel confident," but actual retrieval under test-like conditions: answering questions without the material in front of you, then checking. It works because it's literally rehearsing the exact skill an exam demands, rather than a related but easier skill.
Distributed practice, otherwise known as spaced repetition, came in right behind it. Spreading review sessions out over days and weeks beats cramming the same total hours into one sitting, and the effect holds up across basically every subject researchers have tested it on.
Elaborative interrogation and self-explanation rated well too, which is the academic name for asking "why is this true" and then answering it yourself, out loud or in writing, connecting it to what you already know. This is most of what our AI tutor is built around: you explain something in your own words, it asks the follow-up question that pushes on the weak part of your explanation, and that follow-up is exactly the elaborative step that makes this method work.
Why the gap exists
The methods rated low share something in common: they're passive. You're receiving information, not producing it. Highlighting receives a sentence and marks it. Rereading receives the page again. Even a shallow summary is mostly transcription with some words cut.
The methods rated high share the opposite: they're active. You have to generate an answer, at a delay, without help, and find out whether you were right. That struggle is uncomfortable, which is exactly why people avoid it in favour of the easier methods, and exactly why it works. Difficulty that comes from genuinely retrieving something, rather than difficulty from confusing or badly written material, is called desirable difficulty in the research for a reason. It's supposed to feel harder. That's the mechanism, not a side effect.
A study plan built from what actually works
Practically, this means shifting the balance of your study time:
- Get the material into a reviewable form fast: a summary for the big picture, flashcards for the individual facts.
- Spend the bulk of your remaining time on active recall, not rereading the source material again. Practice questions, flashcards you answer before flipping, and explaining concepts out loud to something that pushes back.
- Space that recall out over days rather than doing it all in one sitting the night before.
None of this means highlighting or rereading are worthless as a first pass through new material. They just shouldn't be where your study time actually lives. The research is fairly blunt about which methods earn the hours and which ones just feel like they do.
The short version
Highlighting and rereading are the most common study methods and among the least effective. Practice testing, spaced review and self-explanation are among the most effective and the least used. The pattern is that active, effortful recall beats passive review almost every time it's been tested, even when it feels worse while you're doing it.