- If you are not finding success at work, it could be because your personality better aligns with a different job.
- A recent study used machine learning techniques on Twitter data to evaluate the personality traits of users and see how they are associated with various jobs.
- The study found that some occupations had distinct patterns in employees' characteristic traits, which may mean some personalities are better suited for certain types of jobs.
- The researchers wrote in the paper that the findings could potentially help recent graduates, the unemployed, or people hoping to change careers.
- Visit Business Insider's homepage for more stories.
If you are a recent graduate with ambitions for your first job but no idea which one is the perfect fit for you, you might look for online career quizzes to assist in narrowing down your search.
However, a new study published in the Proceedings of the National Academy of Sciences suggests that your personal social media accounts could be a surprising way to find a job that matches your personality.
The study, by Margaret (Peggy) Kern of The University of Melbourne, Paul McCarthy and Deepanjan Chakrabarty of University of New South Wales, and Marian-Andrei Rizoiu of The University of Technology Sydney, used IBM's Watson Personality Insights system to analyze tweets and then "infer personality characteristics" of over 128,000 Twitter users that work in over 3,500 occupations.
The study found different types of occupations had varying traits and values scores. In addition, the researchers were able to create groups of similar occupations based solely on Twitter digital fingerprints.
Here are the main results from the study:
SEE ALSO: The 25 major US companies with the happiest employees
The researchers used machine learning techniques to assess Twitter users' personality traits based on their data.
The researchers used the Twitter Application Programming Interface to select users that had their job title listed on their profile.
Afterward, the researchers used IBM Watson to "infer personality characteristics of users" and give each user scores based on their values and personality traits.
Each user's digital fingerprint was classified and given a score for the Big Five personality traits that are frequently used in psychology research, according to Psychology Today: openness, agreeableness, extraversion, emotional stability, and conscientiousness.
Scores of basic values were also assessed: helping others, tradition, taking pleasure in life, achieving success, and excitement.
The researchers made a "vocation map," which puts professions into groups based on the values and traits gathered from the digital fingerprints. That can give a broader view of what jobs have employees with similar traits and values.
The researchers aimed to train an algorithm on one set of users' personality traits and occupation data to try to predict other users' occupations. If the model has a high accuracy rate, then it might be able to predict future users' jobs based on their social media trails, and could thus be a useful tool in figuring out what occupation is right for users based on their personalities.
Predictive modeling showed that it was possible to determine someone's profession based on their digital fingerprint.
The researchers took 10 occupations and over 9,500 user samples to run machine learning algorithms that determined if jobs can accurately be predicted from users' traits or values categorized by Watson. The researchers used five classification predictive modeling algorithms. These algorithms attempt to automatically predict what job a person holds from their traits and values.
When Big Five traits and basic values were used together to predict a user's occupation, the algorithms were more accurate, predicting users' jobs over 70% of the time, than when values and traits were used alone. This accuracy means it could be possible to predict the best occupation for someone based on their digital fingerprint, and the personality traits implied by that fingerprint.
That means a future where people lean on their own social media posts to assist them in their job search could be possible.
"This creates the possibility for a modern approach to matching one's personality and occupation with an excellent accuracy rate," Kern said in a news statement after the paper was published.
The job clusters the model created aligned with a standard classification system used by the US labor department.
The study found that different digital fingerprints can be grouped together by what kinds of jobs social media users held. In total, they classified jobs into 20 clusters using Big Five traits and basic value scores.
Clustering is used in data science to group similar data points together based on an attribute or attributes of the data; the researchers used the job's average for each of 10 values and traits to create clusters.
According to the paper, there was "greater alignment for similar occupations." For instance, managers were clustered in one group of users with similar traits and values, education jobs in another, and developers and programmers in a different one.
Interestingly, the majority of results the researchers found aligned with how they would be categorized in the US Standard Occupation Classification (SOC) system. That means the researchers were able to use Twitter users' digital fingerprints to create groups that were similar to the categories used by the US Bureau of Labor Statistics to classify jobs, suggesting that the clusters found in the study correspond to real-world job categories.
Although the majority of occupations were part of a cluster that was similar to their SOC category, there were some points that fell into a different cluster on the vocation map. The researchers suggested in the paper that this could be because of the skills required in the occupation; those skills might be similar to the requirements of an occupational category found in a different cluster.
The study suggested software engineers and tennis players have very different personalities.
Before the main data clustering study, the researchers first wanted to see if it was even possible to use social media to find differences in personality among people in different jobs. In order to do so, they took a smaller subset of their collected data.
Based on that smaller sample — over 1,000 users and nine occupations — top GitHub contributors, who are usually programmers and software developers, tended to have low scores on agreeableness and conscientiousness but high scores on openness. Meanwhile, professional tennis players tended to be the opposite, with low scores on openness, yet high scores on agreeableness and conscientiousness.
These results suggest that tennis players' personalities can be classified as typically more cooperative and kind, whereas software engineers are more open to new experiences and ideas.
In addition to software engineers, the researchers found that scientists in different fields typically were more open and less agreeable or conscientious.
A similar process could present people looking for a new career with a job field that suits their personality traits.
Maybe you're unhappy with your job, but don't want to change because you're unsure what you can do outside your field.
Research like this study could suggest good alternative career paths for someone who wants to find a new job where they can find success with their skills and personality traits. If social media data can help predict a match between personality and occupation, analyses like these could provide another tool for job-seekers.
"By better understanding the personality dimensions of different jobs we can find more perfect matches," said McCarthy in a news statement.
The researchers said in the published paper that they believe their findings can help people find jobs that match their personality as new occupations emerge and others become obsolete.
Read the full research paper here.
Visit dearJulius.com to get free premium content for all of your lifestyle needs.
COMMENTS