To Be or Not To Be: Personality and Job Performance

Personality and job performance are often very misunderstood. For one thing, there is a tendency to look at personality and make assumptions about job skills. For another, there is a tendency to assume that “style” or “type” tests relate to future job performance. Both assumptions can lead to hiring mistakes. Evidence or Inference? Let’s think about personality-type tests (for the purists, I’m lumping traits, motivations, and attitudes into one big category to simplify this article). An applicant is asked to answer questions like, “Do you like to solve problems? Were you good at math and science? Do you enjoy puzzles?” We add up the scores and miraculously conclude, “This applicant is good at solving problems, excels at math and science, and enjoys solving puzzles!” Wrong? Right? Who knows? All we know is how the applicant described him or herself. Nothing more and nothing less. And it tells us nothing about whether he or she is any good at problem solving! This is the first fundamental error in using a pencil and paper test: we never really know if the person is telling the truth, let alone if the person really has the skills. If you ask managers to rate an employee’s overall performance, independent studies show that about 50% of the manager’s rating can be attributed to personality factors. Why? Here are some research summaries from independent hiring scientists:

• All other things being equal, managers are biased to give higher ratings to subordinates they like than to subordinates they do not like.
• Performance tends to increase when individual personality fits job personality.
• Poor job/person or manager/subordinate “fits” can be a major source of job dissatisfaction.
• Job ratings are about 50% hard ability and 50% schmoozing.
• General personality and job personality theories may have some overlap, but tend to have very different applications.
• A personality difference between employees does not always indicate a performance difference.
• The association between scores on a “big five” type personality test and problem-solving skills are only about 2% (interpersonal skills are only about 8%).
• A personality test developed to predict communications, style, or type is not the same as a personality test developed to predict job performance.
• Personality test users should have statistical evidence that test scores strongly predict job performance.

In summary, personality is important, but only when you use it correctly. Personality as a Hiring Tool The first challenge of using a personality test is to decide which personality factors are important. This sounds easy, but nothing could be further from the truth. For example, there was a truck assembly plant in Canada that hired people based on having high teamwork skills. Within one year their entire workforce was highly team oriented. Problem was, they were so team-centered that they wouldn’t make a decision without 100% agreement. They also wouldn’t meet without 100% attendance and wouldn’t confront problems for fear of upsetting a co-worker. Another organization decided to hire only dominant salespeople. Within a few months, every new salesperson had the sensitivity of a shark, openly challenged management, and tended to offend customers. A third organization wanted to hire a new operating president to run its U.S. facilities. This person had a personality that was highly engaging ó but “dark-side” personality testing showed he would probably be manipulative and undermine corporate authority. The right personality for the job is always a combination of job fit and job attitude that varies with job type and organizational culture. The most significant work on matching personalities to jobs was published about 20 years ago, by Dr. John Holland. Dr. Holland discovered six personality types that could describe all occupations. He categorized these types as follows:

• Realistic (i.e., jobs including the risky, practical, conservative, frank and tangible)
• Investigative (i.e., jobs involving exploration, understanding, scientific, and research)
• Artistic (i.e., jobs involving creativity, openness to experience, intellectual, and innovative interests)
• Social (i.e., jobs that include helping, teaching, agreeableness, and empathy)
• Enterprising (i.e., jobs requiring status, persuasiveness, directing, and gregariousness)
• Conventional (i.e., jobs requiring routine, standards, practical, and orderliness)

Holland’s work was highly comprehensive, but had one shortcoming: it did not address job performance. Personality and job performance studies were all over the block. Some showed personality to be a major part of job performance, while others showed it to have nothing to do with performance. One scientist even wrote an article saying that personality measures were essentially worthless. Explanations for these differences came a few years later when scientists 1) realized that general personality was a poor predictor of job performance, 2) concluded that clinical diagnostic tests were totally inappropriate for predicting job skills, and 3) were finally able to analyze thousands of personality items (thanks to the evolution of computing power) by running them through a statistical “blender”. The blender approach identified five generic personality factors that encompassed the majority of personality items. It was first called the Five Factor model. Several modifications later it became known as the Big Five (B5).

The B5 factors include:

Note. Prototypical characteristics and adjectives taken from McCrae and Costa (1989), Mount et al. (1994), and Hogan (1991); items in parentheses define the opposite pole of each dimension. Table compiled by Harvey, Murry, and Markham for a presentation at the Annual Conference of the Society for Industrial and Organizational Psychology 1995 (May).

If you look carefully through these items, you might notice that some of the characteristics within certain factors are not intuitively clear (i.e., responsible and need for achievement; good-natured and sympathy; or, imaginative and depth of feeling). This is because the B5 is a computer-generated, factor-analytic model. It is not a well thought-out theory of job performance. Unfortunately, these “little technical details” don’t keep people (many so-called “experts” included) from claiming that the B5 is the “holy grail” of personality measures. The major contribution of B5 research, though, is that three of the five factors are consistently associated with job performance ó extraversion, neuroticism, and conscientiousness. In other words, high performers are sociable, not “nuts,” and care about quality. Good sense would advise taking the best from each body of research. So, if you combine Holland’s job fit findings with B5 performance findings, you get a personality test that measures job-related personality, incorporates highly stable (and hard-to-change) factors, and predicts three major areas associated with job attitude. (Oh, yes, you might also want to add a scale that measures faking). In Part 2, we’ll review which personality factors are critical, and present some case studies that show how personality varies with the job.