Columns in this Series:
Part 1
Part 2

Critical Personality Factors

Just what factors should a good personality test include? Translating the six "job fit" factors research and the three "job attitude" factors research into behavioral language yields a combination of ten items that do not fit cleanly into any single definition of personality.

I call these ten factors "attitudes, interests, and motivations," or AIMs:

• Problem solving. This scale represents the applicant's attitude toward solving complicated problems. A high score means he or she prefers jobs requiring a mental challenge and solving complex problems. Low scores mean he or she prefers to avoid mentally challenging positions.

• Idea generation/innovation. This scale represents the applicant's attitude toward free thinking and creativity. High scores indicate that he or she enjoys suggesting new ideas and creative processes. Low scores indicate a preference for systematic traditional work.

• Administration. This scale represents the applicant's preference for following rules and procedures. High scores mean he or she likes to follow precedent and established processes. Low scores indicate a tendency to break rules and work without structure or guidelines.

• Resistance to change. This scale represents the applicant's willingness to adapt and change. High scores represent resistance to change and a desire for stability and consistency. Low scores indicate readiness to change and adapt to whatever conditions might be.

• Teamwork. This scale indicates whether he or she prefers to work alone or with others. High scores indicate a preference for working in close-knit teams. Low scores indicate he or she likes solitary work.

• Expressiveness. This scale represents the applicant's interest in public contact. People who score high on expressiveness label themselves as outgoing and having many social contacts. Low scores indicate the person may not have the interest in being sociable.

• Impulsiveness. Impulsiveness is a measure of how fast a person likes to make decisions. High scores indicate an interest in making fast decisions and quick responses. Low scores mean a preference for slow response and postponed decisions.

• Perfectionism. This scale represents the person's attitude toward producing a perfect product. A little perfectionism goes a long way. But people with high perfection scores may never be satisfied enough with the final product, causing unnecessary delays and reductions in output. People with too little perfectionism, on the other hand, may be sloppy and unconcerned with quality.

• Attitude toward work. This scale represents how an applicant feels about working for an organization. Some people, for example, see the office as a battleground between good (themselves) and evil (everyone else). These people are either unable or unwilling to pull together for the common good, and care little about the customer. People with low scores sap energy and become destructive to both morale and productivity. People with high scores tend to see the organization as a positive place to work and contribute.

• Self-centeredness. This scale represents how much the candidate looks out for him or herself. High scores indicate someone who spends much of their time thinking about themselves and the impact of decisions on them personally instead of worrying about out-producing and outsmarting the competition. People with low scores on this scale indicate that they focus more on what other people feel than on what they consider important.

• Truthfulness. This scale shows whether the person was truthful or not. Scores at either end of the scale or the middle indicate whether the person was trying to make him or herself look good. Exceptionally low or high scores mean that all of the applicant's scores should be very carefully scrutinized.

Determining Which Factors Are Important

The only was to identify which factors are important is to break out the statistics book, crank up the old computer, give the test to jobholders, develop rating criteria, have managers rate the performance of each jobholder, and analyze the results. By the way, before your study can be considered trustworthy, you will need to determine which performance criteria can be trusted, and how many people to use in the sample. You will also need to examine minority groups for adverse impact and be able to understand and apply the basics of statistics. Too much trouble? Sorry, no shortcuts here. No study = hiring mistakes and the possibility of legal challenges.

A note about statistical studies: Statistics is a great way to identify important relationships, a great way to make bad data look respectable, and a wonderful cure for insomnia. Like it or not, though, stats are the only game in town. If you have not used statistics to validate test scores with job performance, or hired someone to do it for you, you might as well use a ouija board to predict performance (I also hear Miss Cleo might still be in business).

We recently had the opportunity to work with AIMs data gathered from two call centers. The first call center specialized in making outbound market survey calls. Their agents contacted people via phone, persuaded them to participate in an interview, asked prepared questions, and entered responses using pencil and paper or keyboards. Managers provided one, overall three-point rating of job performance on 139 agents. We divided people into three groups (i.e., the ones, the twos, and the threes) and examined the groups for differences. We found no differences between the low-rated and mid-rated group. However, the high-rated group had higher, statistically significant resistance-to-change scores, higher expressiveness scores, and higher teamwork scores. Out of ten possible factors, these three factors made the greatest difference in performance.

The second call center study involved 45 people who take inbound sales calls. The managers provided three ratings for this job: 1) a measure of acquired skills, 2) a performance rating, and 3) an overall summary rating.

Acquired skills were strongly related to interest in problem solving (+.33), perfectionism (+.28) and resistance to change (+.27). This means that people who were motivated to acquire more job skills were likely have higher AIM scores in those three areas. When overall performance was analyzed, low scores in idea generation (-.41) and high scores in attitude towards work (+.39) were strongly correlated. The negative relationship with idea generation suggested that people who wanted to "think outside the box" would probably not do very well in this job. Finally, the third manager rating (i.e., overall summary) showed idea generation ( -.36), resistance to change (+.31) and attitude towards work (+.30) were important. Taken together, this means that people rated as being "high performers" tended to score high in problem solving, perfectionism, resisting change, and attitude toward work. There was a negative relationship with idea generation.

These are important findings because they show:

• Personality factors are different depending on the nature of the job
• Personality factors have very strong correlations with manager ratings
• Personality factors can be a valuable source of hiring information
• It takes a formal study to identify personality factors important to hiring decisions.

• Dozens of studies have shown personality is exceptionally stable and slow to change. Twin studies (identical twins separated at birth and raised in different families) suggest that about 50% of personality is genetically influenced. This means that information gathered from a hiring personality test has limited use for coaching or training. What you see is what you get — like it or not.

• By examining patterns among personality scores, the hiring manager can gain valuable insight into applicants' attitudes, interests, and motivations. It provides the capability for building personality diversity (among non-critical factors) while maintaining personality consistency (among performance-critical factors).

• Recent interview research shows that interviews are only accurate predictors of extraversion. Even experienced behavioral interviewers were unable to assess neuroticism and conscientiousness accurately — two of the main attitudes associated with job performance.

• Personality testing does not provide "evidence" of job skills. Personality scores and skills are only slightly related. You should still use simulations or work samples to measure interpersonal skills, and case studies involving problem solving or planning to measure cognitive ability.

• Personality test scores allow hiring managers to make inferences about future employee performance (particularly from the view of the employee's manager). They are reasonably accurate predictors of "will do, " not "can do."