RSW Lectures
Lecture 1 – Introduction and Job analysis
-chapter 1,3 & 7
-Binning validity of personnel decisions
*Binning & Barrett model
How to measure someone’s success potential (relevant characteristics) > and predict their
success at the job (how can we see if a person is successful after being hired)
Snapshot of potential (predictor) > capture how they perform (criterion)
Construct domain Performance domain
-Conceived by researchers with reference Determines/influenced by organizational
to some theoretical framework decision makers & selection specialists
-aim is to explain general regularities in -aim is to translate goals into normative
human behavior/ predictors of certain statements of valued behaviors &
behavior outcomes
-for example extraversion -for example negotiating
So performance domain (coming up with innovative ideas) > measure of performance
(supervisor report) > which construct domain are related/predictors to innovative ideas
(openness to experience or divergent thinking) > measure of construct (brain storming task)
, Show that there is an
empirical relationship
between your predictors and
criterion (behavior) – so the
predictors that you choice
predict how well people
perform (you want to include
as many predictors as
possible) Yellow line
*Binning & Barrett
-validation = the process of gathering
evidence or evaluating the necessary
data
-different kinds of evidence in assessing validity
-degree to which the evidence supports the inferences (expectations) made from the scores
or outcomes of the test/procedure (are you right to infer this and what evidence is needed
to support this inference)
Job Analysis
1. Define the job & specify what employee behavior is necessary to complete it/what
behaviors are needed to perform (job description)
2. Then identify which characteristics underlie this behavior/ which psychological
constructs underlie performance (job specifications) KSASOs = knowledge, skills,
abilities and or other characteristics
-should indicate the minimal requirements/ qualifications
-lies at the heart of almost every HRM program or activity > need accurate information
about a job
- goal is to define each job in terms of behaviors necessary to perform the job and develop
hypotheses about personal characteristics necessary to perform it
-important for; job descriptions and specifications, personnel selection, performance
appraisal, training & development
Skills = practiced acts. They can learn over time to the limits of one’s natural ability
Ability = individuals may be limited in their capacity to change depending on their ability
(certain natural range)
How to establish job specifications
-SMEs 1 (subject matter experts) develop list of tasks & KSAs
-SMEs 2 judge the necessity of tasks and KSAs (should it be a pre or can they be learnt the
skill)
-SMEs 3 judge clarity & adequacy of level, finalize
To what extent is the profile indicated suitable (0-3) so finalize the list
The results of job analysis affect all other aspects of personnel selection. You need to define
the job first before you can even begin to find a person to match that job
,Choices awaiting the job analyst
1. Activities (descriptions) vs Attributions (specification)
-do you want to do a job analyses for training? Then focus on attributions
2. General vs specific levels of detail (flexibility)
3. Qualitative data (narrative/ career planning) vs quantitative data (numeric fixed
scales, time, frequency, importance, criticality)
4. Taxonomy-based (use already existing questionnaires to gather information) vs
blank slate (start with a blank slate and see what data shows in observation) >
looking for general attributes that fit the job)
5. Get information from Observer (trained job analyst) vs incumbents (people on the
job) & supervisor (good to get information from all to get the total picture)
6. KSAs vs KSAOs (O = other personal characteristics like personality traits, values and
attitudes)
7. Single job vs multiple-job comparisons (justify use of same/different selection
system)
8. Descriptive vs prescriptive (does the job already exist or not and do you have to
specify what the job entails)
Reliability and Validity of job analysis
Job description – do they accurately represent job content, environment and conditions of
employment?
-the more systematic the data gathering is the more valid the output
-make sure the sample is big enough
Job specification – do people possessing the personal characteristics believed necessary for
successful job performance in fact do perform more effectively on the job than people
lacking such characteristics
Factors influencing the accuracy of job information
1 group decision making biases (bias towards presenting common information, groupthink,
conformity) especially when there is a hierarchy
2 self-presentation – exaggerations and omissions, especially in ability statements (forget
elements)
3 cognitive – limited ability to process information (information overload)
-Job analysis relies on a lot of subjective judgement. Trick is to make careful choices about
what information to collect and how (depending on the job)
Take in to account all important job information (issues)
-length of job cycle (time in which all instances of job are performed) exceeding observation
period
-no access to job site for personal observation
-no familiarity with job’s tasks, duties & responsibilities (naïve staff members)
-the job changes while investigation
-each method has (dis)advantages so use combinations of methods for greater confidence in
the results of job analysis
> job analysis is only as good as the information you gather
, How to gather information
1 Direct observation
-job analyst observes (live or video) representative samples of work behavior (is it relevant)
-should be unobtrusive (try not to be to visible/present)
-Helpful to use functional job analysis (FJA) > used to record observed tasks, what gets done
(existing tools)
-assumes static jobs (manual standardized, short-cycle jobs) so jobs rewuiring a lot of
concentration and mental work are not good for observation
Example Function Job analysis
2 Actual job performance by analyst
-job analyst actually performs the work (working as the
person performing the job)
-appropriate for jobs that the analyst can easily learn
3 Interviews
-most commonly used technique for establishing tasks,
duties and behaviors at work
-worker acts as his/her own observer & reports
-so the worker has to know the job, including hidden en
long-term aspects
-success is highly dependent on interviewer skill, trust and covering the whole content
domain and honesty of the interviewee (info-distortion) > so corroborate with other sources
or methods
Tips for conducting an interview
-structure the interview (pre-decide questions, check for appropriateness)
-questions should relate to purpose of analysis
-no personal or intimate maters (can result in resentment)
-do not ask about knowledge that interviewees don’t have
-clear and unambiguous wording
-no leading or loaded questions (social desirability)
-run multiple interviews with several workers & supervisors
4 Subject Mater Expert panels (SMEs)
-conduct interviews with groups of 6-10 subject matter experts
-should be representative (race, gender position, experience)
-discuss issues & resolve disagreements openly (who is the best at tasks, think of good and
bad workers and what distinguishes them)
-problem: social elements and groupthink (might not want to disagree with each other)
-especially good for extracting relevant KSAOS
5 Critical incidents
-meet with subject matter experts (SMEs) and ask them to describe especially good/poor
performance
-incidents are categorized according to job dimensions > end up covering the behavioural
domain (also helps with performance management)
Lecture 1 – Introduction and Job analysis
-chapter 1,3 & 7
-Binning validity of personnel decisions
*Binning & Barrett model
How to measure someone’s success potential (relevant characteristics) > and predict their
success at the job (how can we see if a person is successful after being hired)
Snapshot of potential (predictor) > capture how they perform (criterion)
Construct domain Performance domain
-Conceived by researchers with reference Determines/influenced by organizational
to some theoretical framework decision makers & selection specialists
-aim is to explain general regularities in -aim is to translate goals into normative
human behavior/ predictors of certain statements of valued behaviors &
behavior outcomes
-for example extraversion -for example negotiating
So performance domain (coming up with innovative ideas) > measure of performance
(supervisor report) > which construct domain are related/predictors to innovative ideas
(openness to experience or divergent thinking) > measure of construct (brain storming task)
, Show that there is an
empirical relationship
between your predictors and
criterion (behavior) – so the
predictors that you choice
predict how well people
perform (you want to include
as many predictors as
possible) Yellow line
*Binning & Barrett
-validation = the process of gathering
evidence or evaluating the necessary
data
-different kinds of evidence in assessing validity
-degree to which the evidence supports the inferences (expectations) made from the scores
or outcomes of the test/procedure (are you right to infer this and what evidence is needed
to support this inference)
Job Analysis
1. Define the job & specify what employee behavior is necessary to complete it/what
behaviors are needed to perform (job description)
2. Then identify which characteristics underlie this behavior/ which psychological
constructs underlie performance (job specifications) KSASOs = knowledge, skills,
abilities and or other characteristics
-should indicate the minimal requirements/ qualifications
-lies at the heart of almost every HRM program or activity > need accurate information
about a job
- goal is to define each job in terms of behaviors necessary to perform the job and develop
hypotheses about personal characteristics necessary to perform it
-important for; job descriptions and specifications, personnel selection, performance
appraisal, training & development
Skills = practiced acts. They can learn over time to the limits of one’s natural ability
Ability = individuals may be limited in their capacity to change depending on their ability
(certain natural range)
How to establish job specifications
-SMEs 1 (subject matter experts) develop list of tasks & KSAs
-SMEs 2 judge the necessity of tasks and KSAs (should it be a pre or can they be learnt the
skill)
-SMEs 3 judge clarity & adequacy of level, finalize
To what extent is the profile indicated suitable (0-3) so finalize the list
The results of job analysis affect all other aspects of personnel selection. You need to define
the job first before you can even begin to find a person to match that job
,Choices awaiting the job analyst
1. Activities (descriptions) vs Attributions (specification)
-do you want to do a job analyses for training? Then focus on attributions
2. General vs specific levels of detail (flexibility)
3. Qualitative data (narrative/ career planning) vs quantitative data (numeric fixed
scales, time, frequency, importance, criticality)
4. Taxonomy-based (use already existing questionnaires to gather information) vs
blank slate (start with a blank slate and see what data shows in observation) >
looking for general attributes that fit the job)
5. Get information from Observer (trained job analyst) vs incumbents (people on the
job) & supervisor (good to get information from all to get the total picture)
6. KSAs vs KSAOs (O = other personal characteristics like personality traits, values and
attitudes)
7. Single job vs multiple-job comparisons (justify use of same/different selection
system)
8. Descriptive vs prescriptive (does the job already exist or not and do you have to
specify what the job entails)
Reliability and Validity of job analysis
Job description – do they accurately represent job content, environment and conditions of
employment?
-the more systematic the data gathering is the more valid the output
-make sure the sample is big enough
Job specification – do people possessing the personal characteristics believed necessary for
successful job performance in fact do perform more effectively on the job than people
lacking such characteristics
Factors influencing the accuracy of job information
1 group decision making biases (bias towards presenting common information, groupthink,
conformity) especially when there is a hierarchy
2 self-presentation – exaggerations and omissions, especially in ability statements (forget
elements)
3 cognitive – limited ability to process information (information overload)
-Job analysis relies on a lot of subjective judgement. Trick is to make careful choices about
what information to collect and how (depending on the job)
Take in to account all important job information (issues)
-length of job cycle (time in which all instances of job are performed) exceeding observation
period
-no access to job site for personal observation
-no familiarity with job’s tasks, duties & responsibilities (naïve staff members)
-the job changes while investigation
-each method has (dis)advantages so use combinations of methods for greater confidence in
the results of job analysis
> job analysis is only as good as the information you gather
, How to gather information
1 Direct observation
-job analyst observes (live or video) representative samples of work behavior (is it relevant)
-should be unobtrusive (try not to be to visible/present)
-Helpful to use functional job analysis (FJA) > used to record observed tasks, what gets done
(existing tools)
-assumes static jobs (manual standardized, short-cycle jobs) so jobs rewuiring a lot of
concentration and mental work are not good for observation
Example Function Job analysis
2 Actual job performance by analyst
-job analyst actually performs the work (working as the
person performing the job)
-appropriate for jobs that the analyst can easily learn
3 Interviews
-most commonly used technique for establishing tasks,
duties and behaviors at work
-worker acts as his/her own observer & reports
-so the worker has to know the job, including hidden en
long-term aspects
-success is highly dependent on interviewer skill, trust and covering the whole content
domain and honesty of the interviewee (info-distortion) > so corroborate with other sources
or methods
Tips for conducting an interview
-structure the interview (pre-decide questions, check for appropriateness)
-questions should relate to purpose of analysis
-no personal or intimate maters (can result in resentment)
-do not ask about knowledge that interviewees don’t have
-clear and unambiguous wording
-no leading or loaded questions (social desirability)
-run multiple interviews with several workers & supervisors
4 Subject Mater Expert panels (SMEs)
-conduct interviews with groups of 6-10 subject matter experts
-should be representative (race, gender position, experience)
-discuss issues & resolve disagreements openly (who is the best at tasks, think of good and
bad workers and what distinguishes them)
-problem: social elements and groupthink (might not want to disagree with each other)
-especially good for extracting relevant KSAOS
5 Critical incidents
-meet with subject matter experts (SMEs) and ask them to describe especially good/poor
performance
-incidents are categorized according to job dimensions > end up covering the behavioural
domain (also helps with performance management)
