Forgot your password?
typodupeerror
Twitter Cloud Social Networks Stats

Researchers Analyze Twitter To Find Happiest Parts of the United States 160

Posted by Unknown Lamer
from the alternatively-adhd-affliction dept.
Nerval's Lobster writes "If you live in Hawaii, congratulations: according to a new study (PDF) by researchers at the University of Vermont, you live in the happiest state in the union — at least as far as Twitter sentiment is concerned. (Hat tip to The Atlantic for posting about the research.) The researchers — affiliated with the University's Department of Mathematics & Statistics, Complex Systems Center, Computational Story Lab, and Advanced Computing Core — collected 10 million geo-tagged Tweets from 373 urban areas across the United States in 2011 and ran them through a system designed to tag each on a scale from 1 (sad) to 9 (happy). According to the study, the five happiest states include Hawaii, Maine, Nevada, Utah and Vermont; the five saddest are Louisiana, Mississippi, Maryland, Delaware and Georgia. In general, the West and Northeast seemed much happier than the Mid-Atlantic and South—with the exception of Florida, which shaded 'happier' than many of the surrounding states. While the researchers admitted their study's limitations, there are certainly a lot of opportunities for refining the model: for example, if Hawaii's status as a vacation state affects its rate of 'happy' Tweets, or if incorporating languages other than English into the dataset would affect the ultimate results."
This discussion has been archived. No new comments can be posted.

Researchers Analyze Twitter To Find Happiest Parts of the United States

Comments Filter:
  • by CohibaVancouver (864662) on Wednesday February 20, 2013 @01:14PM (#42956635)
    Frankly, I'm not surprised. I amazed many more Americans don't live in Hawaii. I realize there are many reasons people live where they do - Family and friend connections, employment, intertia... But man, if it was easy for me to move to Hawaii I'd be there in a shot. It's just such an agreeable place - Particulary places like Kauai.
    • Indeed, if I could find a job that allowed me to live there and could convince my wife to move to Kauai, I'd do it in a heartbeat.
    • Re:Hawaii (Score:5, Insightful)

      by eksith (2776419) on Wednesday February 20, 2013 @01:29PM (#42956791) Homepage
      Yes, but don't forget, a lot of people *visit* Hawaii as well, which would really add to those Tweets. Who doesn't let the world know they're enjoying their vacation these days (or suffering, whichever the case). I mean it's a lovely place to be, if you're already well off or have the means to support yourself. As you added, employment and inertia are key here, since you can keep in touch with friends and family over Skype.
      • Re: (Score:3, Interesting)

        by theakstonsXB (1075239)
        Great point. Study claim is geolocated tweets *not* that people living there tweet. Seems that the vacation spots have the highest happiness. Who knew.
      • by cayenne8 (626475)
        They obviously didn't do this study a couple of weeks ago in LA during Mardi Gras!!!!

        I think the 'happiness meter' would have blown through the roof if they did....

        Or, perhaps they did this poll on Ash Wednesday, when everyone here was hungover?

        Either way, I have a hard time believing the LA is in the saddest category....everyone here seems to generally be upbeat, happy, and glad we live in (at least in the NOLA area) a place with no open container laws, and you can get a drink 'to go' from a bar...and t

        • Either way, I have a hard time believing the LA is in the saddest category....everyone here seems to generally be upbeat, happy, and glad we live in (at least in the NOLA area) a place with no open container laws, and you can get a drink 'to go' from a bar...and there's always a festival going on year round.

          And mustn't forget the Drive-Thru Daiquiri Shops...

          I think the secret to their analysis is that they didn't consider the possibility that some people are too busy partying to tweet. They're only acc

    • I'm an Australian who has visited Hawaii. I'm not surprised at this result, Hawaii is lovely. The people were nice, lots of attractive people on the beaches but without the pretension of Californian beaches, seemed to be a healthy local economy, good food, relatively low population density (I'm guessing, there seemed to be a lot of open space), good climate and relatively prosperous.
    • by operagost (62405)

      I amazed many more Americans don't live in Hawaii.

      Yeah, I wonder.

      I realize there are many reasons people live where they do - Family and friend connections, employment, intertia... But man, if it was easy for me to move to Hawaii

      Well, those are several good reasons.

    • I thought the place was a shithole, myself. I only stayed to break up the 20+ hr flight across the Pacific. The jarring clash between crass American culture and the beautiful landscape was painful to be around, frankly.

    • Re:Hawaii (Score:5, Informative)

      by Anonymous Coward on Wednesday February 20, 2013 @02:35PM (#42957345)

      Don't believe it. I live in Hawaii, and it's not paradise. There's no work, people here nuts....

      The reason Hawaii is skewed toward being the, "Happiest state" is because of all the tourists that are tweeting. There are not that many computer literate people who live here. It ain't us.

      tom

    • Well, I live in Hawai`i and I'm pretty happy about it :) But if you want to move here --- there's a problem with the high cost of living and relatively low paying jobs, if you even find one. Someone once called it "New York prices on Midwest wages." No so easy go Hawai`i, brah.
    • Re:Hawaii (Score:5, Informative)

      by hawguy (1600213) on Wednesday February 20, 2013 @04:49PM (#42958739)

      Living in Hawaii is not quite the same as visiting there. Hawaii is small - even on Oahu (the most populated island), you can't drive any further than an hour from home (any more than that and you're on your way back home again). If you have family back on the mainland, you're an expensive 5 hour plane ride to the West Coast to visit them - worse if you're going to the East Coast. Housing is expensive too - in Honolulu, expect to pay San Francisco rates for housing, other islands and areas farther from Honolulu tend to be less expensive. Everything is expensive there because nearly everything is shipped in. Salaries tend to be lower than mainland salaries. Electricity is extremely expensive - around 30 cents/KWh. Forget getting good deals online - many places don't ship to Hawaii, or if they do, they use expensive UPS or Fedex shipping (where the least expensive shipping method is 2-day air) - USPS is affordable if you can find a vendor willing to ship via USPS. It'll be expensive to ship your household goods, and you'll pay (mostly) by volume, not be weight, so you probably won't be bringing much furniture... it'll be shipped by boat so you won't have it for a month. You can ship your car for around $1200, and it might be cheaper to do that than to sell it and buy a new one once you get there. Don't expect to be welcomed with open arms by local Hawaiians -- Hawaiians have strong family ties and social bonds, and it's hard to really fit in until you've lived on the Islands for years, and even after you've lived there for a decade or more, you still won't always be treated as a "local".

      That said, for some people Hawaii really is a paradise and they are very happy there, but for others, Island Fever starts to set in after a couple years (or less). Before you pack your bags and decide to move to HI, spend a week (or more), with an eye toward what it would be like to live there (check out grocery and other household goods prices, car and gasoline prices, look at apartments, etc) make sure it's really right for you.

    • by yurtinus (1590157)
      Too small and far too much water. Doncha know people can't breathe underwater?! I'll visit there someday and I'm sure I'll enjoy the hell out of it, but I don't want to live anywhere that I can't just hop in my car or on my motorbike and leave.
  • Duh (Score:5, Insightful)

    by tekrat (242117) on Wednesday February 20, 2013 @01:17PM (#42956671) Homepage Journal

    So, lemme get this straight... If you live in a tropical paradise, you're happier than if you live in a state with a depressed economy and terrible weather.

    In other news, grass is green.

    • Re: (Score:2, Informative)

      by Anonymous Coward

      No if you're posting from a vacation destination you're probably happy.

      • Indeed. Nevada (Las Vegas, Reno) is also a popular vacation destination. All we know about the Tweets in question is that they were "geo-tagged" - not that they came from residents. I'm guessing Florida (Disney World) would have ranked higher if it weren't for all the angry old people populating the rest of the state.
        • by AK Marc (707885)
          Old people don't tweet. They shout randomly at strangers and consider that the same thing.
      • No if you're posting from a vacation destination you're probably happy.

        Exactly. If the researchers didn't account for traveling behavior (i.e., check to see if the person was posting from their typical geographical region) then the results would be heavily skewed by vacations. Hawaii, Maine, Nevada, Utah and Vermont are all popular vacation locations.

    • by Kenja (541830)
      You're missing the most important part, apparently if you dont use Twitter you're not happy!
      • I'm not sure how you came to that conclusion but I assume you don't use twitter so there may be some true that.
        • by Rockoon (1252108)
          Its called selection bias.

          Maybe sad people seek greater attention through tweeting such that P(tweet | happy) is significantly less than P(tweet | sad)

          In such a case the ratio between happy and sad tweets would not contain much information about a geographical area. A better measure would be happy tweets per capita, and sad tweets per capita, both being distinct measures such that the happiest place could also be the saddest place.
    • Actually, Maryland is quite affluent and has been relatively less affected by the economic downturn -- due to its proximity to Washington DC. Perhaps it's the latent self-loathing of all of the folks working for the Federal gov't that's being picked up? ;-)

      (This observation is only partially tongue in cheek and posted by a member of an endangered species, the Maryland Small Business Owner)

      • I was thinking the same thing - what the hell is Maryland doing on that list? Every one of my friends is right now gainfully employed, and none have been jobless for more than a few months over the last couple years.

        I suppose Maryland loses out on the tourist tweeting factor (lots of DC tourists, but rarely do they go north), and I guess also that people will complain about ANYTHING, even if they have a job!

    • No, it doesn't even mean that. It means that people in a tropical paradise use the word happy more often in their twitter posts than people in other states. It has little correlation with actual happiness - or at least, a completely unknown correlation. And even the causation is incredibly suspect. In other words, this is a bullshit study with no informational content.

      Yeah, I'm bitter, because I get to support these same useless tools in our company and get to field questions around "so how does this work"

    • It's even simpler than that: they engaged in massive selection bias.

      Words like "beach" were scored with a high happiness rating, according to the paper. From there, it's trivial to see that people living closer to the beach will score higher on the happiness index, even if they're not necessarily happier, and sure enough, virtually the entire American coast is painted with happy colors on the map contained in the paper (with the exception of the Gulf Coast, which was unsurprisingly painted with unhappy colo

  • We used to get warnings when a link was a PDF. Thar could be danger thar maties!

    • by DaveGod (703167)

      I think it is reasonable to consider that anyone who does not trust Adobe Reader is responsible for disabling it or installing an alternate reader.

      Actually I wouldn't be surprised if the PDF-warning arose not in the interests of security, but from the days of dial-up internet and to advise it opens another application. Back in the day if there was a PDF to read I always used to download it as a file and then open it because Reader was a complete ass at trying to download pages as I was reading through it.

    • by jon3k (691256)
      If only you could see the destination of a link somewhere when you put your mouse over it.
  • Happy Coincidence (Score:5, Insightful)

    by bill_mcgonigle (4333) * on Wednesday February 20, 2013 @01:21PM (#42956709) Homepage Journal

    That the Univerity of Vermont researchers' methodology indicates that Vermont is among the happiest states.

  • Prison population (Score:5, Interesting)

    by Dyinobal (1427207) on Wednesday February 20, 2013 @01:23PM (#42956731)
    Louisiana has the largest prison population and is also the saddest place according to this study, coincidence?
    • by hedwards (940851)

      Yes, they're probably both caused by the same thing, but there's probably not a direct connection between the two. I'm not at all surprised by the South dominating the bottom five spots. The conservative hands off policies on regulation and unmotivated union busting tend not to correlate very well with quality of life.

    • by Bomazi (1875554)

      Given that inmates don't have internet access, yes.

    • by Nyder (754090)

      Louisiana has the largest prison population and is also the saddest place according to this study, coincidence?

      weird, considering you aren't usually allowed internet access in jail/prison.

  • by pesho (843750) on Wednesday February 20, 2013 @01:24PM (#42956749)
    ... that happy people do not use Twitter.
  • by Anonymous Coward

    Anyone else find it interesting that, in general, red states (Republican) are mostly blue (unhappy) on the map, while blue states (Democrat) are mostly red (happy) on the map?

    • by operagost (62405) on Wednesday February 20, 2013 @01:42PM (#42956905) Homepage Journal
      I would find it interesting if that were actually the case. You think that New Jersey, Delaware, Connecticut, Maryland, Michigan, Pennsylvania, and Ohio are Republican-controlled states? And Utah, Nevada, Wyoming, and Florida are Democrat-controlled states? That's enough to take the words "general" out of it.
      • by asylumx (881307)
        Depends if you're talking about the results of the last presidential election, or something else. Michigan, for example, often goes blue in statewide elections but ends up with heavy Republican leanings in their state house & senate.
      • by DeadCatX2 (950953)

        PA has 13 red districts and 5 blue districts. OH has 12 red districts and 4 blue districts.

        Sounds Republican controlled to me. Now, if you want to talk about a Republican popular vote...OH is still Republican controlled (51% R, 46% D in the last election), but PA would be narrowly controlled by Democrats (50% D, 49% R). But having a higher popular vote in today's United States of America doesn't necessarily mean that you control the associated legislature.

    • by characterZer0 (138196) on Wednesday February 20, 2013 @01:48PM (#42956975)

      The republicans enjoy making angry tweets. They are just as happy, but the data analysis does not understand them.

    • by dkleinsc (563838)

      That's not really the relevant correlation, because as a sibling poster points out, that doesn't actually control for anything.

      More interesting to me is happiness [slashdot.org] versus median household income [wikimedia.org]: There may be some sort of relationship between those two, but there appear to be some happy places that aren't rich and some rich places that aren't happy.

    • This is very interesting, as Hawai`i is virtually a one-party state, the Dems have near total control and with a few rare exceptions, just about always have.
    • Anyone else find it interesting that, in general, red states (Republican) are mostly blue (unhappy) on the map, while blue states (Democrat) are mostly red (happy) on the map?

      This could be worthless information, but I'm just throwing my thoughts out into the open here...

      I believe that the liberals have more "hope" than conservatives. Conservatives are more rational and change-limited. Logically, one with hope is more of a happy person with the belief that there is more positive coming in the future. One with rational thought is more likely to know that there are positives and negatives coming in the future which offset each other, thereby limiting overall happiness to a media

  • I think this just means that Utah is the most sarcastic state.
  • by Bob9113 (14996) on Wednesday February 20, 2013 @01:36PM (#42956855) Homepage

    Hawaii, Maine, Nevada, Utah

    Hot, cold, vice, and virtue.

    • by tanujt (1909206)
      Virtue?
      • by Anubis IV (1279820) on Wednesday February 20, 2013 @02:44PM (#42957431)

        Utah is commonly associated with Mormonism, which is generally considered to have followers who are rather courteous and virtuous, if a bit annoying with their missionaries knocking on doors.

        • by Nyder (754090)

          Utah is commonly associated with Mormonism, which is generally considered to have followers who are rather courteous and virtuous, if a bit annoying with their missionaries knocking on doors.

          Hmm, most people i know consider Mormons to be batshit crazy.

          • Utah is commonly associated with Mormonism, which is generally considered to have followers who are rather courteous and virtuous, if a bit annoying with their missionaries knocking on doors.

            Hmm, most people i know consider Mormons to be batshit crazy.

            ...which isn't mutually exclusive with anything I said. ;)

  • The study just did a search for specific words, tabulated the results, and said that because happy words mean happy posters and unhappy words (and swear words) mean unhappy people, that their study was a study on happiness.

    Complete bullshit. I have yet to even see a study that determines what percentage of happy words correlate with an actually happy post (Classic example: "Steak overdone. Not happy"), and extrapolate from there to overall numbers. Instead, it's just some people making shit up so that they

  • by Attila Dimedici (1036002) on Wednesday February 20, 2013 @01:51PM (#42956997)
    Gallup does a "well-being" poll (the factors they use to determine "well-being" correlate pretty well with happiness). While the Gallup poll [gallup.com] agrees that Hawaii is the "happiest", the rest of their poll comes out significantly different. For example, the Twit survey from this article has Florida as above the median for happiness, the Gallup poll has them third from the bottom. Another example, this Twit poll puts Maryland near the bottom, while Gallup puts it near the top. The real problem with the Twit survey is that states that are vacation destinations will have a disproportionate representation of people who are not involved in their daily grind. I suspect That not only are people who are on vacation more likely to be happy, those that are Twits probably tweet more while on vacation.
    • by asylumx (881307)
      That said, you typically vacation to placed that make you happy. Therefore these places were probably "happy places" before the tweeters decided to vacation there. Chicken vs egg.
    • Gallup does a "well-being" poll (the factors they use to determine "well-being" correlate pretty well with happiness). While the Gallup poll [gallup.com] agrees that Hawaii is the "happiest", the rest of their poll comes out significantly different. For example, the Twit survey from this article has Florida as above the median for happiness, the Gallup poll has them third from the bottom.

      This is a study of the happiness of technically savvy people. As cheap and ubiquitous as cell phones are, Florida is full of people who live outside the cities and couldn't care less about twitter. The Gallup poll would, in theory, include those people. And from the sound of it, they are not very happy individuals. :-)

      • No, this is a study of the happiness of Twits, some of whom are technically savvy people, but most of whom are just twits. It does not take any technical savvy to be on Twitter.
      • This is a study of the happiness of technically savvy people.

        No, this is a study of people who twitter.

        Being technically savvy does not equate to "I use twitter".

        Nor does "I use twitter" equate to "I am technically savvy".

    • Perhaps very old people don't tweet. That would explain Florida. It would also explain New Hampshire not being among the happiest, despite being between Maine and Vermont geographically.
  • Don't bother. They looked at 10 million whole tweets (out of a population of 300million) and counted 'happy words' like rainbow, love, beauty, hope, wonderful, wine vs. "sad words" like damn, boo, ugly, smoke, hate, lied. The 'happy' or 'sad' status of words was determined by asking mechanical turk workers.

    • by blueg3 (192743)

      They looked at 10 million whole tweets (out of a population of 300million)

      Population size doesn't actually matter for statistics, even though people often think it does. Systematic bias in selecting your random sample matters (as does sample size and size of effect being measured), but not total population size.

      • by sjames (1099)

        Yes, population size matters. It's why flipping a coin once and declaring that a coin will always land tails up is not statistically valid.

        Of course, given the rest, that's the least of their problems.

        • by blueg3 (192743)

          You actually pick a good example.

          Questions:
          What's the population size for "coin flips"?
          What's the error on a sample size of one?

          • by sjames (1099)

            The question is not only the coin flips, but the coin. What if you just happen to pick an unbalanced coin?

            The correct question is what is the statistical significance of flipping one coin one time. The answer is zero. You don't even get to the point of computing error since it's larger than the sample.

            • by blueg3 (192743)

              You don't even get to the point of computing error since it's larger than the sample.

              It's a population sample, so the error on a single flip is undefined.

              So, statistics correctly says that flipping a coin once is not valid. Population size doesn't matter, it's already true that a single sample doesn't do you any good, statistically. Incidentally, the population size for coin flips is infinite. Yet, you can get an excellent estimate for whether a coin is biased by testing it. Your sensitivity to how much bias you can measure is entirely determined by how many flips you perform. In practice,

              • by sjames (1099)

                So just reading the tweets from one person in each state is enough? What if that one person happens to be a wealthy coke head or a high schooler with a bad case of angst?

                • by blueg3 (192743)

                  No. You seem to be applying some wildly incorrect, reductionist logic that if the total population size doesn't matter, then magically the sample size doesn't matter. The sample size is important. What's not important is how large the population that sample is drawn from. What dictates the sample size is how high you want your confidence and how large the effect size is.

                  Let's say you're measuring the fraction of tweets containing at least one happy word for a geographical area. If, in reality, 50% of tweets

                  • by sjames (1099)

                    Right, so if I read 10,000 tweets from the angst ridden teen who just had a bad breakup, it's just as valid as reading 100 tweets from 100 distinct individuals? I think not. Population matters.

                    You seem to have stepped into a bucket here. Repeatedly denying the bucket's existence won't help.

  • This "happy map" correlates with "twitter user-base map"?

  • even though it is a 1-9 scale the range of this report is really 5.85 to about 6.15 (about .3). So really happy vs sad is really not a huge difference.
    • by blueg3 (192743)

      It's a completely arbitrary scale, so you have no idea how big a 0.3 difference is.

      • by arekin (2605525)

        It's a completely arbitrary scale, so you have no idea how big a 0.3 difference is.

        Indeed, but why make .3 difference on an already arbitrary 9 point scale? All things considered I'd say this is about the most useless report I've ever seen.

  • So the happiest places are the Rocky Mountains, the northern New England states, and Hawaii? Glad I live in the Rockies...

  • The happiest place in the United States doesn't have Twitter.
  • I prefer to believe ( :-) ) that most of us in Massachusetts and NH and VT are so happy we don't even bother with something as pitiful as Twitter in the first place.
    Maybe Hawaii has the happiest twits (or do I have to call them tweeters?). That doesn't mean the GenPop is happiest.

  • All the happiest cities are in locations where marijuana is legal. Coincidence?

    Yeah, probably :)

  • Blue states happy, Red state unhappy. Kind of mirrors their politics. I plan to retire in Maine and I always found it laid back. Hard to be unhappy when you have a beer in hand and a lobster boiling for dinner. A northern version of Jimmie Buffet.
  • I don't think it's blue vs red (as other users have posted), but there are some definite trends. For example, compare to the map of the Post-Election Racist Tweets to see some interesting data: http://geocommons.com/maps/210024 [geocommons.com]
  • Good thing they didn't get as geo-specific as just my company or they'd find the angriest hole of bitterness and hatred on the internet twitternet lol.

Air is water with holes in it.

Working...