Subtle visual cues nudge users to reveal more in online forums | Hacker News Reader

Subtle visual cues nudge users to reveal more in online forums

news.psu.edu

53 pointsrbanffy14 comments

Open articleView on HN

A link to the actual paper with more detailed sample images:

https://dl.acm.org/ft_gateway.cfm?id=3274359&ftid=2013951&dw... [PDF]

NB: this 302-redirects to a URL that only works for you[r IP] for a short period of time.

(Noticed this after ^S after delay resulted in a 403)

https://dx.doi.org/10.1145/3274359 is a more stable way of finding the paper.

> The researchers recruited 218 participants from Amazon Mechanical Turk, an online task platform. The volunteers were randomly assigned to one of six different websites that were designed to reflect a site with a crowd cue or one without the crowd cue, a site with a connectivity cue or no connectivity cue, or a site that reflected the community framing or no framing.

n=218 across 6 groups testing 3 different variables (also why not 2^3=8 groups?). Let me know when it replicates. :-|

> Let me know when it replicates. :-|

Why not just put a (tentative) next to the findings and take it for what it is?

Why not just put a (tentative) next to your newspaper’s astrology column?

Clickbait Crap like this is why there is a replication crisis.

astrology tentative?.. ha!

"Millionaires don't use astrology, billionaires do."

> replication crisis

more like a paradigm crisis... Scientism ftl!

The findings were significant, presumably, and this bitching about replication without doing anything about it isn't helping. You replicate it and let us know what you find if it's such a pain to you.

33 people per group is enough for statistical significance with moderate effect size, which they found here.

If you don't think it's legit, get a counter-study published. Or go into the math and show how they used the wrong statistical test.

Peer-review exists, and the people doing it are a lot smarter than "um sample size lol."

> In a study, researchers found that people using an online sexual health forum featuring computer graphics, called icons

Wow...this reads like news about the "World Wide Web" and "Information Superhighway" from 1995. Takes me back.

I had to double check the publish date!

Sure, this is one reason voting rings exist, to create an illusion of plurality for what is often an extremely fringe viewpoint. This has been weaponized for years.

> The size of the crowd suggested by the icon changed randomly for participants so that they were not merely jumping on the bandwagon of a large crowd, according to the researchers

.... so it was a lie?

Users coming from Mechanical Turk probably don't expect it to be really real.

Reply on news.ycombinator.com