Management Education & Development MED

---

Hi Dan,

 

Finally a meta-analysis that compares feelings of emotional efficacy and performance to G (General mental intelligence) and performance. After one partials out G, little predictable  performance variance remains. Why are we not surprised ?

 

My career research showed when G has too little variance, leader ship making skills predict speed to promotion over 23 years for college graduates.

 

Leadership-Making excellence (LMX) may measure interpersonal efficacy better than self-report. When both agree on shared LMX, skills of  each are validated by self and other reports and predict performance and engagement the best. Clearly, manager's promotions depend on interpersonal efficacy with the leader.

 

Great work.

 

George

 

In a message dated 1/14/2016 10:26:29 P.M. Central Standard Time, d5n@UIUC.EDU writes:

Hi Rob and John,

Good questions. After controlling for cognitive ability, ability EI is an extremely weak predictor of job performance (change in R^2 = .007; see Joseph & Newman, 2010). Also, ability EI does not predict job performance over and above cognitive ability and the Big Five together (change in R^2 = .002, Joseph & Newman; and change in R^2 = .004, O'Boyle et al.). The recently-updated meta-analysis (Joseph, Jin, Newman, & O'Boyle, 2015) did not explicitly address this particular question, but the effect sizes for ability EI were largely the same as in Joseph and Newman (2010), and O'Boyle et al. (2011); so I am confident that ability EI still does not strongly predict job performance, on average, after you control cognitive ability and the Big Five.

If you are looking for exceptions to the average, there might be a subset of jobs (e.g., high emotional labor jobs) for which ability EI predicts job performance a little better (the Joseph & Newman analyses suggest that incremental validity rises to 2% [standardized regression weight for ability EI = .13, controlling for g and Big 5] for high emotional labor jobs). And there might also be incremental validity for outcomes other than overall job performance (although I admittedly do not have data on this).

Mixed EI, on the other hand, robustly predicts job performance after controlling cognitive ability and the Big Five (change in R^2 = .142 = 14%, Joseph & Newman; change in R^2 = .068 = 7%, O'Boyle et al.). The new meta-analysis explained why the O'Boyle et al. estimates for mixed EI differed somewhat from the Joseph & Newman estimates (i.e., O'Boyle et al. had used a more inclusive definition of job performance that coded things like self-rated performance and student performance as "job performance", which allowed them to have a bigger sample size for the meta-analysis; whereas Joseph and Newman had used a narrower operationalization that involved only supervisor ratings of job performance, and therefore had a smaller sample size).

 

One of the primary contributions of the 2015 meta-analysis was to offer one possible explanation for why mixed EI predicts job performance so well, above and beyond cognitive ability and the Big Five. Specifically, it is plausible that mixed EI measures contain some items that measure general self-efficacy and self-rated performance, in addition to items sampled from the personality and ability construct domains.

 

The paper has a ton of additional details, if you are interested.

 

Hope this helps,

Dan 

On Thu, Jan 14, 2016 at 10:14 AM, Tett, Robert <robert-tett@utulsa.edu> wrote:

That's interesting, Dan. Can you offer anything on how ability EI predicts performance after controlling for cognitive ability?

Rob Tett

UTulsa

From: Dan Newman <d5n@UIUC.EDU>
Reply-To: Emotion in Organizations <EMONET-L@AOMLISTS.PACE.EDU>
Date: Thursday, January 14, 2016 at 8:25 AM
To: Emotion in Organizations <EMONET-L@AOMLISTS.PACE.EDU>
Subject: [EMONET-L] Emotional Intelligence and Job Performance, New Meta-Analysis

Hi Emonet,

 

I am writing to let the EI researchers know about a new meta-analysis describing the relation between EI and job performance, recently published in the Journal of Applied Psychology. This meta-analysis (by Joseph, Jin, Newman, and O'Boyle) updates two previous meta-analyses (by Joseph & Newman, 2010, and O'Boyle et al., 2011), and resolves the discrepancy between them.

  

The new meta-analytic estimates of the EI-job performance relationship (from Joseph, Jin, Newman, & O'Boyle) are:

 

Mixed EI-Job Performance (supervisor-rated): r = .23, r_corrected = .29

 

Ability EI- Job Performance (supervisor-rated): r = .17, r_corrected = .20

 

Mixed EI-Job Performance (objective/results perf.): r = .16, r_corrected = .17

   

Beyond the meta-analytic updates, the paper also advances one explanation for why Mixed EI measures predict job performance more strongly than Ability EI measures do. We provide an indirect test of the idea that Mixed EI measures, on average, have likely borrowed/sampled some content from measures of other well-known psychological constructs (e.g., Emotional Stability, Extraversion, Self-Rated Performance, General Self-Efficacy). After controlling for measures of these other psychological constructs, the Mixed EI-job performance relationship goes to zero.

 

FYI,

Dan

 

Joseph, D. L., Jin, J., Newman, D. A., & O'Boyle, E. H. (2015). Why does self-reported emotional intelligence predict job performance? A meta-analytic investigation of mixed EI. Journal of Applied Psychology, 100, 298-342.