Chapter 5- Employee Selection; Reference and Testing

Past performance- best predictor of future performance

● Held jobs that were similar to one for which he is now applying for

Quality of previous performance- difficult to verify

● Verifying previous employment is not difficult

Employer must obtain info about the quality of previous performance by:
1. Relying on applicants references- calling them or asking for recomm. letters

● Reference check- confirming the info

● Reference- expression of an opinion (orally or written) through a written checklist
○ Content and format of a reference are determined by the person or org.
● Letter of recommendation- letter expressing an opinion regarding an applicant’s ability, previous perf. ,
work habits, character, future potential success.
○ Content and format is determined by letter writer

Reasons for using references and recommendations

1. Confirming details on a resume- check references and ask for letters of recomm.
a. Resume Fraud- lying/ faking
2. Checking discipline probs
a. Negligent Hiring- if an org hires an applicant without checking his references & bg. - they can
be liable for NH
- Employer had not taken “reasonable care” in ensuring the well-being of its
customers.
- In determining NH, look at the nature of the job
3. Discovering new info about the applicant
- Former employers and professors can provide info about the applicant’s work habits,
char, pers, skills.
- Opinion provided by any particular reference may be inaccurate or purposefully
untrue.

4. Predicting future performance

- References have not been successful in predicting future employee success
- Average uncorrected validity coefficient for references/letters of recommendation and
performance is only .18 with corrected validity of .29

4 main probs w/ references and LOR

LENIENCY
- Applicants choose their own references
- They could provide favorable references

- Confidentiality concerns
● Coworkers are willing to say negative things about unsatisfactory employees, which can hold
them back
● When a person is writing a reference letter knows that the applicant is allowed to see the letter,
the writer is more inclined to provide a favourable evaluation.

- Fear of Legal Ramifications

● A person providing references can be charged with defamation of character, if the content of
the reference is both untrue and made with malicious intent

- Conditional Privilege
● They have the right to express their opinion provided they believe what they say is true and
have reasonable grounds for this belief
 - One way to avoid losing a defamation suit
● Provide only behavioural information in a reference

- Reference detectives
● Contact the former employer under the guise of being a company considering hiring the former
employee

- Negligent reference
● A former employer also can be guilty of this, if it does not provide relevant information to an
organisation that requests it
● A former employer can be charged with slander or libel if it says something bad about an
applicant that cannot be proven. EMployer can be held liable if it does not provide information
about a potentially dangerous applicant.

KNOWLEDGE OF THE APPLICANT

- Professors
● Does not know the applicant well, has not observed all aspects of an applicants behaviour, or
both

- Supervisor
● Does not see all aspects of an employee’s behaviour

RELIABILITY
● Lack of agreement
- Between two people who provide references for the same person

- There is more agreement between recommendations written by the same person for two
different applicants than between two people writing recommendations for the same person.
Letters of recommendation may say more about the person writing the letter than about the
person for whom it is being written.

● Low level of reliability

- Writer has not seen all aspects of an applicant's behaviour

EXTRANEOUS FACTORS
- Letters written by references who like applicants are longer than those written by references
who do not. The longer the recommendation letter, the more positively the letter was perceived.

● Approach to increasing the validity of references

- To increase the structure of the reference check.
- Can be done by conducting a job analysis. Creating a reference checklist that is tied directly to
the job analysis results.
- References and letters of recommendation often are not great predictors of performance.

3 ETHICAL ISSUES (to avoid..reference providers should follow:)

● Explicitly state your relationship

● Be honest
- A referee has both an ethical and a legal obligation to provide relevant information about an
applicant

● Let the applicant see your reference

- Give him the chance to decline to use it. Should be fair to applicants and reduces the referee’s
liability for any defamation charge.

Predicting Perf using Aplcnt Knowledge

 ● Better educated employees had higher performance, were more likely to engage in organisational
citizenship behaviours, less likely to be absent, and less likely to engage on-the-job substance abuse
● Students GPA can predict job performance
● GPA is most predictive in the first few years after graduation
● Factor Complicating the use of education
- Increased use of homeschooling

Performance Using Applicant Knowledge

● Job knowledge tests

- how much a person knows about a job
- Good predictors of both training perf and on-the-job perf.
- High face validity- + accepted by applicants
- Results in adverse impact and can be used only for jobs in which applicants are
expected to have job knowledge at the time of hire or promotion

Performance Using Applicant Ability

● Extent to which an applicant can learn or perform a job-related skill.

● Applicants are not expected to know how to perform the job at the time of hire
● New employees instead, will be taught the necessary job skills and knowledge

● Cognitive ability
- Oral and written comprehension, oral and written expression, numerical facility, originality,
memorisation, reasoning.
- Important for professional, supervisor, accountant and secretary

● Cognitive ability tests

- Excellent predictors of employee performance
● Predict work performance in two ways
- Allowing employees to quickly learn job-related knowledge
- Processing information resulting in better decision making.

● Draw backs
- High levels of adverse impact, often lack face validity, frequently challenge in court.
- Difficulty of setting a passing score.

- CA is significantly related to performance

● Corrected validity
- Not nearly as high as that reported in the meta-analysis

● Wonderlic Personnel Test

- Short amount of time, administered in a group setting.

● Siena Reasoning Test

- Large race differences in scores on traditional cognitive ability tests were due to
knowledge needed to understand the questions
- Contain nonsense words and words that are commonly known
- Predict college grades and work performance
- Eliminated racial differences in test scores.

● Situational judgement test

- Given a series of situations and asked how they would handle each one
 ● Perceptual ability
- Vision
- Hearing

● Psychomotor Ability
- Dexterity
- Coordination

● Physical ability
- Jobs that require physical strength

● Measured in two ways

- Job simulations
- Physical agility tests

● Job stimulation
- Demonstrates job-related physical behaviours
- Highly content valid
- Often impractical
- 2 categories
● Athletic requirements
- Easy to simulate

● Defensive requirements
- Difficult to safely accurately simulate
- Instead of simulating defensive behaviours, tests are developed that
measure the basic abilities needed to perform these behaviours

● Criticised on three major points

- Job relatedness
- Passing scores
- Time at which they should be required.

● Critics argue that physical agility is not an essential part of the job. This is especially true due to
technological advances in policing

Passing Scores:

● Passing Scores for physical ability tests are set based on one of two types of standards
- Relative
- Absolute

● Relative Standards
- How well an individual scores compared with others in a group.

● Advantage to using relative standards

- That adverse impact is eliminated
● Absolute passing
- Scores are set at the minimum level needed to perform a job

When the ability must be present:

● The applicant does not need the strength or speed until he is actually in the academy or on the job.
● Applicants going through an academy shows significant increases in physical ability and fitness by the
end of the academy

Work Samples
 - applicant performs the actual job related task

Excellent selection tools bc:

1. Directly related to job
- excellent content validity

2. Scores from WS predict actual work perf.

- excellent criterion validity

3. Connection between WS & work performed on job

- excellent face validity
- Less challenged in civil service appeals

4. Lower racial differences

● reason for not using WS

- Expensive, used only by well-paying jobs for which many employees will be hired

Assessment center
- Oberse applicants perform simulated job tasks

● Major advantages
- Job related and multiple trained assessors help to guard against many types of
selection bias.

Requirements:

● Thorough job analysis

● Behaviours must be classified into meaningful and relevant categories
● Multiple assessment techniques must be used
● Must be a job simulation
● Multiple trained assessors
● Must be documented
● Report of their observations
● Overall judgement of an applicant must be based on a combination of information
from the multiple assessors and multiple techniques

Development and Components

● Basic development and types of exercises and fairly standard.

● First step in creating an assessment center

- Job analysis
- Exercises are developed
- Each of these techniques can be used by itself
- Four or five exercises, takes two or three days to complete

● In-Basket technique
- Simulate the types of daily information that appear on a manager’s or employee’s desk

● Two levels
- ‘In’ level holds paperwork that must be handled
- ‘Out’ level contains completed paperwork

● Respond to the paperwork as if he were actually on the job.

● Reliability and validity of the in-basket technique

 - Provides modest support for its usefulness

● Simulation exercises
- Real backbone of the assessment center
- Role-plays and work samples
- Similar as possible to one that will be encountered on the job.

● To be effective
- Must be based on job-related behaviours and should be reasonably realistic.

● Work sample
- Simulation does not involve a situational exercise

● Leaderless Group discussions

- Small groups and are given a job-related problem to solve or a job-related issue to discuss

● Business games
- Allow the applicant to demonstrate such attributes as creativity, decision making, and ability to
work with others.

● Evaluation of Assessment Centers

- Successful in predicting a wide range of employee behaviour
- Has been argued that other methods can predict the same criteria better and less expensively
than assessment centers

● Applicant experience is typically measured in one of four ways:

- Experience ratings
- Biodata
- Reference checks
- Interviews

1. Biodata
- An applicants life, school, military, community, and work experience
- Good predictor of job performance
- Best predictor of future employee
- Measure the difference between successful and unsuccessful performers on a job
Advantages
● Research can predict work behaviour
● Biodata instruments result in higher organizational profit and growth
● Easy to use, quickly administered, inexpensive, not as subject to individual bias as
interviews.

Development of a Biodata Instrument:

● File approach
- Information from personnel files on employees’ previous employment,
education, interests, and demographics.
Major disadvantage
- Information is often missing or incomplete

● Major drawback to questionnaire approach.

- Cannot be obtained from employees who have quit or been fired.

● Criterion
- Is a measure of work behaviour such as quantity, absenteeism, or tenure.
- Be relevant, reliable, and fairly objective.
 - Using biodata can predict a non-relevant criterion would not have saved the organization any
money.

● Criterion Groups
- Determine which pieces of information will distinguish the members of the high criterion group
from those in the low criterion group.

● Vertical percentage method.

- Items that make sense are more face valid and thus easier to defend in court than items that
are empirically valid but does not make rational sense.
- Information is weighted and then summed to form a composite score for each employee.
- Correlated with the criterion to determine whether the newly created biodata instrument will
significantly predict the criterion.
- Fairly easy, although time consuming

● Sample Size
- To create a reliable and valid bioset instrument, desirable to have data from hundreds of
employees

● Derivation sample
- Is used to form the weights

● Hold-out Sample
- Is used to double-check the selected items and weights. It may not be practical when dealing
with a small or moderate sample size.
- Less chance of error when a sample is not split.

Criticisms of Biodata

● Validity of biodata may not be stable, ability to predict employee behaviour decreases with time.

● Some biodata items may not meet the legal requirements stated in the federal Uniform Guidelines,
which establish fair hiring methods.

● The item may result in adverse impact if the organisation is located in a predominantly White area.

● To make biodata instruments less disagreeable to critics:

- Item must deal with events under a person’s control
- Item must be job-related
- Answer to item must be verifiable
- Item must not invade an applicant’s privacy

● Biodata can be faked

- Applicants do in fact respond to items in socially desirable ways

● To reduce faking
- Warning applicants of the presence of a lie scale
- Using objective, verifiable items
- Asking applicants to elaborate their answers, or provide examples

● Bogus Items
- Include an experience that does not actually exist
- Items can be carefully researched

● Bright applicants tend not to fake biodata items as often as applicants lower in cognitive ability
Predicting Performance Using Personality, Interest, and Character

● Personality Inventories

● Tests of normal personality

- Measure the traits exhibited by normal individuals in every-day life.

● Theory-based
Myers-Briggs Type Indicator, theory of Carl Jung

● Statistically based test

16PF (Personality Factors), by Raymond Cattell

● Empirically based
- Grouping answers by people known to possess a certain characteristic.

Minnesota multiphasic personality inventory-2 (MMPI2)

● Items that were endorsed more often by paranoid patients than healthy individuals were keyed under
the para-noia dimension of the MMPI-2

Subfactors for each of the Big Five Factors

- Better predictors of behavior

Objective Pers Inventories are useful in predicting perf.

- Personality can predict performance at low but statistically significant levels
- Can add incremental validity to the use of other selection tests;
- Conscientiousness is the best predictor
- Validity of other four personality dimensions is dependent on the type of job and criterion for which the
test is being validated.

because PI are self-reports, they are relatively easy to fake.

● When they do fake, it has minimal effect on the validity of the rest results.
● People often act differently at work than they do in other contexts.
● Ask the applicant about their work personalities rather than their personality in general

Tests of Psychopathology

● Determine whether individuals have serious or psychological problems

● Seldom used by I/O psychologists expect in the selection law enforcement officers.

● Projective tests
- Questionable of reliability and validity
- Time consuming, expensive, rarely used

● Rorschach Inkblot test

● Thematic Apperception Test

● Objective tests
- Respondent is limited to a few answers

● Interest inventories
● Strong Interest Inventory

● Individual with interests similar to those of people in particular field will more likely be satisfied
in that field.

● Vocational Counselling
- Helping people find the careers for which they are best suited
- Include an interest inventory and a series of ability tests

● Integrity Tests (also called honesty tests)

- Tell an employer the probability that an applicant would steal money or merchandise

● Polygraph
● Voice Stress Analyzer

● Overt Integrity Tests

● Personality-based integrity tests
- Overt tests are more reliable and valid in predicting theft and other counterproductive
behaviours than personality-based tests

- Integrity tests predict a variety of employee behaviours including work performance

- Tend to result in low levels of adverse impact against minorities

● Individual employee behaviour

● If polygraph results are used, the researcher is essentially comparing integrity scores with the test
● Questions about its accuracy.

- If self-admissions are used, the researcher is relying on dishonest people to be honest about their
criminal history

● Shrinkage
- If used, the researcher does not know which of the employees is responsible for theft

● Personal tendency
● Economic Pressure

● Reaction to organisational policy

- To reduce theft caused by situational factors, non testing methods such as increased security,
explicit policy, and availability of appeal and suggestion systems are needed.
- Paper-and-pencil integrity tests, have serious drawbacks

● Conditional Reasoning Tests

- Applicants may not provide accurate responses
- Applicant faking responses
- Not actually being aware of his or her own personality

● Provide test takers with a series of statements and then ask the respondent to select the
reason that best justifies or explains each of the statements.
● Indicates his or her aggressive biases or beliefs

- Harmful intentions behind their behaviour (hostile attribution bias)

- Strength or dominance in social interactions (potency bias)
 - Retaliable when wronged rather than try to maintain a relationship (retribution bias)
- Victimise less powerful individuals (victimisation bias)
- Evil people deserve to have bad things happen to them (derogation of target bias)
- Social customs restrict free will and should be ignored (social discounting bias)

● Counter-productive behaviour is best predicted when conditional reasoning tests are combined with self-
reports of aggressive tendencies

Credit history
Credit checks are conducted for 2 reasons
- People who owe money is likely to steal
- Employees w/ good credit are more responsible and conscientious, thus are better emp.

Graphology
- Handwriting reveals personality
- Graphologists are consistent w their judgments but not in their interpretations
- They are no more accurate and reliable at interpreting handwritings
- Not a useful technique in emp. Selection
- Predicts best when the writing sample is autobiographical
- Prediction is on content than quality of handwriting

Predicting Perf limitations due to medical and psychological probs.

Drug testing
- Illegal users are more likely to miss work
- Appear to reduce drug use
Done in 2 stages:
● Instant screener
● Medical review officer at a testing lab to ensure results were accurate
- DT are not able to determine whether an indiv. Is impaired by drug use

Psych exams
- If applicant fails the exam, the offer is rescinded
Consists of:
- Interview
- Examination of history
- Administration of tests

Medical Exams
- Physician is given a copy of the job desc. And asked to determine if there are any physical condition that
will keep the emp. From safely performing the job

Comparison techniques
Validity
- Unstructured interviews, educ, interest inventories, some pers. Tests are not good predictors of
future emp. perf.
- Ability, work samples, biodata, structured int.,- fairly good at predicting

- Most valid selection battery includes a cognitive ability test and either a work sample, an integrity test or
a structured int.
- All are potentially useful methods for selecting employees

Legal Issues
- In terms of face validity. Int., work samples/ simulations and resumes are the most job-related fair
- graphology , Integrity tests and pers tests are least job-rel./fair
Rejecting Applicants
- Those who are not hired must be notified
- Should be treated well bc they are potential customers or applicants of others

Rejection Letter
- Personally addressed and signed letter
- Company’s appreciation
- Compliment
- Comment about high qualifications of person
- info about the indiv who was hired
- Good luck
- A promise to keep applicants resume on file

Guidelines
- Rejection letters via emails or letters. Failure to send a letter or email results in applicants
feeling - towards an organization
- Don’t sent RL immediately. Applicants react more + if there is a delay
- Be personable and specific
- Include statement about indiv who received the job
- Do not include contact person

● Most imp. thing to consider when writing a LOR is to be HONEST

SUMMARY: