A QUICK GUIDE TO HUMAN RESOURCE TESTING pot

MARIAN POWER
A QUICK
GUIDE TO
HUMAN
RESOURCE
TESTING
Marian Power qualified as a psychologist in 1973, and has been registered since 1987. She has
worked as an organisational psychologist, equal opportunity manager and human resources
manager in federal, state and local government. Her roles encompassed recruitment and
selection, conflict resolution and management of grievances. Marian is currently employed as
consultant psychologist with the Australian Council for Educational Research providing advice
to psychologists and human resource professionals regarding the selection of the most
appropriate assessment instruments for their particular purpose. She also provides
accreditation training in the use of restricted tests.
Marian is an active member for the Australian Psychological Society, the College of
Organisational Psychologists, the Australian Association of Psychological Type, the Australian
Human Resources Institute and the Australian Association of Careers Counsellors.
A QUICK GUIDE TO
HUMAN RESOURCE TESTING
Can you afford to make the wrong recruitment selection decision?
The cost of a wrong selection decision can be up to one-and-a half times the salary of the job,
let alone the time taken in the re-hiring process.
Studies have shown that appropriate assessment tools enhance the chances of making a
good selection and recruitment decision. Testing is also important to the human resources
(HR) practitioner in a variety of other contexts, including team building, change management
and ongoing organisational needs.
A Quick Guide to Human Resource Testing is designed as an introduction, a refresher and a
quick reference guide for HR practitioners who use or plan to use assessment instruments in
any context. It includes explanations, tips, case studies and suggestions to help you get the
most out of your HR testing.
9

ISBN 0-86431-458-2
780864 314581
Quick guide to HR cover.qxd 10/6/04 11:49 AM Page 1
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page iv
MARIAN POWER
A QUICK
GUIDE TO
HUMAN
RESOURCE
TESTING
ACER Press
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page i
First published 2004
by ACER Press
Australian Council for Educational Research Ltd
19 Prospect Hill Road, Camberwell, Victoria 3124
Copyright © 2004 Marian Power
All rights reserved. Except under the conditions
described in the Copyright Act 1968 of
Australia and subsequent amendments, no part
of this publication may be reproduced,
stored in a retrieval system or transmitted in any
form or by any means, electronic, mechanical,
photocopying, recording or otherwise, without
the written permission of the publishers.
Edited by Ronél Redman
Cover and text design by Mason Design
Printed by bpa DIGITAL

National Library of Australia Cataloguing-in-Publication data:
Power, Marian.
A quick guide to human resource testing.
ISBN 0 86431 458 2.
1. Employee selection - Australia - Handbooks, manuals, etc.
2. Employment tests - Handbooks, manuals, etc. 3.
Employees - Recruiting - Australia - Handbooks, manuals,
etc. 4. Employee selection - Law and legislation -
Australia. I. Title.
658.3110994
Visit our website: www.acerpress.com.au
Acknowledgement
In writing this guide, I would like to acknowledge the support I received from my husband, Adrian; and
Dominic, Stephen and Caithlin who freed computer time for me and provided encouragement.
Ralph Saubern, Test Publisher at ACER Press, offered frequent support and constructive advice. My
colleagues in Organisational Psychology and Human Resource Management have shared their experience
over twenty-five years, which has been an invaluable addition to my formal learning. Thank you all.
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page ii
Contents
Introduction 1
Section 1:Test Selection and Administration
When to Use Tests
2
Recruitment/Selection
Organisational Development
Career Planning
Types of Tests 4
Ability and Aptitude Tests
Achievement Tests
Personality Assessments

Interest Inventories
Types of Test Questions 6
Response Type
Content Type
Test Formats 8
Questions
Answers
Score Keys
Computerised Testing
The Manual
Choosing the Most Appropriate Test – the Job 10
Selection Criteria
Job Level
Choosing the Most Appropriate Test –
the Practicalities
12
Time
Budget
Hand Scoring vs. Computer Scoring
Qualifications
Test Administration – Best Practice 14
Environment
Administrators
Getting Started
Using Aids
The Manual/User’s Guide
Collecting Materials
Section II:Test Interpretation
Statistics 1: The Basics
16

Normal Population
Mean
Range
Standard Deviation
Statistics 2: Test Scores 18
Changing a Raw Score to a Standardised Score
Percentile Ranks
Stens and Stanines
T-scores
Statistics 3: Reliability, Validity and
Measurement Errors
20
Reliability
Validity
Margins of Error
Section III: Reporting and Feedback
Reporting Results
22
Report Types
Using Reports for Feedback
Feedback to Candidates
Feedback to Managers
Section IV: Ethical and Legal Issues
Test Security and Candidates’ Rights
24
Equal Employment Opportunity
(EEO) Legislation
Discrimination in Testing
Direct and Indirect Discrimination
Appendix I: Frequently Asked Questions 26

Appendix II: Glossary 28
Appendix III:Tests and Selection Criteria Guide 29
Appendix IV: Annotated List of Tests 30
Ability, Aptitude and Skills
Personality
Interest Inventories
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page iii
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page iv
Introduction
An ounce of prevention is worth a ton of cure.
Those who are put in charge of recruiting and selecting new staff face a decision-making process that needs
to be responsible in its implementation and produce a positive outcome. The cost of a wrong selection
decision has been estimated to be anywhere between one and a half times and five times the salary of the
job in question. Think of the advertising costs, time spent reading through and short-listing applications,
interviewing, testing, sourcing referees and notifying unsuccessful candidates, for a start. Then the
successful applicant commences but is not working at full capacity for a number of months, with existing
staff taking time out to help train the new recruit. After all that effort, what if the person selected turns out
to be unsuitable for the position? The recruitment process starts all over again.
Therefore, getting the right candidate in the first place is well worth the effort!
It is important to investigate the most efficient and effective ways of conducting a selection process to
maximise the chances of a positive outcome. Testing is an important part of this process. Studies have
shown that the chances of making a good decision in recruiting or selecting staff are enhanced when
structured interviews are combined with objective comments from referees as well as appropriately chosen
assessment tools.
Testing is also important in a variety of other contexts, including team building, change management
and other ongoing organisational needs. Good practice and appropriate use of tests are as vital to these
areas as they are in the selection and recruitment of staff.
This book is designed as an introduction, a refresher and a quick reference guide for human resource
practitioners who use, or plan to use, assessment instruments in any context. I hope the explanations, tips,
case studies and suggestions help you form a solid base for sound testing practice, and encourage you to

read further about how testing can help you provide the best possible human resource services.
In the Appendices there is a section on frequently asked questions, a glossary of terms that will clarify
any technical jargon, suggestions for which test to use in particular selection contexts, and an annotated list
of tests together with icons that indicate appropriate usage areas.
Marian Power
1
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 1
In human resource management, many decisions are made based on information that is gathered and
presented – for example, information about strategy planning, leave arrangements, observance of
occupational health and safety regulations. When dealing with important decision making regarding
people and the workplace, information should be gathered from all possible reliable sources.
HR practitioners use testing to collect reliable, objective information in order to optimise the decision-
making in a range of situations. Table 1 details some of these situations, which are further discussed below.
tests
When to use
Section I: Test Selection and Administration
2
Table 1: Using Testing in Human Resource Management
Recruitment/Selection Organisational Development Career Planning
Screening
Recruitment
Order of merit for ongoing selection
Staff development
Organisation development
Promotion
Team building
Change management
Career choice
Career change
Redundancy support

Succession planning
Recruitment/Selection
When a large number of applications is received for
an advertised vacancy, it is usual to eliminate the
first round on the basis of the applicant’s letter and
résumé; that is, those candidates who fail to
demonstrate that they adequately meet the
selection criteria. When there is still an abundance
of possible contenders, a screening test may be
used. Typically, such a test is of reasonably short
duration, may be administered to a large group
and assesses a common skill or ability required for
the position. The examiner, after consulting the
associated test manual, may decide to keep only
those candidates who score above a predetermined
cut-off score in the recruitment pool. Only these
candidates proceed to the next stage.
The next stage involves interviewing and more
specific testing – often two or more additional tests.
If only a small number of applicants applied for the
position, they should proceed directly to this stage.
Depending on individual preferences of the HR
practitioner and the selection committee, testing
can occur before the interviews, with only the best
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 2
A Quick Guide to Human Resource Testing
3
performing candidates then being invited to an
interview. Alternatively, all candidates can be
interviewed and only the best performers tested.

The tests used in this phase are chosen because
they assess abilities, skills and attitudes that are
clearly related to the selection criteria for the job in
question. Verbal reasoning, numerical reasoning,
work style preference, manual dexterity, spatial
reasoning, and personality measures are examples
of tests that are commonly used. (See page 4 for
more detail.)
From these test results, organisations that have
almost continuous recruitment needs can use
candidates’ performance to compile an order of
merit for further reference. This allows for
candidates who achieve scores above a certain
predetermined level to be invited to participate
further in the selection process as their position on
this ‘ladder’ is reached.
Organisational Development
When discrepancies are noted between the skills
or attitudes of current employees and the
requirements of the organisation, a range of testing
protocols is available to assist the HR practitioner
implement change. These are largely chosen from
the same pool of assessments available for initial
recruitment purposes, and should relate to specific
needs as they arise. For example, if an employer is
concerned that staff are experiencing difficulty with
new financial reporting requirements, numerical
reasoning tests may be given to identify those
candidates who would benefit from training to
address any gaps in their skills. This process needs

careful handling to ensure that employees see it
as constructive.
In addition to assessing individual performance,
personality or work style assessments are often
used with great success to identify deficiencies
in team performance, instigate more constructive
work style interactions and develop appropriate
team-building activities.
The change management process is often more
successfully implemented when the results of
assessments can be used to help address staff
members’ individual needs, communication styles
and attitudes to change.
Career Planning
There is a range of assessment tools available that
will assist in identifying employees’ career interests
and help redirect others who are facing
redundancy. Many of these tools are the same as
those used in initial career guidance for school
leavers – interest inventories, measures of values,
card sorts, etc. – whereas others specifically
consider any blockages that someone may be
experiencing in making a career change. Outcomes
enable the career counsellor to work more
effectively with the client in formulating
constructive future plans.
Personality inventories are also helpful in this
scenario – they identify those aspects of personal
style with which an individual is comfortable as
well as those that may cause them distress.

Exploring career paths that accommodate these
preferences is a positive outcome of the process.
In the area of succession planning, tests that
assess abilities and skills required in jobs at a higher
level in the organisation are popular in assisting
managers plan for advancement of their staff.
Sensitivity is required in the management of this
process so that employees see it as constructive.
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 3
Section I: Test Selection and Administration
4
Ability and Aptitude Tests
Ability tests involve questions that require complex
sets of mental processes and are designed to test a
candidate’s natural ability in a particular area. Often
ability tests explore relationships between two or
more words, numbers or pictures and ask the
candidate to extend a pattern or make an assertion
based on an understanding of the relationship.
Aptitude tests are similar to ability tests, but are
designed to give an indication of a candidate’s likely
successful future performance on the attribute that
is being assessed.
In selecting personnel, ability and aptitude tests
are widely used as good general indicators of
someone’s potential to perform the duties of the job
to a satisfactory standard, and to demonstrate an
ability to apply knowledge gained in new situations.
Non-verbal tests are also often used when selecting
staff for positions that demand skills not directly

related to formal education outcomes.
Example 1 on page 5 is an example from a
numerical ability test.
Ability and Aptitude Test Types
The main kinds of ability and aptitude tests are
listed in Table 2 on page 5, along with the common
selection criteria relevant to each test type.
Achievement Tests
Achievement tests are designed to measure what the
individual has learned in the past. Many educational
tests are designed as achievement tests.
Employers can also use achievement tests for
promotional activities within their organisation.
For example, insurance assessors may have to
demonstrate that they have learned risk categories
and appropriate application of policy levels before
being considered eligible for their next promotion.
Example 2 on page 5 is from an educational
achievement test.
Personality Assessments
Personality assessments are designed to provide
information about the way a person typically
behaves in certain situations, their preferences and
personal styles, and how they see themselves and
others,
but care must be taken in their use. It would
be difficult, for example, to argue that only one
personality type may successfully fulfil the
requirements of a particular job. There may be jobs
where particular personality profiles are more or

less desirable. An applicant for the police force
whose personality profile indicated an aggressive
component could be considered highly unsuitable;
positions involving ‘cold calling’ in sales often
attract extroverted personalities.
tests
Types of
Tests can be classified in a number of ways. One option would be to classify them according to what they
are assessing. For example, tests may be assessing optimum performance (as in ability or aptitude tests) or
practical knowledge (as in achievement tests). Alternatively, they may be assessing emotional responses to
gain a picture of typical response patterns or to identify a person’s preferences, likes and dislikes.
The following are some of the major categories of HR tests.
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 4
Table 2: Ability and Aptitude Test Types and Selection Criteria
Ability/Aptitude Type Example Selection Criteria
Numerical reasoning
Numerical estimation
Numerical checking
Verbal reasoning
Verbal checking
Abstract/non-verbal
reasoning
Spatial-visual
reasoning
Mechanical reasoning
ACER Select – numerical
ACER Test of Employment
Entry Mathematics
MOST Numerical estimation
MOST Numerical checking

ACER Applied Reading Test
ACER Select – verbal
MOST Verbal checking
APTS – abstract
SPM
APTS – spatial-visual
ACER Mechanical Reasoning
Tes t
General ability (numerical intelligence)
Ability to work with numbers and calculations
Ability to use, understand and explain information with
numerical content
Attention to detail with numerical elements
Understanding verbal content
General ability (verbal intelligence)
Attention to detail with written materials
General ability (non-verbal intelligence)
General problem-solving skills
Conceptual and planning abilities
Useful for measuring general ability where language may be a barrier
Ability to understand and work with visual representations of the
real world, e.g. maps, designs, plans
Understanding basic laws of physics and mechanics and their
application to the real world
A Quick Guide to Human Resource Testing
5
Personality assessments can also help to decide the
most appropriate management style for a candidate
and the way they are most likely to contribute to an
existing team.

Because of the sensitivity involved in interpreting
personality assessment results and in providing
professional feedback, these assessments are usually
available only to psychologists or people who have
successfully completed prescribed accreditation
training. Accreditation training is usually available to
HR practitioners and other professionals on an
instrument-by-instrument basis. If a psychologist is
not available to assist, a structured interview and
referee comments may provide helpful sources of
information on a candidate’s personal style.
Example 3 is from a personality inventory.
Interest Inventories
Candidates are sometimes asked to complete
vocational interest inventories to assist in placing
them in the most appropriate job. These inventories
are most effective when the person answers honestly
to give the most accurate picture of themselves.
These assessments are very helpful in career
management programs, working with people facing
redundancy or those voluntarily changing career
direction. Example 4 is from a career interest
inventory.
4
3
2
1
I’d rather go to a party than read a book.
Strongly Agree 1 2 3 4 5 Strongly Disagree
Find the two missing numbers in the following sequence.

1 3 ■■ 7 ■■ 11
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 5
Test questions can be categorised by the type of response required from the candidate and by the type of
content presented in the test. Both types are discussed below.
Section I: Test Selection and Administration
6
questions
Types of test
Response Type
HR questions most commonly require either
multiple-choice responses, range-type responses
(for example, ‘Strongly agree’ to ‘Strongly disagree’)
or open-ended responses.
• For multiple-choice questions, candidates are
usually required to select the best answer from
two or more possible answers provided. See
examples 1, 2, 3, 6 and 7 on page 7.
• Range-type responses (sometimes called ‘Likert
Scales’) are often used to indicate preference or
strength of feelings. See example 8 on page 7.
• Open-ended responses require the candidate to
write an answer in a blank area. This can be a
single number or word, or a longer written
response. See examples 4 and 5 on page 7.
Not all test questions are actual questions. For
example, some ‘questions’ are simply a statement
to which the candidate is asked to respond; for
example, they may be asked to respond to the
statement ‘I prefer dogs to cats’. For this reason,
test questions are often referred to as items rather

than questions.
Content Type
Most HR tests present test questions either in
written format or in pictorial form. Written test
questions can either use words (verbal) or numbers
(numerical); pictorial test questions can use a
variety of pictures, diagrams, mazes, maps and
visual puzzles. Examples of the different options for
test content are listed below.
Numerical (or Quantitative)
Numerical test questions require the use of
numbers and numerical symbols and concepts.
They can be used in a variety of test types, including
tests of ability, achievement and aptitude. Some
numerical items require calculations, some require
pattern recognition, while others require the
candidate to check for errors.
Examples 4 and 5 on page 7 are numerical items.
Verbal (or Linguistic)
Verbal test questions are based on words and
textual information. They can be used in a variety of
test types, including tests of ability, achievement,
aptitude and personality. Some verbal items require
word knowledge or logical reasoning, while others
require reading comprehension ability. Others are
simply written statements to which the candidate is
asked to respond by indicating their level of
agreement or disagreement.
Examples 3, 6 and 8 on page 7 are verbal items.
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 6

Non-verbal/Abstract
Non-verbal or abstract test questions usually
contain a series of shapes as the stem or base of the
question. The candidate needs to select another
shape or pattern from a selection of possible
answers in order to continue or complete the series.
Example 2 opposite is a non-verbal item.
Spatial-visual and Mechanical
Spatial-visual and mechanical test questions
require a candidate to look at a visual
representation of a physical object, such as a piece
of equipment, a shape or a geographical map, and
follow some instructions that involve manipulating
the object through space. Some examples are given
below:
• A candidate is shown a complex shape and asked
what it would look like if it were rotated 180
degrees and flipped over.
• A candidate is shown a piece of simple
machinery and asked what direction one part
would move if another, connected, part were
moved down.
• A candidate is given a map with a bird’s eye view
of a landscape and asked questions about what
the landscape would look like from the ground
looking north.
Example 1 opposite is a spatial-visual item.
Example 7 is a mechanical item.
A Quick Guide to Human Resource Testing
7

1
2
3
4
5
6
7
8
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 7
Tests vary in the way they are presented, although the basic elements of questions, answers, scoring and
interpretation/reporting are always present.
The test manual is the primary source of information about the test and often contains administration
guides, score keys, and tables for interpreting data.
formats
Test
Section I: Test Selection and Administration
8
Questions
Questions are usually presented in a test, item or
question booklet. Sometimes these test booklets
also include space for candidates to record their
answers; sometimes the answers are recorded on a
separate answer sheet. When answers are recorded
separately, it means the test booklets can be reused
by the next candidate. The administrator should
ensure that such reusable test booklets have not
been marked or damaged in any way and that all
are collected at the end of each test session.
Often questions are presented in a multiple-
choice format. In this format there is usually a stem

or base question and a number of alternative
answers. The candidate is instructed to choose the
best answer from the alternatives offered.
Sometimes the candidate needs to select two
answers, which together form the correct response.
The answers that are incorrect are known as
‘distractors’. Sometimes distractors provide a correct
answer in one sense, but not the best answer.
Answers
There are many different kinds of answer sheets.
Sometimes the questions and answers are recorded
in the same booklet; sometimes the answers have to
be recorded on a separate answer sheet.
If candidates are asked to record their answers
on a separate sheet, they need to locate the correct
place to record a response (usually by matching the
question number in the test booklet with a number
on the answer sheet). It will also be necessary for
the candidate to record their own details on the
answer sheet.
Often answer sheets are in the form of OMRs
(optical mark recognition) – that is, sheets that are
designed so that they can be read by a computer
scanner. This is particularly useful for large-scale
screening programs. OMR sheets can also be scored
by hand if computer scoring is not required. OMR
answer sheets need to be marked carefully by
candidates so that the computer scanner can read
the responses correctly. Detailed instructions on
how to mark the answer sheets are always included.

Another type of answer sheet is the carbonised
sheet. Once the candidate has finished recording
their answers, the examiner removes the top layer
and the carbon copy of the answers is designed in
such a way as to facilitate simple and immediate
hand scoring.
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 8
A Quick Guide to Human Resource Testing
9
Score Keys
A score key provides the examiner with information
needed to score a candidate’s responses.
Sometimes the examiner needs to count the
number of correct and incorrect answers to obtain
a raw score, in which case a list of correct answers
will be provided. Sometimes each answer is given a
different value and the values need to be added to
obtain the raw score.
Score keys come in a variety of formats:
• in the test manual
• as a separate card or clear plastic overlay
• on the carbonised section of the answer sheet.
More frequently nowadays, score keys are part of a
software system into which the examiner transfers
the candidate’s responses. These systems
automatically score the responses and provide a
report on the candidate’s results.
Interpretation/Reporting
Once the test has been scored, the raw score needs
to be transformed into a standardised score (see

page 18). This is done using tables provided in the
manual, or by the computerised scoring systems.
In the case of ability tests, the standardised score
is usually all that is reported on the candidate. For
personality and work style tests, more complex
reports are often prepared by the examiner or
generated by computer software. These reports will
provide a variety of information and comments to
aid the examiner’s interpretation of the candidate’s
personal style. Figure 1 opposite is an example of a
computer-generated report.
Computerised Testing
Increasingly, tests are available for delivery via the
Internet or otherwise on-screen on a computer. The
candidate sits at the computer and accesses the test
material by using a unique password that has been
allotted to that candidate by the potential employer
at an earlier stage of the selection process.
The software system for computerised testing
usually includes scoring, interpretation and
reporting.
The Manual
The Test Manual (also called a User’s Guide)
contains information on the development of the
test, its purpose, the target audience, precise
administration instructions, conversion of raw
scores to standard scores, and sometimes case
studies and other information to assist
interpretation.
The group(s) of candidates used to obtain a

standard score comparison is also described so that
the examiner may select the most appropriate
comparison sector. See page 18 for more
information about standardised scores.
Personality Interpretive Report Jon Sample
(continued…) 29 February, 2004
Anxiety
According to his responses, Jon Sample is no more or less anxious than most people. He has a
tendency to trust people and therefore may not be as vigilant as others in examining people’s
motives.
Factor Sten 1 2 3 4 5 6 7 8 9 10
Anxiety – general
Stable
Trusting
Assured
Tense
Figure 1: Example of a Computer-generated Report
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 9
Selection Criteria
It is critical that any test can be clearly demonstrated
to relate to one or more of the selection criteria.
This will ensure that the HR practitioner receives
more relevant information from the test session
and that the candidate appreciates the relevance of
that test session to the final hiring decision. As well,
there are important ethical and equal employment
opportunity (EEO) considerations, which require
the tests to relate directly to the selection criteria of
a job. (See page 29 for a list of example selection
criteria and appropriate tests.)

Usually, tests will be chosen from one or more of
the broad groupings below in order to accommodate
the needs of the particular selection exercise.
Verbal vs. Numerical
If a job requires good verbal and written
communication but no real involvement with
numerical work, then tests dealing with reading
comprehension or verbal reasoning may be chosen.
On the other hand, many jobs involve regular work
with numerical calculations but little or no verbal
or written communication. In this case it would be
appropriate to use a numerical assessment and not
a verbal one.
Technicala
In technical fields, skills such as spatial-visual or
mechanical reasoning are often relevant. There are
assessment tools that address these areas.
Decision Making and Problem Solvinga
Many jobs require different degrees of decision-
making and problem-solving skills. A range of
instruments is designed specifically to address
these areas. As well, tests of abstract or non-verbal
reasoning are considered excellent measures of
problem-solving and conceptual-thinking abilities.
Personality and Interest
Personality attributes are another consideration in
making a good selection decision. For example,
does the candidate need to be able to work
effectively in a team environment? Are strong
interpersonal skills critical to successful

performance in the job? Is a person who is very
change-oriented required? These and many other
characteristics may be assessed using a number of
tools including personality assessments and
interest inventories.
The table in Appendix III lists some common
selection criteria and examples of the types of tests
that might be used.
There are two main job-related issues that you need to consider in choosing the right test or tests for your
selection exercise. Firstly, you should consider the selection criteria for the job and secondly, you should
consider the general level of the job.
the job
Choosing the most appropriate test–
Section I: Test Selection and Administration
10
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 10
Job Level
Some basic and entry-level jobs require no formal
education or training, although they may require
on-the-job training. Factory assembly line or shop-
floor roles are in this category. Tests at a lower level
that assess general reading ability, numerical
checking, and speed and accuracy are most
relevant here. Scenario 1 below illustrates a typical
example.
Another class of job involves making decisions
based on the understanding of written and/or
verbal communication. For example, a customer
service officer is required to listen to concerns and
make decisions for future action based on his/her

interpretation of that information. A test that
assesses verbal reasoning at a medium difficulty
level will provide useful data in this selection
decision. Similarly, a team leader in this
environment is required to be comfortable
l
working with figures – estimating stock
requirements, monitoring flow of goods in and out.
Numerical reasoning tests are available that
address assessment at this level. A personality
attribute such as the desire to help others is also a
desirable characteristic in customer service.
Scenario 2 illustrates this.
Management roles (or graduate recruitment for
positions that progress through to management
level) typically demand that the successful
occupant demonstrates sound conceptual and
planning skills. Tests of abstract reasoning are
excellent tools for assessing candidates’ abilities in
this area. When combined with high-level tests of
verbal and numerical reasoning, they provide a
strong base for collecting relevant data on the
capability and capacity of job applicants. Scenario 3
is a typical case.
A warehouse needs to recruit staff
to work in despatch of orders. An
ability to read and understand
fairly routine messages and
accuracy in marking orders against
picking tickets are required.

Candidates are unlikely to have sat
for any tests since leaving school,
and less than 1 hour is available
for this part of the selection
process.
Test selection
An applied reading test is
recommended for the verbal
comprehension component of the
job.This test is used in technical
trade environments where there
is a need to read and understand
a limited range of materials,
such as union and OH&S notices
and company requirements. A
numerical checking test, used to
determine speed and accuracy
when reading numbers, is
recommended for the other major
requirement of the job.
These tests will take around
45 minutes in total to administer.
A team leader in a customer service
environment is being recruited.
Sound communication skills,
together with the ability to monitor
sales figures and report concerns to
management, are required. A desire
to help customers is essential.
Test selection

A verbal reasoning test at a
medium difficulty level, such as
APTS, addresses the first criterion.
A numerical awareness test, which
assesses the ability to do
calculations and detect
discrepancies, will measure that
component of the job. A profile,
such as the Work Aspect Preference
Scale, provides information that
would assist with the assessment of
personal qualities.
Testing time will be around 1 hour.
A travel agency is recruiting a
manager for a busy suburban
location. Sound skills in written and
verbal communication, together
with the ability to manage budgets
and develop marketing strategies,
are required.
Test selection
Verbal and numerical reasoning
tests appropriate for a junior
middle-management level are
recommended for the first two
criteria listed. An abstract
reasoning test will be a suitable
means of assessing a candidate’s
conceptual and planning abilities
required for developing new

strategies.
Tests will take less than
2 hours to complete.
A Quick Guide to Human Resource Testing
11
Scenario 1
Scenario 2
Scenario 3
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 11
Time
Tests vary considerably in their administration
time. The test publisher or distributor typically
provides brief descriptions of their assessment
tools in their catalogues, which include information
on the purpose, cost and time of administering
each test. Obviously, if you have just one hour
available for testing, you cannot choose a test that
takes 50 minutes but addresses only one of the
three criteria you wish to assess. The numerical test
described in Scenario 2 on page 11 takes only
8 minutes to complete, whereas those for the
management role in Scenario 3 take 30 minutes
each.
Getting the best mix of tests that covers the
broadest range of assessable criteria within the
time available takes some thought and planning.
Budget
The cost of a poor staff selection decision will
usually far outweigh the extra money spent in
purchasing a more appropriate recruitment test.

Given this, it is important to consider a range of
issues relating to minimising costs. For example,
tests are presented in a variety of formats (see page
8). If two different tests are designed to assess the
same factors, choosing the test with reusable test
booklets will reduce costs in the long run if you are
planning to do more recruitment for similar
positions. Some tests have reports that can be
produced by hand, while others require
computerised reporting. Usually the costs
associated with computerised reporting (either
purchasing reporting software or paying the test
publisher for a computerised report) are higher,
so if the presentation of reports is not a major
issue, choosing a test with hand reporting may
save money.
There are many practical matters that influence the selection of appropriate test instruments. The main
issues include:
• available time
• the budget for purchasing test instruments
• hand scoring vs. computer scoring
• qualifications required for purchasing and using test instruments.
the practicalities
Choosing the most appropriate test–
Section I: Test Selection and Administration
12
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 12
Hand Scoring vs.
Computer Scoring
Some tests have separate answer sheets that can be

scanned and scored directly by computer (as well as
by hand). These answer sheets are called OMRs (see
also page 8). If the testing program involves a large
number of candidates – for example, a recruitment
screening exercise – it may be appropriate to have
the answer sheets computer scored. This is a service
offered by the test publisher or distributor. Other
tests that do not have OMR answer sheets can also
be scored by the test publisher or distributor on a
fee-for-service basis.
If the testing program involves a small number
of candidates, hand scoring is usually faster and
cheaper.
Qualifications
Table 3 above describes typical levels of
qualification required for test administrators.
Professional development programs are available
that accredit new users to basic levels, and qualify
more experienced users to interpret specific tests.
In most cases, qualified psychologists may use any
published test instrument, although a few products
require specialised training even for psychologists.
A Quick Guide to Human Resource Testing
13
Table 3:Test Levels and Administrators
Test Level Typical Tests Test Administrators
High
Medium
Low
• Personality instruments

• Individual psychological tests
• Individual intelligence tests
• Ability tests
• Aptitude tests
• Achievement tests
• Interest inventories
• Card sorts
• Career instruments
• Registered psychologist or a student under the supervision of
a registered psychologist
• Professionals with experience and post-graduate training in test
use and interpretation who have attended specialist training in
a particular instrument
• Professionals with some tertiary training in human resource
management, psychology, statistics or other relevant discipline
• Professionals with some experience in test use and interpretation
who have attended specialist training
• Professionals with some experience and training in test use
and interpretation
Test
Administration
Course
This one day workshop provides an
introduction to:
→ Psychometric tests
→ The test administration process:
structure & rationale
→ Test administration practice
→ Scoring tests
→ The tasks of the test administrator

→ Basic ethics in testing.
Outcomes for participants will be:
→ A thorough grounding in the principles of test
administration and scoring
→ The opportunity to learn and practise the
administration and scoring of ability and
personality tests
→ The skills to assist qualified test users in
administering and scoring tests, so as to free
their time for interpretation and decision making
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 13
Environment
Let us consider the optimal environment
for conducting a testing session. The room should
be large enough to accommodate comfortably
the anticipated number of candidates. Give
consideration to the ventilation, lighting and
expected external noise levels. A candidate will not
give their best performance if the room is too hot or
cold, too crowded or noisy, or if it is too hard to hear
your instructions or see you or your assistant
administrators.
If you are conducting online testing, it is
essential that all computers are checked and
working well before candidates take their places.
Administrators
If you expect more than fifteen candidates, you
need an assistant to help with distributing and
collecting materials. Add another assistant for each
fifteen to twenty candidates beyond that number.

The logistics of managing an exam session
professionally are well and truly challenged when
you are trying to deal with an inappropriately
supervised group.
Getting Started
A general introductory chat is an important first
step. It serves the dual purpose of providing useful
general information as well as giving nervous
candidates the chance to settle for a few moments
before commencing the actual test(s). This does not
need to be a lengthy exercise. A few facts are useful
– for example, checking that the selection test is in
fact the one the candidates are expecting to do.
There have been occasions when perplexed
candidates have attempted a test for a position they
later found to be not of their choosing – their
session was in another room, on another floor or at
another time of the day! Applicants are often
understandably nervous, so clarifying information
to prevent such scenarios is important – and it
means you have not all wasted valuable time and
energy.
It is also useful to remind candidates where the
test session fits in the selection exercise. Is this an
initial screening? Who will have access to the
results? Where will the results be stored? When and
where may candidates request feedback on their
results? Candidates are well aware of privacy
legislation and their rights under this law.
Remember to let them know how long the test

session should take, whether there will be any
breaks, where toilets are located, etc.
To gain the most useful information from the test session while being completely fair to all candidates, you
need to follow strictly the guidelines set down for the administration of the tests you are using. The Test
Manual or User’s Guide, which is an essential companion to any psychometric test, will contain a section
detailing instructions for the proper administration of the test you are using.
The following are important aspects to consider in setting up a test session.
best practice
Test administration–
Section I: Test Selection and Administration
14
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 14
It is wise to check if everyone can see and hear
you. Does anyone need reading glasses or need to
move to the front of the room because of a hearing
impairment? Is anyone feeling unwell? Again, the
aim is to tap into their best performance, so it may
occasionally be necessary to reschedule a
candidate.
Mobile phones are a distraction. Remind all
candidates that all phones need to be switched off
as a courtesy to everyone.
Using Aids
Many tests do not permit the use of calculators or
such aids during the test session. Again, candidates
should be reminded of this.
The Manual
Ensure that you deliver the formal instructions
exactly as they are printed.
When giving any test from the ability/aptitude

range, it is essential that the administration
instructions in the manual be followed scrupulously
.
A general description will be provided, followed by
a script for introducing the test, giving practice
examples and working through those answers,
starting candidates on the test itself, and finishing
strictly after the allotted time period.
While these rules may sound pedantic, the
whole purpose of a standardised test session is that
candidates’ performance may reasonably be
compared across different test venues, different
administrators and different times. If, for example,
one supervisor is casual about the time frame,
applicants in that session may have an extra one or
two minutes on a 10-minute test. Arguably, this
would give them an unfair advantage over
candidates who are tested according to the
instructions.
There are, of course, some ability tests that are
untimed, and the timing rule is irrelevant in those
situations. However, the general administration
instructions and practice items must still be
followed exactly.
Personality tests usually have no time limit.
Administrators are advised to suggest that
candidates mark the first option that comes to
mind and not to spend too long on any one item.
For practical reasons, it is sensible to schedule
untimed tests at the end of an assessment session

so that people may leave as soon as they have
completed all components of the testing session.
Collecting Materials
It is essential that all materials are collected and
accounted for before candidates leave the test
room. Copyright legislation prohibits the copying of
materials and, to maintain confidence that test
integrity is being preserved, materials need to be
counted in and counted out.
A Quick Guide to Human Resource Testing
15
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 15
Normal Population
The natural world includes many examples of the
so-called ‘normal population’, characteristics of
which are often described by the bell-shaped curve
commonly called the normal curve (see Figure 2
opposite). In a normal population, most of the
people are closer to the average measurement of a
given attribute than to the extremes. For example,
there are many more people who are about average
height than there are extremely tall or extremely
short people. This means that if you measure some
characteristic that is normally distributed, most of
the population will ‘bunch up’ around the middle.
This produces the distinctive bell-shape of the curve.
If you went into a busy suburban street and
measured the height of the first 100 adult males
who passed by (these comprise your ‘sample’
population) and plotted the frequency that each

height occurred, the resulting graph would look
something like that in Figure 2. There is a bunching
up of heights between 165cm and 185cm; far fewer
lie towards each extreme – not many men in this
sample were below 160cm or above 190cm in
height. This is really just commonsense: we all
know that while everyone is a different height, most
adult people are really about the same height, give
or take 20cm. Very few people are extremely tall or
extremely short.
The same is true of intelligence and other
h
uman characteristics that are normally dist
ributed.
Once we understand this concept of the normal
population, we can begin to describe its
characteristics. Three ways that we can do this are
detailing where the centre of the population is (that
is, the mean), what the expected range of results is,
and how quickly (in terms of the unit of
measurement, such as centimetres or IQ points) we
deviate from the centre to the extremes.
Mean
The most commonly used description of the centre
of a population is the mean, or average score. This
is known as a measure of central tendency.
To calculate the mean, all scores gained by all
candidates on a test are summed. The total is
divided by the number of candidates. The resulting
figure is the mean score.

This figure serves as a benchmark against which
other scores may be measured; for example, ‘his
score is well above the mean’, ‘she obtained the
same score as the mean’, etc. This is the most
frequently reported central tendency score cited in
manuals for selection tests.
To interpret test results properly, an understanding of some of the basic terms used in statistics is required.
It is important to be aware of their relevance to help you make sense of test results and decide whether the
data reported is sufficiently valid and reliable for your HR testing exercise.
In this section we will cover four basic statistical concepts:
• the normal population • the range
• the mean • the standard deviation.
the basics
Statistics 1
Section II: Test Interpretation
16
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 16
Range
It is also helpful to know how far candidates’ scores
are dispersed from the centre. Probably the most
commonly used term is ‘range’, which is the
distance between the lowest score obtained by a
candidate on a test to the highest score gained. For
example, on a test of numerical reasoning, the
mean score may be 68, but the range of scores may
be from 12 to 99. You will notice that the mean is not
the ‘middle’ of the range but the average of all
scores obtained. If the mean is much higher than
the middle of the range, it means that more people
are obtaining high scores on the test than people

scoring very low.
Standard Deviation
Once we know the mean and the range, it is useful
to understand how quickly the population moves
away from the mean towards the extremes of the
range (or the ‘spread’ of the scores). One way of
measuring the ‘spread’ of the population is the
‘standard deviation’.
The standard deviation of a normal population
is derived from a mathematical equation, so that
68.2% of the population falls within one standard
deviation higher or lower than the mean, 95.4% of
the population falls within two standard deviations
from the mean, and virtually all the population
(99.6%) falls within three standard deviations of the
mean.
To go back to our height example, if the mean
height was 175cm and the standard deviation was
calculated to be 10cm, then 68% of the population
would fall between the heights of 165cm and
185cm, and 99.8% of the population between the
height of 145cm and 205cm (that is, three standard
deviations either side of the mean). If this sample of
the population was considered to be representative
of the whole male population, it would mean that
only 0.2% (that is, two people in a thousand) would
be taller than 205cm or shorter than 145cm.
If the standard deviation is very small, it means
that scores a small distance from the mean could be
considered extreme or unusual scores, while if the

standard deviation is very large, a score would have
to fall very far from the mean to be extreme. For
example, if in our height example, the standard
deviation was only 1cm, then a person 3cm taller
than the mean (that is, three standard deviations
from the mean) would be in the tallest 0.1% of the
population. If the standard deviation was 20cm, a
person would have to be 60cm taller than the mean
to be in the top 0.1% of the population.
We all have a good understanding of height and
how tall or short someone has to be before they
would be considered taller or shorter than usual.
As each test has its own way of measuring the
underlying attribute or quality that is being
assessed, we need statistics to fully understand the
meaning of any test score compared with the whole
population. Understanding the centre, the range
and the spread of scores is an important first step.
A Quick Guide to Human Resource Testing
17
Figure 2: Normal Curve
mean
height
Number of cases
2.1% 2.1%13.6% 13.6%34.1% 34.1%
standard deviations –3σ –2σ –1σ +1σ +2σ +3σ
height (cm) 145 155 165 175 185 195 205
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 17
Changing a Raw Score to a
Standardised Score

As well as mean, range and standard deviation (see
pages 16 and 17), test manuals usually contain
‘norm tables’, which report standardised scores
such as percentile ranks, stanines, sten scores and
t-scores. Standardised scores are often also referred
to as norm scores, norm-referenced scores and
derived scores.
To obtain one of these standardised scores, the
number of correct responses scored by a candidate
(the raw score) is looked up in a table in order to
locate the corresponding standardised score. This
standardised score allows us to compare each
candidate’s score with the sample population. This
helps us to understand whether 11 is a high score,
an average score or a low score.
Percentile Ranks
Percentile ranks are one of the most commonly
used standardised scores. The following are
examples of how this ranking works.
• Someone with a raw score that converts to a
percentile rank of 50% has scored the mean or
average score. They are right in the centre of the
comparison population; 50% of the sample
population has a higher score than they obtained
and 50% has lower.
• Someone who has a percentile rank of 80% has
scored higher than 80% of the sample population.
• Someone who has a percentile rank of 15% has
scored higher than only 15% of the sample
population.

Percentile ranks are very useful for ranking
candidates in order of merit (especially with ability,
aptitude or achievement tests), and for simple
explanations of where a candidate’s score lies in
relation to the rest of the sample population.
Stens and Stanines
Sten (‘Standard Ten’) scores and stanine (‘Standard
Nine’) scores are other ways of comparing a
candidate’s performance with the whole
population. In both sten and stanine scores there
are numbered categories that cover the whole
population – stens have ten categories: 1 to 10;
stanines have nine categories: 1 to 9. A score of 9 or
10 indicates a very high level relative to the
reference group, while a score of 1 indicates a very
low relative level.
What does it mean if a person obtains a score of 11 correct responses on a test? If there are 11 items, that’s
pretty good! If there are 100 items, perhaps it is not so great.
In order to understand the meaning of a score of 11, the developers of test instruments provide users
with a standard against which they can measure the merits of a candidate’s performance. This puts them in
a much better position to make a fair comparison. The standard is based on a set benchmark that was
determined by assessing a comparison group, or sample, that is selected to represent the population. This
sample is often called the norm group.
test scores
Statistics 2
Section II: Test Interpretation
18
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 18
Raw Score Percentile Stanine T-Score Raw Score
Rank

32 99 9 68 32
31 97 9 66 31
30 94 8 64 30
29 90 8 62 29
28 84 7 60 28
27 77 7 58 27
26 69 6 56 26
25 60 6 54 25
24 53 5 53 24
23 48 5 51 23
22 42 5 49 22
Figure 3: Sample Norm Table
Sten and stanine scores are useful when wanting
to avoid over-emphasising small, unimportant
differences between candidates.
T-scores
T-scores are another common type of standardised
score. A t-score of 50 is the mean score, or 50th
percentile. With t-scores, each 10 points indicate a
standard deviation. Therefore, a t-score of 60 is one
standard deviation above the mean, while a t-score
of 40 is one standard deviation below the mean. In
the same way, a t-score of 70 is two standard
deviations above the mean.
T-scores are a popular method of reporting
candidates’ scores on psychological test
instruments.
The differences between these (and other)
standardised scores are interesting but quite
technical. The important thing to remember in the

use of HR tests is that standardised scores, not raw
scores, should always be used in reporting, so that
the candidates’ scores can be properly compared to
the reference population.
Remember – what does a score of 11 correct
answers really mean?
Figure 3 shows an example of a norm table with
raw scores and three different standardised scores.
The HR practitioner can use whichever of these
standard scores best suits the situation. Often norm
tables only provide one kind of standard score.
Figure 4 is a single graph showing the
relationship between the normal population,
mean, standard deviation and some commonly
used standardised scores.
A Quick Guide to Human Resource Testing
19
Figure 4: Normal Curve and Standard Scores
Number of cases
2.1% 2.1%13.6% 13.6%34.1% 34.1%
standard
deviations
mean
–3σ –2σ –1σ +1σ +2σ +3σ
50 55
20 25 30 35 40 45 60 65 70 75 80
0.1 1 2 7 16 31 50 69 84 93 98 99 99.9
t-score
percentile
rank

56 78 9 10
1234
56 78 9
1234
sten
stanine
Numbers inside the curve
represent the percentage
of the population in that
section of the curve.
Quick guide to HR testing.qxd 1/4/05 10:09 AM Page 19