Search

The general objective of a research program on adaptive testing was to identify several sources of potential error in test scores, and to study adaptive testing as a means for reducing these errors. Errors can result from the mismatch of item difficulty to the individual's ability; the psychological effects of testing and the test environment; the inability to extract enough information from the testee's response; deviations from unidimensionality; and an oversimplistic conceptualization of ability. Several different strategies of adaptive testing are discussed, along with the information level they yield, and the bias that can result from various scoring methods. In a discussion of the unidimentionality of test items, the consistency of the testee's response is analyzed. Finally, group differences are examined in terms of the psychological effects of receiving immediate feedback, especially on low ability groups. The author concludes that adaptive testing and immediate knowledge of results may be able to provide testing conditions more conclusive to each person's ability to demonstrate his/her fullest capacities in test performance. (Author/BW)

Five reports from a 2-year study are presented. Frequencies and descriptions of systematic errors in the four algorithms in arithmetic were studied in upper-middle income, regular, and special education classrooms involving 744 children. Children were screened for adequate knowledge of basic facts and for receiving prior instruction on the computational process. Systematic errors contain a recurring incorrect computational process and are differentiated from careless errors and random errors. Errors were studied within levels of computational skill for each algorithm. Results showed that five to six percent of the children made systematic errors in the addition, multiplication, and division algorithms. The figure was 13 percent for the subtraction algortihm. One year later 23 percent of the children were making either the identical systematic error or another systematic error. (Author/JP)

This paper analyzes a typical school spelling task in terms of an information processing model of spelling performance. Based on principles embodied in a computer simulation program previously reported (SPEL by Simon and Simon) the model proposed here has been expanded to try to account for many more kinds of student error than were possible with the original model. It explores the nature of the speller's internal representation of the word to be spelled, the information one must have in memory about the target word in order to produce a correct spelling, and the organization of the processes available to use the information to produce a spelling and to verify its accuracy. Examples of errors from children's spelling tests are used to discuss sources of errors in terms of the model. Ways in which the model may be adapted to explain performance in other kinds of spelling tasks and some implications for instruction and curriculum design are suggested. (Author/HOD)

This is a study of the kinds of processes involved in learning the meaning of individual lexical items, and in particular how the acquisition of lexical meaning is related to the cognitive structuring of events on the one hand and the ability to produce syntactic paraphrases of a word's meaning and other related constructions on the other. It is proposed that errors involving the use of noncausative verbs in a causative sense come about through inference of a derivational or inclusive relationship between causative verbs and their noncausative counterparts and through a generalization by analogy with these causative verbs. It is further argued that when a child first begins to use causative verbs, they are essentially unanalyzed forms, i.e., the child uses a linguistic form without being aware of its internal structure. The evidence that unanalyzed forms have been analyzed is the creation of novel forms which are made up of same or similar components combined according to the same rules. Thus, a fundamental distinction is made between cognitive knowledge and knowledge of a linguistic structure. (AM)

This paper reports results of a preliminary analysis of the errors made by Russian exchange students learning English at S.U.N.Y. in Albany. Error samples are taken from a taped panel discussion containing prepared and spontaneous speech, from a TOEFL test, and from a quiz. Errors are divided into intralingual, or those reflecting general characteristics of rule learning, and developmental, or those illustrating attempts to build hypotheses about English from limited knowledge. Most of the errors were of the intralingual type. An area of specific difficulty was the use of the article; this is shown to be due to a confusion about English syntax, rather than interference from Russian. Problems of interference from French and Spanish are also examined. The structural approach is suggested as a possible source of error, as well as the practice of having idioms memorized without attention to their syntactic contexts. (AM)

Types of hierarchies of pronunciation difficulty are discussed, and a hierarchy based on contrastive analysis plus informal observation is proposed. This hierarchy is less one of initial difficulty than of error persistence. One feature of this hierarchy is that, because of lesser learner awareness and very limited functional load, errors involving allophones are more persistent than errors involving phonemes. Phonetic difficulty and suprasegmental and spelling interference are aggravating factors that increase the relative difficulty of particular problems in the hierarchy. To test the proposed hierarchy and three other hypotheses, a thorough pronunciation test was administered to 50 English-speaking students of Spanish (almost all of whom also knew some French). The test was given at the end of 45 hours of mostly oral instruction, which included 4 to 5 hours of specifically pronunciation instruction. The resulting data formed the basis for error analysis. The results generally confirmed the proposed hierarchy; they also showed the "sacredness" of word boundaries for speakers of English, the negative effects of spelling on pronunciation, greater frequency of errors in cognate than in non-cognate words, evidence of suprasegmental interference, and more frequent errors with known words than with unknown words. Conclusions include several suggestions for the teaching of Spanish pronunciation. (Author/CLK)

This study analyzes Spanish and German errors committed by adult native speakers of English enrolled in elementary and intermediate levels. Four written samples were collected for each target language, over a period of five months. Errors were categorized according to their possible source. Types of errors were ordered according to their frequency. The hierarchies of relative difficulty thus obtained varied for the target languages. The production of some errors seemed to be based on proposed universals of language acquisition; other errors appeared to be directly related to the languages involved. (Author)

This study focuses on a native speaker of Chinese, in her 40's, who began to learn English in her late 20's when she emigrated to the United States. It was discovered that the subject was able to self-correct nearly every error she made in casual speech when the errors were pointed out to her after their commission. Furthermore, in nearly every case she was able to describe the grammatical principle involved. It is speculated that if second language users are able to apply such conscious rules when sufficient processing time is available, perhaps conscious knowledge is applied only when sufficient processing time is available. There is at least suggestive evidence that subconscious "acquired" linguistic knowledge (that is, internalized before puberty) is involved in ordinary unmonitored speech, while consciously "learned" rules are available to the second-language speaker only as a monitor. (Author/DB)

The data for the study were gathered in the course of a larger sociolinguistic survey carried out among francophones from Welland and Sudbury, Ontario. Among other things, the acquisition of spoken English by bilingual francophone students from these cities was studied in depth, using error analysis. The present study attempts to show that in a bilingual environment, the language most often spoken between parents and children plays a major role in the language acquisition of the children. For the study the spoken English of a sample of 15 Grade 2 children, selected from the French language schools of Welland and Sudbury, was analyzed. This sample was divided in two groups. Group I included children who spoke mostly French with their parents at home. Group II children spoke mostly English. The results of our study show that: (1) Group II children commit a lower proportion of interference errors (errors attributable to the influence of French) than Group I children. On the basis of these results we can say that the language most often spoken between children and parents seems to have a significant influence on the language acquisition of Grade 2 children in a bilingual environment. (Author/KM)

This paper analyzes short stories written in English and French by French immersion pupils at the Grade 3 level. Their English writing skills are compared to those of Grade 3 pupils in a regular English program. Aspects of their writing skills which are examined include vocabulary skills, technical skills (punctuation, capitalization and spelling), grammatical skills and creativity. The errors made by the pupils are discussed in detail. (Author)

This study examines the usage of preposition "pour" in the written and spoken French of two groups of Grade 9 and 12 Franco-Ontarian bilingual students from Welland and from the Sudbury area. Both the students' written and spoken French are shown to include a substantial proportion of constructions with "pour" which deviate from formal Franco-Ontarian French. A comparison of the French of the Franco-Ontarian students with that of unilingual francophone Quebeckers indicates that these constructions are typical of Franco-Ontarian French and may be partly attributed to interference from the English language. (Author)

Previous work by the authors permitted them to hypothesize the existence of certain universal cognitive strategies that play a significant role in child second language acquisition. Forming the basis of the "creative construction process" in L2 learning, these strategies have heretofore remained unspecified. This paper offers new perspectives on the specific nature of the creative construction process: first, by an attempt to use Brown's notions of semantic and linguistic complexity to account for differences between first and second language acquisition; second, by a clarification of the notions of "learning complexity" and "learning strategy"; and finally, by presenting a new analytical procedure and framework which may be useful in research on children's strategies in acquiring L2 syntax. This new procedure involves an adaptation of Bart and Krus'"Ordering- Theoretic Method," which assumes that there is a logical relationship among items in a group, and which is designed to determine those relationships. The ultimate value of this type of analysis lies in its potential to uncover the sorts of empirical facts needed before one can begin to specify with any confidence the nature of language acquisition strategies. (Author/DB)

Participants in a seminar series in second language acquisition held at Harvard University discussed three papers by Dulay and Burt ("Goofing: An Indicator of Children's Second Language Learning Strategies,""Should We Teach Children Syntax?", "Natural Sequences in Child Second Language Acquisition"), and developed several questions and issues regarding specific aspects of those papers. Among the issues raised are: (1) universal language processing strategies; (2) statistical discrepancies between the second and third studies; (3) interference in second language acquisition; (4) theoretical model of morpheme acquisition order; (5) morpheme scoring; (6) statistics and data analysis. Each question posed by the group is followed by a response from Dulay and Burt. (DB)

Three hypotheses are examined in relation to English copula and negative utterances produced by three native Spanish speakers. The hypotheses are interference, interlanguage and L1=L2, which states that acquisition of a language by second language learners will parallel acquisiton of the same language by first language learners. The results of the study are as follows: (A) In the copula construction, utterances such as "Is a book" probably represent interlingual identifications of the English "it's" and the Spanish "es." They do not appear to be produced by interference from the Spanish syntactic device of subject omission. (B) The acquisition sequence of the English negative by the subjects does not correspond to the "stages" described by Klima and Bellugi for children acquiring English as a native language. This result tends to disconfirm a strong version of the L2 = L1 hypothesis which would claim that the acquisition order in both the first and second language should be the same. (C) It is difficult to say whether acquisition of the negative by the subjects is systematic and developmental because these concepts are not clearly defined in any versions of the interlanguage hypothesis. (D) Two of the subjects evince different strategies in acquiring the negative. (Author/DB)

Classic studies in second language (L2) learning offer little evidence for the validity of the notion of universals in L2 learning. The present study investigates this notion in data collected from 15 observational studies of 40 L2 learners who acquired the L2 naturally, that is, they were not taught the language in any formal ways. Interpretation of both input data and production data (e.g., L2 auxiliary development, auxiliary development with negation, auxiliary inversion in question formation, and input frequency and question production) suggests that communication is the principal goal of such learners, and that consequently those parts of the language system which are not important to communication are learned slowly. However, if a structure is extremely frequent in the input data, the learner will produce it. Effects of frequency are modified in a number of ways. If a form has low semantic power, or if it requires changes in word order, it will be learned late, as will structures having a variety of forms. Variations in strategies and speed of L2 acquisition reflect psychological differences, and must be understood before a definitive model of universals in L2 learning can be developed. (DB)

Articles in this volume relate to the Yugoslav Serbo-Croatian-English Contrastive Project: (1) "The Yugoslav Serbo-Croatian-English Contrastive Project at the End of its Second Phase (1971-1975)," Rudolf Filipovic: Pedagogical goals and application of contrastive analysis are best achieved when accompanied by error analysis. Reports, papers and pedagogical materials resulting from the second phase are discussed and a thrid phase projected. (2) Dora Macek's "Relatives in English and their Serbo-Croatian Equivalents": Discusses relatives in both languages through contrastive analysis and notes implications for a Serbo-Croatian learner of English. (3) "On Compiling Pedagogical Materials," Mirjana Vilke: Describes procedures used to test theories and the steps in constructing pedagogical materials. (4) Ralph E. Weber's "Critical Response to the Project": Examines comments made about the study. (5) This section presents biographical and bibliographical data on contributors to the project. (6) Zeljko Bujas, "On Translational Conversion in English: Serbo-Croat Contrastive Analysis": Outlines the problems of a contrastive study of conversion patterns in the two languages. (7) "On Serbo-Croatian Complement Sentences," Collette Craig: Classifies verbs in complement sentences and discusses tenses of embedded verbs. (8) Gordana Opacic, "On Psycholinguistics and its Assumed Relevance to Contrastive Analysis": Discusses theories of language acquisition. (CHK)

For years language teachers have conducted error analysis for remedial purposes. More recently error analysis has assumed a developmental purpose, namely as a clue to the process of acquiring a second language. Causes of learner errors, such as interference from the first language, confusing aspects of the second language, or learners fostering their own errors, are discussed; and teacher responses appropriate for the error type are suggested. Factors influencing the teacher's decision as to whether and when to correct are examined, i.e., the available information about the error, the importance of correction, the ease of correction, and the characteristics of the students. Then the teacher's options as to specific treatment for correcting errors are considered. Finally, some reservations about error analysis are entertained. (Author)

This is a contrastive analysis of British English, American English and Hungarian sentence prosody. The first part is an introduction stating the study's objective, scope and data, and briefly surveying the related literature. It outlines the study's view of prosodic features and its principles of comparison and prediction. Part Two inventories prosodic devices in English and Hungarian, describes them formally and predicts the formal errors learners of either language may make. The next two sections deal with the functional aspects of stress: Part Three compares stress behavior of grammatical phrases and Part Four that of sentence types and sub-types in both languages. Predictions of learners' errors are made. Part Five briefly outlines rhythmical stress modification in sentences. The next three sections describe the intonation of sentences co-extensive with one tone-group, sentences containing final vocatives and quoting clauses and sentences containing tone-group sequences. Again, learners' errors are predicted. The final section summarizes, notes pedagogical implications and suggests further research. (CHK)

Results are presented of an onsite review of data gathered in 1971 by the Consolidated Program Information Report (CPIR), a survey instrument designed to fill the data needs of the Office of Education and state education agencies in relation to federally funded elementary and secondary education programs. This validation and error analysis of CPIR data was undertaken to examine the existence of both systematic and random errors in data reported on the CPIR forms; to suggest changes in procedures, forms, concepts, and instructions; and to identify the needs for further research in validity study areas. In the beginning sections of the study the methodology is described, and the analysis of the case study materials that were collected during site visits is presented, including a discussion of commonalities and trends across local education agencies (LEAs) and states. Chapter 4 presents an analysis of the discrepancies found between the data collected in the study and those originally submitted by the LEA. Chapter 5 contains a series of recommendations for changes in the CPIR instruments, procedures, and instructions. The reasons for the recommendations are documented and the impact on the CPIR data-collection effort is discussed both in terms of ease of collection and data accuracy. (Author/MLF)

Specific evaluation design is a non trivial problem which must be developed through a set of procedures involving considerable study of the enterprise to be evaluated. Three criteria to be considered when an evaluator provides data for a decision maker are: (1) the percentage of the evaluation data that are actually used by the decision maker; (2) the percentage of the decision maker's decisions that are made, in part, with the use of the evaluation data; and (3) the degree to which the data is utilized in the most important decisions, even where all decisions are not made using the evaluation data. These criteria have been named efficiency, completeness, and focus, respectively. Evaluators should avoid errors, and prefer operational to non operational techniques. Evaluator/decision maker interaction patterns should be incorporated into the evaluation design. (Author/BJG)

When classes are the units of analyses, estimates of the reliability of class means are needed. Using classical test theory it is difficult to treat this problem adequately. Generalizability theory, however, provides a natural framework for dealing with the problem. Each of four possible formulas for the generalizability of class means is derived from two points of view. Each of the four generalizability coefficients is shown to be the expected value of a variation on the split-half method for estimating reliability. Finally, the four coefficients are related to estimates of the reliability of class means that have been used previously. (Author)

This paper reports on the development of an error-coding instrument and a computerized continuous feedback system for the diagnostic evaluation and remedial treatment of unstructured second-language performances and the use of such a system in an instructional setting. The system does the following: it allows quick, objective, accurate, and detailed evaluation of "creative" written and spoken German; it provides student performance files in the structural and phonological elements of the language that can be updated as required; it allows accurate performance-to-performance comparisons of progress; it permits the instructor to determine quickly the learning status of an individual student, a group of students, or the entire class, and subsequently to design appropriate exercises for that particular target group; and it allows students to judge their performance from the same criteria which were available to their instructor for marking. Feedback procedures and a tabulation of errors for three levels of German language instruction are presented and their systemic implications for the design of pedagogical grammars are discussed. (Author/KM)

Empirical Bayes point estimates of true score may be obtained if the distribution of observed score for a fixed examinee is approximated in one of several ways by a well-known compound binomial model. The Bayes estimates of true score may be expressed in terms of the observed score distribution and the distribution of a hypothetical binomial test. The latter distribution is found by use of the compound binomial approximation formula and from relationships which exist between Bayes estimates and unconditional probabilities of observed score. Empirical Bayes point estimates are obtained by use of the sample observed score disribution. (Author)

The report describes an analysis of errors yielded by a written test in English given to 424 Swedish 14-year-olds from 24 different classes in the comprehensive school. The analysis discusses the difference in attainment between pupils of high and low proficiency, frequencies and types of errors for regular and irregular verbs, and the occurrence of systematic and nonsystematic errors. It also tries to trace the errors back to the influence of the pupils' first language, to mechanisms regulating the acquisition of skills in general, and to unsatisfactory teaching strategies and teaching material. The investigation is part of a large-scale project trying to establish frequencies and types of errors in oral as well as written production by the age group in question. Informant tests based on the most frequent errors are to be administered in England in order to lay down the communicative value of the errors. (Author)

This is one of a series that is a collection of translations from the extensive Soviet literature of the past 25 years on research in the psychology of mathematics instruction. It also includes works on methods of teaching mathematics directly influenced by the psychological research. Selected papers and books considered to be of value to the American mathematics educator have been translated from the Russian and appear in this series for the first time in English. The aim of this series is to acquaint mathematics educators and teachers with directions, ideas, and accomplishments in the psychology of mathematical instruction in the Soviet Union. This series should assist in opening up avenues of investigation to those who are interested in broadening the foundations of their profession. The seven studies found in this volume are: An Experiment in the Psychological Analysis of Algebraic Errors; Pupils' Comprehension of Geometric Proofs; Elements of the Historical Approach in Teaching Mathematics; Overcoming Students' Errors in the Independent Solution of Arithmetic Problems; Stimulating Student Activity in the Study of Functional Relationships; Psychological Grounds for Extensive Use of Unsolvable Problems; and Psychological Characteristics of Pupils' Assimilation of the Concept of a Function. (Author/MK)

Nine subjects, 5 of whom claimed absolute pitch (AP) ability were instructed to rapidly strike notes on the piano to match randomized tape-recorded piano notes. Stimulus set sizes were 64, 16, or 4 consecutive semitones, or 7 diatonic notes of a designated octave. A control task involved motor movements to notes announced in advance. Accuracy, measured on the basis of deviations of responses from targets, significantly differentiated AP from NAP (non-AP) subjects at all set sizes except the control task. For both groups, accuracy increased as set size decreased. Decision times--that part of total response time before a movement to the response note began--decreased as set size decreased, but did not differentiate AP and NAP subjects. The results are discussed in terms of a two-factor theory of AP ability whereby NAP subjects use only relative tone height as the basis of judgment, but AP subjects also use standards of tone chroma stored in long-term memory. The abnormally high channel capacities and rates of information gain for AP subjects are based on the transmission of information in two channels, whereas NAP subjects transmit information in only one channel. (Author/MV)