POLE lel@Ir AND SELECTED ASPECTS OF PERRORMANCE eres winnie ad Fer Boch g pone Ein INTERNATIONAL ASSOCIATION OF CHEFS OF POUCE THE URBAN INSTITUTE,‘The conclusions and recommendations of this report are those of the authors and not necessartiy those of the Police Foundation, the International Associ~ ‘ation of Chiefs of Police or The Urban Institute, Copyright @ October 1975 Police Foundation Labeary of Congress Catalog Yunber 75-34732AGKHONLEDGMENTS Funding and preliminary planning for the study were provided by the Police Foundation. le would particularly Like to thank Foundation staff nnbers Richard Staufenberger, Joseph Levis and Sharon Winkler for their aasistance. Professor Lee Sechrest, consultant to the Police Foundation, also sesisted by critiquing an early draft, Considerable tine and effort vere expended by the police departments that collected the data for this study. The authors are grateful to the Following departaents and individuals! Dallas, Texas, Police Depertment Captain Leo Savell Dre dia Hallgren Oakland, California, Police Departeent ‘Captain Jim Hedrthur sau County, New York, Police Department Consissioner Lovis J. Frank Vincent G. Thornton, Chief of Inspection Harold J. Wiges, Deputy Inspector, Planning Untt Des ines, Tova, Police Departaent ‘Jol L, Jones, Director, Research and Developaent Dade County, Florida, Public Safety Departnent Harry P. Wendler, Supervisor, Management Analysis Buresy Anne Nicol, Marry E. Bolinger, Supervisor, Managenent Analysis Bureau Im addition to those mentioned by name above, ve wish to thank all other individuals from the police departnents who worked on the data collection effort. The International Association of Chiefs of Police, especially Andrew Crosby and Deborah Kent, was responstble for securing initial cooperation of participating deparenents end for helping revise the draft report. We fare also thankful to Roger Heinke of the IACP for his reviey of the final report, We thank many of our colleagues at The Urban Institute. Katryna Regan, Cynthia Lancer, Jee Gueron, and Montina Pyndell reduced the ray data to meaningful summaries, Myriam Gaviria typed the final report. Alec, out review committee, consisting of John Hall, Suaner Clarren and Don Weldnon, provided a thorough critique. Joseph S. Wholey provided guidance as Director OE Progran Evaluation Studies. Thonas Wi, thtee Peter B. Bloch‘TABLE OF conrENTs ACIOONLEDGMENTS ui = {INTRODUCTION AND siROWARY Summary of Previous Arguments ‘The Purpose of This Study ‘The Data Collection Plan Description of Data that Were Collected Conclusions Recommendations Sunary of Findings Datas Nassau County Other Participating Departments Sunaary of Literature Review Studies Indicating Lack of a Difference Due to Height Studies Indicating 2 Difference Due to Height DALLAS Background of Officere of Different Heights Height of Officers Education Police Acadeay Scores Assignments Civil Service scores Pecforaance of Officers of Different Heights ‘Assaults: Controlling for Sentorsty and Ageignnente Auto Accidents Sick Leave Injuries Departaent Commendations Department Coaplsints Overview of Different Perforaance Measures Other Predictora of Asasult Raves The Relationships Between Assaults and Other Performance Heasures The Interrelationship Between Height and Characteristics of Assapulte Significant and Nonsignificant Relationships Other Characteristics of Assaults ‘The Coot of Assaults: Paid Injury Leave 7 7 2 a 2 23 26 7 a 3 30 ee 2 33 36 35 36 2 39 2 raIEE, NASSAU COUNTY, NEW YORK Relationship Botweon Hetght and Performance Assaults, Injuries Other Performance Data Performance Measures! Ascaulted Versus Non-Aseaulted Officers Iv. OTHER POLICE DEPARTMENTS Des Moses Dade County ‘The Relationshtp Anong Weight, Rank and ‘Assault Experience The Relationship Betveon Height and Other Characteristics of Assaults oaktand Negative Encounters per Officer Man-Honth Injuries and Negative Encomters Industrial Injuries and Vehicular Accidents APPENDICES: Statistical Methodology Data Collection Fora Designed for this Study Review of the Literature 50 31 5L 53 33 60 a 6 65 65 63 78 2 cf10 u 2 3 1“ 15 List OF TABLES Sunnary of Types of Data Collected for this study and of Statistical Significance of Height Comparisons Height of Officers in Dallas for Randoa Saaple of Officers and Number of Assaults Sunmary of Background Characterfetics and Their Relationship to Height Aveoult Rates in Dallas for Less Senior and More Senior Officers Assigned to Patrol and Other Duties ‘Year of Birth for Officers of Different Heights in Dallas ‘Year of Joining the Departaeat for Officers of Different Heights in Dallas Education for Officers of Different Heights in Dallas Police Acedeny Scores for Officers of Different. Heights in Dalles Avsigunents for Officers of Different Heights in Dallas Civil Service Scores for Officers of Different Heights in Dallas Suanary of Performance Characteristics and Their Relationship to Officers” Heights in Dallas ‘The Relationship Between Officers” Heights and Assaults in Dallas When Assignnents and Seniority are Held Constant Height and Auto Accidents for Dalles Sample of ofeicere Sick Leave Used by Office sn Dallas of Different Heights Injuries to Officers of Different Heights in Dallas: 18 20 a 2 23 2% 25 6 2 2 2» 20 a 216 v 18 9 20 a 2 23 um 23 6 a a 29 a Departnent Commendations Received by Officers of Different Heights in Dallas Department Complaints Received by Officers of Different Heights in Dallas Rankings of Officers of Different Heights on Performance Measures in Dallas Relative Strength of Four Possible Assault Rate Predictors Education and Year Jofned Departuent for Random Sanple of Officers in Dall: Comparison of Performance Data for Assaulted Officers and for a Rendon Semple of Officers 1n Dall: Activities of Officers at Tine of Assault in Dallas Height and Activity of Assaulted Officer at Tine of Assault in Dallas Height of Assaulted officer and Assailant’s Mental State in Dallas Height of Assaulted Officers and Injuries to Other Officer in Dallas Characteriatic# of Assailants in Dallas Type of Assault in Dalles Sunnary of Relationships Between Height and Performance Measures in Nassau County Couparison of Samples of Assaulted and Non- Assaulted Officers of Different Heights in Nassau County Injuries to Assaulted and Non-Assaulted Officers of Different Heights in Nassau County Injuries for Combined Sanple of Officers of Different Heights in Naseau Couey. 2 35 37 38 40 a 43 44 45 a 49 st 52 54 352 33 3 33 36 7 38 40 a 2 43 4 45 46 a Department Commendations for Goabined Sanple of officers of Different Heights in Yaseae County Department Complaints (Form 208) for Combined Sanple of Officers of Different Heights in Nat County Sustained Citizen Complaints for Combined Sample of Officers of Digferent Heights in Nasaae County Accidents in Departnent Vehicles for Conbined Semple of Officers of Different Heights in Nassau County comparison of Assaulted with Yon~Assaulted Officers on Performance Neasures in Nassau County Assaults and Injuries to Officers in Des Yoines Suamary of the Assault Experience of Officers, Sergeants, and Lieutenants in Dade County Assaults by Hands, Fiste, Feet, Teeth or Bodily Force on Officers of Different Weights in Dade county Heights of Officers, Sergeants and Lieutenants in Dade County Weapons Used by Assailants and Injuries to Officers and Sergeants in Dade County ALL Types of Assaults on Officers and Officials of Different Heights in Dade County Sunnary of an Analysis of the Charactertetics of Assaults and Their Relationship to aa Officer's Height in Dade County Characteristics of People Who Assaulted Police 4m Dade County Characteristics of Assaults in Dade County Assaults by Intoxicated or Montally Impaired People fon Police of Different Heights in Dade County alt Activity of Officers at Tine of an Au in Dade County 56 37 58 59 6 62 63 66 66 or 68 70 n n a %48 9 50 st 32 33 54 55 56 on Presence of Other Officers During Assaults on Officers of Different Heights in Dede County Sunaary of Types of Injuries in Dade County Sunnary of Oakland, California Police Department Data Months Worked and Numbers of Encounters for Officers of Different Heights in Oakland ‘Analysis of Encounters of Oakland Officers Over 68 Inches Tall Encounters and Injuries for Officers of Different Heights in Oakland Officer Injuries Due to Encounters and Height a Oaldand Industrial Injuries to Police in Oakland Vebiculer Accsdents in the Oakland Police Departaent Summary of Injury Experience for Officers of Different Heights in San Diego Analysis of the Number of Officers Injured and of the Noaber of Injuries for Officers of Different Heights in San Diego Data on Coste of Injuries in San Diego Overview of Portland, Oregon: Sanple of Non-Assaulted Officers and Assault Incidents Distribution of Assaults and Officers by Height in the North Precinct, Afteraoon Relief, Portland, Oregon Distribution of Assaults and Officers by Height in the Central Precinct, Afternoon Helief, Portland, Oregon, Assaulted and NonwAssaulted Male Police Baployees of Different Heights in the Waghington, D.Cey Metropolitan Police Ascaults on Police Bnployees of Different Heights in Washington, D.C. 5 6 n a0 83 87 89 90 9 100 102 106 107 108 43 14Injuries to Police Eaployees of Different Heights in the Los Angeles Police Department Injuries to Male Police Employees of Different Heights tn the Los Angeles Police Department Assaults on Police Maployees of Different Heights in Seattle, Washington Incidents of Resisting Arrest and of Battery Ageinst Police Bsployees of Different Heights in San Francisco Height Distributions of Young Adult Men and Women, us ne 7 us ugLIST OF FIGURES Change in Aesault Rate 4f Shorter Officers ‘Are Dropped from the Dallas Sample Change in Negative Encounters Per Officer Man-Honth Due to Hypothetically Excluding Shorter Officers in Oakland, California Height, Man-Monthe Worked, and Negative Encountera Per Nan-Youths’ in Oakland, California Change in Officer Injuries Per Encounter Due to Hypothetically fxcluding Encounters of Shorter Officers in Oakland, California Change in Citizen Injuries Per Encounter Due to Hypothetically Excluding Encounters of shorter Officers in Oakland, California Height, Encounters, and Injuries in Oakland, california a 2 85 86 88‘UeTRODUCTION AND SURE Te has loog been the shared belief of the Law enforcesent community that the height of a police officer has an important effect on the officer's Job performance. As a result of this belief, 97 percent of @ large sample ‘of the nation’s police departnents had sone nininua height yequirenent in 1973, with the average minimum requirenent being 68 inches. ‘This well-established lav enforcement practice recently has collided with the legal requirements established by equal euployaent opportunity Lave and regulations. The reason for the collision ie chat nminiaun height require- ‘Beats tend to exclude wouen and persons of certain natfonal origins and races, (e.g, persons of Mexican, Puerto Rican, or Oriental ancestry). Under guidelines issued by the Departaent of Justice, {¢ is permissible for a police departuent to apply minimum height standards only if: the recipient of federal funding is able to denonstrate convincingly through the use of supportive data such as professionally validated eudies that such minimum height requireaente lused by the recipient is an operatiopal necessity Hor designated Job caregories. «+ + The Justice Departnent guidelines clarity this standard somevhat by defining “operational necessity” a8? an exploynent practice for which there exists fan overriding legitimate operational purpose ‘such that the practice is necessary to the safe and efficient exercise of Lav enforcenent duties: is sufficiently compelling to override any discrin~ Anatory iapect} 1a effectively carrying out the Terry Eisenberg, Deborah Ann Kent and Charles R, Wall; Police Personnel Practices in State and Local Governments, International Association of Ghiets of Police et al-, Washington, D.C., 1973. 2 Departuent of Justice, "Equal Rights Guidelines: Minium Height Requirenents--Winorities and Wonen," 38 Federal Register 473 (Number 46, Mareh 9, 1973).operational purpose it is alleged to serves and for hich there are available no acceptable alternate policies or practices which vould better accomplish The operational purpose, advance, or accomplish 4 equally well with a lesser discrininatory impact. SUMUARY OF PREVIOUS ARGUMENTS While federal regulations require empirical evidence to support height standards, it {8 nevertheless desirable to review the reasons given by police for maintaining the height standards which are so widely accepted. If the reasons are sufficiently compelling, thon courts night well be persuaded vith lese rigorous eapirical evideace than might otherwise be ‘The princtpal reasons for excluding shorter applicants (typically persons under 68 inches in height) have been stated in numerous ways, but the following stateaents are believed to be representative: 4s Body build is markedly related to strength... and strength correlated sigaificently with height and weight. © Te 4s apparent that many young adult males find mall. body Size a threat to selfvesteen and tend to depregtate their ‘own personal worth based upon this perception, © Taller officers ean see better in crowds and gre there- fore better able to control public disorders. — ‘den. 4 Hunan ngineering Gulde to Equipment Design, edited by Clifford T. Morgen, MeGraw-Hill Book Company, New York, 1963, p. 557, aa cited by Raynond Le Hoobler and J.A, MeQueeney, "A Question of Height,” Police Chief, Novenber 1973, ps 48. E.E, Gunderson, Ph.D., "Body Size, Self-valuation, and Uilitary Bffec~ Journal of Personality and Social Psychology, 1965, Vol. 2, ov 6, pp. S0ES06, ab cited by foobler and Hetueeney, P- 48s 6 Frank M. Verducel, Ph.D., "Hoight and Weight Requtrenents for Police ‘submitted to the Civil Service Commission, City and County of San Francisco, 1974, p. 15 and p. 40.For these arguments to be sufficient to bar shorter applicants from police work, sone logical Links need to be supplied. For example, it might be found that the lesser strength of shorter individuals is reflected 4n lesser perforsance. There could be 2 nunber of reasons for this: shorter officers might be injured more by people trying to take advantage of then; their partners wight be injured more by people trying to take ‘dvantage of the team of officers because they seem nore vulnerable due to the short officer’ alze; their lesser height might encourage attacks Ferulting Ja injuries to their attackers or to bystanders; or their Lesser strength aight discourage then from asking desirable arrests. Tt night also be found that the lesser strength of the shorter officers causes then to have weaker egos and to compensate by making nore frequent unjustifiable attacks on citizens, or that citizens are more likely to attack shorter officers. One probles with these often-heard arguments is that the relative Snportance of physical strength and other desirable characteristics of police officers has not been established. For example, over 95 percent of police deaths in the last decade have been the result of the use of fire~ farms by the aseatlant. No study has established that physical strength would have been effective in proventing these deaths-—although about 15 own gun and might wrevented by greater alertness of agility, Similarly, Sore thorough studien wight even show whether taller, larger s€fteore present larger targets and are thus more vulnerable to attacks with guns Ie 4e widely recognized that physical strength is only one of several tools which an officer may use to perform effectively. Knowing the proper procedures to use vhen erresting a suspect may contribute substan- Efally to officer safety. Knowing how to deal with people, including calming and reassuring then, aay reduce the need to resort to physical force, Selecting officers with these traite may be a more effective way of reducing Anjuries and increasing productivity than may a strict application of standards for physical attributes such as height. When one discusses the height and race of officers in the same breath, the issue becoues even more complicated. One nay ask, for exemple, whether 4 tall White officer or a short Puerto Rican officer is safer oF nore effective in Spanish Harlen ta New York City. In brief, the arguments concerning height and its effect on performance are Sy no neane conclusive, Consequently, it is important to examine enpirical evidence in order to determine the effect of height on Job perforaance.{THE PURPOSE OF THIS STUDY ‘The purpose of thie atudy vas to analyze data from several police departacats to determine whether the height and performance of police officers were correlated. arly in the study, the difficulty of achieving this objective was realized because police agencies rarely keep data in a for which makes it possible to compare the performence of short and tall officers with sinilar patrol experience. There 1s a problem in comparing officers vith similar experience because of @ national trend of gradually Teducing height requirenents as a condition for employaent. The pattern, in the San Diego Police Departaent--ae reported in the Police Chief-~ie typteal ‘The San Diego Police Department lovered its height requirerent fron 3/9" to 5°7-1/2" in July 1968... « On Septenber 15, 1971, the minimm standard was reduced 0 5°6-1/2".7 Since younger and less experienced officers have less seniority, they often receive the least wanted jobs. These jobs often davolve the evening shift (4 pam. to aidnight), veckends, and high-crine sectora. These also are tho Jobs vith the highest risk of injury.g Hence, it Le hot surprising that the strongest deterainent of assaults on officers found in this study, is the officers’ seniority and whether they were aesigned to patrol unite: ‘THE DATA COLLECTION PLAN A telephone survey identified Cincinnati, Ohio; Dade County, Floridas Des Yoines, Tova; New York City; and Oakland, California for inclusion in this study. Nepresentatives from police departments in these cities were invited to a meeting in Washington, D.C. on March 29, 1974, As the result of that meeting, agreement vas reached oa the data requirenents, and a dats collection format vas designed. Cincinnati dropped out of the Study because of the difficulty of assembling the data. New York City was not able to provide the data within the required tine Linit. For the renaining cities, the data collection plan ves adhered to in varying degrees. Maes Hocbler and Mequeeney, Police Chief, November 1973, ps 42. 8 ‘The Hoobler ond ueQueeney study shows that over 57 percent of on police officers occurred on Saturday or Sunday and 55 percent of assaults occurred between 4:00 pan. and midnight (p. 46). o ‘See analysis of data from Dallas, p. 13, below. alesCounty, New York was later added to the study; however, it wis not able to conply fully with the study’s specifications ‘The data format (see Appendix 3) would have permitted an exanination of the relationship between an officer’s height and each of the following variablest # education «civil service score © police acadeay score © felony arrests © noving traffic cttattons given © noa-felony arrests ‘¢ department commendations department complaints © sustained citizen complaints of paid ofc leave © tayaloe paid ineny tone © days on Light duty © days suspended/forfeited pay © tines assaulted @ times in auto accident tines injured on duty. Additional data vere requested concerning the type of activity in vhich an aseaulted officer vas involved, Weapoas vhich may have been used, descriptions of assailants, and the effect of the assault on the officer. Data collection was divided into two phases in order to reduce tthe donands on the participating departments. The first phase was the preparation of a table by each department showing: (1) the distribution of heights of officers assigned to patrol, and (2) the distribution of hhetghts of officers who recently were assculted. The definition of ssault” varied sonevhat depending on the type of data available in each department, Ie wae Intended, however, that Phase Two data collection(see Appendix B) would occur only Af Phase One differences were statistically significant. The purpose of Phase Two data was to provide an in-depth analysis of reasons for differences due to height. DESCRIPTION OF DATA THAT WERE COLLECTED ‘The only departnent vhich adhered to the phased data collection effort was Des Noines. Thetr date shoved that there was no statistically significant relationship between height and assaults. Consequently, Phase ‘Two data were not requested from Des Moines. Dallas, Dade County, Nasesu County and Oakland all provided Phase ‘Two data, which are analyzed in thie report, Table 1 sumarizes the types of data collected by each participating departnent for this study, and it also indicates the statistically signif~ cant differences found in the data that vere provided. ccoNcLusLoxs Police agencies studied have kept their personnel data in vartous Formats which do not persit thex £0 take comparisons of the performance of Aifferent groups of officers. This 19 an indication that police depart- hnents in general are unable at this tine to asseable data on tall and Short officers with comparable field experience and seniority. In addition, data used in this study relate almost exclusively to ‘uales 67 inches and taller. The shortest officer in the Nassau County ‘example vas 68 inches and fever than Eon percent of the sample of Dallas officers were shorter than 68 inches, ‘The dnadequacy of these data makes it dift{eult to address directly the relationsh{p between height and performance of police officers. How ever, neither the empirical study nor the review of literature discovered any data showing an inportant difference in the performance of tall and short officers with similar seniority end aseignnents, Data fron Oakland Indicate that shorter officers are nore Likely to have less senioricy and have nore negative encounters vith eitizens--but without further dat collection the reletive contribution of height and seniority cannot be estinated, Findings from two departments for which a key vartable, officer experience, vas controlled (Nesaau County and Dallas) show that height aifferences have, with one exception, no statistically significant effect fon performance: © No ctatistically significant relationship with height vas found in either departaent for assaults on officers, ute accidents, departnent complaints, injuries on duty, or departnent comendations.able 1 ‘SUMMARY OF TYPES OF DATA COLLECTED FOR THIS STUDY AND OF STATISTICAL SIGNIFICANCE OF HETGHT COMPARTSONS wage a ae ce TEEEEE D enesond/iofe Department Tee mal ‘ee Tove 23, (hnsaina ow ciftcera vo wore av Toure 5 fost 10 fohes Gal wore sore Likely to ects fide responding to stetrvance or seenpting se artes, and seeaults on shovear officers vert ‘het Lilely wesie they uece anfiing petaneerey contacting tear’ steps or engaged in other activities ‘sle'tinle 1) thtemce on beficere ho rere at lente fest tall wote nore ibaly fo be made by soraal‘ctivent~=t intonteted by Liquor of Sra oe tetaliy ipaited--than Weft seeulte co soetel GHluceres tha opposite erend se aceed‘n feotete,t belor) “Ti sen Tle Bis, oldicere viv were 5 fast 9 ches Or sorter vere sore freqnatly the only offices ‘ndnert eg Stas Lr’ Gileaeisace wee taller then segetats and vergeets taller than other officers) see Table AB, (Aneaulte on ofticers sho vere 5 fece 9 inches or shorter were wore Libely to eecst on FF Oettrabic tg."tAloones co science ta were at leat € fost & tach tall ware 2eee 1haty to bs sie sore sce ented Lge a go ety aan thar ese sven oe tet Seer Piahie Det Uortieeecthe ter 3 feet dnchen ov skort ware far sore 2tkaly to beve ustetned ctttsts couplainte thin were sallerefficere) ‘ir ‘Eee gare 2 ead Table is (uring the denple geri, caller officers worked sor auncosthe-t seg eg SBT lc te Uta prin ha fre iematareeant Se Hlgnee jad Gabe 32, . SL S¢ Lfle sl, “totttcere 3 fect 11 tochen or shorter were sever were ithely co be injred dering an esdomtat witha citizen then were taller oifteered)© No statietically significant relationship with height was found for sick leave in Dallas, the only department providing these dat © A stattetically significant relationship vith height was found for sustained citizen conplainte in Nassau County only (one complaint for every 16 man-years worked by officers who were 69 inches or shorter, compared to fone conplaint for every 73 man-years vorked by taller officers). The Dallas data also support two interesting observations. © Officers who were more frequently assaulted also had more auto accidents, comendations, complaints, injuries, and paid aick leave. ‘This finding suggests an "active’ profile for assaulted individuals, which may be related fo high productivity. (Data on productivity vere not collected.) ‘s The cost to the departacat from paid injury leave for officers vas minimal, amounting to an average of sbout 0.08 man-daye per man-year worked, This finding tends £9 de-enphasize the importance of costs in discussions Of aasaulee and ofticer height: For the sites studied here, more control of the data vas possible than in previously reported work, alloving the results to be viewed with ore confidence. However, experienced officers who are shorter than 67 Aches are not frequently enowgh engaged in police patrol work to permit eapiricel evaluation of their performance. ‘The results reported here~~when considered together with arguments: and findings from other research, including professional and legal sources-~ Ihave the folloving operational iaplication ‘Federal regulations require that shorter applicants not be exclided from employment a8 patrol officers unless profes~ onally validated stulies denonstrate an operational necessity. This study found ao such daca. © Height requirements can vastly reduce the pool of applicants who have personal qualities needed by police Gepartmente. For exsnple, fifty-six percent of young fagtlt males and 99 percent of young adult fensles vould bbe excluded fron exployaggt by a minimum height require nent of 5 feet 9 inches. 1 See Table G-13, Appendix ¢.© Police departments will never know whether shorter officers perform differently than their taller counterparts unless Shorter officers are hired as patrol officers and are carefully compared with a properly selected group of taller, "comparison" officers @ There are no data which document that there £2 say difference in performance between short and tall officers ‘who have sinttar seniority and are given eintlar assign RECOMOEENDATIONS The folloving recomendations may assist police departuents to comply with legal requirenents and, simultaneously, to increase their effectiveness: ¢ Hlininate the height requtrenent and use a selection system based on the overall potential of the applicant for successful police work, This would prepare the way for e future evaluation that would resolve the issue of height. © Provide training for officers addressed to skill developaent ip areas thought by police professionals to involve a height perforuance relationship. SOMMARY OF FINDINGS A brief overview of the specific findings era show in Table 1; a nore detailed description of the findings from each department in the Survey is presented below. Dallas ‘wo amples, one consiating of 144 officers and the other consist~ ing of 18] assaults on officers, were submitted for analysis. The data covered roughly thirteen calendar months. In the {ret stage of the analysis, officers of different heights were studied to determine whether their background characteristics or assignments varied with their height. If height was correlated with back= ground characteristics, then it might have been difficult to determine Whether to attribute performance differences to height or to the correlated characteristic, Hovever, the distribution of officers’ heights was elailar for various levels of seniority, agsigonents and background characteristics Hence, the data wore determined to be acceptable for the study of perfora-10 Next, officers of differest heights were compared on the follow ing performance measures: saulte on officers @ auto acetdents «sick Leave ¢ on-duty injuries © doparenent connendetions, end © dopareneat complaints. Yo statistically significant relationship (at the -10 level of significance) was found between hetght and any of these performance measures, ence, if ss concluded that height and performance were not correlated in Dallas, (ata on arrests by officers were not available.) Data were then examined to determine whether factors other than height had a more powerful influence on the Tikelihood of an assault on an officer than did height. The strongest detersinants of assaults were: ¢ the assigment of an officer (either to a patrol or to ‘a non-petrel unit) aad © the sontority of the officer. Patrol officers’ exposure to ascults was influenced by the type of activity im which they vere engaged. Yor example, 64 percent of assaults on officers occurred when they were responding co a disturbance oF were attempting an arrest. Senior officers may have had eseiganeats Which leas frequently exposed them to the risk of assault, and they may hhave acquired skills vhich reduced their exposure to assaults by increa Ang their ability to deal with potentially violent situations. Woxt the data were analyzed to determine the characteristics of officers who were most frequently assaulted, Generally, it was found that they were more active, Assaulted officers weret © more frequently involved in auto accidents © more frequent recipients of department commendations ‘© more frequent recipients of depertaent conplainte ‘© more frequently injured, and ‘© more frequently placed on paid sick leave,n All of these trends were statistically significant at the .0l level. From sone police viewpoints, this picture of assaulted officers would suggest that they were the kinds of officers vho sight be sought for policing. From that viewpoint, the frequency of assaults on officers ought to be rejected as a criterion for determining whether short or tall officers should be hired for patrol. In the next analysis, seventeen aspects of the assaults were examined to deternine vhether officers’ heights were related to the type of aasault ia which they becane tavolved. By chance alone it would be expected that between one and two of these cosparisons would be stats tically significant. (These tests indicate whether each of the 17 variables would be related to height.) Taree aspects of assaults vere found to be significantly related to an officer’s height. Tt vas found that: * a higher proportion of the assaults on taller officers occurred when they were responding to a disturbant attempting an arrest; and a higher proportion of fon shorter officers occurred during other activities, Such as traffic stops of handling prisoners; © assatlants of taller officers were Less Likely to be intoxicated than were assellants of shorter officers; and taller officars vore more Likely to be injured from an saault vhen they were together with another officer than were shorter officers. It te not at all clear what caused these relationships, One aight hypoth= esize that the first relationship occurred because taller officers were assaulted only in relatively tough situations and that shorter officers were assaulted in nore ordinary situations. But, then vhy vas an intox- Seated aseailant less Likely to attack a tall officer? Why were taller officers assaulted sore frequently when there was other backup present? Furthernore, most aspects of assault vere not found to be correl- ated with an officer’s height. These aspects included: use of a weapor officer's duty status (in uniform of not), sox of assailant, race of Sesailent, age of assailant, number of sesatlanta, whether the assailant, as inown by the officer prior to the assault, the direction from which the officer was assaulted, whether the attack was by a sniper or vas some SS ‘The chance that at least one of the 17 tests vould be significant st the «1 level is 1-(.9)17= 83,2 fora of anbush, the type of weapon used by the assailant, vhether the officer was injured, vhether the officer missed work, and whether the officer vas assigned to light duty. The Dallas data indicated that the cost of assaults to the police @epartuent vas aininal. On average, an officer took tvo man-hours of paid injury leave per year due to injuries from assaults. During the 13-aonth sample period there were 182 assaults for an agency vith 750 officers or 0.22 assaults per man-year. Ten percent of the injured officers missed sone vorkdays, with these officers averaging an estimated six workdays Lost. The net result vas a loss of 0.08 days of paid injury leave per an-year. Nassau County. ‘The Nassau County Police Departaent provided summaries of data fon two samples of officers from patrol precincts, consisting of 223 officers Who were agsaulted and 251 officers vho were not assaulted. The da Submitted for analysie did aot contain inforaation on the seniority of the officers or the shifts to which they vere asaigned-—preventing the researchers fron conducting a preliminary analyeie to determine vhether officers” heights were correlated with some other background character= Levtee, Data were analyzed to deteraine whether an officer's height vas related to © assauits, ‘* accidents in department vehicles, ‘© sustained citizen complaints, © injuries on duty, and ‘¢ department comendations, ‘The only statistically significant relationship between height and these measures vas that shorter officers received slightly more citizen conplaints (one per 16 nan years anong the officers who were 69 inches or shorter and fone per 73 man years for the taller officers). Other Participating Departments Des Moines, Tova; Dade County, Florida; and Oakland, California also participated in this study. ‘adhered to the original data collection scheme and provided information Limited to the height of 181 assaulted or injured officers and on 181 non-assaulted officers, only a Phase One analysis of13 this data uae conducted, Tt was determined that an officer’s height vas not significantly related to the likelihood of an assault or injury. Consequently, no further data were requested from the departuent. Dade County provided data on 869 officers and on 249 incidents in which officers were assaulted, It was found that height was not Significantly related to assaults. Further, it vas found that sergeants and Lieutenants, separately identified in this data, were more Likely than non-ranked officers to be taller than 69 inches and that they had much lover assault rate. Clearly, these data cast serious doubt on conclusions which might be drava from data about height vhich does not Aistinguish betveen ranked and unranked officers, in departuents where ranked officers are ept to be taller. (Ranked officers vould tend to hhave more seniority than unranked officers.) ‘In exanining the characteristics of assaults in Dade County, wo statistically significant relationships were found: © assailants of taller officers vere more Likely to be intoxicated than vere assailants of shorter officers; and ‘© assaults on shorter officers were nore Likely to occur hen no other officer was present than vere assaults bn taller offteers, The First relationsh{p was opposite to the relationsh{p found in Dallas, ‘and the second relationship vas similar to the Dallas finding that injuries to shorter officers wore more likely to occur when no other officer vas injured, All other relationships between heights of officers end charac teristice of assauite vere found not to be statistically significant. These non-significant relationships included: assailant’s age, race or sex; che number of assailante and whether they used a veapon; whether the identity of the assailant was imown by the officer prior to the assault; whether the officer used vespon or was injured; the kind of activity ia which the officer was involved; and whether injuries vere sustained by tore people than just the officer vio was assaulted. Data qubmitted by Oakland had the largest saaple sizes of any Jurisdiction in this study, but the data presented serious problens ‘because Oakland recently had reduced its miniaun height requirenents, and short officers in thie saaple had less experience than their talier counterparts. Controlling for experience is necessary for two reasons. First, experience may be Felated to assignaent; officers may be assigned to riskier duties in ways that vary systenatically with experience which in turn ds related to height, Second, experience nay be related to the style of an officer’s performance. Loss experienced officers may be either nore or less cautious, or provocative ia their approach to potential assault situations.“4 The Oakland deta consisted of 12437 “negative encounters" with citizens (i.e., situations involving a charge of resisting arrest, an assault on an officer or other situations considered to be similar), 8,605 officer injuries, 85) citizen injuries and 632 vehicular accidente. The Gata shoved that shorter officers (with less sentority and, perhaps, different assignnents) had a higher number of negative encounters per san~ year. The meaning of this relationshp cannot be adequately understood, however, until data are collected on vhether the officers were assigned to patrol and on what shifte they aay have worked, SUMMARY OF LITERATURE REVIEW A review of research studies dealing with officers” height was conducted. Reviews of individual studies and an analysis of sone of the previously reported data are presented in Appendix C. For the purpose of discussion, these studies and sets of data aay be divided into two groups: (a) studies vbich support the position that there are no adequate Gata: indicating that tall and short officers perform differently, and (b) studies (and sets of data) vhich indicate sone performance advantage for tall officers but which, universally, are based on inadequately controlled data or on faulty analysis. Studies Indicating Lack of a Difference Due to Height ALlunta, Georgia conducted a study of 300 officers, all of wnoa vere Listed on the "watch-duty roster” and who presumably performed patrol duty. Analysis of the Atlanta date indicates that there vas no ifference in the Likelihood that taller or shorter officers would be ‘assaulted of injured. Southern Methodist University Lav School conducted a study of 17 assaults in a sample of 100 officers in Dallas, Texas. Given the small sample size, tt is not surprising that height and assaults vore not cor related in 4 statistically significant way. Frank Verducet conducted a Iiterature reviey da which he did not perform any nev quantitative analysis. He concluded that there vas no definitive study relating height to performance, and he recommended that TZ In a further analysis of the Oakland data, it was determined that shorter officers had slightly fever injuries per negative encounter. This relation ship vas interesting because the nuaber of negative encounters might indicate, in part, the anount of risk to vhich an officer vas exposed during his patrol work, Hovever, this relationship was not quite statistically significant Bt the «10 level.15 fa study be conducted. Tn a simflar vetn, Sam Chapman surveyed 1,143 Aseault incidents (nost of the data cane from cities 1a Oklshoma); bot he realized that in the abeence of a control group he wae unable to draw any conclusions about the assault-height relationship. Studios Tadicating a Difference Due to Height The most widely circulated study in support of the height-per- formance relationship was the Hoobler and MeQueeney study, published in Police Chief. That study of the San Diego Police Departuent found no Statistically significant relationship between officers” heights, the number of arrests they made, the frequency vith which they were assaulted oF their use of sick leave, However, significant relationships were reported to exist betveen an officer's height and: (1) citizen complaints, (2) officer injuries, and (3) accidents with police equipaent, Officers ‘who were 68 inches or shorter were gost Likely to be subject to the three types of occurrences. Im analysing injuries to officers, Hoobler and MeQueeney streased the aunber gf officers who vere injured. Wowever, their date when reanalyzed! show that there was no difference between tall and short officers in the number of injuries per officer. The reason for this apparent contradiction in the data (which depends on which alternative nmothod of data analysis is used) is that the taller officers were less Likely to be injured, but there were more injuries per officer for the officers who were injured, Arguably, injuries per officer is the better easure of the cost of injurtes to the department, and Hoobler and MeQueency vould have done better to stress that measure of cost. Bence, most of the Hoobler and McQueeney data show siatlarities between short and tall officers, The differences boil dow to the Frequency of citizen complaints and the frequency of accidents with police equipnent. It ta not clear why shorter officers should be Gefietent in these respects, The San Diego data are not subject to the objection of lack of coatrol of reassignment to different shifts. Data presented in Chart 2 of the study show that officers who were 68 inches or shorter had very similar assignments to offices vho vere taller (chi-square = .72). No inforaation 1 avatlable from the study about the height of officers assigned to riskier areas of the city or to the riskier shifts. In addition to the Hocbler and MeQueeney study, there wae sn Evansville, Indiana study vhich concluded that shorter officers per~ formed less well than their taller counterparts. That study found that a See ps 99-104.16 shorter officers were more subject to physical abuse complaints~-to sub- Stantiated or unsubstantiated verbal abuse complaints and to injuries Statistical analysis of the Evanaville data corroborated the latter tvo statistical relationships and cast some doubt on the first relationship, Which barely missed statistical significance at the 0.1 level. The prinespal defect in this study 1a its lack of adequate background Infor~ mation on the officers--a seriou problem because ali the officers who were shorter than 69 inches were hired after 1965 (whea the mininun height Fequirments was reduced). Tn addition, the study did not present any data on the number of arrests made or the aumber of commendations received by officers in the sample; and the Dallas data euggest that the officers vho are most coaplained against may also be the aost active and nost Frequently comended. A study by the Portland, Oregon, Bureau of Police concluded that shorter officers were nore Likely to be assaulted ths vere taller officers, Reanalysis of the data in that report indicated that there were important flaws in the statistical analysis (see pages 104-108). Ia addition, the Portland report failed to indicate whether officers had ess seniority than taller officers. In this review of the Literature, comonly cited data fron four cities (see page 112) had been cited and analyzed. While the relation= ‘Ships between height and performance meawarey is thew data cousistently favor the taller officers, aone of these deta sets indicates the seniority oF agsignnents for officers of differest heights. Nor do aay of the data Show any positive performance data, such as nunbers of arrests or numbers of commendations. In Light of general trends in other departaents, it scons Likely that the shorter officers may have less senfority and aay be exposed to greater patrol risks, Despite the statistical significance of these data, this Lack of adequate controls for seniority and exposure to risk depriv. this data of its possible usefulness, Furthermore, the trend in Dallas indicates that officers vho are injured more may aiso be generally nore active, receiving more commendations a5 vell as more couplaints. The absence of positive indicators of performance also helps to deprive the negative indicators of some of their usefulness. If seniority and assignment information and other performance measures (including counendations and numbers of arrests) can be added to these data, the usefulness of the data might be vastly increased.TL. DALLAS Data collected for this study by the Dallas Police Department were tthe most useful for this study because they conformed most closely to the original data collection plan. The saaple period covered May 5, 1973 to June 15, 1974 and included two separate samples. First, there was a random sample Of 144 officers in che departuest during the sample period, Second, there vas a sample of officers who were involved in the 182 asseulte that oceurred during the sample period. Because the officer’s height vas not recorded for seven officers in the randon sample and one officer involved in an assault, the analysis was lintted to 137 officers and 181 assaults. The Dallas data indicated thet there was no stattetically efgntf— cant relationship between Height and six performance measures. The six atocesinn uencan pear ater eer peebten tee ae Saray ae injuries, departaent connendations, and departwent complaints. The Dallas department did not provide data on arreate nade by officers because the tffore of assenbling the data would have been too extensive. Jn Dallas, officers vho were 5 fact 9 inches and shorter were assaulted at a slightly lover rate than taller officers, but the trend ws fot statistically sigetticmt. ‘he distribution of assaults for officers OF different heights 12 shom in Table 2, Figure 1 ohovs vhat would have append’ to he aanault fate during the jaa collection period if shorter TRCTESe Gea re eager eo ccctiaes Coen) eee ctl eer cos meapoee he solid curve in Figure 1 haa been computed directly from Table 2- The doteed carves indicate that the apparent advantage fron hiring shorter officers nay be considered a randon effect, within statistical standards fstablished for this study. for exasple, the 80 percent Lines indicate that {fone vere to érav another set of samples fron « single group (universe) of officers among vhon there were no overall differences due to helghty the chances would be 80 out of 100 that the entire solid curve would fll lenin the dotted Lines, ‘BACKGROUND OF OFFICERS OF DIFFERENT HEIGHTS one of the background characteristics of the randon sample of officers was found to be significantly related (at the 0.10 level) to officer’s height (see Table 3). The civil service score was so frequently oaitted fron the data that thfa report le inclined to disregard the finding— from the available data--that taller officers had significantly higher civil ‘service scores18 ‘Table 2 HEIGHT OF OFFICERS IN DALLAS FOR RANDOM SAMPLE ‘oF OFFICERS AND SUMBER OF ASSAULTS TORRE TN SAUPLES ——COMTCATIVE PERCENT DUPERENCE erat WT assaulted Ail ——Assaateea (inches) Officers Officers officers Officers $ a 38 TT 8 4 BS cy 13 18 19.0 13.2 ° 10 15 26.2 215, a7 70 W 3 38.6 39.7 aa a 2 36 56.1 58.5 26 n 23 23 29 23 0.6 2 3 u 83.8 80.0 38 % 3 2 93.3 OL 22 8 8 5 99.1 93.8 5.3 76 oF nore 100.0 100.0 0.0 Total (known) 137 181 ‘This 4a the greatest difference in the cumulative percents of the number of officers and of assaults. The difference is not statistically Significant even at the 0,20 level of significance.Fe PERCENT CHANGE IN ASSAULT / RATE DUE TO DROPPING / SHORTER OFFICERS FROM THE SAMPLES fe 450% otter Height She +207] +107] 10%] 20%] 302] 503 wore: ~ ag" PERCENT OF OFFICERS 96200 OFFECERS Sample from entize departnent = 137 officers; sample from sll faults = 181 assaults. The 80 percent and 99 percent linits Indicate the range in which the unbroken curve might vary under ‘the condition that the only source of variation vas sampling error and not any eystenatic height trend. Nanely, if one were to draw another set of samples from single source of hheighta, then the chances are 99 in 100 that the entire solid carve would stay within the 99 percent Linits as indicated by the two dashed curves. The derivation of the Linit curves 18 siven in Appendix A, Figure 1 (CHANGE IN ASSAULT RATE IF SHORTER OFFICERS "ARE DROPPED FROM THE DALLAS SAMPLE.20 Table 3 SUMMARY OF BACKGROUND CHARACTERISTICS ‘AND THEIR RELATIONSHIP 10 HEIGHT BACKGROUND CHARACTERISTIC 1S THERE A STATISTIGALLY SIGNIFTONNT™ ‘OF OFFICER RELATIONSHIP BETWEEN AN OFFICER'S HEIGHT AND BACKGROUND CHARACTERISTICS? OO Year of Birth Yo Year Joined Departzent Yo Bducation Yo Police Acadeny Score Yo Assignnent Ne, Civil Service Score ® a ‘At the 0.10 Level, » ‘Too much afssing data; however, based upon the 65 (out of a total of 137) officers with available data, the ansver vould be yes. Taller officers hhad higher scores, Since the seniority and the assignments of officers in Dellas have 4 strong influence on assault rates, the finding thet officers of diftereat hheights did not differ on these variables vas very inportant. For example, the assault rates anong less senior officers on patrol azaiganents was 225 percent of the overall assault rate anong all officers (eee Teble 4).a Table 4 ASSAULT RATES IN DALLAS FOR LESS SENIOR AND NORE SENTOR ‘OFFICERS ASSIGNED TO PATROL AND OTHER DUTIES DATE THE OFFICER JOTNED ASSAULT RATES (AS A PERCENT OF THE ‘THE DEPARTAENT. DEPARTONE’S AVERAGE ASSAULT RATE) (entority) FOR OFFICERS, BY CURRENT ASSIGIIENT Patrol Gy ener (EY ipa EP After 1968 226 66 1968 or earlier oof ast ‘YOUR Sunple alae = 156 randoaly selected officers and 181 assaulted officers. "this category contains 25 percent of the randoa sample of officers and 56 percent of the sample of officers uho were assaulted. » "this category contains 24 percent of the random saxple of officers and 15 percent of the sasple of officers who were assaulted. ‘This category contains 19 percent of the random sample of officers and 13 percent of the sample of officers vho were assaulted. a ‘This category contains 32 percent of the random sample of officers and 15 percent of the sample of officers vho were assaulted.Height of Offte ‘At the tine of the survey, a majority of the officers in the random sauple of officers vere under 35 years old, The distetbution of years of birth for officers of various heights 18 shoim in Table 5. Table 5 YEAR OF BIRTH FOR OFFICERS OF DIFFERENT HEIGHTS IN DALLAS OFFICERS, BY YEAR OF BIRTH ansaar aches) T939-aad 1900-103 ——T5S0 ad So @@ Ow ww 69 ond under 2 9 8 2 Mb 10 35 no-2 ee a eT) 73 and above ao os 8 3 0 ALL heights 35047 53072 16 100135" WOTHS: Data are from May 1973 to June 1974, Cuivsquare = 2.81 with 4 degrees of freedom, probability = 0.59 Excludes two officers whose dates of birth vere miasing.2a Experience in the Department Fifty-eight percent of the officers in the random sample had five or more years experience 1a the department. The taller officers had slightly more work experience than did the shorter officers, but the difference was not statistically significent at the 0,10 level (see Table 6). Table 6 ‘YEAR OF JOINING THE DEPARTMENT FOR OFFICERS OF DIFFERENT HEIGHTS IN DALLAS OFFICERS, BY YEAR JOINED DEPARTIENT HEIGHT Caches) aa Ss T570 aa earlier Later Total Oo @ @ ®© mM @ @ 69 and under 2% 1 ws. 58 21 10036, 10-2 2 18 33a 38 24 10063 73 and above ase 3513 10037 ALL hesghts 243 88S 4258 190 136% WOTE: Data are from May 1973 to June 1974. Excludes one officer vhose date of joining the department was missing. ‘CONCLUSION: As a group, the shorter officers tend to have had fever years of experience in the departments hovever, the dif- ferences vere not statistically significant (chi-square = 7.56, probability = 0.11).ey Education ‘The sample was about evenly split between officers with up to 12 years of education and officers with at least sone additional education, Ofticers of different heights had sintlar educational backgrounds (see Table 7). rable 7 EDUCATION FOR OFFICERS OF DIFFERENT HEIGHTS IN DALLAS OFFICERS, BY YEARS OF EDUCATION EIGHT (ances) a years oF Tess ter TT years Foe — ® a oO oO @ w® 69 and lees 56 20 46 1610036 70-12 8 52 3010058 73 and above 33 19 47 7 10036 ALI heights 32 67 48 63 100 130% WOTH: Data are froa May 1973 to June 1976, Bxcludes seven officers whose years of education were missing. CONCLUSION: There was no statistically significant relationshtp between height and education (chi-square = 0.68, probability = 0.70).3 Police Acadeny Score Police acadany scores vere not reported for 43 of the 137 officers in the random sample of officers. There vag no statistically significant relationship between the heights and acadeay score of officera whose Scores vere reported (see Table 8). Tobie 8 POLICE ACADEMY SCORES FOR OFFICERS ‘OF DIFFERENT HEIGHTS IN DALLAS erent (Inches) oF Tower ‘OF above “Total @ a 2 © © @® 69 and below 2 16 38 101006 0-22 32 a a L100 73 and adove 46 n Sho 13 1002 ALL heights 53 50 47 4h 1009" ‘Excludes 43 officers vhose police acadeny scores vere missing. CONCLUSION: There vas no statistically significant relationship between height and police scadeny scores (chi-square = 1:26, probability = 0.53).26 Aseiganent Since the data collection form anticipated that results vould be obtained only for patrol officers, there was no blank on the fora to iadicate fn officer's assignment. However, Dallas reported information on some officers who did not have patrol agsignnents, and it indicated the aeaiganent by writing At 1m on each of the forns. Using this handwritten data, ve determined that officers of different heights had sin{lar aseignaente (eee Table 9). table 9 ASSIGIMENTS FOR OFFICERS OF DIFFERENT HEIGHTS IN DALLAS OFFICERS, BY ASSIGNMENT nerout (Inches) Special Operations Patrol CID, Vice, Druge Ll Other Total oo OO @ OM wm 69 ond under 9 8 3312 100 36, - 72 4“ 2 0 25 16 100 66 73 and above wo 7 2 2 2 8 100 37 ALL heights rr 2% 36 100137 CONCLUSION: There vas no statietically significant relationshtp between height and assignment (chi-square = 1.39, probability = .86).27 caren vice Scores Civil service scores were not reported for 72 of the 137 officers in the saaple, Among the 65 officers with reported scores, there was a tendency for the taller officers to have had higher scores, The difference vas. statistically significant at the 0,10 level, but the large auaber of aissing cbservations makes it difficult to interpret this statistical relationship (see Table 10). ‘Table 10 (CIVIL SERVICE SCORES FOR OFFICERS (OF DIFFERENT HEIGHTS IN DALLAS HELGHT Caches) 13 oF Tower ‘Wor above Toad o oy 2 oa 69 aad under 1% 2 3 4 100 16 0-12 “8 13 S20 ht00 a 73 and above 36 # 6k t00 2 ALL heights st 33 4932 100s ‘xeludes 72 officers whose civil service scores were missing. ccoxctustoN: ‘Among those officers vith reported scores, shorter officers had lover scores (chi-square = 5.67, probability ~ 0.06). The difference is statistically significant at the 0.10 level; hovever, the large number of officers with missing data is a good reason for viewing the result cauticusly.28 ‘PERFORMANCE OF OFFICERS OF DIFFERENT HEIGHTS ‘The performance of the sample of officers vas not found to be related fo Theit height tn a statistically significant way (eee Table 11). The performance characteristics include: auto accidents, sick leave days, oncduty injuries, departuent comendationa, department coaplaints and ‘aesaulte, Ho data vere available on arrests. Table 11 SUIQURY OF PEFORMANCE CHARACTERISTICS AND THEIR RELATIONSHIP 10 OFFICERS” HEIGHTS IN DALLAS ‘IS THERE A STATISTICALLY SIGNIFICANT © CHARACTERISTIC OF OFFICER ‘RELATIONSHIP BETWEEN THE OFFICERS” HEIGHTS AND PERFORMANCE CHARACTERISTICS? Frequency of Auto Accidents No Stok Leave Days Yo Frequency of On-Duty Tnjury Yo Department Consendations Yo Departneat Complaints Yo Frequency of Assaults fo NOTE: Data are from May 5, 1973 to June 15, 1974, "At the 010 level of significances29 Assaults: Controlling for Seniority and Assignments Because seniority end asaigaente were found to have a strong influence fon assault rates (see Table 4, page 21), an examination was conducted on the influence of height on assaults when seniority and aseignnents are held constant, Hetght was found still not significantly related to the assault ‘rate under any conbination of assignzent and seniority, as shown in Table 12, The sample sizes are suall and only large trends would have been statistically significant. Tabte 12 ‘THE RELATIONSHIP BETWEEN OFFICERS’ HEIGHTS AND ASSAULTS IN DALLAS WEEN ASSTCIMENTS AND SENIORITY ARE HELD CONSTANT ‘OFFICERS, BY CRRENT ASSTOOENT ‘YEAR JOINED THE DEPARTGNT =—_-NEIGUT Patror Other (Seniority) Cinehes) Tandon Assaulted [oa ranted Sample Officers | Sample Officers of Officers of Officers Oo MM 1ae w m Accor 1960 69 ortee 29 100 zz | oe ro aes 4s 6 | ah 68a T3ormore 26 «9033 | oka All heights 100 34 100 102 | 100 32 100 26 1968 or ories i 6 2 6 | 38 26 7 Earlier 0-72 so13 58 ae | 48 2148 13 Wormre 35 9 17 4 | az we 7 ALL heights 100 34 100 102 | 100 32 100 28 CONCLUSION: Anong the four groups of officers determined by assignment (patrol, other) and sentority (Joined before or after 1968) there was no, statistically significant relationship between ascault rates and heights. (The researchers suggest that no inportance be attached to the apparent concentration of assaults anong middle-heighe officere-~70 to 72 inches-—uho hed non-patrol assigazents. While the chi-square was 4,12 [probability~ 0v12], there is no analytical reason for this difference, and it should therefore be ignored).Auto Acekdent: During the sample period, about one quarter of the random sample of officers were tavolved in auto accidents, Taller officers had a slight tendency to be Snvolved In fever accidents; hovever, there vas sbout a one Anvseven probability that the trond was the result of chance alone. The trend is shown in Table 13. Table 13 HEIGHT AND AUTO ACCIDENTS FOR DALLAS SAMPLE OF OFFICERS OFFICERS, BY AUTO ACCIDENTS DURING SAMPLE PERIOD Accrpawrs EIGHT ‘PER OFFICER Caches) one “tae oF Ware Tora ow & MO @ ~ 69 and under m2 2% 2% 10 10 36 0.28 n-n mB 4 27 1710083 0.31 73 and above s 32k t00 86 0.17 ALL heights 77 106 2331100135" 0.26 MOTE: Data are from May 5, 1973 to June 15, 1974, Becludes two officers for vhom dats on numbers of auto accidents vere miseing. CONCLUSION: The trend de for a tall officer to be davolved in slightly fever traffic accidents, although the effect is not static~ tleally significant at the 0.10 level (chi-square = 3.91, probability = 0.15).31 Sick Leave Officers took an average of 2.2 sick leave days during the sample period. Forty percent of the officers took no sick leave. Officers of GiEferent heights took similar mounts of sick leave (see Teble 14). Although there is a trend for shorter officers to take more sick leave days, the statistical significance is strongly dependent on how the data are grouped by height. Table 14 SICK LEAVE USED BY OFFICERS OF DIFFERENT HEIGHTS IN DALLAS ‘OFFICERS, BY DAYS OF PAID SICK SICK LEAVE LEAVE DURING SAMPLE. PERIOD DAYS. PER ‘aexcer OFFICER (Inches) Tine “tae ar Tare oat oo @ @ wm 69 and under 2 10 os 1008s 3.0 n-7 4h 26 363310059 uy 73 and above 46 17 5h 2010037 22 ALL heights 40 53 6078 = 00 ast® ae MOTE: Data are fro May 5, 1973 to June 15, 1974. ‘Excludes efx officers whose sick leave data were missing. CONCLUSION: No statistically significant trend {2 noted Linking the three hesght categories and sick leave (chi-square = 2.83 vith 2 degrees of freedom, probability = 0.24). However, if the taller two height categories are grouped together and a 2 by 2 chi-square test 1s performed, the chi-square value ts 2.80, which 1s statistically significant at the 0.10 level.2 Injurs During the sample period, 14 percent of the officers in the randos sample were injured. Injury experience was the same regardless of officers” heights, as shown in Table 15. Table 15, INJURIES TO OFFICERS OF DIFFERENT HEIGHTS IN DALLAS OFFICERS, BY TIMES TKJURED ON INJURIES FER DUTY DURING SALE PERIOD OFFICER DURLNG Caches) None “the oF Wore Total — ao @ @ @ mH 69 and under ee 00 oe m- 72 8 56 e800 0.16 73 and above e310 a6 ona ALL heights % 16 16 1910013587 as WOES: Data are from May 5, 1973 to June 15, 1974, Chi-square = 0,005, probability = 1.0, Ereludes two officers whose injury data were missing.3 Department Comendatton: ‘The distetbutton of departnent commendations vas fairly broed, with 60 percent of ali the officers ia the sample receiving at least one Renlation during the saxple period, The differences anong officers of Uieferent heights wae not statistically significant (see Table 16.) Table 16 DEPARTMENT COMMFNDATIONS RECEIVED BY OFFICERS OF DIFFERENT HSIGHTS TN DALLAS OFFICERS REGEIVING DIFFERENT MUNBERS OF DEPARTMENT COIOGADATIONS COMMENDATIONS EIGHT PER OFFICER (Inches) Tone “Gna or Wore Total 2m © M @® 69 and under 50 18 © 50 18 100368 n-n 3220 © 681006. Ls 73 and above se 6582010038, 0.8 AL heights 40 56 © 60 8000 #1 OTR: Data are from May 5, 1973 to June 15, 1974. ‘Excludes three officers whose coamendation data were missing. CONCLUSION: Officers of different heights received stetler numbers of commendations (chivequare = 3.33, probability ~ 0.18).Ey Department Complaints Departnent complaints vere lodged against only 24 percent of the officers, and there wero no statistically significant differences among officers of different heights, as shown in Table 17, Table 17 DEPARTMENT COMPLAINTS RECEIVED BY OFFICERS (OF DIFFERENT HEIGHTS 1 DALLAS OFFICERS RECEIVING DIFFERENT COMPLAINTS PER NUNBERS OF DEPARIMENT COMPLAINTS OFFICER DURING wer6aT SAMPLE. PERIOD nehes) Tone ie oF Wore Total oo @ m gm 69 and under 1% 2 2 8 6 0.25 n-2 3 48 271610060 0.43 73 and sbove m2 8 00a 0.38 ALL heights % 101 232,00 aas* 37 WOTES: Data are from May 5, 1973 to June 15, 1976. Chirsquare = 0.41, probability = 0.83, Bxeludes four officers whose complaint dat vere atesing.35 overview of Different Performance Mensur liven though no single performance measure shoved a statistically significant difference for officers of different heights, one might still sk if aay pattern might be found anong all the performance measures, when Considered together. To see if there was a pattern anong six different performance measures, officers were grouped into three height categories Gach of which was aeeigned a rating of best, mid or worst, depending on tthe rankings assigned in Table 2 and in Tables 13 through 17. ‘the rankings constructed in this fashion shoved no pattern, as can bbe seen from Table 18, Had the rankings been assigned randomly, the chance Of getting four or more worst rankings would have been 0.18. ven though the 70 ~ 12 dach height category received four of six vorst ratings, it wae concluded that this pattern, which has no theoretical Justification, ‘ight have occurred by chance alone and should not be considered statistically significant. TABLE 18 RANKINGS OF OFFICERS OF DIFFERENT HEIGHTS 0% TPERFORVANCE MEASURES IN DALLAS RELATIVE RATINGS BY HEIGHT CATEGORY 18 THE TREND enromunce wrasins* staristicalgy Tyiaames JO~ TE TS tases STCNIFICANT Gnd under inches and above Assault rate tet erst Md Xe fico sccidence wa ve ese » Sick lewve toree ma » Oncduty Injuries est ta est nd % Departtent commendations Mid ore » Beparenene coopiainte Best wa * OTE: Data are from May 5, 1973 to June 15, 1974, ‘Average per officer ® ‘At the Ol Level.6 Other Predictors of Assault Rates Had short officers received a disproportionately Large share of gosaults (e.g., 1£ X percent of the officers were short and they received ¥ percent of the assaults, with X being much sualler than ¥), i¢ would have been said that height was a good predictor of assaults. For the purpose of selecting officers in order to reduce the number of assaults in the department, one might then have selected taller officers. lovever, 1f che urpose is to reduce assaults, one might also examine other officer char acteristics to determine whether there are soat potentially more powerful, Predictors of assaults than height. If there are such predictors, then Perhaps they night be used as selection criteria in preference to height. Tn this section of the report, sone other possible predictors are exasined. This portion of the study includes: © the calculated differences between the percent of randomly selected officers vith a given characteristic and the percent of assaulted officers with that characteristic, and © the cosputed percent change in the assault rate that would have occurred if all officers with that characteristic hhad not beon in the department in the sample period Cand walts on then had not occurred to others). Table 19 shows that two officer characteristics, education and police acedeay scores, are better predictors of assaults than height. While the height standard has but a minimal effect on the assault rate, level of education and police acadeny scores might have a substantial effect. Further exam Anation of the education varicble indicates, however, that more highly educated officers have been hired only recently in Dallas, and the relatively higher rate of assaults of this group may be due to ita lack of experience ‘and sentority rather then to education received. Me statistical significance of the charactertatics displayed tn Table 19 has been examined. Height ts the only characteristic in the table whose relationship to assaults is not statistically significant at the 0.10 level.30 Table 19 RELATIVE STRENGTH OF FOUR POSSIBLE ASSAULT RATE PREDICTORS ASSAULTED ‘CHANGE I CHARACTERISTIC OFFICERS OFFICERS ASSAULT RATE wore THE wore THe FROM DROPPING CHARACTERISTIC CHARACTERISTIC OFFICERS WITH (CHARACTERSITIC @ @ @ 69 inches or shorter 26 2 +8 70 inches to 72 inches a 50 -6 More then High ‘Senor Paueation us 66 3h Police Acadeny 3a 70 36 CONCLUSION: Tf police agencies desire to reduce assaults on officers, the height of officers aay be a relatively poor criterion. Tt ‘would be more rational, considering only the Dallas data, to Tetuse to select officers who had more than a high school education or who scored Jess than 90 in the police acadeny Glovever, it {a believed that the conclusion regarding more highly educated officers would be erroneous because the depart Beat has recently begun hiriag more of these officers and they fare among the Less senior and, possibly, the nore exposed to isk (see Table 201.)38 Over the years, the percentage of officers entering the department with more than a high school education has increased sharply (see fable 20). The more recently hired officers have had more years of fornal education but fever years of police experience; and they have suffered more assaultas This puts the more educated Dallas officer in an analagous position to the shorter officers of many ocher departnents. Since no coupericon aay De taade with less educated officers with the sane experience, no conelustons should be ade about the effect of education on the performance of officers. TABLE 20 EDUCATION AND YEAR JOINED DEPARTMENT FOR RANDOM SAMPLE OF OFFICERS TN DALLAS eee YEAR JonED OFFICERS, BY EDUCATION DEPARTMENT Wigh School _~Yore than or Less High School ‘Total @ a @® @ om ee 1964 oF earlier now 23 10038, 1965 - 1969 6s 23 2 ou 10034 1970 or later 2 18 6 40 10056, ALL years 3268 8 62 100 130° eee NOTE: Data are from May 1973 to June 1976, ‘Beelud: seven officers whose education data were missing.33 ‘The Relattonshipe Between Assaults and Other Performance Me In this section of the report, other characteristics of officers ‘who are involved in assaults are examined. The picture appears to be ons df an officer who ts generally nore active than the officer who 18 not assaulted, More specifically, assaults tend to occur to officers who are involved in nore auto accidents, '* use paid injury leave more often, © receive nore department conmendations, ¢ receive more department coaplaints, ¢ are injured ore, ¢ are more frequently assigned to patrol. ALL the above characteristics are statistically significant, as shown in Table 21. ste ‘This section of the report examines the relattonsh{p between officers” heights snd the nature of the assaults in which they were involved. Seven teen dapects of the 161 assaults in Dallas vere examined. Sixty-four per~ cent of the assaults occurred vhen an officer was either responding to a disturbance or attempting an arrest. In 68 percent of the assaults, officers used their hands of feet as Veapone--shorter officers being no more or less Likely to resort fo the use of firearms than vere taller officers. Yost of the assaults occurred to officers who were in uniform, with another officer present, (See Table 22 for a sumaary of the ectivities of the officers ‘at the scenes of assaults.)40 Table 21 ‘COMPARISON OF PERFORCE DATA YOR ASSAULTED OFFICERS. ‘AND FOR A RANDOM SAMPLE OF OFFICERS IN DALLAS sues or sume or kivormems , Assam ‘iaronusce cuanacreustics tern omens oe B88 we cm, vic ame HS sf # 2 we mo ow oie Tt ee a SY 2 $ mo TS wo Ter one eo os om se Oe og wore ao % 8 mm foe To Tt) Too et eye on paid snjury Leave Toe ara Toeale To TT Ter Degartnent commendations one oe » 2 x0 One’ of nore o © 2 Im ora io it To Teo Department conplatats® Tone % ue One of nore 32 fora ey a ‘Tinos Injured on duty “isn % oe 2 One of sore a a. ne Toeal oo |e Bone: Data are frou May 5, 1973 to Je 15, 1974, cored : "k= 1a ‘stactettcally efgatetoant at 0.05 Level. Trelnding sissing data. Seatiatteally aigatftcant at 0.10 Levelsa Table 22 ACTIVITIES OF OFFICERS AT TIME OF ASSAULT IN DALLAS 18 THERE A STATISTICALLY SIGNIFICANT REALTIONSHI? OFFICER'S ACTIVITY o Wim oFFrceR’s uErcHT?® ACTIVITY TYPE Responding to « disturbance 3358 ‘Actenpeing arrest 3155. Handling prisoners 33 Traffic pursuits and stops 8 5 ALL other BS Total® 100178 WEAPON USED BY OFFICER Bone 1m Hands and/or feet 68107 Discharge firearas om No Other 4a 2 total® 100157 ory starus on duty, {n uniform, other officers present 3798 on duty, im uniform, other officers not present 1326 On duty, in unlfors, missing data on other officers presence est No on duty, not in untfors 58 OF duty 6 10 Total® 100173 NOTE: Data are froe May 5, 1973 to June 15, 1974. ‘At the 0.10 level. b See Table 23 for the nature of the relationship between officers” type of activity and height. eluding missing daca.a Significant and Nonsigntficant Relationships Of the 17 aspects of assaults that were exanined, only the follow ing three were found to have @ height: atistically significant relationship to # Most assaults (74 percent) on officers who vere taller than 69 inches occurred while they were responding to a disturbance or vere attempting an arrest. Most assaults (55 percent) on officers who were, at most, 69 inches tall occurred in traffic pursuits or stope, handling prisoners, and other situations (see Table 23). © Only four percent of agsaulte on officers over six fest tall were by intoxicated or otherwise abnormally behaving individuals, as copared to 23 percent of the assaults on shorter officers (see Table 24). 14 Eéghteea percent of assault-related injuries to officers over six feet tall occurred in incidents involving the Injury of more than one officer. Only eight percent of Anjuries to shorter officers involved the simultaneous Anjury of another officer (see Table 25).Tebte 23 ‘HEIGHT AND ACTIVITY OF ASSAULTED OFFICER ‘AT TIME OF ASSAULT IN DALLAS ASSAULTS, BY ACTIVITY OF OFFICER AT ‘TIME OF ASSAULT uercut _egponding Atteapting andLing Privonere, nehes) to Arrest Traffic Pursuits and Disturbance (all types) Stops, and All Other Total oo @ o&® a @ a 69 and under a mw 6 3 a 100 38 n-n2 2 2s 38s 3 a 100 91 73 and over cr ey a B 10049 ALL hesghts 388s 37 5 100 178 Brclad three events for which activity data vere missing. CONCLUSION: Assaults on taller officers occurred more frequently while responding to # disturbance oF atteapting an arrest, coupared to assaults on shorter officers (chimsquare = 12.9, probability = 0,012).4 Table 24 [HEIGHT OF ASSAULTED OFFICER AND ASSAILANTS MENTAL, STATE IN DALLAS ASSAILANTS, BY MENTAL STATE wercuT (Inches) Norsal Impaired Total oo © OM @ © 69 and under 1% 2 2 9 10 38 m-n n 6 2310090 73 and over 9643 4 2 100 45 ALL heights 82 L 18 32100 1738 WOTE: Data are from May 5, 1973 to June 15, 1974. CONCLUSION: OF the assaults made on officers over six fect tall, a much lower percentage are by intoxicated or otherwise absormally behaving individuals as conpared to the percent for shorter officers (chi-square = 8.07, probability = 0,017).Table 25 HEIGHT OF ASSAULTED OFFICERS AKD INJURIES 10 OTHER ‘OFFICERS I DALLAS ASSAULES, BY INJURIES 10 OTHER THAN OFFICER ASSAULTED weroxt (Inches) one Other Officer Tatar oo © M @ om (69 and under 9737 3 1? 10038 m-7 90 80 10 9 100 89 73 and over aa 18 9 100 50 ALL hesehts a9 158 m9 100 a7, NOTE: Data are fron May 5, 1973 to June 15, 1974, ‘The other person who vas injured in this incident wat a firenan, not @ police officer. CONCLUSION: Shorter officers were more frequently the only officers injured in an incident (chi-square = 5.12, probability = 0.078).46 eight was found not to have a statistically significant relation ship to iy of the folloving fourteen characteristics of an assault: officer’s use of a wespon, officor’s duty status, including whether the officer was de uniform of vas vorking with at least one other officer, sex of assattant(s), race of assailent(s), ‘age of assailant(s), number of assatlants, whether the assaflants were know by an officer prior to the assault, Airection of the assault, vhether the assault Savolved # sniper, anbush, or trap, type of weapon used by the assatiant, whether the officer was injured, type of injury to the officer, whether the injury caused the officer to miss work, Whether the injury caused the officer to be assigned to Light duty. Other Characteristics of Assaults Assatlants tended to be young (55 percent between ages 20 and 30) ales (69 percent) who were not affected by intoxication, drugs, or mental Inpatmment. (62 percent); and vho were not previously kno to the assaulted officer (99 perceat) (aee Teble 26).a Table 26 CHARACTERISTICS OF ASSAILANTS TH DALLAS Ts Te a SEATS SIGNIFICANT RELATIONSHIP ASSATLANTS’S CHARACTERISTICS (2) (W)-—=—“METH OFFICER'S HEIGHT?® sex Males 89162 Fensles =, 020 No Total sampic? — 70082 Race Caucastan 5k 88 Black. 75 Xo Other 8 Total sample” T0079 Age 0 to 12 years of age o 2 1-19 ye 20 - 30 55102, Yo over 30 28 Total same? — 00" a7 umber of One 92 164 assailants Two 59 Yo in the Three or morg 3 __6 Aneident Total sample 100 “177 Assailante’ Wornal 821 Dehavior Intoxicated m3 rest Wigh on drugs = 23 Mentally inpgired_2 0 _4 Total sample’ 100173, Identity Yee 12 non by Ko >» 2 4 Yo officer ‘Total sample” 100.172 prior to ssault WOTE: Date are fron May 5, 1973 €0 dune 15, 1974. At the 0,10 Level, » Excluding missing data, ‘See Table 24 for the nature of the relationship between the height of an officer and assailant’ behavior.8 Aimost all assaults (92 percent) vere frontal attacks on the officer, as shom in Table 27. In one third of the incidents, assailants eed oF threatened age of weapons other than parts of the body (o.g-» hands, feet). Plrearms were the most common weapon, but caused only tvo percent of the assault injuries. The use of @ firearm was at least threatened in 17 percent of the incidents. Sixty-two percent of the assaults reported in Dallas resulted in sone injury to an officer, but only ten percent of the injuries caused the officers to be abseat frou work. In ten percent of the assaults in which officers were injured, their partners were also injured. Partners of shorter officers were less frequently injured than were partners of taller officers Table 25). The Cost of Assaults: Paid Injury Leave An average of about two hours of injury Leave wore taken per man- year as the result of assaults in Dallas. Because the average length of paid Injury leave vas aot available ia the Dallas data, an eatinate of the Length of leave vas taken from date contained in the Uniform Crime Reports for 1972. Using the UR data, it was cstinated that injuries lasted between three and nine days, and, therefore, it vas decided to use six days as the nominal Value from which co make the calculations in this section of the report. During the 13-nonth sample pertod i Dallas, there vere 152 as- saults among 750 officers in the departaent, or 0.22 assaults per nan-year Which te elightly higher than the national average of 0-15 assaults per nan~year in 1972, “An estinated 62 percent of the assaults, a higher rate than the ‘national average of 38 percent of 1972, resulted in an injury to an offices and, of the injured officers, ten percont miseed some workdays due to thel injury, Estimating that oix vorldays were lose per injury, there were 0.08 days of injury leave resulting fron assaults per men-year, The calculation Eollowe 0.22 assaults x 0.62 injurtes x 0.10 work-Lose injurtes x § days injury leave injury ‘work-loes injury = 0.08 days of injury leave ‘man-year.‘table 27 ‘THBE OF ASSAULT IN DALLAS TS TRE A STATISTICALLY SIGNIFICANT RELATIONSHIP ASSAULT TYPE. o Q)— WITH THE ovrrceR’s HEIGHT?® officer assaulted érom Front 2168 Side and/or rear 8 1 Xe Total” 100 ur Sniper, aabush, or trap yes 3 4 vo a Xo Total” 100 134 Assatlants used o threatened the use of a weayoo, and, fst, feet, ete. 67 19 Fireara "7 a1 Gateing tostriment ‘ rn % Othera oF other combinations 0 a8 Total” 100 179 WOTR: Data are from May 5, 1973 to June 15, 1974, ‘AE the 0,10 level. » Breluding missing data.30 TIT, NASSAU COUNTY, NEW YORK ‘The Nassau County Police Department (Wow York State) had 3,927 officers during the survey period of January 1, 1972 through June 30, 1974,” During the survey period, there vere 236 asagults on officers. Sone of the assaulted officers vere not essigned to patrol precincts and would, therefore, have been difficult to match to officers in a comparison group. Hence, data vere collected only for the 223 assaulted officers vho wore assigned to patrol precinets. A comparison group of 336 officers who were not agseulted was identified, and 251 of these were "randoaly selected aa they appeared fon rosters of each of eight patrol precincts based on the alphabet."14 The data subuitted for analysis did not contain information on ether the seniority of the officere or the shifts to which they were assigned, ‘which prevented the researchers from conducting a preliminary analysis to determine vhether officers’ heights were correlated vith sone other background charactertatics, Data on the following six performance measures vere collected for the assaulted officers and the couparison group: © assaults (including attempted assaults and assaults Fesulting in deliberate or accideatel tajuries), saute accidents, © citizen complaints, © departnent comendations, ‘* departnont complaints (resulting in at Least the prepa zation of an official departnent form--"Fora 203"), (¢ Anjuries on duty. Of these six measures, only the amber of sustained citizen complaints per officer vas related to height in a statistically significant way—with the number of conplaints against shorter officers being more numerous than against taller officers (see Table 28). Ti Tauia T-Frank, Couatestoner of Police, Nassau County Police Department, to Tox White, The Urban Institute (December 20, 1974).51 ‘Table 28 SUMWARY OF RELATIONSHIPS BETWEEN HEIGHT AND ‘PERFORMANCE MEASURES IN NASSAU COUNTY PERFORMANCE MEASURE DO THE DATA SHOW A STATISTICALLY SIGNIFICANT RELATIONSHIP BETWEEN THELGHT AND THE PERFORMANCE }EASURE?® Assaults No Injuries on Duty No Departzent Commendations Yo Departaent Complaints (Fors 209) Yo Sustained Citizen Complaints yes Department Auto Accidente 0 WOTE: Data are from January 1972 to June 1974. ‘At the 0,10 level of significance, RELATIONSHIP BETWEEN HEIGHT AND PERFORMANCE This section of the report briefly discuss bles and ite relationship to hetght. each of the performance varia~ aulte No significant relationship vas found between height and assaults. The data are displayed in Table 29, which shovs the percent of assaulted and non-assaulted officers of different heights.Table 29 ‘COMPARISON OF SAMPLES OF ASSAULTED AND NON-ASSAULTED ‘OFFICERS OF DIFFERENT HEIGHTS IN NASSAU COUNTY WORSE OF OFFTOHIS ———CIMTTATIVE PERCENT DIFFERENCE IN SAWGLE EIGHT BETEEN ercer oa Fon comuative laches) Assaulted Assaulted Assaulted Assaulted PERCENTS, os a ” 12.3 12a 0.2 6 66 7 38.6 317 ote 70 a a3 56.9 aa n ry 36 29 Ba 0.2 n 2 6 e.7 2 & 10 35.2, ™ 8 6 98.4 8 2 7 99.2 6 2 1 100.0 99.1 0.9 ” ° 2 100.0 100.0, 0.0 Total = 51203 MOTE: Data are from Jan. 1, 1972 to June 30, 1974. This ts the greatest difference in the cumulative percents of non-assaulted and assavlted officers. The difference ts not statistically significant, feven at the 0.20 level, (Note that officers who were 73 inches or shorter Were slightly less likely to be assaulted than were the taller officers.)33 Injuries ‘There was no significant trend relating height and injuries. To distribution of on-duty injuries for officers of different heights ts shown in Table 30, which indicates that there were a0 statiatically sig- Aificant differences between seaailted and non-assaulted officers of different heights. Since the assaulted and non-aseaulted groups have the sane height distribution, the data have been cosbined in Table 31 vhich presents the ausber of injuries (rather than just the number of injured people) for a Combined sample consisting of the assaulted and non-ageaulted groupe. Ths table shove that there were no statistically significant differences in injuries to shorter or taller officers. Other Performance Data Data for commendations, complaints, and auto accidents have been combined, since the rates per officer in the tvo groups (assaulted versus Ron-assaulted) vere not significantly different, and there was no significant difference in the height distributions. ‘The data for these additional performance measures are displayed 40 a series of tables, a8 follows: © department commendations Table 32 (© department complaints Table 33 © citizen compiaints Table 34 © auto accidents Table 35 No significant trends were observed 1a comparing an officer’s height vith department commendations, department complaints, or auto accidents. ‘The efngle significant trend shoved that 2 shorter officer had 2 higher chance of getting sustained citizen couplaint. Although only 38 sFeent of the officers in the sample were 69 inches tall or less, they Feceived 74 percent of all austained citizen complaints. During & two and fone half year period there were 39 sustained citizen complaints anong the 474 officers Am the combined eauple, or one couplaint for every 30 officer yesre worked, The shorter officers (69 inches or less) received one complaint per 16 man-yeare worked, coupared to one couplaint per 73 man-years worked for the taller officersssouvoqsTuBzs 30 Teas 2°9 B42 22 ono ‘queatsTulrs ATTesFaey3e38 tou SF 9ouoz0z37P our “poaTMesse oon OUR S1097;I0 JO PUP peatneess ou 94an oyn 62997330 Jo EIUsoIed DATIETMEND OY UF VVOIITP 369I0IT OY ET STLy sytet f0¢ our 03 zU6T “T “ue Eoxy 9x8 waEq :ALON sh sor ost ee 182 err oo oroot——oroot 3 ° a ° u v0 65 voor , 1 1 z % or 96 66 8 z ‘ z se « or or 8 a on or % 1 st ” ca u 6 of 9 u 9 o ” o oo oc us 9 cy 9s or a © 99 postaesey pearnesey pesynessy poatnesey —poatnesey poarneeey sy “=o sung aATZY IANO TST aT FRE AT PEROT TBS A ar NEGAIGT —GETNMENT ALNG-NO gO OL SSTANENT_ LN suaoTAdO 40 wHeHNN DTH soNmWaagT¢ —INIDWZd SATIVIANND —-No 40 WAG TWLOL ‘AINROO AVSSYN NI SLHOTSH uNawaaaTG 40 SuaOTAAO ORLTAVSSY-NON GRY GHLTAVSSY OF SaTeOrKT oc erger,Table 31 TAJURIES FOR COMBINED SAMPLE OF OFFICERS OF DIFFERENT HEIGHTS IN ASSAII COUNTY aa CoAT PROT uerout —-Oficare DIFFERENCE (laches) in RETWEER combined cuMATIVE Semple Injuries «Officers Injuries —_-PERCENTS. 68 58 n 12.2 a7 0.5 Cy 3 ws 38.0 347 33 10 86 98 56.1 SL 42 n 7 81 Rt 66.2 n 36 soz aa Beat 0.3 B 4 38 91.8 90.8 0.8 ™ 26 3 96.7 96.6 oun 8 9 10 98.5) 98.4 on 16 3 5 99.1 99.3 0.2 ” 2 2 100.0 100.0 0.0 Total 474 566 WOTE: Data are from January 1, 1972 to June 30, 1974. ‘This 1s the greatest difference in the cumulative percents of the the number of officers and of the number of injuries, ‘The difference is not statistically significant, even at the 0.20 ievel. (lote that officers 71 Anches and shorter vere slightly less likely to be assaulted than vere taller officers.)56 Table 32 DEPARTMENT COMMENDATIONS FOR COMBINED SAMPLE OF OFFICERS (OF DIFFERENT HEIGHTS TN WASSAU COUNTY ae LTTE PRCT uerout OF Ficers DIFFERENCE, Geehes) tn SEVERN Cosbined Departnent coxtatvE, Sample Commendations Officers Commendations PERCENTS cy 38 7 122 a ha ° 13 19 38.0 36 2.0 1» a "7 56.1 33 3.0 n 16 na cy 37 n 56 n o.4 0 ah 8 36 2 1.6 ca “08 % ey 3 96.7 35 er 8 9 3 98.5 98 0.5 16 3 1 so. 99 on ” 2 1 100.0 100.0 0.0 Tort 47h 100 MOTE: Data are from January 1, 1972 to June 30, 1974. This {s the greatest difference in the cumulative percents of the number of officers and number of comendations. The difference is not statistically significant, even at the 0,20 level of significance.37 Table 33, DEPARTMENT COMPLATNTS (FORM 209) FOR COMBINED SAMPLE (OF OFFICERS OF DIFFERENT HEIGHTS IN NASSAU COUNTY re Tre weront ‘OF Ficere DIFFERENCE (aches) ia Departaent. ‘BETWEEN Coubined Complaints CUMULATIVE Sample (Form 209) Officers Complaints PERCENTS “ 38 ‘ 12.2 57 6.5 8 123 wv 38.0 30.0 8.0 70 86 10 56.1 443 ust n 19 16 nt ona 5.6 2 36 7 24.6 ma 1 2 34 n a6 92.8 oe % 6 5 96.7 100.0 33 % 8 98.5 100.0, ous) 16 3 99.1 100.0 0.9 n 2 100.0 100.0, 0.0 Total 474 70 WOTE: Data are from January 1, 1972 to June 30, 1974, ‘This 4s the greatest difference in the cumulative percents of the number of officers and number of complaints. The difference is not Statistically significant, even at the 0.20 level of significance.58 Table 34 SUSTAINED CITIZEN COMPLAINTS FOR COMBINED SAMPLE OF OFFICERS (OF DIFFERENT HEIGHTS IN NASSAU COUNTY Ra RTCA TE PRCT user ‘Officers DIFFERENCE (inches) in Sustained BETWERR Combined Citizen Sustained CUMULATIVE Sample Conplaints Officers Complaints PERCENTS oa se 7 1709 ° 123 a Th 0 a 2 79.5 23.4 n 2 ‘ 89.7 17.0 n 58 ‘ te 100.0 -15.6 B 3 91.6 % 2% 96.7 % 9 98.5 6 3 99.1 ” 2 100.0 Total 74 38 NOTE: Data are from January 1, 1972 to June 30, 1974, This 1s the greatest difference in the cumulative percents of the number of officers and number of eustained compleiats, The difference is statistically significant at the 0.001 level of significance, Officers who are 5 feet 9 inches or shorter received more sustained citizen complaints then did taller officers.ACCIDENTS IN DEPARTMENT VEHICLES FOR COMBINED SAMPLE (OF OFFICERS OP DIFFERENT HEIGHTS TH NASSAU COUNTY 59 Table 35 GE TE urront ‘OF Ficers Accidents DIFFERENCE, Inches) in fo BETWEEN Goabined Department, coMATTVE Sample Vehicles Officers Accidents —_-PEROENTS 68 38 3 12 10.4 1.8 cy va 36 38.0, 33.7 0 a6 48 36.1 33.7 2b n % 46 7 28 mo 2 36 3 a8 son our a 4 ” 91.6 93.2 -1.6 % % 2 96.7 98.2 18 6 ° 2 98.5 99.0 5 6 3 2 99.1 100.0 0.9 ” 2 100.0 0.0 Tord ah 240 WOTE: Data are from January 1, 1972 to June 30, 1974. This 12 the Largest difference and 1s not statistically sigaificant, feven at the 0.20 level of significance,so PERFORMANCE MEASURES! ASSAULTED VERSUS NON-ASSAULTED OFFICERS in this section of the report, data on assaulted officers were examined to deteraine how they differ from data on non-assaulted officers, As expected, it was found that assaulted officers vere injured nore fre~ ‘quently than non-aseaulted offtcers—with the injury rate of non- assaulted officers being only 35 percent of that for assaulted officers, dmplying that at most 65 percent of injuries suffered by assaulted officers vere the result of the assaults. There were no significant differences between assaulted and non-assaulted officers in complaints, comendations, or auto ‘accidents (see Table 36). ‘table 36 (COMPARISON OF ASSAULTED WITH HON-ASSMULTED OFFICERS (ON PERFORMANCE MEASURES IN NASSAU COUNTY ASSAULTED NON-ASSAULTED PERFORUNCE MEASURES orricens ‘OFFICERS. Nunber in saaple 223 251 Department coanendations 0.24 0.18 per officer in sample Department complaints (Form 209) 0.17 oa per officer in sample Sustained citizen conplaints 0.076 al 0.083 per officer in sample Depremental auto accidents 0.53 0.49 per officer in sanple Oncduty injuries per 1.82 0.64 officer in sanple NOTE: Data are fron January 1, 1972 to June 30, 1976. Statistically significant at the 0.10 levels61 IV, OTHER POLICE DEPARMIENTS Tis chapter discusses the analysis of data submitted by Des Hoines, ova; Dade County, Florida; and Oakland, California, For various reason: explained in this chapter, the data from these areas were considered to be of Less inportance than the data fron Dallas and Naseau County. DES MOINES Since the Des Moines Departnent of Police adhered to the original data collection schene and provided 1nforaation Linited to the height of 1BL assaulted or injured officers snd 181 non-assaulted officers, only a Phase One analysis was conducted on these data. It was determined that an officer’s height vas not sigaificantly related to the Likelihood of an assault or injury. Consequently, no further deta vere requested fron the department. The data from Des Moines, covering the period of May 15, 1972 through Mareh 13, 1974, are displayed in Table 37. DADE COUNTY ‘The Public Safety Department of Metropolitan Dade County, Florida provided data on assaults (i.e., substantiated verbal assaults, assault and battery, and incidents of resisting arrest). The 355 officers who were assaulted during the sample period of Septeaber 1, 1973 to April 30, 1974, were involved in 253 cases, some of which involved agéaulte on more then one officer. Data included detatled information on the duty status of the officer davolved jault, on the characteristics of the assailant, and on vhether a iS used, Dade County also provided data on a sample of 1,142 evorn personnel that included ranked officers (Lleutenante and sergeants) and uaranked officers. Data on these personsel were for the same sample period, In this data, it was found that sergeants and Lieutenants were taller than unranked officers and were involved in far fever assaults. These trends are shown in Table 38,a Teble 37 ASSAULTS AND INJURIES 70 OFFICERS 1H DES MOINES mS COTATIVE PRON EIGHT Taple Sample Sample (laches) of of of Officers Incidents Of Ficers cy ” 54 20.4 23.8 70 a 3 431 ao n 0 2 33 n 2 19 12.4 38 3 20 3 83.4 0.0 ™ 16 2% 92.3 96.7 bh 5 4 6 100.0 100.0 0.0 ALL heights 181 1st NOTE: Data are from May 15, 1972 to March 13, 1974. ‘Incidents were either b This 4s the greatest difference between the cumulative percents of officers ‘and incidents, It is not statistically significant, even at the 0.20 level Of significance. jsaults on officers or injuries to officers.6 Table 38 SUOIARY OF THE ASSAULT EXPERIENCE OF OFFICERS, SERGEANTS, AND LIEUTENANTS I DADE COUNTY sma ASSAULT EXPERIENCE OFFICERS, SERGEANTS. ‘LERUTENANTS Muaber {2 depertment 869 210 6 Sone 1974 Percent under 70 inches 31.6 20.0 127 Shortest height (aches) 62 63 cy Muaber of assaults 3a 3B 1 Assaults per officer 0.392% 0.062 during sample period Percent of assaults 57.8 38.5 ° where officer injured NOTE: Data are from Soptenber 1973 to April 1974. “The nunber of assaults on unranked officers vas greater than both the ‘number of assaults on sergeants and on Lieutenants. The difference de statistically atgaificant at the 0,01 level of aignificance, ‘The data provided by Dade County on the sample of 1,142 svorn personnel did not permit a comparison of sentority or assignments. Nevertheless, an analysis was conducted of the physical assault experience (excluding verbal assaults) of the unranked officers in this Sample, Based on this analysis it was determined that © the height of officers did not influence the 1ikeli~ hood thet they vould be assaulted; and ¢ the height of officers did not influence the Likeli- hhood that they vould be injured as the result of an assault (see Table 39).9g 03 Ayo4rT S87 20 9x0n oF 2: XINNOO 40VE KI SUNOTHH INGHRAETG 40 $H301420 NO. Mowod ATIG0g YO ‘HIZGL ‘igaa ‘SISTA ‘SUAVE AG SLTaVSSY 6c eae65 Fever renked offie« + sergeants and Lieutenants) than unranked officers were shorter than 69 inches; 20 percent of the sergeants, 13 Percent of the Iteutenants, and 32 percent of the unranked officers vere shorter than 69 inches. This difference in height 1s stattstically vig~ nificant (see Table 40). Sergeants wore much less 1skely to be aseaulted than vere unranked officers. There were 341 people vbo assaulted unranked officers and only 13 who adsauited sergeants. The type of weapon used is these assaults ad the frequency of injuries to unranked officers and sergeants are shown in Table 41. (Because there vas only one assault mong the 63 lieutenants, the sample size was too mall to include in this analysis or in Tablet.) Detailed data on 249 assaulte were provided by Dade County (see Appendix 3). The distributions of the heights of a sample of all officers (including ranked and unranked officers) and of a sample of assaulted officers are sho in Table 42, There vas no significant difference in the two distribu= tions, vhich leads to s conclusion that height did not have any tafluence fon the Likelihood of an officer being assaulted in Dade County. ‘The Relationship Between Height and Other Characteristics of Assaults ‘Thirteen general characteristics of the assaults vere exanined to determine whether short officers vere more or less prone to involvement in different types of assaults. Significant relationshipe vith the height of the assaulted officer would be expected between one and two tines (using independent tests of significance). The folloving two differences were statistically significant: © aesatante of taller officers were sore likely to be intoxicated, © assaults on shorter officers tended to occur more often when the officer was alone. No statistically significant relationships vere found between an officer's height and the folloving characteristics: © race of assatlant(e), © sox of assatant(s),66 ‘Table 40 [HEIGHTS OF OFFICERS, SERGEANTS, AND LIEUTENANTS IN DADE COUNTY. (COMMLATIVE PERCENT BY WEIGHT EIGHT Tamales, (Inches) officers! Sergeants” _Leutenants 62 63 6a 65 66 6 68 6 70 n 2 40 B 37 % 93 5 97 16 99 99) 99; 00, : n "a : 9 100: MOTE: Data are from Septeaber 1973 to April 1974. 869. ® W = 210. We 63, a This 4e the greatest difference between the cumulative percent of Adeutenants and of unranked officers. The difference is statis~ tically significant at the 0.10 level of significance. ‘This s9 the greatest difference between the cumulative percent of sergeants ané of unranked officers, The difference 1s statistically significant at the 0,001 level of sigaificance. CONCLUSION: A much smaller percent of sergeants and Lieutenants than of officers are shorter than 69 inches tall.Table 41 WEAPONS USED BY ASSATLANTS AND TKJURIES To OFFICERS {AND SERGEANTS. IN DADE COUNTY pence OF ASSAULTS WEAPON USED BY ASSAILANT(S) one OF RESULTING ASSAILANTS = IN THUURY. Assaults on Unranked Officers Hands, feet, fists, tecth, bodily force 274 6a.2 Fireare 7 Ine cub 1 85.7 Cutting of stabbing instrument 6 16.7 other, 3 37.8 ‘ALL weapon types 3a 57.8 Assaults on Sergeant Hands, feet, fists, teeth, bodtly force Firearm cub Cutting or stabbing instrument other ‘ALL weapon types 3 38.5 NOTE: Data are from September 1973 to April 1974, Only one incident involving a Lieutenant vas reported.Table 42 ALL, TYPES OF ASSAULTS ON OFFICERS AND OFFICIALS ‘OF DIFFERENT HEIGHTS IN DADE COUNTY mn COMLATIVE PERCENT uercet = AIT asaadted © AT Reeaulted (Inches) Officers Officers —«OFEcers Officers DIFFERENCE 1, or and and CUMULATIVE Officials officials Officials offielale —PERCENTS 2 1 on 63 a 1 012 64 5 1 0:8 6 3 12 66 7 1 20 6 31 3 78 68 Ey a 18:6 o ns 25 ang 70 139 3 8 n 145 a7 64.3 2 133 3 80.3 2 6 1» 87.3 % 57 18 93:9 5 2 2 97.1 16 7 3 gar 7 6 o 99.8 8 1 0 99.9 9 1 4 100.0 100.0 ALL heights 669 24 NOTE: Data are from Septenber 1973 to April 1974. ‘This is the greatest difference in the cumulative percent of all office! and officials and of assaulted officers and officials. The difference Ae not statistically significant even at the 0,20 level. > Data on four eseaulte were missing.6a © nunber of assailants in an incident, © type of weapon used or threatened by assatlant, '* whether the identity of assailant was know to officer, © type of weapon used by officer, 4 officer's activity (naking arrest, ete.) ‘© whether officer vas injured by the a sult (© type of weapon causing an injury to officer, ‘© whether there wore injuries to sore then just officer aceaulted. ‘The above results are summarized in Table 43. Data on the characteristics of aseailante are show in Table 44, Assailants vere most Iikely to be males who were less than 30 yeare old, and who vere acting clone. Just over half of the asasilante were black. The Weapon most commonly used by assailants vas a part of the body. Alaget «ll assailants vere not previously knows to the officer they attacked, These es, along with the percent of assaults made by intoxicated As shown in Table 47, the majority of the assaults occurred in incidents in vhich the officer vas attenpting an arrest. In about half tthe cases the officer reportedly did aot uae a weapon. In 14 percent of the cases the officer vas assaulted when no other officer was presents tthe percentage is higher (212) for shorter officers, as shown in Table 43. The chances of a second officer being assaulted do not seem to be influenced by the height of the first officer assaulted, a2 indicated in Table 49. The majority of assaults in Dade County resulted ia an injury to the officer, and the injurtes vere most often due to the assailants use of bodily force. oaKLAND The Police Department for the city of Oakland, California, provided data on "negative encounters” betveen officers and citizens that resulted in a charge of resisting arrest, or assault on an officer, or in an officially Tecorded negative interaction between police and citizens: The data covered 43.8 year period from January 1, 1970 through October 31, 1973. In addition to negative encounters, the data included all on-duty injuries for vehicular ind industrial accidents. The sample sizes for the Oakland data are sum rized in Table 30.Table 43 ‘SMOURY OF Ai ANALYSIS OF THE CHARACTERISTICS OF ASSAULTS AND THETA RELATIONSHIP TO AN OFFICER'S HEIGHT IN DADE COUNTY TS THERE ANY SEATISTICALEY SIGNIFICANT RELATIONSHIP CHARACTERISTIC OF ASSAULT WITH THE HEIGHT OF THE OFFICER ASSAULTED?® Assailant characteristics Age No Race Xo Sex No Number per incident No ‘Type of weayon threatened or used xo Condition (intoxicated, normal, ete.) Yeo Tdentity known by officer pict to assault No officer’s actions Weapon used or not Xo Activity (arrests, other) Xo Other officers present or not Yee Injuries Officer injured or not 0 ‘Type of veapon causing injury to officer No Tajuriee other than to officer Yo Nore: Data are from Septenber 1973 to April 1974. "at the 0.10 Level,n Table 44 CHARACTERISTICS OF PEOPLE WHO ASSAULTED POLICE TIN DADE. COUNTY Ts Tae STATISTICALLY SIGNIFICANT ASSAILANT CHARACTERISTICS ® RELATIONSBIP, WITH uBrcuT?® ‘Age(s) of assaitant(s) 13 ~ 19 years a 8S 20 - 30 39108, Xo Over 30 3080 Total roo a Raco(s) of assailant(s) Caucasian 54 Black 46 113 Yo Total 100? Sex(es) of assattant(s} Male 2 Fenale 18 Wo ‘Total 100 Momber of aseatlants per incident one 86215, wo 9 2 No Over two. 5 Total 1002s” NOTE: Data are from Septeuber 1973 to April 1974 ‘At the 0-10 level. b Bxeluding atesing data; counting the umber of aseatlante, which some tines is more than one per assault.n Table 45 (CHARACTERISTICS OF ASSAULTS IN DADE COUNTY Ts THERE STATISTICALLY SIGNIFICANT cunacrenrst1c ® 0 RELATIONSHIP, Woe wErcHT?® Weapon used or threatened Renda, fiste, feet, bodély force 73 194 Firearms 3 oi Cutting instrument as to Other Weapon ue ‘Total 100 263° Condition(s) of assailants) Normal n 179 Intoxicated 38 High on druge 6 ves Mentally impaired 27 Hone of these Soa Total 100 asi Identity of assailant inom by officer prior to assault Yes “a Yo 95 238 Xo ‘Total S 100249 NOTE: Data are from Septenber 1973 to April 1974. ‘At the 0-10 level. Bven at the 0.01 level,B Table 46 ASSAULIS BY INTOXICATED OR MEWTALLY IMPAIRED PEOPLE, ‘ON POLICE OF DIFFERENT HEIGHTS IN DADE COUNTY ASSAULES, BY CONDITION OF ASSAILANT Intoxicated, Tigh om EIGHT Drugs, Megeally Tnpatred, (Inches) Normal and Other™ oral ae @ oo ow 69 and ender B55 23 8 10073 n- 73 73 105 a 38 100 146 7h and over 5619 4 1 10034 ALI heights nous Fd n 100 251 NOTE: Data are froa Septeaber 1973 to April 1974, Intoxicated only, W = 58. CONCLUSION: Assaults on tall officers are more Likely to involve an intoxicated assailant (chi-square = 4,72, probability = 0.095).Table 47 ACTIVITY OF OFFICERS AT TIME OF ASSAULT TN DADE COUNTY Ts aE STATISTICALLY SIGNIFICANT ‘Tire oF activiTY o® ‘RELATIONSHIP, WITH HEIGHT? lias a weapon used by the officer? Yes ssa Yo a 105 Yo Total assaults 100 se Way an arreut betng sttenpeedt Yes 61 ast Yo 3 8 Yo ‘Total assaults roo 24s? ere other officers present? ves 86 201 Wo wo 8 Yes Total assaults too 230? NOTE: Data are from Septenber 1973 to April 1974. “At the 0-10 level. » Excluding missing data.5 Table 48 PRESENCE OF OTHER OFFICERS DURING ASSAULTS ON OFFICERS OF DIFFERENT HEIGHTS IK DADE COUNTY other EIGHT Officer (inches) Present @ 69 and under 35 m-73 90 120, 14 and over Be 28 ALL heights 86 201 No Other. officer Present total 2 Mm @ @ 2 1510070 113100 133, 16 Ce 16 331002348 WOTE: Data are from September 1973 to April 1974. ‘eeluding missing date. CONCLUSION: A slightly Larger fraction of assaults on shorter officers occurred when no other officer was present (chi-square = 5.23, probability = 0.07).Table 49 SMOURY OF TYPES OF INJURIES IN DADE COUNTY Ts Tene STATISTICALLY SIGNIFICANT STATUS OF THJURIES o RELATIONSHIP ‘wETH HEIGHT?® xee 57a Xo 4 106 to Total assaults 100 ag Cause of injury to officer Hands, fists, feet, bodily force 64 116, Other’ weapon’ 38S No ‘Total use of weapon 100181? Were there injuries to other than the assaulted officer! None a0 201 Officer's partner wm Other officer 2 Yo citizen 2 Total 100 asi OTE: Data are from September 1973 to April 1974. ‘AL the 0,10 level. » Counting miltiple occurrences and excluding missing data,n Table 50 SUMMARY OF OAKLAND, CALIFORNIA POLICE DEPARTWENT DATA SAMPLE SIZE FOR ‘THE BUREAU OF ‘oral FIELD OPERATIONS SwM@LE (PATROL, TRAFFIC & ‘rvees oF pata SIZE SPECIAL” OPERATIONS) Number of Officers as Men-Honthe Worked 30,279 Officers taving Encounters 693 638 Negative Encounters 17,497 ni,o75 Encounter Injuries to Citizens 353 756 Encounter Injuries to Officers 8,505 1,352 Officers with Encounter Injuries 626 563 Industrial Injuries to officers 866 _ Officers with Industeial Injuries aa Vehicular Accidents 82 officers with Vehicular Accidents 367 NOTE: Deta are from January 1, 1970 through October 21, 1972.1 ‘The OakLand data failed to provide # directly comparable group with which to compare the performance of the shorter police officers Since shorter officers had less seniority than taller officers. officers ‘ho were 5 feet 7 inches tall vere working an average of 47 percent of the tine during the aaaple period, and 56 percent for officers 5 fect 8 inches, as compared to the department average of 73 percent.!5 This trend reflects the history of height requirenents a the departaent. Ia July 1968 the minimum height requirement was dropped to 5 feet 8 inches, and in July 1970 the requirenent dropped to 5 feet inches. Shorter officers were more Likely to have joined the force during the seaple period and therefore to have been present for a lover percent of the tine during the sample period. Because of the seniority differences, fev clear inferences can be draw from the Oakland data. Tor example, vhile it ts unclear vhat meaning should be attached to then, the folloving statistical findings can be noted: (1) officers who wore 5 fect 8 taches or shorter (shorter officers) had twice as many negative encounters with citizens as did taller officers, (2) shorter officers ere nore frequently tajured then taller officers, (3) officers vho wore 5 feat 9 inches tall--and were therefore eligible to join the deparement at any tine during the study poriod--had nore negative encounters than taller officers, aad (4) officers who vere 5 fect B inches or shorter vere more Likely than taller officers to be involved 4a vehicular accidents, If one accepts “negative encounters” as @ aeasure of risk rarner chen a5 a performance measure, then one would be most interested in the frequency With Uhich officers sustained injuries per negative encouater, By this Reasure, shorter officers veré no nore or less iajury-prone than taller officers. (Shorter officers sustained fever injuries per nogative encounter, but the difference just missed being statistically significant at the 0.10 Level, with the probability being 0.11.) There was no statistically significant relationship between officers* heights and the munber of induatrial injuries, Considering oaly officers who were 69 inches or taller--and were eligible to Join the departaent ducing the entire study period—there vas no relationship between height and vehicular aceidente. Negative Eacounters per Offtcer Man-Month During the 3.8 year study period, shorter officers (5 feet 8 inches or shorter) vere involved in more negative encounters per aan-soath than taller officers. However, the taller officers included supervisors, managers, and fa Larger munber of experienced patrol officers. is Data on man~nonthe worked during the sanple period are summarized tn Pigure 2.» ‘The exposure to negative encounters may have been greater for the short officers for the following three reason © the duttes of supervisory and managenent personnel (all of whom vere "tall") involved a reduced probability of involvenent in citizen officer conflicts; # taller officers were sore ox better able to avoid unneces having offictel records m © taller officers vere more experienced and may have been sonewbat more likely to receive low-risk assiguaeats (e.8-, Station oF traffic duty). ‘The nuaber of negative encounters per san-noath for officers of different heights {8 shom in Table 51, Officers shorter than 69 inches hhad almost twice as many negative encounters per man-nonth as other officers. How auch of this difference vas due to height alone (as compared to type of assignment and years of experience) cannot be detersined from Oalland’e computerized data base. The relationship betwoen height and encounters per man-nonth is statiatically significant. Figure 2 shove the expected change by hypothet~ Aeally excluding shorter officers. The computed impact vas far Larger than What could be attributed €0 chenec slonc. Figure 2 indicates that by flininating officera vho were 68 inches or shorter (and eliminating the Aseaulte om these officers) the average number of encounters per nan-nonth would have dropped about five percent. Te further shows that officers nthe Oakland Police Department who were 68 inches or shorter worked five percent of the man-nonths during the sanple period. Moreover, according to the figure, the elininstion of all officers who vere 71 inches or Shorter would have excluded half the man-monthe and vould have reduced the rate of negative encounters by more than ten percent. Another nethod of enalyzing the sane data is to exclude from the sample officers shorter than 69 inches, because these officers were not eligible to Join the department during the entire sanple period, Figure 3 shows that even for this restricted group the taller officers worked ore nonths during the saxple period—indicating thet taller officers vere Likely to have greater seniority than shorter officers, The taller (sore experienced) the officers, the fever negative encounters they had. Table 52 shows that, of the officers 69 inches and above, 49 percent of the man- ‘nonths worked were worked by officers shorter than’6 feet, but 54 percent of the encounters involved officers wader 6 feet, This difference in percentages is statistically significant. Injuries and Negative Bncounters In this section of the paper, the number of encounters in which an officer vas involved will be used as an estimate of risk of injury80 ‘soNmaaaTa axvDwo xt ‘SUHOTEH INEWAHAIO 40 SWAOTEAO ¥OX SAZINNOONA 40 SUSHAAN ANY GACH SHINOK 1s stars,a PERORNT CHANGE IN NEGATIVE [RNCOUNTERS PER OFFICER MaN-MONTH DUE TO EXCLUDING SHORTER OFFICERS +507] +204] 207] 101] -203} ~sor! wore: PERCENT OF ‘TOTAL, 002 DEPARTMENT ‘MAN-MONTHS EXCLUDED Munber of man-nonths = 30,279; this aumber is for sll 892 officers between Jans 1, 1970 and Oct. 31, 1973. The number of negative encounters = 12,437; thie nuiber 4s for all negative interactions between an officer and a citizen resulting in "Resisting Arrest" or e similar charge. Confience Linite on this figure indicate the probability that the entire solid curve would fall within the Linite show by the dotted Lines. Figure 2 ‘CHANGE, IN NEGATIVE ENCOUNTERS PER OFFICER KAN-MONTH DUE ‘70 HYPOTHETICALLY EXCLUDING SHORTER OFFICERS 1M OAKLAND, CALIFORNIAoa AVERAGE PER OFFICER: AVERAGE PER OFFICER: MAN-HONTES WORKED WEGATIVE ENCOUNTERS PER DURING SAMPLE PERIOD MAN-NONTH. WORKED 6 8 35 % 3A 2 3 ° 32 66 3 63 20 60 29 37 28 54 a 31 2% 4s 25 45 2% 2 23 8 2 36 a 2 20 30 o 68 68 1 Mm 8 Om OS HEIGHT OF OFFICER (raches) WNTE: ate are From January 1970 to October 1973. Figure 3 HEIGHT, MAN-MONTHS WORKED, AND NEGATIVE ENCOUNTERS PER MAN-WONTH IN OAKLAND, CALIFORNIA83 Table 52 ANALYSIS OF ENCOUNTERS OF OAKLAND OFFICERS OVER 68 INCHES TALL ‘COMILATIVE PERCENT DIFFERENCE. erent Wae-tonghe —egative , CUMULATIVE PERCEITS (Inches) Worked® Encounter 6 7 16.3 36 70 29.6 33.5 n n 2 " 90.9 90.3 0.6 8 100.0 100.0 0.0 NOTE: Data are from January 1970 to October 1973. W = 28,597; 824 officer: i N= 11,169, This 1s the point of greatest difference between the cumulative percent fof man-monthe worked and of encounters, The difference is statistically significant at the 0.05 level of significance. Officers vho were at luost 71 inches tall had somewhat sore negative encounters than did the teller officers.24 For the purpose of this discussion, it is assumed that officers aay have sone control over the sumber of injuries during the nogative encounters im vhich they become involved. This procedure of examining the number of injuries per encounter seaas to be a logical vay of trying to control some- vhet for the great disparity that vas found in the experiment involving the shorter end taller officers. The effect of eliminating shorter officers on injuries per encounter 4s show in Figures 4 and 5. Figure 4 shows the effect on officer injuries and Figure 5 shows the effect on citizen injuries. Eliminating shorter officers slightly increased officer injuries per encouster; hoever, the observed trend vas not statistically sigaificant (probability = 0-11). The reported nunber of civilians injured was too suall to observe any trends. The number of injuries per incident for officers of different heights As show in Table 53, The number of officer injuries was about 10 tines ‘as great as citizen injuries, ‘wo trends have been observed: shorter officers were involved in more encounters per man-aonth, but in fever injuries per encounter. The net effect on shorter officers was that they had a higher rate of injuries er man-nonth (as is illustrated in Figure 6). The heightwinjury relation Ship 1s statistically significant (as show ia Table 54), but the sual] Gitference (Less than -1 encounters per man-month worked) due to height nay have resulted from the lesser experience of the shorter officers. Industriel Injuries and Vehtewlar Accident: in Oakdand hed 0.023 vehicular accidents per sen-sonth ‘and 0,029 industrial injuries per man-moath, No statistically significant relationship was found between height and industrial injuries. There was 42 statistically significant relationship between hefght and vehicular accidente--shorter officers vere more frequently involved in vehicular seeidente,as PERCENT GHANGE IN OFFICER TAJURIES PER NEGATIVE COUNTER DUE 70 EXCLUDING ENCOUNTERS OF SHORTER OFFICERS a = a feist Se a mcent oF a nocarive xooz ENCOUNTERS EXCUIDED ~ -2 a = MOTE: Number of negative encounters ~ 12,437; thie number is for all officers between Jane 1, 1970 and Oct. 31, 1973. Mmber of injuries = 8,605; this number is for injuries sustained by officers in the encounters. Confidence Limits on this figure indicate the probability that the entire solid curve would fall within the Linits shown by the dotted Lines. The solid curve on this figure would be expected to occur by chance alone 11 tines out of 100. Figure 4 ‘CHANGE 18 OFPICER INJURIES PER ENCOUNTER DUE. TO HYPOTHETICALLY EXCLUDING ENCOUNTERS OF SHORTER ‘OFFICERS TX OAKLAND, CALTFORNIA8 PERCENT CHANGE 28 CYTTZEN TRUURIES PER NCOUNTER DUE TO EXCLUDING ENCOUNTERS OF SHORTER OFFICERS 454 a officer Height a soz 100% PERCENT OF ENCOUNTERS EXCLUDED NOTE: Confidence Limits on this figure indicate the probability that the entire solid curve would fall within the linite shown by the dotted Lines. Figure 5 CHANGE IN CITIZEN INJURIES PER ENCOUNTER DUE 70 HYPOTHETICALLY HXCLUDING ENCOUNTERS OP SUORTER ‘OFFICERS IN OAKLAND, CALIFORNIAa wo oO ee oro oso “ wo we ° sooo ® oro 880 om zu foro sto we u woo a oe wo xr © «ee ® ut on s semaancom eoseun navns ives oman: ‘une auvnwvo ar ‘SUHOTAH GAUUAEAIG JO SAHOLEAO HOE SAMALNT ANY SHALNNOOKE es orges,‘vrmoaTv9 ‘ONVDIVO AE SUTUMCMK GHY “S¥aLNNOONE ‘IHOTSH 9 canes HENOR- HR WE SeEIOOL on a os ost or ot ox sezunnonk MoUs saranrer 242089 Table 54 OFFICER TAIURTES DUE TO ENCOUNTERS AND HEIGHT IN OAKLAND OFFICER TAUURTES comattvs — commative uexcat FROM ENCOUNTERS PERCENT OF PERCENT OF (Eoches) PER MONTH WORKED worras "URES. 20 Wonke ‘OFFICERS o 123/347 = 0.35 Mw a 6s 648/1,335 = 0.49 35 9.0 o 1,267/3,650 = 0.36 17.6 23.5 70 1,385/4,846 = 0.29, 33.6 39.5 n 1,718/5,355 = 0.31 n 1,179/5,481 = 0.22 70.1 73.2 2 865/3,612 = 0.25, a 83.2 ™ 693/3,115 = 0.22, 916 o13 8 749/2,538 = 0.30 100.0 100.0 Total _8, 605/30,279 = 0.28 ‘This is the greatest difference between the cumulative percent of months worked and cumulative percent of injuries to officers. The difference is statistically significant at the 0.001 level of significance.90 ‘The data for industrial injuries and vehicular accidents are show in Tables 55 and 56, If officers under 69 inches (who were eligible to join the department only during the Latter part of the sample period) Were excluded, there vas no statistically significant relationship between height and vehteular accidents (even at the «2 level of significance Table 55, INDUSTRIAL INJURIES TO POLICE IN OAKLAND CuAATIVE PERCENT wEIGRT INDUSTRIAL THIURIES Man-Months Industrial (inches) ‘PER MAN-HONTH Worked Injuries or 13/347 = 0.037 68 42/1,335 = 0,031 6 111/3,650 = 0.030, 70 131/43846 = 0.027, nm 161/5,555 = 0,029 n 145/5,481 = 0.026 a 107/3,412 = 0.031 % T2/3,115 = 0.023 5 82/2,538 = 0.032 Total 866/30,279 = 029 ‘This is the point of greatest difference between the cumulative percent of months worked snd of industrial injuries. The difference ts not Statistically significant, even at the 0,2 level of significance,En Table 56 ‘VEWICULAR ACCIDENTS IN THE OAKLAND POLICE DEPARTMENT mmLaTive PeaceNTS CCUMMILATIVE PERCENT EXCLUDING OFFICERS HEIGHT VERICULAR ACCIDENTS SHORTER THAN’ 65. THCHES Clnches) PER MAN-WONTE Youths Vesicular” ~-Montha Vehicular Worked Accidents Worked Accidente 67 15/3k7 = 0.083 682/335 = 0.03155 69 95/3,650 = 0.025 17.6 2.3" 2815.2 7095/4486 = 0.020 336 36.2 2.7303 TL 10/5555 = 0.022 52.1 53.8 wa 86 72 98/5,481 = 0.018 7041 68.2 653° 73 93/3412 = 0.027813 BB a0.1 7H 6/3,115 = 0,022 91.6 91.6 90.8 7S $6/2,838 = 0.022 100.0 100.0 100.0 100.0 Total 682/30,279 = 0.023 ‘This 1s the greatest difference in the cumulative percents of san-nonths worked and vehicular accidents. The difference is statistically significant at the 0,02 level of significance, e ‘This 12 the greatest difference in the cumulative percents of man-aonthe Worked and of vehicular accidenta, considering only officers who are at least 5 fect 9 inches tall. The difference 19 not statistically significant, ‘even at the 0.20 level of significance.APPENDIX A STATISTICAL METHODOLOGY93 STATISTICAL METHODOLOGY ‘wo statistical techntquegywere used: the Kolnogorov-Smirnov To Sample Test, and chi-square test. In cases vhere data could be displayed by one-inch height intervals, a Folnogorov-Suirnov test vas used to compare the cumulative percentage Of the observations in each of two categories, This test a sensitive to any kind of difference in the distributions fron vhich the two sample: re dram. Tho Kolnogorov-Sairaov test vas used for the folloving three '* heights do not have to be partitioned into categories, which are alvays somewhat arbitrary; hen compared 2 a totaat, the Kyjpogoror-tnizoy test Te'highy efficient (about 900)51? and Le leo fo sore several cs SL ageace teeth 4¢ the Kolmogorov-Suiraov procedure permits computation of the statistical significance of the estimated impact of dropping officers who are below a given height. To use this test, compare the X, percent of officers who were less than or equal to A, inches tall and the'T, percent of che assauite chat were ade on officer ai inches tall or less. ‘A height h, is found chat produces the largest absolute differences between X, and Y,, dad the largest difference ax |X,- 1,] 19 used to accept or reject thé hypothesis that the two samples, assuted independent, were dram from the sane distribution of heights [Ae the 0.10 level of significance, the hypothesis te rejected if: 4 > 1a 7 12 Y 1 ‘where N, and, are the sunber of observations in each of the tw samples. 1 ‘See S. Siegel, Nonparametric Statintice for the Behavioral Sciences, 7 Teid., ps 136.6 when data vere arranged in groupe with greater than one-inch intervals in tables where the row percentages added to 100 percent, = ‘Chivaquare test for ke independent samples vas used (k {s the number of categories represented in the rove of the table). The hypothesis being tested was chat the percent of observations in each column was the sane for every row. If the chi-equare valve computed vas Large enough, che hypothesis was rejected at the 0.10 level of significance, (Chi-square teats must be performed on frequencies of observations and not on percentage distributions, as bat gonetines been done in previous police studies.) In none of the chi-square teste were any of the possible uodifications used, such as the correction for continuity or Fisher's exact probability test. he sensitivity of various rates (assault, injury, ete.) to height, vas sonetines examined by computing the percent change in'the Tate that could be expected by excluding all officers under a height h,. By excluding the X, percent of the shorter officers the resulting rate andng the reanin- ing officers differs froa the rate for all officers by the percentage 100% which can be computed aa a curve for various values of Xj Under the hypothesis that there is no systenatic height effect, ch® observed curve should be entirely contained in the following Limite with @ 90 percent confidence for all values 0 pe assaults filed between Jenuary 1, 1972 and Deceaber ‘From 2 random samples of 100 officers in the department uho vere not assaulted. ‘From the sample of assault reports.107 Table C-5 DISTRIBUTION OF ASSAULTS AND OFFICERS BY HEYGHT IN THE [NORTH PRECINCT, AFTERNOON RELIEF, PORTLAND, OREGON sR ‘CUMULATIVE PERCENT Uificers Aesaults “Officers Assaultw (Assaulted AVERAGE ‘and non DIFFERENCE ASSAULT PER BELGHT assaulted OFFICER 18 RANGE samples, ‘COMBINED (Inches) combined) SAMPLE, e702 2 no 48.2. 9.17 Meme1/2 B WE 4.29 7371/2 6 2 59.8 : 3.67 Above 75 3 2 100.0 a:67 ‘total 38 28 6.00 ‘This ds the Largest difference in the cumulative percents end is not statistically significant even at the 0.2 level, using a two-sample test. The chi-square value is 4.85 with 3 degraes of freedom using a tvo-sanple test, which is not significant, ‘even at the 0.1 level (probability = 0.19).108 Table 0-6 DISTRIBUTION OF ASSAULTS AND OFFICERS BY HEIGHT IN ‘THE CENTERAL PRECINCT, AFTERNOON RELTEP, PORTLAND, OREGON NUMBER TS SMIPLE CUMULATIVE PERCENT mance ———"——ttcers officers not. Assaults assaulted assaulted 69-70-1/2 13 7 2 12s Men1/2 2 2 3 ae W3e7H1/2 1 1 4 10:0 100 Above 750 Total 16 10 2 This 1s the Largest difference in the cumulative percents and te statis teally significant at the 0.01 level weing » two-sample Kolmogorov-Seirnov “A STUDY OF POLICE HEIGHT REQUIREMENTS, “Geepared by Coke Deapaayy Tonas Depat= ent of Public Safety, April 1974.) Inquiries were mated to 403 agencies in the United States, and 193 responses were received. Anong the data provided aa a result of the inquiries were studies and related information from state and city police departments. Seven studies provided the bulk of the supportive evidence cited; these studies include material from departaents in the following cities: © San Diego, California, © Portland, Oregon, © Bansvilte, Indiana, # Seattle, Washington, © Washington, Di109 ¢ Beauont, Texas Mami, Florida, © Cincinnati, Ohio and (¢ Dee Moines, Iowa. Data were provided by seven state agencies, 39 cities, and 11 other agencies. A valuable service was rendered by collecting data. Unfortunately, the Rethodological and arithmetical errors in the study are so frequent and Serious that the reader cannot judge vhether many of the conclusions are valid, Bleven tables wore presented in vhich the distribution of police officers and incidents were displayed for officers of different heights. Four of the tables vere taken directly froa the San Diego study, and the statistics appear to be correctly computed. Six other tables contain improperly computed statistical resulte-—the chi-square values have been incorrectly cosputed by weing percentages of the observations rather thas the auaber of observations. In another table, the level of significance of ‘the Fesults is incorrectly interpreted because the wong degrees of freedom were attributed to the chi-square statistic. Tn sumary, serious errors ‘are apparent in seven of the eleven tables for which chi-square tests vere 1H EVANSVILLE POLICE DEPARDIN’s MENIMM {EIGHT ReQUIREVENT: A BONA FIDE GOCUPATIONAL TALIFICATION," (Prepared by the Evansville Patane Police boportaeat, Personnel sod Training Diviston, Planning and Research Section; Novesber 6, 1973,) This {s a comparatively thorough effort to exanine the relationship between an officer's height snd perfersance. The principal defect in the eudy ie that Evansville, like other departnents in the country, has a history of lovering its height standards (t.e., the height standard vas Lowered to 68 inches in 1965). Hence, the most senior officers also are amore likely to be the taller officers and the aizple height-performance Telationships shown in the study aay be Very aisleading. 229 officers studied over 21 months, 35 physical sbuse complaints, 71 verbal abuse couplaints, and 50 injuries. The study presente data on the relationship between height ‘and three measures of performance: physical abuse complaints, substan Ciated or unsubstantiated verbal abuse complaints, and injuries, Statis~ tical analysis of these data, using a Xolnogorov-Smirnov two-sample test, indicates that’110 © there was no statistically significant relationship between height and physical abuse complaints (the test is almost statistically significant at the 0-10 level), with officers 69 inches or shorter comprising 18 percent Of the departnent and receiving 40 percent of the complaints; ¢ officers who were 69 inches or shorter received sore sub- stantiated verbal abuse couplainte (substantiated and un substantiated) than taller officers (significant at 0.05 level); with these shorter officers conpriaing 18 percent of the deparment and receiving 44 percent of the complaints; * officers who vere 69 taches or shorter vere nore Likely to be injured during encounters vith citizens (aignificant at 0,10 Level); with these shorter officers comprising 18 percent of the department and receiving 40 percent of the injuries. ‘The study did not present any data concerning the ausber of arrests sade or number of commendations received by officers in the sample (earlier, it was reported that officers who receive complaints also are 1ikely to receive commendations) « 4 STUDY OFTHE MINIM HEIGHT REQUIRRIENT FoR ‘TR, Cochran, Arizona Departnent of Public Safety, Planning and Research (undated) .) Although undated, thts short paper was apparently written some- time after March 1973. Te consists primarily of a review of about a dozen documents related to height of police officers and does not make any clain to being a complete research project on the subject. The report makes an interesting observation: the Phoenix Police Depertacnt reports that the average height of sale suspects aseaulting police officers vas 691/2 daches, ‘compared to an average height of the officers involved of 71 inches (the aseailants are, on average, shorter then the officers assaulted). "A DESCRIPTIVE PROFILE OF THE ASSAULT INCIDENT ‘Wrepared by Samual C. Chapnan and Cheryl G. Swanson, April 30, 1974 (abstracted from a prograa report, "Assaults on Police Research Project,” at the University of Oklahona).) An extensive study of assaults on police officers was being conducted by Samuel Chapman of the University of Oklahoma. (According to T.R, Cochran, this study has been discontinued.) The study is based on 4a 'sample of 1,143 assault incidents, for vhich data were collected on the following fout dimensions:mm © officer characteristics, jantlant characteriatics, assault environnent, * dmnantes of the assault event. Most of the deta came fron cities in Oklahoma, Although the study provided data on the distribution of height of officers assaulted, 1¢ did not reach any conclusion about the 1ikelshood of assaults on officers of different heights, because no data were available on the heights of nos-assaulted officers, A correlation analysis was performed on the heights of the officer and of the assailant, and the correlation was found to be very low (0,001, hot statistically significant, even at the 0.2 level £ significance). This Finding suggests, that chere vas no connection between height differences and an assailant’s decieion to attack an officer. Assaults occurred auch more frequently (86,2 percent) between 4 P.M. and 4 A.M. then during the remaining 12 hours of the day: only 13,8 percent Gecurred during the rensining 12 hors, ‘This shows chat the exposure co. assaults can be highly dependent on the hours of the day which the officer works. More recently Swanson and Hale have published an articlel9 on thete alysis of the data, Tesults cover « survey of 1900 police officers in 13 ‘municipal police agencies?0 during the calendar year 1973, By cosparing the 376 officers who were assaulted one or more tines vith the remaining 1524 who vere noe assaaulted during the one year sample period, the authors conclude that"... the data do aot support the premise that chorter officers hhave an overall greater probability of being assaulted than taller police persoanel." w Cheryl G, Swanson, charlé Assaults on Police,” Journ: No. 2, ppr 183-188. 20 Abilene, Texas; Anarillo, Texas; Austin, Texas; Bossier City, Louisiana; Galveston, Texas; Lake Charles, Louisiana; Lauton, Oklahoua; Sonroe, Lovisianas Norman, Oklahoma; North Little Bock, Arkansas; OkLchona City, Oklahoma; Pine BIUEf, Arkansas; Tulsa, Oklahoma. D, Hale, "A Question of Height Revisttes of Police se 4 Adninistration, Vol. 3,uz ‘Although data vere collected on such variables as training, education, tenure, and age of officers, the authors only report that these variables will be’ the cubject of subsequent reports. The conclusions would be significantly strengthened once the authors exanine these variables for possible indications of correlations between height and seniority or iseignnent, end report the recent history of height standards used in the jepartaent® studied. No statistical tests of sigatficance were utilized by Swanton and Hale when comparing assaulted and non-aesaulted officers. liovever, using the Kolnogorov-Snirnov test on the saaples of assaulted versus nonassaulted officers, analysis shows the height distributions of the two populations fare not statistically different even at the 0.1 level of significance, The article neglects to provide the reader vith any justification on whether the data fron 13 different cities can be aggregated. Are the height distributions similar across cities? Te the definition of “assault” common to all cities? Until a nore complete analysis of the data hecone available, 8 trong ct issue of height and FREQUENTLY CITED DATA Some sets of data frequently mentioned in various studies on height are presented here for easy reference, They consist of data fron four p.lice departnents, as follow © Metropolitan Police Department, D.C. 0-7, 6-8), “Assaults (Tables + tos Angeles Foltce Departaent--Injuries (Tables 6-9, C-10), # Soattte Police Depsrtnent~—Assaults (Teble C11), * San Francisco Police Departnent—Infurtes (Table 6-12). Since these data sets are not acconpanted by any indication of sentority or assignments, they cannot be used to reach ay definitive conclusions relating to height. The distribution of young adult men and vonen in the U.S. by hhetght according t0 a 1960-1962 survey 4 show in Table C-13,us ‘Table 0-7 ASSAULTED AND NON-ASSMULTED MALE POLICE EMPLOYEES OF DIFFERENT HEIGHTS IN THE WASHINGTON, D.C., METROPOLITAN POLICE conpuarive PERceyt DIFFERENCE 1 ueroHT Tawatted Won-Assaulted COmULATIVE Ciaches) Employees Eaployees PERCENTS 66 0.0 0.0 0.0 67 0.5 ae 23 8 19.9 12.2 nT 6 40.6 26.4 14.2 1 a2 a2 n ns 60.0 1.6 n 83.4 m6 #i 1. 93.2 ee 68 % 974 93.6 3.8 5 98.7 97h 13 16 99.5 99.3 02 n 100.0 99.8 02 8 100.0 100.0 0.0 SOURCE: Sergeant Mary Ellen Abrecht. NOTE: Data are from 1971. CONCLUSLON: “Shorter officers have © higher probability of being assaulted. N= 236, > No 4,436, ‘This is the greatest difference in the cumulative distributions of officers assaulted and of officers not assaulted. It 4s statistically significant {St the 0,0001 level of significance.us Table C-8 ASSAULTS ON POLICE IAPLOYEES OF DIFFERENT HEIGHTS TN WASHINGTON, D.C. CUMULATIVE PEROENT DIFFERENCE IN Assaulted sulted CUMULATIVE Employees Eaployees Employees PERUEATS. BELGHT (inches 66 2 ° 0.2 0.0 0.2 6 187 2 aan an 0.6 “ 496 7 14.8 19.1 6 706 5 29.4 70 ms a rey n 228 36 6.7 n 209 30 78.3 ay 430 23 87.5 o1.8 43 ™ a4 m9 94.0 97.3, 3.3 5 170 7 97.6 98.4 “0.8 6 86 2 99.4 99.2 0.2 ” 6 2 98.9 100.0 “on 8 5 ° 100.0 100.0, 0.0 SOURCE: Analysis by the authora of this report of data reported by Frank Verducci, p. 26 (see a review of the report on page 104 of this, report). Wore: ata are for 1971. CONCLUSION: “Shorter personnel have a higher at ult rates This ts the greatest difference ia the cunulative percent of all officers ‘and assaulted officers. The difference is statistically significant at the .001 level.us Table 0-9 INJURIES TO POLICE BAPLOYERS OF DIFFERENT HEIGHTS ‘TW THE LOS ANGELES POLICE DEPARTMENT wnee CUMULATIVE PERCENT DIFFERENCE 14 EIGHT Tafared Tajured — CUMULATIVE (Inches) Bmployees Eaployees _Eaployees Euployees PERCENT 68 320 163 8.5 16.3 6 538 184 BI aT ” 887 178 46.1 52.6 n my 216 65.3 74.2 n 350 96 79.9 33.8 o39 B 366 6 29.5 50.1 0.6 % 396 9 100.0 100.0 0.0 Total 3,784 998 SOURCE: Analysis by the authors of this roport of data reported by Prank Verducci, p. 35 (see a review of the report on page 104 of this report). Wore: Data are irom 1965. CONCLUSION: Shorter personnel have a Aigher injury rate, ‘this de the greatest difference in the cumulative percents of employees and of injured eaployees, The difference 1s statistically significant at the 0,001 level.116 Table C-10 TRIURIES TO MALE POLICE RAPLOYEES OF DIFFERENT HEIGHTS ‘DH THE LOS ANGELES POLICE DEPARTIENT comuarive peacest purvenmice 1 aeear aplovegs Employees cumativE Cinehes) Injured ot Insured Pencents os 10.5 ba 8.4 o 386 AL Ey 40.6 9.6 n na 59.6 a7 n aus 76.8 48 2 29.7 s8.3 Ls ” 9541 95.4 -0.3 6 97.2 90.2 -1.0 % 98.7 99.7 1.0 ” 100.0 99.9 ot ” , 100.0 0.0 SOURCE: Police Foundation, Washington, D.C. WOTE: Data are from 1960. CONCLUSION: “Shorter officers have a higher injury rate. = 539. b W = 2,930; however N could be 2,828, due to uncertainty in reading aunbers rnakiag up'the total, ‘This is the greatest difference in the cumulative percents of officers injured and of officers not injured. The difference is statistically significant at the 0.001 level.aELGaT (Inches) Eaployees Employees aployees Eaploys uy Table C-1L ASSAULTS ON POLICE EMPLOYEES OF DIFFERENT HEIGHTS 1 SEATTLE, WASHINGTON NUMER, UANATIVE PERCENT _ DIFFERENCE IN Teawulted ‘Assaulted CUMULATIVE ‘PERCENTS. ° %6 ob 23.8 0 mm 108 46.3 n 2 8 63.7 n us 78.3 49 n m8 ae 86.3 ous 1 oe oa 23.5 as 8 o 2 97.2 98.3 -1.0 1 2 a 9.2 99.0 0.2 1” 5 s 9.7 100.0 0.3 * ‘ ° 100.0 0.0 Tom 15879 SOURCE: Analysis by the authore of thie report of data reported by Frank Verdicel, pe 24 (aee a review of the Feport on page 104 of this ort: pate'are for 1971. CONCLUSION: “Shorter perooanel have a Mgher probability of being assaulted, ‘this de the greatest digference between the cumolative percents of employees and of asgaulted employ -001 Level. the 0. The difference ie statistically significant atus Table 0-12 TACIDENTS OF RESISTING ARREST AND OF BATTERY AGAINST POLICE MPLOYEES OF DIFFERENT HEIGHTS IN SAN FRANCTSCO mene DIFFERENCE EIGHT and Battery im CUMULATIVE, Inches) Maployees Incidents Employees Incidents! PERCENTS 67-68 8 29 3.0 0.9 68-69 243 0 10.1 65 6-70 268 38.1 28 10.3 70-71 367166 sn 43.2 Be an 35887 59 66.9 no 13 ne 87.8 TB. 9.4 116 16 7 93.9 88.6 3.3 1675 65 2 97.3 96.3 1.0 15-76 43 Foy 99.5 98.9 0.6 16-77 5 ° 99.8 99.9 02 18 ‘ 1 100.0 100.0 0.0 Total 1,917 951 SOURCE: Analysis by the authors of thie report of data reported by Frank Verducci, p. 22 (see a review of the report on page 104 of this report). NOTE: Data are from July 1, 1972 to August 30, 1972. CONCLUSION: “Shorter personnel have a lover probability of being involved in an incident of resisting arrest of battery (assault) Incidents include resisting arrest or battery on a police officer, » This 18 the greatest difference between the cumulative distribution of ceaployees and of enployees involved in incidents. The difference is statistically significant at the 0,001 level.us Table C=13 EIGHT DISTRIBUTIONS OF YOUNG ADULT MEN AKD WoxEN CUMDLATIVE PERCENT OF U.S. (1960-62) POPULATION, EIGHT AGE 18-24 YEARS, LESS THAN OR EQUAL TO GIVEN HEIGIT inches) Men omen, 60 wa or 02 23.9 62 us 40.1 63 37 513 66 rn 70.6 6 va an 66 28.6 919 67 ana 95.3 68 56.1 98.8 6 68.7 98.5 10 81.0 99.3 n 86.2 n 94.7 B ore % 99.2 6 99.8 SOURCE: U.S. Departuent of H.E.W., Public Health Services, National Center for Health Statistics, Weight by Height and Age of Sdulte, dotted states lo00ce), Sestes 11, Naber Topniay 1966, ‘Gore tecent data have not been published by HEW as'of January 1975.)[BIBLIOGRAPHY ‘Atlanta Regional Commission, Goverment Services Department, Technical ‘Assistance Division, "A study of the Police Officer Height Reguirenent," October 1973. Chapman, Senuel G., and Cheryl G, Swanson, "A Deseriptive Profile of the Assault Incident,” University of Oklahona, April 30, 1974, Cochran, T. Re, Arizona Departaent of Public Safety, Planning and Research, ‘The Minimum Height Requirenent for the Classification (undated). Cochran, Ws Gz, "Some Methods for Strengthening the Common Chi-Square “Tests,” Biometrics, Vol. 10, 1954. Denpsey, 0. A., "A Study of Police Height Requirements," Texas Department Of Public Safety, April 1974. Eisenberg, Terry, Deborah Ann Kent and Charles R, Wall, Police Personnel, ‘Practices in State and Local Governments, Intersational Ass- ‘tation of Chiefs of Police at al-, Washington, D.C., 1973. Bvansville, Indiana, Police Department, Personnel and Training Diviston, Planning and Research Section, “the Hvansville Police Departments Minimum Height Requirenentt A Bona Fide Occupational Qualification," Yovenber 6, 1973. Hoobles Raymond L., and J. A. MeQuesney, "A Question of Height, Ghiet, Noveaber 1973. The Poids Portland, Oregon, Bureau of Police, Planning and Research Divieton, "Analyste ‘of Assaulted and Non-Assaulted Officers," February 1973. Stegle, Sidney, Nonparasetric Statistics for the Behavioral Sciences, oe Southers Methodist University Lav School, Center for Police Development, “an Analysis of Physical and Bducational Requirements,” (undated, prepared for the Dallas Police Departuest). United States Department of H.E,jW., Public Health Services, National Center for Health Statistica, Weight by Weight and Age of Adults, SELES Stuets iogos19e2" serlee tie Wasber er tay eee United States Departaent of Justice, "Equal Rights Guidelines: Minimum Height Requiresente--Minorities and Vouen," 33 Federal Register 472 Caumber 46, March 9, 1973). Verducet, Frank M., "Height and Weight Requirements for Police Officers,’ (Gubmitted to the Civil Service Commission, City and County of San Francisco, 1974).