HRM

Personality faking research

"Can" people distort their responses (i.e., fake)? Yes, most research has indicated that scores can be altered, such as by instructing subjects to fake "Do" people distort their responses (e.g., applicants vs. incumbents)? 14 studies, in 7 applicants scored higher than non applicants (suggesting "yes"), 4 found that it does occur, but not as much as often thought, and 3 found no differences between applicants and non applicants Does faking affect criterion-related validity (i.e., how well personality scores predict performance)? 17 studies, 8 found that response distortion did affect the validity of the measures, 9 studies found that it did not

key steps to effective selection

3. establish a systematic process create checkboxes for the example ksaos and the selection measure and which one lines up most with each ksao=check ppt this shit make no sense Compensatory model - all applicants complete all selection stages and assessments; high score in one area can make up for a low score in another area Multiple hurdle model - only applicants with sufficiently high scores at each selection stage go on to a subsequent stage

Biographical Data Bio Data

A broad spectrum of an individual's background, experiences, interests, attitudes, and values With the context reasonably constant, the best predictor of job applicants' future behavior is their past behavior Only relevant (empirically significant) items are selected

cognitive ability tests

A quick clarification on terminology and definitions... IQ is a historical term that stands for "intelligence quotient"; term is popular with laypersons but is generally not used by scientists Mental ability and cognitive ability are current terms that scientists often use interchangeably General mental ability or intelligence refers to general intellectual capacity (often called "g" or GMA) "g" (intelligence) - the ability to reason, plan, solve problems, comprehend complex ideas, and learn from experience

big five and work performance

All jobs? Conscientiousness Sales performance? Extraversion Training performance and expatriate success? Openness to experience Customer service and teamwork? Agreeableness Many jobs (e.g., management, safety/security)? Neuroticism

work samples and situations

Ask applicant to complete some job activity, either behavioral or verbal, under structured testing conditions Tied directly to KSAs that underlie job tasks Pros: Strong validity (early estimate of .54, more recently .33); high acceptability; reduces the ability to fake or misrepresent skills Cons: Development difficulties (complex jobs); can be costly

personality work context

Bing, Davison, and Smothers (2014) Field investigation of the frame-of-reference effect at the item level within an organizational context Work-specific scales had higher validities than the noncontextual scales in predicting specific job performance dimensions e.g., persuasion NC with sales performance (r = .16) versus persuasion WS with sales performance (r = .24) Work-specific scales obtained incremental validity above and beyond the noncontextual scales in predicting overall job performance and a performance composite

Forecasting

Determine labor demand Derived from product/service demands Often based on specific skills or job categories e.g., aging population in U.S. would suggest greater demand for skills related to elder care Determine labor supply Internal - succession planning, internal movements caused by transfers, promotions, turnover, retirements, etc. External - several agencies make projections on labor market e.g., Bureau of Labor Statistics, DOL, NSF, etc. Determine labor surplus or shortage Compare supply and demand

Education

Does education predict performance? Years of education = .10 What about academic achievement (i.e. GPA)? Roth et al. (1996) - meta-analysis of 71 studies GPA and performance = .32 Time moderates: GPA and performance (1 year later) = .45 GPA and performance (2-5 years later) = .30 GPA and performance (6+ years later) = .11 Type of organization moderates: Education (.39), military (.29), business (.27), scientific (.24), medical (.23) Be careful though, can result in adverse impact due to differences in group means

Experience

Does job tenure predict performance? Studies range from .18 to .24 Does "experience" predict performance? Quinones et al. (1995) - meta-analysis of 22 studies Overall "work experience" and performance = .27 Amount (.43) and task-level (.41) based measures of experience had the highest correlations Does "prehire" experience predict performance? Iddekinge et al. (2019) - meta-analysis (k = 44, n = 11,785) Weak relationships with job performance (.06), training performance (.11), and turnover (.00) The authors, "caution organizations from selecting employees based on such measures unless more positive evidence emerges."

Expected labor surplus

Downsizing - planned elimination of large numbers of personnel to enhance organizational competitiveness Three common reasons for downsizing: Reduce labor costs Technological changes reduce need for labor Organizations change location for economic reasons Cutting hours or pay Can sometimes avoid layoffs if can get employees to take a cut in hours/pay "Furloughs" - cut in hours for salaried workers

Biodata examples

During high school, how many times did you make the honor roll? How much freedom or independence did your parents allow you in grade school? How important did your favorite high school teachers stress discipline in the classroom? How many times did you change schools before you were sixteen years old? Compared to other people in high school, how many friends did you have? How old were you when you spent your first week (or more) away from your parents? How bothered are you if you a job is left undone? How often do you read craft and mechanics magazines? In most ways, is your life close to ideal? How well do you feel you can understand the feelings of others? How well do you tolerate performing routine tasks?

Expected labor surplus early retirement

Early retirement programs The average age of U.S. workforce is increasing Baby boomers are not retiring early due to: Improved health Fear that Social Security will be cut Mandatory retirement is outlawed Collapse of the financial and housing markets made it economically unviable to retire Many employers try voluntary attrition among older workers through early retirement incentive programs

Expected labor shortage

Employing temporary workers and independent contractors Offers operational flexibility Frees the firm from many administrative tasks and financial burdens (health insurance, pension, worker's compensation, life insurance, etc.) Outsourcing - using an outside organization for specific services Jobs that are proprietary or require tight security should not be outsourced Offshoring - a special case of outsourcing, where jobs leave one country and go to another Get more hours out of the existing workforce Some companies may be reluctant to hire new full- or part-time employees, but instead give more hours to current employees

factors affecting the validity of personality tests

Faking Knowledge/image of self Work context/situations De-contextualized test items Situation strength Trait relevance Broad/narrow trait measures Non-linear relationships with criteria

selection decisions

False positive Applicant accepted but performed poorly False negative Applicant rejected but would have performed well True positive Applicant accepted and performed well True negative Applicant rejected and would have performed poorly

other research on sources

Few meta-analyses, but see other reviews by Breaugh (2000; 2008) Recruiting through informal mechanisms (e.g., referrals) resulted in: Lower turnover Higher levels of job satisfaction Higher quality (were more likely to be offered a job) Formal ads and employment agencies are least effective sources Rynes et al. (2002) Only ½ HR managers surveyed knew that job ads produce higher turnover than employee referrals

workforce planning and forecasting

Forecasting - attempts to determine the supply and demand for various types of human resources Predict future labor shortages or surpluses Aligned with business strategy Predicting the skills the employer will need to execute its strategy

generalizability

Generalizability - degree to which the validity of a selection method established in one context extends to other contexts a.k.a., "validity generalization" e.g., different companies, samples of people, time periods, etc. Meta-analyses very useful for establishing validity generalization Many employers won't have the resources or expertise to conduct a validation study Organizations must identify tests/screening tools that have been shown to be valid in other settings (e.g., companies), in hopes that they'll be valid in their own organization

Methods of estimating reliability

Generally estimated by examining the relationship (i.e., correlation) between two sets of measures that measure the same thing among the same people Test-retest reliability - give same test twice to same people Equivalent forms reliability - two forms of same test given to same people Internal consistency reliability - intercorrelation between test items e.g., statistics - Cronbach's coefficient alpha, split-half reliability, Kuder-Richardson (KR-20) Inter-rater reliability The magnitude of similarity in scores is usually expressed as a reliability coefficient (rxx, ryy, etc.) Ranges from 0.0 to 1.0 How high is acceptable? The more critical the decision to be made, the greater the need for precision and the higher the reliability required

Faking personality tests research

Griffith et al. (2007) - repeated measures design Administered a Conscientiousness scale to job applicants and again, posthire, in an honest condition ensuring confidentiality Mean shift from faking (d = .61) Applicant-honest correlation was only .50, a 44% drop from the .89 correlation observed for the same measure with a similar time interval between two honest condition assessments Hough et al. (1990) - participants told to purposely fake responses to make themselves look good Predictive validity remained the same Hough (1998) - meta-analysis (k>700) Predictive studies (using applicants) yielded mean validity estimates about half as large as those from concurrent studies (using incumbents)

Biodata-Do they work?

How reliable and valid are biodata? Reliabilities range from .60 to .90 Validities range from .20 to .37 Potential issues Need large sample to construct properly Pure empirical approach (e.g., versus content approach) Legal issues (e.g. adverse impact, validity, reliability) Lacks face validity

enhancing application accuracy

Inform applicants (verbally and in writing), that information will affect their employability Inform applicants that data will be thoroughly checked Require applicants to sign a statement certifying the accuracy of the information they provided on the form Include warnings of penalties (not being hired or termination upon discovery) for deliberate falsification Include a statement that the application does not create a binding obligation of employment for any specific period of time

improving recruitment sources

Internal Job Postings Emphasize detail and specificity (e.g., tasks and KSAs) HRIS - update regularly and standardize data so people are comparable Rehiring Fair and cordial treatment during exit Maintain and support relationships (e.g., "alumni" networks) Employee Referrals Reward referrals (i.e., $$ or other benefit) Make it "visible" in the organization Track where best referrals originate Set "referral goals"

recruitment sources

Internal sources Human Resources Information Systems (HRIS) Job Postings (listing of vacancies within an organization) External sources Referrals Advertisements (e.g., newspapers and magazines) Public and private employment agencies Colleges and universities Electronic recruiting (e.g., CareerBuilder, LinkedIn, etc.) Temp agencies and employee leasing firms Executive recruiters ("headhunters")

cognitive ability tests studies

Is "g" important at work? Schmidt & Hunter (2004) - meta-analysis "Yes", g predicts performance across many jobs ↑ job complexity = ↑ predictive value of tests of g Pros: 85 years of research shown g to be one of the best predictors of future performance and training (r = .51) Generalizable Low costs and high utility Cons: Adverse impact

examples of work sample tests

Job-- Civil Engineer Salesperson Clerical Employee Customer Svc. Rep Programmer Professor test-- CAD drawing Sales call Typing test Customer role-play Debugging test Teaching a class while on campus interview

beware certain personality test

MBTI=most widely used but sucks because of no predictive validity and retest changes result common DISC=On the surface level, it appears that the instrument is useful in personal or professional development. However, the test suffers from questionable reliability and unknown validity. The measure is also remiss in its lack of reliability and validity for diverse ethnic/racial populations and professional occupations. Therefore, the use of the DiSC is not recommended."

personality tests

Managers intuitively believe personality traits matter at work There are some connections between aspects of personality and various work behaviors Big Five Personality Model a.k.a., Five Factor Model (FFM) Taxonomy of five personality factors Result of statistical and conceptual analysis of decades of personality test data

personality tests- taking by self v other

Oh, Wang, and Mount (2011) Used "observer" reports (e.g., subordinates, supervisors, customers, friends, etc.) instead of self-report Stronger predictive validity (e.g., ρ=.32 for conscientiousness) Increases with additional observers

the big five

Openness - imaginative, curious, willing to experiment, creative Conscientiousness - dependable, organized, reliable, responsible, persevering Extraversion - sociable, ambitious, assertive, talkative Agreeableness - good-natured, tolerant, courteous, cooperative Neuroticism (emotional stability) - secure, not easily upset, calm, sedate

The Case for Systematic Selection

Org. Effectiveness Good business sense because poor selection decisions lead to: Increased turnover Increased absenteeism Lower productivity Lower commitment Legal Compliance Good business sense because poor selection decisions lead to: Increased turnover Increased absenteeism Lower productivity Lower commitment Mgr. Effectiveness Reducing human bias Most time is spent trying to confirm our first impressions Increasing precision Need for more data Remove "guess work" Improve predictions

Legal implication of applications

Organizations routinely request information on applications that are "inadvisable" 1980's research: All application forms reviewed had at least 1 inadvisable question, average = 7 to 9 Fortune 500: 38% > 10 inadvisable items Most frequently used inadvisable items: Have you ever been arrested for felony? What dates did you attend grammar school? High School?

personality tests-legal issues

Personality tests have not been shown to have disparate impact among demographic groups Personality tests may create problems under the ADA if considered a pre-employment medical examinations (e.g., used to assess depression, paranoia, etc.) If using a multi-hurdle selection program, it is recommended to use the personality test first, and then the cognitive ability test, and then the interview, etc.

Validity - Criterion-Related

Predictor Test purchased or designed to assess identified KSAOs Criterion Outcome variable measuring important aspects of the job (performance) Correlation between our test (the predictor) and some future behavior (the criterion) rxy = the criterion validity coefficient We wish rxy = 1.0, which will mean...... In practice rxy = .30 - .50 Rarely above .60 (Cronbach, 1990)

situational judgment tests

Present candidate with written scenario, then ask candidate to choose best response from series of alternatives A "low-fidelity" simulation Different scoring rubrics Can be written or video based Pros: Job-related; valid predictor of task performance (.21 - .50), contextual performance (.21 - .35), and managerial performance (.29 - .36) Well-accepted by test takers Reduced adverse impact compared to other devices Cons: Scenarios are brief, therefore can only be but so realistic Responses options may not cover all possibilities

Applications

Principles for effective use: Good for "weeding out" applicants without minimum requirements Know that information is not always accurate Law assumes any and all information on applications is used for selection Keep demographics separate

selection

Process whereby decisions are made about who will or will not be allowed to join the org.

personality assessments-summary

Pros: Predictive validity (some) Better with work contextualized items and if matched to a specific job (i.e., trait relevant) No/less adverse impact Cons Lower predictive validity than cog. ability tests Social desirability and faking

assessment centers

Provide simulations in which candidates perform realistic management tasks, under the observation of experts who appraise each candidate's potential Common simulated exercises In-basket - candidate must take action on memos, reports, emails, phone calls, etc. Leaderless group discussion - group given a task, candidates observed and evaluated in terms of interpersonal skills, leadership potential, etc. Individual presentations - candidate gives an oral presentation on assigned topic, good for evaluating communication skills and persuasiveness Role-plays Pros: Realistic, job related, well accepted, valid predictor (.36) Cons: Very time consuming and expensive to both develop and administer

Decision Accuracy

Quality of decisions change as validity changes Better decisions (more true positive and true negatives) with higher validity

Recruiting Job candidates

Recruitment - any practice or activity carried on by the organization with the primary purpose of identifying and attracting potential employees Intended to increase pool of qualified applicants (decrease pool of unqualified applicants) More complex than many people think Should be aligned with strategic plans Some methods better than others Legal issues e.g., if already a non-diverse workforce, then EEOC may view word-of-mouth referrals as barrier to EEO

forecasting methods continued reg and HRIS

Regression analysis - statistical technique that identifies the relationship between a series of predictor variables and an outcome variable e.g., relationship between multiple, past employment characteristics and revenues or productivity Enables managers to build more complex statistical models in making personnel projections Human Resource Information System (HRIS) - provides a database of employee qualifications and other employment characteristics ("talent inventory")

Criteria for effective predictors

Reliability - degree to which a selection method is free from random error (shows consistency) Validity - accuracy of inferences made based on a selection method Generalizability - degree to which the validity of a selection method established in one context extends to other contexts Utility - degree of practical and economic value provided by a selection method Legality - degree to which a selection method conforms to existing laws and legal precedents

improving the interview and the interviewer

Restrict the scope of the interview primarily to selection (not recruiting) Focus on the most job-relevant characteristics (preferably just 2-3 KSAOs) Limit the use of pre-interview data about applicants Adopt a structured format Use multiple questions for each KSAO Rely on multiple independent interviewers Use a scoring rubric (developed a priori) Train interviewers

Resumes

Resume screening - incredibly important activity Major problems Resumes are self-reported, unstandardized applications Suffer same distortions as applications Resume readers (recruiters) infer KSAs and screen using idiosyncratic methods Moderate interrater reliabilities are reported (.60s to .70s) Little evidence regarding predictive validity Advice - choose small # of KSAs and assign point value

discrimination in interviews

Review of 158 federal cases Unstructured interviews involved in 57% of the cases, structured only 6% None of the 6% found "discrimination"; 41% ruled as "discrimination" for unstructured Huffcutt & Roth (1998): Meta-analysis, 31 studies of racial group differences in interviews Overall differences - d = .25 High structure, d = .23; Low structure, d = .32 Legal interviews are...job related

Effect of moving the cutoff score

Specifically, assuming you have a valid test, raising the cutoff score will result in fewer false positives, but the price you pay for that is more false negatives. Conversely, lowering the cutoff score will result in fewer false negatives, but the price you pay then is more false positives.

Interpreting Validity Coefficients

Squaring a correlation (r), provides the variance in the criterion accounted for by the predictor Caveat: 100% of the variance will never be explained by a single predictor...human behavior is too complex Interpretative guidelines for r and r2

correlation

Statistic or measure of association ("r") Index of the relationship between the 2 variables 2 key features in interpreting the "r": - direction of relationship (+ or -) - strength of relationship (-1 to +1) Does NOT necessarily indicate a cause-effect relationship

how to validate a test

Step 1 - Analyze the job and write job descriptions and job specifications. Step 2 - Choose the tests that measure the attributes (predictors) important for job success. Step 3 - Administer the test selected to old or new employees for concurrent and/or predictive validation. Step 4 - Relate test scores and criteria through a correlation or regression analysis, which shows the degree of statistical relationship between (1) scores on the test and (2) job performance. Step 5 - Cross-validate and revalidate by performing Steps 3 and 4 again on a new sample of employees.

examples of cognitive ability tests

Tests of general mental ability (in the workplace) e.g., Wonderlic Personnel Test (WPT) - 50 items to assess verbal, numerical, and spatial abilities; 12 minute administration High reliability, strong correlations with more elaborate intelligence tests Used by the NFL - Do you think g matters for performance in football? Lyons, Hoffman, & Michel (2009) - WPT scores unrelated to future NFL performance Tests of specific abilities e.g., Bennett Test of Mechanical Comprehension Other tests measure perceptual accuracy, memory, spatial relations, etc. Cognitive test batteries Assess a variety of cognitive aptitudes or abilities e.g., Armed Services Vocational Aptitude Battery (ASVAB)

reliability

The degree of dependability, consistency, or stability of scores on a measure used in staffing decisions (rxx) The degree to which a measurement instrument is free from random error Greater reliability leads to greater precision

Validity

The degree to which available evidence supports inferences made from scores on selection measures Indicator of whether a test accurately measures what it intends to measure, or accurately predicts what it intends to predict Test might be valid for one purpose, but not for another Can't say, "this is a valid test" - must be valid for its intended purpose High reliability without validity is possible, high validity without reliability is not possible Validity evidence Content validity - degree to which the test content/items represent an adequate sample of important work behaviors and activities or worker KSAOs defined by job analysis Construct validity - degree to which an actual measure (e.g., test of intelligence) is an accurate representation of its underlying psychological construct (e.g., intelligence) Criterion-related validity - degree to which test scores accurately predict something else (e.g., job performance)

big five model-self report measures

There are many self-report questionnaires available NEO-PI-R, 16PF, CPI, Jackson's PRF, Hogan's HPI Well-developed self-report measures of personality, when applied appropriately (i.e., when based on a job analysis), can be useful for selection decisions Barrick and Mount (1991) - meta-analysis; conscientiousness was positively related to success in all aspects of work for all occupations But... criticized because not highly correlated (ρ=.22) with work behaviors

What about the handshake?

There is a commonly held belief that a firm handshake bears a critical nonverbal influence on impressions formed in the employment interview But does your handshake impact hiring decisions? (Yes or No) Is there any evidence? Stewart et al., (2008) 98 undergrads put through realistic job interviews Rated on 5-point scales for grip strength, completeness, duration, and vigor (also rated degree of eye contact) Independently, real HR professionals shook hands with and interviewed the participants, but were NOT asked to rate or explicitly evaluate the handshake Quality of handshake was positively related to hiring recommendations (.29) Women benefited more than men from a strong handshake

cohen's d

To help you interpret the above, Cohen's d reflects the standardized mean differences across groups Cohen's d: "small" = .25, "medium" = .5, "large" = .80 Compared to cognitive ability (+/- 1 SD), interviews create much less adverse impact

Forecasting methods

Trend analysis - studying a firm's historical trends over a period of time to predict future needs e.g., compute number of employees each year over previous 5 years, to help identify impending needs Ratio analysis - making forecasts based on the ratio between (1) some business factor, like sales volume, and (2) number of employees required, such as the number of salespeople e.g., if you want to generate additional $3 million in sales, and typical salesperson generates $500,000 in sales, then would need 6 new salespeople (6 x $500K)

What about now in legal implications of applications

What about now? 2004 study of 109 customer service applications Average = 5.35 inadvisable items 2005 review of 300 federal court cases involving use of applications Questions associated with sex and age most likely to lead to litigation (53%) Plaintiff won > 40% of these cases

Types of interviews and the reliability of them

Which approach is more reliable? Conway et al. (1995) Unstructured reliability = .37 Structured reliability = .66 Remember, reliability creates an upper limit on predictive validity

validity of interviews

Which approach is more valid in predicting job performance? Unstructured interview Corrected validity coefficients from .20 - mid .30s lower validity Structured interview Corrected validity coefficients from .40s - .60s higher validity

research on recruitment sources

Zottoli and Wanous (2000) reviewed 28 studies "Effective" = most likely to yield long-term employees Most effective: In-house job postings Rehired workers Referrals by current employees Least effective Newspaper ads Employment agencies School placement services Results suggested companies can increase their job survival rate by 25% or more by using "effective" sources...

interpretive guidelines for r and r2

correlation r variance explained r2 interpretation 0.10 0.01 small effect 0.3 0.09 medium 0.5 0.25 large effect

what about brainteaserS?

google hr=waste of time

concurrent validation

measure all current job incumbents on attribute + combined with measure all current job incumbents performance too== obtain correlation between these two sets of numbers

predictive validation

measure all job applicants on attribute and then hire some applicants and reject others and wait for some time period and then measure all newly hired job incumbents performance to ====then obtain correlation between these two sets of numbers

physical ability tests

only valid for certain jobs Validity coefficients reported by Fleishman: Electrical workers, r = .53 Pipeline workers, r = .63 Correctional officers, r = .64 Army (enlisted men), r = .87 Warehouse workers, r = .39 ` Used to measure ability for physically demanding jobs Tied directly to physical-related KSAs Since we were talking about the NFL... Do tests of physical abilities from the NFL combine (i.e.,40-yrd dash, 20-yrd shuttle run, 3-cone drill, vertical jump) predict future performance? Lyons et al. (2011) - Very weak relationships between these physical tests and performance Pros: Can improve prediction of job success for physically demanding jobs e.g., firefighter - stairway climb, hose pull, ladder pull, etc. Reduces the incidents of work-related injuries Cons: Increased adverse impact for women and those with physical disabilities

Selection Ratio

ratio of the number of open positions to the number of job applicants Smaller ratios are better from the company's standpoint, as smaller ratios (closer to zero) indicate that there are many applicants for any one position (i.e., lots of applicants to choose from, to hopefully identify the high performers When the selection ratio is close to one (or worse), most applicants will need to be hired in order to fill the available positions. And therefore, knowing that one applicant is likely to perform better than another is of limited value (i.e., because you're going to have to hire them all anyway).

situation vs behavioral interviews

situational Applicant is given a hypothetical situation and asked how he/she would respond to it Each question links to one knowledge or skill Most effective when based on work analysis Work well because intentions predict future behavior Reliability .94; validity .43 to .45 behavioral Applicant is asked to explain a previous time when he/she took a certain action Each question links to one knowledge or skill Most effective when based on work analysis Work well because one of the best predictors of future behavior is past behavior Reliability .97; validity .51 to .56