Hacker News Reader

Comment by neilv | Hacker News Reader

> The researchers randomly changed the “People You May Know” recommendations algorithm that LinkedIn shows its users, so that the prevalence of weaker and stronger connections varied among people on the site. The experiment showed that weaker ties (where a pair of users had only one mutual friend, say) were more likely to lead to job applications and job moves than those where people had 25 mutual friends or more.

Was this research experiment with people's careers and livelihoods done with informed consent of human subjects?

View on HN

Not in the sense you're thinking, and there was some controversy about this earlier this year

https://www.nytimes.com/2022/09/24/business/linkedin-social-...

> “The findings suggest that some users had better access to job opportunities or a meaningful difference in access to job opportunities,” said Michael Zimmer, an associate professor of computer science and the director of the Center for Data, Ethics and Society at Marquette University. “These are the kind of long-term consequences that need to be contemplated when we think of the ethics of engaging in this kind of big data research.”

However I strongly agree with a sibling comment that these platforms are constantly varying this kind of thing without any producing any kind of systematic learning, and it's not clear why this particular manipulation should come under more scrutiny than any other. At least this manipulation produced published research.

This sounds like a fair question, and it might be (including from a legal standpoint - IANAL), but the way I figure it LinkedIn will be constantly experimenting with their recommendations anyway, whether those results are shared for research purposes or not, and they're not going to ask your permission to do that. It's what you sign up for when you join the site.

And it's the same with other social media. Perhaps it seems more seriously because it's LinkedIn and it's a "professional" network. I'm not really defending them: it's more that I've always taken the view that I am responsible for my career and, whilst LinkedIn's tools and recommendations might be helpful, they're also not definitive, and I'm not going to hold them responsible for either opportunities that don't work out or for failing to show me all possible opportunities that might be a good fit.

On top of that, I'm not convinced it means what they think it means.

To me, having used LinkedIn quite heavily, the people I have 25 mutual "friends" or more with are almost all "super connectors" that spam huge numbers of connection requests (and I, like many others on LinkedIn, accept a lot of connections from people I don't know, "just in case" it leads somewhere).

And so they are frequently the weakest ties of all.

I'm not convinced you can measure "weak" vs "strong" ties on LinkedIn by number of connections, because so much tends to depend on circumstances. E.g. is there a culture for using LinkedIn in your job, or does your boss look at it with fear? (I had a manager once who kept an eye on which recruiters people connected to...) Was there an exodus that triggered people to connect to huddle together? Are the people you care about actually active on LinkedIn?

Was your previous company big? E.g. I have many connections from big past employers that I've never spoken to. They're not strong ties, and certainly not stronger than the persons I have only a couple of shared connections to because they're far away and the company we worked together at was tiny, but whom I've hired twice and spoken to many times over many years.

If I write a blog post about how I recommend people use PostgreSQL instead of MySQL, and it so happens that I am a PostgreSQL consultant, do I need to require informed consent from every reader before displaying the recommendation?

It reminds me of Facebook's experiments where they tweaked people's feeds to alter their moods - sometimes for the worse.

Trying to find people more job opportunities is very different from trying to make people depressed.

> Trying to find people more job opportunities is very different from trying to make people depressed.

Yeah, but having suffered with depression myself, having had a friend with depression commit suicide, and having seen my dad's life utterly ruined by it, I'll take LinkedIn fucking around with potential job opportunities every day of the week and twice on Sundays. And, to be absolutely clear: no, I am not joking. Facebook manipulating peoples' mood and thinking is a far worse and more pernicious problem. A person's career is serious business, but it's not the most serious business.

Why would it need it? LinkedIn doesn’t owe its users anything and A and B testing is quite common.

Reply on news.ycombinator.com