Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

blog.effectcheck.com

Post reply on HN
Print view View original

31–40 of 57 posts

Re: Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

#31

Post by sfk »

tansey wrote: post #7
noelsequeira wrote: post #6

I don't mean to come across as hostile or undermine the effort that's been invested, but I can't help but be a tad skeptical about these visualizations, especially given the fact that the author (in what might be construed as cavalier fashion) presents seemingly nebulous metrics like they are absolute matter-of-fact ("Anxiety/Confidence Ratio", "Hostility/Compassion Ratio", "Depression/Happiness Ratio"). It would cer…

>And with that, I throw down the gauntlet. Analyze this! Sure. Scored with EffectCheck: Anxiety - Very High Hostility - High Depression - Very High Confidence - Low Compassion - Low Happiness - Very Low I will post a longer explanation detailing how/why it works, since others have had this question as well.

I'm not sure that I understand. You have just demonstrated that the algorithm does not work.

Re: Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

#32

Post by tansey »

sidww2 wrote: post #20

Are the number of votes for a posts available? It could be interesting to use the metrics given to compare posts with high vs low number of votes and could also help validate the use of given metrics.

I did originally have a similar idea, where I started out by looking at the top 10% of comments each month (ranked by upvotes). However, the resulting trends looked pretty similar to the overall community: http://blog.effectcheck.com/top_dep_hap.png

Re: Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

#33

Post by tansey »

pgroves wrote: post #29

What jumps out at me is that the community curves all peak in the months following May 2009, which is when the stock market bottomed. The community was least happy when the economy was at it's "darkest before the dawn" moment. In other words, I wonder if these graphs are just tracking general sentiment, and would look the same for any site.

It's a good question, and I guess we'll see how well the correlation holds. It's worth noting that the stock market bottomed in March 2009, so this may be a lagging indicator.

Re: Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

#35

Post by jarrett »

tansey wrote: post #9

Hi everyone, So there seems to be a lot of skepticism about whether our algorithm can actually measure emotional impact accurately. For the long answer, I'll refer you to the about page[1] for EffectCheck. For the short answer: My co-founders are an AI PhD and a Clinical Psychology PhD. They spent three years curating a huge dictionary of words using a methodology similar to the Harvard Psychosocial Dictionary [2], b…

Check out the results for their own webpage. Click "Try EffectCheck on this page!" here:

http://effectcheck.com/tour

I'm a little skeptical.

It rates "anxiety" as "medium-high." Yet I don't feel the slightest bit anxious reading that page. Nor does the author come off as anxious. It all feels very positive to me. (I'm counting the text alone, not the graphic design.)

It rates "compassion" as "very high." I wasn't able to find a description of what exactly the compassion measurement means in EffectCheck. But I'll go out on a limb and assume a high compassion rating means a text will either make you feel like the author is compassionate, or make you feel compassionate yourself. I don't feel either for that page. Not that it's hostile--I just don't feel anything on that axis when I read the page.

It's entirely possible that I've misunderstood how EffectCheck's ratings should be interpreted. It's also possible that I'm missing the point by comparing how I think I feel against EffectCheck's results. Maybe we're unreliable observers of how a text emotionally affects us, and EffectCheck more accurately predicts what's really going on deep in our brains. In which case, forgive my unwarranted naysaying.

Re: Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

#36

Post by brudgers »

Hypothesis: PG's mood is improving because YC of which HN is a part is increasingly successful.

Hypothesis: The net mood of HNer's is deteriorating because it the ratio of financially successful members is lowered as a more people join. In addition as HN grows the ratio of preexisting and external relationships among members gets smaller, e.g. the number of HN'ers affiliated with YC companies has not kept pace with the overall growth in HN membership.

Re: Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

#37

Post by tansey »

michaeldhopkins wrote: post #17
tansey wrote: post #7

Earlier quoted context omitted.

>And with that, I throw down the gauntlet. Analyze this! Sure. Scored with EffectCheck: Anxiety - Very High Hostility - High Depression - Very High Confidence - Low Compassion - Low Happiness - Very Low I will post a longer explanation detailing how/why it works, since others have had this question as well.

This is excellent. Would you be willing to check this text from my blog please? "It's said that great military commanders, chess players and Go players feel physical pressure on their stomachs when their game pieces are threatened, and the pressure indicates the moves to make. This full-body thinking communicates much more rapidly than purely deductive mental reasoning. The intuition is the result of thousands of pri…

Sure. Here you go:

Anxiety - High

Hostility - High

Depression - High

Confidence - Medium

Compassion - Very Low to Low

Happiness - Low to Medium

Re: Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

#38

Post by tansey »

jarrett wrote: post #35
tansey wrote: post #9

Hi everyone, So there seems to be a lot of skepticism about whether our algorithm can actually measure emotional impact accurately. For the long answer, I'll refer you to the about page[1] for EffectCheck. For the short answer: My co-founders are an AI PhD and a Clinical Psychology PhD. They spent three years curating a huge dictionary of words using a methodology similar to the Harvard Psychosocial Dictionary [2], b…

Check out the results for their own webpage. Click "Try EffectCheck on this page!" here: http://effectcheck.com/tour I'm a little skeptical. It rates "anxiety" as "medium-high." Yet I don't feel the slightest bit anxious reading that page. Nor does the author come off as anxious. It all feels very positive to me. (I'm counting the text alone, not the graphic design.) It rates "compassion" as "very high." I wasn't abl…

>It's also possible that I'm missing the point by comparing how I think I feel against EffectCheck's results. Maybe we're unreliable observers of how a text emotionally affects us, and EffectCheck more accurately predicts what's really going on deep in our brains. In which case, forgive my unwarranted naysaying.

Bingo. Our algorithm focuses on the lingering, subliminal emotions that you aren't necessarily aware you're feeling. Though not related to the emotions we measure, one commonly-referenced study shows that if you're walking down the street and you see a big sign with the word "ELDERLY" written on it, you'll actually walk slower without realizing it [1]. Lots of words have these subconscious impacts and you simply cannot accurately poll your own brain to determine how you feel.

[1] See "Stumbling on Happiness" for a great overview of this study: http://www.amazon.com/gp/product/1400077427/ref=as_li_ss_tl?...

Re: Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

#39

Post by pg »

"Thus, if we look at the PG analysis as a measure of his mental state, we can speculate this is due to the prosperity he’s experienced in these last few years."

Actually it's because the site has become so large that I have to be more diplomatic. When it was small I could say harsh-sounding things without worrying about being misinterpreted. Now I have to be more of a politician.

Re: Do Social News Sites Deteriorate? (Analysis of 1.8M HN Comments)

#40

Post by ma2rten »

sfk wrote: post #31
tansey wrote: post #7

Earlier quoted context omitted.

>And with that, I throw down the gauntlet. Analyze this! Sure. Scored with EffectCheck: Anxiety - Very High Hostility - High Depression - Very High Confidence - Low Compassion - Low Happiness - Very Low I will post a longer explanation detailing how/why it works, since others have had this question as well.

I'm not sure that I understand. You have just demonstrated that the algorithm does not work.

Depends if you actually ment what your wrote. I think his point is that if you did, you would have chosen a different wording from a psychological point of view.
Post reply on HN