Shuichiro Ogawa
日本語

Notes · updated 2026-11-13 · Markdown

Does the Quantity of Work and Practice Build Skill? Evidence on Repetitions, Case Volume, Years of Experience, Practice Scheduling, Working Hours, and Workplace Experience

The belief that doing more makes people better was examined for everyday work and practice, separately for six kinds of quantity (repetitions, case volume, years of experience, practice time and its scheduling, working hours, and workplace experience), using 46 academic sources and 1 report.

Contents (10)
  1. Which quantity does “doing more makes you better” refer to?
  2. Scope and method
  3. How much faster and more accurate does repeating the same task make people?
  4. Do surgeons who handle more cases achieve better outcomes?
  5. Do years of experience indicate skill?
  6. Does practicing longer make people better?
  7. Do longer working hours help people acquire routines?
  8. What makes workplace experience a source of learning?
  9. Which forms of the claim are supported?
  10. Gaps

Which quantity does “doing more makes you better” refer to?

Workplaces often recommend quantity. Get more cases under your belt. Work long hours while you are young and learn the routines. Spend 10,000 hours on one thing and you will reach the top. All of these sayings share one form: accumulating quantity raises skill.

However, the quantity in these sayings is not a single thing. How many times a task was repeated, how many cases were handled, how many years a person has held a job, how many hours were spent practicing, how many hours were worked per week, and what kind of experience was accumulated have each been studied in separate research streams, each linked to a different indicator of skill. The indicators also differ across studies: task speed, error counts, patient mortality, supervisor ratings, sales, and test scores. Even when people with more years of experience perform better, the correlation does not show that quantity built skill unless three explanations are separated: the same person improved over the years, less skilled people left while skilled people stayed, or work was assigned to people who were already skilled.

The number of creative works was examined in Does Quantity Breed Quality? Hits and Hit Rates, Idea Counts and the Best Idea, How Prototypes Are Compared, Practice Time, and the Origin of the Ceramics Class Story, which concluded that producing more works increases the number of hits, whereas whether each work becomes better depends on how quality is measured. How much practice time explains differences in expertise, and how this relates to heritability, was examined in Is the Ability to Make an Effort Also a Talent? The Heritability of Grit, Self-Control, and Motivation to Learn, and How Far These Traits Can Be Changed. This note focuses on the quantity of everyday work and practice. Do people with more work or more years of experience become more skilled? Do longer working hours help people acquire routines (standard procedures and judgments)? Does practicing longer make people better?

Scope and method

  • Target: peer-reviewed articles, meta-analyses, systematic reviews, and reviews that measured the relationship between quantity (counts, cases, years, hours) and skill, performance, or quality. Six streams were covered: learning curves in production and the labor economics of learning by doing, surgical volume and outcomes, years of experience and job performance, practice time and scheduling, working hours and output and health, and experiential learning in the workplace. The period was 1979 to 2021. Japanese peer-reviewed articles were included.
  • Collection: 71 candidates were gathered across academic databases (OpenAlex, Crossref, Europe PMC, J-STAGE), and Crossref was used to check for retractions, corrections, and predatory journals (0 retractions, 0 suspected predatory journals). Studies of the number of creative works and studies of heritability and talent were excluded because existing notes cover them.
  • Verification: of the 48 references (46 academic sources, 1 report, and 1 book), the full texts of 20 academic sources and the report were obtained (including working-paper versions, institutional repositories, PMC, and J-STAGE), and numbers and claims were checked against the relevant passages. For 26 academic sources, only the abstract was available, and their content is reported as the authors’ claims. For the book, only the bibliographic record was checked.
  • Reporting: for each study, the note distinguishes what was treated as quantity, what was treated as the indicator of skill, and whether the comparison was between people or within the same person or organization over time.

How much faster and more accurate does repeating the same task make people?

In psychology, the relationship between repetition and improvement has been described as the law of practice, a curve in which response time decreases with the number of practice trials. According to the abstract of Evans, Brown, Mewhort, and Heathcote (2018), this curve was first described as a power function, but the power form had been shown to be a bias produced by averaging, across participants, individual curves that are better described by exponential functions. In both the power and the exponential forms, the rate of speed-up declines steadily with practice. The authors proposed a law that can also describe curves that are slow at first, then faster, then slow again, and a delayed exponential law was supported in a clear majority of tasks. The gain per repetition becomes smaller as repetitions accumulate.

Output at work also shows large early gains. Shaw and Lazear (2008, working-paper version) compared individual output records with months of tenure at an automobile windshield installation company. Output per worker was 53% higher after one year than at hire. This estimate followed the same workers over time, so it excludes the apparent gains that would arise from less able workers leaving. However, the company weeded out less able workers in the first few months, and the curves that compared different people were steeper than the curves within the same person. Among older workers with more than two years of tenure, output declined with tenure. Because the work itself did not change, the authors considered declining effort (boredom) more likely than declining skill.

The gains are not necessarily stored in workers’ skills. Levitt, List, and Syverson (2013, working-paper version) used detailed records from an assembly plant of a major automaker and showed that defects per vehicle fell by more than 80% in the first eight weeks of production. Yet the second shift, which started nearly two months later with new workers, had fewer defects than the first shift from the very beginning. Worker absenteeism also had only a small effect on defects. The authors concluded that the improvements were stored in the plant’s physical and organizational capital rather than in individual workers. In the same paper, the estimated stock of production experience was lost at 3% to 7% per week.

Depreciation of experience was also reported for aircraft production. Benkard (2000, working-paper version) used production records of the Lockheed L-1011 and showed that the cost per aircraft did not always fall with cumulative production. The analysis supported organizational forgetting, in which a firm’s production experience depreciates over time, and also supported incomplete transfer of experience from one generation of an aircraft to the next. In his abstract, Thompson (2012) revisited Wright’s (1936) cost curve for aircraft and noted that only one of Wright’s three explanations is unambiguously organizational learning, whereas the others are also consistent with static economies of scale. Against the Boston Consulting Group’s experience curve, which claimed that costs fall by 20% to 30% each time cumulative production doubles and continue to fall without limit, Thompson stated that direct empirical evidence does not sufficiently support reading the relationship between cost and cumulative production as an organizational learning curve.

When repeating the same task helps also differed. Staats and Gino (2012, working-paper version) analyzed 56,227 home loan applications processed by 140 workers over two and a half years at Shinsei Bank in Japan. Within a single day, workers were more productive on days when they stayed on the same task. Across days, workers who had experienced a variety of tasks became more productive and were also more likely to stay with the firm.

At least for simple tasks, the quantity of repetition has a large effect early on. The gains slow down, however, and part of the gains comes from the selection of workers and from organizational arrangements, and the gains depreciate unless production continues.

Do surgeons who handle more cases achieve better outcomes?

Surgery is one of the fields where the relationship between quantity and outcomes has been studied most extensively. In their abstract, Luft, Bunker, and Enthoven (1979) reported, for 12 surgical procedures in 1,498 hospitals, that hospitals performing 200 or more operations per year had case-mix-adjusted death rates 25% to 41% lower for procedures such as open-heart surgery and coronary bypass. For total hip replacement, mortality had nearly reached its low at 50 to 100 operations per year, and for cholecystectomy there was no relationship between volume and mortality. The authors wrote that the results may reflect the effect of experience or the referral of patients to institutions with better outcomes.

This volume-outcome relationship was later confirmed widely in systematic reviews. According to the abstract of Halm, Lee, and Chassin (2002), among 135 studies that met the inclusion criteria, 71% of the studies of hospital volume and 69% of the studies of physician volume reported a significant association between higher volume and better outcomes. However, the size of the association varied greatly across procedures, and studies that adjusted for risk with clinical data were less likely to report significant associations. According to the abstract of Birkmeyer et al. (2002), mortality decreased with hospital volume for all 14 procedures in about 2.5 million operations in the United States, from 16.3% to 3.8% for pancreatic resection and from 1.7% to 1.5% for carotid endarterectomy.

Much of the effect of hospital volume was explained by the volume of the operating surgeon. According to the abstract of Birkmeyer et al. (2003), surgeon volume was inversely related to mortality for all eight procedures among 474,108 patients. Surgeon volume accounted for 100% of the apparent effect of hospital volume for aortic-valve replacement and 24% for lung resection. According to the abstract of the systematic review by Chowdhury, Dagash, and Pierro (2007), high-volume surgeons had significantly better outcomes in 74% of the studies and specialists outperformed general surgeons in 91%, whereas the benefit of hospital volume was limited to 40% in prospective studies.

The studies above are correlations showing that hospitals and physicians with more cases achieve better outcomes. If patients go to skilled physicians, outcomes produce volume rather than the reverse. Hentschker and Mennicken (2018) used the number of potential patients living near each hospital and the number of nearby hospitals as instruments and estimated the causal effect of volume on mortality for 89,541 hip fracture patients in 1,262 hospitals in Germany in 2007. Volume significantly reduced in-hospital mortality, and the authors treated this result as evidence for the practice-makes-perfect hypothesis. However, the only outcome measured was in-hospital mortality, and at low volumes the shape of the relationship varied considerably with the estimation method. This result is at the hospital level and does not measure individual skill.

The experience curve of an individual surgeon can also be long. According to the abstract of Vickers et al. (2007), the curve for recurrence among 7,765 patients who underwent radical prostatectomy by 72 surgeons at four centers was steep and did not begin to plateau until the surgeon had completed about 250 operations. The predicted probability of recurrence at five years was 17.9% for surgeons with 10 prior operations and 10.7% for surgeons with 250.

However, accumulated case volume is not always portable skill. Huckman and Pisano (2006, working-paper version) used records of cardiac surgeons who operated at more than one hospital in the same year and showed that a surgeon’s outcomes at a given hospital improved with annual volume at that hospital but were not related to volume at other hospitals. The authors concluded that part of a surgeon’s performance is specific to the hospital and is not carried over to another hospital.

What people learn from experience also depended on the kind of outcome. According to the abstract of KC, Staats, and Gino (2013), in ten years of data on 71 surgeons who performed more than 6,500 cardiac procedures with a new technology, surgeons learned more from their own successes than from their own failures, and more from others’ failures than from others’ successes. Accumulating one’s own failed cases did not easily turn into learning by itself.

Skill also varies widely among surgeons who perform the same operation. According to the abstract of Birkmeyer et al. (2013), when peers anonymously rated videos of operations by 20 bariatric surgeons, mean skill ratings ranged from 2.6 to 4.8 on a five-point scale, and patients of surgeons in the lowest quartile had more complications (14.5% vs. 5.2%) and higher mortality (0.26% vs. 0.05%) than patients of surgeons in the highest quartile.

Case volume was supported as a predictor of surgical outcomes by correlations, by a hospital-level causal estimate, and by individual surgeons’ experience curves. However, part of this skill was specific to the hospital, and learning differed depending on whether successes or failures accumulated.

Do years of experience indicate skill?

Years of experience are the easiest measure of quantity. However, years do not show how many times, or what kind of work, a person performed during those years. According to the abstract of Quiñones, Ford, and Teachout (1995), a meta-analysis of 44 studies (25,911 people) estimated the correlation between work experience and job performance at .27. The correlation was highest when experience was measured as the number of times a task was performed (.43) or at the task level (.41), and it was .39 for objective measures such as work samples and .24 for ratings.

The correlation between experience and performance is strongest at early stages of experience. According to the abstract of McDaniel, Schmidt, and Hunter (1988), in data on 16,058 people, the correlation was highest in groups with low mean experience and in jobs with low cognitive demands. Sturman (2003) meta-analyzed the relationships of job experience, organizational tenure, and age with performance and showed, for low-complexity jobs, an inverted U-shaped relationship in which performance first rises and then falls for all three. For high-complexity jobs, the relationship was not an inverted U, and experience became more predictive of performance as it accumulated. Sturman concluded that the inverted U-shaped relationship between time and performance does not hold universally.

In medicine, many studies report lower quality with more years in practice. According to the abstract of Choudhry, Fletcher, and Soumerai (2005), 32 of 62 evaluations (52%) reported decreasing performance with more years in practice for all outcomes, 13 (21%) for some outcomes, and 13 (21%) reported no association. Learning by Practice with Knowledge Withheld, Learning by Knowledge with Practice Withheld: The Literature on Whether Withholding One Side Has Meaning reads the same review in a discussion of whether to withhold knowledge or practice.

This decline sometimes disappears when quantity, rather than years, is considered. Tsugawa et al. (2017) compared 736,537 hospital admissions of older patients in the United States by the age of 18,854 hospitalists within the same hospital. Adjusted 30-day mortality was 10.8% for physicians under 40 and 12.1% for physicians 60 or older. However, among physicians who treated more than 201 admissions per year, there was no association between age and mortality. The authors suggested that treating many patients may maintain clinical skills, but they also noted, as limitations, that the relationship could run in the opposite direction (physicians whose skills decline may treat fewer patients) and that the cross-sectional design could not separate aging from cohort differences.

For the accuracy of clinical judgment, the effect of experience was small but positive. According to the abstract of Spengler et al. (2009), a meta-analysis of 75 studies on mental health judgments (4,607 clinicians) found an effect size of d = .12 for educational or clinical experience and judgment accuracy, corresponding to an increase in accuracy of about 13%. The authors described this result as contrary to the general conclusion of earlier clinical judgment research that accuracy does not improve with experience.

Whether experience sharpens judgment depends on the environment. In their abstract, Kahneman and Klein (2009) concluded that evaluating the likely quality of an intuitive judgment requires assessing how predictable the environment is and whether the person had an opportunity to learn its regularities, and that subjective confidence is not a reliable indicator of judgment accuracy. Ericsson (2008) wrote in his abstract that length of experience and reputation are only weakly related to actual performance and that expert performance can be traced to deliberate practice with immediate feedback and opportunities for repetition. In the opening of Ericsson (2004), he stated that most professionals reach a stable, average level of performance within a relatively short time and maintain it for the rest of their careers.

In some occupations, performance continues to improve with years. Kini and Podolsky (2016) at the Learning Policy Institute reviewed 30 studies on teachers’ experience and student achievement in the United States and reported that 28 studies found a positive and significant association between experience and teacher effectiveness, and that all 18 studies that followed the same teachers found a positive association. The gains were steepest in the first few years but remained significant in the second, and often third, decade of a teacher’s career. The gains were larger when teachers accumulated experience in the same grade, subject, or district, and when they worked in supportive and collegial environments. This report is a policy institute document that did not undergo peer review.

Among Japanese salespeople, the effect of years also differed by period. Matsuo, Hosoi, Yoshino, and Kusumi (1999) divided 108 salespeople at a car dealership by years of experience. The correlation between years of experience and sales was significant in the 3 to 5 year group (r = .498) and the 6 to 9 year group (r = .623) and not significant in the group with 10 or more years (r = .218). In the group with 10 or more years, many items of procedural knowledge about selling correlated positively with sales. The authors interpreted these results as showing that, beyond 10 years, adding years alone does not raise sales and that the amount of selling knowledge determines sales. They related the results to a model of cognitive skill acquisition in which knowledge moves from a declarative stage to a procedural stage and to the 10-year rule in expertise research. This was a single-time survey and did not follow the same people over time.

Years of experience are a rough measure that mixes the number of times and the kinds of work. The effect of years was large early in a career and in simple work, and over longer periods performance diverged depending on what people did and how much, even at the same number of years.

Does practicing longer make people better?

How much practice time explains differences in expertise was examined in Does Quantity Breed Quality? Hits and Hit Rates, Idea Counts and the Best Idea, How Prototypes Are Compared, Practice Time, and the Origin of the Ceramics Class Story as part of the debate over the meta-analyses by Macnamara and colleagues, using the values after the 2018 correction. This section examines two points outside that debate: the origin of the 10,000-hour figure and how the same amount of practice is scheduled.

The “10,000-hour rule” is not a research conclusion. According to Ericsson and Harwell (2019), Gladwell (2008) devoted a chapter of Outliers to the “10,000-hour rule,” cited Ericsson, Krampe, and Tesch-Römer (1993) as the primary empirical evidence, and called 10,000 hours “the magic number for true expertise.” Ericsson and Harwell wrote that their research showed that international-level performance requires an extended period of training but provided no evidence for a magic number, and they cited an estimate that about 25,000 hours would be more accurate for winning international piano competitions. In addition, Gladwell never used the term deliberate practice in his book.

The same amount of practice produces different retention depending on how it is scheduled. Cepeda et al. (2006) meta-analyzed the effect of distributed practice on verbal recall tasks, collecting 839 assessments from 317 experiments in 184 articles. Spacing learning episodes led to better retention than massed repetition, and the interval that maximized retention increased as the retention interval before the test increased. According to the abstract of Donovan and Radosevich (1999), a meta-analysis of 63 studies (112 effect sizes) on task performance found that spaced practice led to higher performance than massed practice, with a mean effect size of 0.46. However, the size of the effect varied greatly with the nature of the task and the intertrial interval, and studies with lower methodological rigor showed larger effects.

Overlearning, which means continuing to practice after a skill has been acquired, was effective but conditional. According to the abstract of Driskell, Willis, and Copper (1992), overlearning had a moderate effect on retention, and the size of the effect depended on the degree of overlearning, the type of task, and the length of the retention period. According to the abstract of Rohrer and Taylor (2006), in mathematics practice by 216 college students, increasing the problems in one session from three to nine did not change scores one week or four weeks later. When the same ten problems were completed either in one session or in two sessions one week apart, there was no difference on a test one week later, whereas the distributed group scored much higher on a test four weeks later.

Performance during practice is not an index of how much has been learned. In their abstract, Soderstrom and Bjork (2015) reviewed research showing that what can be observed during practice is performance, which differs from durable learning, that improvements in performance can fail to produce learning, and that some manipulations move performance and learning in opposite directions. They also wrote that people often mistake their performance during practice for a guide to learning. Practicing in a massed way can look like improvement at the time and still leave little a few weeks later.

Do longer working hours help people acquire routines?

Long working hours first reduce output per hour. Pencavel (2015, working-paper version) analyzed records of munition workers, most of them women, in British factories during the First World War. Output was proportional to hours up to 49 hours per week, rose more slowly beyond that point, and reached its maximum at about 63 hours. Output at 70 hours per week differed little from output at 56 hours. In weeks that included Sunday work, output was about 10% lower for the same hours. Collewet and Sauermann (2017, working-paper version) used day-to-day variation in working hours set by the employer according to expected demand at a call center in the Netherlands and showed that a 1% increase in working hours raised the number of calls answered by only about 0.9%. The effect of fatigue appeared even in a workplace where most employees worked part-time, and the decreasing returns were stronger for workers with shorter tenure.

Long shifts also increase errors. According to the abstract of Landrigan et al. (2004), interns in intensive care units made 35.9% more serious medical errors (136.0 vs. 100.1 per 1,000 patient-days) on a traditional schedule that included shifts of 24 hours or more every other shift than on a schedule that eliminated such shifts and reduced weekly hours. On the traditional schedule, they also made 5.6 times as many serious diagnostic errors. According to the abstract of Lockley et al. (2004), which studied the same interns, weekly work hours averaged 84.9 on the traditional schedule and 65.4 on the intervention schedule; on the intervention schedule, interns slept 5.8 hours more per week and had less than half the rate of attentional failures during night work.

There is also a cost to health. Kivimäki et al. (2015) compared working 55 hours or more per week with 35 to 40 hours, estimated relative risks of 1.13 for coronary heart disease and 1.33 for stroke, and reported a dose-response association for stroke. In the WHO/ILO estimates (Pega et al. 2021), 488 million people worldwide worked 55 hours or more per week in 2016, and 745,194 deaths from ischemic heart disease and stroke were attributable to this exposure.

Does shortening duty hours, then, increase learning? Regulation of residents’ duty hours provided an occasion to test this question. Ahmed et al. (2014) systematically reviewed 135 articles on duty hour restrictions for surgical residents. Among moderate- to high-quality studies, only one reported improved educational outcomes after the restrictions, and the rest reported worsened outcomes or no change. For the 16-hour limit, studies repeatedly reported less operative exposure, more patient handovers, and poorer integration into teams.

In randomized trials, relaxing duty-hour rules did not worsen patient outcomes. According to the abstract of the FIRST trial (Bilimoria et al. 2016), which randomly assigned 117 US general surgery residency programs to flexible or standard duty-hour policies, the flexible group did not have more deaths or serious complications (9.1% vs. 9.0%), and the proportion of residents dissatisfied with overall education quality did not differ (11.0% vs. 10.7%). Residents in the flexible group were less likely to report leaving during an operation (7.0% vs. 13.2%) or handing off active patient issues. In the iCOMPARE trial in 63 internal medicine programs, 30-day mortality in the flexible programs was noninferior to that in the standard programs (Silber et al. 2019). Silber et al., citing the report on the educational outcomes of the same trial, stated that there was no significant difference in the proportion of time interns spent on direct patient care or education, but that satisfaction with educational quality and overall well-being was lower in the flexible programs. In the sleep measurements of the same trial (Basner et al. 2019), sleep per 24 hours was 6.85 hours and 7.03 hours and met the noninferiority criterion, but the results for alertness did not.

The studies of residents showed that cutting hours does not increase learning. At the same time, relaxed schedules did not raise satisfaction with education. Within the scope of the search, no study was found that directly measured, with an indicator of skill, whether longer working hours speed up the acquisition of routines. What is known is that long hours increase opportunities for experience (the number of operations attended and continuity of patient care) while costing output per hour, attention, errors, and health.

What makes workplace experience a source of learning?

The claim that 70% or more of workplace learning happens informally has come to be treated as a basic fact in human resource development under the name 70:20:10. In his abstract, Clardy (2018) examined five literature traditions that converge on this claim, stated that its evidential basis is weak, with sloppy scholarship, inconsistent conceptualizations, and problems in research protocols, and concluded that the rule needs to be set aside.

The size of experience also has diminishing returns. According to the abstract of DeRue and Wellman (2009), in 225 on-the-job experiences of 60 managers, the relationship between developmental challenge and leadership skill development showed diminishing returns. However, access to feedback offset the diminishing returns at high levels of challenge.

Japanese research on workplace learning has also examined how the actions of others combine with experience. Nakahara (2012) introduced his 2010 survey and divided the support people receive from others at work into three types: advice and instruction on tasks (task support), prompts for reflection (reflection support), and encouragement (emotional support). In a multiple regression analysis explaining self-perceived improvement in ability, reflection support had the strongest effect. This outcome was self-rated improvement, not externally measured performance. Tanaka and Nakahara (2017) analyzed interviews with 15 middle managers who had led new businesses and proposed a model in which learning proceeds through four phases: a phase of blaming others, a phase of accepting reality, a phase of reflective thinking, and a phase of perspective transformation. Difficult experience became learning after a phase of reflecting on one’s own thinking and behavior.

Research on workplace experience has shown that the presence of feedback and reflection, more than the amount of experience, distinguishes experience that leads to learning. However, most of this research relies on self-ratings or interviews, and few studies measured changes in performance. How the route of learning alongside experts changes when generative AI reduces entry-level work is examined in What Should Juniors Do When Generative AI Takes Away the Entry-Level Work? (Academic Research), and the lack of time to learn among managers who also carry their own workload is examined in What Is a Playing Manager? Gains and Losses for the Appointee and the Appointing Side in Academic and Industry Sources.

Which forms of the claim are supported?

When the claim that doing more makes people better is divided by kind of quantity, the strength of support differs by form.

Repetition produces large gains at first: strongly supported. Windshield installers’ output rose by 53% in one year (Shaw and Lazear 2008), defects at an assembly plant fell by more than 80% in eight weeks (Levitt et al. 2013), and recurrence after prostatectomy fell until the surgeon had performed about 250 operations (Vickers et al. 2007). However, the gain per repetition becomes smaller as repetitions accumulate (Evans et al. 2018).

Accumulating quantity keeps improving performance indefinitely: supported only under conditions. In low-complexity jobs, performance rose with years and then fell (Sturman 2003), and in medicine many studies reported lower quality with more years in practice (Choudhry et al. 2005). Continued improvement was found in high-complexity jobs (Sturman 2003), among teachers (Kini and Podolsky 2016), and among physicians who continued to treat many patients (Tsugawa et al. 2017).

Accumulated quantity becomes portable skill: only partly supported. Improvements at a plant were stored in the organization rather than in workers (Levitt et al. 2013), and part of cardiac surgeons’ skill was specific to the hospital (Huckman and Pisano 2006). Experience depreciated when production stopped (Benkard 2000, Levitt et al. 2013).

People with more years of experience are more skilled: years are a weaker measure than counts, and over long periods years alone do not distinguish performance well. Years correlated less with performance than counts did (Quiñones et al. 1995), and the correlation was weaker in groups with long experience (McDaniel et al. 1988). Among Japanese salespeople, beyond 10 years, knowledge rather than years distinguished sales (Matsuo et al. 1999).

Practicing longer makes people better: not determined by the number of hours. The figure of 10,000 hours is not a number that research established (Ericsson and Harwell 2019), spaced practice was retained longer than the same amount of massed practice (Cepeda et al. 2006, Rohrer and Taylor 2006), and additional practice after mastery was not retained in mathematics (Rohrer and Taylor 2006).

Working longer helps people acquire routines: there is no direct evidence. Long hours reduced output per hour (Pencavel 2015, Collewet and Sauermann 2017) and came with costs in errors and health (Landrigan et al. 2004, Kivimäki et al. 2015). On the other hand, shortening residents’ duty hours did not improve educational outcomes (Ahmed et al. 2014, Bilimoria et al. 2016).

The literature repeatedly identified several conditions for turning quantity into skill. These were an environment that is regular and in which outcomes become known quickly (Kahneman and Klein 2009), accumulation within the same organization and the same kind of work (Huckman and Pisano 2006, Kini and Podolsky 2016), spaced repetition (Cepeda et al. 2006), the opportunity to learn from one’s successes and others’ failures (KC et al. 2013), and feedback and reflection (DeRue and Wellman 2009, Nakahara 2012). The advice to get more cases under one’s belt fits the literature well for newcomers and for work in which outcomes return quickly. The advice to work long hours costs output per hour and health, and no study has yet examined whether it helps people acquire routines.

Gaps

  • No study was found that measured, with the same people’s performance or tests, how changing working hours (weekly hours or overtime) affects the acquisition of skills or routine judgments. Studies of residents focused on patient outcomes and self-reported satisfaction with education, and skill test results were limited to observational studies.
  • Within the scope of the search, no empirical study was found that measured, with indicators of skill, the up-or-out system of professional service firms (consulting, law, and accounting firms) and development under long working hours.
  • Most studies of quantity and performance came from medicine and manufacturing, and no study was found that measured the number of projects or years of experience against the quality of deliverables in knowledge work such as design.
  • Most research on experiential learning at work relies on self-ratings or interviews, and no study was found that re-measured the 70:20:10 proportions themselves.

Unverified items

  • No major claim remains unverified.
  • For 26 sources, the full text could not be obtained and only the abstract was used (Evans et al. 2018, Thompson 2012, Luft et al. 1979, Halm et al. 2002, Birkmeyer et al. 2002, 2003, 2013, Chowdhury et al. 2007, Vickers et al. 2007, KC et al. 2013, Quiñones et al. 1995, McDaniel et al. 1988, Choudhry et al. 2005, Spengler et al. 2009, Kahneman and Klein 2009, Ericsson 2008, Donovan and Radosevich 1999, Driskell et al. 1992, Rohrer and Taylor 2006, Soderstrom and Bjork 2015, Landrigan et al. 2004, Lockley et al. 2004, Bilimoria et al. 2016, Clardy 2018, DeRue and Wellman 2009, Ericsson 2004). In the text, their content is attributed to the authors with “according to the abstract” or “in the abstract.” For Ericsson (2004), only the opening paragraph of the body text, which OpenAlex stores in its abstract field, was read.
  • Shaw and Lazear (2008), Levitt et al. (2013), Benkard (2000), Huckman and Pisano (2006), Pencavel (2015), Collewet and Sauermann (2017), and Staats and Gino (2012) were read in working-paper versions from NBER, IZA, and Harvard Business School and were not checked against the published full texts. The working-paper version of Huckman and Pisano has a different title (The Effect of Organizational Context on Individual Performance, 2003).
  • Kini and Podolsky (2016) is a report by a policy institute that did not undergo peer review; the full text of the journal version (Podolsky et al. 2019) was not obtained.
  • For Gladwell (2008), only the bibliographic record (Open Library) was checked, and the content of the book is reported through the quotation in Ericsson and Harwell (2019).
  • The educational outcomes of the iCOMPARE trial cited by Silber et al. (2019) are reported within Silber et al.’s description; the original report was not read.

References

All sources were accessed on 2026-11-13.

Repetition and learning curves

  • Evans, N. J., Brown, S. D., Mewhort, D. J. K., & Heathcote, A. (2018). Refining the law of practice. Psychological Review, 125(4), 592–605. https://doi.org/10.1037/rev0000105
  • Thompson, P. (2012). The relationship between unit cost and cumulative quantity and the evidence for organizational learning-by-doing. Journal of Economic Perspectives, 26(3), 203–224. https://doi.org/10.1257/jep.26.3.203
  • Benkard, C. L. (2000). Learning and forgetting: The dynamics of aircraft production. American Economic Review, 90(4), 1034–1054. https://doi.org/10.1257/aer.90.4.1034
  • Levitt, S. D., List, J. A., & Syverson, C. (2013). Toward an understanding of learning by doing: Evidence from an automobile assembly plant. Journal of Political Economy, 121(4), 643–681. https://doi.org/10.1086/671137
  • Shaw, K., & Lazear, E. P. (2008). Tenure and output. Labour Economics, 15(4), 704–723. https://doi.org/10.1016/j.labeco.2007.08.004
  • Staats, B. R., & Gino, F. (2012). Specialization and variety in repetitive tasks: Evidence from a Japanese bank. Management Science, 58(6), 1141–1159. https://doi.org/10.1287/mnsc.1110.1482

Case volume and outcomes

  • Luft, H. S., Bunker, J. P., & Enthoven, A. C. (1979). Should operations be regionalized? The empirical relation between surgical volume and mortality. New England Journal of Medicine, 301(25), 1364–1369. https://doi.org/10.1056/NEJM197912203012503
  • Halm, E. A., Lee, C., & Chassin, M. R. (2002). Is volume related to outcome in health care? A systematic review and methodologic critique of the literature. Annals of Internal Medicine, 137(6), 511–520. https://doi.org/10.7326/0003-4819-137-6-200209170-00012
  • Birkmeyer, J. D., Siewers, A. E., Finlayson, E. V. A., Stukel, T. A., Lucas, F. L., Batista, I., Welch, H. G., & Wennberg, D. E. (2002). Hospital volume and surgical mortality in the United States. New England Journal of Medicine, 346(15), 1128–1137. https://doi.org/10.1056/NEJMsa012337
  • Birkmeyer, J. D., Stukel, T. A., Siewers, A. E., Goodney, P. P., Wennberg, D. E., & Lucas, F. L. (2003). Surgeon volume and operative mortality in the United States. New England Journal of Medicine, 349(22), 2117–2127. https://doi.org/10.1056/NEJMsa035205
  • Chowdhury, M. M., Dagash, H., & Pierro, A. (2007). A systematic review of the impact of volume of surgery and specialization on patient outcome. British Journal of Surgery, 94(2), 145–161. https://doi.org/10.1002/bjs.5714
  • Hentschker, C., & Mennicken, R. (2018). The volume–outcome relationship revisited: Practice indeed makes perfect. Health Services Research, 53(1), 15–34. https://doi.org/10.1111/1475-6773.12696
  • Vickers, A. J., Bianco, F. J., Serio, A. M., Eastham, J. A., Schrag, D., Klein, E. A., Reuther, A. M., Kattan, M. W., Pontes, J. E., & Scardino, P. T. (2007). The surgical learning curve for prostate cancer control after radical prostatectomy. Journal of the National Cancer Institute, 99(15), 1171–1177. https://doi.org/10.1093/jnci/djm060
  • Huckman, R. S., & Pisano, G. P. (2006). The firm specificity of individual performance: Evidence from cardiac surgery. Management Science, 52(4), 473–488. https://doi.org/10.1287/mnsc.1050.0464
  • KC, D. S., Staats, B. R., & Gino, F. (2013). Learning from my success and from others’ failure: Evidence from minimally invasive cardiac surgery. Management Science, 59(11), 2435–2449. https://doi.org/10.1287/mnsc.2013.1720
  • Birkmeyer, J. D., Finks, J. F., O’Reilly, A., Oerline, M., Carlin, A. M., Nunn, A. R., Dimick, J., Banerjee, M., & Birkmeyer, N. J. O. (2013). Surgical skill and complication rates after bariatric surgery. New England Journal of Medicine, 369(15), 1434–1442. https://doi.org/10.1056/NEJMsa1300625

Years of experience and performance

  • Quiñones, M. A., Ford, J. K., & Teachout, M. S. (1995). The relationship between work experience and job performance: A conceptual and meta-analytic review. Personnel Psychology, 48(4), 887–910. https://doi.org/10.1111/j.1744-6570.1995.tb01785.x
  • McDaniel, M. A., Schmidt, F. L., & Hunter, J. E. (1988). Job experience correlates of job performance. Journal of Applied Psychology, 73(2), 327–330. https://doi.org/10.1037/0021-9010.73.2.327
  • Sturman, M. C. (2003). Searching for the inverted U-shaped relationship between time and performance: Meta-analyses of the experience/performance, tenure/performance, and age/performance relationships. Journal of Management, 29(5), 609–640. https://doi.org/10.1016/S0149-2063%2803%2900028-X
  • Choudhry, N. K., Fletcher, R. H., & Soumerai, S. B. (2005). Systematic review: The relationship between clinical experience and quality of health care. Annals of Internal Medicine, 142(4), 260–273. https://doi.org/10.7326/0003-4819-142-4-200502150-00008
  • Tsugawa, Y., Newhouse, J. P., Zaslavsky, A. M., Blumenthal, D. M., & Jena, A. B. (2017). Physician age and outcomes in elderly patients in hospital in the US: Observational study. BMJ, 357, j1797. https://doi.org/10.1136/bmj.j1797
  • Spengler, P. M., White, M. J., Ægisdóttir, S., Maugherman, A. S., Anderson, L. A., Cook, R. S., Nichols, C. N., Lampropoulos, G. K., Walker, B. S., Cohen, G. R., & Rush, J. D. (2009). The Meta-Analysis of Clinical Judgment Project: Effects of experience on judgment accuracy. The Counseling Psychologist, 37(3), 350–399. https://doi.org/10.1177/0011000006295149
  • Kahneman, D., & Klein, G. (2009). Conditions for intuitive expertise: A failure to disagree. American Psychologist, 64(6), 515–526. https://doi.org/10.1037/a0016755
  • Ericsson, K. A. (2004). Deliberate practice and the acquisition and maintenance of expert performance in medicine and related domains. Academic Medicine, 79(10 Suppl.), S70–S81. https://doi.org/10.1097/00001888-200410001-00022
  • Ericsson, K. A. (2008). Deliberate practice and acquisition of expert performance: A general overview. Academic Emergency Medicine, 15(11), 988–994. https://doi.org/10.1111/j.1553-2712.2008.00227.x
  • Kini, T., & Podolsky, A. (2016). Does teaching experience increase teacher effectiveness? A review of the research. Learning Policy Institute. https://learningpolicyinstitute.org/product/teaching-experience-report
  • Matsuo, M., Hosoi, K., Yoshino, Y., & Kusumi, T. (1999). Procedural knowledge and sales performance: The mediating effect of years of experience and the process of knowledge acquisition (営業の手続的知識と業績：経験年数の媒介効果と知識獲得プロセス). Journal of Marketing & Distribution (流通研究), 2(1), 43–57. https://doi.org/10.5844/jsmd.2.43 (in Japanese)

Practice time and scheduling

  • Ericsson, K. A., & Harwell, K. W. (2019). Deliberate practice and proposed limits on the effects of practice on the acquisition of expert performance: Why the original definition matters and recommendations for future research. Frontiers in Psychology, 10, 2396. https://doi.org/10.3389/fpsyg.2019.02396
  • Gladwell, M. (2008). Outliers: The story of success. Little, Brown and Company. https://openlibrary.org/works/OL5749847W
  • Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354–380. https://doi.org/10.1037/0033-2909.132.3.354
  • Donovan, J. J., & Radosevich, D. J. (1999). A meta-analytic review of the distribution of practice effect: Now you see it, now you don’t. Journal of Applied Psychology, 84(5), 795–805. https://doi.org/10.1037/0021-9010.84.5.795
  • Driskell, J. E., Willis, R. P., & Copper, C. (1992). Effect of overlearning on retention. Journal of Applied Psychology, 77(5), 615–622. https://doi.org/10.1037/0021-9010.77.5.615
  • Rohrer, D., & Taylor, K. (2006). The effects of overlearning and distributed practise on the retention of mathematics knowledge. Applied Cognitive Psychology, 20(9), 1209–1224. https://doi.org/10.1002/acp.1266
  • Soderstrom, N. C., & Bjork, R. A. (2015). Learning versus performance: An integrative review. Perspectives on Psychological Science, 10(2), 176–199. https://doi.org/10.1177/1745691615569000

Working hours, output, errors, and health

  • Pencavel, J. (2015). The productivity of working hours. The Economic Journal, 125(589), 2052–2076. https://doi.org/10.1111/ecoj.12166
  • Collewet, M., & Sauermann, J. (2017). Working hours and productivity. Labour Economics, 47, 96–106. https://doi.org/10.1016/j.labeco.2017.03.006
  • Landrigan, C. P., Rothschild, J. M., Cronin, J. W., Kaushal, R., Burdick, E., Katz, J. T., Lilly, C. M., Stone, P. H., Lockley, S. W., Bates, D. W., & Czeisler, C. A. (2004). Effect of reducing interns’ work hours on serious medical errors in intensive care units. New England Journal of Medicine, 351(18), 1838–1848. https://doi.org/10.1056/NEJMoa041406
  • Lockley, S. W., Cronin, J. W., Evans, E. E., Cade, B. E., Lee, C. J., Landrigan, C. P., Rothschild, J. M., Katz, J. T., Lilly, C. M., Stone, P. H., Aeschbach, D., & Czeisler, C. A. (2004). Effect of reducing interns’ weekly work hours on sleep and attentional failures. New England Journal of Medicine, 351(18), 1829–1837. https://doi.org/10.1056/NEJMoa041404
  • Kivimäki, M., Jokela, M., Nyberg, S. T., Singh-Manoux, A., Fransson, E. I., Alfredsson, L., et al. (2015). Long working hours and risk of coronary heart disease and stroke: A systematic review and meta-analysis of published and unpublished data for 603 838 individuals. The Lancet, 386(10005), 1739–1746. https://doi.org/10.1016/S0140-6736%2815%2960295-1
  • Pega, F., Náfrádi, B., Momen, N. C., Ujita, Y., Streicher, K. N., Prüss-Üstün, A. M., et al. (2021). Global, regional, and national burdens of ischemic heart disease and stroke attributable to exposure to long working hours for 194 countries, 2000–2016: A systematic analysis from the WHO/ILO Joint Estimates of the Work-related Burden of Disease and Injury. Environment International, 154, 106595. https://doi.org/10.1016/j.envint.2021.106595
  • Ahmed, N., Devitt, K. S., Keshet, I., Spicer, J., Imrie, K., Feldman, L., Cools-Lartigue, J., Kayssi, A., Lipsman, N., Elmi, M., Kulkarni, A. V., Parshuram, C., Mainprize, T., Warren, R. J., Fata, P., Gorman, M. S., Feinberg, S., & Rutka, J. (2014). A systematic review of the effects of resident duty hour restrictions in surgery: Impact on resident wellness, training, and patient outcomes. Annals of Surgery, 259(6), 1041–1053. https://doi.org/10.1097/SLA.0000000000000595
  • Bilimoria, K. Y., Chung, J. W., Hedges, L. V., Dahlke, A. R., Love, R., Cohen, M. E., et al. (2016). National cluster-randomized trial of duty-hour flexibility in surgical training. New England Journal of Medicine, 374(8), 713–727. https://doi.org/10.1056/NEJMoa1515724
  • Silber, J. H., Bellini, L. M., Shea, J. A., Desai, S. V., Dinges, D. F., Basner, M., et al. (2019). Patient safety outcomes under flexible and standard resident duty-hour rules. New England Journal of Medicine, 380(10), 905–914. https://doi.org/10.1056/NEJMoa1810642
  • Basner, M., Asch, D. A., Shea, J. A., Bellini, L. M., Carlin, M., Ecker, A. J., et al. (2019). Sleep and alertness in a duty-hour flexibility trial in internal medicine. New England Journal of Medicine, 380(10), 915–923. https://doi.org/10.1056/NEJMoa1810641

Experiential learning at work

  • Clardy, A. (2018). 70-20-10 and the dominance of informal learning: A fact in search of evidence. Human Resource Development Review, 17(2), 153–178. https://doi.org/10.1177/1534484318759399
  • DeRue, D. S., & Wellman, N. (2009). Developing leaders via experience: The role of developmental challenge, learning orientation, and feedback availability. Journal of Applied Psychology, 94(4), 859–875. https://doi.org/10.1037/a0015317
  • Nakahara, J. (2012). The workplace as a learning environment: Where management research and learning research intersect (学習環境としての「職場」：経営研究と学習研究の交差する場所). The Japanese Journal of Labour Studies (日本労働研究雑誌), 618, 35–45. https://www.jil.go.jp/institute/zassi/backnumber/2012/01/pdf/035-045.pdf (in Japanese)
  • Tanaka, S., & Nakahara, J. (2017). An empirical study on learning by middle management through the experience of new business incubation (新規事業創出経験を通じた中堅管理職の学習に関する実証的研究). Japanese Journal of Administrative Science (経営行動科学), 30(1), 13–29. https://doi.org/10.5651/jaas.30.13 (in Japanese)

Author: Shuichiro Ogawa (Design Researcher / Design Consultant) About me →