The pictures and comments that people post on Social Media can be used to get an accurate assessment of their personality. Discuss:
Social media provides a plethora of voluntary data. Users interact, share emotions and opinions directly onto the World Wide Web in a naturalistic fashion, because we can upload our thoughts blog instantaneously, we do so as they happen. Thereby offering computer and social scientists, economists, and statisticians, among others, new opportunities for research (Schoen, Gayo-Avello, Takis Metaxas, Mustafaraj, Strohmaier, and Gloor. 2013). Social media accounts are open books to some of the users most private thoughts and reflections. We express feelings and opinions that we may not otherwise say out loud. Therefore, it would seem like a valid platform for Psychologists to take advantage and use this data to study and assess personalities. This essay aims to discuss how accurate are these assessments are.
Personality psychologists have reached a consensus that the domain of personality can best be described by the Big-Five dimensions of the Five Factor Model (FFM) (Moore & McElroy, 2012). However now social media has become the norm for broadcasting our daily and deepest thoughts, researchers have combined the data from these sites and measured them against the FFM. A study by Liu, Tov, Kosinski, Stillwell & Qiu (2015) compiled the written language from 66,732 Facebook users and their self-report questionnaire based on the FFM. This data was used to build a predictive model of personality and tested on a separate sample of 4,824. Compared with previously used language based predictive models the study found that their method resulted in state-of-the-art accuracy and found that the analysis of the predictive language provided a ample representation of personality traits and could be used as a cost effective additional measure to assess larger samples otherwise unobtainable. The advantages of computer based prediction methods are evident. Studies can gather data from large sample sizes with very little effort. Collecting the data from over 60,000 users took seconds (Liu, Tov, Kosinski, Stillwell & Qiu. 2015). Its low cost to run this studies as resources are minimal. These progressive points are favourbe considering the accuracy some studies have produced. If we are able to predict personalities using this readily available data the purpose of the knowledge could be worthwhile. Particularly so to advertising and marketing companies. Facebook adverts could be adjusted to the personality of the particular user. The layout, skin and features of the users account could specific to the personality of said user. Incorporating ones personality into their social media page would set out to enhance the users experience. By integrating these findings into Facebooks friend recommendation feature we would be able to more accurately predict who else on Facebook a user might be more likely to add (Golbeck, Robles, and Turner. 2011).
While there are advantages, like with all social science studies there are also disadvantages. A study examining how personality correlates with the use of the two largest social network sites, Facebook and Twitter, (Hughes, Rowe, Batey and Lee. 2012) focused on the personality traits of the Big-Five (Neuroticism, Extraversion, Openness-to-Experience, Agreeableness and Conscientiousness). The study had 9 hypotheses, one which stated that Neuroticism would be positively correlated with social use of both Facebook and Twitter. However, their findings relating to the other personality traits did not correlate across the board. They concluded that people who displayed a more social and high neuroticism personality seek more contact via Facebook than those who don’t. Although personality was related it could be concluded that it was not as simple as previous research may suggest and there was evidence to suggest that some users behave differentially on their Twitter accounts than they do on their Facebook accounts suggesting that they use each account for different reasons. Further studies were carried out by Gou, L., Zhou, M., & Yang, H. (2014). This study focused on a two part investigation, examining and collecting data from 256 twitter accounts. Researchers used an automated personality predictor method called KnowMe and compared its accuracy with psychometric test scores and users perception. One of the studies hypotheses was that people are likely to share more high-value positive traits such as friendliness. Reasoning for this is that trait value would play a large part in what people shared. People are less willing to share negative traits such as emotional instability. Disadvantages to to this type of exploration comes in the form of self preservation. Participants are more likely to self report reflected self views than actual behavior (Schoen et al. 2013).
Despite limitations there is evidence to suggest the that personality predictions using online output could supplement already existing methods. Qiu, Lin, Ramsay and Yang, (2012) collected microblog tweets from 142 participants over a 1 month period. The language data was measured against the Big-Five personality traits and 8 observers also rated participants personally based on their tweets. The findings replicated previous studies but also uncovered new relationships between linguistic cues in tweet microblogs. The results show that in particular agreeableness and neuroticism can be consistently recognized by observers who do not know the blogger but instead judgments made based on the language content. The study is another good example of how we can further research personality with minimal resources.
Social media allows scientists to study individuals as they reveal themselves, uninterrupted. It provides a window to observe psychological characteristics in a real life environment. It could be argued that the very fact social media a curtain in which keyboard warriors come into their own and develop a personality untrue to their real self. However the state of the art technology developed to analyze the vast data it collects does so on the most part accurately, with data it is given. It should be considered a complementary tool to aid personality research. Studies have concluded that used in conjunction with historic measurements such as the Big-Five questionnaire we can predict and connect personality to an extent. We can make use of the language used in portrayal and study individuals on a whole new level