What EQ gets right
This isn't a hit piece on emotional intelligence. The concept pointed at something real.
Emotional awareness — the ability to recognise your own emotional states and those of others — is genuinely valuable. Empathy and social skills predict interpersonal effectiveness. The researchers who developed the original EQ construct (Mayer, Salovey, and Caruso) were describing something meaningful: emotional life matters for cognition, and ignoring it has costs.
The problem is not the observation that emotions affect reasoning. The problem is that the subsequent scientific case — particularly for EQ training and its incremental predictive power — has never matched the cultural enthusiasm.
Where the EQ evidence falls apart
The foundational meta-analysis came from Van Rooy and Viswesvaran in 2004: 69 studies, and EQ predicted job performance at ρ = 0.23. That's modest — below the predictive power of general cognitive ability (ρ ≈ 0.5) and conscientiousness (ρ ≈ 0.3). But the more damaging finding came when researchers controlled for known predictors.
Locke (2005) and Schulte et al. (2004) both found that after controlling for the Big Five personality traits and general cognitive ability, EQ's incremental validity in predicting performance dropped to approximately 1–2% additional variance — statistically marginal, practically negligible. Joseph and Newman's 2010 meta-analysis confirmed the pattern: EQ measures are largely a repackaging of existing personality and intelligence constructs, not a distinct, independently predictive skill.
The construct validity problem runs deeper than job performance. Brackett and Mayer (2003) found that self-report EQ measures and ability-based EQ measures correlate at only r ≈ 0.18— meaning they barely measure the same thing. Self-report EQ, the dominant format in workplace training, is predominantly a personality measure. Ability-based EQ is closer to an intelligence measure. Neither is capturing a cleanly distinct, trainable skill.
Antonakis (2009) reviewed the leadership research specifically and concluded that "EQ is not a valid predictor of leadership" beyond what IQ and personality already predict. The leadership industry's enthusiasm for EQ has outpaced the science by a significant margin.
On training specifically: Mattingly and Kraiger's 2019 meta-analysis of EQ training programmes found an average effect size of d ≈ 0.35, and noted that most effects were short-lived. Training designs rarely produce durable change because they're attempting to alter what is, in large part, a stable dispositional trait — not a trainable process skill.
The metacognition evidence
Against that background, the metacognition research is striking.
De Boer, Donker, and van der Werf (2018) reviewed 67 studies and found metacognitive training produced an effect size of g = 0.63 on academic performance — larger than the effect of homework, larger than class-size reduction, and comparable to one-on-one tutoring. Critically, this effect was measured weeks and months after training concluded — not just immediately after.
Hidayat et al.'s controlled study found an even larger effect — ES ≈ 1.11 — in structured training environments. At this effect size, the trained group outperformed approximately 86% of the untrained group. This is among the largest effects ever measured in cognitive training.
The clinical evidence is equally compelling. Normann and Morina (2021) meta-analysed studies comparing metacognitive therapy (MCT) to Cognitive Behavioural Therapy — the acknowledged gold standard for anxiety and depression. MCT outperformed CBT with an effect size of g = 0.69.
Applied to real-world decision-making: the Good Judgment Project found that structured metacognitive practice produced a +14% gain in forecasting accuracy across 3,000 forecasters — a sustained, measurable improvement in real-stakes prediction.
The structural reason metacognition outperforms EQ
The deeper explanation isn't just that the effect sizes are larger. It's structural.
EQ tells you what you feel. Metacognition determines what you do with it.
Adrian Wells' Metacognitive Model (2009), supported by decades of clinical research, demonstrates that emotional distress — anxiety, rumination, depression — is not primarily driven by emotional states themselves but by metacognitive beliefs about thinking: the conviction that worrying is protective, that rumination is necessary to solve a problem, that intrusive thoughts must be controlled. This is why Metacognitive Therapy outperformed CBT in Normann and Morina's analysis — it targets the operating system, not just individual outputs.
James Gross's process model of emotion regulation (2015) converges on the same conclusion: emotional regulation is itself a metacognitive process. You don't regulate emotions by having feelings about feelings — you regulate them by deploying metacognitive strategies to redirect attention, reappraise situations, and modulate responses.
The practical implication: you can have precise emotional awareness and still make systematically poor decisions if your metacognition is weak. You can be highly empathetic and still be profoundly overconfident. You can recognise your emotions clearly and still fail to update your beliefs when contradicting evidence appears. Metacognitive skill is the layer that determines whether emotional intelligence is actually used effectively.
Think of it this way: EQ is input. Metacognition is the operating system.
What this means practically
If you've spent time and money on EQ workshops, the effects were probably real — particularly for interpersonal awareness and empathy. But the evidence suggests they were likely short-lived, and they did not address the more trainable, more impactful upstream skill.
Metacognition training is specific, measurable, and durable. The gains in calibration, reasoning quality, and belief-updating that structured metacognitive practice produces are not a temporary performance lift — they reflect genuine changes in how people think about their own thinking. And the research suggests those changes persist.
MindFrame trains metacognition directly — and the Emotional Radar mode in particular trains the metacognitive skill most relevant to emotional intelligence: recognising how emotional states distort your reasoning quality in real time, so you can compensate rather than just feel.