ManageTime

Article

Print it or read it on screen: the document you have to take in

Print it when the text is dense and you are reading against the clock. A 2018 meta-analysis of 54 studies and 171,055 readers found a small advantage for paper, about a fifth of a standard deviation, and that advantage grew under time pressure and held on informational text, not narrative. Reading at your own pace, the medium barely registered. This article covers what that research measured: which of your documents the paper gain applies to, why a deadline changes the answer, and what the evidence does not reach, since nearly all of it is students taking comprehension tests.

Does reading on screen make you take in less?

A little, and the size of the gap is the part you can use. Delgado, Vargas, Ackerman and Salmeron published a meta-analysis in Educational Research Review in 2018. It pulled together studies from 2000 to 2017 in which people read comparable texts on paper and on a digital device. Thirty-eight between-participants designs and sixteen within-participants designs, fifty-four in all, covering 171,055 participants. Both kinds of design produced the same answer: paper held a small advantage, Hedge's g of -0.21, where the minus sign points at paper.

A fifth of a standard deviation is not a dramatic number, and it is not nothing. On one document it means the paper version leaves you slightly better able to answer questions about what you just read. Across a year of documents it means no difference you would ever notice on most of them, and something real on a few. Which few is the question you actually care about, and the meta-analysis answers it through its three moderators.

Take the time frame first, because it is the one that will change what you do tomorrow. The paper advantage increased when reading was time-constrained and shrank when readers worked at their own pace. Text genre came second: the advantage held across studies using informational texts, or informational mixed with narrative, and did not hold in the studies that used narrative text alone. The third moderator is the odd one. The paper advantage grew over the 17 years of studies the analysis covered, rather than fading as screens and readers both got better.

The later the study, the bigger the gap.

Read the base under all of this before you carry any of it to your desk. These were mostly students, reading for a comprehension test, in a study, for minutes rather than an afternoon. What got measured was what they could report afterwards, not a decision they got right. Nobody watched somebody sign off a contract. What travels to your own week is the shape of the finding and its moderators, not the number.

Which of your documents are worth printing?

The dense ones you have to get right today. Two separate pieces of evidence converge on that description, and it covers far less of your reading week than the printer queue suggests.

Fontaine and colleagues ran the narrower of the two, a meta-analysis in health professional education published in 2021 in the American Journal of Pharmaceutical Education. Ten controlled studies, 817 participants between them. Overall they found a slight advantage for paper that did not reach significance: a standardized mean difference of -0.08, with a confidence interval running from -0.28 to 0.12. An interval that crosses zero is the honest headline there, and the authors put it plainly as little to no difference.

One subgroup moved, though. When the texts were relevant to the students' own professional discipline, the paper advantage did reach significance, at -0.36 with an interval of -0.69 to -0.03. Material people had a reason to use behaved differently from material they were handed to read. The reviewers rated the quality of that evidence low, because of risks of bias across the studies, so read the subgroup as a lead rather than a result.

Line it up against the 2018 genre moderator and you get a test you can apply per document instead of a policy. Print the ones that are informational, dense, and yours to act on: the contract you have to mark up, the specification you will be held to, the forty-page report you have to summarize for somebody else by Friday. Read on screen the ones you are monitoring rather than absorbing, which is the status update, the thread you are catching up on, the deck whose second half you have already seen.

Most of what arrives is the second kind.

Printing costs you things, and the honest version of this advice names them. You lose search. You lose the ability to copy a line straight into your reply, you carry paper between rooms, and you cannot usefully print a document that will change twice before the meeting. That is why this is a per-document call: the version of it that survives a real week is printing one or two things and reading everything else where it arrived.

Why do you skim a long report on screen?

Twenty minutes to take in forty pages before the meeting starts. Delgado and Salmeron built an experiment around exactly that squeeze and published it in Learning and Instruction in 2021. They split 140 undergraduates across four conditions: print or screen, crossed with free reading time or time pressure. Then they measured how much readers mindwandered, what they predicted they would score, and what they actually comprehended.

The result worth your attention is the adjustment that did not happen. Readers in print mindwandered less under time pressure than they did with free time, which is what responding to a deadline looks like from the outside. On-screen readers made no such shift; their mindwandering did not drop when the clock started. And the on-screen readers under pressure comprehended less than all three of the other groups did.

The deadline changed the paper readers and left the screen readers where they were.

With free reading time, the medium changed neither mindwandering nor comprehension. The screen made nobody a worse reader. It made your deadline harder to answer, which is a narrower claim and a more useful one.

One more measure from that study is worth a second read. How accurately readers predicted their own test performance did not differ by condition, so the experiment found no sign that anybody could tell the medium was costing them. That is a null result in a 140-person study, and it was not built to hunt for overconfidence, so do not stretch it. It does mean your own sense of how well the reading went is not evidence. You would not feel it happening.

The trigger, then, is your deadline rather than the page count. A long document you can read at your own pace on Tuesday afternoon does not need a printer, and neither of these studies gives you a reason to find one. Hand you the same document at two for a three o'clock meeting and the answer flips. That is the precise condition where paper measured best, in the 2018 analysis and in the 2021 experiment both.

The other half of a rushed read is what lands in the middle of it, which this site covers separately in its articles on focus-block length and on overheard office talk. Putting your printed document inside a Deep Work block appears in none of these three studies. Treat that as the mechanism applied, not a measured result. The reasoning is plain enough: an advantage measured over minutes of uninterrupted reading has little to give you if your forty pages get read in six interrupted pieces.

How does your working style change the answer?

These three studies never asked how anybody organises a week, so read this section as reasoning and not as evidence. Architect, Sprinter, Visionary and Improviser are descriptions of how you arrange your work, and none of them predicts how your reading holds up on a screen. What follows sets the findings beside the week each style tends to have.

If you plan the week before it starts, the Architect's habit, you hold the one advantage this evidence rewards: you can see the dense document coming. A specification that lands on Monday for a Thursday decision can be printed on Monday and read at your own pace, which puts you in the condition where the medium stopped mattering at all. Print it early and the question dissolves.

Burst workers, closest to the Sprinter, live in the condition where paper won. One strong block against a deadline is time-constrained reading by definition, so the document you carry into that block is the one worth printing. The cost lands later. You retype every quote by hand.

The Visionary reads for the shape of an argument rather than for its detail, and here the genre moderator is a warning rather than a permission. Paper measured ahead on informational text, which is most of what a working week hands you; the narrative reading where it did not measure ahead is not what a forty-page report is. Skimming a dense document for the gist on screen is the pattern with the least support anywhere in this evidence.

If you cannot predict next week's shape, the Improviser's case, a printing habit will not survive but a two-part question will. Is it dense, and is it due? Two yeses mean print it. Anything else means reading it where it already is. That question holds up on a home day, on an office day with a broken printer, and at a client site.

Whichever of those sounds like you, the experiment is small enough to run this week. Take the next document that is both dense and due, print it, read it on paper, and see how much of it you can reconstruct afterwards without looking down.

One document will not settle anything, because the measured effect is about a fifth of a standard deviation and your week is not a controlled trial. What two or three tries will tell you is whether you read paper differently at all, which is the part no meta-analysis can answer on your behalf.

If you are not sure which style is yours, the working style assessment takes a few minutes. And if what breaks a long read is what happens around it rather than what it is printed on, the focus pillar sorts through the rest of the causes.

Frequently asked questions

Is it better to read a long document on paper or on screen?

Paper holds a small measured advantage, about a fifth of a standard deviation in a 2018 meta-analysis of 54 studies and 171,055 readers, and it was largest when reading was time-constrained and did not hold on narrative text at all. Reading at your own pace, the medium made little difference. So the useful version of the question is not paper or screen in general, but whether this particular document is dense and due soon.

Which work documents are worth printing?

The dense, informational ones you have to act on under a deadline: a contract to mark up, a specification you will be held to, a long report you have to summarize by a set time. A 2021 meta-analysis of ten studies in health professional education found no significant paper advantage overall, but did find one on texts relevant to the readers' own discipline. Material people had a reason to use behaved differently from material handed to them. A status update you are monitoring is not in that category.

Does reading on screen make you take in less?

Slightly, on average, and mostly when you are rushed: the 2018 meta-analysis found a small overall advantage for paper that grew under time pressure and held on informational rather than narrative texts. It also grew across the 17 years of studies the analysis covered, rather than shrinking as screens improved. Read the base honestly, though: these were mostly students taking comprehension tests, not people making decisions out of work documents.

Why do you skim when you read a report on screen?

Because a deadline does not change screen reading the way it changes paper reading: in a 2021 experiment with 140 undergraduates, print readers mindwandered less under time pressure while on-screen readers made no adjustment at all. The on-screen readers under pressure comprehended less than the other three groups. With free reading time, the medium changed neither mindwandering nor comprehension. The same study found no difference by condition in how accurately readers predicted their own scores.

Sources