Call number: LD2668 .T4 1946 E3 / Master of Science

The problem of this study was to develop an instrument to measure the achievement in modern mathematics in the primary grades.

The purpose of this study was to determine if selected variables were significant predictors for the academic success or failure of Algebra I students. Two methods of data gathering were used in the study. The Student Attribution Inventory (SAI) was first administered to the classes during one class period. One hundred eighty-five students participated. Only two students could not because of a language barrier. The second method of data gathering was done by the researcher. First, a letter was sent home to the students' parents for permission to use the cumulative folders of their sons and daughters. Letters were collected daily by the teacher and weekly by the researcher. A follow-up letter was sent home several weeks later to those parents who had not previously responded. Once permission was granted, the researcher received the cumulative folders and obtained the necessary information. Out of 185 students, 110 were used in the study. Stepwise multiple regression was used for four of the eight hypotheses. These were tested at the .05 level of significance. This level of prediction was used to determine if selected variables were significant predictors for the academic success or failure of Algebra I students. The findings showed that achievement motivation (a resultant behavior of an emotional conflict between hopes for success and fears of failure), luck (external, unstable, unintentional factor which is related to the perceived randomness in events), and previous mathematics achievement (as measured by the California Achievement Test (CAT)), were the variables found to be statistically significant. Ability, previous reading achievement, and age were not entered into the regression equation by the Statistical Package for the Social Sciences (SPSS) because of insufficient predictability. The remaining hypotheses employed t-tests as the statistical method of testing. These four hypotheses were also tested at the .05 level of significance. The results showed that neither sex differences nor ethnic groups were significant in prediction of Algebra I final grades and achievement motivation. Further research is necessary to increase the understanding of achievement motivation and luck in classroom situations, counseling sessions, and group dynamics.

by Lee Fong Lok. / Chinese title: / Bibliography: leaves 103-111 / Thesis (M.A.Ed.)--Chinese University of Hong Kong

No description available.

A confirmatory approach based on a multidimensional model (Douglas, Roussos, & Stout, 1996; Shealy & Stout, 1993; Stout & Roussos, 1995) was used to identify sources of differential item functioning (DIF) and differential bundle functioning (DBF) for boys and girls on the British Columbia Principles of Mathematics Exam for grade 12 (PME12). Data consisted of a total of 9404 examinees; 4335 girls and 5069 boys. There were 45 multiple choice items in the exam. Analyses were completed in two stages. In stage 1, patterns present in the gender DIF research in mathematics were identified. Stage 2 was the statistical confirmation of these patterns. Sources of gender DIF were confirmed for the content areas: polynomial, quadratic relations, logarithms and exponents. Items tapping higher cognitive levels dealing with patterns and relation, word problems, and items containing visuals were also confirmed as a source of DIF. Exploratory analyses indicated that computation items for which no equations are provided may be a source of DIF along with trigonometry items. This study contributes to an increased understanding of sources of gender DIF that may assist test developers to ensure that mathematics items measure the construct that they are intended to measure and that the test as a whole measures that which it purports to measure. The findings of this research provide an additional source of information about the differential performance of boys and girls that may be used to develop guidelines and test construction principles for reducing gender DIF in mathematics. This research also contributes to a greater understanding of gender differences in mathematics learning and achievement.

The purpose of this research was to determine if teacher written evaluations in arithmetic and the frequency of administering them would produce a significant difference in arithmetic achievement. The following null hypotheses were set up to guide the statistical analysis:1. There will be no significant difference between the achievement of students receiving no teacher written evaluation and those receiving teacher written evaluation every day.2. There will be no significant difference between the achievement of students receiving teacher written evaluation once every two weeks than those who receive no teacher written evaluation.3. There will be no significant difference between the achievement of students receiving teacher written evaluation every day than those having no evaluation and those evaluated once every two weeks.Three groups of sixth grade children in Anthony Elementary School in Muncie, Indiana were used in this study. Each group contained fourteen boys and eight girls. The mean intelligence quotients determined by the Otis Lennon Dental Ability Test of the three groups were approximately the same (110). Each group had a different teacher. Arithmetic was taught at the same time for fifty minutes each day for eighteen weeks. The researcher met each week with the teachers to determine pace and material to cover the next week.Group I did not use tests of any type during the study. Group II was given teacher written evaluations once every two weeks. Group III was evaluated every day by teacher written tests.The Standard Achievement Test, Modern Mathematics Concepts Tests, Form X was given at the beginning of the study. The analysis of variance and the "F" test was applied to the raw scores resulting in the value of "F" as 1.286. A difference at the .C3 level of confidence would necessitate an "F" value of 3.14. We therefore concluded that there was no significant difference in the achievement of the three groups at the beginning of the study.Form W of the Stanford Achievement Test, Modern Mathematics Concepts Tests, was given at the conclusion of the study. Again using the analysis of variance and the 'F" test the value of "F" was 1.284. A difference at the .05 level of confidence would necessitate en "F" value of 3.14. We therefore concluded that there was no significant difference in the achievement of the three groups at the conclusion of the study.The analysis of variance and the "F" tests was used to compare the group because of the involvement of more than two groups. This means of comparison eliminates the necessity of comparing subsamples one by one which is nearly impossible because of the calculation involved.Null hypothesis number one which stated, "There will be no significant difference between the achievement of students receiving no teacher written evaluation and. those receiving teacher written evaluation every day," can not be rejected at the .05 level of confidence and must be accepted with the limits of this study as valid end reasonable.Null hypothesis number two which stated, "There will be no significant difference between the achievement of students receiving teacher written evaluation once every two weeks than those who receive no teacher written evaluation," can not be rejected at the .05 level of confidence and must be accepted within the limits of this study as valid and reasonable.Null hypothesis number three which stated, "here will be no significant difference between the achievement of students receiving teacher written evaluation every day than those having no evaluation and those evaluated every two weeks," can not be rejected at the .05 level of confidence and must be accepted within the limits of this study as valid and reasonable.

The purpose of this study was to establish and describe the validity and reliability of an instrument based on a model of teaching behaviors conceived as magnetic lines of force.Content validity was established by reviewing the literature and identifying possible teaching behaviors which researchers felt contributed to increased pupil learning. These behaviors were presented to a panel of judges from faculty in the Teachers College, Ball State University, to ascertain behaviors which they felt increase the likelihood pupils would increase comprehension in reading. After the panel had made their suggestions, the list of behaviors was revised and used with two middle school classes at the Burris Laboratory School. Pupils were asked to identify the behaviors which they felt were associated with their increased learning. A final list of twenty items became the revised instrument for use in the study.A group of practicing teachers critiqued the list of twenty behaviors. This group was directed to identify how strongly they agreed with the placement of each of the behaviors under one of the four scales postulated. Construct validity was established by discussing the research supporting each of the scales, subscales and behaviors listed in the instrument.Each of the twenty items was judged as valid by a minimum of 70 percent of the selected panel members and two classes of selected middle school pupils. Each of the twenty items included was recognized as a valid example of the constructs identified with the exception of those items assumed to reflect firmness.I t was possible to relate each of the twenty items in the instrument to research studies accepted by editors and authors in the field despite inherent weaknesses in definitions, experimental designs, and theoretical orientation. In this sense, the instrument has acceptable construct validity.The instrument was then used to gather data on a group of teacher trainees assigned to a laboratory school at Ball State University. After each teacher trainee had taught a short unit, the pupils taught filled out the instrument which had been devised. Reliability measures consisted of computing coefficient alpha for various major scales and subscales of the instrument. Computations were run by the Ball State University Center using the reliability program from the SPSS Statistical Package for the Social Sciences DEC system -10.The coefficient alpha reliability indexes of .90 for the total scale, .86 for the cognitive scale, and .78 for the affective scale were judged acceptable for use in additional research studies in the field. The coefficient alpha reliability index of .82 for the clarity subscale, .81 for the flexibility subscale, and .78 for the warmth subscale were also judged acceptable for use in additional research studies in the area. The coefficient alpha reliability index of .68 for the firmness subscale suggests questionable reliability for all but the most general research purposes.Recommendations were formulated to encourage further development of the model and refinement of the instrument (particularly the "firmness" subscale).

A confirmatory approach based on a multidimensional model (Douglas, Roussos, & Stout, 1996; Shealy & Stout, 1993; Stout & Roussos, 1995) was used to identify sources of differential item functioning (DIF) and differential bundle functioning (DBF) for boys and girls on the British Columbia Principles of Mathematics Exam for grade 12 (PME12). Data consisted of a total of 9404 examinees; 4335 girls and 5069 boys. There were 45 multiple choice items in the exam. Analyses were completed in two stages. In stage 1, patterns present in the gender DIF research in mathematics were identified. Stage 2 was the statistical confirmation of these patterns. Sources of gender DIF were confirmed for the content areas: polynomial, quadratic relations, logarithms and exponents. Items tapping higher cognitive levels dealing with patterns and relation, word problems, and items containing visuals were also confirmed as a source of DIF. Exploratory analyses indicated that computation items for which no equations are provided may be a source of DIF along with trigonometry items. This study contributes to an increased understanding of sources of gender DIF that may assist test developers to ensure that mathematics items measure the construct that they are intended to measure and that the test as a whole measures that which it purports to measure. The findings of this research provide an additional source of information about the differential performance of boys and girls that may be used to develop guidelines and test construction principles for reducing gender DIF in mathematics. This research also contributes to a greater understanding of gender differences in mathematics learning and achievement. / Education, Faculty of / Educational and Counselling Psychology, and Special Education (ECPS), Department of / Graduate

This study seeks to determine the nature of the intellectual demands of the Higher Grade course in Mathematics with a view to early, more accurate prediction of individual pupil success in this course. The need for such early prediction is clearly indicated by the interest shown on the part of parents and pupils alike during the Standard Seven year where the realisation exists that Matriculation Mathematics is a subject sometimes found to be "overwhelmingly difficult". The "drop out" figure from the Higher Grade course to the Standard Grade course in most schools further demonstrates the need for more careful selection at the Standard Seven level. Both old (1973) and new (1984) syllabuses are analysed to determine the nature of the content and the intellectual level at which this should be taught. In addition, a series of past Cape Senior Certificate examination papers are investigated to reveal information about the nature and level of examining. Mental processes involved in the examination items are classified and the general composition of the examination papers is discussed. A test device suitable for Standard Seven pupils is developed on the basis of the composition of the Higher Grade Matriculation examination papers analysed. The object of this test is to provide that early indication to pupils of their ability to cope with the level of mental process required by the Higher Grade course in Mathematics. The investigation describes the construction, administration and further development of the test device and, furthermore, seeks to show its predictive validity for the Matriculation examination in Mathematics by comparing test results with successive school examination results over a three year period. The possibility of sex differences in Mathematics achievement and prediction are also investigated on the basis of the results obtained during the course of this experiment. General conclusions are drawn, the difficulties encountered are discussed and some suggestions for further research are offered.