Computing AFQT Scores From Historical Data

4m ago
3.30 MB
81 Pages

AD-A263 893cRM 92-80/August 1992Computing AFQT ScoresFrom Historical DataPaul W. MayberryCatherine M. HiattDTICELECTEMAY07 1993E D50YearspA-C A1992ubfl4IIbCENTER FOR NAVAL ANALYSES4,91 FordiAvenue * Post Offict Box 1CZ268 * Alexa nra, Virginia22302-026893-09916o004

IForm ApprovedREPORT DOCUMENTATION PAGEOPMi No. 0704.0188ost bisbdr smka-te aoso k. ciesa fmoraos.addnfor "iast coleutfioczaon os cstrAWto Sesadeawosaro,Po icmpal bhedameddamuueti.n#stlo 5 ofafaaet .ts.collectioa opdsesses-f'ootmft olleum of eusormasdsmorst datasmPwds. ibmand lewlAMUMfar Icrctonss Operatow and Reports. 12115kifersot Dao i tumay.51.00 524. Aiiitpeni. VA U2220-32. wri toh bur&x. soWbmio qursrrs Ser-amt DurectoruzeWdaWofte 1c o ion &M R olasotvAftmonOffixe of Wrumn es snd IL4es. Wassiborw. DC 2%033. REPORT TYPE AND DATES COVERED2. REPORT DATE1. AGENCY USE ONLY (Leave Slads)IFinalAugust 19925. FUNDING NUMBERS4. TITLE AND SUBTITI.E*ComputingAFQT Scores FmcnIlistoric-Ai Data6. AUTo-OR(S)Paul W. Maybenry. Catherine M. MilC-N00014-91-C-002PE-65153MR0S. PERFORMING ORGANIZATIONREPORT NUMBER7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)CRM 92-80Center for Naval Analyses4401 Ford AvenueAlexandria. Virginia 22302-026810. SPONSORINGIONITORING AGENCYREPORT NUM1BER9. SPONSORINGIMIONITORING AGENCY NAME(S) AND ADDRE:SS(rES)Commanding GeneralMarine Corps Combat Development Command (WF 13F)Studies and Analyses BranchQuansico, Virginia 22134It. SUPPLEMENTARY NOTESM2.DISTRIBUTONIAVAILAILITY STATEMENT12b. DISTRIBUTION CODEApproved for Public Release; Distnibution Unlimited13. ABSTRACT (Maximum 200 words)The Armed Forces Qualificai ion Test (AIFQT) serves as an initial screen for military selection. AFQT scores must accurately reflect the quality of recruits whoenter the services and must mnaintain the same -neaning and interpretation over time. Unfortunately, several changes in the AFQT and other problems makesuch histonical computations and ftcore interpretations difficult. This research memorandum details these past problems and presents solutions that willallow for she proper computation of curent AFQT scores from historical databases.ij SUBJECT TERMSArQT (Armed for-cs qualific,3tion test), Aptitude tests, AS VABl (Armed scrvicei vocational aptitude battery). Computerprograms, Quality, Problems, Scoring, ITables (data), lest score%17. SECURITY CIA LiFinCAYF4CF'IS SL.Ct'RITY CLASSIFICATION1O Ti IIlls PAGEI19SECLURIT*YCLASSIFICATION01IBTAT NU\IJllR 01: PAGESi16 PRICE COD)EI20 LIMITATION 'M AIJSTRAI"rPrrsar.hed by ANSI SO 239 11N9901

5 OYears5'A1992CENTER FOR NAVAL ANALYSES4401 Ford Avenue 9 Post Office Box 16268 0 Alexandria, Virginia 22302-0268 * (703) 824-20005 October 1992SMEMORANDUM FOR DISTRIBUTION LISTSubj:CNA Research Memorandum 92-80Encl:(1) CNA Research Memorandum 92-80, Computing AFQT Scores FromHistorical Data, by Paul W. Mayberry and Catherine M. Hiatt,Aug 19921. Enclosure (1) is forwarded as a matter of possible interest.2. The Armed Forces Qualification Test (AFQT) serves as an initialscreen for military selection. AFQT scores must accurately reflect thequality of recruits who enter the services and must maintain the samemeaning and interpretation over time. Unfortunately, several changes inthe AFQT and other problems make such historical computations and scoreinterpretations difficult. This research memorandum details these pastproblems and presents solutions that will allow for the proper computation of current AFQT scores from historical databases.Donald J. bykrotDirectorManpower and Training ProgramLAcceonl ForDistribution pageDTIG TABUnannouncedCRA&IUJUiStiticatiofl"'By .Dist, ibution IAvailability CodesDistAI, 'Avail and Ior

Subj:Center for Naval Analyses Research Memorandumn 92-80Distribution ListSNDLAlAlHA2AASASA5ASA6AftA6DASN - MANPOWER (2copies)ASSTSECNAV MRACNRCHNAVPERSPERS-2PERS-SPERS-11CG MCRDAC - WashingtonHQMC AVNHQMC MPR &RAAttn: Code MRAttn: Code UPAttn: Code MMAttn: Code MA (3copies)Attn: Code MPP-54HQMC PP&OA6FF?8 USNAAttn: Nimitz LibraryFF42 NAVPGSCOLFF44 NAVWARCOLFJA1 COMNAVMIL.PERSCOMFJAl13 NAVPERSRANDCENAttn: Technical Dirctor (Code 01)Attn: Technical LibraryAttn: Dir., Manpower Systems (Code 11)Attn: Dir., Personnel Systems (Code 12)FJB 1 COMNAVCRUITCOMCNETFrlCG MAGTECV12CG MCCDCV12Attn: Studies and Analyses BranchAttn: Director, Warfighting CenterAttn: Warfighting Center, MAGTF Propenency and Requiremewnts B nnh (2zopie,;)CG MCRDAC - QuanticoV12OTHERJoint Service Job Performance Measurement Working Group (12 copies)Military Accession Policy Working Group (17 copies)

CRM 92-80 / August 1992Computing AFQT ScoresFrom Historical DataPaul W. MayberryCatherine M. HiattOperationsand Support DivisionI50)Years*0191992CENTER FOR NAVAL ANALYSES4401 Ford Avenue * Post Cffie Box 16268 9 Akwmdra, Virginia 22302-0268

ABSTRACTThe Armed Forces Qualification Test(AFQT) serves as an initial screen formilitary selection. AFQT scores must accurately reflect the quality of recruits whoenter the services and must maintain thesam6 meaning and interpretation over time.Unfortunately, several changes in the AFQTand other problems make such historicalcomputations and score interpretations difficult. This research memorandum detailsthese past problems and presents solutionsthat will allow for the proper computationof current AFQT scores fro. agtoricaldatabases.-iii-

EXECUTIVE SUMARYThe Armed Services Vocational Aptitude Battery (ASVAB) is theselection test administered to military applicants to determine theireligibility for service and to classify individuals into jobs they arelikely to perform successfully. The Armed Forces Qualification Test(AFQT) is a part of the ASVAB and is the initial screen for militaryselection. Applicants must exceed certain minimum AFQT standards to beconsidered for enlistment. This test is also used as the primaryindicator of recruit quality, assisting military manpower managers inmonitoring the enlisted force and establishing recruiting goals forfuture accession cohorts.Military selection and classification testing has an extensive history dating back to World War I. The AFQT was initially developed andimplemented in 1950 as a joint-service selection instrument. During theearly 1970s, the services departed from using a common AFQT. When a newtesting program for aLl military applicants began in 1976, all servicesagain began using a common AFQT.Over time the AFQT experienced many changes and problems: addition/deletion of ASVAB subtests, changes in the ASVAB score scale, changes insubtest score metrics used to compute the AFQT, and changes in AFQT subtests. These changes are detailed in table I. Despite efforts to maintain the consistency of AFQT scores, such dramatic changes and problemshave negatively affected the historical consistency of AFQT scores. Thescores recorded in historical databases tend to be incompatible. Therefore, AFQT scores of record terd not to be directly comparable over time,should not be aggregated, and may result in inconsistent researchfindings.In order to obtain an accurate AFQT score that is comparable acrosstime and ASVAB forms, it is necessary to use a common set of subtests todefine the AFQT and to compute all scorIs on the same score scale. Thecurrent AFQT definition (2VE AR MK) on the 1980 score scale is thecommon base to which all other AFQT form-score scale combinations shouldbe equated.The only accurate means of obtaining AFQT scores equivalent to thecurrent definition and score scale is to work with subtest raw scores(not standard scores) and recompute subtest standard scores and the correct AFQT percentile score. One cannot adjust the AFQT scores of recordso that they reflect the current definition and score scale. Computerprograms and detailed instructions for computing the AFQT are presentedin this paper.1. VE is the verbal composite, which consists of the sum of the WordKnowledge and Paragraph Comprehension subtests. AR and MK are theArithmetic Reasoning and Math Knowledge subtests, respectively.-o

1.40003.410to4)"44101.4 000U4)140001.40000 4co400 0H c.0C(- 4U(U(r4 1440 00to4)4)co4-Ir4(0U 4(U1.40UDc:V32U)0cc.,.I4V4r.(UC)CA44)4).140 U i0))1404)04 0) N4N "H 4) 4) 0H4 C %U (Ut0tWC414.) Cd4JW0 04) V) 04-'0 2 (44(1)1-.4)U4WH40.4)ul9z)C4)4)1.40%0%%00"H0%0%-A* H 0H04.)441)CUbO' 04)4)14A 04.4 U4U0 P0 0t44jCO-r4 14000r44' 4)U202-4f0ANtNl1VHr4.01%00%NE-44c(AU)H-'-vi-1 .4%NrNHNH-1-0

CONTENTS*PageTables. ixIntroduction.1Background .Content of ASVAB .Equating of ASVAB Forms.ASVAB Score Scale .Subtest and AFQT Score Metrics .Changes in AFQT Definition and Computation .122567Problem. .9Solution . 910Computing Current AFQT From Historical Subtest Raw Scores .ASVAB Forms 5/6/7. 11All Other ASVAB Forms . 14Summary. 14Appendix A: Conversion Tables Required for ComputingCurrent AFQT from Historical Subtest Raw Scores . A-1-A-22Appendix B: Computer Programs To Compute Current AFQTFrom Historical Subtest Raw Scores. B-1-B-27Appendix C: Documentation of CNA-Generated ConversionTables To Compute Current AFQT from Historical SubtestRaw Scores . C-1-vii--C-4

TABLES"*Page1Subtest Composition of ASVAB by Test Form .32Dates of Administration for Each ASVAB Form .43Definition of AFQT Categories .74Historical Changes in AFQT Definition, Score Scale, andSubtest Metric .8Computing Current AFQT from Historical SubtestRaw Scores .125-ix-

INTRODUCTIONThe Armed Services Vocational Aptitude Battery (ASVAB) is theselection test administered to military applicants to determine theireligibility for service. Scores from the ASVAB are also used to classify individuals into jobs they are likely to perform successfully. TheASVAB is administered each year to about 1 million people who apply forenlistment and to about another million students in high schools andpost-secondary institutions who are making career decisions and explor!.ng vocational possibilities.The Armed Forces Qualification Test (AFQT) is a part of the ASVABand is the initial screen for military selection. Applicants mustexceed certain minimum AFQT standards to be considered for enlistment.The test is also used as the primary indicator of recruit quality,assisting military manpower managers in monitoring the enlisted forceand establishing recruiting goals for future accession cohorts.This research memorandum documents the history of the AFQT andnotes events that have adversely affected its interpretation and computation. Detailed information and computer programs are provided so thatconsistent AFQT scores can be computed from historical databases thatreflect the current AFQT subtest definition and score scale.BACKGROUNDMilitary selection auid classification testing has an extensivehistory, dating back to Uorld War I. However, the first joint-servicerequirements for testing were not mandated until 1948 with the passageof the Selective Service Act. Subsequently, the AFQT was developed and,in 1950, was implemented as the joint-service selection instrument todetermine the eligibility of military draftees and volunteers. Thoughthe services used a common AFQT for selection decisions, each servicedeveloped and administered its own test battery for job classification.In the 1960s, the Department of Defense (DOD) developed a commonselection and classification test (the ASVAB) for use in the nation'shigh schools. The intent was to provide the high schools with a viableand informative testing program while supplying the services with information that would be helpful for recruiting high school students. TheASVAB high school testing program began in 1968 with the introduction ofASVAB form 1.Between 1973 and 1975, the services departed from using a commonAFQT. Instead, each service was allowed to develop its own proceduresfor computing AFQT scores from its respective battery of aptitude tests.In 1974, efforts were begun to consolidate the testing programs acrossthe services to eliminate duplicate testing for individuals applying tomultiple services, to facilitate interservice referrals of applicants,and to allow researchers to focus their attention on developing a single

accession test. Based on the success of ASVAB testing in the highschools, a joint-service program, called the production testing program,was developed to replicate the high school approach of a common testbattery to address the testing requirements for all military applicants.The new production testing program introduced two ASVAB forms in1976 (forms 6 and 7). Based on these new ASVAB forms, all servicesagain used a common set of tests to define the AFQT. The production andthe high school testing programs have continued to produce revised formsof the ASVAB and to conduct research to monitor the quality of the testsand the stability of score interpretation over time.Content of ASVABThe ASVAB is composed of a number of subtests that measure specificaptitudes. The content of the ASVAB has evolved over time to reflectadvances in the measurement of mental aptitudes and the findings fromextensive validation research. Table 1 documents the subtest composition of all ASVAB forms. The most extensive revisions in contentoccurred wi