Transcript Document

PSY 6430 Unit 6
Job Analysis
Tonight and Monday: Lecture
Monday, 3/27: Exam
1
Introduction: Job Analysis (NFE)



Job analysis is the bedrock of any selection
procedure and is required from a legal
perspective if the selection procedure results in
adverse impact and is challenged in court
There are several different types of job analysis
procedures
I am going to focus on two


Task analysis (and task questionnaire)
Position Analysis Questionnaire (PAQ)


Commercially available structured job analysis questionnaire;
an off-the-shelf job analysis questionnaire
From the hand-out I gave you from the 6th edition of the text
(just need a little intro before I talk about the project)
2
Introduction: Job Analysis (NFE)

The authors first talk about a job analysis interview





But that is rarely used by itself
Used by itself only when there are only a few job incumbents in
the position
Usually combined with the task analysis and used only as the
first step in the process
The task analysis and PAQ are the most popular types
of job analysis
The authors also talk about the Critical Incidents
Technique, but I am not going to focus on that


Critical incidents technique focuses on excellent and poor ways
of handling important job-related situations/scenarios
It’s great to use for developing interview questions and
situational tests but does not result in a comprehensive job
analysis
(always started with the exemplary performers; great to use after a comprehensive analysis to develop some of the
3
instruments/methods for selection; main usefulness after a comprehensive analysis – supplement; face validity)
Relationship of job analysis to type
of validity procedure*

As I indicated last week, the type of validity
procedure you use is dependent upon several
factors




Sample size
Whether the KSA is observable or not
Whether you use an off-the-shelf test or write the test
What type of cut-off score procedure you want to use
*Learn for the exam! Next slide shows the relationship.
4
Relationship of job analysis to type
of validity procedure

Once you decide what validity procedure you want to use, then you
must make sure that your job analysis procedure is appropriate for
that type of validity procedure
Validity Procedure
Type of Job Analysis
Content validity
Task analysis only
Criterion-related validity
Task analysis, PAQ, any type
Validity generalization
Any type is OK, should be the
same as the type used
by the other company
(required by Uniform Guidelines; you cannot use the PAQ or any other type - I’ll explain why you cannot use the PAQ a bit later; predictive
and concurrent; any type because the proof is in the correlation unlike content solely expert judgment; job component - PAQ)
5
Task Analysis Project





Most companies are now using content
validity, so I am focusing on the task
analysis
Due Monday, April 22 (final exam week)
Worth 70 points (Units 9 and 10)
Get started now
This is a time consuming project
(but I obviously couldn’t assign it until we got to this point in the course)
6
Task Analysis Project

Task analysis is the first step in the development
of a task inventory (job analysis questionnaire)



I am not going to have you develop the questionnaire
or get ratings for the task statements or KSAs
Project consists of just the first step in the
process - the identification of tasks and KSAs
linked to each of the tasks
Three models in the course pack

Lindsay Street, Thor Flosason, & Julie Slowiak
7
Task Analysis Project

Task analysis is the most versatile job
analysis procedure because


It can be used for both content and criterionvalidity studies and
Is legally required for content validity (if tests
are challenged in court)
8
Task Analysis Project


Resource on the web that will make this project a
lot easier and a lot less time consuming for you
O*NET




Occupational Information Network – Dept. of Labor
Web site address is in the instructions
To start with, go to O*NET and use
“Bartender” as the example position
Also use this for your own job search
Related titles and ability “buzz words”
for your resumes
(more detailed instructions in the description of project)
9
Task Analysis Project


You may conduct the analysis for one job
incumbent
You cannot do the task analysis for a job you
currently have or have had in the past


I want you to have the professional experience of
collecting information about a job that you are not
familiar with - this is much harder than you think
The job may be part-time or full-time
10
Task Analysis Project

The analysis will consist of





Task statements
KSAs linked to each task statement
Work activities: Physical activities and requirements
(required for ADA)
Work Context: Environmental conditions (also good for
ADA)
Work Styles: Typical working incidents (again good for
ADA)


This is where many job analysts put things like “persuading
individuals, working under stress, getting along with others”
that are critical for ADA
Any licensing or certification requirements
(don’t forget job attendance -getting to work on time, and regular attendance, if that is required - ADA; Tasks MUST BE EXACT FORMAT)
11
Task Analysis Project

Task statement format



Gatewood, Feild, & Barrick, Figure 7.1, page 256
My lecture tonight
Worksheets provided in course pack
(Tasks MUST BE EXACT FORMAT)
12
Task Analysis Project

Caution! The task statement format differs in
O*NET


O*NET is the end result of a task inventory, not the
starting point
O*NET has 8 sections after the tasks and KSAs

I am requiring only the following three




work activities
work context
work styles
But adding

professional licensing and certification requirements
13
Task Analysis Project

You should have between 15-30 task
statements


Part of the problem with writing task statements is
determining how specific or how broad they should
be
You will understand this difficulty when you begin to
write the task statements
14
Task Analysis Project: Steps
See the project description which indicates
where you find instructions for




Writing task statements
Writing KSAs
Documenting work activities, work context, work
styles and licensure/certification requirements
15
Task Analysis Project: Steps
Preparing for the meetings


Before the meeting, obtain any written
documentation that exists about the job




Job descriptions, organizational chart, etc.
Locate job title/position using O*Net (if you can find a similar
position)
Meet with the job incumbent three times
First meeting: Task statements only (1.5-2 hrs)



Ask the incumbent just to identify/state the major aspects or
areas of his/her job
After you have that list, go back to the first major area and
ask for the tasks in that area
Move onto the next major area, etc.
(different interview procedure from GFB; they provide interview guides on pages 254&259 are very good, but 3 meetings rather than 1;
Also strongly recommend that you record the interview – never going to remember everything, miss things )
16
Task Analysis Project: Steps

Second meeting (1.5-2 hours)

Confirm accuracy of task statements, obtain KSAs,
and perhaps work activities, work context, work
styles and licensure/certification (time permitting)

Before meeting with the incumbent, type up a list of the
task statements



Identify any missing components to the tasks that you need to
ask about
Have the incumbent review the task statements for accuracy
and completeness
Take each task statement one at a time and ask the
incumbent what KSAs are required to perform that
particular task

GFB interview guide, Figure 7.3 on page 254 may be helpful
here
17
Task Analysis Project: Steps

Third meeting: Review of entire document



Type up the entire document and have the job
incumbent review it.
If revision is required, then you should ask the
incumbent to review it one more time
Some students have done this last step virtually which
is fine – but it is not good to do the first two steps
virtually
18
Task Analysis Project: Format

See models in the course pack

Description of the company and position


Include the name of the position in the company
and the corresponding O*Net position name, if there
is one
Description of the methods/steps you used to
develop the task analysis.
How many times did you meet with the incumbent?
 What did you do during each meeting?
 How long was each meeting?

19
Task Analysis Project: Format



Task Statements with KSAs linked to each task
statement
Work Activities*
Work Context*


Work Styles*


Environmental conditions and physical
requirements
Key ADA requirements - getting along with others,
dependability, coming to work regularly, etc.
Licensure/Certification Requirements
*See bartender example on O*Net for examples
20
Questions
about the project?
21
SO2: Job Analysis (review)

Discrepancy between the professional and the
legal requirements regarding job analysis

Job analysis is not required professionally for predictive
and concurrent validity (it is required for content and
validity generalization)


The proof of job-relatedness rests with a statistically significant
correlation between test scores and job performance measures
It is, however, required legally for all types of validity
procedures
22
SOs 7&8: Task Statement Format

What activity is performed?


To whom or what?


The object of the verb
Why?


Use a single present tense action verb - verb ends in
an “s”
What is produced - what is the expected outcome of
the action
How?

Using what materials, tools, procedures, guidelines,
or equipment
(moving to Sos 7&8; four components of a task statement; two tasks for a job you have held)
23
Task Statement Examples
Welfare Eligibility Examiner

What activity is performed?


To whom or what?


Of client on eligibility form
Why?


Asks questions, listens, and records answers
In order to determine eligibility for food stamps
How?

Using eligibility form, eligibility criteria in manual,
interviewing techniques
(from GFB; note more than one action verb that are all related to the same WHY; Why is the most frequently
omitted component when students write task statements; clean up the task statement – next slide)
24
Task Statement Examples
Welfare Eligibility Examiner

Clean it up
Asks client questions, listens, and records
answers on standard eligibility form to gather
information from which the client’s eligibility for
food stamps can be determined using knowledge
of interview techniques and eligibility criteria.
(notice, several different verbs, actions that go together for the same “in order to”; often a question students have, next
Slide another example)
25
Task Statement Examples
Welfare Eligibility Examiner

What activity is performed?


To whom or what?


Of applicant
Why?


Determines eligibility
In order to complete client’s application for food
stamps
How?

Using regulatory policies as a guide
(another example - same job; once they obtain information from the client, then they must decide whether the person is
eligible using policies; next slide different topic)
26
SO9: Problems with Job Interview as only
source (this should not surprise you)

Lack of standardization across job analysts


Likely to get different information
Not practical for large numbers of incumbents
because it’s one-on-one (2-3 hrs at least)


Once you have more than 5-10 job incumbents, it
gets very time consuming and hence expensive
Legally, you must have a representative sample
stratified by geographical region (if relevant), and
protected classes, thus you may end up interviewing
more incumbents than would otherwise be necessary
(again, rarely, if ever used alone; I have just asked you to list any three of the ones given; more on next page)
27
SO9: Problems with job interview as only
source, cont.

Easy to miss things when you are interviewing someone
due to the quick verbal exchanges (as you will find out
when you do your project)



Thus, unless you thoroughly document each interview, the legal
requirements for the job analysis may not be met
Results are highly dependent upon the skills of both the
interviewer and interviewee
Text lists some others - but clearly it is not good to use
interviews alone in most cases
28
SOs 10 & 11: Task Statement and
KSA Rating Scales


The Uniform Guidelines discusses the technical
standards for a job analysis if it is used for
content validation
 Google or see www.ipmaac.org/files/ug.pdf
All of the following four rating scales must be
included for task statements




Frequency of task performance
Task importance or criticality (add essential function?)
Task difficulty
Whether the task can be learned on the job within a
relatively short period of time (within 6 months)
(KSA rating scales next)
29
SOs 10 & 11: Task Statement and
KSA Rating Scales

All of the following rating scales must be included
for KSAs (if, again, used for content validation)



Is the KSA necessary for successful job performance?
Is the KSA required upon entry to the job or can it be
learned within a relatively short period of time (within 6
months)
How difficult is it to obtain the KSA?

What type of education/experience/licensure/certification is
necessary?
(as you can see, the task inventory/questionnaire becomes quite extensive; 4 scales for tasks at least 3 for KSAs)
30
SO12: Obtaining names of the job incumbents

I strongly disagree with the authors about
obtaining names on questionnaires



Employees tend to be very nervous and concerned
about how the information will be used
When I have interviewed individuals, even one-onone, I have almost always had to assure them that
they would not be identified - rather the information
would be combined with the information of others,
with no identifiers.
Most employees just don’t trust the “corporate office,”
HR or their bosses not to use the information against
them (evaluation, downsizing, elimination of job)

You will probably be interviewing people you don’t know
(you have no established relationship with these individuals and if you are from “corporate”
you are immediately distrusted)
31
SO12: Obtaining names, cont. (this slide NFE)

You should handle this just like you handle a participant
list for research



You should create a master list with names and code
numbers that correspond to your stratified sample
Keep it under lock and key - you are the only one who
should have access to it (or if you are not in charge of the
analysis, the person that is)
Put the code numbers on the questionnaire, so you know
whether you have a representative sample when the
questionnaires are returned
(Why codes, next slide)
32
SO12: Obtaining names, cont. (this slide NFE)

Why codes on questionnaires?



There is a legal requirement to have a representative
sample for your job analysis
The legal requirement pertains to who participated in
the job analysis, not simply who you sent questionnaires
to
If you don’t get the sample you need, contact everyone
in the underrepresented categories, with the standard,
“We are contacting you because we recently sent you a
questionnaire to complete. If you have not completed
the questionnaire, we would greatly appreciate it if you
would do so. If you have, please accept our thanks,
and ignore this message.”
33
SO13: Advantages and Disadvantages
of the Task Inventory

Advantages


It is an efficient way to collect data from a large
number of individuals, particularly if they are
geographically dispersed
Data can be easily quantified due to the rating
scales
34
SO13: Advantages and Disadvantages
of the Task Inventory

Disadvantages



Development is time consuming and expensive
Motivation to complete the questionnaire can be a
problem because it is often very long due to the rating
scales that must be included
Ambiguities and questions may arise when individuals
complete the form that can’t be dealt with, which may
result in inaccurate and less reliable information
35
SO14: Court case related to job analysis,
the two implications: intro

Kirkland v NY State Department of Correctional Services,
1974





The litigated test was a promotional exam for correctional officers
The passing rates for whites was 31% while the passing rate for
blacks was 8%
Adverse impact had been demonstrated thus the burden of proof
shifted to the organization
Organizational defended the exam on the basis of content validity
The court ruled in favor of the plaintiffs ruling that


The job analysis was not adequate, but more importantly because job
analysis had been stressed in other cases
The process used to construct the test and the content of the test was
not adequate
(very important case related to job analysis when used for content validity; not in text)
36
SO14: Court case, the two implications*

Not only do you have to document the job analysis
process, but you also must document the process used
to construct the test



Did you develop a matrix that linked the KSAs to specific items on
the test?
Who participated in that process and when did those meetings
occur?
Remember, court cases often occur years after the development
of the test and you are simply not going to remember the dates of
the meetings, who attended, or the outcome of those meetings
*For the exam
(first one this slide, second next slide)
37
SO14: Court case, the two implications

The court may look at the test itself, not just the job
analysis procedure and ask why specific items were
included and why they were weighted the way they were
given the job analysis

This is why it is important/essential to construct a test if you use
content validity rather than using an off-the-shelf test, and why I
have been stressing that point
38
Introduction: Position Analysis
Questionnaire (PAQ) (NFE)

The PAQ is a structured, standardized job analysis
questionnaire that can be used to analyze virtually any
job except managerial and professional positions



There is no other job analysis procedure that has been
researched more extensively



McCormick, Jeanneret, & Mecham
Professional and Managerial Position Questionnaire (PMPQ)
Research started in 1974, over 30 years ago
Over 330,000 different jobs have been analyzed in the
past 30 years
Can also be used for job evaluation (salary
determination) and development of competencies
(onto the PAQ)
39
Introduction: Position Analysis
Questionnaire (PAQ) (NFE)





The authors moved a description of the PAQ to the
validity chapter – job component validity*
Job components being identified by the PAQ analysis
and then using validity data/information from the PAQ
database, selecting test instruments
Using them without doing a local validity study
I don’t recommend that for legal reasons although at
least one court has upheld its use this way
I recommend a more conservative approach, using it to
identify selection tests and then conducting a local
empirical validity study

Cannot be used for content validity – job/worker attributes, not
tasks
(relates job components to skills measured by selection tests; transportability for jobs for which job tasks
are not similar – requirement for validity generalization; building block approach )
40
Introduction: Position Analysis
Questionnaire (PAQ) (NFE)

The PAQ analyzes jobs in terms of 187 job
elements or attributes that are broken down into
six categories

Information input


Mental processes


When and how a worker gets information needed to perform
the job
The reasoning, decision making, planning, and information
processing activities that are involved in performing the job
Work output

The physical activities, tools, and devices used by the worker
to perform the job
41
Introduction: Position Analysis
Questionnaire (PAQ) (NFE)

Relationships with other persons


Job context


What relationships with other persons are required to
perform the job?
The physical and social context in which the work is
performed
Other job characteristics

The activities, conditions, and characteristics other than
those already covered
42
Introduction: Position Analysis
Questionnaire (PAQ) (NFE)

Information input (broken into visual, nonvisual,
sensory & perceptual processes, and estimation
activities)
 Visual sources of job information - examples




Written material (books, reports, office notes, signs)
Quantitative materials (graphs, accounts, tables of numbers)
Pictorial materials (drawings, blueprints, diagrams, maps)
Nonvisual sources of job information - examples




Verbal sources (instructions, requests, interviews, orders)
Nonverbal sources (noises, engine sounds, sonar, music)
Touch (pressure, pain, feeling the texture of a surface)
Taste (food tasting, wine tasting)
(break down Information input: each element/attribute is rated on a 6 point scale from 0, does not apply to 5, very substantial)
43
Introduction: Position Analysis
Questionnaire (PAQ)




The PAQ focuses on worker attributes rather than job
tasks
For example, Arvey & Begalla conducted PAQ analyses
of a home maker’s position and a police officer’s position,
and found the profiles to be very, very similar
Because of that, it cannot be used for content validity*,
(which is its main disadvantage)
If you use the PAQ you must conduct a criterion-related
validity study or use validity generalization* (or job
component validity)
*SO23 asks you to learn these two points
44
SO20A: Underlying Rationale of the
PAQ (2 assumptions*)
1. There are a limited number of generalized human
behaviors**/attributes/job elements and all jobs can be
characterized by the extent to which any particular job
requires each of those job attributes/elements


That is, there are a certain number of job attributes/elements that
are the “building blocks” for all jobs that exist (the PAQ authors
identified 187), and each job is constructed from a set of these
building blocks
Jobs can be analyzed by identifying which of the “building blocks”
are required and the extent to which each is required in
comparison to all other jobs that exist
*The study objectives ask you to learn the two assumptions
(**the authors refer to the “job elements” as generalized human behaviors, but they are not using the term the way we as
Behavior analysts use it, so I prefer to talk about the job attributes or elements)
45
SO20A: Underlying Rationale of the
PAQ (2 assumptions)
2. Each job attribute/element is normally distributed across
all possible jobs that exist

Each of the 187 job attributes/elements has its own normal
distribution across jobs (I’ll show a slide of this in a moment)
 e.g., (a) the use of manually powered tools, (b) planning and
scheduling, (c) the amount of oral communication, etc.
46
SO20B: How are the important job attributes
for each job determined by the PAQ?


Each job analyzed is compared to the normal
distribution for each of the 187 job attributes/elements to
determine how much of this job attribute/element is
required in comparison to all of the other jobs that exist
Data are reported in terms of percentiles


If a particular job attribute/element falls at the 90th percentile, it
means that 90% of all other jobs require less of this attribute
than the current job and thus it is a very important job
attribute/element for this job
What would it mean if planning and scheduling fell at the 10th
percentile?
47
Each job attribute is
normally distributed
across all jobs
There is a normal
distribution curve for
each of the 187 attributes
in the PAQ
Percentiles
The percentage of jobs
that require this attribute
less than the target
job position
48
Percentiles
Use of manually powered
tools: 80th percentile
This job requires more use
of manually powered tools
than 80% of all jobs
Percentiles
Planning and scheduling:
10th percentile
This job requires more
planning and scheduling
than only 10% of all jobs
49
PAQ (NFE)



Obviously, the PAQ analysis requires a
very large data base of other jobs so you
can compare your job to those jobs
This means if you use the PAQ, you must
use the on-line data analysis of PAQ
Services
See www.paq.com
50
SO23: PAQ and type of validity study


Because the PAQ does not provide task
statements and KSAs that are operationalized, the
PAQ cannot be used as the basis for content
validity
If you use the PAQ, you must conduct a criterionrelated validity study (job component
notwithstanding)


Concurrent
Predictive
(redundant, but students have had trouble with this in the past; or job component - )
51
SO24: PAQ Disadvantages

First, the PAQ requires the reading level of a
college graduate



It is a very complicated questionnaire, with different
rating scales used for different sections
In most settings, you can’t just hand over the PAQ to
someone and ask him/her to complete it
The best way to administer it is individually, with a
trained job analyst reading the questions and
recording the answers

Also may be more reliable because incumbents may rate
certain elements as being more significant than they are, just
because they haven’t been exposed to a lot of jobs (and
remember the PAQ analysis relies on a comparison to other
jobs)
52
SO24: PAQ Disadvantages

Second, because the PAQ analyzes jobs in
terms of job elements and does not actually
identify task statements, it cannot be used for
some purposes that job analyses are typically
used for



Content validity
Job descriptions
Performance appraisals
53
SO25: PAQ Major ADA Disadvantage

Since the passage of ADA, the fact that the PAQ cannot
be used to develop job descriptions has become a major
disadvantage because ADA requires a job description for
each and every job prior to selection (PAQ was
developed in 1974, ADA was passed in 1990)

How else is the organization and applicant going to know what
the essential job functions are and whether or not a disabled
individual can perform them with or without reasonable
accommodation?
54
SO25: Major ADA disadvantage, cont.
but this slide NFE

Selection specialists usually do not develop job
descriptions, rather that responsibility usually falls to the
compensation specialists who have to justify salaries
assigned to positions in terms of their tasks and
responsibilities



Common organizational “silo” in organizations
But, a job analysis/study can be used for both purposes
Some organizations use the same procedure for selection
and compensation purposes, but very, very few
55
SO26: PAQ Advantages



26A: First, it provides a standardized, quantitative
means for collecting job information that permits you to
assess all jobs the same way, thus enabling you to
compare across positions/jobs
26B: Delete (in the SO, I refer to validity generalization,
not job component validity)
25C: Third, from a professional technical perspective it
provides very valid and reliable job information



There are extensive data attesting to its validity and reliability
Makes it easy to defend the job analysis process if selection
tests are questioned in court
25D: Fourth, although not stated in the text, it has a
very important legal advantage, in that it has stood up
in court and the EEOC has strongly recommended its
use
(now, let’s look at its advantages; delete SO 26B)
56
PAQ Advantage (NFE)

Another advantage is that the PAQ has been shown to
produce very reliable and valid job information with as
few as three job incumbents!
 I used to have students learn this as an advantage but
it tended to confuse them because:


Usually you need to have more than three job incumbents
participate in a job analysis in order to have a legally and
professionally acceptable representative sample (stratified by
gender and ethnic/racial background)
You can’t use it for content validity, so you can’t use it for
positions that have a small number of incumbents as a basis
for justifying the job-relatedness of a selection procedure
57
SO27: Selecting tests after a PAQ Analysis


Analyze PAQ data on-line using PAQ Services
Based on the PAQ report, select tests



Empirically validate the tests


PAQ indicates tests in the GATB that are most relevant
You cannot use these, but you can locate similar offthe-shelf tests or ask PAQ to send you a list of similar
tests
Remember you cannot use the PAQ for content validity
Check the correlation for statistical significance
58
Sample PAQ Report:
Senior Shipping and Receiving Clerk
Mean
Score
Predicted
Validity
Coefficient
General Intelligence
95.1
.19
.33
82.2
Verbal Aptitude
96.0
.10
.21
80.3
Numerical Aptitude
95.6
.18
.24
78.7
Spatial Aptitude
96.5
.19
.47 <
79.3
Clerical Perception
98.3
.12
.34 <
83.0
GATB Test
1A
Probability Predicted
Low Score
Of Use1
“<“ indicates the test should be considered as a selection measure
Lists the nine tests in the GATB
• These tests are restricted by the government
• They are not available to private employers
• Thus, PAQ services provides a list of alternative tests
59
Sample PAQ Report:
Senior Shipping and Receiving Clerk
GATB Test
Mean
Score
Predicted
Validity
Probability Predicted
Low Score
Coefficient
Of Use1
General Intelligence
95.1
.19
.33
82.2
Verbal Aptitude
96.0
.10
.21
80.3
Numerical Aptitude
95.6
.18
.24
78.7
Spatial Aptitude
96.5
.19
.47 <
79.3
Clerical Perception
98.3
.12
.34 <
83.0
1A
“<“ indicates the test should be considered as a selection measure
Mean score on the exam that is predicted for your position.
The higher the score, the more relevant the test is
60
Sample PAQ Report:
Senior Shipping and Receiving Clerk
Mean
Score
Predicted
Validity
Coefficient
General Intelligence
95.1
.19
.33
82.2
Verbal Aptitude
96.0
.10
.21
80.3
Numerical Aptitude
95.6
.18
.24
78.7
Spatial Aptitude
96.5
.19
.47 <
79.3
Clerical Perception
98.3
.12
.34 <
83.0
GATB Test
1A
Probability Predicted
Low Score
Of Use1
“<“ indicates the test should be considered as a selection measure
This is the predicted validity coefficient for your position if you
correlated the test scores with a measure of performance
• But remember, this is only predicted; you still must
conduct an empirical validity study
61
Sample PAQ Report:
Senior Shipping and Receiving Clerk
Mean
Score
Predicted
Validity
Coefficient
General Intelligence
95.1
.19
.33
82.2
Verbal Aptitude
96.0
.10
.21
80.3
Numerical Aptitude
95.6
.18
.24
78.7
Spatial Aptitude
96.5
.19
.47 <
79.3
Clerical Perception
98.3
.12
.34 <
83.0
GATB Test
1A
Probability Predicted
Low Score
Of Use1
“<“ indicates the test should be considered as a selection measure
This indicates whether the US Employment Service would be likely to
use this test in a selection test battery
•
The higher the score, the more likely the USES would use it
• Notice that the tests that are selected are not always the ones
with the highest predicted validity coefficients
• The tests are intercorrelated, thus you don’t get unique
information from each one
(click)
62
Sample PAQ Report:
Senior Shipping and Receiving Clerk
Mean
Score
Predicted
Validity
Coefficient
General Intelligence
95.1
.19
.33
82.2
Verbal Aptitude
96.0
.10
.21
80.3
Numerical Aptitude
95.6
.18
.24
78.7
Spatial Aptitude
96.5
.19
.47 <
79.3
Clerical Perception
98.3
.12
.34 <
83.0
GATB Test
1A
Probability Predicted
Low Score
Of Use1
“<“ indicates the test should be considered as a selection measure
This shows the potential cutoff or passing score for the test
• It is set at 1 SD below the mean score listed in column 2
• If used, approximately 1/3 of the current employees would
have been eliminated when they applied for the position
• The 1/3 is based on the normal distribution (1/3 of scores
fall below 1 SD below the mean)
63
PAQ: Alternatives to GATB Tests

General Intelligence




Verbal Aptitude




Wonderlic Personnel Test
Adaptability Test
Test of Learning Ability
Personnel Tests for Industry - Verbal
Short Employment Tests - Verbal
Employee Aptitude Survey - Verbal
Numerical Aptitude






Personnel Tests for Industry - Numerical
Short Employment Tests - Numerical
Employee Aptitude Survey - Numerical
Arithmetic Index
Flanagan Industrial Tests - Arithmetic
Arithmetic Fundamentals Test
(cont.)
64
PAQ: Alternatives to GATB Tests

Spatial Aptitude




Revised Minnesota Paper Form Board
Employee Aptitude Survey - Spatial
Flanagan Industrial Tests - Assembly
Clerical Perception



Employee Aptitude Survey - Visual Speed and Accuracy
Short Employment Tests - Clerical
Minnesota Clerical Tests - Names
Note that the names of the tests vary even though they measure
approximately the same thing.
It is also true that even though tests may be named the same,
they may measure different things.
(last slide)
65
That’s All Folks
Questions on U6?
66