pith. sign in

arxiv: 1906.09463 · v1 · pith:RU5VTOJRnew · submitted 2019-06-22 · 💻 cs.SE

The Connection Between Burnout and Personality Types in Software Developers

Emanuel Mellblom , Isar Arason , Lucas Gren , Richard Torkar This is my paper

Pith reviewed 2026-05-25 18:10 UTC · model grok-4.3

classification 💻 cs.SE
keywords burnoutpersonality traitsFive Factor Modelneuroticismsoftware developersBayesian regressionopen source
0
0 comments X

The pith

Neuroticism is strongly linked to burnout in software developers while other Five Factor Model traits add no power to the model.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper tests whether the Five Factor Model personality traits relate to burnout levels among software developers by surveying open source contributors. It runs a Bayesian linear regression on the responses and reports that only neuroticism shows a strong connection, matching patterns found in other fields. A sympathetic reader would care because the result suggests personality could help flag developers who might need support to avoid burnout. The study collected 47 valid responses using a short personality questionnaire and a standard burnout scale but did not examine work quality or other possible influences such as teamwork.

Core claim

The results from a Bayesian Linear Regression analysis indicate a strong link between neuroticism and burnout confirming previous work, while the other Five Factor Model traits were not adding power to the model. Employers could be aware of, and support, software developers with high neuroticism.

What carries the argument

Bayesian Linear Regression relating Five Factor Model personality traits (from miniaturized IPIP) to burnout scores (from Shirom-Melamed Burnout Measure).

If this is right

  • Neuroticism can be treated as a usable predictor for burnout risk in software development settings.
  • The remaining four Five Factor Model traits do not improve burnout prediction once neuroticism is included.
  • Employers may consider providing targeted support to developers who score high on neuroticism.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same personality-burnout pattern might appear in non-open-source teams if the psychological mechanisms are comparable.
  • Interventions aimed at emotional stability could be tested for their effect on burnout rates in developer populations.
  • Repeating the survey with mandatory rather than voluntary participation would help check for self-selection effects.

Load-bearing premise

The miniaturized International Personality Item Pool and Shirom-Melamed Burnout Measure validly capture personality and burnout in open source software developers and the 47 self-selected responses are free from major selection bias or unmeasured confounders such as teamwork.

What would settle it

A larger study of software developers using the same measures that finds no significant association between neuroticism and burnout would falsify the reported link.

Figures

Figures reproduced from arXiv: 1906.09463 by Emanuel Mellblom, Isar Arason, Lucas Gren, Richard Torkar.

Figure 1
Figure 1. Figure 1: Boxplots of all the constructs in this study. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
read the original abstract

This paper examines the connection between the Five Factor Model personality traits and burnout in software developers. This study aims to validate generalizations of findings in other fields. An online survey consisting of a miniaturized International Personality Item Pool questionnaire for measuring the Five Factor Model personality traits, and the Shirom-Melamed Burnout Measure for measuring burnout, were distributed to open source developer mailing lists, obtaining 47 valid responses. The results from a Bayesian Linear Regression analysis indicate a strong link between neuroticism and burnout confirming previous work, while the other Five Factor Model traits were not adding power to the model. It is important to note that we did not investigate the quality of work in connection to personality, nor did we take any other confounding factors into account like, for example, teamwork. Nonetheless, employers could be aware of, and support, software developers with high neuroticism.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper examines the connection between Five Factor Model personality traits and burnout among software developers. An online survey using a miniaturized International Personality Item Pool (IPIP) and the Shirom-Melamed Burnout Measure was distributed to open source developer mailing lists, obtaining 47 valid responses. Bayesian linear regression analysis is reported to show a strong link between neuroticism and burnout (confirming prior work), while the other four FFM traits did not add explanatory power to the model. The authors explicitly note that work quality was not assessed and that confounders such as teamwork were unaccounted for, with a suggestion that employers support developers high in neuroticism.

Significance. If the reported association is robust, the work provides a domain-specific validation of personality-burnout links in open source software development and offers a practical implication for employer awareness. The choice of Bayesian regression is a methodological strength for handling small samples, and the explicit acknowledgment of limitations (confounders, no work-quality measures) is transparent.

major comments (3)
  1. [Methods (data collection)] Methods section (survey and sample): The central claim rests on Bayesian linear regression with five predictors on n=47 self-selected responses from OSS mailing lists. This sample size is marginal for reliable estimation of multiple coefficients; the paper does not report power analysis, effective sample size after priors, or checks for influential observations, directly threatening whether the neuroticism coefficient reflects a population relationship.
  2. [Results (Bayesian Linear Regression)] Results section (Bayesian regression): No model details, priors, convergence diagnostics, posterior predictive checks, or sensitivity analyses are described. The abstract states the regression 'indicate[s] a strong link' and that other traits 'were not adding power,' but without Bayes factors, credible intervals, or model-comparison metrics, the strength of evidence for the neuroticism effect and the nulls for the remaining traits cannot be evaluated.
  3. [Discussion] Discussion: The authors note unaccounted confounders (e.g., teamwork) and lack of work-quality measures, yet the interpretation that employers 'could be aware of, and support' high-neuroticism developers treats the observed coefficient as actionable. This inference is load-bearing for the applied claim and requires explicit qualification or additional robustness checks against selection bias (neuroticism potentially correlating with survey participation).
minor comments (2)
  1. [Abstract] Abstract: The description of the Bayesian analysis could include at least one quantitative result (e.g., credible interval or R² analogue) rather than a qualitative statement alone.
  2. [Methods] Measures: Brief discussion of the validity and reliability of the miniaturized IPIP and Shirom-Melamed Burnout Measure specifically within open-source developer populations would strengthen the methods.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed comments, which highlight important areas for improving methodological transparency and interpretive caution. We address each major comment below and outline revisions to the manuscript.

read point-by-point responses

  1. Referee: [Methods (data collection)] Methods section (survey and sample): The central claim rests on Bayesian linear regression with five predictors on n=47 self-selected responses from OSS mailing lists. This sample size is marginal for reliable estimation of multiple coefficients; the paper does not report power analysis, effective sample size after priors, or checks for influential observations, directly threatening whether the neuroticism coefficient reflects a population relationship.

    Authors: We agree that n=47 is modest for estimating five coefficients and that the self-selected nature of responses from OSS mailing lists introduces potential bias. Bayesian methods were selected specifically to stabilize estimates in small samples via priors. No pre-collection power analysis was performed, as the study was exploratory. In revision we will add an explicit limitations paragraph on sample size, report posterior effective sample sizes, and include checks for influential observations (e.g., via Pareto-smoothed importance sampling or Cook's distance analogs). revision: partial

  2. Referee: [Results (Bayesian Linear Regression)] Results section (Bayesian regression): No model details, priors, convergence diagnostics, posterior predictive checks, or sensitivity analyses are described. The abstract states the regression 'indicate[s] a strong link' and that other traits 'were not adding power,' but without Bayes factors, credible intervals, or model-comparison metrics, the strength of evidence for the neuroticism effect and the nulls for the remaining traits cannot be evaluated.

    Authors: The submitted version omitted these details for brevity. We will expand the Results section to document the exact model (including priors, e.g., weakly informative defaults), convergence diagnostics (R-hat and ESS), posterior predictive checks, and 95% credible intervals for all coefficients. We will also add Bayes factors or LOO-CV comparisons between the full model and reduced models to quantify evidence for the neuroticism effect versus the null contributions of the other traits. revision: yes

  3. Referee: [Discussion] Discussion: The authors note unaccounted confounders (e.g., teamwork) and lack of work-quality measures, yet the interpretation that employers 'could be aware of, and support' high-neuroticism developers treats the observed coefficient as actionable. This inference is load-bearing for the applied claim and requires explicit qualification or additional robustness checks against selection bias (neuroticism potentially correlating with survey participation).

    Authors: We concur that the employer recommendation must be qualified more explicitly as tentative and correlational. The original text already flags unmeasured confounders; we will revise the Discussion to state that the suggestion is exploratory, does not establish causation, and is subject to selection bias in survey participation. We will add a sentence acknowledging that neuroticism may correlate with response propensity and recommend future work use alternative sampling frames to test robustness. revision: yes

Circularity Check

0 steps flagged

No circularity; purely empirical survey and regression with data-driven results

full rationale

The paper reports an online survey of 47 OSS developers using standardized IPIP and SMBM instruments followed by Bayesian linear regression on the five FFM traits as predictors of burnout. No equations, derivations, or predictions are present that reduce to fitted parameters by construction, nor are there self-definitional loops, load-bearing self-citations, uniqueness theorems, or ansatzes imported from prior work. The central claim is a statistical association observed in the collected data, with explicit caveats about unmeasured confounders; the analysis chain is self-contained against external benchmarks and contains no circular steps.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim depends on the validity of self-report psychological instruments and the representativeness of a small self-selected sample; no new entities are postulated.

free parameters (1)
  • Bayesian regression coefficients
    Fitted to the 47 survey responses to quantify trait-burnout associations.
axioms (1)
  • domain assumption The Five Factor Model traits and Shirom-Melamed Burnout Measure are valid and appropriate for measuring the constructs in this population.
    Required to interpret the regression output as evidence of a meaningful link.

pith-pipeline@v0.9.0 · 5677 in / 1397 out tokens · 39943 ms · 2026-05-25T18:10:13.978603+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

48 extracted references · 48 canonical work pages

  1. [1]

    Ahola, K., Toppinen-Tanner, S., & Sepp ¨anen, J. (2017). Interventions to alleviate burnout symptoms and to support return to work among em- ployees with burnout: Systematic review and meta-analysis. Burnout re- search, 4, 1–11

    work page 2017

  2. [2]

    Armon, G., Shirom, A., & Melamed, S. (2012). The big five personal- ity factors as predictors of changes across time in burnout and its facets. Journal of Personality, 80:2, 403–427

    work page 2012

  3. [3]

    B., Van Der Zee, K

    Bakker, A. B., Van Der Zee, K. I., Lewig, K. A., & Dollard, M. F. (2006). The relationship between the big five personality factors and burnout: A study among volunteer counselors. The Journal of Social Psychology , 146(1), 31–50

    work page 2006

  4. [4]

    Cobb, S. (1976). Social support as a moderator of life stress. Psychoso- matic medicine,

    work page 1976

  5. [5]

    B., Oswald, F

    Donnellan, M. B., Oswald, F. L., Baird, B. M., & Lucas, R. E. (2006). The mini-ipip scales: Tiny-yet-e ffective measures of the big five factors of personality. Psychological Assessment, (pp. 192–203)

    work page 2006

  6. [6]

    Goldberg, L. R. (1992). The development of markers for the big-five factor structure. Psychological Assessment, 4, 26–42

    work page 1992

  7. [7]

    Grant, S., & Langan-Fox, J. (2007). Personality and the occupational stressor-strain relationship: The role of the big five. Journal of occupa- tional health psychology, 12, 20

    work page 2007

  8. [8]

    Jocic, D., Djonovic, N., Krajnovic, D., Stefanovic, S., Stojkov, S., & Ko- cic, S. (2018). Cultural adaptation and examination of metric charac- teristics shirom-melamed burnout questionnaire (smbq) on a sample of pharmacists in serbia. Indian Journal of Pharmaceutical Education and Research, 52, 166–180

    work page 2018

  9. [9]

    A., Heller, D., & Mount, M

    Judge, T. A., Heller, D., & Mount, M. K. (2002). Five-factor model of personality and job satisfaction: A meta-analysis. Journal of Applied Psychology, 87, 530–541

    work page 2002

  10. [10]

    Karasek Jr, R. A. (1979). Job demands, job decision latitude, and mental strain: Implications for job redesign. Administrative science quarterly , (pp. 285–308)

    work page 1979

  11. [11]

    L., & Westerlund, H

    Leineweber, C., Baltzer, M., Magnusson Hanson, L. L., & Westerlund, H. (2012). Work–family conflict and health in swedish working women and men: a 2-year prospective analysis (the slosh study). The European Journal of Public Health, 23, 710–716

    work page 2012

  12. [12]

    Pallant, J., & Ahlborg Jr, G

    Lundgren Nilsson, A., Jonsdottir, I., H. Pallant, J., & Ahlborg Jr, G. (2012). Internal construct validity of the shirom-melamed burnout ques- tionnaire (smbq). BMC Public Health, 12:1, 1–8

    work page 2012

  13. [13]

    McAdams, D. (2006). The five-factor model in personality: A critical appraisal. Journal of personality, 60, 329–361

    work page 2006

  14. [14]

    A., & McGeary, D

    McGeary, C. A., & McGeary, D. D. (2012). Occupational burnout. In R. J. Gatchel, & I. Z. Schultz (Eds.), Handbook of Occupational Health and Wellness chapter 9. (pp. 181–201). New York: Springer

    work page 2012

  15. [15]

    Robert, F., Lefteris, A., Richard, T., & Maria, S. (2010). Links between the personalities, views and attitudes of software engineers. Information and Software Technology, 52, Issue 6, 611–624

    work page 2010

  16. [16]

    Shirom, A. (2018). Burnout and vigor scales (smbm & smvm)

    work page 2018

  17. [17]

    Singh, P., & Suar, D. (2013). Health consequences and bu ffers of job burnout among indian software developers. Psychological Studies , 58, 20–32

    work page 2013

  18. [18]

    T ¨ornroos, M., Hintsanen, M., Hintsa, T., Jokela, M., Pulkki-Råback, L., & Hutri-K ¨ah¨onen, N. (2013). Associations between five-factor model traits and perceived job strain: A population-based study. Journal of Oc- cupational Health Psychology, V ol. 18, No. 4, 492–500

    work page 2013

  19. [19]

    Williams, D. G. (1992). Dispositional optimism, neuroticism, and ex- traversion. Personality and individual differences, 13, 475–477. 5 Appendix .1. Mailing Lists • Apache Open Office, Development Mailing List:https://openoffice. apache.org/mailing-lists.html#development-mailing-list-public • KDE Development: https://www.kde.org/support/mailinglists/ • KDE C...

    work page 1992

  20. [20]

    Are you currently employed as a software engineer? (Yes / No)

    work page

  21. [21]

    Items have the following answer options: Very Inaccurate — 1, 2, 3, 4, 5 — Very Accurate

    How many years of professional software development experience do you have? (None, 1 year, 2-4 years, 5-9 years, 10+ years) IPIP Section. Items have the following answer options: Very Inaccurate — 1, 2, 3, 4, 5 — Very Accurate

    work page

  22. [22]

    I am the life of the party

    work page

  23. [23]

    I sympathize with others’ feelings

    work page

  24. [24]

    I get chores done right away

    work page

  25. [25]

    I have frequent mood swings

    work page

  26. [26]

    I have a vivid imagination

    work page

  27. [27]

    I am not interested in other people’s problems

    work page

  28. [28]

    I often forget to put things back in their proper place

    work page

  29. [29]

    I am relaxed most of the time

    work page

  30. [30]

    I am not interested in abstract ideas

    work page

  31. [31]

    I talk to a lot of di fferent people at parties

    work page

  32. [32]

    I feel others’ emotions

    work page

  33. [33]

    I have di fficulty understanding abstract ideas

    work page

  34. [34]

    I keep in the background

    work page

  35. [35]

    I am not really interested in others

    work page

  36. [36]

    I make a mess of things

    work page

  37. [37]

    Items have the following answer options: Never/Almost Never — 1, 2, 3, 4, 5, 6, 7 — Always /Almost Always

    I do not have a good imagination SMBM Section. Items have the following answer options: Never/Almost Never — 1, 2, 3, 4, 5, 6, 7 — Always /Almost Always

    work page

  38. [38]

    I have no energy for going to work in the morning

    work page

  39. [39]

    I feel physically drained

    work page

  40. [40]

    I feel like my ”batteries” are ”dead”

    work page

  41. [41]

    My thinking process is slow

    work page

  42. [42]

    I have di fficulty concentrating

    work page

  43. [43]

    I feel I’m not thinking clearly

    work page

  44. [44]

    I feel I’m not focused in my thinking

    work page

  45. [45]

    I have di fficulty thinking about complex things

    work page

  46. [46]

    I feel I am unable to be sensitive to the needs of cowork- ers and customers

    work page

  47. [47]

    I feel I am not capable of investing emotionally in cowork- ers and customers

    work page

  48. [48]

    I feel I am not capable of being sympathetic to coworkers and customers 6

    work page