See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/276894677

The Use of Twitter Profiles to Assess Personality and Hireability

Article  in  Academy of Management Annual Meeting Proceedings · October 2014

DOI: 10.5465/AMBPP.2014.17496abstract

CITATIONS READS
2 1,135

5 authors, including:

Peter A. Rosen Benjamin D. Mclarty

Pittsburg State University Mississippi State University
24 PUBLICATIONS   747 CITATIONS    18 PUBLICATIONS   211 CITATIONS   

SEE PROFILE SEE PROFILE

Shelby Solomon
University of West Florida
18 PUBLICATIONS   141 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Peter A. Rosen on 25 January 2016.

The user has requested enhancement of the downloaded file.

Running Head: TWITTER, PERSONALITY & HIREABILITY 17496 1

The Use of Twitter Profiles to Assess Personality and Hireability

ABSTRACT

The use of social media screening in the hiring process is becoming commonplace by HR

practitioners, but has yet to be thoroughly studied by academics. Using 152 active Twitter users,

we examined whether personality and hireability could be assessed solely by using Twitter

profiles. Other-rated personality and hireability ratings using Twitter profiles were conducted by

six qualified raters, and the results demonstrated sufficient inter-rater reliability and internal

consistency. Results showed that all correlations between self-reported and other-rated

personality were statistically significant and ranged from .19 for conscientiousness to .54 for

openness. The three other-rated personality traits that significantly correlated with hireability

were conscientiousness, agreeableness, and emotional stability, indicating that these

traits determined whether the HR professionals used in this study found the candidates

worthy of further consideration.

Keywords:

Twitter; selection; personality; hireability

TWITTER, PERSONALITY & HIREABILITY 17496 2

INTRODUCTION

Social media (SM) screening can provide organizations with a “treasure trove of

information that potential employers will not find on an application or learn about in an

interview” (Saylin & Horrocks, 2013). More often than not, this information is used to eliminate

a candidate, but now more than ever, it can also bolster the chances of his or her hire. Since

organizations are looking to keep hiring costs to a minimum, hiring the right candidate the first

time is of utmost importance. The Society for Human Resources Management conducted a

survey in 2008 and found 26% of organizations used social media screening to assess person-

organization fit (SHRM, 2008). A more recent study conducted by Reppler (2011) found that

91% of all organizations surveyed practice social media screening, with 76% using Facebook

and 53% using Twitter to glean information about job candidates. Twitter and Facebook are used

more commonly than LinkedIn, because they contain information that would give a glimpse into

the personal life of the candidate, while the latter is akin to an online resume.

In a 2012 Career Builder.com study, 2000 hiring managers and HR personnel were polled

and 37% admitted to conducting social media screening. While 34% of those surveyed indicated

that candidates had been eliminated after the screening, 29% said they hired candidates based on

their social media information. Nearly 60% of those indicated that they could get a good feel for

the candidate’s personality by viewing their social media sites; while 55% indicated that the

information was a good indicator of the professionalism of the candidate (Grasz, 2012). To date,

most academic studies have focused on the use of the social media website Facebook in the

screening process, as there are over 1.11 billion registered users, and copious amounts of

information about each user (e.g., Van Iddekinge, Lanivich, Roth & Junco, in press; Schwartz et
TWITTER, PERSONALITY & HIREABILITY 17496 3

al., 2013; Kluemper, Mossholder & Rosen, 2012; Back et al., 2010; Karl, Peluchette &

Schlanegel, 2010; Kluemper & Rosen, 2009).

Fewer studies have focused on Twitter as an indication of personality and for use in the

screening process. Twitter is a microblogging website where users can post short messages for

other users to see with a maximum limit of 140 characters. These messages, or tweets, are

available for viewing by anyone who is following a Twitter account, and can be seen by others

who perform key-word searches on the topics that are being discussed (Twitter.com, 2013). For

studies that have already been conducted using Twitter, initial evidence suggests that one’s

profile can be an indication of one’s personality (Quercia, Kosinksi, Stillwell & Crowcroft, 2011;

Golbeck, Robles, Edmondson & Turner, 2011; Sumner, Byers, Boochever & Park, 2012; Qiu,

Lin, Ramsay & Yang, 2012). However, to our knowledge, no research has explored the link

between personality and hirability using Twitter profiles, which is the aim of this current

research effort.

There are several contributions to research in the field of human resources within this

manuscript. This is the first study to use raters to assess both the personality and hireability of

subjects using Twitter profiles. This study differs from past research in that we use two sets of

multiple raters to assess both personality and hireabilty mimicking what would be found in

organizations; (i.e., machine learning was not a factor in determining the results). Inter-rater

reliability for all six personality traits and hireability were consistent with past studies that used

SM technologies other than Twitter, supporting the validity of using Twitter to assess personality

and hireability. Hence, we help to establish that Twitter is another form of SM that can be a

useful alternative for assessing these organizationally-relevant constructs. A further contribution

of our work is insight into how raters develop their assessments. Our raters were asked to
TWITTER, PERSONALITY & HIREABILITY 17496 4

provide statements regarding which aspects of the profiles were influential in their ratings of

subjects, so that the researchers could better understand how assessments were made. Knowing

what aspects of the Twitter profile lead to the value judgments about the subject could be

beneficial in two ways: for hiring organizations looking to implement social media screening

using Twitter, and for job candidates who can make themselves look attractive to hiring

organizations.

We begin our manuscript by conducting a literature review that explores how

socioanalytic theory can be used to explain the actions of Twitter users and how these users are

perceived by others viewing these profiles. The methodology section explains the study design,

using 152 degree-seeking college students from a private Midwestern, U.S. university, and how

qualified raters, in the form of organizational behavior Ph.D. students and HR professionals,

were used to assess personality and hireability respectively. The results indicate adequate inter-

rater reliability, strong correlations between self-reported and other-rated personality, and

meaningful correlations between other-personality traits and hireability. Finally a discussion of

the utility and importance of the findings and study limitations leads into the conclusion.

LITERATURE REVIEW

Twitter is among the three largest U.S. social media sites with over 645 million active

registered users, who tweet on average 58 million times per day (Statistic Brain, 2014). Surveys

like the ones conducted by SHRM, Career Builder.com, and Reppler mention Twitter as one of

the major social media sites that employers screen. Twitter is one of the top 12 most visited

websites on the planet as tracked by Alexa, The Web Information Company (2014), making it

the #2 ranked social media website in the world. Hiring organizations can potentially use this site

to determine one’s level of creativity, the strength of one’s communication skills, an

TWITTER, PERSONALITY & HIREABILITY 17496 5

understanding of one’s professional qualifications and whether one has adequate organizational

fit (Reppler, 2011).

Past SM research studies have indicated that these websites can give hiring organizations

a glimpse into a candidate’s personality (Schwartz et al., 2013; Back et al., 2010; Kluemper et

al., 2012; Kluemper & Rosen, 2009) which is often an important indicator of future job

performance (Barrick & Mount, 1991). Given the depth of information that can be found on

Twitter, including tweets, pictures, videos, favorite tweets, retweets (reposting of another’s

tweet), lists, and public personal conversations, it is not unreasonable to expect that HR

practitioners might use Twitter to assess personality and potential hireability.

Personality does not occur in vacuum, and Hogan and Shelton (1998) use socioanalytic

theory to argue that there are two major components to one’s personality, personality from the

outside, or reputation, and personality from the inside, or identity. They argue that “observers

can describe a target person, and their descriptions define personality from the outside” (Hogan

& Shelton, 1998: 131). They go onto suggest that reputation is built on one’s action over a period

of time, so “reputations are a valid way of forecasting people’s future behavior” (Hogan &

Shelton, 1998: 131). Reputation is “defined in terms of traits, and is used to predict a person’s

performance” (Hogan & Shelton, 1998: 131). Identity “reflects the strategies that a person has

developed to get along, get ahead, and find meaning,” and “is used to explain a person’s

performance” (Hogan & Shelton, 1998: 132).

Applying socioanalytic theory (Hogan, 1996, 2007) to the present study, self-reported

personality, in the form of the Big Five personality traits plus narcissism, will be used to assess

identity. The tweets themselves provide evidence for behaviors, actions, and thoughts that can be

assessed by qualified raters, to create an other-rated personality, and as such, will be a proxy for
TWITTER, PERSONALITY & HIREABILITY 17496 6

reputation. In this way, it can be determined whether identity and reputation are meaningfully

similar, and whether reputation influences how hireable a candidate is deemed.

Personality Traits

The Big Five personality traits were used as measures in this study. Extroversion

measures how outgoing one is, and those who score high on the extroversion scale are often

considered “sociable, gregarious, assertive, talkative, and active” (Barrick & Mount, 1991: 3).

These individuals tend to be optimistic, enthusiastic, and have a can-do attitude. Agreeableness is

associated with likeability and friendliness. This personality trait is associated with being

“courteous, flexible, trusting, good-natured, forgiving, cooperative, soft-hearted and tolerant”

(Barrick & Mount, 1991: 4). Conscientiousness “reflects dependability; that is, being careful,

thorough, responsible, organized, and planful” (Barrick & Mount, 1991: 4). Commonly

associated with traits such as “being imaginative, cultured, curious, original, broad-minded,

intelligent, and artistically sensitive,” openness to experience helps distinguish creative from

more conventional people (Barrick & Mount, 1991: 5). Neuroticism is a trait commonly linked to

feelings of “anxiety, worry, insecurity, depression,” among others (Barrick & Mount, 1991: 4).

Narcissistic individuals tend to have a “grandiose view of one’s self-importance; entitlement,

expecting special favors without reciprocal responsibilities; and a lack of empathy” (Raskin &

Hall, 1979: 45). Hireability is the “suitability of a job candidate and probable interview outcome”

(Stevens & Kristoff, 1995: 592).

Hypotheses

Furr and Funder (2007) identify rater consensus as a primary concern with judge-rated

behavioral observation. Inter-rater reliability of all six personality traits being studied a)

extroversion, b) agreeableness, c) conscientiousness, d) openness to experience, e) neuroticism,

TWITTER, PERSONALITY & HIREABILITY 17496 7

and f) narcissism, needs to be present in this study or personality assessments would be fully

dependent on the rater performing the assessments. Similar to what is required of personality

raters, the hireability raters need to be consistent across raters for the use of Twitter to be

validated in this study.

Hypotheses 1a-1g: ratings of the Big Five personality traits, narcissism, and hireability will be
consistent across raters.

Based on the rich variety of information found on Twitter, we propose that it will be

useful a mechanism to assess personality. Kluemper & Rosen (2009), and Kluemper et al.,

(2012) found initial support in a set of similar studies using Facebook profiles. This leads to our

second set of hypotheses:

Hypotheses 2a-2f: Other-ratings of the Big Five personality traits and narcissism will correlate
with self-ratings.

To be consistent with past studies done on the impact of personality and job performance

(Barrick & Mount, 1991) the researchers examined whether self-reported personality has links to

organizationally relevant constructs like hireability. The next set of hypotheses are as follows:

Hypotheses 3a-3f: Self-reported Big Five personality traits and narcissism will correlate with
hireability, with neuroticism and narcissism correlating in the negative direction.

Our final set of hypotheses examine the other-rated personality and hireability

relationship. Do assessments of those making personality judgments differ from those making

hireability decisions? We believe that by assessing this relationship will lead to further insight

on what personality traits most impact hireabiliy.

Hypotheses 4a-4f: Other-rated Big Five personality traits and narcissism will correlate with
hireability, with neuroticism and narcissism correlating in the negative direction.
TWITTER, PERSONALITY & HIREABILITY 17496 8

METHOD

Participants

This study was conducted at a small private university in the midwestern U.S.

Undergraduate students were invited by the research team via email to voluntarily participate.

The email indicated that it would be a survey of personality and social media use. Both users and

non-users of social media sites were encouraged to participate. An incentive of $250 was

awarded to those who completed the survey, with five randomly drawn winners each receiving

$50. The online survey consisted of 62 questions, including: 1) what is your gender? 2) which of

these ten popular U.S. social media sites do you have accounts on? 3) 50 questions assessing Big

Five personality traits 4) 9 items addressing narcissism and 5) a final question that asked for the

subject’s Twitter handle.

The online survey was open for a total of one week during the fall 2013 semester, and

567 surveys were completed. The directions indicated that it was designed to determine the

personalities of those using various social media websites, and to determine the personality

differences between users and non-users of social media sites. Of the 567 completed surveys,

397 subjects (70%) were female, and the average number of social media sites out of the ten

presented was 4.38 sites. 152 subjects were selected for the final sample which included those

that had verified Twitter accounts, and which were public (accessible by anyone), and that had

Tweeted 200 or more times.

Measures

Self- and other-rated personality traits – Subjects completed the International Personality

Item Pool (Goldberg et al, 2006) to assess the Big Five personality traits. The 50-item scale

contains ten items for each of five dimensions (extroversion, agreeableness, conscientiousness,
TWITTER, PERSONALITY & HIREABILITY 17496 9

openness to experiences and neuroticism). Subjects answered questions such as “I feel

comfortable around people” (extroversion), “I have a good word for everyone” (agreeableness),

“I am always prepared” (conscientiousness), “I believe in the importance of art” (openness to

experience), and “I often feel blue” (neuroticism) using a 5-point Likert-type scale ranging from

1 (strongly disagree) to 5 (strongly agree). Narcissism was assessed by having subjects complete

the nine-item narcissism subscale of the short dark triad (SD3) with statements like “People see

me as a natural leader” (Jones & Paulhus, in press). The personality raters completed the same 59

items as the subjects to ensure consistency between subjects and raters.

Hireability ratings – This commonly used organizational practice to assess a job

candidate’s attractiveness was measured using three items drawn from Stevens and Kirstof

(1995). Questions included “How highly do you regard this applicant?”, “How likely would you

be to offer this person a job?”, and “How attractive is this applicant as a potential employee?”

The items were rated on a 5-point scale, with 1 representing low and 5 representing high. Raters

were instructed to not base hireability assessments on any particular job or organization, but as a

general measure of how likely they would be to employ the subject based on their Twitter

profile.

Raters

Three qualified raters were used to assess personality, and separately three qualified

raters were used to assess the hireability of the 152 subjects. The three raters recruited to assess

personality were first-year Ph.D. students at a large public southeastern U.S. university who

completed a seminar in organizational behavior in which personality traits were extensively

covered. These raters were active users of social media sites and familiar with how personality

traits were typically assessed. The personality raters appear as authors on this study. To limit any
TWITTER, PERSONALITY & HIREABILITY 17496 10

bias they might have when conducting the assessments, they didn’t learn the purpose of the study

until after the assessments were completed.

The personality raters were given written instructions saying that the study was to

determine whether personality could be assessed by reviewing one’s Twitter profile. In addition

to the written instructions, each rater was asked to watch a six-minute video created by the lead

author to demonstrate the various aspects of a Twitter profile that could be used to make

personality assessments, including: whether or not the subject had a profile picture and/or header

picture, the header information supplied by the subject, the number of Tweets, the number of

people the subject was following, the number of people that were following the subject, the

subject’s favorite tweets, the lists the subject were on, the pictures and videos uploaded by the

subject, the conversations the subject engaged in, and most importantly the content of the

subject’s Tweets and Retweets. Raters were asked not to communicate with each other about the

project until the assessments were complete to ensure rater independence.

The personality raters were given a schedule of when they were to assess the Tweets,

their job taking place over a two-week period, to ensure that all three raters were seeing almost

identical information for each subject. To minimize rater fatigue, they were asked to conduct

their assessments for no more than one hour at a time, and for no more than two hours a day,

once in the morning and once in the evening. The raters answered the same online survey as the

subjects in the study, so raters were instructed to review the survey prior to making the first

assessment, in order to be familiar with the questions they themselves would complete. Raters

reviewed the subject’s Twitter profile for five minutes, at which point they then assessed the

subject’s personality based on their impressions by answering the online survey.

TWITTER, PERSONALITY & HIREABILITY 17496 11

The three raters recruited to assess the hireability of each subject all worked in the field

of human resources and had experience in the recruiting/hiring of candidates. The hireability

raters all held business degrees and were active users of social media sites. They represented

three different industries: energy, information systems and transportation. These raters were

given written instructions and asked to consider the hireability of the subjects with no particular

job category or organization in mind. This more general assessment of hireability was done using

the same aspects of the Twitter profile that the personality raters used. Therefore, the hireability

raters were also instructed to watch the video produced by the lead author. The complete purpose

of the study was not presented to the hireability raters, and they did not know of the existence of

the personality raters. Similar to the personality raters, the hireability raters were asked not to

communicate with each other about the project until the assessments were complete to ensure

rater independence. For each subject in the study, the hireability raters visited their Twitter

profile and were asked to skim it for three minutes. After that time, the raters answered three

questions from Stevens and Kirstof (1995) to assess hireability. The hireability raters had a

schedule that coincided with the personality raters to try to ensure that all raters were seeing the

same basic information when making their assessments.

RESULTS

In order to assess inter-rater reliability for the raters’ assessments of personality and

hireability, average measures intraclass correlation coefficients (ICCs) were used. The two-way

mixed model was selected as the same three raters rated all 152 subjects, and the consistency

measure was used over the absolute agreement measure, as the mean value was of more

importance than any individual rater’s assessment (Shrout & Fleiss, 1979; McGraw and Wong,

1996). In a similar study in which raters used Facebook profiles to make personality and
TWITTER, PERSONALITY & HIREABILITY 17496 12

hireability assessments, Kluemper et al. (2012) argued for the use of ICC threshold values of .50

for other-rated personality traits, and reported ICC values ranging from a low .47 for

extroversion to a high of .67 for agreeableness.

----------------------------------------------------------------

Insert Table 1 about here

----------------------------------------------------------------

For the six personality variables examined in this study, all ICC values were greater than .50,

with a low of .53 for neuroticism and a high of .79 for extroversion. This is deemed acceptable

for three raters using an unstructured method of personality assessment as shown in Table 1.

H1a-f, which stated that there would be acceptable levels of inter-rater reliability among the

three personality raters for the six personality traits using Twitter profiles as the means to assess

personality, was supported. H1g examined whether or not acceptable levels of inter-rater

reliability would be reached for the hireability construct, again with an ICC threshold value of

.50 as the target. In the Kluemper et al. (2012) study, an ICC value of .64 was found using the

three raters. In this study, the ICC value for hireabilty across raters was .69 (see Table 1),

providing support for H1g.

----------------------------------------------------------------

Insert Table 2 about here

----------------------------------------------------------------

H2a-2f examined whether raters can predict one’s personality from a view of his/her

Twitter profile. Table 2 shows the intercorrelations among study variables and shows that for all

six personality variables, self- and other-rated personality were significantly correlated, with a

low value of .19 for conscientiousness to a high value of .54 for openness to experience as shown

in Table 2.
TWITTER, PERSONALITY & HIREABILITY 17496 13

H3a-3f examined whether self-rated personality traits and hireability were related. It was

hypothesized that self-rated extroversion, agreeableness, conscientiousness, and openness to

experience would positively correlate, and that self-rated neuroticism and narcissism would

negatively correlate with hireability. The hypotheses for self-rated agreeableness and neuroticism

were supported, while the others were not. In the case of self-rated openness, the trait was

significantly but negatively correlated with hireability, which was contrary to the hypothesized

direction of the relationship.

H4a-4f examined whether other-rated personality traits and hireability were related. It

was hypothesized that other-rated extroversion, agreeableness, conscientiousness, and openness

to experience would positively correlate and that other-rated neuroticism and narcissism would

negatively correlate with hireability. Three of the six hypotheses were supported with other-rated

agreeableness and conscientiousness correlating positively and other-rated neuroticism

correlating negatively to hireability.

----------------------------------------------------------------

Insert Table 3 about here

----------------------------------------------------------------

Nineteen of the twenty-five hypotheses were supported, as shown in Table 3,

demonstrating inter-rater reliability between the three raters on the six personality traits and

hireability, showing support for the ability of other-raters to predict self-rated personality traits,

and linking both self-rated (agreeableness and neuroticism) and other-rated personality traits

(agreeableness, conscientiousness and neuroticism) to hireability.

DISCUSSION

What insights can hiring organizations gain about you from your Twitter profile? This

study was an attempt to discover whether one could use Twitter to accurately predict personality,
TWITTER, PERSONALITY & HIREABILITY 17496 14

and how these assessments may impact a candidate’s hireability. Using the framework of

Hogan’s socioanlaytic theory, it was argued that self-reported personality is akin to one’s identity

and reputation is similar to other-rated personality. In the present study, the link between identity

and reputation was shown by correlating self-reported and other-rated personality, and the

impact of one’s reputation on hireability was examined. This is important, because surveys of

practitioners show that social media sites like Twitter and Facebook are being used extensively

in the screening process. For example, a Reppler (2011) study showed that 91% of the 300 HR

professionals surveyed use social media sites for this purpose.

Using a sample of 152 subjects, we found that personality can be accurately predicted

solely using Twitter profiles. Three personality raters came to general agreement about the

subjects’ profiles and accurately predicted the Big Five personality traits and narcissism from

information gathered from Twitter. Unlike prior studies that used machine learning to draw

conclusions from processed tweets (Qiu et al., 2012; Golbeck et al., 2011; Querica et al., 2011;

Sumner et al., 2012), this study used trained raters to draw its conclusions. This is more

representative of what you would find in hiring organizations. HR staff rarely have access to the

computerized tools used in past studies, and would, in practice, make their assessments about the

suitability of job candidates. Unlike other studies, the raters in this study also used all available

information on each subjects’ profile, including pictures and videos, in addition to the Tweets, to

come to their conclusions. Using all available information gives HR personnel a deeper

understanding of each candidate being assessed.

Raters were asked to provide the researchers with a brief description of the criteria they

used to evaluate the subjects. Personality raters indicated that a subject who was high in

extroversion, for example, would have a high number of followers, group pictures, and
TWITTER, PERSONALITY & HIREABILITY 17496 15

conversations that were directed at many different people. A highly agreeable subject was one

that positively interacted with others, sent congratulatory or friendly tweets, and showed

empathy to others. Those who were highly conscientious were involved in numerous activities,

put work ahead of social activities, showed self-discipline, and followed professional Twitter

accounts such as companies and/or news organizations. Someone who was high in openness had

a customized background, interesting header and profile photos, discussed travel and arts,

followed wisdom-sharing accounts, and had photos of interesting activities and travel

destinations. Those high in neuroticism posted critical comments about themselves, and

frequently talked about their bad mood, negative feelings towards life in general, and discussed

high levels of stress and mental anguish. Narcissistic individuals made self-absorbed comments

that involved bragging and self-promotion, had pictures that focused solely on themselves,

followed a large number of celebrities, and often one-upped their followers in conversations.

Hireability raters all indicated that they looked at the profile picture and header picture

for signs of professionalism. This and the level of activity and responsibility one tweeted about

pointed to how conscientious the subject was. Two raters also mentioned that if the subjects were

following news organizations and companies, they were deemed as more conscientious. Mood of

the tweets was important to these raters, and revealed how neurotic the subject was. Profanity

was universally mentioned as a negative to hireability and lowered the rating for those subjects

that used it frequently in their tweets. Such subjects were found to be low in agreeableness. Not

surprisingly, the other-rated personality traits that significantly correlated with hireability were

neuroticism, agreeableness, and conscientiousness.

In summary, our contributions are as follows: 1) this is one of the first studies to use

raters to assess both the personality and hireability of subjects using Twitter profiles, 2) this
TWITTER, PERSONALITY & HIREABILITY 17496 16

study differs from past research in that there were two sets of multiple raters to assess both

personality and hireabilty; machine learning was not a factor in determining the results of the

study (mimicking what would be found in organizations), 3) inter-rater reliability for all six

personality traits and hireability were consistent with past studies that used SM technologies

other than Twitter, proving the validity of using Twitter to predict personality and hireability,

and 4) raters were asked to provide statements regarding which aspects of the profiles were

influential in their ratings of subjects, so that the researchers could better understand how

assessments were made.

There are some limitations to our research. While providing support for the impact of

Twitter on personality prediction and hireability ratings, we do not suggest that Twitter should be

used in isolation, but rather as a supplement to traditional screening methods. Knowledge, skills

and abilities are still important factors in the hiring process, and should be assessed in other

manners. This study also asked the hireability raters to provide a general assessment of subjects

and not think of any particular job or organization when providing their ratings. In practice,

hiring managers would know what positions candidates had applied for, so they could use

Twitter as a supplement to see if the candidate demonstrates person-organization and/or person-

job fit. All subjects in this study were degree-seeking college-aged students, so there could be

different results should this study be applied to different age groups or those with different

education levels. It is important to note that hiring solely on the basis of SM profiles can also

lead to adverse impact, as suggested by Kluemper et al. (2012), Van Iddekinge et al. (in press),

and Acquisti and Fong (2013), and should not be done. Not all subjects in this study had Twitter

accounts (62% did), and not all of the accounts were public or had enough tweets to make
TWITTER, PERSONALITY & HIREABILITY 17496 17

accurate assessments, so about 27% of our original sample was used to draw the conclusions in

this study.

In conclusion, we found valuable support for using Twitter profiles to assess personality

and hireability. Our results show acceptable levels of inter-rater reliability, highly correlated self-

reported and other-rated personality, and highly correlated other-rated conscientiousness,

agreeableness, and neuroticism (in the negative direction) to hireability.

TWITTER, PERSONALITY & HIREABILITY 17496 18

Table 1 - Descriptive Statistics for Self- and Other-Rated Study Variables

Variable M SD Alpha ICC

Extroversion
Other-Rated 3.44 .67 .93 .79
Self-Rated 3.60 .72 .90
Agreeableness
Other-Rated 3.63 .56 .93 .68
Self-Rated 3.79 .53 .81
Conscientiousness
Other-Rated 3.42 .54 .93 .58
Self-Rated 3.65 .59 .84
Openness
Other-Rated 3.31 .57 .91 .70
Self-Rated 3.68 .68 .84
Neuroticism
Other-Rated 2.60 .44 .90 .53
Self-Rated 2.56 .74 .88
Narcissism
Other-Rated 3.04 .48 .88 .64
Self-Rated 3.16 .57 .74

Hireability 3.39 .71 .96 .69

ICC = Intraclass correlation coefficient two-way random effects model average measure
reliability is reported.
 TWITTER, PERSONALITY & HIREABILITY 17496 19

Table 2 – Intercorrelations among Study Variables

1 2 3 4 5 6 7 8 9 10 11 12 13
1 --- .09 .59** .59** -.12 -.46** .05 -.03 .23** .12 -.19* -.20* .04
2 --- .10 -.05 -.18* -.51** .81** .34** .01 .11 -.24** -.18* .30**
3 --- .59** .00 -.49** -.04 .06 .27** .20* -.11 -.28** .07
4 --- .10 -.38** -.07 -.06 .17* .19* -.03 -.11 .03

5 --- .21* -.25* .13 .09 .00 .54** .07 .08

6 --- -.46** -.11 -.14 -.17* .21** .34** -.13
7 --- .31** -.08 .02 -.25** -.17* .30**
8 --- .05 .11 .12 -.38** .63**
9 --- .18* .02 -.35** -.09
10 --- -.20* -.25** .19*
11 --- .10 .05
12 --- -.30**
13 ---

Significance Levels:
p < .05 *
p < .01 **

Variable List

1) hireability
2) other-rated extroversion
3) other-rated agreeableness
4) other-rated conscientiousness
5) other-rated openness to experience
6) other-rated neuroticism
7) other-rated narcissism
8) self-reported extroversion
9) self-reported agreeableness
10) self-reported conscientiousness
11) self-reported openness to experience
12) self-reported neuroticism
13) self-reported narcissism
TWITTER, PERSONALITY & HIREABILITY 17496 20

Table 3 – Summary of Hypotheses and Results

Inter-rater Reliability Trait ICC value Conclusion

H1a Extroversion (EXT) .79 Supported

H1b Agreeableness (AGR) .68 Supported
H1c Conscientiousness (CON) .58 Supported
H1d Openness to Experience (OPEN) .70 Supported
H1e Neuroticism (NEURO) .53 Supported
H1f Narcissism (NAR) .64 Supported
H1g Hireability .69 Supported

Relationship Correlation Conclusion

H2a Self-rated EXT  Other-rated EXT .38** Supported

H2b Self-rated AGR  Other-rated AGR .27** Supported
H2c Self-rated CON  Other-rated CON .19* Supported
H2d Self-rated OPEN  Other-rated OPEN .54** Supported
H2e Self-rated NEURO  Other-rated NEURO .34** Supported
H2f Self-rated NAR  Other-rated NAR .30** Supported

H3a Self-rated EXT  Hireability -.03 Not Supported

H3b Self-rated AGR  Hireability .23** Supported
H3c Self-rated CON  Hireability .12 Not Supported
H3d Self-rated OPEN  Hireability -.19* Not Supported
H3e Self-rated NEURO -  Hireability -.20* Supported
H3f Self-rated NAR -  Hireability .04 Not Supported

H4a Other-rated EXT  Hireability .09 Not Supported

H4b Other-rated AGR  Hireability .59** Supported
H4c Other-rated CON  Hireability .59** Supported
H4d Other-rated OPEN  Hireability -.12 Not Supported
H4e Other-rated NEURO -  Hireability -.46** Supported
H4f Other-rated NAR -  Hireability .05 Not Supported

Significance Levels:
p < .05 *
p < .01 **
TWITTER, PERSONALITY & HIREABILITY 17496 21

REFERENCES

Acquisti, A. & Fong, C.M. 2013. An experiment in hiring discrimination via online social
networks. Working paper 2031979, Carnegie Mellon University, Pittsburgh, PA.

Alexa, The Web Information Company 2014. The top 500 sites on the web. Retrieved, January
11, 2014 from http://www.alexa.com/topsites

Back, M.D., Stopfer, J.M, Vazire, S., Gaddis, S., Schmukle, S.C., Egloff, B. & Gosling, S.D.
2010. Facebook profiles reveal actual personality, not self-idealization. Psychological Science,
21: 372-374.

Barrick, M.R. & Mount, M.K. 1991. The Big Five personality dimensions and job performance:
A meta-analysis. Personnel Psychology, 44: 1-26.

Furr, R.M. & Funder, D.C. 2007. Behavioral observation. In Robins, R.W., Fraley, R.C. &
Kruger, R.F. (Eds.), Handbook of Research Methods in Personality Psychology, New York:
Guildford Press.

Golbeck, J., Robles, C., Edmondson, M. & Turner, K. 2011. Predicting personality from
Twitter. Paper presented at the IEEE International Conference on Privacy, Security, Risk, Trust,
and IEEE International Conference on Social Computing.

Goldberg, L.R., Johnson, J.A., Eber, H.W., Hogan, R., Ashton, M.C., Cloninger, C.R. & Gough,
H.G. 2006. The international personality item pool and the future of public-domain personality
measures. Journal of Research in Personality, 40: 84-96.

Grasz, J. 2012. Thirty-seven percent of companies use social networks to research potential job
candidates, according to new CareerBuilder survey. CareerBuilder. Retrieved, January 11, 2014
from
http://www.careerbuilder.com/share/aboutus/pressreleasesdetail.aspx?id=pr691&sd=4%2f18%2f
2012&ed=4%2f18%2f2099

Hogan, R. T. 1996. A socioanalytic perspective on the five-factor model. In J. S. Wiggins (Ed.),

The five-factor model of personality (pp. 163–179). New York: Guilford Press.

Hogan, R. T. (2007). Personality and the fate of organizations. Mahwah, NJ: Erlbaum.

Hogan, R. & Shelton, D. 1998. A socioanalytic perspective on job performance. Human

Performance, 11: 129-144.

Jones, D.N. & Paulhus, D.L. (in press). Introducing the short dark triad (SD3): A brief measure
of dark personality traits. Assessment.
TWITTER, PERSONALITY & HIREABILITY 17496 22

Karl, K., Peluchette, J. & Schlanegel, C. 2010. Who’s posting Facebook faux pas? A cross-
cultural examination of personality differences. International Journal of Selection and
Assessment, 18: 174-186.

Kluemper, D.H. & Rosen, P.A. 2009. Future employment selection methods: Evaluating social
networking websites. Journal of Managerial Psychology, 24: 567-580.

Kluemper, D.H., Mossholder, K.W. & Rosen, P.A. 2012. Social networking websites,
personality ratings, and the organizational context: More than meets the eye? Journal of Applied
Social Psychology, 42: 1143-1172.

McGraw, K.O. & Wong, S.P. 1996. Forming inferences about some intraclass correlation
coefficients. Psychological Methods, 1: 30-46.

Quercia, D., Kosinski, M., Stillwell, D. & Crowcroft, J. 2011. Our Twitter profiles, our selves:
predicting personality with Twitter. Paper presented at the IEEE International Conference on
Privacy, Security, Risk, Trust, and IEEE International Conference on Social Computing.

Qui, L., Lin, H. Ramsay, J., & Yang, F. 2012. You are what you tweet: Personality expression
and perception on Twitter. Journal of Research in Personality, 46: 710-718.
Raskin, R.N. & Hall, C.S. 1979. A narcissistic personality inventory. Psychological Reports, 45:
590-590.
Reppler. 2011. Managing Your Online Image Across Social Networks. Retrieved, January 11,
2014 from http://blog.reppler.com/2011/09/27/managing-your-online-image-across-social-
networks/

Saylin, G.M. & Horrocks, T.S. 2013. The Risks of Pre-employment Social Media Screening.
Society for Human Resource Management. Retrieved, January 11, 2014 from
http://www.shrm.org/hrdisciplines/staffingmanagement/Articles/Pages/Preemployment-Social-
Media-Screening.aspx

Schwartz, H.A., Eichstaedt, J.C., Kern, M.L., Dziurzynski, L., Ramones, S.M., Agarwal, M.,
Shah, A., Kosinksi, M., Stillwell, D., Seligman, M.E.P. & Ungar, L.H. 2013. Personality, gender,
and age in the language of social media: The Open-Vocabulary approach. PLoS ONE, 8:
e73791.

Shrout, P.E. & Fleiss, J.L. 1979. Interclass correlations: Uses in assessing rater reliability.
Psychological Bulletin, 86: 420-428.

Society for Human Resource Management. 2008. Online technologies and their impact on
recruitment strategies. SHRM Staffing Research, July-September. Retrieved, January 11, 2014
from
http://www.shrm.org/research/surveyfindings/documents/social%20networking%20flyer_staffin
g%20conference_final1.pdf
 TWITTER, PERSONALITY & HIREABILITY 17496 23

Statistic Brain 2014. Twitter Statistics. Retrieved, January 11, 2014 from
http://www.statisticbrain.com/twitter-statistics/

Stevens, C.K. & Kirstof, A.L. 1995. Making the right impression: A field study of applicant
management during job interviews. Journal of Applied Psychology, 80: 587-606.

Sumner, C., Byers, A., Boochever, R., & Park, G.J. 2012. Predicting dark triad personality
traits from Twitter usage and a linguistic analysis of tweets. Paper presented at the IEEE 11th
International Conference on Machine Learning and Applications.

Twitter.com 2014. Getting started with Twitter. Retrieved, January 11, 2014 from
https://support.twitter.com/articles/215585#

Van Iddekinge, C.H., Lanivich, S.E., Roth, P.L. & Junco, E. (in press). Social media for
selection? Validity and adverse impact potential of a Facebook-based assessment. Journal of
Management.

View publication stats