Writer's Block: What the Research Actually Shows
The research on writer’s block is thinner than the advice built on it. The most recent survey I could find is Sarah Ahmed and C. Dominik Güss’s 2022 study in the Creativity Research Journal, which collected answers from 146 writers online. The experiment cited most often for fixed writing sessions is Robert Boice’s from 1983, and 27 people finished it, nine in each of three groups.
This guide goes through each study with its sample, its method and its limit, and it follows three numbers that circulate in English-language advice back to their sources. At the end there is a four-week test you can run on your own word counts. Where I could not open a full text, the text says so.
What the research on writer’s block consists of
Four sources carry most of the weight. Mike Rose published an analysis of writer’s block in College Composition and Communication in 1980, and the ERIC record describes its subject as the negative relationship between the structural rules taught in writing courses and writer’s block. I could read only that record, so I cannot report Rose’s sample or method.
Boice’s 1983 trial in Behaviour Research and Therapy compared three ways of scheduling writing. Helen Sword re-examined his studies in 2016 in the International Journal for Academic Development (Sword 2016), and I read her paper in full. Ahmed and Güss’s survey (Ahmed and Güss 2022) is the fourth source, and I read its abstract on the University of North Florida’s repository because the full text did not load.
Three figures that circulate, and where they lead
A guide on Best Writing, updated on 23 October 2026, states that 94 percent of writers experience writer’s block at some point. The link behind that sentence leads to a blog post on Science of People, not to a study. I found no survey with that result among the sources I could open.
A Final Draft article from 5 September 2025 says writers who tried walking or exercise succeeded 100 percent of the time and attributes the figure to Sarah Ahmed’s 2019 thesis. The thesis did not load for me, so I cannot confirm it. The abstract of the 2022 paper by the same author contains no percentages.
A third figure comes from the productivity literature. Gray (2010) is cited for 17 pages a year in a control group against 157 pages in a group with accountability. Sword reports that these numbers do not appear in Boice’s 1989 paper and show up, at most, in a hard-to-read chart on page 609.
What writers say causes it and what they say helps
Ahmed and Güss, both at the University of North Florida, asked 146 writers to complete an online mixed-method survey about their experiences with writer’s block. The abstract sorts the causes into four groups, physiological, motivational, cognitive and behavioral, and reports that physiological and motivational causes were the most frequent. Professional and semi-professional writers were both in the sample. The paper appeared online on 27 January 2022 and in volume 34, issue 3, pages 339 to 354.
The abstract lists twelve solutions the writers found effective. The four most useful were taking a break from writing, working on a different writing project, forcing oneself to keep writing and discussing ideas with others. Two of these point in opposite directions, because a break stops the writing and forcing it continues the writing. The abstract does not match solutions to causes, so it cannot tell you which one to try when.
Crossref lists no correction or retraction for the paper. The data come from self-reports, which means they describe what writers believe worked for them, and the abstract mentions no comparison group. A Psychology Today post from 2 March 2022 summarizes the study. It was written by Güss, a co-author.
What Boice’s 1983 experiment measured
Boice recruited 42 volunteers in Albany, New York, who described themselves as procrastinators. They were split into three conditions. A control group did no non-essential writing for ten weeks and kept a log of ideas. A second group wrote when in the mood. A third group kept five scheduled writing sessions a week and paid a 15 dollar check to an organization they disliked whenever they skipped one (Sword 2016, pp. 314 and 316 to 317).
The scheduled group wrote 3.2 pages a day on average, the spontaneous group 0.8 and the control group 0.2. Dividing the means gives a factor of 4 against the spontaneous group and 16 against the control group. Sword summarizes the result as more than three times the output of the spontaneous group and 16 times that of the control group, and she cites the means from Boice (1983, p. 538). Boice also reported more creative ideas in the scheduled group, a claim Stephen Krashen questioned in a 2002 re-analysis.
Fifteen of the 42 volunteers dropped out, three from the scheduled group, two from the spontaneous group and ten from the control group. Boice gave the burden of charting and listing as the reason for nearly all of them. The 27 who remained made up nine per group. Sword notes that the studies were limited in demographic scope and have never been replicated, and she cannot tell from the paper whether the pages were typed or handwritten. For converting pages into words, the guide to the handwritten page has the math.
In 1989 Boice worked for a year with three groups of new faculty who had written little. Ten of them, chronic procrastinators he had identified among 108 new faculty at his own university, went through the intervention. According to Sword, the group that also received visits twice a week wrote more than twice as many pages as the group with daily sessions and more than nine times as many as the control group. In a 1997 paper, binge writers who waited for inspiration produced about one sixth as many pages as regular writers.
Does writing every day prevent writer’s block?
Sword interviewed 100 researchers and collected 1,223 questionnaires between 2011 and 2015 at 45 universities in 15 countries. Of the interviewees, 12.8 percent write or aim to write every weekday of the academic year, and so do 11.5 percent of the questionnaire respondents. Roughly seven in eight do not write daily. Only two of the 100 interviewees described routines that closely resemble Boice’s regime.
The respondents were not randomly selected, since many attended her workshops, and the paper measures no productivity. Sword states that daily writing is not a bad idea and that she writes nearly every weekday before breakfast herself. Her conclusion reads, “Productivity, it turns out, is a broad church that tolerates many creeds.”
Boice and Johnson (1984) are cited by Sword for the statement that around 12 percent of academic researchers suffer from writer’s block. I could not open the 1984 paper, so that figure stays attributed to Sword. The guides to building a writing habit and to writing streaks cover what a daily routine costs when a day is missed.
What the studies say about walking and freewriting
Marily Oppezzo and Daniel Schwartz compared thinking while seated and while walking in the Journal of Experimental Psychology: Learning, Memory, and Cognition in 2014. According to the abstract on PubMed, walking on a treadmill in Experiment 1 raised the scores of 81 percent of participants on Guilford’s alternate uses test of divergent thinking, but only 23 percent on the compound remote associates test of convergent thinking. In Experiment 2 the effect lasted when participants sat down again after walking, and in Experiment 4 walking outside produced the most novel analogies. All four were tests of creative thinking and none was a writing task, and the sample sizes are only in the full text, which I could not open. Crossref lists no correction or retraction.
Linda Li of the University of Canberra ran a two-week course in the 2006 winter break with 27 students, who produced 190 freewriting texts of 130 to 350 words in response to ten prompts (Li 2007). She analysed course feedback and a teacher’s log qualitatively and reports more insight into academic writing and more confidence. The paper describes no control group and no measure of writing output, and it studies academic writing in general rather than writer’s block. For Julia Cameron’s morning pages, a fixed form of freewriting, PubMed lists no study at all, as Morning Pages: The Rules, the Word Count and the Evidence sets out.
| Measure | Source | Type and sample | Limit |
|---|---|---|---|
| Break from writing | Ahmed and Güss 2022 | Survey, 146 writers | Self-report, no comparison group mentioned |
| A different writing project | Ahmed and Güss 2022 | Survey, 146 writers | Self-report, no comparison group mentioned |
| Forcing yourself to keep writing | Ahmed and Güss 2022 | Survey, 146 writers | Self-report, no comparison group mentioned |
| Discussing ideas with others | Ahmed and Güss 2022 | Survey, 146 writers | Self-report, no comparison group mentioned |
| Scheduled sessions with a penalty | Boice 1983 | Trial, 27 academics in 3 groups | Nine per group, 15 dropouts, never replicated |
| Walking | Oppezzo and Schwartz 2014 | Four experiments on creative thinking | No writing task, sample sizes only in the full text |
| Freewriting | Li 2007 | Course of 27 students, 190 texts | Qualitative, no control group described |
A four-week test on your own numbers
Boice’s design can be rebuilt on a small scale if you count your words. You need one fixed writing time, a threshold for what counts as a writing day and four weeks. In weeks 1 and 3 you write on five weekdays at a fixed hour, for example for 30 minutes. In weeks 2 and 4 you write whenever it fits. Alternating the conditions keeps a growing habit from favoring only one of them.
For each condition you divide the total words by the ten weekdays and count the writing days. A threshold of 50 words works as a writing day, which is the first color step on the heatmap in the WriteZone apps and is explained in the guide to reading a writing heatmap. Also note how many sessions you actually started, and keep the log to one number a day, because Boice reports the effort of charting as the main reason for dropouts. Days of illness or travel are removed from both conditions and replaced by additional weekdays.
| Condition | Total words | Weekdays | Words per weekday | Writing days |
|---|---|---|---|---|
| Fixed session (weeks 1 and 3) | 4,500 | 10 | 450 | 10 |
| Free (weeks 2 and 4) | 2,100 | 10 | 210 | 6 |
In the example the ratio is 450 to 210, which is about 2.1, compared with the factor of 4 that Boice’s means give. Your own ratio rests on twenty weekdays and proves nothing beyond your own routine. It does show whether the fixed session gives you more words and how often the free weeks produce no writing at all. A writing tracker or the word count of your editor supplies the totals, and the guide to words per day shows how to turn a total goal into a daily one.
When this advice does not apply
Ahmed and Güss found physiological causes among the most frequent. The abstract does not say what falls into that group, and the full text did not load. When the body needs rest, one more fixed session probably adds little, which is my reading and not a finding of the studies.
The penalty check in Boice’s trial is not a recommendation here. It comes from an experiment with 27 completers. For a thesis with a hard deadline, the guide to catching up on a dissertation shows how to recompute the plan. If a block lasts for weeks and weighs on you beyond your writing, a university writing center or a doctor is a better next step than another tip from an advice article. Timed writing sprints apply the idea of a fixed session without penalties, and I found no study on their effect on writer’s block.
How do you fix writer's block?
In Ahmed and Güss (2022), 146 writers described twelve solutions they found effective. The abstract lists “the four most useful strategies” as taking a break from writing, working on a different writing project, forcing themselves to keep writing and discussing ideas with others. These are self-reports without a comparison group. The four-week test above shows whether a fixed session raises your own word count.
What causes writer's block?
Ahmed and Güss (2022) sorted the answers of their 146 writers into four groups, physiological, motivational, cognitive and behavioral. According to the abstract, physiological and motivational causes were the most common and “interfered with the composition process of writing more often than idea generation.” Mike Rose linked blocking to the rigid rules of composition classes in 1980.
Is writer's block real?
The journal College Composition and Communication published an analysis of it in 1980 by Mike Rose, and Ahmed and Güss surveyed 146 writers about their experiences in 2022. The sources I read give no single definition or diagnostic test.
How common is writer's block?
The only figure I could trace is Boice and Johnson’s roughly 12 percent of academic researchers, quoted by Sword. The 94 percent figure in advice articles leads to a blog post, not to a survey.
How long does writer's block last?
I found no reliable figures. The abstract of Ahmed and Güss (2022) mentions duration only to say that it did not differ significantly between professional and semi-professional writers, and it gives no values, and Boice’s 1983 trial measured pages per day.
Does walking help with writer's block?
Oppezzo and Schwartz (2014) found higher scores while walking on tests of creative thinking in four experiments, for 81 percent of participants in the first one. None of the tests was a writing task, so the study says nothing direct about writer’s block.
Does writer's block affect professional writers?
The sample of 146 in Ahmed and Güss (2022) included professional and semi-professional writers. The abstract reports no significant difference between the two groups.