.6- .,'r '95, A ; l n ‘ h. 7- , .39 ”f“, D 252*. I . «a; in I b l’\"‘ wgiu‘u * 1 :- v, «r .‘ r; ‘ .L' Q "2‘32, . L a u": 3." o 4 h > a ‘ Y: 35‘ _ \§'fi$k.; . 7. :VUI‘H‘TJ: . .. ur . ' . aw : . , . *3 A ' , ' - ~ * SIM-3‘5": . ‘ ' 'r‘.“ ‘ . . -‘ "’ 1" 33»? 'H 5‘3 t": u ‘ .“ " ’ 4t“, ‘ - («2a - 4" 2.. f '%$‘ 1“” g; ‘ .9. .1? m ”.9 \ ifijvs~um NCJfi-m‘ 1». .wn-J m ’ «film- it ha #5. , ‘ {($3.26 ‘ ' «L ‘A --n...Iu in? and-{4% A": .v‘. WI 7 '2 “#553 ‘ ’3; 132;; . I}. .- k’fii‘ (a. v .‘4 u‘ ; 5 'r 1» . ,_. '{s‘sfv Q EVE 4.1.5575 .. . . H vs at; 'I ; :fifizafir . (“5’3“ '2 _' mix?" -. . . ‘x-r-nf - ‘1. W l c an ‘llllllllllllllll‘llllllllllll " 3 1293 01031 7794 This is to certify that the thesis entitled Usage, Practices and Policies of Polygraph ("Lie Detector" ) Screening in Police Agencies in the United States presented by Robert Thomas Meesig - y has been accepted towards fulfillment of the requirements for Master of Science Criminal Justice degree in (ajor professor Date June 101 1994 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State Unlverslty PLACE II’REI’URN Boxmmwomhchodtomfrunmncord. TO AVOID FINES "mm on or “or. dd. duo. I DATE DUE DATE DUE DATE DUE , flair—H.“ USAGE, PRACTICES AND POLICIES OF POLYGRAPH ("LIE DETECTOR") SCREENING IN POLICE AGENCIES IN THE UNITED STATES BY Robert Thomas Meesig A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE School of Criminal Justice 1994 ABSTRACT USAGE, PRACTICES AND POLICIES OF POLYGRAPH ("LIE DETECTOR") SCREENING IN POLICE AGENCIES IN THE UNITED STATES BY Robert Thomas Meesig The purpose of this study was to determine the extent of use and perceptions of pre-employment polygraph screening (PEPS) among police agencies in the United States. A questionnaire was mailed to administrators of 699 of the largest police agencies and a random sample of the remaining smaller agencies, and 2108 (73%) usable responses were received. PEPS was used by 62% of the large and 13% of the small agencies. These agencies employed 32% of all sworn officers in the United States. The primary reasons for using PEPS were that it reveals information not otherwise available and that it deters undesirable applicants. The main benefits were better and more honest employees. Agencies using PEPS were more likely to be located in the Western United States and in states with no laws regarding PEPS, and they tended to use more additional screening techniques than agencies that did not use PEPS. Copyright by ROBERT THOMAS MEES IG 1994 This research was carried out under the auspices of the Polygraph Research Center, School of Criminal Justice, Michigan State University, partially supported by the College of Social Science, MSU, and a grant from the American Polygraph..Association. All statements, opinions and conclusions are those of the author and do not necessarily represent the opinions or official position of the American Polygraph Association or its members ACKNOWLEDGEMENTS I wish to express my gratitude to my instructor, Professor Frank Horvath, who had the patience and the persistence to teach me in this learning endeavor, and to my wife of 25 years, Pyong—hwa Meesig, who had the patience and persistence to support me. iv TABLE OF CONTENTS LIST OF TABLES ........................................... viii LIST OF FIGURES ............................................ Xi Chapter I ................................................... 1_ Statement of the Problem ............................... 1 Police Applicant Selection ........................ 1 The PEPS Technique ..................................... 2 Public Opinion .................................... 3 Employee Rights ................................... 4 Extent of Use of PEPS ............................. 6 Need for the Study ..................................... 7 Purpose of the Study .................................. 8 Study Overview ........................................ 8 Chapter II Review of Literature ............................. 9 Section A — Polygraph Testing ........................ .9 The Examination Process ......................... 10 Specific Issue vs. PEPS Exams ................... 11 Research ........................................ 12 Accuracy. .................................. 13 Utility .................................... 14 Section B — Studies .................................. 15 Section C - Users .................................... 21 Characteristics of Users ........................ 21 Size. ...................................... 22 Type. ...................................... 23 Geographic region. ......................... 23 State laws. ................................ 24 Summary. ................................... 24 Why Police Agencies Use PEPS .................... 24 Summary. ................................... 25 How Agencies Use PEPS ........................... 25 Circumstances of use. ...................... 25 Policies and procedures. ................... 26 Summary. ................................... 28 User Evaluations of PEPS ........................ 28 V Summary. ................................... 29 Other Screening Techniques ...................... 30 Comment ......................................... 30 Section D - Nonusers ................................. 30 Former Users .................................... 31 Nonusers ........................................ 31 Summary ......................................... 34 Comment ......................................... 34 Chapter III Methodology .................................... 35 Research Questions ................................... 35 Variables ............................................ 36 Design of the Instrument ............................. 36 Population Under Study ............................... 37 Sample .......................................... 38 Data Collection ...................................... 39 Questionnaire Distribution ...................... 40 Size Characteristics of Respondents ............. 42 Agency Type of Respondents ...................... 44 Chapter IV Results ......................................... 46 Section A — Overview of PEPS Usage ................... 46 Section B - Description of Users ..................... 49 Characteristics of Users ........................ 49 Size. ...................................... 49 Type. ...................................... 50 Why Police Agencies Use PEPS .................... 51 Reasons for use. ........................... 51 Issues of greatest importance. ............. 53 How Police Agencies Use PEPS .................... 55 Circumstances of use. ...................... 55 Policies and procedures. ................... 58 User Evaluations of PEPS ........................ 62 Applicant pass/fail proportions. ........... 63 Admissions. ................................ 64 Evaluations of PEPS. ....................... 66 Other Screening Techniques ...................... 70 Section C - Description of Former Users .............. 72 Characteristics of Former Users ................. 72 Size. ...................................... 72 Type. ...................................... 73 When and Why Former Users Discontinued PEPS ..... 74 Past circumstances of use. ................. 74 Why Former Users discontinued use. ......... 75 vi Former User Evaluations of PEPS ................. 77 Screening Techniques Currently Used ............. 79 Future Plans to Use PEPS ........................ 81 Section D — Description of Nonusers .................. 85 Characteristics of Nonusers ..................... 85 Size. ...................................... 85 Type. ...................................... 86 Screening Techniques Currently Used ............. 87 Future Plans to Use PEPS ........................ 89 Section E - Comparisons Between Users,Former Users and Nonusers ....................... 93 Characteristics of Users and Nonusers ........... 93 Size. ...................................... 93 Type. ...................................... 94 Geographic region. ......................... 95 State laws. ................................ 98 User and Former User Evaluations of PEPS ....... 101 User and Nonuser Screening Techniques .......... 104 Former User and Nonuser Future Plans to Use PEPS ..................................... 107 Chapter V Discussion ...................................... 113 Section A - Extent of Use of PEPS ................... 113 Section B - Major Agency Characteristics Related to PEPS Usage ................... 115 Section C — Why and How Police Agencies Use PEPS ....119 Why Police Agencies Use PEPS ................... 119 How Police Agencies Use PEPS ................... 120 Why Former Users Discontinued PEPS ............. 121 Evaluations of PEPS ............................ 122 Other Screening Techniques ..................... 124 Future Plans to Use PEPS ....................... 125 Section D - Summary and Future Research Issues ....... 126 APPENDICES A. Independent Variables ............................ 132 B. Large Agency Survey Questionnaire ................ 134 C. Modified Questions Used in the Small Agency Survey Questionnaire ................ 144 D. ANOVA Tables ..................................... 146 vii 10. 11. 12. 13. 14. LI ST OF TABLES Extent of Police Agency Use of PEPS as Reported in Ten Surveys, 1962—1991 ........... Association of Agency Characteristics with the Use of PEPS by Police Agencies ...... Rank Order and Percentages of Nonuser's Reasons for Not Using PEPS ................... Number and Percent of Questionnaires Distributed and Received in the Large and Small Agency Surveys ..................... Size Characteristics of Agencies that Provided Usable Responses in the Large and Small Agency Surveys ..................... Agency Types of Agencies that Provided Usable Responses in the Large and Small Agency Surveys ......................... Comparison of Large and Small Agencies by PEPS Usage ................................ Size Characteristics of Large and Small Users Number and Percent of Large and Small Users Presented by Agency Types .................... Large and Small User Reasons for Using PEPS Presented by Mean Scores in Rank Order ...... Large and Small User PEPS Testing Issues of Greatest Importance Presented by Mean Scores in Rank Order ............................... Circumstances of Use of PEPS by Large and Small Users ................................. PEPS Policies and Procedures of Large and Small Users ............................. Proportions of Large and Small User Applicants Who Passed and Did Not Pass PEPS During a 12 Month Period ............... viii .......... 19 .......... 22 .......... 32 .......... 41 .......... 43 .......... 45 .......... 48 .......... 50 .......... 51 .......... 52 .......... 54 .......... 56 .......... 59 .......... 63 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. Large and Small User Applicant Admissions During PEPS Examinations ................... 65 Large and Small User Evaluations of PEPS .............. 66 Other Applicant Screening Techniques Used by Large and Small Users .............................. 70 Size Characteristics of Large and Small Former Users .................................... 72 Number and Percent of Large and Small Former Users Presented by Agency Types ................ 73 Past Circumstances of Use of PEPS by Large and Small Former Users ....................... 74 Large and Small Former User Reasons for Discontinuance of PEPS .................... 76 Former User Evaluations of PEPS ....................... 78 Screening Techniques Currently Used by Large and Small Former Users ....................... 80 Large and Small Former User Plans to Implement PEPS ............................... 82 Circumstances in which Large and Small Former Users Indicated that the Use of PEPS would be Considered ................................... 84 Size Characteristics of Large and Small Nonusers ...... 86 Number and Percent of Nonusers Presented by Agency Types ............................. 87 Screening Techniques Currently Used by Large and Small Nonusers ........................... 88 Large and Small Nonuser Plans to Implement PEPS ....... 9O Circumstances in which Large and Small Nonusers Indicated that Use of PEPS would be Considered ........ 92 Comparisons of Large and Small Users and Nonusers by Agency Type ............................... 94 Comparisons of Users and Nonusers by U.S. Census Bureau Regions ......................... 96 Comparisons of Large and Small Users and Nonusers by U.S. Census Bureau Regions........... ..... 97 ix 34. Comparisons of Users and Nonusers by the Existence of State Laws Affecting PEPS ................ 99 35. Comparisons of Large and Small Users and Nonusers by the Existence of State Laws Affecting PEPS ........ 100 36. Comparisons of Users and Former Users Regarding Their Levels of Confidence in PEPS ................... 101 36A. ANOVA Tables Testing Differences in Agency Mean Confidence Level Scores with Respect to PEPS Usage and Agency Size ................................ 146 37. Comparisons of Users and Former Users Regarding Their Estimates of Accuracy of PEPS .................. 103 37A. ANOVA Tables Testing Differences in Agency Mean Accuracy Level Scores with Respect to PEPS Usage and Agency Size ................................ 147 38. Comparisons of Users and Nonusers Regarding Screening Techniques Used .................. 105 39. Comparisons of Former Users and Nonusers Regarding Plans to Implement PEPS .................... 109 40. Comparisons of Former Users and Nonusers Regarding Circumstances in which PEPS would be Considered .................................. 111 1. LIST OF FIGURES Number and Percent of Users, Former Users and Nonusers in the Large and Small Surveys and in Both Surveys Combined ........................... 46 xi Chapter I The screening and selection of applicants for police officer positions is an important yet difficult function of police administrators. This thesis addresses the extent to which pre—employment polygraph screening (PEPS) is used by police agencies, the major factors related to its use, and why and how agencies use it as an applicant selection technique. Statement of the Problem According to the Bureau of Justice Statistics (BJS)(1992), there are 15,430 general purpose law enforcement agencies in the U.S. A total of 12,288 (79.7%) are local (operated In! municipal cur county' governments) jpolice agencies, 3,093 (20%) are sheriff agencies and 49 (0.3%) are state police and highway patrols. These agencies employ 741,195 full—time personnel. About 62% (459,891) are employed at the local level, 28% (203,974) by sheriff agencies, and 10% (77,330) by state agencies. Approximately 75% (556,791) are "sworn" officers who have full arrest powers for all types of offenses. The combined annual budgets of all these agencies was approximately $28 billion in fiscal year 1990 (Bureau of Justice Statistics [BJS], 1992). Police Applicant Selection The overall effectiveness of a: police agency is initially impacted at the employee entry level. The selection of personnel is one of the most important functions 2 of agency administrators as it affects the entire organizational mission. Administrators essentially attempt to predict future performance every time a police applicant is selected. They try to do this by basing their prediction on an evaluation of information regarding the applicant's qualifications and past history. In order to evaluate performance, a variety of personnel selection methods and techniques can Ina used t1) obtain information regarding applicants (Ash, Slora and Britton, 1990). Administrators then select applicants based on the information that the selection techniques have provided. The PEPS Technique PEPS is a type of screening procedure in which a job applicant undergoes a polygraph examination to determine whether he or she meets an employer's requirements for employment. The use of PEPS in both the private and public sector to determine a prospective employee's suitability for employment has been and continues to be an extremely controversial matter of public policy. Although part of the reason for this controversy stems from the lack of solid scientific research about polygraph testing in general, and PEPS in particular, there seems to be differing perceptions about its role in employee selection. Primarily, there appears to be far greater support for PEPS in the public sector than in private industry (Horvath, 1987a). This can be seen particularly in the areas of public opinion and employee rights. Pu i inion With regard to public opinion, the available surveys of the U.S. population reflect that the majority of the public favors polygraph usage in specific situations involving high degrees of trust such as law enforcement. However, there seems to be much less public sentiment for polygraph usage without sufficient justification (Phannenstill, 1983; Horvath, 1987a). A review of the literature disclosed seven surveys of the attitudes of persons who have actually taken PEPS in commercial settings (Horvath and Phannenstill, 1987). In these seven surveys, carried out in different cities and at different times between 1973 and 1987, 82% (representing a total of about 1200 respondents) of the people reported that based on their experiences in taking the PEPS exam, they thought that polygraph screening was fair. A similar percentage reported that the test was not objectionable and was not an invasion of their personal privacy, and about 84% agreed to take such a test again to gain employment. A similar study done by Putnam (1978) on applicants for police work in Washoe county, NV, showed similarly highly favorable views (Putnam, 1978). In another study of applicants for the National Security Agency, 74% of the respondents reported that polygraph screening was important to protect security and 91% agreed to further testing if necessary (Department of Defense, 1984).» Finally, in a national poll, using a sample generalizable to 161 million adults in the U.S., 65% of the 4 respondents reported that they did not object to polygraph screening and 81% felt that polygraph testing should be used for screening persons in sensitive situations, such as those who handle classified information (Horvath, 1987). However, as Horvath (1987) pointed out in his discussion of the poll results, there was much less public support for the arbitrary or routine use of such testing. In summary, the available research data do not show that polygraph screening is generally viewed by the public as an unfair‘ or objectionable employment. practice. CHI the contrary, it indicates considerable public support for polygraph testing when it is used with adequate justification in sensitive positions of public trust such as law enforcement. Employee Rights In 1988 the U.S. Congress passed the Employee Polygraph Protection Act (EPPA). This law essentially prohibits the great majority of private employers in the U.S. from using polygraph testing to screen applicants for employment (EPPA, 1988). However, the EPPA also attempted to balance the prospective employee's privacy rights against the employer's right to protect his or her business by including a number of exemptions for private security services and other special classes of private employers. Further, it exempted all federal, state and local government employers, and national defense and security positions from its provisions (Cross, 1989). Similarly, although 44 state legislatures and the 5 District of Columbia have instituted restrictions limiting the use of the polygraph in some form or another or have laws requiring licensure of polygraph examiners, there are provisions in almost all state statutes exempting public agencies from proscriptions against testing in employment situations (Ansley and Beaumont, 1992). Hence, in creating this double standard for private and public employees, legislative bodies that have considered the issue of polygraph testing have essentially acknowledged that a properly conducted PEPS process does not infringe on applicant rights, and has value in applicant selection, particularly in positions involving high public trust (Cross, 1989). In concert with the legislatures, the courts also have recognized the need for public law enforcement agency administrators to select only the most highly qualified applicants to fill sensitive positions of public trust. They have held that applicants for police positions can be required to submit to PEPS exams or be denied employment (O'Leary, 1977). Although the U.S. Supreme Court has never addressed this issue, in 1987, the U.S. Third Circuit Court of Appeals heard an appeal by applicants for the Philadelphia, PA, Police ZDepartment. alleging' that the department's PEPS requirement was a violation of their due process and equal protection rights. The court ruled that, provided that the PEPS test is administered to all applicants in a fair and reasonable manner without discrimination, and 6 that all questions are job—related, the technique did not violate either their due process or equal protection rights (Anderson v. City of Philadelphia, 1987). Extent of Use of PEPS Until the passage of EPPA in 1988, private sector employers used PEPS tests for over 40 years to help select applicants for employment or to assist in the investigation of employees suspected of involvement in acts of employee dishonesty. While exact figures are difficult to obtain, there are several surveys which indicate that prior to passage of EPPA in 1988, about 20% of all major businesses in the U.S. made use of PEPS. In certain industries involving high levels of employee trust, the figures were much higher; for instance, approximately 50% of all commercial banks and over 60% of all retail operations reportedly used PEPS (Buckley, 1988). In the public sector, the first reported use of the polygraph in screening police applicants was in the early 19505 in a California police department (Gugas, 1962). Since that time, the available surveys of state and local police agencies show that PEPS examinations of police applicants has grown steadily from fewer than 20% in the 1960s (Yeschke, 1962), to at least 44% in the mid—19808 (Roper, 1981; Horvath and Shelton, 1982), and more recently, to 75% of selected agencies in the U.S. (McCloud, 1991). At the federal level there has also been a dramatic increase in the use of PEPS in recent years. The Federal 7 Bureau of Investigations and the U.S. Secret Service have initiated PEPS programs, and the Department of Defense reported that the polygraph is a valuable screening device in its applicant selection process, even though it uses many others, including the most expensive and time—consuming procedures (Department of [miense, 1984). Further, the nation's most important investigative and counterintelligence agencies have supported the continued use of the polygraph as an indispensable ‘weapon against crime, espionage and international terrorism (Heinley, 1986; Cross, 1989). Need for the Study The literature shows that the use of PEPS as a police applicant selection technique by law enforcement agencies in the U.S. has increased significantly over the past thirty years. Yet, as Meesig and Horvath (1993) point out, the literature reflects surprisingly little information regarding the effectiveness of the technique cu luwv it is used by police agencies, or even what may lead one agency to implement it and another to reject it. Further, the need for more thorough and sound research has been clearly identified for more than a decade as one of the primary reasons for the reluctance to use it. Additionally, because of the lack of existing research on PEPS, police administrators considering implementation have minimal information available upon which to evaluate it and base their decision. Yet as the increased use of PEPS among agencies would indicate, such decisions are being made with greater frequency. 8 There is a need for a current and comprehensive description regarding the use of PEPS. The results of such a report can be used by researchers, the legislatures and the courts to address the conflicting and controversial issues surrounding PEPS, and it can also be used by police administrators to assess the technique and make more informed decisions regarding its legitimate and proper use in their agencies. Purpose of the Studv The purpose of this study is three-fold: (1) To obtain current information regarding the extent of use of PEPS as an applicant selection technique for police agencies in the U.S. (2) To identify the major factors related to PEPS usage. (3) To provide an analysis of why and how U.S. police agencies use PEPS. Study Qverview In Chapter II, after a brief discussion of polygraph testing, the available literature pertaining to police agency use of the PEPS technique will be reviewed. The survey methodology used in this study will be set forth in Chapter III. Ina Chapter IV, the purposes of this study will be accomplished by presenting a description of the findings of the survey findings. Chapter V will consolidate the study results and compare them with previous studies. Chapter II Review of Literature This chapter is divided into four sections. It begins with a general discussion of polygraph testing. Then, in the second section, studies found in the literature specifically reporting on the extent of use of PEPS by police agencies are identified. Following' that, 1J1 Section (3 additional information those studies contain pertaining to agencies that use PEPS is presented. Finally, in the fourth section, the information the studies contain pertaining to agencies that do not use PEPS is presented. Section A - Polygraph Testinq It was not until about 1895, when Cesare Lombroso, an Italian physiologist, and his student, Mosso, published their work on the use of the hydrosphygmograph, that objective measurement of physiological changes became associated with the detection of deception. Since that time, substantial improvements and refinements have been made in the process of detecting deception, and in our knowledge about the process (Ansley, Horvath and Barland, 1983). The polygraph instrument of today, after almost a century of development, is a sophisticated instrument which consists of a ndnimum of three measures of physiological functions. The basic components are the cardiograph, which monitors changes in blood volume and heart rate; the pneumograph, which measures respiration and related movements in the abdominal and thoracic area; and the galvanometer, 10 which records the changes in resistance to electricity on the surface of the skin. The constant changes in physiological measurements sensed by these components are transmitted to a pen and recorded on a moving chart (Ansley and Abrams, 1980). The Examination Process A polygraph examination normally consists of a pretest interview, polygraph testing and a post-test phase. During the pre-test interview, the examiner explains to the examinee the nature of the polygraph instrument, the purpose of the test and the pertinent issues. The examiner also formulates about eight to ten questions on relevant issues and reviews all test questions with the examinee at this time (Horvath, 1987a). During the polygraph test phase, the second component of the examination process, the examiner attaches the polygraph to the examinee and asks the examinee the previously reviewed questions while the polygraph instrument records physiological changes. Relevant questions (those pertinent to the issues to be resolved) are asked at least once in separate repetitions of the question list (Horvath and Reid, 1981; Bureau of National Affairs, 1985). The results of the testing, that is, the charts on which the physiological measurements are displayed, are reviewed by the examiner. There are a number of accepted methods for carrying out this review, depending on such things as the testing procedure, the particular training orientation of the examiner, and the type of examination at hand. In any event, the review of the data consists of a comparison of the 11 nature, magnitude and consistency of physiological changes which occur'txa the different questions asked during the testing. The examiner then makes a determination regarding the examinee's truthfulness to the questions (Nagle, 1984; Horvath, 1987a). Often, particularly in federal agencies, the examiner's review may be followed by at least one other review carried out by another examiner or a polygraph supervisor as ea quality control measure. If there is disagreement, additional testing may be carried out. In the post-test phase following the testing phase the examinee is advised of the test results and given an opportunity to explain or clarify any issue which may be of concern. The test and post—test interview may be repeated in an effort to clarify the examinee's responses to specific questions (Garwood, 1985). After the examination process is completed, the test results (i.e., information provided by the examinee) are provided to the authority that requested the exam to assist in a decision—making process regarding the examinee. Steam—WM The polygraph is used to conduct specific issue examinations to investigate involvement in criminal offenses and it::hs also used to conduct PEPS examinations of job applicants to determine suitability for employment. Although each type of examination is typically administered in the phases described above, Horvath (1987a) described several important differences between them. First, in a specific 12 issue examination the pretest interview focuses only on the offense at hand and, aside from the collection of demographic information, there is In: questioning regarding unrelated matters. In contrast, in the PEPS examination the pretest interview is essentially an information gathering process. The applicant is asked questions regarding a number of different areas of concern without focus on any one particular issue. A second difference in procedure between the specific issue and PEPS examinatbmms is that in the former the purpose is to determine the examinee's truthfulness to one specific issue, e.g., a murder, a burglary, an arson, etc. In the latter instance, the purpose of polygraph testing is to verify the applicant's truthfulness regarding several different issues. Third, in a specific issue examination all relevant test questions pertain to the same offense, whereas in a PEPS examination each relevant test question pertains to a separate area of inquiry (Horvath, 1987a). Research Although the use of PEPS by police agencies has grown considerably over the past 40 years, Surprisingly little research had been conducted regarding the technique. Both proponents and opponents of the polygraph agree that there has not ‘been. sufficient empirical research (n1 either polygraph testing in general, or on the use of the polygraph in screening situations in particular (Office of Technology Assessment, 1983; Horvath, 1985). 13 Accuragy. Specifically with regard to the accuracy of polygraph testing, the research evidence is not well developed, even though since the 19705 over 100 studies have been conducted regarding the subject (Buckley, 1988). Although there is general agreement 1J1 the scientific community that the accuracy of polygraph testing is greater than chance (Office of Technology Assessment, 1983; Nagle, 1984), the issue that separates the opponents from the proponents is not whether polygraph testing "works" (i.e., has a detection rate above chance), but how well it works (Horvath, 1987a). Opponents of polygraph testing claim essentially that it does not work well enough to justify its use (Office of Technology Assessment, 1983). Proponents point out, however, that there is presently no research to show that PEPS is inaccurate, and moreover, that all of the scientific data now available show that its accuracy is equal to or better than other selection devices (Nagle, 1984; Horvath, 1987a). Proponents typically maintain that the accuracy of polygraph testing is about 90%; opponents maintain that it is about 70%, but with a high "false positive" error rate (Buckley, 1988). (A false positive error is a polygraph outcome that indicates a person is deceptive when the person is, in fact, telling the truth. Conversely, a "false negative" error is a polygraph outcome of truthfulness, or no deception, for an examinee who is actually lying). Opponents claim that false positive errors occur more often than do false negative errors (Bureau of 14 National Affairs, 1985), and as a result, honest people could undeservedLy be denied legitimate employment or otherwise discriminated against. The great majority of the research studies on polygraph testing pertain to specific issue polygraph examinations (e.g., investigation of a robbery, homicide, theft, etc.) and are not applicable to PEPS testing. As previously noted, there are important differences of opinion regarding how to interpret research evidence relating to accuracy even if it is confined only to specific issue testing. In addition, it is generally agreed that generalization from the research base on specific issue testing to PEPS can only be done with great caution (Correa and Adams, 1981). Aside from the accuracy issue, both Nagle (1984) and Horvath (1987a) report that in the literature in general there is a consensus that polygraph testing has a utilitarian value. It serves a utilitarian purpose in that the information collected during PEPS exams is often not obtainable through other methods. One of the few key pieces of empirical research specifically addressing this PEPS issue was a field study conducted by Blum (1967) to determine its utility in real- life situations. Blum's study was conducted in a California sheriff department and included an assessment of PEPS together with information obtained by a pmysical exam, a physical agility test, a written application, a preliminary 15 interviewu written. psychological tests, a jpsychiatric interview, a background investigation, a civil service exam, and an oral board exam The PEPS results were not made available for consideration in the selection process. The study showed that of the 57 applicants who had completed all phases of screening, the oral board selected 31 of them. Of these 31 selectees, 17 of them (55%) had admitted during PEPS testing to having committed serious crimes in their past. Only one of them had been identified by other selection methods as having committed a serious crime. The study showed that PEPS has utility in revealing an applicant's involvement in serious crime not revealed by other selection procedures, and that other procedures do not reveal a criminal history not also revealed by PEPS (Horvath, 1987). However, follow—up observations of how these selectees subsequently performed as sheriff deputies were never reported, and the usefulness of PEPS as a means of actually predicting subsequent on—the-job performance and honesty remains untested. Section B — Studies In this section ten studies found in the literature specifically reporting on the extent of the use of PEPS by police agencies are identified. One of the earliest reports regarding the extent of PEPS testing of police applicants was by Gugas (1962). He reported the testing of applicants in California police departments since the early 19508. Gugas stated he was aware 16 of about 35 police agencies using PEPS as part of their applicant screening programs. In 1962, Yeschke conducted a survey of 180 large law enforcement agencies to examine their use of PEPS and to study PEPS advantages and limitations. His survey population included all state police agencies and police departments in the U.S. serving cities with populations of 100,000 or more. A total of 116 agencies (64%) responded and 19 agencies (16% of respondents) reported using PEPS. Twenty—six (22%) were considering its use in the future. In 1964, Gooch conducted a national survey designed to determine the extent of the use of PEPS by U.S. law enforcement agencies. In this study, Gooch selected a purposive sample of 167 federal, state and municipal police agencies, including all state agencies (excluding Alaska and Hawaii) and municipal agencies serving populations of 50,000 and above. A total of 118 usable responses were received, which was a response rate of 71% (118/167). Twenty-three agencies (19% of respondents) reported using PEPS together with other techniques, and 13 (11%) were considering its use. Eisenberg, Kent and Wall (1973) conducted a survey of various personnel practices in large police agencies across the U.S. They sent questionnaires to all state (47), county (140) and municipal (481) police agencies in the U.S. having 50 or more sworn personnel. Based on a return rate of 74% (493/668), they found that 31% of the respondents (153 agencies) reported using PEPS to assess aptitudes and l7 characteristics of sworn officer applicants. Roper (1981) conducted a national survey of all state departments of public safety, county sheriff departments and municipal police departments with 100 or more sworn employees to determine police applicant selection requirements and procedures currently used. A total of 508 (75%) of the 675 agencies included in the study responded and, of those, 221 (44%) reported using PEPS in their applicant selection battery. In 1982, Horvath and Shelton reported the results of a national survey which extended the Gooch (1964) study and sought to determine changes in the use of PEPS during the 18 years since the Gooch report. Horvath and Shelton surveyed all federal and state police agencies and a stratified random sample of 270 local agencies. From a total of 340 agencies, usable responses were received from 237 (70% response rate) and a total of 105 agencies (44%) reported using PEPS. Fourteen agencies (6%) reported that they had used PEPS in the past but had since discontinued its use, and 76 agencies (65% of 117 responses) said they would consider its use. Kendrick (1983) reported a survey in which 91 state and local law enforcement agencies across the U.S. responded to a questionnaire regarding the use of the polygraph (the total number of agencies included in the survey and the manner in which they were selected were not reported). About 43% (39) of the respondents reported using the polygraph as part of their applicant screening process. Thirteen of the 91 18 agencies (14%) reported that they had used PEPS in the past but had since discontinued its use, and 11 (12%) said it was being considered for future use. In the mid—19805, the Oceanside, CA, Police Department sent out questionnaires to 405 California police agencies to determine the extent of use of the polygraph in screening police applicants (Lopez, undated). The manner in which the 405 agencies were selected was not specified. Of the 218 agencies that responded (54% response rate), 113 (52%) reported using it. Ash, Slora and Britton (1990) conducted a survey of large police agencies to determine selection procedures used in screening police applicants. A total of 99 agencies (49 state police agencies and police departments in the 50 largest cities in the U.S.) were surveyed and 62 agencies (63%) responded. Thirty—five agencies (56%) reported using the polygraph as part of their screening programs. McCloud (1991) conducted a survey of all state police agencies and selected municipal police agencies in 1989 to determine the extent and utilization of polygraph testing in law enforcement. Out of 406 agencies in the sample, 308 (76%) responded and at least 75% of them reported using the polygraph for applicant screening. In Table 1 below, extent of use information regarding the studies reviewed are set forth for ease of comparison and reference. 1 9 Table 1 Extent of Police Agency Use of PEPS as Reported in Ten Surveys, 1962-1991 No. & Type No. (%) No. (%) urve car A encies am le A enc Size Re onse Users Gugas 1962 Unspecified Unspecified Unspecified Unspecified 35 ( - ) CA only Yeschke 1962 180 Population Large 1 16 (64) 19 ( 16) State, Local (100,000+ pep) Gooch 1964 167 Purposive Large 1 18 (71) 23 ( 19) Federal, State, Local Eisenberg, 1973 668 Population Large 493 (74) 153 (31) Kent & State, (50+ sworn Wall County, employees) Local Roper 1981 675 Population Large 508 (75) 221 (44) State, (100+ County, sworn Local employees) Horvath & 1982 340 National All Sizes 237 (70) 105 (44) Shelton Federal, Random State, Local Kendrick 1983 Estimated Unspecified Unknown 91 (80) 39 (43) 114 State, Local Lopez Undated 405 Unspecified Unknown 218 (54) 1 13 (52) (19805) CA only Ash, Slora 1990 99 Population Large 62 (63) 35 (56) & Britton State, (50 largest Municipal cities) McCloud 1991 406 Purposive Unknown 308 (76) Estimated State, 231 (75) Municipal As the table reflects, the studies were conducted over a thirty year period. The survey sample sizes of the studies 20 ranged from 99 to 675 agencies and the reported survey response rates ranged from 54 to 80%. The percentage of agencies that reported using PEPS increased consistently from 16% in 1962 to 75% in 1991. Commencing with the Yeschke (1962) study; almost. every' subsequent. survey' found an1 increase in usage among police agencies, even though the surveys themselves varied in size, sample selection criteria and definitions of agency size and type. This would indicate that the use of PEPS increased generally throughout the U.S. law enforcement agency population during the three decade time period. However, it is important to note here that, because of the various methodologies employed 1J1 the studies, the results of only one of them may be considered to be generalizable to all agencies in the U.S. Only the Horvath and Shelton (1982) study used a nationally representative random sample of all federal, state and local agencies and, therefore, the study results may be generalizable to all such agencies in the U.S. The Eisenberg, Kent and Wall (1973), Roper (1981) and Ash, Slora and Britton (1990) studies limited their surveys to specific populations of large agencies, and as a result their study findings may be generalizable only to those specified populations. For the remaining six studies, generalization of findings beyond the study sample groups may not be appropriate as the manner of selection of agencies for inclusion in the studies was either not clearly specified or was not random. 21 Section g — Users For clarity in discussion, police agencies that reported they currently used PEPS in their applicant selection programs are referred to as Users. In this section additional information reported in the above studies pertaining to PEPS Users is reviewed. The information is presented in the context of the characteristics of Users, why and how they use PEPS, User evaluations of it an; a screening device based on their experience, and other screening techniques employed by Users. Characteristics of Users Five of the above studies reported information comparing police agency characteristics with the use of PEPS. Agency characteristics included the agency size (which was measured by the number of agency sworn employees and also by the size of the population within the agency's area of jurisdiction), the type of agency (i.e., municipal, county, state), the geographic region in which the agency was located in the U.S., and the existence of state laws regulating the use of PEPS. The data reported by the studies on these characteristics are summarized in Table 2 below. 2 2 Table 2 Association of Agency Characteristics with the Use of PEPS by Police Agencies Agency Size (sworn Agency Size Geographic Survey ermns) (pop) Agency Type Region State Lflvs Eisenberg, Small less - Not related - - Kent & Wall likely to use (1973) Roper (1981) Not related Not related Not related Difference by Trend mea Horvath & Small less - Not related - - Shelton (1982) likely to use Lopez Small less - - - - (undated) likely to use Ash, Slora & - - Municipal - - Britton use more (1990) than state Size. Roper (1981) reported finding no relationship between the use of PEPS and agency size (by either number of sworn employees or size of population served). However, her survey sample included only larger-sized agencies (all state, sheriff and municipal agencies employing 100+ sworn personnel). On the other hand, the Eisenberg, Kent and Wall (1973) study found that the polygraph. was used less frequently in smaller agencies than in larger agencies. The reported usage by agency size was as follows: 100 or fewer sworn employees — 25%; 101-300 sworn employees — 34%; 301-500 sworn employees — 35%; 501—1,000 sworn employees - 32%; and 1,000+ sworn employees - 52%. The Lopez (undated) study, 23 which did not report agency sizes, and also the Horvath and Shelton (1982) study, which included agencies in a wide range of sizes based on both the number of sworn employees and city population sizes, also noted that smaller agencies were less likely to use PEPS than larger agencies. This would indicate that agency size is related to usage of the technique. EYES- Eisenberg, Kent anni Wall (1973), Roper (1981) and Horvath and Shelton (1982) all reported finding no relationship between agency type and the use of PEPS. However, Ash, Slora and Britton (1990) reported that in their survey municipal agencies were significantly more likely to use the technique than state police agencies. This finding may have been affected by the small number of User respondents in the study (a total of 35 state and municipal agencies - reference Table 1). In any case, it appears likely that agency type has at most had little relationship with the use of PEPS during at least the past twenty years. Seggraphic regign. Only the Roper (1981) study compared police agency PEPS usage with geographic regions of the U.S. When she analysed PEPS usage by U.S. Census Regions, she reported that usage was most prevalent in the Western (60% of respondents) and Southern (50%) regions of the country. In contrast, agencies in the Northeastern U.S., including New York and New Jersey, were least likely (about 8%) to use PEPS. 24 State laws. Again, only the Roper (1981) study mentioned that the use of PEPS may be tied to statewide requirements regarding its legality. However, the matter was not further pursued. Summapy. In sum, the literature indicates that agency size and geographic region may be associated with police agency use of PEPS, whereas agency type and state legal requirements are not associated with usage. Whv Police Agencies Use PEPS In this section information available in the studies that provides insights as to why police agencies use PEPS is presented and analyzed. Gugas (1962) stated that PEPS examinations of 510 applicants in seven police agencies resulted in the agencies rejecting 233 (46%) of them based on the uncovering of information that had not previously been detected by other techniques. The type of information that was revealed related to applicant admissions regarding health problems, arrests, military service, prior work records, abnormal sex behavior and credit problems. Yeschke (1962) also reported that a properly administered PEPS exam could be used to check hidden unlawful activities of applicants, with the intent to eliminate the less desirable applicants from further consideration. Horvath and Shelton (1982) reported the primary reason agencies used PEPS was that it reveals information about 25 applicants not available through other screening methods (94%). The second and third ranked reasons were that it "deters undesirable applicants" (54%) anxi its "speed in obtaining results" (42%). Ranked fourth was that PEPS "saves money" (37%). The type of information that the PEPS exams revealed that were not otherwise detected were admissions related to both felonies and misdemeanors committed, drug and alcohol abuse, dishonesty on jobs, homosexual activities and finance/credit matters. Summapy. Based on the available data, police agencies use PEPS primarily' because it if; most effective sh) revealing unfavorable information run: otherwise available. Such information can then be used to screen out the less suitably qualified applicants. Hog Agencies Sse PEPS In this section the studies are reviewed for information regarding circumstances of use of PEPS and agency policies and procedures governing its use. gircumstances of use. Horvath and Shelton (1982) reported the following information: - Almost 94% of their 105 User respondents reported using just the polygraph instrument to conduct PEPS exams. About 2% reported using a voice analyzer instrument instead of a polygraph instrument. Almost 5% reported using some combination of the instruments. 26 — About 80% used their own examiners to conduct PEPS exams; 18% used an outside firm and 2% used both. When asked why they used an outside firm, 45% said it was due to a lack of trained examiners, 25% said agency size precludes inhouse programs, 5% said it was less expensive and 25% indicated a combination of reasons. Kendrick (1983) reported that 95% of his 91 respondents stated they had their own examiners and that 92% of them conducted PEPS exams for other agencies: only 15% charged a fee for their services. He also reported that 12 of the User respondents did not have their own examiners and arranged for other agencies or firms to conduct their PEPS exams: five of them paid a fee for the exams. Horvath and Shelton (1982) reported that: — Respondents claimed that a mean of 242 PEPS exams were conducted annually by or for their agency (Range=0 to 2100; Median=125). - About 54% began using PEPS exams prior to 1972; 45% began between 1973—1981 (1% unknown). Policies and procedures. Kendrick (1983) reported that 66% of his respondents had written policies regarding PEPS. Additionally, 62% stated they had quality control or review procedures (primarily reviews by other examiners or a supervisor) relative to PEPS. With respect to which categories or groups of applicants are required to undergo PEPS testing, Eisenberg, Kent and Wall (1973) reported that many of their respondents specified 27 that PEPS was not always used for all applicants and that it often was used only for questionable applicants. Horvath and Shelton (1982) reported that 74% of their respondents tested applicants for sworn positions and 14% tested applicants for civilian positions. Horvath and Shelton (1982) reported that 79% of the respondents stated they advertised or gave some other form of notice to the public that a PEPS exam was required. A total of 61% reported that applicants were made aware of the PEPS requirement before they were given application form packages, 26% made individuals aware during the time the packages were filled out, and 14% made them aware after the packages were filled out. Several studies reported information regarding the sequence in which PEPS should be administered in combination with other techniques. Gugas (1962) reported that PEPS should supplement psychology tests, and Yeschke (1962) reported that it could be used in conjunction with a background investigation. Kendrick (1983) reported that of the 91 respondents in his survey, 22 indicated they used it before the background investigation in their selection process, and 12 indicated they used it after the background investigation. Lopez (undated) reported tflun: over 97% (102/105) of his respondents said the sequence and proper scheduling of polygraph exams in the screening process was important. About 69% ranked oral boards and interviews as the most important first step in the process, and 44% ranked 28 PEPS as the most important second step. Psychological testing was considered by 45% as the most important third step. Horvath and Shelton (1982) reported that PEPS was generally administered after written tests and oral interviews but before psychological and/or psychiatric exams, background investigations and physical exams. Horvath and Shelton (1982) reported that about 82% of the respondents said that an applicant's refusal to undergo a PEPS exam leads to automatic rejection of the application. Horvath and Shelton (1982) reported that the two primary purposes of PEPS exams were to verify information derived from the application form and/or the background investigation (82%) and to develop new information not revealed by other selection devices/steps (67%). Only 3% reported using PEPS only when questionable or apparently incomplete information is provided or discovered. Summa y. In sum, the information available regarding how agencies use PEPS is minimal and some information is based on limited samples of agencies. However, when studies reported on similar areas of use (most agencies use their own examiners; not all applicants are tested; PEPS is normally administered after interviews and before background investigations), their findings were in general agreement. W In this section, information available in the studies reflecting the perceived agency benefits and results obtained 29 by using PEPS is reviewed. In Gooch (1964), agencies reported that the three primary benefits they received from using PEPS were "higher quality employees" (74%), "fewer problems" (57%) and "lower turnover" (48%). Additional. benefits ‘were "increased employee efficiency" (43%), "reduced training costs" (35%), "increased public confidence" (35%) and a "reduced number of applicants" (22%). Eighteen years later, Horvath and Shelton (1982) reported the same three primary benefits in the same rank order (88%, 43% and 36%, respectively). "Fewer citizen complaints" ranked fourth (33%) and several other miscellaneous benefits were also listed. Gugas (1962) did not specifically evaluate the importance of the types of information revealed by PEPS, but he did report that the largest cause for the 46% applicant rejection rate based on PEPS exams was admissions of health problems, and that other causes included admissions regarding arrests, military service, prior work records, abnormal sex behavior and credit problems. Lopez (undated) reported that when asked to evaluate how successful they felt PEPS was, 81% of their respondents rated it above average or outstanding. SEEEQEY- In sum, Users perceived positive and tangible benefits, primarily in the area of higher quality employees, based on their use of PEPS. 30 cher Screening Technigges Of the studies reviewed, only Roper (1981) presented information regarding PEPS usage and the number of other techniques used by agencies (both Users and Nonusers) which indicated a possible relationship between the two variables. She reported that agencies in the Western U.S., where the highest number of Users in her survey were located, tended to incorporate the most screening devices in their selection programs, and that agencies in the Eastern U.S., where the fewest Users in her survey' were located, tended to incorporate the fewest. However, the association between the two variables was not further pursued. QQEEQEL It is important to recognize that all of the above reported information regarding Users is at least 10 years old. It was published well before the U.S. Congress passed the EPPA in 1988, which proscribed the use of PEPS in the private sector. Thus, the data does not reflect any potential impact of EPPA, or of any other judicial or state legislative actions in recent years. Section D - Nonusers In this section, information available in the studies is reviewed pertaining to police agencies that reported they had used PEPS at some time in the past but had since discontinued its use (referred to as Former Users) and agencies that reported they have never used PEPS (referred to as Nonusers). 31 Former Users Horvath and Shelton (1982) reported that 14 agencies (6% of the total number of respondents) were Former Users of PEPS. Thirteen of them indicated that they had used the polygraph instrument to administer PEPS exams. Four said they had discontinued its use prior to 1972 and eight discontinued.:U: between 1973 - 1981. The reasons they discontinued its use included enactment of state prohibitory legislation (four agencies), revision of selection process (three agencies), cost (two agencies) and dissatisfaction with results (one agency). Kendrick (1983) was the only other study that reported Former User respondents. In that study 13 agencies (14% of 91 respondents) identified themselves as Former Users. NQQQEQLE Gooch (1964) and Horvath and Shelton (1982) both asked agencies why they did not use PEPS and in both studies the agencies that responded included both Former Users and Nonusers. The reasons for nonuse were tabled and compared by Meesig auui Horvath (1993). Their findings regarding the number and percent of agency scores for each reason in the two studies are presented in Table 3 below, and the reasons are rank ordered according to the Gooch (1964) study percentages. 3 2 Table 3 Rank Order and Percentages of Nonusers' Reasons for Not Using PEPS Gooch Horvath & Shelton (1964) (1982) Reasons (N=95) (N=1251 MI I! Rafi % Satisfied with Other Methods 95 l 39 125 l 54 Never been approached! unaware of program 95 2 17 125 7 1 1 Cost Involved 95 3 14 125 2 34 Considering Use 95 4 14 N/A2 Agency Size 95 5 9 125 5 15 Lack Confidence in Polygraph Examiners 95 6 6 125 8 6 Resentment that Would Result 95 7 6 125 9 6 Lack Confidence in Polygraph Technique/Equipment 95 8 5 125 6 l4 Shortage of Applicants 95 9 4 125 10 5 Lack Trained Examiners N/A 125 3 22 Legislative Problems N/A 125 4 16 1Corrected for missing data. Percentages rounded to nearest whole percent. 2N/A=Question or option not asked or not available in this study. It can be seen that in both the Gooch (1964) and Horvath and Shelton (1982) studies, the primary reason given was that the agencies were satisfied with their current (other) methods of screening; in Gooch (1964), 39% and in Horvath and Shelton (1982), 54% of the agencies cited this reason. In Gooch (1964), the second—ranked reason was that the agencies had never been approached about initiating a program or were 33 generally unaware of polygraph screening programs, cited by 17% of the agencies. The third reason was that such a program was viewed as being too costly (14%). In the Horvath and Shelton (1982) study, the cost of polygraph screening was the second-ranked (34%) factor and the lack of trained examiners was third (22%). It is of some interest to note that in the Horvath and Shelton (1982) study the lack of awareness of polygraph screening was the seventh-ranked factor (11%), showing, perhaps, a growing awareness of the use of such screening over the 18 year period of time between that and the Gooch (1964) study. In addition to the above, Lopez (undated) reported that many small agencies that did not use PEPS listed budgetary constraints as the primary reason. Several studies asked agencies about the possibility of using PEPS in the future. Yeschke (1962) reported that 26 (27%) of his 97 Nonuser respondents said they 'were considering its use in the future. Gooch (1964) reported that of the 95 responding Nonuser agencies, 13 (14%) were considering its use. Kendrick (1983) reported that 11 (25%) of 44 Nonuser respondents stated that ,PEPS was being considered for the future. In the Horvath and Shelton (1982) study, 132 responding agencies were identified as Nonusers (including Former Users) and were asked under what conditions they ‘would. consider' implementing' PEPS exams in 'their applicant screening program. Among the 177 responding agencies (89%) the top three reasons were (1) evidence of 34 effectiveness (37%); (2) court acceptance of the validity of PEPS (36%); and (3) if law/policies permit its use (17%). Other reasons included "if funds available" (11%), "increase in applicants" (11%), "systems failure" (9%), "improved training and/or requirements" (8%), "considering its use" (5%) and "if cost effective" (4%). Summapy Two studies (Horvath and Shelton, 1982, and Kendrick, 1983) identified relatively small groups of Former Users among their respondents. Two studies (Gooch, 1964, and Horvath and Shelton, 1982) reported that Nonusers (including Former Users) stated their primary reason for not using PEPS was that they were satisfied with other methods. Unfortunately, no information was available regarding which other methods they were referring to. One study (Horvath and Shelton, 1982) reported that the primary factors that would influence Nonusers to consider using PEPS were research showing that it was effective and court acceptance of the validity of PEPS. ommen As noted at the end of the Users section above, it is important to recognize that all of the above reported information regarding Former Users and Nonusers is at least 10 years old and was published well before the U.S. Congress passed the EPPA in 1988. Chapter III Methodology This study was conducted between 1989—1991 to address the use of PEPS by police agencies as an applicant selection procedure. The methodology of the study included mailing to police agencies survey questionnaires asking questions regarding PEPS usage. Due to funding limitations, two separate consecutive surveys were conducted rather than one large survey. Because agency size had been identified in three previous studies (reference Table 2) as a major factor associated with PEPS usage, it was decided that the samples of the two surveys would be selected based on agency size. The first survey, referred to as the large agency survey and conducted in 1989-1990, addressed the largest police agencies in the U.S., and the second survey, referred to as the small agency survey and conducted in 1991, addressed the remaining agencies. Research Questions Three research questions are investigated in this study. They are as follows: (1) To what extent do police agencies in the U.S. currently use PEPS as an applicant selection technique? (2) What major factors are related to police agency PEPS usage? (3) Why and how do police agencies use PEPS? 35 36 Variables Based on the literature review, the following independent variables were identified and included in this study: PEPS usage, agency size, agency type, geographic region of location, and the existence of state laws affecting the use of PEPS. The values for each of these independent variables are set forth in Appendix A. The above independent variables will be used to assess the dependent variables set forth below: Why Agencies Use PEPS Reasons for using PEPS; issues of greatest importance; reasons for discontinuing use How Agencies Use PEPS Circumstances surrounding usage; policies and procedures regarding usage Agency Evaluations of PEPS Applicant pass/fail proportions; admissions during PEPS examinations; evaluations (benefits, confidence levels, accuracy estimates, absolute importance, relative usefulness) Other Screening Techniques Used Extent of use of 13 common pre-employment tests and procedures Future Plfils to Use PEPS Plans to implement PEPS in the next one to three years; circumstances in which implementation of PEPS would be considered Design of the Instrument The questionnaire designed for this survey was a self— administered” mail survey' questionnaire consisting of questions organized into the following general categories: — Agency Demographic Data 37 - Usage (User, Former User, Nonuser) - User Questions - Former User Questions - Nonuser Questions — Other Techniques Used (User, Former User, Nonuser) An initial draft questionnaire was developed in 1988. It underwent several modifications after review by a number of prominent and knowledgeable polygraph examiners who were members of, and in a number of instances, on the Board of Directors of the American Polygraph Association and the American Association of Police Polygraphists. In addition, the questionnaire was pretested on a number of police officers and law enforcement officials. The final document, consisting of 41 questions, was completed in 1989 and used in the large agency survey. A copy of the questionnaire is located in Appendix B. In 1991, the same questionnaire was used in the small agency survey, with the exceptions that the answer choices to one question (#24) were increased and modified for clarification; one additional question (#41) was included to determine agency policy regarding release of PEPS testing results; and one additional answer choice was added to one question (#42) to determine the extent of use of drug testing as an applicant selection technique. These modified questions are presented in Appendix C. W In the large agency' survey’ the study' population consisted of a listing of 699 of the largest general purpose 38 (having full arrest powers) police agencies in the U.S. This listing was compiled by the Police Executive Research Forum (PERF) based (Hi the agency eligibility' criteria for membership in that organization (Carter and Sapp, 1990). The listing included all 49 state police/highway patrol agencies; all consolidated police agencies (26); all county sheriff departments with 100 or more swornemployees (169); and all municipal police departments serving populations of 50,000 or more (455). In the small agency survey, the study population was approximately' 16,000 agencies identified. in the 1986 Directory Survey of Law Enforcement Agencies, which is a listing of U.S. law enforcement agencies maintained by the BJS. S_a_mp_le In the large agency survey, the entire population of 699 agencies was surveyed. The sample in the small agency survey was derived through a series of steps as described below. In 1990, the BJS conducted a survey of U.S. law enforcement agencies as part of its Law Enforcement Management and Administrative Statistics (LEMAS) program. Using its 1986 Directory Survey of Law Enforcement Agencies, BJS developed a nationally representative sample of two groups of large and small agencies for its survey. The large agency group consisted of all agencies in the directory with 100 or more sworn employees, including all state agencies. The small agency group consisted of a systematic random 39 sample of all the remaining agencies in the directory stratified first by size of population served, and then by number of sworn officers. The resulting sample consisted of a total of 2,931 general purpose agencies, including 721 large and 2,210 small agencies (Hubble, 1990; Sweet, 1990). The LEMAS listing identified 721 large agencies; however, the PERF listing had identified 699 large agencies. The difference of 22 agencies was due to the different criteria used to define large agencies. The LEMAS and PERF listings were not compared to identify the 22 agencies. In the present study, only the large agencies in the PERF listing were included in the sampling process. Thus, it is possible that the 22 agencies identified only in the LEMAS listing were not included in the present study sample. In 1990, the BJS made the listing of 2210 small agencies in their sample available for use in the present study. A review of these agencies disclosed that due to the difference between the LEMAS and PERF definitions regarding agency size, 18 of the 2210 LEMAS small agencies had already been included in the PERF large agency listing and large agency sample. Therefore, they were excluded from the small agency sample. This left a total sample of 2192 small agencies (697 sheriff and 1495 local) for the present study's small agency survey sample. W The data for both surveys were collected by the staff of the American Polygraph Association Research Center, School of 40 Criminal Justice, Michigan State University, in the following manner. A letter of transmittal was prepared describing the nature and purpose of the study and reflecting that it was jointly sponsored by the American Polygraph Association and the American Association of Police Polygraphists. The letter assured that individual agency responses would be held in confidence and requested the. agency's support in participating in the survey. The letter also explained that all questionnaires were numbered to insure that follow—up mailings could be done efficiently and with minimal cost. Copies of the letter and the questionnaires, together with stamped, self—addressed return envelopes, were mailed to all of the chief law enforcement administrators identified in the two samples. In the large agency survey, the first mailing of questionnaires was made in August, 1989, with follow-up mailings to non-respondents in November, 1989, and February, 1990. In the small agency survey, the first mailing was made in January, 1991, with follow—up mailings to non—respondents in April and July, 1991. W In Table 4, the number and percent of questionnaires distributed and received in each survey are summarized. 41 Table 4 Number and Percent of Questionnaires Distributed and Received in the Large and Small Agency Surveys esti naires Lgrge Small Combined n_ E n_ 5’71 n_' fab. Number Distributed 699 100 2192 100 2891 100 Overall Responses 635 9 1 15 12 69 2147 74 Usable Responses 626 90 1482 67 2108 73 In the large agency survey, questionnaires were distributed to 699 agencies and 635 were returned (overall return rate of 91%) Eight of the respondents declined to participate in the survey and one respondent was eliminated as it was determined to be outside the population parameters. As a result, there were 626 usable responses (usable return rate of 90%). In the small agency survey, questionnaires were distributed to 2192 agencies and 1512 were returned (overall return rate of 69%). Nonusable responses included 14 agencies that declined to participate in the survey and 16 respondents who replied that the police agency to which the questionnaire was addressed no longer existed. As a result, there were 1482 usable responses (usable return rate of 67%). A total of 25 of the 1482 agencies reported that the number of sworn officers assigned were now in excess of 100 personnel. However, they were not excluded from the small agency survey. The reason for this was that the large agency 42 population had already been defined and identified by the PERF listing and the LEMAS sample, and the study population from which the small agency survey sample was drawn was defined as "all remaining agencies." Therefore, changes in agency manpower did not affect the large and small agency category definitions. The "combined" column in Table 4 displays the data for large anxi small agencies when i1: is added together, or combined. ERMA; combined. column. will also appear in subsequent tables t1) serve jprimarily' as 51 basis for organizing agency responses. Size Sharacteristics of Respondents Agency responses were reviewed to determine the size characteristics of the large and small respondents. Table 5 reflects size characteristics in terms of number of sworn employees and size of population served. 43 Table 5 Size Characteristics of Agencies that Provided Usable Responses in the Large and Small Agency Surveys Large Small Combined characteristic (N=626) (N: 14821 LN =2 108) B u r f Sworn Em lo ees n 621 1454 2075 Range 17 - 26,000 1 - 291 l - 26,000 Mean 491 19 160 Median 187 10 20 Total 305,21 1 26,962 332,173 By ngulation Served n 620 1455 2075 Range 50,000 - 28M 50 - 860,000 50 - 28M Mean 614,405 19,941 197,564 Median 135 ,000 7,800 18,000 Table 5 reflects that the number of sworn employees among large agencies ranged between 17 to 26,000 and among small agencies ranged between 1 to 291. There is an overlap between the two size groups and this is due to two reasons. First, as previously discussed, the large agencies were defined as all state agencies, sheriff agencies employing 100 or more sworn personnel and municipal agencies serving populations cflf 50,000 or nmme. Agencies meeting these criteria were documented in the PERF agency listing. However, the definition of small agencies was based on the LEMAS survey sample. The LEMAS sample used the BJS directory of 16,000 agencies as its population and defined large agencies as all agencies with 100 or more sworn employees as reported in the directory. Small agencies were then defined 44 as all remaining agencies in the directory. Second, in addition to the different size definitions, it was found that some respondents reported higher or lower numbers of sworn employees than the numbers that had been reported in the PERF and BJS directory listings. Table 5 also reflects that the populations served ranged between 50,000 to 28 million among the large agencies and between 50 to 860,000 among the small agencies. Here again the overlapping of population sizes between the two surveys was caused by the same factors that were described in the preceding paragraph regarding the overlaps in numbers of sworn employees. Agencv Type of Respondents Table 6 below presents the number and percent of usable responses received from the Large and Small Agency surveys broken down by agency type. 4 5 Table 6 Agency Types of Agencies that Provided Usable Responses in the Large and Small Agency Surveys Large Small Combined Qharacteristic (N=626) (N=1482) JN=2 108) n. :72 n 20. a 34 Local 406 65 89 1 60 1 297 62 County 1 66 26 469 32 635 30 State 49 8 2 -* 5 l 2 gm _4__l __1_1_1__§ _1_1_5__6 Total 625 100 1473 100 2098 100 *Less than 1%. Although all state agencies were thought to have been included in the large agency survey, the table reflects that two :respondents in tflua small agenCy’ survey' identified themselves as state level agencies. Chapter IV Results In this chapter the results are presented in five sections. Section A sets forth an overview of PEPS usage. Sections B, C and D then provide separate descriptions of Users, Formers Users and Nonusers, respectively, based on the study results. Section E presents comparisons between Users, Former Users and Nonusers. Segpion A - degyiew of PEPS usege Figure 1 below sets forth the number and percentages of agencies that reported they were Users, Former Users and Nonusers of PEPS in the large and small agency surveys and when the two surveys were combined. 83% 9o _, [:1 Users 1227 : 677 :3 . 62% I 532:,“ ° ‘ 386 - Nonusers 60 a I .. 50‘. 31% g 40 - 196 a: 30 j 13% 20: 199 4% 10 4 56 O “’5 Small Combined (N=1482) (N=2108) Figure 1 Number and Percent of Users, Former Users and Nonusers in the Large and Small Surveys and in Both Surveys Combined 46 47 In the large agency survey, 386 of 626 respondents (62%) indicated they were current Users of PEPS, 44 (7%) were Former Users and 196 (31%) were Nonusers. 1n: the small agency survey, 199 of the 1482 respondents (13%) indicated they were current PEPS Users, 56 (4%) were Former Users and 1227 (83%) had never used it. When the respondents from the two surveys were combined, they totaled 2108 agencies, 13% of all general purpose law enforcement agencies in time U.S. (BJS, 1990). Of the combined 2108 agencies, 585 (28%) were PEPS Users, 100 (5%) were Former Users and 1423 (67%) were Nonusers. To determine whether PEPS usage and agency size were associated, the large and small agency groups were compared regarding whether they were PEPS Users or Nonusers. In each size group, the Former Users were first combined with the Nonusers as they were not currently using PEPS. Then each size group was divided into Users and Nonusers. The results are reflected in Table 7 below. 48 Table 7 Comparison of Large and Small Agencies by PEPS Usage Large Small Combined Usage (N=626) (N=1482) LN=2108) a E1 n. E n. E Users 386 62* 199 13 585 28 Nonusers 240 38 1283 87 1523 72 Total 626 100 1482 100 2108 100 1Corrected for missing data. Percentages rounded to nearest whole percent. *Significant difference by agency size regarding PEPS usage. Analysis showed that there was a significant difference by agency size regarding PEPS usage [X2(l)=510.7, p=.0001, Phi=.49] . Large agencies were more likely to be PEPS Users and small agencies were :more likely t1) be INonusers. The Phi coefficient of .49 indicated that the relationship between the two variables was moderate. Because of the demonstrated relationship between PEPS usage and agency 5 ize , agency size is treated as an independent variable in Sections B, C and D when presenting data for Users, Former Users and Nonusers. For example, the data regarding Users in the Section B tables are presented in separate columns for Large and Small Users. And similarly, the data regarding Former Users and Nonusers in the Section C and D tables, respectively, are presented in separate columns for Large and Small Former Users and Nonusers. Additionally, the tables in all the sections contain a third column which displays the data for the large and small agencies when they 49 are added together or "combined." This column is used primarily as a basis for organizing agency responses in the tables. In some of the tables where individual item responses are set forth in percentages, the total responses may exceed 100%. This is due to the fact that some of the questions that agencies were asked in the survey questionnaires provided for multiple independent responses. In other words, some of the questions provided several response choices and instructed respondents to mark all that applied. Section B — Description of Users In this section, Large and Small Users are described in terms of User characteristics, why and how PEPS is used and User evaluations of PEPS. gheracteristics of users Presented below are the data regarding size characteristics of Large and Small User agencies. ize. Table 8 reflects Large and Small User data compared by two specific size characteristics: first, by the number of sworn employees reported by the Users in each survey, and then by the populations served by the Users in each survey. 5 0 Table 8 Size Characteristics of Large and Small Users Large Small Combined Size (N=386) (N=199) (N=585) By Number of Sworn Employees n 383 196 579 Range 17 - 8,414 2 - 170 2 - 8,414 Mean 447 35 307 Median 215 28 133 Total 171,094 6,836 177,930 By ngulation Served n 381 198 579 Range 50,000 - 12.4M 950 - 700,000 950 - 12.4M Mean 522,105 31,267 354,254 Median 142,000 17 ,500 85 ,000 As reflected in the table, the mean number of sworn employees in Large User agencies was 447 (Median=215) and the mean number of sworn employees in Small User agencies was 35 (Median=215). The size of populations served was 522,105 (Median=142,000) in Large User agencies and 31,267 (Median=17,500) in Small User agencies. Type. Table 9 below reflects Large and Small Users compared by agency type . 51 Tmfle9 Number and Percent of Large and Small Users Presented by Agency Types Large Small Combined Agency Type (N=386) (N: 199) (N=585) a E1 n E n E Ixxxd 262 68 156 79 418 72 County 91 24 35 18 126 21 Sand er 32 8 7 3 .§2___Z 'Toufl 385 100 198 100 583 100 Table 9 reflected that 68% of the Large Users were local agencies, 24% were county agencies, and 8% were state or other type agencies. Additionally, 79% of the Small Users were local, 18% were county, and 3% were state or other types. Why Police Agencies Use PEPS In this section data are presented regarding reasons for using PEPS and issues of greatest importance in PEPS. Reasons for use. Users were provided a list of 10 different reasons for using PEPS. They were asked to indicate, based on their own experiences, the extent of their agreement with each statement using the following scale: 1=Strongly Disagree, 2=Disagree, 3=Undecided, 4=Agree, 5=Stroneg Agree. The Large and Small User mean scores for each reason are reflected in Table 10, and the reasons are rank ordered according to the "combined" mean score values. 52 Table 10 Large and Small User Reasons for Using PEPS Presented by Mean Scores in Rank Order Large Small Combined fiason (N=386) (N=fi9) (N=585) 11 Mean1 R k n_ Mean Rank 11 Mean Rank Reveals info. not available otherwise 384 4.3 1 190 3.9 3 574 4.2 l Deters undesirables 382 4.1 3 191 4.1 1 573 4.1 2 Background easier to establish 384 4.2 2 190 4.0 2 574 4.1 3 Faster 384 4.0 4 190 3.8 4 574 4.0 4 More useful info 383 3.7 5 190 3.4 7 573 3.6 5 Identifies problem persons better 383 3.6 6 189 3.5 6 572 3.6 6 Easierto administer 383 3.5 8 167 3.5 5 550 3.5 7 Less expensive method 380 3.5 7 191 3.4 8 571 3.5 8 More effective than background investigation 381 3.4 9 191 3.3 10 572 3.4 9 Less faulty info than background investigation 379 3.3 10 191 3.3 9 570 3.3 10 1Corrected for missing data. Mean=Mean score of responses scored as 1=Strongly Disagree, 2=Disagree, 3=Undecided, 4=Agree, 5=Strongly Agree. Table 10 reflects that the highest three mean score values of both Large and Small Users were for the same three reasons. The reasons were that it reveals information not available otherwise (Large User M=4.3; Small User M=3.9); it deters undesirables (both Large and Small User M=4.1); and it makes background information easier to establish (Large User 53 M=4.1; Small User M=4.0). This indicates that the primary reasons why Large and Small Users used PEPS were similar. To determine the degree of similarity between the Large and Small User rankings of all the reasons, they were compared by using the Spearman rank correlation coefficient. This correlation coefficient, based on the ranks of the two sets of scores, reflected a very strong relationship between the two rank orders [rs=.86]. This indicates very high agreement on the reasons for using PEPS for both large and small agencies. It is noteworthy that the mean scores for all of the reasons ranged between Undecided (3) and Strongly Agree (5), indicating that on average both Large and Small Users agreed that all the listed reasons, regardless of rank order, were positive reasons which contributed to agency decisions to use PEPS. Issues of greatest importance. Users were provided a list of 16 different issues to be investigated in an applicant's background. They were asked to indicate, based on their experience, how important it is to use PEPS to investigate each of the issues using the following scale: 1=Very Unimportant, 2=Unimportant, 3=Important, 4=Very Important. The results are reflected in Table 11 by the mean score of agency responses to each issue and the issues are rank ordered in the table according to the "combined" mean score values. 54 Table 11 Large and Small User PEPS Testing Issues of Greatest Importance Presented by Mean Scores in Rank Order Large Small Combined Test Issues (N=386) (N=199) (N=58SL Q Mean] Rank 11 Mean Rank 11 Mean Rank Illegal drug use 383 3.9 1 190 3.7 l 573 3.8 l Feloniescommitted 381 3.6 2 191 3.5 2 572 3.6 2 Dishonesty prior employment 381 3.6 3 191 3.5 4 572 3.6 3 Accept/pay bribes 381 3.5 4 190 3.5 3 571 3.5 4 Use of excessive force 381 3.4 5 187 3.4 5 568 3.4 5 Alcohol abuse 380 3.4 6 191 3.3 6 571 3.4 6 Illegal sexual activity 375 3.3 8 188 3.2 8 563 3.3 7 Employment history 381 3.3 7 191 3.1 10 572 3.2 8 Misdemeanors committed 382 3.3 9 189 3.1 11 571 3.2 9 Involvement in subversive organizations 375 3.2 10 189 3.2 9 564 3.2 10 Mentalproblems 379 3.1 11 189 3.2 7 568 3.1 11 Medical problems 380 3.0 12 190 3.0 12 570 3.0 12 Physical disabilities 381 2.9 13 190 2.9 13 571 2.9 13 Finance/credit problems 380 2.9 14 188 2.9 14 568 2.9 14 Traffic violations 381 2.7 15 187 2.6 15 568 2.7 15 Homosexual activity 360 2.5 16 187 2.6 16 547 2.5 16 1Corrected for missing data. Mean=Mean score of responses scored as 1=Very Unimportant, 2=Unimportant, 3=Important, 4=Very Important. 55 As shown in the table, the two issues that received the highest mean score values by both Large and Small Users were illegal drug use (Large Users M=3.8; Small Users M=3.7) and felonies committed (Large Users M=3.6; Small Users M=3.5). This indicated that both Large and Small Users considered these two issues to be the most important PEPS testing issues. The Large and Small User mean score rankings of all of the issues were compared in order to determine the degree of similarity between them, and analysis reflected a very strong relationship [rs=.96]. This indicates very high agreement on the ordering of the importance of the issues for both large and small agencies. It is noteworthy that the mean scores for all of the issues ranged between Important (3) and Very Important (4), indicating that on average both Large and Small Users agreed that PEPS was of positive value in addressing all of them. Hog Pglice Agencies gse PEPS In this section, data are set forth regarding the circumstances of use of PEPS and User policies and procedures governing PEPS. W. Table 12 below reflects Large and Small User results regarding the circumstances of use of PEPS by Users. 56 Table 12 Circumstances of Use of PEPS by Large and Small Users Large Small Combined We (N=386) 4N=l99l IN=5851 Iym Instrument Q 5721 g E p % Polygraph 374 98 187 97 561 98 Voice stress 5 l 5 3 10 2 Beth 2 1 1 - 3 - Total 381 100 193 100 574 100 wt 'v n 3/91 a E u E Own examiner 268 70 46 24 314 55 Outside examiner 116 30* 146 76 262 45 Total 384 100 192 100 576 100 Nemper gf Owg Examiners Range 1-26 1-2 1-26 Median 2 l 2 Mean 2 1 2 Number gf Exams Last Five Years Range 5 - 10,000 1 - 400 1 - 10,000 Mean 779* 54 537 Median 400 28 200 Total 292,896 10,1 10 303,006 Nemmr of Exams Last 12 Months Range 0 - 1,764 0 - 120 0 - 1,764 Mean 176* 13 125 Median 100 6 45 Total 67,278 2,361 69,639 De Exams for Other Agencies 11 f/_01 g E e E Yes 87 23* 20 10 107 19 119 296 77 174 0 4]Q__8_1_ Total 383 100 194 100 577 100 Number of Exams Done for Other Agencies Last 12 Months Range 0-516 0—21 0-516 Mean 47* 9 41 Median 19 7 15 Total exams 4041 139 4180 57 Table 12 (cont'd) Wh se rA encies fl 1721 Rank 3 £1 Rank 11 $1 R k No examiners 85 82 l 135 94 1 220 89 1 Less expensive 46 54 2 49 56 2 95 55 2 Better trained 29 41 4 43 51 3 72 46 3 Confidential 33 42 3 32 36 4 65 39 4 se er encies u E1 3 9'2 g % Commercial 96 79* 69 47 165 62 Other police 12 10 66 45 78 29 ngbipation 13 11 12 8 25 2 Total 121 100 147 100 268 100 1Corrected for missing data. Percentages rounded to nearest whole percent. *Significant difference between Large and Small Users. Table 12 shows that Large Users were more likely than Small Users to use their own polygraph examiners to conduct PEPS exams [X2(l)2108.4,p2.0001;Phi=.43]. The Phi coefficient of .43 indicates that the relationship between agency size and use of their own examiners is moderate. Large Users employed more examiners per agency than Small Users (M22 vs. M21). Large Users conducted a significantly greater number of exams than Small Users during the five year period preceding the surveys (M2779 vs. M254) [one-tailt28.6,df2562,p=.0001] and during the 12 months preceding the surveys (M2176 vs. M213) kme4m1 t28.6, df=557, p2.0001] . Large Users were more likely than Small Users to conduct PEPS exams for other law enforcement agencies [X2(1)213.1, p=.0003;Phi=.15]. However, the Phi coefficient of .15 indicates that the relationship between agency size and conducting exams for other agencies is very slight. Nevertheless, Large Users conducted a larger number of exams for other agencies 58 than Small Users (M247 vs. M29) [one-tail t=1.9, df2100, p2.0312] during the 12 months preceding the surveys. The Users that did not use their own examiners to conduct PEPS exams, but instead used outside examiners, were asked why they did so. Both Large and Small Users reported that their primary reason was because they had no examiners of their own (Large Users — 82%; Small Users - 94%). It was found that Large Users that used outside examiners were more likely than Small Users to use commercial agencies and that Small Users were more likely to use other police agencies [X2(2)239.7, p2.0001] . Policies and procedures. Survey results pertaining to a number of User policies and procedures regarding PEPS are set forth in Table 13. Table 13 PEPS Policies and Procedures of Large and Small Users Large Small Combined Relicy/Procedure (N=386) (N2199l (N=585) wheisflfested fl %1 n E Q E All sworn 380 99 180 90 560 96 Allcivilian 206 54 65 33 271 46 Critical/special 206 54 N/A2 N/A Some sworn N/A 13 7 N/A Some civilian N/A 33 17 N/A Other - - 10 5 10 2 When Tgld of Exam Q %1 Rank n % Rank p % Rank Upon inquiry about job 306 80 l 139 70 l 445 76 1 With application form 230 60 2 85 43 2 315 54 2 Inmedia announcement 129 34 3 71 36 3 200 34 3 After turn in application 64 17 4 40 20 4 104 18 4 Afterallstepsdone 52 14 5 29 15 5 81 14 5 When Test Administered n %1 Rank 11 E Rank p E Rank Before medical exam 288 79 l 115 66 1 403 75 1 Before psychological interview 261 75 2 104 64 2 365 71 2 Before background investigation 212 60 3 66 37 4 278 52 3 Before oral board 163 47 4 65 38 3 288 44 4 Refusaltg Take Exam 11 $1 3 E 11 E Automatic reject 346 92 152 89 498 91 Use other device 11 3 11 7 22 4 No penalty 15 4 2 1 l7 3 Delay process 5 1 5 3 _10__2 Total 377 100 170 100 547 100 No policy N/A 20 N/A 20 N/A 6 0 Table 13 (cont'd) Re-exam Pglicy Q 5721 p E p E With approval 124 33 58 30 182 32 By exception 132 35 48 25 180 32 Never 49 13 33 17 82 14 Automatic 40 11 24 13 64 11 Applicant request _3_2_ _8_ __2_9_ _15 E1 _11_ 'Toufl 377 100 192 100 569 100 How Testing Used 11 fil Rank p E Rank p E Rank Verify application/ background info 366 97 l 184 98 1 550 97 1 Develop unique info 338 91 2 149 81 2 487 88 2 Verify questionable/ incomplete info 299 82 3 135 76 3 434 80 3 Substitute for background inves. 7 2 4 2 1 4 9 2 4 Release gf Test Results p % [1 %1 p % To applicant N/A7- 106 60 N/A Not outside agency 89 57 No policy 52 40 To other agencies- exceptional 56 34 To other agencies upon request 24 14 1Corrected for missing data. Percentages rounded to nearest whole percent. 2N/A2Question or option not asked or not available in this study. As reflected in Table 13, 99% of the Large Users and 90% of the Small Users require all applicants for sworn positions to take PEPS exams, indicating strong consensus among Users regarding the routinized use of PEPS in police applicant screening. However, the percentages of both Large and Small Users requiring all applicants for civilian positions to take PEPS exams were not nearly so high (54% and 33%), indicating 61 that Users in general were motivated to use PEPS primarily to screen police officer applicants. Users were asked to identify at which of five stages in the screening process they made applicants aware of the need to complete a PEPS exam, and to check as many stages as applied. As indicated in Table 13, the rank orderings of both Large and Small Users were identical. The great majority of USers said the PEPS exam requirement was set forth in the initial stage when information is distributed upon inquiry about job openings. The lowest ranked stages were after applications were turned in and after all steps were completed. Upon being asked when they administered PEPS exams in their applicant screening process, the majority of both Large and Small Users administered PEPS before a medical exam (79% and 66%) or a psychological interview (75% and 64%). While a majority of Large Users (60%) administered it before a background investigation, only a minority of Small Users (37%) did so. However, only a minority of both Large and Small Users administered PEPS before an oral board (47% and 38%). When asked what their policy was regarding an applicant's refusal to take a PEPS examination, the great majority of both Large (92%) and Small Users (89%) responded similarly that refusal leads to automatic rejection. This indicates a consensus that in most User agencies, regardless of size, PEPS was considered to be a mandatory requirement 62 for applicants, and that alternatives to PEPS were not offered. In the small agency survey, agencies were provided an additional option to indicate that they had no policy regarding refusals, and 20 agencies selected that option. When asked what their policy was with respect to a re- examination of an applicant who is reported to be "deceptive" to one cut more important issues (without significant admissions) in an initial polygraph examination, only a very small percentage of Large and Small Users said a second exam was administered automatically (11% and.13%n. The great majority of both Large and Small Users (89% and 87%) placed at least some restrictions on retesting applicants. When asked how they used PEPS testing, the rank ordering of the four answer choices was identical for both Large and Small Users. The great majority said it was used to verify basic information derived from tflua applicant (Large297%, Sma11298%); to develop unique information not revealed by other selection devices (Large291%, Small281%); and to verify questionable or incomplete information provided by other selection devices (Large282%, Small276%). Very few Large or Small Users (Large22%, Small21%) indicated they used PEPS as a substitute for a background investigation. User Evaluations of PEPS In this section, data are presented regarding applicant PEPS pass/fail proportions, applicant admissions during PEPS exams and agency evaluations of PEPS. 63 Applicant passlfail proportions. Users were asked to indicate the approximate percentages of applicants who "passed" and "did not pass" PEPS exams during the 12 months preceding the surveys. Their responses are displayed by number and mean percentage in Table 14 below. Tmflel4 Proportions of Large and Small User Applicants Who Passed and Did Not Pass PEPS During a 12 Month Period Large Small Combined Item (N2386) (N2199) (N2585) p m1 a ME .11 ME Passed all/eligible 332 49 170 63 502 54 Disqualified/did not pass 337 25 164 11 501 21 Deceptive 152 21 166 11 318 16 Refused/discontinued 332 1 165 1 497 1 Passed PEPS N/A2 177 75 N/A 1Corrected for missing data. Percentages rounded to nearest whole percent. 2N/A2Question or option not asked or not available in this study. With regard to applicants who "passed" the PEPS exam, Table 14 discloses that 49% of all Large User and 63% of all Small User applicants passed all selection techniques, including PEPS, and were found to be eligible for employment. About 25% of the Large User applicants and 11% of the Small User applicants were reported to have been disqualified from consideration because of the PEPS exam outcome. Additionally, 21% of the Large User applicants and 11% of the 64 Small User applicants were found to be "deceptive" during PEPS testing. Finally, only 1% of both Large and Small User applicants either refused txa undergo PEPS testing' or discontinued an examination in progress. In the small agency survey only, agencies were asked what percentage of their applicants had successfully "passed" the PEPS, regardless of whether or not they passed other techniques or were eligible for hire. A total of 177 agencies responded that an average of 75% of their applicants had done so. Admissions. In the large agency survey, Users were presented with a list of the crime types of burglary, arson, robbery, rape and homicide and were asked whether they ever had an applicant admit to any unsolved crimes in these categories during PEPS exams. In addition to selecting as many of the listed crime types as applied, a number of agencies wrote in additional types of crimes in an "other" answer choice. The ones most frequently written in were drug abuse, larceny and sex offenses. Therefore, in the subsequent small agency survey, the list of crime types was expanded to include these also. The results are set forth in Table 15 below. All results are presented by the number and percentage of agencies that responded to each category and they are rank ordered by percentage . 65 Table 15 Large and Small User Applicant Admissions During PEPS Examinations Large Small Combined Crime (N=386) (N239) (N=585) p E1 m ll 24 m n 72 Rank. Burglary 250 70* 1 56 31 3 306 57 N/A Arson 150 44* 2 19 11 5 169 33 Robbery 129 38* 3 9 5 7 138 27 Rape 113 34 4 10 6 6 123 24 Homicide 31 10 5 2 1 8 33 7 Drug abuse 76 N/A7- 124 673 1 N/A Larceny 60 N/A 113 62 2 N/A Sex offenses 30 N/A 33 19 4 N/A 1Corrected for missing data. Percentages rounded to nearest whole percent. 2N/A2Not offered as a "Yes" or "No" option. Responses were written in by agencies in the "Other" option. 3Offered as a "Yes" or "No" option in the small agency survey only. *Significant difference between proportions of Large and Small Users. As exhibited in Table 15, both Large and Small Users reported that applicants had admitted to unsolved serious crimes in all the crime categories during PEPS exams. It is noted that significantly larger proportions of Large Users than Small Users reported admissions in three of the five originally listed categories [burglary - 70% vs. 31%, 225.57; arson - 44% vs. 11%, 222.75; robbery - 38% vs. 5%, 222.02; rape - 34% vs. 6%, 221.8; homicide - 10% vs. 1%, 220.43] . Eyaluations gf PEPS. Table 16 below sets forth User evaluations of PEPS based on their experience with the technique. Tmfle16 Large and Small User Evaluations of PEPS Large Small Combined Item (N2386 (N2199) (N2585) Emma p E1 Rank 3 E Bank .11 E MI More honest applications 314 86 1 160 86 1 474 86 l Higherquality hires 293 83 2 143 78 3 436 81 2 Fewer undesirable applicants 270 76 3 158 85 2 428 79 3 Fewer misconduct problems 191 60 5 120 69 4 311 63 4 Fewer internal problems 193 60 4 116 66 5 309 62 5 Fewercomplaints 156 51 6 82 53 6 238 52 6 Lower turnover 144 45 7 74 43 7 218 44 7 Qenfidence in Testing 3 %1 p E p E Minimal (0-50%) 6 2 5 2 11 2 Fair (51-75%) 23 6 29 15 52 9 Moderate (76-85%) 86 22 63 33 149 26 Hi h 86-1 % 266 70 96 50 362 63 'Toud 381 100 193 100 574 100 Estimate of Accuracy of Testm' g p £791 3 (.79. D. % Less than 50% 2 1 4 2 6 1 50to 75% 20 5 23 12 43 8 76to 85% 53 14 24 13 77 13 86 to 95% 168 44 93 49 261 46 96 to 100% 135 36 46 24 M 'Toud 378 100 190 100 568 100 67 Table 16 (cont'd) Immrtapce pf PEPS in Selection Process ’ (Range from 1=Very Little Importance to 100=Extreme Importance) Number responses 380 191 571 Range 10 - 100 5 - 100 5 - 100 Mean 84 75* 81 Median 90 80 90 Relative Usefulness of Selection Procedures (1=Less Useful than Polygraph, 2=About the Same Usefulness, 3=More Useful than Polygraph) 11 Mean _R__k p Mean Rank p Mean Rank Background investigation 384 2.24 l 190 2.41 1 574 2.29 1 Psychology test 375 1.93 2 189 2.11 2 564 1.99 2 Psychologist interview 370 1.92 3 186 2.00 3 556 1.95 3 Psychiatrist interview 327 1.83 4 173 1.91 6 500 1.86 4 Personal interview 378 1.75 5 189 2.01 4 567 1.84 5 Selection board interview 374 1.69 6 189 1.94 5 563 1.77 6 1Corrected for missing data. Percentages rounded to nearest whole percent. As shown in Table 16, Users were provided with a list seven potential benefits of using PEPS and were asked indicate, based on their experience, whether each one was was not considered to be a primary benefit. The majority of to or of both Large and Small Users selected all but one (lower turnover) of the seven benefits as primary benefits. In addition, when their responses were rank ordered and compared there was a very strong correlation between the rankings of the two groups [r5293], indicating a very high agreement on the ordering of benefits for both Large and Small Users on agency 68 size. When the Large and Small User mean scores were rank ordered, it was noted that the top three benefits for both Large and Small Users (more honest applications, higher quality hires and fewer undesirable applicants) all concerned perceived effects of PEPS on the applicant screening process. The remaining four benefits for both Large and Small Users all related to perceived effects of PEPS subsequent to the hiring process. Users were asked how much confidence they felt they had in the results obtained from PEPS testing by selecting one of four answer choices (minimal20—50%; fair251-75%; moderate276- 85%; high286—100%). The great majority of Users said they had moderate to high (76—100%) levels of confidence in the results (Large Users=92%, Small Users=83%). To determine whether there was a relationship between agency size and level of confidence, the four answer choices were collapsed into two categories (minimal to fair20-75%, and moderate to high276—100%) and Large and Small User responses were compared. Analysis showed that Large Users were more likely than Small Users to have moderate to high confidence in PEPS testing [X2(1)=13.1,p=.0003;Phi=.15]. However, the Phi coefficient of .15 indicates that the relationship between agency size and level of confidence is very slight. Users were asked to estimate the accuracy of PEPS testing results as it was used in their agency by selecting one of five answer choices (less than 50%; 50—75%; 76-85%; 86-95%; and 96—100%). The great majority of Users estimated 69 the level of accuracy to be between 76-100% (Large Users=94%, Small Users=86%). 1k) determine ‘whether there: was a relationship between agency size and estimates of accuracy, the five answer choices were collapsed into two categories (less than 50%—75%, and 76-100%) and Large and Small User responses were compared. Analysis showed that Large Users were more likely than Small Users to have high accuracy estimates of PEPS testing [X2(l)211.3,p=.0008;Phi=.14]. However, the Phi coefficient of .14 indicates that the relationship between agency size and estimate of accuracy is very slight. When asked to rate on a 100 point scale, with 1 indicating very little importance and 100 indicating extreme importance, how important they considered PEPS to be in their selection process, both Large and Small Users indicated relatively high levels of importance (M284 and M275). However, the rating of Large Users was significantly higher than that of Small Users [one-tailedt24.8,df2569,p=.0001]. Finally, Users were asked to rate the usefulness of six other‘ selectitul procedures relative txa PEPS ‘using' the following scale: 12Less Useful than Polygraph; 2=About the Same as Polygraph; 3=More Useful than Polygraph. When the mean scores of the Large and Small User responses were rank— ordered and compared, a very strong relationship between the two orders was observed [r3283]. This indicated very high agreement on the rank ordering of ratings between Large and Small Users. The only technique that both Large and Small Users ranked as relatively more useful than PEPS was a 70 background investigation (mean scores CHE 2.24 and 2.41, respectively). Other Screening Techniques Users were presented with a list of 13 common tests and procedures other than PEPS that police agencies use in pre- employment screening. For each technique all respondents were asked to indicate whether or not they used them as a regular part of their applicant screening process. The responses by Users are set forth in Table 17 below by the number and percent of agencies that used them and they are rank ordered according to the "combined" percentage values. Tmflel7 Other Applicant Screening Techniques Used by Large and Small Users Large ' Small Combined Technique (N2386) EN2199) (N=585) p flngmk a fihlgmk 11 EZREE; Criminal history 363 99 2 194 99 1 557 99 1 Reference check 374 99 1 192 98 2 566 99 2 Background investigation 375 99 3 192 98 3 567 99 3 Medical exam 373 99 4 187 96 5 560 98 4 Application form 363 96 5 188 96 4 551 96 5 Oralboard 337 90 6 178 92 6 515 91 6 Psychologist/ Psychiatrist interview 323 86 7 145 76 7 468 82 7 Personality test 317 85 8 137 71 8 454 80 8 Physical agility 310 83 9 122 65 10 432 77 9 71 Table 17 (cont'd) Knowledge test 289 78 10 133 70 9 422 75 10 Aptitudetest 210 58 12 112 60 11 322 58 11 Civil service exam 227 61 11 68 36 12 295 53 12 Honesty test 48 13 13 40 22 13 88 16 13 Drug test 9 N/A2 132 693 N/A N/A 1Corrected for missing data. Percentages rounded to nearest whole percent. 2Not offered as a "Yes" or "No" option. Responses were written in by agencies in the "Other" option. 3Offered as a "Yes" or "No" option on the small agency survey only. Table 17 shows that at least 90% of all Users used the same six techniques in addition to PEPS in their applicant screening processes, and that at least 58% of them used 11 of the 13 listed techniques. When the rank orders of the Large and Small User responses were compared, there was a very strong correlation between the rankings of the two groups D§298L This indicated very high agreement on the ordering of the techniques by use for both Large and Small Users. In the large agency survey, in addition to selecting from the 13 listed techniques, a number of agencies wrote in additional procedures in an "Other" answer choice. The ones most frequently written in were drug tests, credit checks and driving records. As the drug test technique was of particular interest, the list of 13 techniques was expanded in the subsequent small agency survey to include drug tests as a "Yes" or "No" option. Drug tests were used by 69% of the responding Small Users. 72 Section C — Description of Former Users In this section, Large and Small Former Users are described in terms of Former User characteristics, when and why they discontinued PEPS, their evaluations of PEPS, what techniques they currently use, and future plans to use PEPS. Characteristics of Former Users Presented below are data regarding size characteristics and agency types of Large and Small Former Users. ize. Table 18 reflects Large and Small Former User data compared by two specific size characteristics: first, by the number of sworn employees reported by the Former Users in each survey, and then by the populations served by the Former Users in each survey. Table 18 Size Characteristics of Large and Small Former Users Large Small Combined Size (N244) (N256) LN2100) By Number of Sworn Employees n 43 55 98 Range 60 - 4,783 2 - 112 2 - 4,783 Mean 500 28 235 Median 17S 17 58 Total 21,489 1,540 23,029 P a 'on S ed n 44 55 99 Range 50,000 - 6.6M 950 - 175,000 950 - 6.6M Mean 466,932 23,518 220,591 Median 140,000 8,000 55,000 73 As displayed in the table, the mean number of sworn employees was 500 (Median=175) in Large Former User agencies and 28 (Median2l7) in Small Former User agencies. The mean size of populations served was 466, 932 (Median2140,000) in Large Former User agencies and 23,518 (Median28,000) in Small Former User agencies. IXQE- The number and percent of Large and Small Former Users compared by agency type are set forth below in Table 19. Table 19 Number and Percent of Large and Small Former Users Presented by Agency Types Large Small Combined Agepcy Type (N244) ( N256) (N2100) a E1 n. E n E Local 32 73 40 73 72 73 County 10 23 14 25 24 24 S er _;__5 __L__; 3 3 Tom] 44 100 55 UK) 99 100 Table 19 reflected that 73% of the Large Users were local agencies, 23% were county agencies and 4% were state or other level agencies. Additionally, 73% of the Small Users were local, 25% were county and 2% were state or other level agencies. 74 When and Why Former Ssers Discontinued PEPS In this section data are presented regarding the past circumstances of use of PEPS by Large and Small Former Users and the reasons why they discontinued using PEPS. Past circumstances of use. The type of polygraph instrument previously used by Former Users and the time frames during which they began and discontinued the use of PEPS is shown in Table 20. Table 20 Past Circumstances of Use of PEPS by Large and Small Former Users Large Small Combined Circumstance (N244) (N256) (N2100) f1fyp_e Instrument 3 %l p E p E Polygraph 43 98 53 96 96 97 Voice stress - - - - - - Bod: 1 2 2 4 3 3 'Toufl 44 100 55 100 99 100 Yem Began BEPS Number of responses 33 50 83 At least half began 1940 - 73 (52%) 1959 - 81 (54%) 1940 - 81 (53%) Remainder began 1973 - 89 (48%) 1981 - 90 (46%) 1974 - 90 (47%) Yea Qpit PEPS Number of responses 37 46 83 At least half quit 1965 - 82 (54%) 1972 - 87 (52%) 1965 - 87 (53%) Remainder quit 1982 - 89 (46%) 1987 - 90 (48%) 1982 - 90 (47%) 1Corrected for missing data. Percentages rounded to nearest whole percent. The years the Former Users reported they began using PEPS ranged from 1940 to 1990. Large Former Users started using the technique as early as 1940 and at least half (52%) 75 started using it In; 1973, whereas the earliest date that Small Former Users started using it was almost two decades later (1959) and at least half (54%) started using it by 1981. The years the Large Former Users reported they discontinued use ranged from 1965 to 1990, and about half (54%) had done so by 1982, whereas the earliest reported date that Small Former Users discontinued it was 1972, and at least half (52%) had done so by 1987. Thus, the majority of Large Former Users started and discontinued the technique earlier than the majority of Small Former Users. Whv Former Users discontinued use. Former Users were presented with a list of 11 reasons for discontinuing the use of PEPS and were asked to indicate whether or not they were reasons why they discontinued use. Table 21 displays the number and percent of agency responses to each reason and the reasons are rank ordered in the table according to the "combined" percent value. 76 Table 21 Large and Small Former User Reasons for Discontinuance of PEPS Large Small Combined Reason (N244) (N256) (N2100) p 5E1.BEM§ .fl EZIQEK fl ZZIQEL Lack confidence in test 9 25 4 20 43 1 29 35 1 Too controversial 12 33 2 16 33 5 28 33 2 Lack positive results 9 26 3 18 38 3 27 33 3 Cost too high 6 17 8 19 39 2 25 30 4 Accuracy not satisfactory 7 19 7 16 34 4 23 28 5 Prohibitive legislation 17 47 1 4 9 9 21 27 6 Lack of qualified examiners 9 26 4 12 27 6 21 27 7 Results not useful 6 l7 9 7 16 7 13 16 8 Too much resentment 9 25 6 3 7 11 12 15 9 Lackofapplicants 2 611 7 15 8 9 11 10 Prohibitive court order 5 14 10 4 9 10 9 11 ll 1Corrected for missing data. Percentages rounded to nearest whole percent. There were some clear differences between Large and Small Former Users regarding their reasons for discontinuing PEPS. The top reason for Large Former Users was prohibitive legislation (47%); yet this reason was ranked ninth (9%) by Small Former Users, indicating that it had minimal impact on their decisions to discontinue PEPS. On the other hand, the main reason the Small Former Users gave for discontinuing PEPS was a lack of confidence in the testing (43%), which was ranked fourth (25%) by Large Former Users. It is interesting 77 to note the distinctive natures of these two reasons. Prohibitive legislation implies that the agencies may have been forced to discontinue using PEPS, whereas a lack of confidence in the testing indicates that agencies may have discontinued the technique based on their own individual evaluations of its effectiveness. In order to determine the degree of similarity between the Large and Small Former User mean score rankings of all of the reasons for discontinuing PEPS, they were compared and the results reflected a very slight relationship ngaan. This indicates that there was very low agreement on the ordering of reasons for both Large and Small Users. Former User Evaluations of PEPS Table 22 below sets forth Former User evaluations of PEPS based on their experience with the technique. 78 Table 22 Former User Evaluations of PEPS Large Small Combined Item 119244) (N256) (N2100) Cenfidence ip Testing fl %1 p E p E Minimal to Fair (0-75%) 16 37 28 52 44 45 Moderate to High (fiZ6—100%) 27 63 26 48 53 55 Total 43 100 54 100 97 100 Estimate of Accuracy of Testing 3 9&1 p E p % Lessthan50%to75% 13 31 23 44 36 38 to 100% 29 62 30 56 5 62 Total 42 100 53 100 95 100 1Corrected for missing data. Percentages rounded to nearest whole percent. Former Users were asked to indicate how much confidence they felt they had in the results obtained from PEPS testing by selecting (Hue of four answer choices (minimal20—50%; fair251—75%; moderate276—85%; high286-100%). The majority of Large Former Users (63%) said they had moderate to high (76— 100%) levels of confidence in the results, but only 48% of the Small Former Users said so. To determine whether there was a :relationship Ibetween agency' size and level of confidence, the four answer choices were collapsed into two categories (minimal to fair20-75%, and moderate to high276— 100%) and the collapsed data are set forth in Table 22 above. Large and Small Former User responses were compared, and analysis showed no significant difference between responses based on agency size [X2(l)=2.1,p2.1501]. 79 Former Users were asked to estimate the accuracy of PEPS testing results as it was used in their agency by selecting one of five answer choices (less than 50%; 50-75%; 76-85%; 86-95%; and 96—100%). The majority of Former Users estimated the level of accuracy to be correct between 76-100% of the time (Large Former Users=69%, Small Former Users=56%). To determine whether there was a relationship between agency size and estimates of accuracy, the five answer choices were collapsed into two categories (less than 50%—75%, and 76- 100%) and the collapsed data are set forth in Table 22 above. Large and Small Former User responses were compared, and analysis showed no significant difference between responses based on agency size [X2(1)21.5,p2.2144;Phi2.13]. Screening Technigges Surrently Used Former Users were presented with a list of 13 conunon tests and procedures (other than PEPS or voice stress analysis) that gxfljxma agencies 'use 1J1 pre—employment screening. For each technique all respondents were asked to indicate whether or not they used them as a regular part of the applicant screening process. The results are set forth in Table 23 below by the percentages of agency responses to each technique and the techniques are rank ordered in the table according to the "combined" percentage values. 8 0 Table 23 Screening Techniques Currently Used by Large and Small Former Users Large Small Combined Technique (N244) (N256) (N2100) Q E1 _Rmiig n E EEK n E Bani Reference check 40 98 2 53 96 1 93 97 1 Criminal history 36 97 4 53 96 2 89 97 2 Background investigation 39 98 3 52 95 3 91 96 3 Medicalexam 41100 1 50 93 5 91 96 4 Application form 40 95 5 52 95 4 92 95 5 Oral board 38 90 6 47 87 6 85 89 6 Knowledge test 33 83 7 36 65 7 69 73 7 Physical agility 31 76 8 35 65 8 66 69 8 Psychologist/ ‘ Psychiatrist interview 29 74 9 29 57 9 58 64 9 Personality test 29 74 10 28 56 10 57 64 10 Aptitude test 26 65 11 24 47 12 50 55 11 Civil serviceexam 23 59 12 18 36 13 41 46 12 Honestytest 12 30 13 7 14 14 19 21 13 Drug test 1 N/A2 29 56311 30 N/A 1Corrected for missing data. Percentages rounded to nearest whole percent. 2Not offered as a "Yes" or "No" option. Responses were written in by agencies in the "Other" option. 3Offered as a "Yes" or "No" option on the small agency survey only. The table shows that at least 87% of both the Large and Small Former Users used the same six techniques in their applicant screening processes, and that at least 47% of both Large and Small Former Users used 11 of the 13 listed 81 techniques. When the rank orders of all of the Large and Small Former User responses were compared, there was a very strong correlation between the rankings of the two groups h§29fl, which indicated very high agreement on the ordering of techniques by use by both Large and Small Former Users. In the large agency survey, in addition to selecting from the 13 listed techniques, a number of agencies wrote in additional procedures in an "other" answer choice. The ones most frequently written in were drug tests, credit checks and driving records. As the drug test technique was of particular interest, the list of 13 techniques was expanded in the subsequent small agency survey to include drug tests as a "Yes" or "No" option. Drug tests were used by 56% of the responding Small Former Users. Future Plans to Use PEPS Former Users were asked several questions regarding their possible future use of PEPS. Specifically, they were asked to indicate whether they planned to implement a PEPS program in the next one to three years, and if so, for what reasons they were planning to do so. The number and percent of agency responses are reflected in Table 24 below. For agencies that indicated they were planning to implement PEPS, the percentages of their responses to each of the listed reasons are rank ordered in the table according to their "combined" percentage values. 82 Table 24 Large and Small Former User Plans to Implement PEPS Large Small Combined Isspe (N244) (N256) (N2100) l t lement in ne tpflfhmeYears Q E1 n E Q E YES 3 7 8 15 11 12 119 38 93 46 85 84 88 'Touu 41 100 54 100 95 100 Reesgns Why Planning 19 Implement fl %1 Rank _11 E Rank p E Rank Assist background investigations 4 67 1 9 9O 1 13 81 1 Reduce undesirable applicants 2 33 4 8 80 2 10 63 2 Save money 3 50 2 5 56 4 8 53 3 Increase in number of applicants 1 l7 5 6 60 3 7 44 4 Background inves. not feasible 1 17 6 5 50 6 6 38 5 Background inves. restricted - - - 5565 5336 Legislative! judicial actions 3 50 3 1 ll 8 4 27 7 Citizen complaints increased - - - 2 25 7 2 15 8 Budget increased - - - l 11 9 l 7 9 1Corrected for missing data. Percentages rounded to nearest whole percent. As displayed in the table, 7% of the 41 responding Large Former Users and 15% of the 59 responding Small Former Users planned to implement PEPS in the next one to three years. The primary reason for doing so for both Large and Small 83 Former Users was to assist in background investigations by having knowledge of possible problem areas before the background investigation starts. Former Users that were not planning to implement PEPS were asked whether there were any circumstances in which they would consider the use of PEPS, and if so, which circumstances. The number and percent of agency responses are reflected in Table 25 below. For agencies that indicated they would consider using PEPS, the percentages of their responses to (%Hfll of the listed circumstances are rank ordered in the table according to their "combined" percentage values. Circumstances in which Large and Small Former Users Indicated that the Use of PEPS would be Considered 84 Table 25 Large Small Combined Qimpmstance LN244) LN=56) (N 2100) W 1d onsi er Implementing _Q 721 p E p E Yes 17 46 25 61 42 54 Me 20 54 l 39 36 46 Total 37 100 41 100 78 100 Circpmstances in which wgulg Qonsider Q E1 Rank p E Rank p E Bank Research showing effectiveness 11 69 1 19 76 1 30 73 1 Background inves. restrictions 9 60 2 18 72 2 27 68 2 Favorable court decision 9 56 3 14 54 5 23 55 3 Improved examiner qualifications 5 33 5 15 63 3 20 51 4 Increased citizen complaints 6 43 4 13 57 4 19 51 5 Budget increase 5 33 5 10 43 6 15 39 6 Licensing legislation 4 27 6 9 38 7 13 33 7 Increase in applicants 2 14 7 6 25 8 8 21 8 1Corrected for missing data. Percentages rounded to nearest whole percent. As exhibited in the table, 46% of the 37 responding Large Former Users and 25% of the 41 responding Small Former Users said they would consider use of the technique. The two main circumstances in which both Large and Small Users reported they would consider doing so were (1) research evidence showing that polygraph testing is effective as a 85 pre-employment screening device; and (2) if restrictions were placed on background investigations. When the rank orders of all of the Large and Small Former User responses were compared, there was a very strong correlation between the two groups [r5290], which indicated very high agreement on the ordering of the responses by both Large and Small Users. Section D — Description of Nonusers In this section, Large and Small Nonusers are described in terms of characteristics of Nonusers, what techniques they currently use, and future plans to use PEPS. characteristics of Nonusers Presented below are survey results regarding size characteristics and agency types of Large and Small Nonusers. ize. Table 26 below reflects Large and Small Nonuser data compared by two specific size characteristics: first, by the number of sworn employees reported by the Nonusers in each survey, and then by the populations served by the Nonusers in each survey. 86 Table 26 Size Characteristics of Large and Small Nonusers Large Small Combined Size (N2196) (N21227) (N21423) By Numlmr of Sworn Employees n 195 1203 1398 Range 50 - 26,000 1 - 291 l - 26,000 hfian1 578 15 94 Median 165 8 10 Total 112,628 18,586 131,214 B Po ulation Served n 195 1202 1397 Range 50,000 - 28M 50 — 860,000 50 - 28M Mean 828,021 17,912 131,991 Median 125,000 6,300 9,000 As reflected in the table, the mean number of sworn employees was 578 (Median=165) in Large Nonuser agencies and 15 (Median28) in Small Nonuser agencies. The mean size of populations served was 828,021 (Median2125,000) in Large Nonuser agencies and 17,912 (Median26,300) in Small Nonuser agencies. _T_‘yp_e. Table 27 below reflects the number and percent of Large and Small Nonusers compared by agency type. 87 IaMe27 Number and Percent of Nonusers Presented by Agency Types Large Small Combined AgeneyType (N2196) (N21227) (N21423) n E1 11. E n. E Local 112 58 695 57 807 57 County 65 33 420 34 485 34 ta er 12 2 105 9 20 2 Total 196 100 1220 100 1416 100 1Corrected for missing data. Percentages rounded to nearest whole percent. The table disclosed that 58% of the Large Nonusers were local agencies, 33% were county agencies, and 9% were state or other level agencies. Additionally, 57% of the Small Nonusers were local, 34% were county and 9% were state or other agencies. Screening Technigges Currently used Nonusers were presented with a list of 13 common tests and procedures (other than PEPS or voice stress analysis) that police agencies use in pre-employment screening. For each technique all respondents were asked to indicate whether or not they used them as a regular part of the applicant screening process. The results are set forth in Table 28 below by the percentages of agency responses to each technique and the techniques are rank ordered in the table according to the "combined" percentage values. 88 Table 28 Screening Techniques Currently Used by Large and Small Nonusers Large Small Combined Technqu: LN=196) (N21227) (N21423) n E 1 Rank 11 E m a E BEE Reference check 186 99 4 1142 97 1 1328 98 1 Criminal history 173 97 5 1142 97 2 1315 97 2 Background investigation 187 99 1 1125 96 3 1312 96 3 Application form 187 99 3 1086 92 4 1273 93 4 Medical exam 188 99 2 967 84 6 1155 86 5 Oral board 168 89 6 979 86 5 1147 86 6 Physical agility 147 80 8 496 46 7 643 51 7 Knowledge test 127 73 10 479 45 8 606 49 8 Psychologist! Psychiatrist interview 149 83 7 452 42 9 601 48 9 Aptitude test 94 55 12 391 37 10 485 40 10 Personality test 129 74 9 341 33 11 470 39 11 Civil service exam 113 63 11 225 22 12 338 28 12 Honesty test 36 22 13 104 10 13 140 12 13 Drug test 11 N/A2 477 453 N/A 488 N/A 1Corrected for missing data. Percentages rounded to nearest whole percent. 2Not offered as a "Yes" or "No" option. Responses were written in by agencies in the "Other" option. 3Offered as a "Yes" or "No" option on the small agency survey only. The table shows that at least 86% of both the Large and Small Nonusers used the same six techniques in their applicant screening processes, and that at least 42% of both Large and Small Nonusers used nine of the 13 listed 89 techniques. When the rank orders of all of the Large and Small Nonuser responses were compared, there was a strong correlation between the rankings of the two groups D3273L which indicated high agreement on the ordering of the techniques by use for both Large and Small Users. In the large agency survey, in addition to selecting from the 13 listed techniques, a number of agencies wrote in additional procedures in an "other" answer choice. The ones most frequently written in were drug tests, credit checks and driving records. As the drug test technique was of particular interest, the list of 13 techniques was expanded in the subsequent small agency survey to include drug tests as a "Yes" or "No" option. Drug tests were used by 45% of the responding Small Nonusers. Future Plans to Use PEPS Nonusers were asked several questions regarding their possible future use of PEPS. Specifically, they were asked to indicate whether they planned to implement a PEPS program in the next one to three years, and if so, for what reasons they were planning to do emu The number and percent of agency responses are reflected in Table 29 below. For agencies that indicated they were planning to implement PEPS, the percentages of their responses to each of the listed reasons are rank ordered in the table according to their "combined" percentage values. 90 Table 29 Large and Small Nonuser Plans to Implement PEPS Large Small Combined Isspe (N2196L (N21227) (N21423) Plan to Implement in One to flhmYears _11 E1 D. E n E Yes 8 4 56 5 64 5 119 183 26 1155 5 1338 25 Tbml 191100 1211100 1402100 Reasons Why Planning to Implement Q E1 R k p E Rank p E Rank Assist background investigations 8 89 2 50 86 1 58 87 1 Reduce undesirable applicants 9 100 1 49 84 2 58 87 2 Save money 6 60 3 29 50 3 35 51 3 Increase in number of applicants 5 56 4 25 44 4 30 45 4 Background investigation restricted 2 22 6 23 38 5 25 36 5 Background investigation notfeasible 3 33 5 20 36 6 23 35 6 Iegislativd judicial actions 2 22 7 17 30 7 19 29 7 Citizen complaints increased 1118 489 588 Budget increased - - 9 5 9 8 5 8 9 1Corrected for missing data. Percentages rounded to nearest whole percent. As set forth in the table, only 4% of the 191 responding Large Nonusers and 5% of the 1211 responding Small Nonusers planned to implement PEPS in the next one to three years. The two primary reasons for doing so were to assist in 91 background investigations by having knowledge of possible problem areas before the background investigation starts, and to reduce undesirable applicants. Nonusers that were not planning to implement PEPS were asked whether there were any circumstances in which they would consider the use of PEPS, and if so, which circumstances. The number and percentage of agency responses are displayed in Table 30 below. For agencies that indicated they would consider using PEPS, the percentages of their responses to Efiflfll of the listed circumstances are rank ordered in the table according to their "combined" percentage values. 92 Table 30 Circumstances in which Large and Small Nonusers Indicated that Use of PEPS would be Considered Large Small Combined Circumstance (N2196) (N21227) (N21423) W id Implementing fl 5721 p E p E Yes 37 21 283 25 320 24 _NE 142 72 847 75 289 76 Total 179 100 1130 100 1320 100 Circumstances in which would Conside_r n E1 Rank p E Rank p E Rank Favorable court decision 25 78 2 204 78 2 229 78 1 Background inves. restrictions 24 73 3 205 78 l 229 77 2 Research showing effectiveness 24 80 1 193 75 3 217 75 3 Budget increase 16 53 4 183 71 4 199 69 4 Increased citizen complaints 11 37 7 177 69 5 188 66 5 Improved examiner qualifications 16 52 5 141 60 6 157 59 6 Licensing legislation 15 50 6 140 58 7 155 57 7 Increase in applicants 4 14 8 111 46 8 115 42 8 1Corrected for missing data. Percentages rounded to nearest whole percent. As reflected in the table, 21% of the 179 responding Large Nonusers and 25 % of the 1130 responding Small Nonusers said they would consider use of the technique. The top three circumstances in which they would consider doing so were (1) a major court decision favorable to such screening; (2) a restriction on their ability to do an adequate background 93 investigation; and (3) research evidence showing that polygraph testing is effective as a pre-employment screening device. When the rank orders of all the Large and Small Nonuser responses were compared, there was a very strong correlation between the two groups [r3286], which indicated very high agreement on the ordering of the circumstances for both Large and Small Nonusers. Section E - Comparisons Between Users, Former Users and Nonusers In this section, comparisons are made between Users, Former Users and Nonusers regarding agency characteristics, evaluations of PEPS, the use of screening techniques aside from PEPS, and future plans to use PEPS. Sharacteristics of gsers and Nonusers Users and Nonusers were compared below by size, agency types, geographic region of location and the existence of state laws affecting PEPS usage. For purposes of analysis, Former Users are combined with Nonusers. Sal—2e. As previously reflected in Figure 3 of Section A, there were a total of 585 USers (28%) and 1523 Nonusers (72%) (including 100 Former Users — 5%) in the study. As previously reflected in Table 7, comparison of large and small agencies by Users and Nonusers disclosed that Users were more likely to be large agencies and Nonusers were more likely to be small agencies, and that the relationship between PEPS usage and agency size was very strong. 94 13:92- In Table 31 below, the number and percent of Users and Nonusers are compared by agency type. Large agencies are distinguished from small agencies because only the large agency survey included state agencies and the small agency survey included all other (miscellaneous) agencies. Table 31 Comparisons of Large and Small Users and Nonusers by Agency Type Users Nonusers Combined _Agmy Type (N=585) (N21523) (N22108) Lerge Agencies 11 %1 p E p E Local 262 68 144 60 406 65 County 91 24 75 3 1 1 66 26 EEQEZEEQQI 32 8 21 ' 9 53 9 Total 385 100 240 100 625 100 Small Agencies ILocal 156 79* 735 58 891 60 County 35 1 8 434 34 469 32 Miscellaneous 7 3 106 8 1 l 3 8 Total 198 100 1275 100 1473 100 1Corrected for missing data. Percentages rounded to nearest whole percent. *Significant differences based on PEPS usage and agency type. For the large agencies, the table reflects that 68% of the Users and 60% of the Nonusers were local agencies; that 24% and 31%, respectively, were country agencies; and that 8% and 9%, respectively, were state or other level agencies. 95 Large Users and Nonusers were compared by the three agency types, and analysis showed that agency type did not differ significantly between them [X2(2)24.7,p2.09]. For the small agencies, the table reflects that 79% of the Users and 58% of the Nonusers were local agencies; That 18% and 34%, respectively, were country agencies; and that 3% and 8%, respectively, were miscellaneous agencies. Small Users and Nonusers were compared by the three agency categories, and analysis showed a sdgnificant difference between the agency types of Small Users and Nonusers [X2(2)=32.2,p2.0001]. Small Users were more likely to be local agencies and Small Nonusers were more likely to be county level and miscellaneous agencies. In sum, among large agencies, agency type was not associated with PEPS usage; however, among small agencies, agency type was associated with usage. Seogrephic region. In Table 32 below, Users and Nonusers are compared by U.S. Census Bureau geographic region of location. Displayed in the table are the number and percentage of Users and Nonusers located in each of the four U.S. Census Bureau regions (Northeast, North Central, South and West) identified previously in Chapter III. 9 6 Table 32 Comparisons of Users and Nonusers by U.S. Census Bureau Regions U.S. Census Bureau Users Nonusers Combined Region (N=585) (N21523) LN22108) n E1 D. E p E Northeast 62 11* 294 19 356 17 North Central 122 21 571 38 693 33 South 214 36 503 33 717 34 West 1M 155 10 342 16 Total 585 100 1523 100 2108 100 1Corrected for missing data. Percentages rounded to nearest whole percent. *Significant difference between Users and Nonusers. The table discloses a significant difference in the distribution of Users and Nonusers [X2(3)=180.O,p2.0001]. Users were more likely to be located in the West Region and Nonusers were more likely to be located in the Northeast and North Central Regions. In an effort to determine whether agency size was associated with these results, the same tests were conducted controlling for size. In Table 33 below, the number and percent of Large Users and Nonusers in each of the four Census Bureau regions are set forth, followed by the number and percent of Small Users and Nonusers in each of the four regions. 97 Table 33 Comparisons of Large and Small Users and Nonusers by U.S. Census Bureau Regions U.S. Census Bureau Users Nonusers Combined Region (N=585) (N21523L (N22108L W a E1 11 E a E Northeast 35 9* 65 27 100 1 6 North Central 57 15 83 35 140 22 South 162 42 57 24 219 35 West 132 34 35 14 167 27 Total 386 100 240 100 626 100 Sm 1A ncies a E11 a E 11 E Northeast 27 13 * 229 18 256 17 North Central 65 33 488 38 533 37 South 52 26 446 35 498 34 West 55 28 120 2 175 12 Total 199 100 1283 100 1482 100 1Corrected for missing data. Percentages rounded to nearest whole percent. *Significant difference between Users and Nonusers. Table 33 reflects that among the large agencies, there was a higher frequency of Users in the South and West Regions and Nonusers in the Northeast and North Central Regions [X2(3)291.4, p2.0001]. Among the small agencies, there was a higher frequency of Users in the West Region and Nonusers in the South Region [X2(3)255.8,p2.0001]. 98 In sum, the geographic region in which agencies are located appears to be associated with usage of PEPS and agency size. State laws. In Table 34 below, Users and Nonusers are compared by state licensing and limiting laws regulating the use of PEPS. State licensing laws are laws which essentially require polygraph examiners to obtain some type of license (requirements vary among states) which authorizes the conduct of polygraph exams with the state. State limiting laws are laws which place limitations on the use of the polygraph in various cases in the state. They may range from limiting its use in certain specified situations to prohibiting the use of the polygraph in all contexts. Displayed in the table are the number and percent of Users and Nonusers located in states with no licensing or limiting laws, in states with either type of laws, and in states with both types of laws, as identified in Appendix A. 99 Table 34 Comparisons of Users and Nonusers by the Existence of State Laws Affecting PEPS Users Nonusers Combined StateLews (N=585) (N21523) (N22108) p 2&1 E E2 E 22 No Laws 152 26* 200 13 352 17 Limiting Laws 82 14 326 21 408 19 Licensing Laws 155 26 531 35 686 33 Limiting and Licensing Laws 1 6 34 466 31 662 31 Total 585 100 1523 100 2108 100 ————_— 1Corrected for missing data. Percentages rounded to nearest whole percent. *Significant differences based on PEPS usage and agency size. The table displays a significant difference in the distribution of Users and Nonusers [X2(3)=64.0,p2.0001]. Users were more likely to be located in states with no licensing or limiting laws and Nonusers were more likely to be located in states with only licensing laws or only limiting laws. In an effort to determine whether agency size was associated with these results, the same tests were conducted controlling for size. In Table 35 below, the number and percent of Large Users and Nonusers in each of the four categories of state laws are set forth, followed by the number and percent of Small Users and Nonusers in each of the four categories. 1 0 0 Table 35 Comparisons of Large and Small Users and Nonusers by the Existence of State Laws Affecting PEPS Users Nonusers Combined State Laws (N=585) (N21523) (N22108L nge Agencies n E1 a E .11 E No Laws 86 22* 24 10 110 18 Limiting Laws 52 14 59 24 111 18 Licensing Laws 116 30 62 26 178 28 Limiting and Licensing Laws 132 34 25 40 227 36 Total 386 100 240 100 626 100 Small Agencies 1; E1 n E n E No Laws 66 33* 176 14 242 16 Limiting Laws 30 15 267 21 297 20 Licensing Laws 39 20 463 36 508 34 Limiting and Licensing Laws (A 32 371 2 435 30 Total 199 100 1283 100 1482 100 1Corrected for missing data. Percentages rounded to nearest whole percent. *Significant differences based on PEPS usage and agency size. Table 35 shows that, among the large agencies, there was a higher frequency of Users in states with no laws and Nonusers in states with only limiting laws [X2(3)=25.1,p=.0001]. Among the small agencies, there was a higher frequency of Users in states with no laws and Nonusers in states with only licensing laws [X2(3)=57.8,p=.0001]. 101 In sum, state laws regulating the use of PEPS are associated with the usage of PEPS and agency size. User and Former User Evaluations of PEPS Both Users and Former Users were asked how much confidence they had in the results obtained from PEPS testing. They were asked to select one of four answer choices (minimal20-50%; fair251—75%; moderate276—85%; and high286—100%) and the number and percent of their responses are displayed in Table 36 below. Table 36 Comparisons of Users and Former Users Regarding Their Levels of Confidence in PEPS Users Former Users Combined Confidence Level (N2585) (N2100) (N2685) n E1 Q E n E Minimal (0-50%) 1 l 2 21 21 32 5 Fair (5 1-75%) 52 9 23 24 75 1 1 Moderate (76-85%) 149 26 26 27 175 26 High (86—100%) 362 63 27 28 38 58 'Toud 574 100 97 100 671 100 1Corrected for missing data. Percentages rounded to nearest whole percent. As set forth in the table, the great majority of the Users (89%) and a majority of the Former Users (55%) indicated moderate to high (76—100%) levels of confidence. To determine whether agency confidence levels differed with respect to the use of PEPS and agency size, a two factor Analysis of Variance (ANOVA) was carried out. The two factors were Use of PEPS (User and Former User) and agency Size (large and small) and the dependent variable was the raw 102 confidence scores selected by the agencies. Their scores were treated as an interval level variable with a scale of 1 tt> 4, and. with. the ihigher scores indicating greater confidence. The main effects for Use showed that the mean confidence score of Users was significantly higher than for Former Users [F(1,667)287.9,p2.0001]. The main effect for Size reflected that the mean confidence score of large agencies was significantly higher than for small agencies HKhtfi7fid15, p2.0008]. The interaction between PEPS usage and agency size was not significant [F(1, 667)20.02, p2.8873]. The ANOVA table reflecting these results is set forth in Appendix D. Users and Former Users were asked what they thought was a reasonable estimate of the accuracy of PEPS results as it was used in their agencies for pre—employment screening. They were asked to select one of five answer choices (less than 50%; 50—75%; 76—85%; 86—95%; and 96—100%) and the number and percent of their responses are displayed in Table 37 below. 1 0 3 Table 37 Comparisons of Users and Former Users Regarding Their Estimates of Accuracy of PEPS Users Former Users Combined Aecuracy Estimate (N=585) (N2100; (N 2685) n EH E E2 h E; Less than 50% 6 1 26 27 22 3 50 to 75% 43 8 20 21 63 10 76 to 85% 77 l3 17 18 94 14 86 to 95% 261 46 31 33 292 44 Mg 181 32 11 11 192 2 Total 568 100 95 100 663 100 1Corrected for missing data. Percentages rounded to nearest whole percent. As set forth in the table, the great majority of the Users (91%) and a majority of the Former Users (62%) estimated the accuracy of PEPS results to be between 76 and 100%. To determine whether agency accuracy estimates differed with respect to the use of PEPS and agency size, a two factor ANOVA was carried out. The two factors were Use of PEPS (User and Former User) and agency Size (large and small) and the dependent variable was the raw accuracy scores selected by the agencies. Their scores were treated as an interval level variable with a scale of 1 to 5, and with the higher scores indicating greater accuracy. The main effects for Use showed that the mean accuracy score of Users was significantly higher than for Former Users [F(1, 659)270.2, p2.0001]. The main effect for Size reflected that the mean accuracy score of large agencies was significantly higher than for small agencies [F(1,659)=8.7,p2.0032]. The interaction 104 between PEPS usage and agency size was not significant Hfll, 6i”=013,p=7143L The ANOVA table reflecting these results appears in Appendix D. Use; egg Monuser Screening Techniques All agency respondents were presented with a list of 13 common tests and procedures (other than PEPS or voice stress analysis) that gxfljxm: agencies use 111 pre-employment screening, and they were asked to indicate whether or not they used each technique as a regular part of their applicant screening process. The issue of interest concerning this question was whether or not there was any difference between PEPS Users and Nonusers regarding the number of techniques that they used. Former Users were first combined with Nonusers and then Nonusers were compared with the PEPS Users that used each technique. It was noted that six of the techniques were used by at least 80% of both PEPS Users and Nonusers, and that the remaining seven were used by fewer than 80%. Therefore, in Table 38 below, the techniques are presented in two separate groups based on percentage of use (those used by at least 80% of the agencies and those used by fewer than 80%). The number and percent of agencies that used each of the techniques are also displayed and the techniques are rank ordered in each group according to the "combined" PEPS User and Nonuser percentage values. 105 Table 38 Comparisons of Users and Nonusers Regarding Screening Techniques Used Users Nonusers Combined Technique (N=585) EN21523) (N22108) p E1 Reply D E BEL 11 E El: Used by at Least 82% of All Agencies Surveyed Reference check 566 99 2 1416 93 l 1987 98 1 Criminal history 557 99 1 1399 92 3 1961 98 2 Background investigation 567 99 3 1398 92 2 1970 97 3 Application form 551 96 5 1360 90 4 1916 94 4 Medical exam 560 98 4 1243 82 5 1806 90 5 Oral board 515 91 6 1230 81 6 1747 88 6 e Fewer than 0% of 1A encies Su e ed Physical agility 432 77 3 707 47 l 1141 59 l Psychologist/Psychiatrist V interview 468 82 l 658 36 3 1127 59 2 Knowledge test 422 75 4 673 44 2 1097 58 Personality test 454 80 2 526 35 5 981 52 4 Aptitude test 322 58 5 534 35 4 857 46 5 Civil service exam 29S 53 6 379 25 6 674 36 6 Honesty test 88 16 7 159 10 7 247 14 7 1Corrected for missing data. Percentages rounded to nearest whole percent. In viewing the table, it can be seen that for each of the techniques in both groups the percentage of Users is higher than the percentage of Nonusers. To determine whether the differences were significant, the data were tested in several ways . 106 The group of six techniques used by at least 80% of all agencies surveyed was collapsed into a variable with two categories - agencies that used aflj. six techniques and agencies that used fewer than six (zero to five) techniques. Analysis showed a significant difference between the two categories of Users and Nonusers [X2(l)=39.7,p=.0001; Phi=.l4]. Users were more likely to use all six techniques and Nonusers were more likely to use fewer than six techniques. The Phi coefficient of .14 indicated that the relationship between the two variables was very slight. In an effort to determine whether agency size was associated with these results, the same tests were conducted controlling for size. Among large agencies there was no difference in the number of techniques used by PEPS Users and Nonusers [X2(1)=O.2,p=.6375]. However, among small agencies it was determined that PEPS Users were more likely to use all six techniques and that Nonusers were more likely to use fewer than six of the techniques [X2(1)=28.4, p=.0001;Phi=.14]. The Phi coefficient of .14 indicated that this relationship was very slight. In sum, among the six techniques used by at least 80% of all agencies, Large Users and Nonusers were likely to use similar numbers of techniques, but Small Users were likely to use more techniques than Small Nonusers. The group of seven techniques used by fewer than 80% of all agencies surveyed was collapsed into a variable with two categories - agencies that used more (five to seven) 107 techniques and agencies that used fewer (zero to four) techniques. Analysis showed a significant difference between the two categories of Users and Nonusers [X2(l)=190.2,p=.0001; Idm:30L Users were more likely to use more (five to seven) techniques and Nonusers were likely to use fewer (zero to four) techniques. The Phi coefficient of .30 indicated that the relationship between the two variables was slight. In an effort to determine whether agency size was associated with the results, the same tests were conducted controlling for size. Among both large and small agencies it was found that Users were more likely to use more (five to seven) techniques and that Nonusers were more likely to use fewer (zero to four) techniques. [Among large agencies, X2(1)=6.6,p=.0102;Phi=.10. Among small agencies, X2(1)=63.5,p=.0001; PhkaZl]. The Phi coefficient of .10 among large agencies indicated the relationship is very slight; however, the Phi coefficient of .21 among small agencies indicated the relationship is slight. In sum, among the seven techniques used by fewer than 80% of all agencies, PEPS Users were more likely than Nonusers t1) use more (five tx3 seven) other techniques, regardless of agency size. Former User and Nonuser Future Plans to Use PEPS Former Users and Nonusers were asked several questions regarding their possible future use of PEPS. Specifically, they were asked to indicate whether they planned to implement a PEPS program in the next one to three years. Agencies that 108 responded in the affirmative were then presented with a list of nine reasons and asked to indicate whether or not they would be factors in their decisions to implement PEPS. The number and percentages of agencies that responded to these questions are displayed below in Table 39. The reasons for planning to implement PEPS are arranged in rank order according to the Nonuser percentage values. l O 9 Table 39 Comparisons of Former Users and Nonusers Regarding Plans to Implement PEPS Former Users Nonusers _s_sue LN=100) (N=1423) Plan :9 Implement in One to Three Years 3 fl 1 n % Yes 11 12* 64 5 1:19 84 88 1338 95 Total 95 100 1402 100 For Agencies that Answered Yes to Amve - Reasons Why Planning to Implement g £1 3% n 17g Rank Assist background investigations 132 100 1 58 91 1 Reduce undesirable applicants 10 77 2 58 91 2 Save money 8 62 3 35 55 3 Increase in number of applicants 7 54 4 30 47 4 Background inves. restricted 5 38 6 25 39 5 Background inves. not feasible 6 46 5 23 36 6 Legislative/judicial actions 4 3 1 7 19 30 7 Citizen complaints increased 2 15 8 5 9 8 Budget increased 1 8 9 5 9 9 1Corrected for missing data. Percentages rounded to nearest whole percent. 2Although only 11 Former Users indicated they planned to implement PEPS in one to three years, a total of 13 Former Users responded to the question asking to indicate their reasons for doing so. *Significant difference between Former Users and Nonusers. As presented in Table 39, 12% (n=ll) of the Former Users and 5% (n=64) of the Nonusers indicated they planned to implement PEPS in the next one to three years. Although only small minorities of both groups were planning to do so , analysis disclosed that Former Users were more likely than Nonusers to have such plans [X2(1)=9.2,p=.0024;Phi=.08]. The Phi 110 coefficient of .08 indicated that the relationship between the two variables is very slight. The top three reasons given by both Former Users and Nonusers for planning to implement PEPS were (1) to assist in background investigations by having knowledge of possible problem areas before the background investigation starts; (2) to reduce undesirable applicants; and (3) to save money. Further, the Former User and Nonuser percentage rankings of all nine reasons were compared in order to determine the degree of similarity between them, and analysis reflected a very strong overall relationship D§:98L This indicated very high agreement between Former Users and Nonusers regarding the ordering of reasons why they plan to implement PEPS. As was shown in Table 39 above, the great majority of Former Users and Nonusers (88% and 95%, respectively) indicated that they were not planning to implement PEPS in one to three years. These agencies were then asked if there were any circumstances in which they would consider the use of PEPS. Agencies that responded in the affirmative to this question were then presented with a list of eight circumstances and asked to indicate whether or not they would be factors in their decisions to implement PEPS. The number and percentages of agencies that responded to these questions are displayed below in Table 40. The circumstances in which agencies would consider using PEPS are arranged in rank order according to the Nonuser percentage values. l l 1 Table 40 Comparisons of Former Users and Nonusers Regarding Circumstances in which PEPS would be Considered Former Users Nonusers Circumstance (N=100) LN=1423) Would Cpnsider Implementing 3 %1 p Z; Yes 42 54* 320 24 _N_o 36 46 989 76 Total 78 100 1309 100 For Agencies that Answered Yes to Above — Circumstances in Which They Would Cpnsider Implementing PEPS Q 221 Rank p :72 Rank Favorable court decision 23 55 3 229 78 1 Background inves. restrictions 27 68 2 229 77 2 Research showing effective 30 73 l 217 75 Budget increase 1 5 39 6 199 70 4 Increased citizen complaints 19 51 4 188 66 5 Improved examiner qualifications 20 51 S 157 59 6 Licensing legislation 1 3 33 7 155 58 7 Increase in applicants 8 21 8 1 15 42 8 1Corrected for missing data. Percentages rounded to nearest whole percent. *Significant difference between Former Users and Nonusers. As shown in Table 40, a little more than half (54%, n=42) of the Former Users and a little less than one—fourth (24%, n=320) of the Nonusers indicated they would consider implementing PEPS. Analysis disclosed that the difference between Former Users and Nonusers was significant [X2(1)=33.0, p=.0001; Phi=.15] and that Former Users were more likely than Nonusers to consider doing so. The Phi coefficient of .15 112 indicated that the relationship between the two variables is very slight. The top three circumstances given by both Former Users and Nonusers in which they would consider implementing PEPS were (1) a major court decision favorable to such screening; (2) if a restriction were placed on their ability to do an adequate background investigation; and (3) research evidence showing that PEPS testing is effective as a pre-employment screening device. Further, the Former User and Nonuser percentage rankings of all eight circumstances were compared in order to determine the degree of similarity between them, and analysis reflected a very strong overall relationship [g=83]. This indicated very high agreement between Former Users and Nonusers regarding the ordering of the circumstances in which they would consider implementing PEPS. Chapter V Discussion In Chapter I, the purpose of this study was identified as being three-fold: (1) To obtain current information regarding the extent of use of PEPS as an applicant screening technique by police agencies in the U.S. (2) To identify the major factors related to PEPS usage. (3) To conduct an analysis of why and how police agencies use PEPS. The first three sections of this chapter discuss the study results corresponding to each of the three purposes, and comparisons are made with the findings of pertinent previous studies. The fourth and final section addresses the direction of future research needs regarding PEPS. Section A — Extent of Use of PEPS Table 1. in Chapter II listed ten studies conducted between 1962—1991, that reported on PEPS usage among police agencies in the U.S. Although the sample sizes among these surveys varied widely, none of them approached the scope of the present study, which included 2,891 agencies (699 large and 2192 small - see Table 4), or 19% of all 15,430 state and local general purpose police agencies 1J1 the U.S. (BJS, 1990). The combined usable response rate CH? 73% in the present study (90% large and 67% small - see Table 4) included a total of 2108 agencies (626 large and 1482 small) that employed a total of 332,173 sworn officers (305,211 in 113 114 large agencies and 26,962 in small agencies — see Table 5). The responding agencies in the present study represented 14% of the 15,430 state and local general purpose police agencies and employed 60% of the total number of 556,791 sworn officers employed by those agencies (BJS, 1990). Based on its size and its sample selection procedures (the entire population of 699 of the largest agencies and a nationally representative random sample of the remaining agencies), the present study can be considered as the most representative estimate of PEPS usage and the study results are generalizable to all state and local police agencies in the U.S. The studies in Table 1 reflected that PEPS usage among police agencies increased consistently over the past three decades from a low of 16% of the sampled agencies in 1962, to an estimated high of 75% in one purposive sample in 1991. In the present study, however, the reported 62% (386/626) usage rate among large agencies vs. the 13% (199/1482) usage rate among small agencies (Table 7) clearly reflects that any meaningful estimate of the extent of PEPS usage must take agency size into account. As an example, it is important to recognize that the 386 Large User respondents employed 31% (171,094) of the 556,791 sworn officers in the U.S., and that the 199 Small User respondents employed only 1% (6,836). Additionally, although the combined 585 Large and Small Users accounted for only 4% of the total of 15,430 state and local 115 police agencies in the U.S., they employed 32% (177,930/556,791) of all sworn police officers in the U.S. Section B — Major Agency Characteristics Related to PEPS Usage The four agency characteristics that were examined in relation to PEPS usage in the present study were agency size, type, geographic region of location and the existence of state laws 511 the U.S. regulating PEPS usage. .All four characteristics were found to be associated with PEPS usage, and these findings are compared below with prior research results. With respect to size, three previous studies (Eisenberg, Kent and walls, 1973; Horvath and Shelton, 1982; Lopez, undated — see Table 2) reported that small agencies were less likely to use PEPS than large agencies. That finding agreed with the results of the present study, which found that agency size was clearly associated with the use of PEPS. Large agencies were significantly more likely than small agencies to use PEPS (62% vs. 13%) (Table 7). Because this relationship was so strong, size was controlled for when the relationship between PEPS usage and the type and geographic location characteristics were examined. With respect to type, three previous studies (Eisenberg, Kent and Walls, 1973; Roper, 1981; Horvath and Shelton, 1982) reported finding no relationship between agency type and the use of PEPS. However, one study (Ash, Slora and Britton, 1990), which was based on a relatively small sample of large 116 agencies, reported that municipal agencies were significantly more likely to use the technique than state police agencies. Although the categorizations of agency types used by the prior and present studies varied somewhat, the present study agreed with the majority of previous studies in finding that type was not associated with PEPS usage among large agencies. However, it did find a relationship between type and PEPS usage among small agencies, which had not been previously reported (Table 31). The relationship between geographic region of location and PEPS usage was examined by only one previous study (Roper, 1981), which involved large agencies only. That study found that PEPS usage was most prevalent in the South and West U.S. Census Bureau Regions, and the least prevalent in the Northeast Region. The present study reported the same findings among large agencies (Table 33). However, it also found that the geographic regions of locations of Small Users and Small Nonusers were different than for large agencies, and again this difference had not been previously reported. Thus, the present study findings were similar to most prior research regarding the relationship between PEPS usage and agency size. However, the present study also demonstrated that the size factor was related to agency type and geographic location characteristics regarding PEPS usage. The relationship between state laws regulating PEPS usage and the use of PEPS by police agencies was briefly addressed in only one previous study (Roper, 1981), which 117 involved large agencies only. That study mentioned that the use of PEPS may be tied to statewide requirements regarding its legality; however, the issue was not further pursued. The only other available research regarding this issue was conducted in private industry by Belt and Holden (1978), who surveyed the personnel directors of 400 major U.S. corporations nationwide regarding the use of PEPS in their personnel selection programs. The corporations were selected from Fortune's list of largest companies and usable responses were received from 143 (35.7%) personnel directors. A total of 29 (20.3%) of them said their firms used PEPS and 114 (79.7%) replied they did not. When Belt and Holden cross— tabulated the respondents' locations with the existence of state laws regulating polygraph examiner licensing and training requirements, they found that significantly greater proportions of firms in regulated states used PEPS than firms in nonregulated states. They concluded that the existence of state licensing and training laws regulating PEPS may be an important factor related to the use of PEPS in corporate personnel selection programs, and that such laws may serve to enhance the overall credibility of the technique. In the present study, however, quite different results were obtained (Tables 34 and 35). It was found that police agency Users were more likely to be located in states with no laws regarding PEPS, and that Nonusers were more likely to be located in states with only limiting laws or only licensing laws, depending on agency size. 118 In considering the divergent findings between the Belt and Holden (1978) and present study, it should be noted first that the Belt and Holden (1978) study was conducted more than 25 years ago. Additionally, as pointed out earlier in Chapter I, most legislative actions have distinguished between PEPS usage in the private and public sectors. Although the number of states regulating PEPS has increased since 1978, most states have exempted public agencies from proscriptions. against. testing 1J1 employment situations (Bureau of National Affairs, 1985; Ansley and Beaumont, 1992). Further, although the use of PEPS by corporations in the private sector was severely restricted by the EPPA in 1988, public sector police agencies were exempted from the EPPA limitations. It might be expected that the legislative exemptions made for police agencies would have encouraged them to use PEPS. However, almost half (17/44=47%) of the large Former Users reported that the primary reason why they discontinued PEPS was prohibitive legislation (Table 21). Thus, while some state licensing laws may serve to enhance the credibility of PEPS testing, other state regulatory laws limiting use of the technique may deter usage. Moreover, many Former Users and Nonusers indicated that one of the main circumstances under which they would consider implementing PEPS was if further restrictions were placed on their ability to do an adequate background investigation (Table 40). This suggests that agency decisions to use PEPS can be influenced 119 by legislation regarding other personnel selection techniques as well as PEPS. Finally, as the limited cross—sectional data in this present study is the only information available on police agencies regarding this issue, changes in agency usage over time related to legislative actions cannot be detected. It appears that this issue requires further study in order to more fully understand all of the related factors and their interactions. Section C - Whv and How Police Agencies Use PEPS In this section, issues are examined relating to why police agencies use PEPS, how they use it, why Former Users discontinued PEPS, User and Former User evaluations of PEPS, other screening techniques police agencies use, enui the future plans of Former Users and Nonusers for using PEPS. Why Police Agencies Use PEPS Four previous studies (Gugas, 1962; Yeschke, 1962; Blum, 1967; Horvath and Shelton, 1982) indicated that police agencies used PEPS because it had utility in revealing information about applicants that was not obtainable by any other means. In the present study, which reports agency experiences based on over 300,000 PEPS exams conducted during a five—year period, this also was the case. 'Nma primary reason why agencies reported using PEPS was that the type of information revealed was not available by any other means (Table 10). Thus, for three decades police agencies have consistently 120 identified the utilitarian value of PEPS as the main reason why they use it. Although previous studies did not specifically address the type of information that agencies considered to be of greatest importance in PEPS testing, the present study reported that the three primary issues were illegal drug use, felonies committed and dishonesty in prior employment (Table 11). It is noteworthy that all three of these issues bear directly on applicant honesty and integrity areas of concern regarding positions of public trust. When this is considered in conjunction with the consistently reported use of PEPS because it reveals information that cannot be derived from any other techniques, it suggests that PEPS contributes importantly to providing information in key areas of concern in the police applicant screening process. How Police Agencies Use PEPS In three previous studies (Eisenberg, Kent and Walls, 1973; Horvath. and. Shelton, 1982; Kendrick, 1983) the information reported regarding how PEPS is used was minimal. However, it was noted that when studies reported on similar areas of use, their findings were generally supportive. In areas where comparisons could be made, the present study findings were similar to the previous studies. For example, in comparing the results of the present study with the Horvath and Shelton (1982) study it was noted that both studies reported that the primary purposes of PEPS exams were to verify information and to develop new information. This 121 indicates consistent agreement among agencies during the past decade in this regard. Additionally, the Horvath and Shelton (1982) study reported that 74% cflfvall Users tested all sworn officer applicants and 14% tested all applicants for civilian positions. The present study percentages were 96% and 46%, respectively (Table 13), which were supportive of the Horvath and Shelton (1982) findings, and also reflecbad a large increase in the use of PEPS within User agencies during the past decade. The present study also conducted. more extensive inquiries than previous studies regarding how police agencies use PEPS. As a result, it was found that agency size was a major factor associated with the circumstances of use of PEPS (Table 12). However, it was also found that large and small agencies generally had similar policies and procedures regarding the use of PEPS (Table 13). This suggests that the differences between large and small agencies regarding how they use the technique is more a function of agency size that the way they view the utility of the technique itself. Why Former Users Discontinued PEPS Two previous studies reported that some agencies had identified themselves as Former Users of PEPS (6% of a nationally representative random sample in Horvath and Shelton, 1982, and 14% of a purposive sample in Kendrick, 1983). One of them (Horvath and Shelton, 1982) reported that of the 14 Former Users identified, four reported they 122 discontinued PEPS because of state prohibitive legislation, three because of revisions in their applicant selection process, two because of cost, and one because of dissatisfaction with results. The present study findings were similar. A total of 100 Former Users were identified (about 5% of the total sample — Figure 1); thus, the percentage of agencies that discontinued the use of PEPS during the past decade has remained low. Their reasons for discontinuing PEPS, although related to agency size (Table 21), were generally similar to those mentioned in the Horvath and Shelton (1982) study. However, 12% of the Former Users reported they planned to re—implement PEPS in the next one to three years (Table 24), and 54% of the remaining responding Former Users indicated they would reconsider implementing PEPS under certain circumstances (Table 25). This suggests that among the small group of identified Former Users, agency decisions regarding PEPS usage can be influenced by a variety of factors both related and unrelated to their experiences with the technique. Evaluations of PEPS Four previous studies reported that Users perceived positive and tangible benefits based on their use of PEPS. Gugas (1962) reported a 46% applicant rejection rate based on admissions made during PEPS exams. Gooch (1964) and Horvath and Shelton (1982) both reported the same three primary benefits perceived by PEPS Users (higher quality employees, fewer problems and lower turnover), even though the two 123 studies were conducted 18 years apart. Lopez (undated) reported that PEPS was rated above average or outstanding by 81% of the Users. The present study reported agency evaluations of PEPS in greater depth and included evaluations by both Users and Former Users. Regarding Users, both Large and Small Users rated the top three benefits of PEPS as more honest applicatiOns, higher quality employees and fewer undesirables (Table 11H. Significantly, the higher quality employees choice was the first ranked benefit in both the Gooch (1964) and Horvath and Shelton (1982) surveys mentioned above. The first and third ranked benefits in the present survey were not provided as options in the previous studies. Thus, higher quality employees has consistently been considered a major benefit of PEPS by police agencies for a period of time spanning three decades. In addition to the perceived benefits of PEPS, Users evaluated PEPS with regard to levels of confidence, accuracy and importance. Overall, the great majority of ratings were moderate to high (76—100%); however, Large Users were more likely to provide higher evaluations than Small Users (Table 16). This may be related to the higher number of applicants processed by Large Users, which typically tend to draw from larger and more widely diversified populations in large cities than Small Users, which generally recruit applicants from smaller, less diversified areas. Additionally, the tendency toward higher PEPS fail rates (Table 14) and the 124 higher percentages of admissions to unsolved crimes (Table 15) reported by Large Users could also contribute to higher levels of evaluations of PEPS. cher Screening Technigpes One previous study (Roper, 1981) indicated that Users tended.tx>'use more screening techniques than anusers in their applicant screening programs. The present study expanded this issue to include consideration of agency size also. The number of other techniques that agencies use was found to be related to both PEPS usage and agency size. While Large Users were more likely than Large Nonusers to use more of the less frequently used techniques, Small Users were more likely than Small Nonusers to use more techniques overall. This indicates that PEPS is not used as a substitute for other techniques. Instead, it appears that the more techniques that agencies (especially small agencies) use, the greater the likelihood that they use PEPS. That PEPS is used predominantly in conjunction with other screening techniques is supportive of the finding that PEPS provides information not otherwise revealed by other screening techniques. It also suggests that agencies may use PEPS in their screening programs not to be more efficient, but to be more thorough and complete in their applicant reviews. 125 Future Plans to Use PEPS Past research regarding future plans of Former Users and Nonusers to use PEPS essentially addressed two separate questions. The first question was whether they were considering its use. The second question was, if they were not considering its use, whether there were any circumstances under which they would do so. With respect to the first question, three previous studies asked Nonusers about the possibility of using PEPS in the future and found that between 11-25% were considering its use (Yeschke, 1962 — 22%; Gooch, 1964 — 11%; Kendrick, 1983 — 25%). In the present study, only a small proportion of the Former Users (12%) and Nonusers (5%) indicated that they had plans to implement PEPS in the next one to three years (Table 39). However, the two primary reasons that both Former Users and Nonusers gave for doing so (to assist in background investigations and to reduce undesirable applicants) were very similar to the primary reasons for using PEPS that were cited by Users (reveals information not otherwise obtainable, deters undesirables anxi makes background investigations easier t1) establish). 'Fhis finding further supports the reported perceived utilitarian value of the PEPS technique in personnel selection. With respect to the second question regarding future plans to use PEPS, one previous study (Horvath and Shelton, 1982) asked Nonusers (including Former Users) that were not planning to implement PEPS under what circumstances they 126 would consider doing so. The top three circumstances were (1) evidence/research showing it was effective; (2) court acceptance or a favorable court decision; and (3) if law/policies permit. The findings of the present study regarding the second question were supportive of the Horvath and Shelton (1982) findings. About 54% of the Former Users and 24% of the Nonusers that were not already planning to implement PEPS stated they would consider doing so (Table 40). Both groups cited circumstances that were similar to the Horvath and Shelton (1982) study (research, courts and further restrictions (n1 background. investigations). Yet, as discussed in Chapter I, the interim period between the two studies was when the judicial and legislative controversy over PEPS testing culminated in the EPPA, which imposed restrictions on PEPS usage in the private sector. It seems that, in spite of the legal activity, these issues remain paramount in the view of many Former User and Nonuser agencies. Section D — Summagy and Future Research Issues In Chapter I, it was stated that a current and comprehensive descriptive analysis regarding PEPS usage was needed so that it could be used by researchers, the legislatures and the courts to address the conflicting and controversial issues surrounding the technique, and also by 127 police administrators in making decisions to implement it in their agencies. The present study provides information regarding PEPS based on the experiences of police agencies that employ 60% of all the sworn officers in the U.S. Of these agencies, PEPS Users employed almost one-third (32%) of all the sworn officers in the U.S. and the information provided was based on more than 300,000 PEPS exams that they conducted during a five-year period. Some of the key findings were that agencies use PEPS primarily because it has utilitarian value in revealing information not otherwise obtainable regarding critical aspects of applicant honesty and integrity, and that it results in higher quality officers. Further, agencies that use PEPS also use a greater number of other selection techniques in their applicant screening programs than agencies that do not use PEPS. The use of PEPS in conjunction with other techniques is supportive of the finding that PEPS reveals unique information that other techniques do not provide. The study findings regarding the utilitarian value of PEPS have been reported with remarkable consistency by PEPS Users over a period of time spanning three decades. As the use of PEPS has increased significantly during this period (see Table 1), this indicates that the Nonusers that have implemented PEPS have experienced the same benefits as the longer-term Users. Additionally, these findings do not appear to have been significantly affected by the controversy 128 over PEPS or the legislative or judicial actions that have occurred during this period. All of these factors suggest strongly that the motivating reasons for the use of PEPS are well-established and that the benefits are repeatedly validated as more and more agencies implement it in their selection programs (Meesig and Horvath, 1993). The only data reported in this study that did not support the utility of PEPS was provided by a minority proportion of the group of 100 Former Users (Table 21). Some of the reasons they gave for discontinuing the technique included a lack of confidence in PEPS (25% of the Large Former Users and 43% of the Small Former Users), a lack of positive results (26% and 38%, respectively), and unsatisfactory accuracy (17% and 39%, respectively). Although the reasons cited by these agencies cannot be ignored, it is important to consider them in the context of the overall findings regarding Former Users. For example, other reasons given for discontinuing PEPS were related to issues broader than utility, such as the technique being too controversial (reported by 33% of both Large and Small Former Users,), the costs being too high (17% and 39% of Large and Small Former Users), and prohibitive legislation (47% and 9%, respectively). Additionally, about half of the Former Users (large agencies=63%; small agencies=48%) indicated they had moderate to high confidence in PEPS, and more than half (large agencies=69%; small agencies=56%) felt tflun: the accuracy of PEPS was between 76-100% (Table 22). Moreover, 129 more than half of the Former Users indicated they were either planning to implement it again in the near future (large agencies=7%; small agencies=15%)(Table 24) or that they would consider doing so given the proper circumstances (large agencies=46%; small agencies=61%)(Table 25). It seems clear that in general many Former Users retain a positive perception of PEPS, and that their reasons for having discontinued it were related to a variety of factors extending beyond the specific utility of the technique. While the information in this study provides a current and comprehensive description of PEPS usage among police agencies in the U.S., and it is the best research data presently available in this regard, its value beyond description of usage is limited. Although the study results are highly supportive of PEPS usage, they are necessarily based only on experiential data and they do not empirically prove or disprove the relative utility or validity of PEPS. As pointed out earlier in Chapter I, empirical data directly testing these issues is sorely lacking, despite the fact that the need for such research has been clearly identified for more than a decade as one of the primary reasons for agency reluctance to use the technique (Meesig and Horvath, 1993). Thus, while this study reports that police agencies in general have positive experiences using PEPS, and while it shows that these experiences are clearly different than what the critics might suggest, it cannot resolve these differences empirically. 130 The research findings of this present and prior studies that have been reported so consistently over the past three decades suggests strongly that these results have merit, and because of this they deserve further scrutiny. If the experiential data regarding PEPS is correct, then the proper use of PEPS should be encouraged and developed. One need only be reminded of the riots in Los Angeles following the Rodney King verdict in 1992 ("30 Days," 1992), and of the Michael Dowd corruption scandal in the New York Police Department in 1992 ("Corruption Scandal," 1992), to recognize the importance of issues such as honesty and integrity among police officers, and to see the impact that these issues have on public perceptions of the police. However, if the real value of PEPS has been overstated in the experiential data, then this must be recognized also, and police administrators must take this into account in developing their selection procedures. It is primarily through the conduct of empirical research that the contribution of PEPS as an applicant selection technique can be most accurately explored. In this respect, the experiential data in this study, in addition to describing current police PEPS usage, practices and policies, also helps to clarify and articulate some of the critical issues that must be addressed next in the PEPS research area. Two central research issues are framed here as a guide to future research regarding the relative importance of PEPS (Horvath and Meesig, 1993). 131 First, although the experiential data overwhelmingly reports that PEPS reveals unique information, what is the relative utility of that information? Compared to information provided by other techniques, is the PEPS information pertinent and important with regard to the selection of applicants? Second, what is the relative value of the information produced by PEPS and other screening techniques in predicting honesty and integrity in job performance? Is the information produced by PEPS more or less effective as a predictor than information provided by other techniques? It is hoped that the results of this study and the above research questions will serve as a useful reference guide in the future for researchers and for judicial, legislative and police decision—makers in evaluating the effectiveness of PEPS as a screening technique for police applicants in the U.S. APPENDICES APPENDIX A Independent Variables 132 APPENDIX A Independent Variables PEP sa e Users - police agencies that reported they currently used PEPS in their applicant selection programs Former Users - police agencies that reported they had used PEPS at some time in the past but had since discontinued its use Nonusers — police agencies that reported they have never used PEPS Agency Size Large Agencies — includes all state police, all consolidated countywide police agencies anmi all sheriff departments with 100 or more sworn employees, and all municipal police departments serving populations of 50,000 or more Small Agencies - includes all county sheriff departments and local (operated by municipal or county governments) police agencies with fewer than 100 sworn personnel. Agency type Local Police Agencies (includes city and township) County Agencies State Agencies (includes state police and state highway patrol) Other Agencies Geographic Region All agencies were grouped into four regions constructed by the U.S. Census Bureau to sectionalize the U.S. (U.S. Bureau of the Census, 1989), as follows: Northeast (9 states) — CT, ME, MA, NH, NJ, NY, PA, RI, VT North Central (12 states) — IA, IL, IN, KS, MI, MN, MO, NE, ND, OH, SD, WI 133 South (DC and 16 states) — AL, AR, DC, DE, FL, GA, KY, LA, MD, MS, NC, OK, SC, TN, TX, VA, WV West (13 states) — AK, AZ, CA, CO, HI, ID, MT, NV, NM, OR, UT, WA, WY Existence of State Laws State licensing laws are laws which essentially require polygraph examiners to obtain some type of license (requirements vary among states) which authorizes the conduct of polygraph exams with the state. State limiting laws are laws which place limitations on the use of the polygraph in various cases in the state. They may range from limiting its use in certain specified situations to prohibiting the use of the polygraph in all contexts. States that have laws regarding the licensing of polygraph examiners or limiting the use of polygraph testing are identified by Ansley and Beaumont (1992). Based on their information, all the states, including the District of Columbia, are grouped into four categories as follows: States with no licensing or limiting laws (7 states)— CO, FL, MS, NH, NY, OH, WY_ States with limiting laws only (DC and 13 states) - AK, CA, CT, DE, DC, HI, ID, MD, MN, NJ, PA, RI, WA, WI States with licensing laws only (16 states) - AL, AZ, AR, IN, KS, KY, LA, MO, NC, ND, OK, SC, SD, TN, TX, UT States with both licensing and limiting laws (14 states) - GA, IL, IA, ME, MA, MI, MT, NE, NV, NM, OR, VT, VA, WV APPENDIX B Large Agency Survey Questionnaire 134 THE USE OF DRE-EMPLOYMENT POLYGRAPH TESTING BY LAW ENFORCEMENT AGENCIES A joint research project of the: AMERICAN POLYGRAPH ASSOCIATION AMERICAN ASSOCIATION O? POLICE POLYGRAPHISTS APA POLYGRAPH RESEARCH CENTER, 560 Baker Hall, Michigan State University, East Lansing, Michigan 48824 GENERAL INSTRUCTIONS: This questionnaire 1&5 designed tx> collect information about the extent of, and circumstances in which polygraph testing is 'used in tflua pre-employment screening' process in law enforcement agencies. Most responses may be indicated by merely placing an 'X' or a check mark in the appropriate space; some however, require write-in responses. All responses are to be recorded directly on the questionnaire itself. ALL RESPONSES TO THE QUESTIONNAIRE WILL BE HELD IN STRICT CONFIDENCE AND INDIVIDUAL AGENCIES WILL NOT BE IDENTIFIED. ALL DATA WILL BE REPORTED ONLY IN THE AGGREGATE. A PARTICULAR RESPONSE OR SET OF RESPONSES WILL NOT BE ASSOCIATED WITH A PARTICULAR AGENCY. PART I 1. In what state is your agency located? 2. What is the approximate population your agency's jurisdiction serves? 3. Which term best describes your law enforcement agency? a ) City [ l d.) State Agency b.) County [ ] (Highway Patrol) [ ] c ) State Agency e ) Township [1 (Police) [ ] f.) Other [] 4. Approximately how many people are employed by your agency? Sworn Civilian 5. Does your agency now use polygraph testing or "voice stress analysis" in the applicant screening process? Yes [ ] (Skip to Part II, page 4) No [ ] (Go to question #6) 6. Has your agency ever used polygraph testing or voice stress analysis to screen applicants for employment? Yes [ ] If yes, go to question #7 No [ ] If no, skip to question #13 7. What type of equipment did your agency use for pre-employment screening? a.) Polygraph only b.) voice Stress Analyzer only (e.g., Psychological Stress Evaluator) c.) Both Polygraph and Voice Stress Analyzer [ ] f—‘H t—lL—J 8. 9. 10. 11. 135 In what year did your agency begin use of A. Polygraph testing B. Voice Stress Analysis In what year did your agency discontinue use of A. Polygraph testing B. Voice Stress Analysis For each of the statements below, not one of the reasons why your agency discontinued use of polygraph testing or voice stress analysis. A. B. C. The testing caused too much resentment Yes The testing was too controversial Yes The testing and/or equipment was prohibited by legislation Yes The cost of processing applicants was too high Yes The information developed was not useful in hiring decisions Yes The accuracy of the results was not satisfactory Yes The testing and/or equipment was prohibited by court order Yes Our agency developed a lack of confidence in this testing Yes The testing did not offer enough positive results to continue its use Yes Demand for, and shortage of applicants prohibited such high selectivity Yes Persons qualified to do this testing were no longer available Yes Other (please explain) [ indicate whether it was or ] l9__ __ l9 l9__ 19 was N0[] N01] N01] N0[] N0[] N01] N01] N01] N01] N0[] How much confidence do you feel you were able to place in the results obtained from polygraph testing when used by your agency? (Check only one) Minimal degree of confidence (0—50%) Fair degree of confidence (51%-75%) Moderate degree of confidence (76%-85%) high degree of confidence (86%-100%) (LOUD) I—wr—sr—nv—w HHHH 12. 13. 14. IF YOU COMPLETED QUESTION 14, PLEASE SKIP TO PART III, PAGE 10, 136 What do you feel is a reasonable estimate of the accuracy of polygraph testing results as it was used in your agency for pre-employment screening? (Check only one) Correct less than 50% of the time Correct between 50% and 75% of the time Correct between 76% and 85% of the time Correct between 86% and 95% of the time Correct between 96% and 100% of the time (DQOU‘O’ Does your agency have any plans to implement a polygraph screening program in the next one to three years? Yes [ ] If yes, go to question #14 No [ ] If no, skip to question #15 I—‘F‘HHH HHHL—JL—l For each statement below, indicate whether it is or is not a reason why your agency is planning to implement a polygraph screening program. A. Belief that such a program would save money Yes [ ] B. There has been an increase in the agency's budget appropriated for applicant screening Yes [ ] C. There has been an increase in the number of applicants Yes [ ] D. Your ability to do adequate background investigations has been restricted due to decreased funding Yes [ ] E. Sometimes background investigations are not feasible Yes [ ] F. To assist in background investigations by having knowledge of possible problem areas before the background investigation starts Yes [ ] G. To reduce the number of undesirable persons applying for positions Yes [ ] H. Such forms of testing have become more acceptable because of either legislative or judicial action Yes [ ] I. There has been an increase in citizen complaints about the behavior of your officers Yes [ ] J. Other (please specify) NO No No No No No No COMPLETE THE QUESTIONNAIRE. 15. Is there any circumstance in which your agency would consider the use of polygraph testing to screen applicants? Yes [ ] If yes, go to question #16 No [ ] If no, skip to Part III, page 10 TO 137 16. For each statement below, indicate whether it would or would not be a factor in your agency's decision to use polygraph screening. A. A major court decision favorable to such screening Yes [ ] No [ l B. Legislation that would provide for licensure of examiners Yes [ ] No [ ] C. A dramatic increase in the number of applicants for jobs in your agency Yes [ ] No [ ] D. A restriction on your ability to do an adequate background investigation Yes [ ] No [ J E. An increase in the number of citizen complaints about the behavior of your agency's officers Yes [ ] No [ ] F. An in crease in your agency's budget devoted to screening applicants Yes [ ] No [ ] G. Research evidence showing that polygraph testing is effective as a pre-employment screening device Yes [ ] No [ ] I. Other (please specify) PLEASE TURN TO PART III, PAGE 10, TO COMPLETE THE QUESTIONNAIRE ************************************************************************ PART II Instructions: Questions #17 to #40, which follow, are to be answered iny by Qgencies who NOW USE POLYGRAPH TESTING OR VOICE STRESS ANALYSIS IN THE APPLICANT SCREENING PROCESS. However, the term "polygraph examination" is used here in a generic sense to refer to testing with either a polygraph instrument or a "voice stress analyzer." Thus, we are interested in your responses regardless of the type of instrumentation used. l7. What type of instrumentation does your agency use for the purpose of pre-employment screening? a.) Polygraph only [] IL) Voice Stress Analyzer only (e.g., Psychological Stress Evaluator, Mark II, etc.) c.) Both Polygraph and Voice Stress Analyzer f—‘H 18. 19. 20. 21. 22. 23. 24. 138 Are all or most of the pre-employment polygraph examinations for your agency administered by your agency's own examiners? Yes [ ] If yes, go to question #19 No [ ] If no, skip to question #20 How many examiners does your agency employ who do pre—employment screening? [Skip to question #22] Who does this testing for your agency? a.) Examiners employed by another law enforcement agency kn) Examiners in a commercial (private) testing firm c.) A combination of both internal and outside examiners Why does your agency choose to have examinations conducted by examiners not employed by your agency? (respond to each statement) A. Agency does not employ its own examiner Yes [ ] No [ B. Outside examination is less expensive Yes [ ] No [ C. Outside examiners are better trained in pre—employment screening Yes [ ] No ] ] D. To protect the confidentiality of the information Yes [ ] No [ ] E. Other (please specify) In the last five years, what is the approximate number of pre-employment polygraph examinations conducted by (or for) your agency? In the last 12 months, what is the approximate number of pre—employment polygraph examinations that were conducted by (or for) your agency? In the last 12 months, what is the approximate percentage of applicants who were given pre—employment polygraph examinations who: A. were disqualified from consideration because of the polygraph examination results (either admissions or a deceptive test outcome) B. were found to be "deceptive" to one or more of the important issues covered during the polygraph examination C. either refused to undergo polygraph testing or discontinued an examination in progress D. who "passed" all selection devices, including polygraph test, and were found to be eligible for employment 25. 26. 28. 139 Does your agency conduct pre-employment examinations on applicants for other law enforcement agencies in your area? Yes [ ] If yes, go to question #26 No [ 1 If no, skip to question #27 Listed below are reasons some law enforcement agencies have reported using pre-employment polygraph screening. Based on your agency's experience with pre-employment polygraph testing, please indicate your extent of agreement with each statement using the following scale: Strongly Disagree (SD), Disagree (D), Undecided (U), Agree (A), or Strongly Agree (SA) with each statement. Polygraph Testing . . . SQ Q g A S A. is less expensive than other comparable screening devices [ l [ l [ l [ ] [ B. is faster in producing results than other methods [ ] [ l [ l [ l l C. reveals information not available by any other means [ l [ l [ l [ l [ D. makes background information about applicants easier to establish [ ] [ l [ ] l l [ E. is more effective than traditional background investigations [ ] [ ] [ ] [ ] [ F. produces more useful information about applicants than other methods [ ] [ ] [ ] [ ] [ G. produces less faulty information than background investigations [ ] [ ] [ ] [ ] [ H. deters some undesirable applicants from applying for police employment [ ] [ ] [ ] [ ] [ I. identifies potential problem personnel better than other methods [ ] [ ] [] [ ] [ J. is generally easier to administer than other screening devices [ l [ l [ ] [ l [ K. Other (please explain) Which applicants for positions in your agency, as indicated below, are asked to undergo pre—employment polygraph examinations? (Check all that apply) Applicants for civilian positions Applicants for sworn positions Applicants for "critical" or special positions Other (please specify) CLOUD) I—JI—lh—J 29. 30. 31. 32. What is your agency's policy regarding an applicant's refusal to take a pre—employment polygraph examination? 140 a.) Refusal to take the examination leads to automatic rejection for employment IL) Refusal to take the examination leads to use of other screening devices (i.e., background investigation, psychological exams) not normally required but does not lead to automatic rejection c.) Refusal to take examination results in no additional procedures but may delay processing d.) Refusal to take the examination is not penalized in any way e.) Other (please specify) (Check only one) ] What is your agency's policy with respect to a re-examination of applicant who is reported to be "deceptive" to one or more important issues (without significant admissions) polygraph examination? (Check one only) second exam second exam second exam second exam second exam (DQOU'W 3’3’3’3’3’ is is is is is administered (scheduled) automatically administered at the applicant's request given only with administrative approval given only in exceptional circumstances never provided The following statements are some of the ways in which law enforcement agencies make use of polygraph testing. these statements, your agency uses polygraph testing. A. To verify basic information derived from the application form and/or the background investigation Yes [ ] B. To verify questionable or incomplete information provided by other selection devices Yes [ ] C. To develop information not revealed by other selection steps or devices Yes [ ] D. To substitute for a background investigation Yes [ ] E. Other (please specify) in an initial For each of indicate whether or not it is one of the purposes No No an [ [ l—‘f-q A—dh—lh—Ih-JI—J I—Jt—I When are applicants for employment made aware of the need to complete a pre—employment polygraph examination? apply) a.) In newspaper or other media announcements of openings b.) In the information distributed to all persons inquiring about job openings c.) In the information distributed with agency or civil service application forms 0.: (Check all that After the application material is filled out and turned in e.) Other (please specify) [ [ [ [ l l 1 ] 33. 34. 35. 141 The following statements have been reported as being among the benefits of using pre-employment polygraph screening. Indicate for each of these whether or not your agency's experience shows it as a primary benefit. A. Application forms tend to be more honestly completed Yes [ ] No [ ] B. There are fewer undesirable applicants Yes [ ] No [ ] C. Persons hired are of higher quality Yes [ ] No [ ] D. There is lower turnover of personnel Yes [ ] No [ ] E. There are fewer internal personnel problems Yes [ ] No [ ] F. There are fewer misconduct problems Yes [ ] No [ ] G. There are fewer citizen complaints against officers Yes [ ] No [ ] H. Other (please specify) Pre-employment polygraph testing may be used to investigate a number of different issues in an applicant's background. Given your agency's experience, indicate how important it is to use polygraph testing to investigate each of these issues using the following scale: Very unimportant (VU), Unimportant (U), Important (I), or Very Important (VI). 15 IC 1H lfi Medical problems Mental problems Employment history Dishonesty in prior employment Traffic violation history Financial/credit problems Illegal drug use Misdemeanors committed Felonies committed Physical disabilities Abuse of alcohol Involvement in illegal sexual activities Involvement in homosexual activity Involvement in subversive organizations/activities Acceptance or payment of bribes Use of excessive force 0 O I O O O O O I I l I I O I I f—! H '_| '—‘ '— '—| f—I '—_‘ F—W I—‘ f_‘ f_‘ f—‘ '—\ '—‘ H l—l L—l L—.J L—l H H I—l H I—l |—l |—l L—J I—J I—J L—l H '—‘ '— H '—‘ '— H H '—‘ f-‘i 'n—u' '— '-_‘ '—‘ '——.‘ H f_‘ A—l L—J H L—l L—l H H H I—l I—l I—l L—J 0—1 H h—l I—J f—‘ f—! P“ '—| f—I f—‘ H P—Q f_| l—I f—fi f—‘ H f—W H H ‘—J H I—l t—l L—J H L—J h—J L—J E—J H H I—l I—l H H H H H H H fl fl i—I H F! H I'_‘ H F, I—‘ H lu—l I—nl H H H t—l E—l t—l \—l &_J E—J I—l h—J b—J H H In your agency, is pre-employment polygraph testing administered to applicants Yes No Does not apply before a background investigation before a psychological interview before a medical examination before an oral board 0 I C I l—J H I—J A—J f—‘ H '—‘ '—‘ l—l H H L—J F" '_V '_‘ l—I L—l |—l k—l |—J 36. 37. 38. 39. 40. 142 Overall, on a 100 point scale, with 100 indicating extreme importance and 1 indicating very little importance, how important do you consider polygraph testing to be in your selection process? Assume that the overall usefulness of polygraph testing is used as a standard of comparison for other selection procedures and devices. Use the following scale to show the usefulness of each of these other procedures relative to polygraph testing: (1) Less More useful than polygraph; (2) About the same as polygraph; (3) useful than polygraph. Background investigation Written psychological testing (MMPI, CPI) Interview with psychologist Interview with psychiatrist Interview with selection board Personal interview WMUOU‘JIP‘ '—.f—.'_"_.'_.r—‘ L—lI—JQ—lHt—JL—O In your agency's experience, how much confidence do you feel you can have in the results obtained from polygraph testing? (Check only one) a.) Minimal degree of confidence (O-SO%) b.) Fair degree of confidence (51%—75%) c.) Moderate degree of confidence (76%—85%) d.) High degree of confidence (86%—100%) In your agency's experience, what do you feel is a reasonable estimate of the accuracy of polygraph testing results as it is used in your agency for pre-employment screening? (Check only one) a.) Correct less than 50% of the time b.) Correct between 50% and 75% of the time c.) Correct between 76% and 85% of the time d.) Correct between 86% and 95% of the time e.) Correct between 96% and 100% of the time Has your agency ever had an applicant admit to an unsolved A. homicide Yes [ B. arson Yes [ C. forcible rape Yes [ D. armed robbery Yes [ E. burglary Yes [ F. Other serious felony crime(s) t—lHl—JN—lL—J No No No No r—Ho—Ht—vr—sv— HHF—‘f—Wl—N L—lt—I—Jh—J HHL—IHL—J HHHH—Jl—J 143 ************************'k*********************************************** ELBI_;LLL 41. Listed below are some of the common tests and procedures (other than polygraph testing or voice stress analysis) law enforcement agencies use in pre-employment screening. For each procedure, indicate whether or not your agency uses it as a regular part of the applicant screening process. A. Detailed application form Yes [ ] No [ ] B. Civil Service examination Yes [ ] No [ ] C. Written test of basic knowledge Yes [ ] No [ ] D. Written test of aptitude for police work Yes [ ] No [ ] E. Written personality test (e.g., MMPI, CPI, Inwald) Yes [ ] No [ ] F. Written "honesty" test Yes [ ] No [ ] G. Interview with police (oral) board Yes [ ] No [ ] H. Interview with psychologist or psychiatrist Yes [ ] No [ ] I. Physical agility test Yes [ ] No [ ] J. Medical examination Yes [ ] No [ ] K. Reference check Yes [ ] No [ ] L. Background investigation, including neighbors and employers Yes [ ] No [ ] M. Criminal history check Yes [ ] No [ ] N. Other (please specify) Thank you for your participation in this study. Place your completed questionnaire into the enclosed, stamped, self—addressed envelope and return it to the APA Research Center, Frank Horvath, Ph.D., Director, 560 Baker Hall, Michigan State University, E. Lansing, Michigan 48824. If you would like to receive a summary of the results when they are available, indicate in the space below the name and address where you would like the results sent. PLEASE SEND A SUMMARY OF THE RESEARCH RESULTS TO: NAME: AGENCY: ADDRESS: APPENDIX C Modified Questions Used in the Small Agency Survey Questionnaire 144 APPENDIX C Modified Questions Used in the Small Agency Survey Questionnaire 24. In the last 12 months, what is the approximate percentage of applicants who were given pre-employment polygraph examinations who A. Either refused to undergo polygraph testing or discontinued an examination in progress % B. Successfully "passed" the polygraph testing % C. Did not "pass" the polygraph testing % D. Were found to be "deceptive" during polygraph testing % E. "Passed" all selection procedures, including polygraph testing, and were found to be eligible for employment % 41. What is your agency's policy regarding release of pre-employment polygraph screening results? (Check all that apply) ' A. Results are released to the applicant upon request Yes [ ] No [ ] B. Results are released to other law enforcement agencies upon request Yes [ ] No [ ] C. Results are released to other law enforcement agencies only in exceptional circumstances Yes [ ] No [ ] D. Results are not released to outside agencies Yes [ ] No [ ] E. Do not have a policy on this issue Yes [ ] No [ ] F. Other (please specify): 42. Listed below are some of the common tests and procedures (other than polygraph testing or voice stress analysis) law enforcement agencies use in pre-employment screening. For each procedure, indicate whether or not your agency uses it as a regular part of the applicant screening process. A. Detailed application form Yes [ ] No [ ] W. C: t" 145 Civil Service examination Written test of basic knowledge Written test of aptitude for police work Written personality test (e.g., MMPI, CPI, Inwald) Written "honesty" test Interview with police (oral) board Interview with psychologist or psychiatrist Physical agility test Medical examination Reference check Background investigation, including neighbors and employers Criminal history check Drug testing Other (please specify) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes NO No NO No No No No No No No No No NO APPENDIX D ANOVA Tables 146 APPENDIX D Table 36A ANOVA Tables Testing Difference in Agency Mean Confidence Level Scores with Respect to AN OVA Table for a Two-Factor Analysis of Variance PEPS Usage and Agency Size on Confidence Source df Sum of Squares Mean Square F-test P my Usage (A) 1 55.092 55.092 87.85 .0001 Size (B) 1 7 . 187 7.187 1 1.46 .0008 AB 1 0.013 0.013 0.02 .8873 Error 667 418.267 0.627 AB Incidence Table Source Small Large Totals 193 381 574 User 3.295 3.606 3.502 54 43 97 former User 2&1 2.167 2&8 Total 247 424 671 3.117 3.521 3.373 147 APPENDIX D Table 37A ANOVA Tables Testing Difference in Agency Mean Accuracy Level Scores with Respect to PEPS Usage and Agency Size AN OVA Table for a Two-Factor Analysis of Variance on Accuracy Source df Sum of Squages Mean_Sgu_are F-test P value Usage (A) 1 67.104 67.104 70.236 .0001 Size (B) 1 8.351 8.351 8.74 .0032 AB 1 0.128 0.128 0.134 .7143 Error 659 629.614 0.955 AB Incidence Table Usage S_mall Large Totals 190 378 568 User 3.81 1 4.095 4 53 42 95 Former User 2.8—42 m 3.04 Total 243 420 663 3.601 4.007 3.858 BIBLIOGRAPHY 148 BIBLIOGRAPHY Anderson et al., v. City of Philadelphia et al., 668 F. Supp. 441 (E.D. Pa. 1987) Ansley, N., & Beaumont, L.S. (1992). Quick reference guide to polygraph admissibility licensing laws, and limiting laws (16th ed.). Severna Park, MD: American Polygraph Association. Ansley, N., & Abrams, S. (1980). The polygraph profession. Linthicum Heights, MD: American Polygraph Association. Ansley, N., Horvath, F., & Barland, G. H. ((1983). Truth and sm2ience. Linthicunt Heights, MD: .American Polygraph Association. Ash, P., Slora, K. B., & Britton, C. F. (1990). Police agency officer selection practices. Journal of Police Science and Administration, 11(4), 258-269. Belt, J. A., & Holden, P. B. (1978). Polygraph usage among major U.S. corporations. Personnel Journal, :1, 80—86. Blumq R. H. (1967, November—December). The polygraph examination in law enforcement personnel selection. Police, pp. 60—70. Buckley, J. P. (1988). The use of the polygraph in the workplace: The American Polygraph Association's View. Polygraph, 11, 80-90. Bureau of Justice Statistics (BJS). (1992). State and local police departments, 1990. (NCJ Publication No. 133284). Rockville, MD: Justice Statistics Clearinghouse, National Criminal Justice Reference Service. Bureau. of National Affairs. (1985). Polygraphs and employment: A BNA special report (Product Code 02—GERR— 96). Rockville, MD: The Bureau of National Affairs, Inc. Carter, D.L., & Sapp, A.D. (1990). The evolution of higher 149 education in law enforcement: Preliminary findings from a national study. Journal of Criminal Justice Education, 1(1), 59-85. Correa, E. I., & Adams, H. E. (1981). The validity of the pre-employment polygraph examination and the effects of motivation. Polygraph, 10, 143-155. Corruption scandal rocks NYPD; Internal affairs unit under fire. (1992, June). Law Enforcement News, pp. 1, 11. Cross, J. L. (1989). The Employee Polygraph Protection Act of 1988: Background and implications. Labor Law Journal, 40, 663-671. Department of Defense. (1984). The accuracy and utility of polygraph testing. Washington, D.C.: U.S. Department of Defense. Eisenberg, T., Kent, D. A., & Wall, C. R. (1973). Police personnel practices in state and local governments. Washington: Police Foundation. Employee Polygraph Protection Act of 1988 (EPPA), -- U.S.C. S (1988). {June 27, 1988. Public Law 100—347 (H.R. 1212)} Garwood, M. (1985). Two issues on the validity of personnel screening polygraph examinations. Polygraph, 13, 209—216. Gooch, C. (1964). An ingpipy into the use of the polygraph in applicant gvaluation and personnel screening. Unpublished master's thesis, Michigan State University, Lansing. Gugas, C. (1962, JUly-August). Better policemen through better screening. Police, 6, pp. 54-58. Heinley, D. (1986, June). The polygraph vs. the U.S. Congress. Security World, pp. 56-59. Horvath, F. (1985, September—October). Job screening. Society, 22, pp. 43—46. Horvath, F. (1987). Public attitudes on polygraph testing: A national survey. Polygraph, 1g, 1—17. Horvath, F. (1987a). The selection. of candidates for employment in the Philadelphia Police Department: An analysis of the major issues in the use of polygraph screening. Unpublished manuscript. Horvath, F., & Meesig, R. T. (1993). The relative utility and effectiveness of pre-employment polygraph screening for 150 positions of trust: A (concept 'paperu ‘Unpublished manuscript. Horvath, F., & Phannenstill, R. J. (1987). Pre—employment polygraph testing: The attitudes of applicants and their relationships to personal characteristics. Polygraph, 1S, 18—32. Horvath, F., & Reid, J. E. (1981). The reliability of polygraph examiner diagnosis of tnnfifli and deception. Journal of Criminal Law. Criminology and Police Science, §;(2), 276-281. Horvath, F., & Shelton, E. (1982). The use of polygraphic pre—employment testing in police agencies in the united States. Paper presented at a meeting of the American Polygraph Association, Vancouver, Canada. Hubble, D. (1990, May). 1990 Law Enforcement Management and Aflinigtrative Statistics Survey Sample. Unpublished memorandum, dated May 14, 1990, to Diana M. Cull, Chief, Governmental Organization and Criminal Justice Statistics Branch, by the author, Statistical Methods Division, Bureau of the Census, U.S. Department of Commerce. Kendrick, D. H. (1983). use of polygraph by law enforcement agencies. (Report available from Fact Finders Polygraph Service, Alvin, TX.) Lopez, E. (undated). Oceanside Police Department police applicant survey. Unpublished manuscript. McCloud, D. G. (1991, September). A survey of polygraph utilization. Law and Order, pp. 123-124. Meesig, R.T., & Horvath, F. (1993). Changes in usage, practices and policies in pre-employment polygraph testing in law enforcement agencies in the United States: 1964- 1991. Polygraph, 22, 1-16. Nagle, D. E. (1984). The polygraph in employment: Applications and legal considerations. Polygraph, 1;, 1— 20. Office of Technology Assessment (OTA). (1983). Scientific validity of polygraph testing: A research review and evaluation (Technical memorandum OTA-TM—H-lS). Washington, DC: U.S. Government Printing Office. O'Leary, J. (1977, May). Polygraphs and police. The Police Shief, p. 18. 151 Phannenstill, R. J. (1983). The polygraph passes the test. Security Management, 11, 58-60. Putnam, IR. I“ (1978). Polygraph screening (Hf police applicants: Necessity or abuse? Polygraph, 1, 257-262. Roper, I“ E. It (1981). Law enforcement selection reguirements: A national analysis. Unpublished doctoral dissertation, Sam Houston State University, Huntsville, TX. Sweet, E. (1990, June). Sample Selection Specifications for the 1990 Law Enforcement Management and Administrative Stapistics (LEMAS) Survey. Unpublished memorandum for documentation, dated. June 7, 1990, by’ the (author, Statistical Methods Division, Bureau of the Census, U.S. Department of Commerce. United States Bureau of the Census (1989). County and city data book 1988. Washington DC: U.S. Government Printing Office. Yeschke, C. (1962, August). The advantages and limitations of police applicant testing mdth tflua polygraph. Report prepared for American Academy of Polygraph Examiners, Chicago, IL. 30 days of urban agony: In L.A. and other cities, a powderkeg explodes. (1992, May). Law Enforcement News, pp. 7, ll. "111111111111111155