Reading is studied extensively in cognitive science and education research, and several findings have practical implications for how people read.
Decoding and comprehension
Recognising words and understanding meaning are separable skills.
Which is the basis of the simple view of reading used in education research.
Fluent decoding frees attention for comprehension, which is why fluency matters.
Background knowledge
Comprehension depends heavily on what a reader already knows about the subject.
Which is one of the most consistent findings in the literature.
A reader with subject knowledge outperforms a stronger general reader without it.
Vocabulary
Strongly associated with comprehension and largely acquired through reading itself.
Which produces a compounding effect over time.
Wide reading is the main route to vocabulary growth after early childhood.
Screen versus paper
Meta-analyses have generally found a modest advantage for paper on comprehension of longer informational text.
Which appears to relate to reading behaviour and to overconfidence about comprehension on screen.
The effect is smaller or absent for narrative text.
Rereading and testing
Retrieval practice produces better retention than rereading.
Which is well established in learning research.
Pausing to summarise what you have read outperforms reading the passage again.
Speed reading claims
Research finds substantial comprehension loss at very high speeds.
Which contradicts the claims made for commercial speed reading programmes.
Skimming is a legitimate skill for specific purposes and is not reading.
Reading for pleasure
Associated with comprehension, vocabulary and, in longitudinal studies, with broader outcomes.
Which is a correlation with plausible mechanisms rather than a demonstrated cause.
Practical implications
Read widely to build knowledge, read attentively rather than quickly, and pause to recall rather than rereading.
Reading aloud to children
Associated with vocabulary and later reading outcomes.
Which is one of the more robust findings in developmental research.
Dialogic reading, involving discussion, appears more effective than reading alone.
Phonics and instruction
Systematic phonics instruction has substantial evidence for early decoding.
Which has been the subject of prolonged educational debate.
Decoding instruction and knowledge building address different parts of the same problem.
Annotation and note-taking
Active engagement with text improves retention.
Which is supported across studies.
Handwritten notes have shown advantages over verbatim typing in several studies.
Audio comprehension
Listening comprehension is comparable to reading for narrative material.
Which supports treating audiobooks as reading for most purposes.
Dense informational text shows more divergence.
What to take from it
Read widely, read attentively, pause to recall, and build knowledge in the subjects you want to read about.
Attention and interruption
Fragmented reading affects comprehension of extended argument.
Which is a plausible mechanism behind screen reading differences.
Reading environment matters more than medium in several studies.
Difficulty and desirable difficulty
Text that requires effort can produce better retention.
Which is a documented effect in learning research.
Making material easier does not always improve learning from it.
Prior expectations
Readers interpret text through existing beliefs.
Which affects comprehension and recall measurably.
This is well documented and applies to everyone rather than to other people.
Reading in a second language
Comprehension is affected by vocabulary and by cultural knowledge assumed by the text.
Which is separate from language proficiency.
Building the habit
Regular reading of increasing difficulty in subjects you care about, which is what the research supports and what almost no advice says.
Why background knowledge dominates
Understanding a text requires connecting it to what you already know, and readers with relevant knowledge outperform stronger general readers without it.
This finding has substantial implications for how reading is taught, and it argues for wide reading across subjects rather than for comprehension exercises in isolation.
Reading and wellbeing
Associations between reading and various wellbeing measures appear in observational research.
Which is correlational and is frequently reported as causal.
The plausible mechanisms include attention, absorption and reduced screen use.
Practical takeaway
Consistency matters more than volume, and reading things you actually want to read is what produces consistency.
Reading in translation
Cultural references and idiom affect comprehension independently of language.
Which translators address through choices about domestication.
Notes and introductions serve exactly this purpose in scholarly editions.
Reading difficulty measures
Readability formulas measure sentence and word length rather than actual difficulty.
Which makes them crude proxies that are widely misapplied.
A text can score as easy and be conceptually demanding.
A closing observation
The most robust finding in reading research is also the least commercially useful: comprehension depends overwhelmingly on what you already know about the subject.
Which means the route to reading better is reading widely over years, and there is no technique that substitutes for it.
The one practical habit
Pause at the end of a chapter and summarise it to yourself. It outperforms rereading and takes thirty seconds.
The finding holds across studies and is consistently ignored in favour of techniques that promise faster results.
Nothing about reading well is quick, which is presumably why so much advice pretends otherwise.