The Rising Return to Non-Cognitive Skill [pdf] | Hacker News Reader

The Rising Return to Non-Cognitive Skill [pdf]

ftp.iza.org

135 pointsluu60 comments

I had to look the defninition of Non-Cognitive skills. Below is an example.

https://blog.udemy.com/non-cognitive-skills/

Once I understood what they were I realize that I have encountered the behavior in my life anecdotally. Also, I realize the ones that I am deficient in.

EDIT: The study has a definition that I missed when reading it. See tboyd47's comment below.

Thanks for this. I was wondering the same thing. Terrible term, for sure.

It doesn't seem that the skills in that link are relevant to this study:

> The non-cognitive score is based on behavioral questions in a 20-minute interview with a trained psychologist. On the basis of the interview, the draftee is scored along four separate dimensions (see Mood et al. (2012)): (i) social maturity, (ii) psychological energy (e.g., focus and perseverance), (iii) intensity (e.g., activation without external pressure) and (iv) emotional stability (e.g., tolerance to stress). There is also an overall non-cognitive score on a Stanine scale, which ranges from 1 to 9.

To clarify, this quote (posted by tboyd47) is the text describing "non-cognitive skills" from the body of the linked PDF. It seems somewhat related to the concepts shared on Udemy, but it's obviously better to use the definition provided in the paper itself.

They seem to define "non-cognitive skills" as "social skills". Not manual skills, such as those needed for carpentry. This is confusing. Is this a translation from Swedish?

Not just social skills: for example, intensity and stress tolerance. I think "emotional skills" would be a better way to describe what they're getting at.

Yes, it's confusing and no, it isn't due to a translation. There is quite a bit of other unnecessarily convoluted jargon and grammar in the report as well, so I think it's just that one or more of the authors has misguided ideals about academic writing style.

>Not manual skills, such as those needed for carpentry

Those skills are still cognitive. You need to know what you're doing to do carpentry. It's not like your hand moves by itself.

In terms of these I find that (ii), (iii), and (iv) do seem to fit the term "non-cognitive" in that they're things we build up with experience and practice but don't really think about consciously. But social maturity surprises me, perhaps it's a technical term that means something different than I think, because at least for me maturing took quite a bit of thought. The "overall" score seems kind of mysteriously thrown in.

Thanks, I initially took it to mean soft skills more along the lines of interpersonal relations.

The definition of "non-cognitive skill" actually used by the researchers does seem to include some interpersonal aspects such as "social maturity."

Wait, how are problem solving and critical thinking not cognitive skills?

In my experience, non-cognitive skills are definitely a big boost to your career. However, the real need is for people with both cognitive AND non-cognitive skills: if you can understand something intrinsically through your cognitive skills, you can leverage your non-cognitive skills to lead a group of people with high cognitive skills.

That leadership piece is what's in demand; and IMO that requires strength in both areas.

Not from my experience.

How so? I've heard what OP said often enough I'd call it folk knowledge. What part do you disagree with?

> That leadership piece is what's in demand; and IMO that requires strength in both areas.

I'm not the person you asked but for me, the idea that leadership is more in demand than intelligence doesn't ring true (anecdotally, of course). In the course of my career, I've met more than enough capable leaders, but finding people who meet the bar for intelligence is much harder. At some level, this is probably due to the fact that you need far less leadership than you do intelligence: by definition, leaders are a fraction of the workforce, but intelligence is useful at pretty much every level in the org (for white-collar work).

I guess I'm lucky enough that I've been in companies that were generally quite well-run; I'm sure the demand for leaders would seem higher in dysfunctional companies where competence in leadership isn't routinely achieved.

Yeah if great leaders are plentiful in your organization I would say that's an outlier in my experience. To be clear, there are often many decent managers in any stable organization...however, if we interpret "leader" to be something more, including the ability to drive change and execute, I think those types are few and far between.

In my experience leaders are seldom judged on being right, and that's the cognitive portion of their job. A leader may be organized, sociable, disciplined, but a leader also has to make choices. A leader in particular must make a choice if there's no consensus, and so there is not an agreement on the right course of action, and presumably there is no identifiable "right" choice. In fact, though, there is usually a right choice, it's just that no one knows for sure what it is. But as events transpire we learn about the effect of decisions (though we still lack the counterfactual), and often it becomes clear if a decision was right or not. I seldom see leaders judged based on the result of their cognitive decisions.

I find there's a strong bias against even attempting to make that judgement. "In hindsight" is used to dismiss criticism: sure, you can say that in hindsight, but I had to make the decision at the time with the information I had. But very few outcomes are actually entirely surprising, we weigh information and risk and probabilities, and these can be improved. An example I remember with particular frustration was Hillary Clinton, repeatedly confirming her decision about the Iraq war given the information she knew at the time. There was lots of information available, she made particular choices about how to weigh the information and risk, and while within her decision framework at the time the choice may have seemed right, clearly her framework was wrong even given her own conclusions. (It's also hard because we expect leaders to always be right, and if they aren't then they shouldn't be leaders, which leaves no room for improvement.)

Culturally, we don't seem to consider information gathering and filtering to be a leadership skill. Leaders are just assumed to be the best informed people at any given moment and we don't look how much work they are doing to get that information.

If a ruler of a country never steps down off the throne and walks through the marketplace to ask the people what they need, how else will he know? Information gathering is connected to humility and manners in my mind. In order to be well-informed you have to be able to interact with different kinds of people and that requires a lot of social maturity, which according to the researchers, is apparently a "non-cognitive" skill.

Information filtering is absolutely a leadership skill. It's the core component of managing, honestly. You don't want your boss to know all of the mistakes that people make, because it's your job as a manager to fix those. If there are structural or systemic problems, yes. But if one of your team members screwed up, you shouldn't be pointing fingers at your subordinates. As an engineering manager, it's your main job to be the crap filter between your executives and your engineers.

Most leaders don't have intelligence. They have political acumen. Or, their intelligence is focused 98% on gaining positions of leadership, not the act itself.

I'd like to work where you work. My experience with management has been "barely hanging on" to "actively damaging the company" with a few "adequate" and 1 or 2 "inspiring" thrown in the mix.

I guess my point is that intelligent leadership is what's really in demand. :)

And yes, I agree. In most white-collar jobs, intelligence is assumed. It's the non-cognitive skills (let's call it emotional intelligence) that are sorely missing.

Like a very strong person can simply lift heavy objects, a very smart person can solve problems through brute intelligence. A smart person can simply employ the Feynman Algorithm, and Think Real Hard at a level that most people physically cannot. Most people must be more clever and crafty, leveraging tangential or unrelated skill sets, and often cooperating.

I refuse this terrible term and invite everyone to use (following zitterbewegung's udemy link) to use "metacognitive skills" instead.

Seconded. Non-cognitive has a pretty clear intuitive meaning so to then define term to mean almost the polar opposite doesn't make any sense to me.

> Non-cognitive has a pretty clear intuitive meaning so to then define term to mean almost the polar opposite doesn't make any sense to me.

FWIW, I didn't interpret it as the polar opposite. I don't see that much of an issue with a model of skills as physical, cognitive, and emotional.

Why do you think "metacognitive skills" is better? The definition of "metacognition" is "awareness and understanding of one's own thought processes." How does that relate to what the researchers were studying?

Meta means self-referential. How are emotional skills "self-referential cognitive skills?"

Meta-x = x about x. E.g. meta-theory is a theory about theories, not a theory about itself.

https://en.wikipedia.org/wiki/Meta

(Douglas Hofstdadter made a career of cheerfully pointing out that a theory about theories must include itself as the kind of theories it is about -- "I am a strange loop" and all that. But still, what the language means is that.)

Right but that doesn't apply here. Meta almost means "higher-level" as you're using it, but emotional processing isn't a higher-level cognitive skill. It's just a different type of cognition.

Meta implies a hierarchy.

I see the problem here. It's too strong to say that being good in using your emotions (say, as an actor) is a form of level 1-cognitive grasp of a level-2 cognitive process.

"Non-epistemic" might be an alternative to "non-cognitive".

GNU/cognitive skills

(but really, the term "cognitive" used in this context boils down to g-loading, so "non-kernel processing" is as apt as anything else — an improvement over "non-cognitive" at the very least.)

How about "precognitive skills"?

Usually refers to precognition in the psychic sense, in english.

I like the para- prefix here. Para-congntive: alongside, auxiliary to cognitive skills.

Paracognitive is good, but I feel it understates the role of emotional/social/etc. skills in performing the tasks we've been over-crediting to cognitive skills altogether. The point that's being made by users of "non-cognitive" is not that emotional skills are good for emotional affairs while cognitive skills are good for computer programming, but rather than being a effective programmer requires dipping on those other skills as well.

I'm beginning to like "non-epistemic". Not everything about being a succesful intellectual or knowledge worker is about discovering the Truth.

It seems I can't edit my own post.

"Metacognitive" has problems. My new favorite is "non-epistemic".

The danger area is conflating lack of success in a particular social setting with noncognitive skills.

I.e., "this person lacks persistence" versus "this person has given up because this social setting is pathological."

I worry that this trend is used to justify unethical behavior by making trait-based attributions of dissenters.

I say this without denying the importance of social skills--but there's a fine line between "socially skillful" and "psychopath", especially as you move up the ladder.

Fascinating study & highly relevant to most people here (since many of us "hackers" are in highly cognitive job roles)

Conclusion:

> Thus, the labor market appears to increasingly value individuals possessing high non-cognitive relative to cognitive skills over time.

How do you increase your stress tolerance? Imagine you are working on bleeding edge stuff at some top company with crazy pace and you don't want to end up with burnout. What would you do in order to improve this aspect? I noticed ramping up fitness training actually decreases my stress resistance as I stress my body and consequently brain much more and am more prone to getting mentally exhausted than when I am not training at all.

Fitness training is a very personal and individual thing. If you're feeling mentally exhausted after working out, you are definitely doing something wrong. I would recommend working with a qualified physical trainer with actual professional qualifications to get a tailored workout regime (be careful since there are many "personal trainers" who really don't know what they're doing).

Also, consider getting involved with yoga and getting a massage. While working out is great for building muscle and endurance, stress your body too much and it will give away.

It's actually quite weird, I do HIIT/strength/cardio training for a few days, I feel awesome, and then one day suddenly I hit a cliff and my brain is shutting down. The thing is I can't predict when it happens; I saw some relation to NO (sometimes adding NO-inducing supplements alleviates issues) though it doesn't seem to work 100% And if you read fitness articles, I am not the only person that has this issue. I mean "neural fatigue" is pretty common, so probably it goes further up with some people like me...

Get a trainer who will take the time to understand your needs and your goals and customize a proper workout routine for you that will modulate in type and intensity throughout the period of a week.

What you're describing sounds like over-training.

Actually the GP is incorrect. It's pretty well understood [1] these days that your body is like a "cup" that can only take a certain amount of stress before it starts getting counter productive. Mental stress and center-nervous-system stress (what weight training does) both contribute so if you already have a lot of one you're forced to have less than the other or everything suffers.

[1] I'm not in a position to hunt links but lookup "Alan Thrall stress" on youtube.

Many thanks, this seems to be pretty close! Anyway, do you know any way to increase the "stress threshold", so that I can withstand more mental + physical stress? Mental stress is given by work, and one doesn't want to look flabby either but rather ripped without getting ill/fatigued etc. I understand a lot of this is genetic, though I am sure there are some ways to help body with it.

Well, this is all new to me as well but the way I've seen it presented (from multiple sources) it seems to be person-specific. If there is a way to increase you "cup size" either I haven't yet seen it (wouldn't be surprising) or it isn't yet known.

Oh well, I guess it's trial and error then. I was actually doing 5 minute cold showers for a year, alongside HIIT/strength training/cardio and bleeding edge mental work, and I guess I was too optimistic I could handle it and went over my limits (which seem to be pretty high already but still firmly in place...)

I'm finding good tips about this in this book: https://www.amazon.com/Peak-Performance-Elevate-Burnout-Scie...

Is your training mostly sub-max cardio-vascular stuff? I think that can increase low-level stress and cortisol.

I have experienced weight training has been a great tool for mood regulation.

(I am not interested in a cardio v. weights flame war)

I was doing near-max HIIT training all the time and the problems started to be evident when I was adding a lot of strength training around it as well. There were some amazing results though - after a few months of HIIT I could drink 3 bottles of milk at once with no issues but prior to that after a few glasses I had some gastrointenstinal issues, likely from lactase deficiency. Also skin gets better, hair grows denser etc.

Try experimenting with dropping milk entirely. I discovered that my stomach/gut health has a huge impact on my mood. When I did away with a number of foods that seemed to annoy it, and ate more fermented stuff, I definitely noticed an improvement. Try a six week vacation from milk, you might be surprised.

You also might not get enough sleep to sustain that load of hiit and weight training. People like to brag about surviving on 6 hours sleep etc, but I realized, while unemployed but training heavily for three months, that my desirable baseload is 9hrs sleep. I admit it sounds indulgent, and I normally make do with 8.

Good luck

The buzzword a year or two back was "grit".

From my own experience non-cognitive skills are becoming rarer, so increasing returns make sense.

Increased internet use seems to be leading to decreased availability of non-cognitive skills.

> Workers with an abundance of non-cognitive skill were increasingly sorted into occupations that were intensive in: cognitive skill; as well as abstract, non-routine, social, non-automatable and offshorable tasks. Such occupations were also the types of occupations which saw greater increases in the relative return to non-cognitive skill.

Isn't this a contradiction? "Increasingly sorted into occupations that were intensive in cognitive skill" and simultaneously "greater increases in the relative return to non-cognitive skill"? So are we just saying they saw the heaviest rise in skills in general?

I didn't read past the abstract, maybe the body is clearer.

It sounds like the quoted excerpt is saying: for occupations intensive in cognitive skill, both the number of people with non-cognitive skills entering the job and their relative salaries increased during the examined time period.

Can these non-cognitive skills be taught? Does anyone know of good resources for building these non-cognitive skills?

So then surely we need some measures to prevent discrimination against those lacking in non-cognitive skills?

Or our education systems need to actually train people in those skills instead of ignoring them.

Reply on news.ycombinator.com