Personality and Remote Work: What the Effect Sizes Say

Short answer: remote and hybrid work are not a personality type. Across 46 studies and 12,883 employees, telecommuting had small, mostly beneficial effects—more perceived autonomy, slightly higher job satisfaction, slightly less work–family conflict—and no general damage to coworker relationships. That coworker cost showed up when people were away more than about half the week. Personality changes the picture at the margins, not as a pass/fail. Extraversion’s usual advantages in engagement and satisfaction shrink as remote days increase, and loneliness in general is strongly tied to lower Extraversion and higher Neuroticism. Conscientiousness still predicts performance in ordinary jobs, but enforced remote work and fully remote weeks can erode that advantage. Read the trait bars as questions about social load, structure, and stress, then judge the actual arrangement.

This post puts the classic telecommuting meta-analysis next to later personality studies, maps the traits onto the letters PersonalityCareers reports, and shows how to use a result when a listing says remote or hybrid.

What remote work does, on average

Gajendran and Harrison (2007) meta-analysed telecommuting in natural settings—mostly managers and professionals, and mostly before smartphones and pandemic lockdowns. Estimated correlations corrected for unreliability were:

A correlation of .10 with job satisfaction is real and small: about 1% of the variance. The uncorrected satisfaction correlation was .09, with a 95% confidence interval from .07 to .11. Perceived autonomy fully accounted for the satisfaction link in their mediation tests, and partly accounted for the links with supervisor-rated performance, turnover intent, and role stress. The arrangement helps when it actually increases control over where and when the work gets done.

Two results are easy to over-read. Self-rated performance was essentially zero (ρ = .01). The larger performance figure (ρ = .19) comes from four studies and 484 people, and supervisors may have offered telecommuting to people who were already performing well. Coworker relationships were about zero overall (95% confidence interval for the uncorrected correlation −.03 to .03). Intensity changed that. High-intensity telecommuting—more than 2.5 days a week—accentuated the drop in work–family conflict and harmed coworker relationships (uncorrected r = −.19, 95% confidence interval −.30 to −.08). Low-intensity telecommuting was unrelated to coworker relationship quality. Intensity did not change the autonomy benefit or the supervisor-relationship benefit.

That is the hybrid cutoff in this literature: a day or two away can carry the small autonomy and satisfaction gains without the coworker cost that shows up past half the week. It is an average from older studies. It is not a rule that every team should be in the office three days.

Isolation is a work condition, and a personality prior

Professional isolation is the worry the average correlations do not capture, because those correlations were near zero for coworkers. Golden, Veiga, and Dino (2008) surveyed 261 professional teleworkers and their managers. Greater professional isolation predicted lower manager-rated performance (β = −.13, about 2% of the variance) and, contrary to the authors’ hypothesis, lower turnover intentions (β = −.27). More days teleworking strengthened isolation’s effect on those outcomes. More face-to-face contact and access to communication technology weakened it. Isolation here is a feature of the arrangement and the contact pattern, not a trait score.

Who feels lonely in general is a trait question, and it is a large one. Buecker, Maes, Denissen, and Luhmann (2020) meta-analysed 113 studies (N = 93,668). Bivariate correlations with loneliness were:

Extraversion and Neuroticism were the large associations in that review’s own benchmarks. When all five traits were entered together, every trait except Openness still had a unique link with loneliness. This meta-analysis is not a remote-work study. It is the prior: people lower in Extraversion and higher in Neuroticism report more loneliness across settings, so a remote week that removes casual contact is a sharper test for them. It does not mean they will be less effective. Golden and colleagues’ performance link was with felt professional isolation, and that link was small.

Extraversion: the advantage is harder to keep

In ordinary jobs, Extraversion’s clearest performance links are in socially interactive work, and its link with job satisfaction is one of the two Big Five correlations that generalised (true-score r = .25; Judge, Heller & Mount, 2002). Remote studies ask what happens to that advantage when the social cues thin out.

Evans, Meyers, van de Calseyde, and Stavrova (2022) followed 974 employees working from home because of COVID-19, in four waves from May to August 2020. They used the HEXACO model. HEXACO extroversion is a close relative of Big Five Extraversion; it is not an MBTI letter, and HEXACO emotionality is not the same scale as Neuroticism. Traits and outcomes were standardised. Time ran from 0 at the first wave to 1 at the last, so a Time × trait coefficient is the extra change across the study for someone one standard deviation higher on the trait.

At the first wave, higher extroversion and higher conscientiousness went with better outcomes. Standardised coefficients were .16 and .19 for self-reported performance, .16 and .15 for engagement, .16 and .10 for job satisfaction, and −.09 and −.07 for burnout (extroversion first, then conscientiousness). Across the next three months those advantages eroded:

People at or above the average on those traits deteriorated. People slightly below average did not change reliably. People lower still—for example, a standard deviation or more below the mean—improved. Turnover intentions did not follow the same pattern. On average, self-reported performance itself declined (time coefficient −.11 in that model); engagement, satisfaction, burnout, and turnover did not show a reliable average trend.

Read this as a pandemic lockdown study, not as a hybrid policy. Remote work was enforced. Two of the seven extroversion and conscientiousness interactions were no longer significant when people who failed an attention check were dropped, so the pattern is consistent and not equally precise for every outcome.

A later, non-lockdown panel points the same way for engagement, with a thinner fully remote group. Olsen, Fu, and Jensen (2024) followed 801 Norwegian employees for a year. Extraversion predicted work engagement (β = 0.25), innovative work behaviour (β = 0.26), and general health (β = 0.16). More remote days weakened the extraversion–engagement link (interaction B = −0.11, 95% confidence interval −0.20 to −0.02). Comparing five remote days with none, high-extraversion employees showed a large drop in engagement (B = −0.94). That contrast sits on a skewed sample: 65.9% never worked remotely, and only 2.6% did so five days a week. Treat the five-day figure as a warning about full-week remote work in a small subgroup, not as a population effect size.

Wanting remote work is a different question, and Extraversion does not answer it. Clark, Karau, and Michalisin (2012) asked 333 business students about telecommuting attitudes. Controlling for demographics, Extraversion’s standardised coefficient was .025 and not significant. Agreeableness was .133 and emotional stability was −.144 (students lower in emotional stability viewed telecommuting more favourably). Personality as a set added 2.9% of variance, and that increment was only marginally significant (p = .082). Students who had actually telecommuted before saw more challenges and fewer benefits. Attitude studies are not performance studies, and a student sample is not a workforce.

Among people already teleworking, communication channel still matters, and the published account does not support a letter-to-medium rule. Smith, Patmos, and Pitts (2018) surveyed 384 U.S. teleworkers. Extraversion, Openness, Agreeableness, and Conscientiousness were positively correlated with job satisfaction; the abstract does not report the coefficients, so this post does not invent them. Openness moderated the links of phone and video communication with job satisfaction, and Agreeableness moderated the phone link. That is evidence that the useful medium is not the same for everyone. It is not a finding that introverts should avoid video or that extraverts should live on it.

Conscientiousness, structure, and the fully remote week

In ordinary jobs, Conscientiousness is the Big Five trait with the most consistent performance correlation. Barrick and Mount (1991) estimated true-score correlations of about .20 to .23 across occupational groups, with a mean of .22 across criteria. The detailed breakdown is in Big Five vs MBTI: What Each One Predicts for Career Decisions. Remote studies do not replace that result. They ask whether the link holds when the office’s external structure is thinner.

Evans and colleagues’ Time × Conscientiousness coefficients above are the sharpest published answer for enforced remote work: conscientious employees started ahead on performance, engagement, and satisfaction, then lost ground over three months. The performance interaction (−.29) was the largest trait-by-time effect in that table. Olsen and colleagues found a narrower version for health. Conscientiousness predicted general health overall (β = 0.11), but at five remote days a week that relation was negative (B = −0.52, 95% confidence interval −0.96 to −0.08). The same sample-composition caveat applies: almost everyone in the panel was not fully remote. Remote days did not moderate Conscientiousness’s other outcomes in a way the authors highlighted, and they did not moderate Agreeableness, Neuroticism, or intellect/imagination.

Clark and colleagues did find that Conscientiousness predicted a preference for the work-style side of telecommuting (their Work Preferences subscale), even though it did not predict overall attitudes. Preferring the arrangement and keeping a performance advantage inside it are separate results. A conscientious person can want the autonomy and still miss the shared schedule that makes follow-through easier.

Neuroticism, Openness, and the letters

Neuroticism is the trait this site reports outside the four letters, as the Assertive/Turbulent identity. It has the strongest generalisable Big Five link with job satisfaction in ordinary jobs (true-score r = −.29; Judge, Heller & Mount, 2002), unpacked in Does Personality Predict Job Satisfaction?. It also has one of the two large loneliness correlations above (r = .36). In the Norwegian panel, Neuroticism predicted lower engagement (β = −0.16), worse general health (β = −0.21), and more sick leave (β = 0.23), and remote days did not change those links. The practical reading is not “turbulent people cannot work from home.” It is that the stress and loneliness questions do not go away when the commute does, and the arrangement did not, in that study, turn them off.

Openness is the trait behind S/N, and it is not a remote-performance score. Anderson, Kaplan, and Vega (2015) followed 102 employees at a U.S. government agency and compared days. People reported more job-related positive affect and less job-related negative affect on telework days than on office days. The positive-affect benefit was stronger for people higher in Openness, lower in rumination, and more socially connected outside work. That is a within-person mood finding in one agency. It is a reason to ask whether a remote day feels like useful variety, not a reason to treat an N letter as a remote-work qualification.

The letters on a PersonalityCareers result are a vocabulary for four of these traits, not a fifth set of findings. The free test scores a Big Five–style questionnaire, with items from the public-domain International Personality Item Pool (Goldberg, 1999), and maps four scores onto MBTI-style letters using McCrae and Costa’s (1989) correspondences. E/I tracks Extraversion closely. S/N tracks Openness closely. T/F and J/P line up with Agreeableness and Conscientiousness only partly, and J/P also shares variance with Openness. Neuroticism is not in the classic four letters. The correlations and the reason a midpoint cut throws information away are in Big Five vs MBTI, and the scoring detail is on Test Methodology and in What Our Personality Test Measures.

Reviews of MBTI-style type find little evidence that the four-letter code predicts job performance (Furnham, 2018). The Manual presents type as preference, not capability or vocation (Myers, McCaulley, Quenk & Hammer, 1998). Nothing in the remote-work studies above promotes the code to a location filter.

How to use this when a job is remote or hybrid

1. Do not treat a type as a remote-work verdict. The average telecommuting effects are small, and the personality effects that are specific to remote time come from longitudinal studies, not from a large remote-only meta-analysis. Your manager, the task, and whether you control your schedule will usually matter more than a letter.
2. Use Extraversion for a social-load question, not a productivity label. Higher Extraversion predicts less loneliness in general (r = −.37) and a satisfaction advantage in ordinary jobs. Those advantages were harder to keep as remote time rose, especially in enforced and full-week arrangements. On the Career page for your letters (from Browse Types), read the work-environment section and ask how much of the week is actually shared with other people.
3. Use the half-week cutoff as a hypothesis about coworkers, not a law. In Gajendran and Harrison, coworker relationships suffered past about 2.5 days away and did not, on average, below that. Ask the listing whether “hybrid” means one or two days away or means a full week with a badge swipe on Fridays.
4. Use Conscientiousness, and only partly the J/P letter, for a structure question. The ordinary-job performance correlation is about .22. Enforced remote work and fully remote weeks are where that advantage shrank or, for health in one small fully remote contrast, reversed. If your Conscientiousness bar is high, ask who sets the week’s plan when nobody is in the same room. A J or P is a partial read of that bar.
5. Use Neuroticism / Assertive–Turbulent for stress and loneliness, whether or not the role is remote. The satisfaction and loneliness correlations are larger than the telecommuting main effects. Remote days did not remove Neuroticism’s links with engagement, health, or sick leave in the Norwegian panel. Read the stress section on the Career page, and ask what recovery looks like when work is in the same rooms as the rest of life.
6. Treat autonomy as the mechanism. Telecommuting’s satisfaction benefit ran through perceived control. A remote role that specifies hours, tools, and presence on camera all day is not the arrangement in the 2007 meta-analysis. A hybrid role with real control over focus time is closer to it.
7. Notice the medium instead of assigning one from a letter. Phone and video did not relate to teleworker satisfaction the same way for every trait (Smith, Patmos & Pitts, 2018). Professional isolation’s small performance cost was lower when people still met face to face and had usable communication tools (Golden, Veiga & Dino, 2008).
8. Decide from the listing. How to Job-Search by MBTI Type Without Boxing Yourself In keeps letters as environment hypotheses. Location—office, hybrid, or remote—is one more environment hypothesis, checked against the trait bars and the actual week.

Frequently asked questions

Does personality predict who does well working remotely?

Modestly, and not as a type. Telecommuting itself has small average benefits, on the order of ρ = .10 for job satisfaction and ρ = .22 for perceived autonomy (Gajendran & Harrison, 2007). Extraversion and conscientiousness predict better engagement and satisfaction at a starting point, and those advantages have shrunk as remote time increased in a lockdown panel (Evans et al., 2022) and, for extraversion and engagement, in a later Norwegian panel (Olsen, Fu & Jensen, 2024). There is not yet a remote-only meta-analysis on the scale of the ordinary-job performance reviews.

Are introverts better at remote work?

Not as a rule. People lower in extroversion improved over three months of enforced remote work, while people higher in extroversion started ahead and then lost that lead (Evans et al., 2022). Extraversion’s positive link with later work engagement also weakened as remote days increased (Olsen, Fu & Jensen, 2024). Lower Extraversion still predicts more loneliness in general (r = −.37; Buecker et al., 2020). The useful question is whether the week’s contact matches your social energy, not whether an I letter is a remote qualification.

Does conscientiousness help when you work from home?

In ordinary jobs, yes, at about ρ = .22 with performance (Barrick & Mount, 1991). Under enforced remote work, more conscientious employees became less productive, less engaged, and less satisfied over three months relative to their strong start (Evans et al., 2022). In one panel, the conscientiousness–health link was negative among people remote five days a week, a small slice of that sample (Olsen, Fu & Jensen, 2024). Keep the ordinary-job finding, and ask whether a fully remote week still has enough structure for follow-through.

Does remote work cause isolation, and who feels it?

The average coworker-relationship effect of telecommuting was about zero, and it turned negative past about 2.5 days away (Gajendran & Harrison, 2007). Among teleworkers, feeling professionally isolated predicted slightly lower manager-rated performance (β = −.13), more so with more remote time, and less so with face-to-face contact and better communication tools (Golden, Veiga & Dino, 2008). Separately, loneliness in general is strongly associated with lower Extraversion and higher Neuroticism (Buecker et al., 2020). Isolation is both a design problem and a trait prior.

Can my MBTI type tell me whether to take a remote job?

No. The four letters are a shorthand for four trait scores, and the trait with the strongest satisfaction and loneliness links—Neuroticism—is outside those letters (McCrae & Costa, 1989; Judge, Heller & Mount, 2002; Buecker et al., 2020). E/I is the letter closest to the remote-time findings, and even that evidence is about slopes and changes, not a hiring cutoff. Use the bars, then the listing.

Start here

• Take the Free Test and read Extraversion, Conscientiousness, and the Neuroticism / Assertive–Turbulent result before the four letters
• Browse Types and open the Career page work-environment and stress sections for your letters
• Read Test Methodology for how scores map to letters
• Read What Our Personality Test Measures for what the bars are and are not
• Read Does Personality Predict Job Satisfaction? for the ordinary-job satisfaction correlations
• Read Big Five vs MBTI for the letter–trait correlations and the performance evidence
• Read How to Job-Search by MBTI Type before treating location as a type filter

Personality sets a prior about contact, structure, and stress. The week’s autonomy and the people you still see decide most of whether remote or hybrid will fit.