Ch 14 of text
Urine Tests
(a) Relatively inexpensive (b) results returned quickly (24 hrs) (c) testing convenient for applicant -used most often
Research shows the typical employee-thief:
(a) is more tempted to steal (b) engages in many common rationalizations for theft (c) would punish thieves less (d) often thinks about theft-related activities (e) attributes more theft to others (f) shows more inter-thief loyalty (g) is more vulnerable to peer pressure to steal
CWBs represent a wide array of illegal or unethical conduct
-Theft of money, equipment, other resources -Illicit drug use -Violence -Cursing at others -Sabotaging company equipment -Purposely doing work incorrectly
No adverse impact found with use of personality tests, but two legal issues:
1. ADA prohibits pre-employment medical exams; because psychologists can use personality tests to diagnose mental health deficiencies, such tests could be seen as medical exams EEOC: If test includes scales with psychiatric disorders, likely to be a medical exam If test focuses on general traits, not a medical exam 2. Privacy is an issue for employees for public institutions, as the are protected by the U.S. Constitution in their employment Several states have constitutional privacy protection acts or a statutory right to privacy; these apply to both public and private employers In one case, a personality test asking questions about an applicant's sexual inclinations or religious views for store security positions were banned
A large meta-analysis, focused on predicting job performance, concluded:
1. Conscientiousness & emotional stability were valid predictors of overall work performance over all studies; these are universal predictors across all jobs 2. Extraversion was a valid predictor for some occupational groups such as managers, and specific criteria, such as training performance
One 2006 study found regarding drug use:
1. Marijuana used by 11.3% of workforce 2. Cocaine used by 1.0% of workforce 3. One of four psychotherapeutic drugs used by 4.9% of the workforce 4. Marijuana use had caused impairment to10.6% of workforce 5. Cocaine use had impaired 0.9% of workforce 6. One of four psychotherapeutic drugs impaired 2.2%
Other meta-analysis, examining unique criteria concluded:
1. When selecting effective leaders from ineffective ones, candidates high in extraversion, openness to experience and emotional stability experience the greatest success on the job 2. Conscientiousness, emotional stability and agreeableness useful to predict teamwork and performance 3. All Five-Factor Model traits except openness to experience predicted success as an expatriate or an entrepreneur; extraversion is a particularly important predictor 4. Deviant behavior less likely from those scoring high in conscientiousness and agreeableness 5. These studies investigated the process through which personality affects job performance; the primary means through which personality operates is motivation and interpersonal competence; conscientiousness and emotional stability related to 3 different measures of performance motivation
Major questions about drug testing
1. it represents an invasion of privacy True, but only an issue if the employer is a governmental institution 2. it constitutes unreasonable search & seizure Limited application to private employers; but does apply to governmental institutions 3. is a violation of due process If authorized, professionally conducted testing not a violation 4. drug users are protected under the ADA 5. Testing may violate the Civil Rights Act 6. Testing my violate the National Labor Relations Act
Guidelines for Drug-Testing with Job Applicants
1. limit testing to positions which have major safety implications; testing may be required for those 2. Private employers have more latitude unless they are subject to federal regulation 3. Private employers have more flexibility with job applicants than with existing employees 4. a combination of screening and confirmatory tests are necessary 5. obtain written consent of the job applicant prior to testing; if positive result, allow person to explain; physician's verification may be required; offer test by a lab of their choosing 6. use standardized procedures, applied to all job applicants; do not exempt anyone 7. program should be designed and reviewed periodically to ensure privacy for those tested; results must be kept confidential, especially from supervisors of existing employees, unless a defined need to know
Issues have arisen concerning the scoring and use of information obtained from these tests:
1. reliability of an individuals responses at two different times 2. impact on person's score due to total number of responses, especially on tests with free response 3. scoring of information, especially for free response 4. few HR specialists trained in these tests
We have drawn on personality psychology to help you use personality data appropriately; this requires understanding
1. traits vary greatly in the extent to which they influence behavior 2. the situation also has an important influence on a person's behavior 3. measuring personality is challenging; try to reduce "faking good" or cheating on Web examsThree most commonly used methods:
Phase 2: This is Not Working, and It's Your Fault, Not Mine!
1981, research study focused on failure of ACs to show correlations among dimension ratings If same dimension tested on 3 exercises, correlation between ratings for each should be high (convergent validity); but it wasn't Many possible reasons Correlations among measures of what are thought to be separate dimensions are not error but an accurate measure of how well the participant is performing
A large meta-analysis, focused on predicting job performance, concluded:
3. Agreeableness and openness to experience demonstrated modest validity overall; but each trait was related to a specific criterion: Agreeableness was related to teamwork Openness was associated with training performance
False Positives
40 to 70% of applicants commonly fail integrity tests, and a large number of those may be incorrectly classified t may be in the interest of the publisher to avoid false negatives, so may set low cutoff score It may be in the interests of the employer to avoid false positives, so may set high cutoff score. -Because of implications of false positives, use of a personality inventory better than integrity test
Procedures
A pretest discussion with examinee that covers all questions in the test to ensure understanding Polygraph is attached, exam conducted; questions may be repeated
Phase 1: We Love Assessment Centers!
ACs' ratings predictive of movement into middle-level management Also reduced racial and gender differences Some evidence ACs could produce adverse impact For 25 years after development of Acs at AT&T, selection specialists loved them
What is Genetic Testing?
Analysis of human DNA, chromosomes, proteins or metabolites that detect genotypes, mutations, more Produces genetic information on (a) individuals (for our purposes, job applicants) (b) family members of these individuals (c) the health histories of the individuals' family members
Determining Dimension and Overall Evaluation Ratings
Assessor combines the data on the same dimension across exercises Uses PEDRs to develop DRs
Categorizing Participant Behavior
Assessor must record behavior of participant under the proper dimension
Determining the Rating of Participant Behavior
Assessors must use consistent rating scales
Realizing Prediction Gains over Time
Benefits will improve with employee's age
Accuracy of Chemical Tests
Chemical tests should be completely accurate False positives may occur in an initial test, but should be eliminated by follow-up test
Appropriate to think of integrity tests as measuring three of the Big-Five Factors:
Conscientiousness Agreeableness Emotional stability
Using "At-Work" Personality Tests
Contextualize personality items so that they measure the test-taker's behavior at work; leads to superior predictive validity
Last step: of training assessors
DRs used to form the OAR of the participant's ability Most critical dimensions weighted more heavily
Overt types ask such questions as
Do you think it is okay for workers to use "soft" drugs at work if it does not cause poor job performance? In past 6 months, how often have you used marijuana at work? In past 6 months, have you brought cocaine to work even though you did not use it at work? Almost no research on validity of this test
Drug use:
Drug & alcohol use of concern to employers since 1960's Drug testing increasingly common, especially for those with safety sensitive jobs or government contracts negatively associated with job performance Positively correlated with accidents, injuries, absences, involuntary turnover, job-withdrawal behaviors
Faking in Personality Inventories
Empirical evidence is contradictory as to whether candidates intentionally alter their responses to get hired Contradiction may be caused by design of research Little evidence that faking affects the predictive validity of personality inventories Self-reports may be less accurate than anticipated for reasons other than faking
Faking "Good" during Selection
Faking good does not substantially alter predictive validity; but including warnings in the instructions that faking on the test can be detected tends to discourage it
Define Work Effectiveness as Retaining Productive Employees
Focus on the Five-Factor Model to comprehensively measure an applicant's personality, ensuring predictive validity Other researchers recommend using facets to identify a more specific personality trait Relevance of the Work Context Both personality and the situation determine behavior
Genetic Testing
In 2003, Human Genome Project completed mapping of 25,000 genes Specific genes linked to certain diseases For some diseases, link is strong; for others, weak These results will continue to have important implications for health care and employment Implications of genetic testing for employers when selecting among job applicants Some employers have used genetic information to predict which most likely to miss work due to illness, disease
Observing the Behavior of Participants
Initial reaction of assessor is to immediately make judgment, but must instead focus on recording behavior
Three types of questions used in a polygraph test
Irrelevant, non-emotional questions Emotional control questions Questions about behavior of interest
Define Personality Traits in Terms of Job Behaviors
It is necessary to have a job-related basis for measuring personality Job analysis information can identify job tasks, interactions, working conditions, etc. From these data, an HR specialist should be able to infer personality traits needed
The Appropriate Use of the Interview
Limit assessment of personality to traits needed for job through job analysis Behavior description interview appropriate
Determining the validity of integrity tests is difficult; which criterion variable should we correlate?
Many CWBs not reported in employee records Self-reporting may not be an accurate measure Integrity tests valid for prediction of theft, general CWBs and work performance
Validity across Jobs
Mental ability tests are valid across a large variety of jobs and can serve as useful selection instruments Table 11.5 reports the results Mental ability tests increase the predictability of job performance as the job becomes more complex
Implications for Selection
No longer necessary to conduct validity studies within each organization (contra Uniform Guidelines); the same predictor can be used across all organizations Cognitive ability tests are valid predictors for many jobs, but magnitude of correlation varies from job to job Scores from cognitive ability tests as good a predictor of job performance as scores from a composite test of specific abilities for a job Large differences exist in mean scores between white and black respondents and between white and Hispanic respondents Using these tests as selectors usually results in a higher number of whites being hired - the very definition of adverse impact A pareto analysis can lead to improvement in diversity hiring without reducing the validity
Source of Measurement
Observer ratings predict performance better than self-reports; use the interview, letters of reference, and headhunters who talk to former associates
Observer Ratings of Personality
Observers can rate an applicant's personality even after brief conversation, with modestly high agreement on rating by one who knows applicant Observer ratings show a persons' public self, the way we are perceived by others Self-reports reflect our perceptions of self Observer ratings enhance prediction of job performance, often twice as well as self-reporting
Oral Fluid Tests
Oral swab on inside of check; saliva analyzed for drugs Barring lab error, this test is always accurate
Integrity tests aimed toward identifying job applicants predisposed to engage in counterproductive behaviors have two forms
Overt integrity tests directly asking for information Personality-oriented measures, inventories measuring traits linked to certain CWBs
Trade-offs between Validity and Disparate Impact
Personality tests produce few large differences between demographic groups, avoiding disparate impact; continue to monitor results
examples of Overt Integrity Tests
Personnel Selection Inventory, the Reid Report, the Stanton Survey
Limitations of chemical tests
Positive result means only presence of drug above threshold level set for detection Does not determine how much of the drug was used, how frequently, how long ago, the circumstances or the level of impairment in performance People differ in absorption, metabolism and tolerance Not correct to assume that presence of drug means it affected performance
Ratings from Other Observers
Ratings better from those who know the subject Obtain reputation-infused ratings from supervisors or coworkers Perhaps a headhunter can visit with people who worked with the candidate A letter of reference may also served as a way to assess the candidate's personality Study: A structured letter of reference predicted job performance better than self-reporting
Validity of Personality Tests as a Predictor
Recent studies have challenged the notion that more of a trait is always better for work performance It is possible to have too much of a good thing Validity may be highest in the middle range of scores
Personality-oriented tests correlated more highly with the Big-Five dimensions than overt tests:
Recent work postulates a sixth personality dimension: Honesty-Humility, which measures fairness, sincerity, lack of conceit and lack of greed The Six-Factor model outperformed the Five-Factor
Projective Techniques
Require verbal responses, but major differences; questions and answers are intentionally ambiguous, instructions brief and vague Respondents presented with inkblots or pictures and asked to make up a story about them Or given sentence stem and asked to complete: "My father..." or "My favorite..." Respondents' projections expose personality
Hair Analysis
Samples of hair including the root are taken from applicants to be tested More expensive, but use rising Drug use can be detected over a longer period of time than urine testing Each inch of hair shows usage for 90 days For cannabis, 1 to 3 months
Three most commonly used measurement methods:
Self-report inventories (produce insignificant or lower validity coefficients than other source) Judgment of interviewers Judgment of other observers (especially former coworkers or supervisors)
Incremental Validity over Other Selection Measures
Superior prediction results when personality is combined with other predictors
There are a variety of countermeasures examinees can use to avoid detection
Tensing muscles, biting tongue, flexing toes, etc. Mental countermeasures like relaxation
Forced-Choice Inventories
Test-takers must choose the most liked item of 2-4 equally desirable items; faking reduced A few studies yield predictive validities as good as or better than Likert-scale self-report scales Issues to be resolved, but another option
Relationship Between Overt and Personality-Oriented Tests: Do these measure the same constructs?
The integrity tests showed a great deal of similarity in what they measured Overt tests correlated more highly with honesty and supervision attitudes; personality tests correlated more highly with self/impulse control, home life/upbringing, risk taking, diligence and emotional stability
Paper-and-Pencil Integrity Tests
These have been used for 65 years These are legal under federal law, but a few states have prohibited them or restricted use
Phase 3: Well I Might Have Made A Mistake And We May Be Okay After All
This phase began early 2000s Major theme is that the measurement model was incorrect. PEDRs are not the appropriate measure; focus should be on the DRs. AC OAR scores were found to be correlated consistently with various measures of work performance
The Genetic Information Nondiscrimination Act of 2008 (GINA)
Title II prohibits employers, and others from discriminating against employees & job applicants because of genetic information: "An employer may never use genetic information to make an employment decision because genetic information is not relevant to an individuals' current ability to work" But complaints filed with the EEOC
The Appropriateness of the Selection Instrument
Use only tests with enough developmental information to show that it measures the personality traits it is said to assess Review reliability, item statistics, etc. Some tests sold expensive but useless Can respondent learn "correct" responses? Better to use instrument for which appropriate responses are not apparent
Cheating during Selection
We recommend that even web tests be proctored to enhance standardization of the testing procedures and to secure test security
Role of Technology and Global Trends
Web-based personality tests result in high levels of internal consistency; scores are equivalent to paper-and-pencil scores Also evidence of equivalence between proctored and unproctored examinations, but professional standards require proctoring Further innovation in progress Monitor innovations Monitor EEOC implications
"Bad" behaviors
are categorized as counterproductive work behaviors (CWBs)
Situational Judgment Tests (SJTs)
are verbal simulations, regarded as low-fidelity because they are almost exclusively descriptions of work situations rather than actual replications Ask the respondent to choose among multiple-choice alternatives
Health insurers used genetic information to
deny insurance or refused to pay for treatment, calling genetic information a "pre-existing" health condition Case involving employers' unauthorized testing of black applicants for sickle cell anemia raised issue in federal government
Before federal law, based on genetic information:
employers denied jobs, fired existing employees insurers denied coverage based on pre-existing condition
Before protective laws were passed
employers refused to hire affected applicants to avoid health insurance costs
Neuroscientists studying the brain have begun to
examine relationships between neural circuits and development of diseases (Alzheimer's), behaviors (racism, cooperativeness) and personality characteristics (extraversion, introversion) One day, it will be possible to determine the probability that certain neural wirings are linked to specific outcomes The same concerns arise here as did with genetic testing and the human genome
Reactions other than
guilt can trigger an emotional response
Integrity tests a:
hierarchical model with 3 levels and can measure all three levels
A polygraph (lie detector)
is a machine that measures physiological responses of examinees that accompany their verbal responses to direct questions -This data, and opinion of polygraph operator, used to evaluate truthfulness regarding the respondent's past or present behaviors
Major drawback to polygraph
is frequency of false-positive results and there can also be false negatives
Faking
may be especially possible for overt tests that directly ask about attitudes and conduct, but faking does not seem to make a difference in the validity of the test But this may not indicate the full effect of distortion on selection devices
Under ADA (integrity testing)
personality inventories designed to detect mental health issues are a prohibited medical exam (Ex: MMPI) Tests may not include questions about previous drug or alcohol use
Employee Polygraph Protection Act of 1988
restricts use of the polygraph for selection to police work, security, and the like
Assumption is that
there is a common personality pattern, and it can be identified - subset of items are identified as an integrity scale, and these are used to measure the test-taker's integrity Test less transparent; reduces faking Rejected applicants need not be rejected for dishonesty
Assuming a certain cost and applicant pass rate:
use of an integrity test would lead, on average, to a savings of $201,470 per 1,000 job applicants
Integrity testing is
useful in testing for certain undesirable work behaviors -But there is some disagreement regarding the strength of the relationships between integrity test scores and certain undesirable employee behaviors