Finite state test theory models the response behavior of an examinee and establishes the relationship between the ability of the examinee and the observed responses on a multiple-choice test. In finite state modeling, various assumptions about item characteristics and the examinees’ response strategies are made to estimate an examinee ability, and willingness to guess. Twelve sets of plausible assumptions about identifiability of distractors and examinee guessing strategies were adopted and the corresponding finite state models were actualized. Three consequences of the adoption of the 12 sets of assumptions were investigated: 1) the extent to which the resulting ability estimates rank ordered the examinees similarly, 2) variation in the magnitude of ability estimates and the estimated willingness to guess across the 12 models, and 3) the extent to which conclusions about examinees subgroups would differ according to the model employed. Also, conventional number-right scores were compared with the finite state scores with respect to the three outcomes just listed. All scoring methods rank ordered the examinees essentially the same. The magnitude of the finite state scores varied considerably across models mainly due to differing assumptions about the identifiability of distractors. Differing assumptions about examinee guessing strategy had surprisingly little effect on the magnitude of the ability estimates, though estimates of willingness to guess varied consistently according to the assumed strategies. Conclusions about group differences also varied across the models as a result of differing assumptions about both item characteristics and examinee guessing strategies. / Ph. D.

Several alternatives to the Kuder and Richardson formula number 20 (KR20) were compared for accuracy using simulated and actual data sets. Coefficients by Loevinger (1948), Horst (1954), Raju (1982), and Cliff (1984) as well as the Kuder and Richardson formulae numbers 8 and 14 were examined. These alternative reliability coefficients were compared by (1) simulation of tests with varying degrees of item difficulty dispersion, subject proficiency, reliability, and length, and (2) use of the norming samples of the Curriculum Referenced Tests of Mastery (Charles E. Merrill Publishing Co., publisher) for grades four, six, and eight. Most of the coefficients examined proved no more accurate than the KR20 and several were decidedly worse. All coefficients, with the exception of Loevinger's, were affected by item difficulty dispersion. Two coefficients, the KR8 and Horst, were found to have potential as KR20 substitutes. These two coefficients are discussed with recommendations made as to the appropriate use of each one.

M.Ed. (Educational Psychology) / Please refer to full text to view abstract.

M.A.(Psychology) / Please refer to full text to view abstract

M.Cur. / Although clinical evaluation and examination is regarded as the basis of scientifically founded nursing care, it is inherently a subjective process. This process is complicated by the variability of four inter-related systems, namely the eyaluator, the clinical environment, the student nurse and other individuals. The purpose of the study was the analysis of current evaluation practices and the OSCE and to determine the perception of clinical educational personnel and student nurses in this regard. Consequently a descriptive exploratory survey was done by means of a questionnaire. For the purpose of this study, clinical teaching. Personnel and student nurses associated with the nine affiliated hospitals of one nursing college in the Transvaal were approached. From the obtained data it appears that clinical evaluation and examination remains, to a certain degree, a subjective process. Preconceptions probably influence human judgement. In spite of the occurrence of identified variables the OSCE appears to be a more acceptable approach to clinical examination than traditional long examination methods.

Thesis (PhD)--Stellenbosch Universityh, 1995. / ENGLISH ABSTRACT: This study investigates the influence of readability of mathematics examination questions on achievement. The aim of any mathematics examination is to assess whether the aims of a specific mathematics programme have been realized. Readability factors that unnecessarily prevent a clear understanding of questions could jeopardize this aim. The important issue is, therefore, whether there are indeed readability factors in mathematics examination questions that cause comprehension problems for students and, if there are, do they hav~ any effect on test scores? The issue of readability is of even greater importance for second language readers. In the South Mrican context, the reading problems of second language readers are of particular importance as most students at school are second language learners. An important question would therefore be: What readability factors cause comprehension difficulties for second language students, especially those whose mother tongue is not kindred to English? Furthermore, what is the influence of cultural factors on readability? This study provides answers to these and other related questions for mathematics text at senior secondary school level. Protocol analysis was used to ascertain what readability problems are experienced by students when reading examination questions in mathematics. Three different language groups, comprising 17 -18-year-old students, were used in the study: English First Language students and two groups who had English as a second language. One second language group had Mrikaans as first language while the other group comprised Mrican students whose mother tongue is unrelated to English. A framework was developed to analyse the protocols and it comprised five categories: unfamiliar vocabulary structural problems obscure information visualization difficulties non-verbal factors Mter the protocol study, students were asked to adapt the examination questions to a more comprehensible form. Students' adaptations addressed lexical, syntactical, discourse and non-verbal factors. Most of the readability problems identified in the literature study were verified in the empirical study. However, the empirical study generated additional readability problems that are mainly restricted to mathematics text and relate to nonverbal factors like mathematical expressions. During the last phase of the empirical study a composite test was used to test the hypothesis that improved readability of the common language used in mathematics examination questions' will improve achievement. Nine socalled "word problems" from previous examination papers were set in three different versions: original, adapted and non-verbal. The hypothesis was confirmed in a number of important cases. A significant finding of the study was, therefore, that readability factors not only influence the comprehension of mathematics examination questions, but also have a marked influence on students' achievement levels. The results of the empirical study are reported quantitatively as well as qualitatively. Other results include the following: Not only second language students, but also first language students experienced a variety of readability problems. All three language groups demonstrated the same level of competency on the non-verbal versions. When comparing test scores of the verbal versions, differences in achievement levels between the different language groups were often caused by linguistic and cultural factors. Cultural thought patterns, typical of a mother tongue but absent in a second language, were often a source of comprehension difficulties for second language readers. This study has led to certain conclusions for teaching and examination practice. For example, factors influencing the readability of ordinary English should be considered with other factors when writing mathematics examination questions. Furthermore, the distinctly different reading needs of second language students suggest that examination papers be set, so that the language needs of second language learners are accommodated. Guidelines for writing more readable examination questions were developed and are presented as a readability checklist. Suggestions for further research include the investigation of the influence of readability on achievement in authentic examination conditions. / AFRIKAANSE OPSOMMING: Hierdie studie ondersoek die invloed van leesbaarheid van wiskunde eksamenvrae op prestasie. Die doel van enige wiskunde eksamen is om vas te stel of die doelwitte van 'n spesifieke wiskunde program bereik is. Leesbaarheidsfaktore wat die volledige begryp van vraestelle onnodig belemmer, kan die bereiking van hierdie doel verhinder. Dit is dus belangrik om vas te stel of daar wel leesbaarheidsfaktore in wiskunde eksamenvrae bestaan wat vir leerlinge begripsprobleme veroorsaak en, indien wel, of hulle enige effek op toetspunte het. Vir tweedetaallesers is die kwessie van leesbaarheid van nog groter belang. In die SuidAfrikaanse opset is die leesprobleme van tweedetaal lesers 'n uiters aktuele saak aangesien die meeste skoolleerlinge tweedetaal leerders is. Belangrike vrae is dus: Watter faktore veroorsaak leesbaarheidsprobleme vir tweedetaalleerlinge, veral diegene met 'n nie-verwante moedertaal en, watter invloed het kultuur op leesbaarheid? Hierdie ondersoek bied antwoorde op hierdie en ander verwante vrae ten opsigte van wiskunde eksamenvrae op senior sekondere vlak. Protokol analise is gebruik om vas te stel watter leesbaarheidsprobleme ondervind word wanneer leerlinge wiskunde eksamenvrae lees. Drie verskillende taalgroepe, bestaande uit 17 -18-jarige leerlinge, het aan die ondersoek deelgeneem: Engels Eerstetaalleerlinge en twee groepe wat Engels as tweede taal gehad het. Een van die tweedetaal groepe het Afrikaans as eerste taal gehad terwyl die ander groep bestaan het uit Afrikane wie se moedertaal nie aan Engels verwant is nie. 'n Raamwerk is ontwikkel om die protokolle te analiseer en het uit die volgende vyf kategoriee bestaan: onbekende woordeskat strukturele probleme onduidelike inligting visualiseringsprobleme nie-verbale probleme Gedurende die tweede fase van die ondersoek is leerlinge gevra om die vrae tot 'n meer verstaanbare vorm aan te pas. Leerlinge se aanpassings het leksikale, sintaktiese, diskoers en nie-verbale faktore aangespreek. Sommige leesbaarheidsprobleme wat in die literatuurstudie gei'dentifiseer is, is in die empiriese ondersoek geverifieer. Die empiriese ondersoek het egter addisionele leesbaarheidsprobleme uitgelig wat meerendeels wiskundig van aard is en verband hou met nie-verbale faktore soos wiskunde uitdrukkings. Gedurende die laaste deel van die empiriese ondersoek is 'n samegestelde toets gebruik om die volgende hipotese te toets: Verbeterde leesbaarheid van die gewone taal wat in wiskunde eksamenvrae gebruik word, sal die prestasie van leerlinge verb et er. N ege sogenaamde woordsomme is op verskillende maniere gestel: oorspronklik, aangepas en nie-verbaal. Die hipotese is in 'n hele aantal belangrike gevalle bevestig. Een van die bevindinge van die ondersoek was dus dat leesbaarheidsfaktore nie slegs be grip ten opsigte van wiskunde eksamenvrae bei:nvloed nie, maar ook 'n beduidende invloed op prestasie het. Die resultate van die empiriese ondersoek word kwalitatief sowel as kwantitatief weergegee. Ander resultate sluit die volgende in: Nie slegs tweedetaal leerlinge nie, maar ook eerstetaal leerlinge het 'n verskeidenhied van leesbaarheidsprobleme ondervind Al drie taalgroepe het op die nie-verbale weergawe dieselfde bekwaamheidsvlak getoon. Verskille in die prestasievlakke tussen die verskillende taalgroepe op die verbale weergawes is baiekeer deur taalkundige en kulturele faktore veroorsaak Kulturele denkpatrone wat tipies is van 'n leerling se moedertaal, maar nie in die tweedetaal voorkom nie, het dikwels tot begripsprobleme by tweedetaal lesers gelei. Hierdie ondersoek het sekere gevolgtrekkings v1r o~derrig- en eksamenpraktyk. Byvoorbeeld, faktore wat die leesbaarheid van gewone Engels bei:nvloed, behoort saam met ander faktore in ag geneem te word wanneer wiskunde eksamenvraestelle opgestel word. Verder is daar duidelike aanduidings dat aparte vraestelle vir eerste- en tweedetaal leerlinge opgestel moet word sodat die taalbehoeftes van tweedetaal leerlinge in ag geneem kan word. Riglyne vir die skryf van meer leesbare eksamenvrae is saamgestel en as 'n oorsiglys aangebied. Voorstelle vir verdere navorsing sluit in die ondersoek na die invloed van leesbaarheid op prestasie in ware eksamen omstandighede.

D.Ed. (Education) / In this research the core syllabuses, departmental syllabuses and the matriculation examination papers were evaluated. For the purpose of evaluation certain criteria were developed based on inherent curriculum design principles, including a situation-analysis. Cognizance was taken of the data obtained in an earlier investigation by the Human Sciences Research Council on the needs.and demands of the work situation. Criteria were developed in the areas of Education, Linguistics and Language Didactics. Linguistic criteria are based on a Pragmatic language perspective, including a Pragmatic component as part of language structure. In order to obtain a clearer criterial focus, the different structure oriented, content based and educational criteria were integrated in a criterial structure. In the application of the criterial structure to the abovementioned curriculum documents, it was found that the language syllabuses do not adequately equip the learner to become adult educated users of the language. The learner is being inadequately equipped for the demands of the various social contexts in which he or she is required to use the language. The essential nature of language according to the dynamic interaction between language use and structure within a given context is negated. A theoretical basis is lacking resulting in the negation of the essentially communicative nature of language and the role of context in the functional use of language. Guidelines, based on the stated criteria, have been put forward with a view to accountable curriculum and syllabus innovation.

M.Ed. / The grade 12 Physical Sciences students of 2008 were the first group of South African students to write a National Senior Certificate (NSC) on the new outcomes-based education (OBE) curriculum – the National Curriculum Statement (NCS). Society scrutinised the performance of students in this high stake examination. The outcome was disappointing: 71,3% of the students achieved a mark of less than 40%, and 45% of the group achieved a mark of less than 30%. Concern amongst the educational community, specifically the Department of Education (DOE), initiated a request for research into the possible causes of the poor performance by students in this examination. There are many factors that affect the performance of students, including the misconceptions held by students regarding subject content. This study aims to contribute knowledge about the common misconceptions held by science students regarding Physics. It also investigates the performance of students in explanation-type questions and what explanation-types reveal about student misconceptions. The research design for this study is a content analysis which was carried out qualitatively in two phases. In the primary phase, a sample of student examination scripts was analysed. During the secondary phase, interviews were conducted with grade 12 Physical Sciences students and teachers from one school. The findings of this study are that the following misconceptions are commonly held by students: • Heavier objects exert more force on lighter objects during a collision; • Total external resistance decreases when an external resistor, connected in parallel, is removed; • Energy is lost in certain situations; • A split-ring is found in an AC generator; • The voltage increases when appliances are added to a multi-plug.

This study had four purposes. They were to determine whether accounting tests used during the 1989 Virginia FBLA regional accounting competitive events were reliable, valid, comprised of test items at appropriate difficulty levels, and to ascertain if the curricular drift was present. Secondary-level high school business students, all of whom were members of Virginia FBLA, participated in this study. Nine of the 11 Virginia FBLA regions were represented in this study. Test scores for 152 students were obtained for Accounting I and 81 students for Accounting II. The researcher examined the tests for reliability by applying Kuder-Richardson Formula 20 (KR-20). Both the Accounting I and Accounting II tests were found to have low reliability estimates, .81 and .78 respectively. / Master of Science

杨志明. / 呈交日期: 2003年2月. / 论文(哲学博士)--香港中文大学, 2002. / 参考文献 (p. 115-120). / 中英文摘要. / Cheng jiao ri qi: 2003 nian 2 yue. / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. Ann Arbor, MI : ProQuest Information and Learning Company, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Mode of access: World Wide Web. / Yang Zhiming. / Zhong Ying wen zhai yao. / Lun wen (zhe xue bo shi)--Xianggang Zhong wen da xue, 2002. / Can kao wen xian (p. 115-120).