DUST Search - Who are the trolls?

Pseudogenesis wrote:

DeadlyAztec11 wrote:

Knight Solitaire wrote:

Does anyone expect trolls to fill out a survey/test honestly? Trolls gonna troll.

It didn't ask if people are trolls. It took random people's comments and habits from Amazon and analyzed them. The basic result is that people who comment more tend to be worse people and people who don't comment tend to be better people.

Erm, didn't you look at the paper? You've gotta do your homework yourself man, can't just take what some youtuber says at face value. The full paper is here. The studies they conducted were made in survey format, and samples were only taken from Amazon's Mechanical Turks website. Using data from users without their consent in the way that you're describing is against academic ethics and such a thing wouldn't make it past peer review.

Anyways, I looked over it and it's shaky at best. As I said before, the sample was very limited. It was from mechanical turks, which is basically a service that hires users to perform small, quick tasks for tiny sums of money. They are not likely to represent the entirety of the internet. Also, if I were somebody on that website I would be looking to bullshit this survey ASAP in order to get that 50 cents. The sample size is also way too small to describe the population with any sort of accuracy (assuming the population is the 3 billion people who now use the internet). Also there's voluntary response bias in play, and this study doesn't look like it's been replicated, etc etc

Basically the study is way too limited to draw any conclusions from. I would be interested in seeing more research, but this isn't a research topic that's likely to get the funding it needs to be accurate.

Man, this is making me miss my statistics classes.

Pages: 1 :: [one page]

Author	Thread Statistics \| Show CCP posts - 0 post(s)
DeadlyAztec11 Ostrakon Agency 7554	Posted - 2015.02.20 23:22:00 - [1] - Quote https://www.google.com/url?sa=t&source=web&rct=j&ei=HsHnVNWjEY3msATyuIDoBQ&url=http://www.youtube.com/watch%3Fv%3DZYbVS76BAfE&ved=0CB8QtwIwAQ&usg=AFQjCNFZ6K72vgRSGHCx7vFwRVGCspZDGA TL;DW: Trolls are actually horrible people in real life. Put your flags up in the sky. And wave them side to side. Show the world where you're from. Show the world we are one.
Damodred Matari Heavy Support INC 15	Posted - 2015.02.21 03:00:00 - [2] - Quote If making an 8 minute video filled with random guesses isn't a victim of trolling and a lack of lifting, i don't know what is. That kid should lift, rekt. GG's rekt in 1v1's, easy wins broham. 720 no scoping steroids for days. I lift bro, rekt.
Knight Solitaire 761	Posted - 2015.02.23 00:07:00 - [3] - Quote Does anyone expect trolls to fill out a survey/test honestly? Trolls gonna troll. Justice4Redberry
DeadlyAztec11 Ostrakon Agency 7597	Posted - 2015.02.23 00:32:00 - [4] - Quote Knight Solitaire wrote: Does anyone expect trolls to fill out a survey/test honestly? Trolls gonna troll. It didn't ask if people are trolls. It took random people's comments and habits from Amazon and analyzed them. The basic result is that people who comment more tend to be worse people and people who don't comment tend to be better people. Put your flags up in the sky. And wave them side to side. Show the world where you're from. Show the world we are one.
Damodred Matari Heavy Support INC 17	Posted - 2015.02.23 02:36:00 - [5] - Quote DeadlyAztec11 wrote: Knight Solitaire wrote: Does anyone expect trolls to fill out a survey/test honestly? Trolls gonna troll. It didn't ask if people are trolls. It took random people's comments and habits from Amazon and analyzed them. The basic result is that people who comment more tend to be worse people and people who don't comment tend to be better people. How would you know whether the people who didn't comment are bad or lift?. Rekt. I lift bro, rekt.
JL3Eleven Ancient Exiles. 2176	Posted - 2015.02.23 09:07:00 - [6] - Quote I iz troll if you post stupid ****. Can I haz ur ISK? Do you have officer gear or rare items to sell or trade? Message me in game with details.
Pseudogenesis Nos Nothi 1832	Posted - 2015.02.23 10:05:00 - [7] - Quote DeadlyAztec11 wrote: Knight Solitaire wrote: Does anyone expect trolls to fill out a survey/test honestly? Trolls gonna troll. It didn't ask if people are trolls. It took random people's comments and habits from Amazon and analyzed them. The basic result is that people who comment more tend to be worse people and people who don't comment tend to be better people. Erm, didn't you look at the paper? You've gotta do your homework yourself man, can't just take what some youtuber says at face value. The full paper is here. The studies they conducted were made in survey format, and samples were only taken from Amazon's Mechanical Turks website. Using data from users without their consent in the way that you're describing is against academic ethics and such a thing wouldn't make it past peer review. Anyways, I looked over it and it's shaky at best. As I said before, the sample was very limited. It was from mechanical turks, which is basically a service that hires users to perform small, quick tasks for tiny sums of money. They are not likely to represent the entirety of the internet. Also, if I were somebody on that website I would be looking to bullshit this survey ASAP in order to get that 50 cents. The sample size is also way too small to describe the population with any sort of accuracy (assuming the population is the 3 billion people who now use the internet). Also there's voluntary response bias in play, and this study doesn't look like it's been replicated, etc etc Basically the study is way too limited to draw any conclusions from. I would be interested in seeing more research, but this isn't a research topic that's likely to get the funding it needs to be accurate. Man, this is making me miss my statistics classes. Stabby-stabber extraordinaire Gû¼+¦GòÉGòÉn¦ñ I stabbed Rattati once, you know.
Damodred Matari Heavy Support INC 18	Posted - 2015.02.24 13:21:00 - [8] - Quote Pseudogenesis wrote: DeadlyAztec11 wrote: Knight Solitaire wrote: Does anyone expect trolls to fill out a survey/test honestly? Trolls gonna troll. It didn't ask if people are trolls. It took random people's comments and habits from Amazon and analyzed them. The basic result is that people who comment more tend to be worse people and people who don't comment tend to be better people. Erm, didn't you look at the paper? You've gotta do your homework yourself man, can't just take what some youtuber says at face value. The full paper is here. The studies they conducted were made in survey format, and samples were only taken from Amazon's Mechanical Turks website. Using data from users without their consent in the way that you're describing is against academic ethics and such a thing wouldn't make it past peer review. Anyways, I looked over it and it's shaky at best. As I said before, the sample was very limited. It was from mechanical turks, which is basically a service that hires users to perform small, quick tasks for tiny sums of money. They are not likely to represent the entirety of the internet. Also, if I were somebody on that website I would be looking to bullshit this survey ASAP in order to get that 50 cents. The sample size is also way too small to describe the population with any sort of accuracy (assuming the population is the 3 billion people who now use the internet). Also there's voluntary response bias in play, and this study doesn't look like it's been replicated, etc etc Basically the study is way too limited to draw any conclusions from. I would be interested in seeing more research, but this isn't a research topic that's likely to get the funding it needs to be accurate. Man, this is making me miss my statistics classes. Anything that tries to tldr or generalise upon humanity is flawed. It's far too compplex for some silly little survey, what a joke. Anyway, i just 1080 no scoped dat survey, rekt for days. I lift bro, i lift 1kg more than anyone, goml. ROFLYCOPTER. I lift bro, rekt.


Pages: 1 :: [one page]

Justice4Redberry