Is social media, like Facebook and Twitter, really controlling
our lives? I think so. Now even research has been conducted through Twitter to
analyze our every day human behavior.
It is a Cornell University scientific study that just
recently have been published, that analyzes the “tweets” on Twitter based on
what is said and what mood is expressed.
This study that has been conducted across cultures within
the frames of English speaking tweets from 2.4 million twitter users over a two
year period, and is suppose to visualize people’s lives and behaviors in their
everyday ups and downs. The researchers were able to reach all of the data of
400 messages from the millions of twitter users; you could have been one of
those analyzed tweeters.
A computer analyzed the chosen tweets by positivity and
negativity wordings, noting specific times of day and of the season. As a
result, the research found normal human mood patterns, where the mood of the
twitter users were rising in the end of the week, in comparison to the
beginning of the week. Also, a fun fact, they found out that seasonal mood
changes are due more to a “diminishing of positive emotions in anticipation of
short days” – not a negative view towards the seasonal changes.
I don’t see this is a hundred percent accurate
research in the sense of evaluation human behavior. It is, first of all,
analyzed by a computer system, noting only positive and negative words and
symbols, so all the sarcasm that is expressed in the tweets are faulty
analyzed. And second of all, from my personal experience regarding me and my friends, we write random non-relevant anecdotes and “over exaggerate”
things, and write just sometime when in the mood to. The analyzation of human behavior on social websites, like twitter, are therefore not valid because we humans are so much more than just symbols on the computer screens.
I'm not alone... In the article
in New York Times, that describes this research, psychologists do not fully
support this either, by saying:
“Tweets may
tell us more about what the tweeter thinks the follower wants to hear than
about what the tweeter is actually feeling,” said Dan Gilbert, a Harvard
psychologist… “In short, tweets are not a simple reflection of a person’s
current affective state and should not be taken at face value.”
I have heard of this research before! It was in response to the hashtag #nyfw for New York Fashion Week—I think they tried to analyze how collections were received from critics and people online.
ReplyDeleteI agree with your NYTimes article, Twitter is comprised of a more affluent set than you would think. It assumes people have access to the internet, which a lot of people do not, especially those of lower socioeconomic status. Therefore, it is in no way a reflection of public opinion/mood. Yes social media has come to be a prominent part of modern life, with people spending a significant amount of time on it every day, and yes it represents an incredible marketing opportunity for businesses wanting to get “in” with the younger generation. (I was sitting in on a conference call at work where a bunch of advertising executives were discussing how to better increase their presence in social media the other day actually) But let’s not overestimate its ability to tell us the feelings of its users. Short of a poll when you log in asking how you’re feeling today, I think that this analysis is heavily skewed by the demographic of people that use social media, and definitely needs to be taken with a grain of salt.
Great post!
I agree that I don't think that analyzing people's posts to find emotion would be very accurate. I would venture to say that unless the study can completely understand sarcasm, it is completely invalid. More people probably use sarcasm in tweets than people who actually use twitter to emote their feelings. People definitely put on a persona when using social media, and until that is decoded, studies like that won't be very useful
ReplyDeleteWith the advancement of technology and the rise of social media sites, I feel as though researchers are looking for new outlets to expand their research on. Unfortunately, this research is one that may not be completely accurate. I agree with you on the notion that a person's emotions and feelings cannot be simply evaluated and analyzed by a computer system. I have twitter account and half my tweet exchanges between my friends and I are embedded with sarcasm. If this system tried to analyze my tweets, it would probably conclude that I was a bipolar weirdo roaming on twitter.
ReplyDeleteI think Facebook, Twitter, and other social media sites provide great opportunities for scientists in the field to examine social trends and patterns. However, as you stated, with a computer analyzing these messages, there will inherently be some margin of error. On top of that, the psychologist you quoted explains how Tweets usually tell us more about what the Tweeter thinks the audience wants to hear, rather than what the Tweeter is actually feeling. I do agree with that to an extent, especially in regards to celebrity and political figure Tweeters who have a huge number of followers. Although very time consuming and probably not feasible, it would be interesting to see what kind of information could be pulled from Tweets with human researchers actually doing the interpreting. Social media provides a window into the lives of the prevailing attitudes of the people, and as such it has the potential to be a gold mine of data for marketers, political analysts, etc.
ReplyDelete