EV-0036
The happy-worker claim, at its actual size
Judge, T. A., Thoresen, C. J., Bono, J. E., & Patton, G. K. (2001). Psychological Bulletin, 127(3), 376-407 · Source
What this entitles us to say
A meta-analysis of 312 samples and 54,417 people estimating the job satisfaction-job performance relationship at about rho = .30 - real and moderate, and far smaller than the popular version of the claim.
What it found
312 samples, combined N of 54,417. The satisfaction-performance relationship was estimated at rho = .30. The paper also organizes the prior literature around seven distinct models of how the two might relate, rather than assuming satisfaction causes performance.
Why it matters here
Attitude is the trait most exposed to overclaiming, because the popular literature around it is almost entirely made of overclaims. This entry is the disciplined version: a real, moderate relationship, honestly sized.
The seven-models framing is worth carrying into the module. It is entirely plausible that doing good work produces satisfaction rather than the reverse, or that both follow from something else. A program teaching people to choose their attitude should be candid that the arrow is not settled - and it can be, because the book's claim is not that a good attitude produces results. It is that your attitude is your mental coach: an internal fact you control, whatever it does downstream.
The limits
Correlational, and the direction is genuinely unsettled - the paper's own qualitative half organizes the literature around seven competing models of how satisfaction and performance relate, several of which run the arrow the other way or route both through a third cause. Satisfaction is not the same thing as the book's Attitude, which is closer to a chosen orientation than to contentment; the mapping is ours. Verified to abstract level.
What we do not say
"Satisfaction drives performance." The paper itself lays out seven competing models of the relationship, several of which reverse the arrow.