[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-seo-529710-105":3,"detail-sidebar-cat-0-en-105":81,"doc-detail-529710-en":130},{"code":4,"msg":5,"data":6},0,"ok",{"site_id":7,"language":8,"slug":9,"title":10,"keywords":11,"description":12,"schema_data":13,"social_meta":74,"head_meta":76,"extra_data":78,"updated_unix":80},105,"en","some-measurement-characteristics-of-aggregated-versus-individual-scores","Some Measurement Characteristics of Aggregated Versus Individual Scores","","Not infrequently, investigators claim that reliability for groups exceeds reliability for individuals and/or that error variance for groups is smaller than for persons. Generalizability theory shows that this conventional wisdom is not necessarily true. The discussion draws on examples from course evaluation research and performance testing literature, identifying specific conditions under which the usual comparisons break down and reliability patterns can contradict expectations.",{"@graph":14,"@context":73},[15,34,56],{"@type":16,"itemListElement":17},"BreadcrumbList",[18,23,27,31],{"item":19,"name":20,"@type":21,"position":22},"https://docshare.wps.com","Home","ListItem",1,{"item":24,"name":25,"@type":21,"position":26},"https://docshare.wps.com/document/","Document",2,{"item":28,"name":29,"@type":21,"position":30},"https://docshare.wps.com/document/research-report/","Research & Report",3,{"item":32,"name":10,"@type":21,"position":33},"https://docshare.wps.com/document/some-measurement-characteristics-of-aggregated-versus-individual-scores/529710/",4,{"url":32,"name":10,"@type":35,"image":36,"author":41,"headline":10,"publisher":44,"fileFormat":47,"inLanguage":8,"description":12,"dateModified":48,"datePublished":49,"encodingFormat":47,"isAccessibleForFree":50,"interactionStatistic":51},"DigitalDocument",{"url":37,"@type":38,"width":39,"height":40},"https://docshare.wps.com/thumbnails/some-measurement-characteristics-of-aggregated-versus-individual-scores/529710.png","ImageObject",300,407,{"name":42,"@type":43},"jay","Person",{"url":19,"name":45,"@type":46},"DocShare","Organization","application/pdf","2026-11-10","2026-10-09",true,{"@type":52,"interactionType":53,"userInteractionCount":55},"InteractionCounter",{"@type":54},"ViewAction",7,{"@type":57,"mainEntity":58},"FAQPage",[59,65,69],{"name":60,"@type":61,"acceptedAnswer":62},"Why is the conventional claim about higher group reliability sometimes wrong?","Question",{"text":63,"@type":64},"The paper uses generalizability theory to show that reliability for groups is not always larger than reliability for individuals, and error variance relationships can reverse under specific conditions.","Answer",{"name":66,"@type":61,"acceptedAnswer":67},"What measurement design is denoted by (p:g)×i?",{"text":68,"@type":64},"The notation describes persons nested within groups and crossed with items, producing an interaction term confounded with sources of error that define variance components.",{"name":70,"@type":61,"acceptedAnswer":71},"How do the generalizability coefficients differ when persons are objects of measurement within groups versus across groups?",{"text":72,"@type":64},"Equation 3 focuses on persons within a single randomly selected group, while Equation 4 generalizes across all persons across groups; both involve the items facet as part of the generalization universe.","https://schema.org",{"og:url":32,"og:type":75,"og:title":10,"og:site_name":45,"og:description":12},"article",{"robots":77,"canonical":32},"index,follow",{"doc_id":79,"site_id":7},529710,1791715792,{"code":4,"msg":82,"data":83},"success",[84,88,92,96,101,106,110,114,119,122,126],{"id":22,"doc_module":4,"doc_module_name":25,"category_name":85,"show_sort_weight":86,"slug":87},"Story & Novel",90,"story-novel",{"id":26,"doc_module":4,"doc_module_name":25,"category_name":89,"show_sort_weight":90,"slug":91},"Literature",80,"literature",{"id":33,"doc_module":4,"doc_module_name":25,"category_name":93,"show_sort_weight":94,"slug":95},"Exam",70,"exam",{"id":97,"doc_module":4,"doc_module_name":25,"category_name":98,"show_sort_weight":99,"slug":100},5,"Comic",60,"comic",{"id":102,"doc_module":4,"doc_module_name":25,"category_name":103,"show_sort_weight":104,"slug":105},6,"Technology",50,"technology",{"id":55,"doc_module":4,"doc_module_name":25,"category_name":107,"show_sort_weight":108,"slug":109},"Healthcare",40,"healthcare",{"id":111,"doc_module":4,"doc_module_name":25,"category_name":29,"show_sort_weight":112,"slug":113},8,30,"research-report",{"id":115,"doc_module":4,"doc_module_name":25,"category_name":116,"show_sort_weight":117,"slug":118},9,"Religion & Spirituality",20,"religion-spirituality",{"id":117,"doc_module":4,"doc_module_name":25,"category_name":120,"show_sort_weight":117,"slug":121},"World Cup","world-cup",{"id":123,"doc_module":4,"doc_module_name":25,"category_name":124,"show_sort_weight":123,"slug":125},10,"Lifestyle","lifestyle",{"id":127,"doc_module":4,"doc_module_name":25,"category_name":128,"show_sort_weight":97,"slug":129},19,"General","general",{"code":4,"msg":82,"data":131},{"doc_id":79,"user_id":132,"nickname":42,"user_avatar":133,"doc_module":4,"category_id":111,"category_name":29,"doc_title":10,"doc_description":12,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":55,"is_deleted":4,"is_public":22,"is_downloadable":22,"audit_status":22,"page_count":139,"language":140,"language_code":8,"site_id":7,"html_lang":8,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":12,"update_tm":144,"read_time":145,"preview_page_count":33,"preview_urls":146},3985747858866,"https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd","# Some Measurement\n\nCharacteristics of AggregatedVersus Individual Scores  \nRobert L.Brennan  \nDecember 1993  \nFor additional copies write:ACT Research Report SeriesP.O.Box 168lowa City,lowa 52243  \n# Some Measurement Characteristics ofAggregated Versus Individual Scores\n\nRobert L.Brennan  \nAmerican College Testing  \n2  \n## Abstract\n\nNot infrequently,investigators assume that reliability for groups is greater thanreliability for persons,and/or error variance for groups is less than error variance forpersons.Using generalizability theory,it is shown that this \"conventional wisdom\"is notnecessarily true.Examples are provided from the course evaluation literature and theperformance testing literature.  \n3  \n# Some Measurement Characteristics ofAggregated Versus Individual Scores\n\nIt is often stated that if a test is not reliable enough for making decisions aboutindividuals,or if error variance for individuals is unacceptably large,then the test shouldbe used only for making decisions about groups.Implicit in such statements is anassumption that reliability for groups is necessarily larger than reliability for persons,anderror variance for groups is necessarily smaller than error variance for persons.In thispaper,such statements or assumptions will be called the \"conventional wisdom.\"Thepurpose of this paper is to show that this conventional wisdom is not necessarily true,andto identify specific conditions that lead to contradictions of this conventional wisdom.  \nThese issues are considered in the context of generalizability theory(Cronbach,Gleser,Nanda,&Rajaratnam,1972),without a full explication of all the details of thetheory.Readers unfamiliar with generalizability theory can consult Cronbach et al.(1972),Brennan(1992a),or Shavelson and Webb(1991).Also,many aspects ofreliability(or generalizability)of group means have been treated by Kane and Brennan(1977).A brief introduction to generalizability theory is provided by Bennan(1992b).  \nGeneralizability Coefficients for Groups  \n   \nwith Two Random Facets  \n   \nWhen persons(p)are nested within groups(g)and crossed with items(i),thedesign is denoted (p:g)xi,and the linear model is  \n(1)  \nThe terms to the right of the equality(except μ)are uncorrelated score effects withexpectations of zero,and the p:g~term is the interaction effect confounded with othersources of error.The variances of these score effects are called variance components.  \n4  \nFor this design,if groups are the objects of measurement,then the universe ofgeneralization consists of the p and i facets.If,in addition,p and i are both random,thenthe generalizability coefficient for generalizing over samples of k items and n personswithin each group is  \n(2)  \nwhere  \nkand  \nIt is important to note that n in Equation 2 is the number of persons within a group,notthe larger number of persons across all groups.  \n   \nAssuming that p and i are both random implies that replications of themeasurement procedure would involve different sets of persons and items or,  \nequivalently,that an investigator wants to generalize to a larger set of persons and itcmsthan those in a particular measurement procedure.This assumption seems especiallysensible for programs such as NAEP,which uses a type of matrix sampling design.  \nFrom the perspective of generalizability theory,traditional measurement error ismost closely associated with generalizing over samples of items.It is important to note,however,that traditional measurement error is not necessarily the only,or even the mostimportant,source of unreliability for inferences about group means.As noted by Feldtand Brennan(1989):  \nThe test results for any given year reflect not only the character of theinstrictional program but also the character of students enrolled at thatspecific moment.These individuals must be regarded as a sample,in alongitudinal sense,from the population that flows through the districtschools over a period of years.A curricular judgment can be in error if aparticula","cbCait1nybtKsza1","https://ap.wps.com/l/cbCait1nybtKsza1","pdf",477011,28,"English","# Abstract\n## Reliability assumptions and the “conventional wisdom”\n## Generalizability coefficients for groups with two random facets\n## Reliability for persons within groups vs across groups\n## Implications for test length and decision making","[{\"question\":\"Why is the conventional claim about higher group reliability sometimes wrong?\",\"answer\":\"The paper uses generalizability theory to show that reliability for groups is not always larger than reliability for individuals, and error variance relationships can reverse under specific conditions.\"},{\"question\":\"What measurement design is denoted by (p:g)×i?\",\"answer\":\"The notation describes persons nested within groups and crossed with items, producing an interaction term confounded with sources of error that define variance components.\"},{\"question\":\"How do the generalizability coefficients differ when persons are objects of measurement within groups versus across groups?\",\"answer\":\"Equation 3 focuses on persons within a single randomly selected group, while Equation 4 generalizes across all persons across groups; both involve the items facet as part of the generalization universe.\"}]","Some Measurement Characteristics of Aggregated Versus Individual Scores | PDF",1791521791,71,[147,151,155,159],{"page_no":22,"url":148,"width":149,"height":150},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/96752f635064a2a6fc033b8930e43498.webp",983,1255,{"page_no":26,"url":152,"width":153,"height":154},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/41b7a0abcca4f36b83c3fdfb8c18aed4.webp",982,1259,{"page_no":30,"url":156,"width":157,"height":158},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/0e3be8327d45c756603e0fc94049a0a8.webp",981,1283,{"page_no":33,"url":160,"width":161,"height":162},"https://res-academy.cache.wpscdn.com/upload/kdocshare/2026-11-17/52e102293c6be36f5ca8e55d7fd7bd53.webp",980,1302]