Using Psychological Screening to Select Top Performers for Firefighting

© Paschal Baute, Ed. D., FACFE, DABPS

  Executive Summary This article will report on a seven step assessment design process used for a Southeastern mid-sized metro government for the Fire Department, since 1982. Nineteen factors in a battery of three instruments have proved predictive of above average performance. Ongoing means for development of norms in firefighter recruit screening with small populations is discussed. Comments of two Fire Chiefs and the HR director are noted.

In the past, the most important question to answer in selecting personnel is "How well will this person perform if we hire him or her?" The most common method used to answer this question traditionally has been the interview. Many studies have shown the interview to be biased and unreliable as a performance predictor. Organizations are increasingly using well-developed, high quality tests as part of their selection process to select top performers, reduce turnover, reduce recruitment and training costs, to help make fair hiring decisions, and to build productive, competitive workforces. Professional practice -and in some states local law-dictates that measures used to select recruits be validated.

Since 1990, both the Americans With Disabilities Act of 1990 (ADA) and the Civil Rights Act of 1991 (CRA) have impacted procedures used in preemployment screening. In addition the EEOC has clarified these requirements. The purpose of this article is to provide a description of the development of fire recruit testing program that may serve as a model for others who want to adopt an effective and standardized ADA/CRA/EEOC-compliant approach. Specifically, the process of psychological screening used for a fire Department by a Southeastern mid-sized metro city / country merged government is reported. Psychological screening began in 1982 and has undergone several revisions and studies of predictive validity. .

Assessment design has at least four primary determining variables: the job competencies required, the objectives of the organization or the department, the organizational culture and climate, and the pool of applicants. These four variables affect all components such as measures employed, e.g. paper and pencil psychological tests, structure of the screening, outcomes desired and obtained, and appropriate follow up. Assessment design may need also to address critical local HR issues: need for diversity, excessive turn-over, problem employees, legal costs, morale, avoidance of adverse impact on minority groups, and administrative effectiveness and efficiency. In order to be ADA/CRA/EEOC compliant, legally defensible, objective, and fair, professional standards must be observed to make sure that assessment design is appropriate for the particular position and the particular company or organization. Distinction must be made between essential and useful knowledge, skills, attitudes and other behaviors (KSAOs). The assessment design steps reported here were developed to follow these guidelines and to help the organization select applicants who were more likely to succeed. Insert Figure 1 about here.

Figure 1 illustrates the organizational variables: factors influencing assessment design, components, and local HR issues that impact the standards and safeguards by which the conditions and processes are weighed and resolved.

Even still one does not simply suggest a test or tests out of some bag of potential tests. Questions that need be addressed in the selection design process are: What does the organization require? What behaviors and competencies are most relevant, most essential? Does that behavior / competency meet the organizational requirements? What behaviors, KSAO's (Knowledge, Skills, Attitudes and Other characteristics) are most critical? What assessment methods are the most practical procedures, given the decision-making time frames and a cost versus benefits analysis? Have potential sources of bias and error been thoroughly reviewed? Are there cross-cultural issues to be resolved–according to ADA/CRA/EEOC guidelines? Can we demonstrate that the assessments used are actually predictive of performance? Insert Figure 2 about here. This shows the flow of the decision questions that determine assessment outcomes, from a cross-cultural, diversity point of view.

These questions illustrate the need for a thorough and professional decision making model for assessment. They illustrate the type of questions addressed in this selection design. It is the responsibility of the assessor, in this case, an organizational psychologist, to properly select, use, and interpret assessment instruments consistent with the organizational context, job requirements, and assessment purpose. In addition, an ethical responsibility is proposed for both for the assessor and the organization to provide and encourage feedback and research regarding assessment processes. Seven steps in assessment design are described in this report beginning in 1982 and continuing today.

Assessment Design #1, 1982-88. During the first six years following the beginning of psychological screening, a well validated personality test, the 16PF . was employed together with background inventories. The 16PF questionnaire has been widely used for selecting personnel into jobs for over 30 years. As a comprehensive measure of normal adult personality, the questionnaire has proved effective in predicting performance with concentrated use in selection of public-safety personnel such as police officers, corrections personnel and security personnel. Feedback from training officers indicated that the specific predictive remarks in the report of psychological screening proved accurate and helpful to the training supervisors.

Assessment Design #2, 1988. Further validation was begun in 1988 when the use of the 16PF was validated as predicting top performers on a number of scales. This also began a more thorough research into job performance as it related to screening criteria. The first validation study was conducted on 115 firefighters hired between 1982 and 1988. These individuals had all been screened using the 16PF Form A and performance evaluations were collected on the entire group.

The pre-hire primary scale scores found to be correlated with later performance and several significant validity coefficients were found. Specifically, better performers were more warm (A+), intelligent (B+), emotionally stable (C+), conscientious (G+), socially bold (H+), trusting, (L-), practical (M-), self-assured (O-), self-reliant (Q2+), exacting(Q3+), and relaxed (Q4-). This validation provided empirical evidence that the use of the 16PF was predictive for this early stage of psychological screening, 1982-1988.

Assessment Design #3, 1994. The American with Disabilities Act (ADA) effectively outlawed "medical" tests in pre-employment screening. Written before ADA, it was not perfectly clear that Form A of the 16PF used previously was appropriate under the ADA. Also, the 1991 Civil Rights Act outlawed within--group norms which may cause a problem in the use of some same sex norms with Form A of the 16PF questionnaire. There was a window of opportunity in which test publishers were seeking post-ADA verification of their instruments and in some cases revisions to eliminate bias. Some test vendors were willing to assist in the statistical evaluation of the results at reduced cost, which proved to be an incentive here for re-validation.

The Fire department furnished a select population of 20 low performers and 20 high performers for the study, called group 1 and group 2. They were assigned numbers so that no names were reported in the screening. Nine profiles including the 16PF were employed. Four of the low performers did not show up for the assigned time for profiling, leaving 20 for group 2 and 16 for group 1, for a total of 36 firefighters screened. Nine profiles were employed, but five proved non-predictive of differences in this small group. Four test profiles were demonstrated to yield significant differences between low and high performers in a total of 26 factors. The profiles were the 16PF (5th edition) the IS5 and the HPP/SQ both of Hilson Research, and the Behaviordyne scoring of the MMPI-2. Differences were also noted on the Wonderlic Basic Skills Test (WBST) and the Comprehensive Personality Profile, but there were not statistically significant. These results are now discussed.

For the 16PF results, the two groups differed on four primary scales. Higher performers were more emotionally stable (C+), utilitarian (I-), practical (M-), and traditional (Q I-). A regression analysis resulted in a slightly different perspective: more emotionally stable (C+), less dominant (E-), skeptical (L), traditional (Q I-), and exacting (Q3+). These significance levels are based on a p<. 10 level. Thus, this study replicated the general result that primaries of the 16PF were quite effective at distinguishing between effective and less effective firefighters. High performers were more tough-minded, resolute, unempathic (.001), more utilitarian, objective, unsentimental (.004), more grounded, practical solution oriented (.01), more emotionally stable, adaptive, mature (.03), more traditional, attached to the familiar (.06), and more accommodating, selfless, agreeable (.07), for total of six factors. That one profile can correctly predict six factors, several at a highly significant statistical difference is remarkable with such a small population total of 36 actual profiles, 20 high performers and 16 low performers.

Further, in this same 1994 validation study, three other profiles proved predictive. Inwald Survey 5 (IS5) scales that correctly predicted overall work performance included lack of Sensitivity (LS), Frustration/Anger Patterns (FR, low), Introverted Personality (IP, higher), Distrust of others (DI, low), and Work Effort Concerns (WE, above average), for a total of five factors. On these scales, group means for those rated as "Needs Much Improvements" were higher than those who were rated as "Exceptional/Average".

These results suggest that a successful firefighter may be more concerned than others about how his/her behavior is judged, may be less likely than others to express anger/hostility, may be more trusting of fellow workers, and may tend to strive for the completion of tasks or to "go the extra mile". Discriminate function equations correctly classified 83.3% of these firefighters on overall work performance.

Hilson Personality Profile / Success Quotient (HPP/SQ) scales used to correctly predict overall work performance included Achievement History (AH), Popularity (PO), Competitive Spirit (CO), Drive (DR), Goal Orientation (GO), and Anxiety About Organization (AX).xxx higher in all of these measures, for a total of five factors.

This suggests that a successful firefighter generally has a history of higher achievement in school and jobs, is well-liked by others, is not overly competitive, tends to work harder than his/her peers, tends to set personal goals, and is concerned about completing assigned tasks in a correct and timely manner. Discriminate function equations developed from HPP/SQ scales correctly classified 75% of the firefighters on overall work performance.

Moreover, when the IS5 and HPP/SQ were combined to predict work performance, discriminate function equations from both tests correctly classified 94.4% of the firefighters. In addition other scales used in the equation to correctly predict overall work performance suggest that firefighters who are successful on the job also tend to organize their work and meet responsibilities on time, and generally have fewer job adjustment difficulties. Therefore we could correctly identify the high performers early in the screening by using the combination HPP/SQ and IS5 or the 16 PF. As the result of Assessment Design (AD#3). the use of the Hilson Research profiles became standard. This battery was later expanded to the entire Hilson Research battery of six profiles developed specifically for occupational screening of candidates for all Public safety and Security positions, police, fire, corrections and other security.

Further differences were found in the Wonderlic Basic Skills Test (WBST) and the Comprehensive Personality Profile (CPP) between group 1 (low performers) and group 2 (above average performers). The WBST verbal, quantitative and composite scores of Group 2 are higher that those of Group 1. This indicates that higher levels of basic verbal and quantitative skills may be associated with higher levels of performance in the required duties of a firefighter. Also the results also show that the sample means CPP Assertiveness and Trust scores of Group 2 are lower that those of Group 1. This indicates that lower levels of assertiveness and trust may be associated with higher levels of performance in the required duties of firefighter. Further analysis involving larger samples of test and performance data would be necessary to identify what statistically significant relationships exist between job performance and WBST and CPP test scores.

Assessment Design #4. 1994-95. Whereas EEOC did not specifically require job analyses for specific positions, professional standard practice suggests distinguishing essential from important job requirements prior to selecting employment tests. In 1995, new job analyses were completed. These were the Hilson Job Analysis Questionnaire and the Managing for Success Work Environment version. These have served to confirm that our assessment profiles are measuring the "right stuff." As reported by a training supervisor and two chiefs, "essential" attitudes needed for firefighting were: Work Patterns, Admission of Shortcomings, Loyalty to the Organization, Lack of Procrastination, Conscientiousness and Frustration Tolerance. Rated as "Important" characteristics were: Self-confidence, Learning Ability, Social Awareness, Communication Skills and Trust. Insert Table 1, HJAQ, about here.

On the Managing for Success Work Environment version, the desired behavior of the firefighter is shown to be characterized on the Success Wheel as a Supportive Coordinator. This means that the performing firefighter will demonstrate steady behavior: systematic, deliberate, team player, patient, and a good listener. This job analyses is represented in figure 3 and figure 4. Table 2 Learning the DISC Language indicates clearly the differences in the four DISC styles. . Insert Figure 3, 4 and table 2 about here. Figure 4 together with Table 2 shows that the composite DISC profile of the successful firefighter S and C above the midline, and D and I slightly below the line is a graphic portrayal of the HJAQ Job Analysis. The DISC profile is Methodical, Reliable, Amiable, Steady, (S) Thoughtful, Analytic, (C) Friendly (I) and Decisive (D).

Assessment Design # 5, 1996. An indication of how feedback from training supervisors has been useful in developing the psychological screening process was found in the early 90's. It was found that too many were getting though the process who had difficulty coping with the math and reading part of the training manuals. We began by adding a problem solving measure of general mental ability, the Wonderlic Personnel Test (WPT) in the early ‘90s to the final screening, (See box C in Figure 3) Training supervisors feedback was that the WPT raised the bar and helped some, but too many still were being selected who did not have sufficient math skills. The Wonderlic test of Basic Skills (WBST) in math and verbal were added, but first only to the final psychological screening. Then it was discovered too many could complete the screening process, from pre-screening and through the interviews only to fail the WPT or the WBST at the end.

In 1995 the pre-screening qualifications testing were re-examined to eliminate the bias thought to be present because of number of the recruits came from a volunteer firefighter background. It was decided to use at this early stage both a general problem solving test, the Wonderlic Personnel Test and a measure of basic math and reading skills, the Wonderlic Basic Skills test as Admissions testing, or what is called Pre-screening on figure 3. Because of the excellent predictive validity of the 16PF it was decided to use the 16PF also in the pre-screening. Cut off scores on the WPT were set at 21, the national average of the high school graduate, and for the WBST, 12th grade level verbal skills and 10th grade level math skills. These two instruments alone, the WPT and the WBST eliminated about one half of the candidates from being accepted for the interviewing process, but resulted in a higher quality of competencies on the part of the recruits being interviewed.

Assessment Design #6. The new job analyses AD#4, confirmed the previous assessment decisions and indicated that the larger, more comprehensive battery of six instruments of Hilson Research (validated for public safety and security screening) was appropriate to use. These inventories included (with number of scales listed after each): IPI - 26, HSRI - 14, IS5 - 11, HPP - 16, IS2 - 10, and HLAP - 14, Several parallel assessments were tested but these did not prove predictive. Further, the scanned score results of the Hilson battery are sent via modem to the HR computer which scores all profiles against local norms that the sending agency previously has established. The HR computer holds the data bank of scores from the particular locality and compares each answer with the answers of the previous applicant groups entered (currently 389 total in the local norm group). We can immediately view the difference between this candidates answers with the norms previous established and stored on a total of 91 HR scales. We can target for careful exploration in the interview those whose score scales are outside one and two standard deviations of the means (statistical norms) of the groups of fire recruits who have already successfully passed.

The use of this particular program at Hilson Research provides a constant feedback loop and on-going "instant validation." Each group of successful firefighter recruits is added to the previous groups to provide a larger norming base for comparing the answers of each applicant of each new group to be tested. Overall the data base established for all profiles for both fire, police, and public school bus drivers (all public safety and security arenas) in this metro area since 1982 is about 1000. Currently this battery is also used with police screening in central Kentucky. New research indicates that there are no significant differences in the primaries on profiles measuring fire as compared with police candidates. Software menu allows the choice of comparing each applicants scores with the group local norms. This is one outcome of the ongoing feedback, research and monitoring supported by the Fire Department and the Human Resource Department of the metro government.

Assessment Design #7. Because it has been ten years since the last large validation study, a third study begun in 1997 is now underway that requires more data to be collected on the performance of firefighters. We are currently using the 16PF in pre-screening but not employing the results in selection so as to not bias the selection outcomes. Attendance, safety/accidents/ training success, turnover/tenure and other behavioral measures will be collected. Data collection is scheduled six months after they start active duty and will include a separate research performance evaluation. These behavioral indications of success will then be correlated with 16PF scores to further "prove" the validity of the 16PF questionnaire for us in screening firefighter recruits.

An overall graphic is presented here to explain how these assessment steps fit together and came together over a period of 17 years. If the previous steps are sufficiently understandable, the reader may wish to skip this section. Insert figure 5 about here, which is now explained by means of the color coding.

In assessment design #1, The green boxes pre-screening and Interviews by Fire staff indicate the selection process before the installment of psychological screening. The blue box at the top of Figure 1, Assessment Design Step #1 indicates the beginning of the implementation of psychological screening for this metro department. A profile validated in other settings, the 16PF, (step A) was selected for use (B, yellow arrow) in the final psychological screening (C, blue box). The original on-site validation of this profile was determined after training (D, blue scroll) by feedback from training supervisors (E. blue arrow, back to box blue block arrow box, A), that is, by a "goodness of fit" evaluation that specific recommendations concerning recruits were accurate. This was our first feedback loop illustrated by the blue and yellow colors at the top of Figure 4.

Assessment design #2 is shown in figure 1, beginning with Duty after training (tan, step F), Assessment Design #2, with the 16PF scales being validated by job performance ratings of supervisors (tan, step G), which affirmed the usefulness of the 16PF at the final psychological screening (C). This second feedback loop in the low right corner of figure 4, A.D. #2, by steps F, G and C.

Assessment design step #3 is represented by the orange box in the lower center and AD#3, and boxes F, H, and the arrow back to C in figure 1. The two groups were preselected and tested. Assessment design step #4 is represented on figure 5 by the rose colored upper left graphic, AD#4, step J, B, C. Step #5 is represented by C,D,F,I, and by the lime colored box arrow pointing back to the green pre-screening box in the lower left of figure 4.

Step #6 The expansion of the full Hilson Battery of six instruments is represented on Figure 5 by Assessment Design # 6, upper left rose multi-document graphic, step K, B, C, and the red arrow L, above "Interviews by Fire Staff," going back to K indicates the continuous loop of adding scores of each new group to the data base for future norming. It is important to note that the K, B, C, K loop is now continuous with each new group of applicants successfully accepted. Step #7 is represented in figure 5 by the red box arrow in the lower left, AD#7, step M.


Perhaps the most effective method of relating job analysis components and assessment research can be by means of a Matrix, identifying KSAO's as essential and which assessment is appropriate for measuring the KSAO. This is here defined in Table 3.

Author's Note: This table is being reconfigured for web display. In the meantime, please email author for hard copy by snail mail.

The critical KSAOs are listed as the rows, the available assessments are listed on the columns, and check marks are entered in the matrix to show effective measure of the KSAOs by the specific assessments. Note that some of the KSAOs can be predicted by more than one assessment; such redundancy is especially advisable when feasible. Selection of assessments is also driven by practicalities; in this case, an interview may be too costly or if the total testing time is too long then the personality test or background questionnaire might be dropped. For high-volume assessment systems, several assessments might be validated and then some dropped based on the validation results. This matrix can illustrate both the range of assessment strategies useful and the usefulness of a particular assessment tool. This completes our survey of our assessment design strategies.

Although ADA/CRA/EEOC does not require the establishment of local norms, professional practice in psychological assessment does, particularly practice that is forensically defensible. Applicants for public safety positions who are not recommended may appeal, and may obtain a lawyer to represent their appeal. To be able to say that a particular person does not meet nationally established norms for a particular profile or test used in job screening is one thing, but to be able to say that this candidate did not meet the norms established locally by those successful candidates over a period of some years in this particular setting is another very different thing. To think you are rated as wanting in psychological health is one thing. To be told that you did not fit the profile established in local research for successful job performance is another. The " forensically-aware" psychologist in pre-employment screening will go the extra step to make sure that everything that is done follows professional standards in the field, is legally defensible, and that the organization that he or she works for is protected from liability in possible forensic challenges. When the sponsoring organization is willing to sponsor a research orientation of the industrial/organizational psychologist, possible forensic challenges to hiring decisions are deterred, risk management is strengthened, and legal vulnerability is reduced.

In employment screening, it is important to note that use of clinical diagnosis, or measures of psycho-pathology while they are often used to exclude candidates, cannot determine occupational fitness. Screening-out profiles can reveal which are high risk candidates. Screening-in profiles are used to identify which candidates will be a "good fit" to the particular job demands. In addition, the 1992 Supreme Court Daubert decision and its Kumho decision progeny will make it harder to defend the use of psychological concepts and measures in the courtroom. (I have a paper in forensic psychology now being considered for a special fall issue in The Forensic Examiner.) Screening should be accomplished by using profiles and tests that have been developed specifically for occupational uses. Those developed in clinical settings are inappropriate for occupational use and make the organization more legally liable. The Daubert decision is changing the landscape of expert witness, and judges are given the power to determine what is good science and given many new guidelines. () Kentucky has adopted the Daubert guidelines. Therefore, it is now regarded as forensically unsafe to use instruments developed for clinical uses and clinical settings. These type of tests such as the MMPI, developed for clinical purposes, which have been commonly used for occupational screening, have not been employed in any of these assessment steps.

Comments by the current Fire Chief:

I feel that it is very important for any administration, especially those involved in public safety occupations, to be able to predict with reliability, job performance. Job performance is considered paramount as a measure of productivity in any work place and the predictability associated with these studies has been extremely valuable.

Comments by the most recent retired Chief:

While serving as Fire Chief from 1987 to 1997, the staff and I evaluated the firefighter recruitment process with each class of firefighters assigned to the academy, usually twice yearly.

I can state without fear of contradiction that the process designed and implemented by Dr. Baute was, and is, unusually effective in predicting which candidates will be successful. Specific comments provided regarding an individual candidate's strengths and weakness have been proven to be extremely accurate.

Perhaps the most important fact is that after each process Dr. Baute also evaluated the results and suggested changes, if necessary, to continually improve the outcome.

Comments by the former Director of Human Resources:

In mid 1982, a series of meetings was held with top level police and fire personnel, university police and fire science faculty and industrial psychologists from the private sector. As a result of these meetings, a decision was made to develop a valid selection process for police and fire that would predict which applicants would be successful police and/or fire recruits.

It was envisioned that the benefits of a valid selection process would be the hiring of individuals who were more suited for police and fire work, resulting in a higher level of service to the community and the eventual creation of a larger pool of well-trained officers available for promotion to higher ranks. The three major goals established for the selection process were as follows: 1. Conduct a valid job analysis to determine the frequency of the essential tasks performed by police and fire personnel and determine the necessary (job elements), knowledge, skills and abilities needed to perform the tasks. 2. Based on the results of the job analysis, determine the most appropriate testing profiles in order to predict those candidates who could perform the essential job elements. 3. Validate via a comparison of training test scores, performance scores and peer feedback that the selection instruments accurately predicted future successful employees. The above goals were accomplished by contracting with public university and private sector industrial psychologists and statisticians. As indicated in the report, the process yielded a very high correlation between predictability of performance and actual performance. Such has resulted in a more cost effective recruiting and retention process with high job satisfaction, less turnover, more quality personnel available for promotions, fewer employee relationship problems and, most of all, an increased quality of sustained professional service to the public.


Psychological screening began in 1982 in a mid-Atlantic metro government of a mid-sized city. Assessment design has followed a number of revisions. Current profiling provides clear evidence of skills, attitudes and other characteristics that are predictive of above average performers. A total of 17 factors on three profiles have been found to be predictive. Ongoing research is underway to re-validate one of the key instruments found to be useful in predicting successful performance. Ongoing validation is also obtained by the use of a vendor dedicated to research and providing instant scoring by computer with comparison with local norms established. Several job analysis profiles were established and shown. These were integrated into a KSAO Matrix showing how various kinds of assessments are useful in the screening process. Finally comments of the just retired and current Fire chiefs and those of the HRD Director who support the initial and later research 1982-1996 were noted. Further research is indicated and is ongoing.


American Psychological Association Task Force. (1991). Questionnaires Used in the Prediction of Trustworthiness in Pre-Employment Selection Decisions. Washington, D.C. Robin Inwald

Americans with Disabilities Act of 1990, 42 U.S.C. 1200-12113.

Asquith, Nancy and Feld, Daniel. Employment Testing Manual, is a compendium of up-to-date guidelines on changing laws, technologies, and practices that affect employment testing. 1994 Cumulative Supplement, 1993 Cumulative Supplement. Warren, Gorham and Lamont, ISBN 0-7913-1855-9

Baute, Paschal. "ADA compliance Use of Psychological Tests in Occupational Screening." Workshop given at the Kentucky Psychological Association annual convention, Louisville, Ky, August, 1995.

Baute, Paschal. "ADA / EEOC Update For Personnel Screening, Fall," 1994. PERSONNEL PROMPTINGS published privately for corporate clients, November 11, 1994,

Baute, Paschal. "How Does An Organization Decide Which Selection Tests Are Appropriate? PERSONAL PROMPTINGS, published privately for corporate clients, May, 1995

Baute, Paschal. "Is Your Expert Witnessing Prepared for a Daubert/Kumho Challenge?" in press, The Forensic Examiner.


Baute, Paschal. "Practical Solutions to Selection Pitfalls in Personality Testing." Practitioner Forum and panel. 14th Annual Conference of the Society for Industrial and Organizational Psychology, April 30-May 2, 1999, Atlanta, Ga.


Baute, Paschal. "The Role of the Psychologist in Expert Witnessing: Falsiability." The Forensic Examiner. May-June, 1999. 25-27.

Enforcement Guidance: Pre-employment Disability-Related Inquiries and Medical Examinations Under the Americans with Disabilities Act of 1990 (#915.002, 5-19-94).

Robert Ramsey. The Testing Manual. Ramsey Corporation, PIttsburg, Pa., 1990.

Principles for the Validation and Use of Personnel Selection Procedures. Society for Industrial and Organizational Psychology, Inc., 1987.

Standards for Educational and Psychological Testing. American Educational Research Association, American Psychological Association, National Council of Measurement in Education.

Buros Mental Measurement Yearbook series, is the standard reference for balanced information about specific instruments.

Cascio, Robert. Applied Psychology in Personnel Management (ISBN 0-13-041427-1) text emphasizes assessment in its discussion of all phases of human resources management.

Congress of the United States, Office of Technology Assessment. (1990) The Use of Integrity Tests for Pre-Employment Screening. Washington, D.C. Robin Inwald.

Enforcement Guidance: Pre-employment Disability-Related Inquiries and Medical Examinations Under the Americans with Disabilities Act of 1990 (#915.002, 5-19-94).

Fairness in Selecting Employees by Arvey & Faley (Addison-Wesley) is a good treatment of the legal context of personnel selection up to about 1988; another edition is rumored to be in the works.

Guion, Robert. Assessment, Measurement, and Prediction in Personnel Decisions, (Erlbaum 1998) is the most recent and comprehensive compendium for the field.

Inwald, Robin, (1989) The Hilson Research Model. A Guide for Conducting Interviews and Preparing Psychological Screening Reports. Paper presented at the American Psychological Association Convention, Ne Orleans, La.

Inwald, Robin and Jody Resko. "Preemployment Screening for Public Safety Personnel." Innovations in Clinical Practice, A Source Book. vol 14.

Jeanneret, Richard and Silzer, Robert. Individual Psychological Assessment. (Predicting Behavior in Organizational Settings). Jossey-Bass, 1998.

Practical Issues in Employment Testing by Rose (ISBN 0-911907-09-2) is an excellently written book focusing on the use of assessment in personnel selection including easy-to-understand discussions of basic and advanced topics, self tests, and numerous resources.

Principles for the Validation and Use of Personnel Selection Procedures. Society for Industrial and Organizational Psychology, Inc., 1987.

Ramsey, Robert. The Testing Manual. Ramsey Corporation, PITTSBURGH, Pa., 1990.