MIT Unravels the Secrets Behind Collective Intelligence | Hacker News Reader

MIT Unravels the Secrets Behind Collective Intelligence

singularityhub.com

95 pointshinto_ize23 comments

This came up on LW before (http://lesswrong.com/lw/73w/study_on_group_intelligence/); I pointed them (http://lesswrong.com/lw/73w/link_study_on_group_intelligence...) to a previous comment I had seen dissecting the study: http://hbr.org/2011/06/defend-your-research-what-makes-a-tea...

tl;dr: as one would expect, it's not that great and certainly doesn't justify the title.

The study was done in only one perspective, wich is adding women to a group of men. In other words, men perform better in the presence of women. Next should be trying to add women to more women, or maybe adding men to a group of women, in order to have a bigger picture about how the collective intelligence works when mixing genders.

Did you read the article?

They studied a number of things and "... when they controlled for the number of women in a group, it was shown that it was the emotional sensitivity scores which won out."

Whatever its merits or flaws, the study was not simply about adding women to groups of men.

1) I read the article 2) Did they try adding women to a women only group ? 3) Did they try adding men to a women only group ? 4) Did they try adding more women to a mixed group ? 5) Did they try adding more men to a mixed group ? 6) Did they try to remove all men from a group and replace them by women ? 6) Does this have to do with women being in a group, or does this just have to do with the group having elements from both genders collaborating ? The conclusions made in the article ,to me, seems more buzz-seeking than enriching.

I think you're severely misunderstanding what it is they did in this experiment. The overall emphasis was not at all on the gender of the participants. As the article took pains to point out, they formed the groups based on a variety of intelligence scores - not gender.

Initially, they discovered that the most significant factors were "the high average social sensitivity of group members, a high rate of sharing who gets to communicate, and more females". Then they controlled for the number of females in each group, and discovered that the number of females actually didn't matter so much - the women in the study were just more likely (than the men) to have high social sensitivity, so a team with many women would be more likely to do well. But a team consisting of all men would also do very well, as long as you made sure to pick men with high social sensitivity.

These studies always leave out the negative side of social sensitivity: bullying and exclusion. In a task worked on for 5 hours everyone still has on their party manners.

This dovetails to some degree with recent studies showing female asset managers both are less risk-tolerant and provide better overall returns than male asset managers, on average.

Do you have a citation for this?

If true, it's counterintuitive - one would expect the risk-tolerant investors to have higher returns.

The way I remember it, it was simply that men traded more and so the fees (combined with some bad choices) damaged their returns: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.145...

This paper is based on retail investors at a discount brokerage, not asset managers.

Who we would expect to be immune to the issues of their gender... why? I bet if you look, there are similar results for managers, that those who trade more damage their returns.

Because the asset managers who are not immune to the problems of retail investors get fired very quickly.

Just as it is hard to generalize from your average facebook user to programmers, it is hard to generalize from retail investors to professional traders.

I vaguely remember what the GP is talking about. IIRC it basically came down to women being more likely to 'let go' and not pour more good money after bad. I don't remember it being about 'risk-tolerance' specifically. More about women having less 'ego' in the game.

Sorry I can't find the links.

Recipe for success = just add women.

Mad Men for the win. I knew replacing our secretaries with computers would destroy America. Your team doesn't need more developers, you need a Peggy Olson.

Did they try to control for Tea Party members or Christian fundamentalists? I'm not saying this just for snark value; if you don't believe in science or the scientific method, for instance, it would seem a bit hard for you to contribute much in the way of intelligence on such questions. Same thing with "social sensitivity"; I would guess that's not a strength among isolationists or fundamentalists of various types.

Wow I'd really like to see some data to back up these assertions.

Did they correct for poor welfare moms, blue collar union workers, or Muslims?

To think that only Tea party members and Christian fundamentalists are the only ones who might not understand or have a problem with the scientific method is absurd. I really doubt your average 20 year old wearing a Che shirt cares any more about the scientific method than your average Tea Partier.

What a ridiculous question.

Not ridiculous at all, although very predictable that this would be downvoted and controversial.

I didn't say "only" Tea Party folks or fundies, so I'd appreciate not being misquoted. I also didn't say anything about "understanding" the scientific method. I think a lot of those folks understand it quite well; they just choose to cynically lie about it for political reasons, which is the crucial difference here.

You bringing up Muslims or welfare moms says all we need to know about your particular attitudes. It's really not my fault if you have a problem acknowledging the factual reality here. The facts are, the Tea Party is actively hostile to science, the scientific method, and scientific conclusions in general. This isn't my opinion, it's not controversial, and it involves no tricky judgment calls. It's a fact.

This wasn't an argument on whether a Tea Party backed political party would be detrimental to science in general. Your assertion was that we should control for Tea Party members in a particular scientific study.

If what you say is true --That a certain subset of the population, that you find undesirable, will negatively impact the results of the study --Then, I will point out another subset of the population who, given you're apparent political leanings you won't find undesirable, that will impact the study in the same way.

I did this so that hopefully you would see the problem with arbitrarily deciding to exclude a group of people who you disagree with from consideration.

Btw I'm certainly not a conservative. I would also ask you to read my response to the comment below.

I wouldn't use the term "controversial" to describe your post. Instead, I'd say you made a pointless, unnecessary, partisan attempt to drag emotional baggage into the thread. You've been on Hacker News for over a year; you should know better than to engage in such douchebaggery.

>To think that only Tea party members and Christian fundamentalists are the only ones who might not understand or have a problem with the scientific method is absurd. I really doubt your average 20 year old wearing a Che shirt cares any more about the scientific method than your average Tea Partier.

even a 20 years old wearing a Che shirt would see a connection [possibly after several trials] between stepping on a rake and a strange immediate sensation on the forehead and will cease to step on the rake thus manifesting basic grasp of the scientific method. Not so for a Tea Partier - he (or she) will every time think that it is God's will or somebody attacking them and their freedom to step on the rake and the Tea Partier will continue to step on the rake to assert their freedom and will also force others to do so.

I believe that political biases are blinding you to the even distribution of cognitive dissonance--no one political party has a monopoly.

Conservative: "There's no way I'm going to trust those idiots in washington with my healthcare" "I do, however, trust those idiots in washington to fight endless wars, and keep the streets free of the evils of Marijuana"

Liberal: "I don't trust those idiots in Washington. They lied to all of us, so that we would support an unjust war." "I do, however, trust those idiots in Washington to run our healthcare and education"

People on both sides will continue to make decisions without using past experience to attempt to determine the outcome.

Meh, I think it is that people really just don't feel like they have an optimal choice. So they decide based on what matters to them most. Some think healthcare and education are more important, others security and freedom. Either way you can't have it all.

Singling out a group of people and then identifying their thoughts as not worthwhile seems to undermine the conclusion of the study.

Reply on news.ycombinator.com