Search

This article reports on research designed to respond to a set of problems normally associated with the assessment of school performance. It is more concerned with specific resource effects in specific schools than with effectiveness and efficiency in the use of individual school resources in the assessment of a system 's performance over time. The emphasis is on the overall performance of the entire set of elementary schools in the Chicago Public School System rather than on the resource use in each school. Joining standard regression techniques with the results of data envelopment analysis, the basic questions asked are the following: Are there discernible changes in the performance of Chicago elementary schools following the fundamental school reforms of 1988-1989? Are these changes in the desired direction? The results suggest encouraging improvement in schooling outcomes over the period of study.

Prior school effectiveness research was extended in this study by an examination of classroom processes in urban elementary schools that have been identified as effective/efficient (E/E) and ineffective/inefficient (I/I), based on both data envelopment analyses (DEA) (Anderson & Walberg, 1994) and multiple-regression analyses. Achievement test scores and school demographic data were used initially to determine the efficiency and effectiveness ratings of 167 elementary schools from a large urban school district in the south central region of the United States. Four schools were randomly selected from the population of I/I schools. Systematic classroom observation and classroom learning environment measures were collected from each of these schools. The results indicated several classroom process differences between these E/E and I/I urban schools. In the I/I schools, for example, students were observed interacting with their teacher only about 47% of the time and there were no classroom interactio...

To assess the relative influences of education, social environment, and current activities on the practical literacy of young adults in the United States, data were drawn from the 1985 National Assessment of Educational Progress survey of 3,618 21–25 year olds. Oral and written examinations provided reliable measures of comprehension of written language, documentary and writing skills, pictorial and chart understanding, and numeracy; oral and some additional writing skills were assessed with single-item rating scales. A multidimensional model of literacy distinguished 20 early background characteristics of education, home, and family from 16 literacy-promoting current adult activities—all of which were coded into scales from respondent reports. Regression and canonical analyses showed that both early experience and current activities powerfully influence young-adult literacy.‘Matthew effects,’ however, show early environmental advantages yield subsequent advantages that lead to wide disparities in adult literacy skills. RESUME Les pratiques de lecture-ecriture de jeunes adultes: antecedents pedagogiques et influences On a tire de l'evaluation nationale faite aux Etats-Unis en 1985 par la Evaluation Nationale du Progres Pedagogique sur des personnes âgees de 21 a 25 ans des donnees permettant d'evaluer les influences relatives de l'education, de l'environnement social et des activites habituelles sur les pratiques courantes de lecture-ecriture de jeunes adultes americains. Des examens oraux et ecrits ont fourni des mesures fiables de la comprehension de la langue ecrite, des capacites a se documenter et A ecrire, de la comprehension des images et des graphiques, et de la numeration. D'autres competences ecrites et verbales ont eteevaluees a l'aide d'echelles a item unique. Un modele multidimensionnel du savoir lire-ecrire a fait apparaitre 20 caracteristiques relatives a l'education initiale, a la maison, A la famille provenant de 16 activites adultes de promotion de la lecture codees sous forme d'echelles construites a partir des reponses fournies a plus de 400 questions. Les scores des facteurs derives d'une analyse en composantes principales de ces caracteristiques et activitks ont ete utilises comme predicteurs dans trois modeles statistiques. Les premieres caracteristiques, comportant la presence d'objets a lire a la maison, l'education parentale et le revenu, ainsi que le type de scolarite secondaire, ont ete maintenus ‘constantes’ pour des raisons statistiques et theoriques. Les activites adultes courantes, comme la lecture en contexte de travail, A l'ecole, et a la maison, ainsi que les activites associatives, la participation a la vie sociale, et les activites sociales liees a la maitrise de l'ecrit, ont ete considerees comme ‘variables'. On a effectue des analyses canoniques en trois etapes, la premiere pour explorer les effets des facteurs constants sur le lire-ecrire, la seconde pour leurs effets sur les activites variables, et la troisieme sur les variables relatives au controle du lire-ecrire sur les variables constantes. On a adopte ce plan afin de disposer d'informations qui suggerent les domaines de comportements au cows de l'enfance, de la jeunesse et des debuts de l'âge adulte susceptibles de repondre A des interventions de promotion des pratiques de la lecture-ecriture. Les facteurs constants ont revele une association entre une position socio-economique avantageuse au depart et un patron general de performance en lecture-keriture qui se caracterise surtout par des capacites de comprehension en lecture et en documentation. Le second modele, qui examinait les effets des facteurs constants sur les variables, a indique que l'education du repondant et ses avantages familiaux etaient associes avec le niveau de lecture, la poursuite d'etudes longues, et le montant du salaire. Le test des influences independantes des activites courantes sur la lecture-ecriture a montre que tous les effets etaient generalement predits par la perception des capacites de lecture-keriture likes au travail et la lecture de livres varies, de magazines et de journaux. Prise dans son ensemble, cette recherche suggere que tant l'experience initiale que les activites courantes influencent la lecture-ecriture des jeunes adultes.‘L'effet Matthew’ montre cependant que les avantages sociaux de depart produisent des avantages qui conduisent aelargir les ecarts ulterieurs de capacite de lecture-ecriture des adultes.

A meta-analysis of 43 studies of the effects of computer-based instruction (CBI) in secondary schools showed that CBI programs produce, on average, small but significant improvement in students' sc...

To assess the impact of the innovative precollege science curricula of the past twenty years on learning, a search was conducted using the computer-assisted Bibliographic Retrieval System (BRS), the ERIC Annual Summaries of Research in Science Education, and Dissertation Abstracts International. A total of 151 effect sizes were obtained from 33 studies representing 19,149 junior and senior high school students in the United States, Great Britain, and Israel. Study-weighted analysis yielded an overall mean effect size of 0.31 significantly favorable to the innovative curricula [t(25) = 2.183, p < 0.05] on all outcomes. Student performance in innovative curricula averaged in the 62nd percentile relative to the control norm. Tabulation of signed comparisons indicated that sixty-four out of eighty-one unweighted outcomes were favorable to the innovative curricula. Separate analyses for test content bias, methodological rigor, type of learning, and student characteristics showed no significant differences across these categories.

Forty-eight studies on computer based instruction (CBI) in the elementary schools were coded on 45 variables using meta-analytic techniques. Results of the synthesis indicated computer assisted instruction (CAI) raised outcome measures by .45 standard deviation units. This is interpreted to mean that students in CAI treatment groups performed on the 65th percentile of the control group distribution. Results indicated that younger and lower achieving students profit most from exposure to CAI. Drill and practice is most effective in raising student achievement scores. The effects of computer managed instruction are minimal. The computer functions best when the task to be learned is relatively simple paired association.

This paper reviews eight theories or models presenting holistic conceptions of student learning in classroom settings (Bennett, 1978 ; Bloom, 1976 ; Bruner, 1966 ; Carroll, 1963 ; Cooley & Leinhardt, 1975 ; Gagné, 1974 ; Glaser, 1976 ; Harnischfeger & Wiley, 1976 ). To be included, a model or theory was required to describe variables important to the performance of individual learners or single instructional tasks. Most models also derived implications for the organization of curriculum and/or group instruction. Following discussions of the eight models, major constructs posited by different theorists are cross-tabulated and related to factors of the model of educational productivity (Walberg, 1980 ).

This paper synthesizes empirical studies of homework and of various homework strategies on the academic achievement and attitude of elementary and secondary students. A computer search yielded 15 p...

This paper examines the literature concerning computer-assisted instruction (CAI) among exceptional children. Two approaches to meta-analysis were used: a vote count and a computation and analysis of effect sizes. Using 22 studies, the effectiveness of CAI with exceptional children was determined to be largely effective on the basis of the vote count, which ascertained whether or not a study reported positive results. Six studies were eliminated from the second analysis—that of calculating effect sizes—because they did not report sufficient statistical information. The median effect size was .52, a moderate positive effect. On the basis of the meta-analyses, CAI appears to be effective with exceptional children, especially those at lower levels of learning. Suggestions for future research regarding CAI use with exceptional children were made.

To probe a psychological theory of educational productivity, social studies achievement and attitude test scores of 2,001 17-year-old high school students (from a National Assessment of Educational Progress sample) were statistically related to each other and to indicators of constructs that prior research shows are associated with learning outcomes. In eight linear and log-linear, ordinary least-square regressions with from 7 to 15 independent variables, several production factors are significant-socioeconomic status, home environment, traditional instruction, time or amount of study, and (negatively) leisure-time television viewing. Also, item-learning correlations yield clues for improving learning productivity. Systems regressions, however, reveal ambiguity in the causal relations in that measurement error and exogenous and reverse causes may account for the significant findings.

To determine the dependence of adult outcomes on connections, certification, and verbal competence, 26 indexes of parent socioeconomic status and other variables (years of education, diplomas and degrees, and scores on a vocabulary test) were related to seven characteristics (occupational prestige, income, health, number of memberships in organizations, confidence in institutions, overall happiness, and interest in life) using simple, multiple, and canonical correlations in a stratified random national sample of 1,800 adults. The analyses revealed that each set of variables brings a small but significant and unique influence to the predictions of the outcomes, but a substantial amount of the variance in the outcomes, especially of happiness, remains unexplained.

Computer searches of three data bases and hand searches of four major review articles yielded 35 studies of the association between academic and occupational performance in various fields completed since 1950. The study-weighted mean of the 209 correlations is .155, and the standard deviation is .176. Although highly significant, academic indicators such as grades and test scores account on average for only 2.4% of the variance in occupational performance criteria such as income, job satisfaction, and effectiveness ratings. Predictions of occupational performance from academic indexes were somewhat higher in business and nursing, somewhat lower in teaching and engineering, and not significant for MD’s and PhD’s. The correlation magnitudes depend on several characteristics of the studies and their populations.

A recent meta-analysis showed large effects of teacher use of higher order questions on student achievement (Redfield & Rousseau, 1981) and contradicted Winners (1979) previous review which showed the research to be inconclusive. For this reason, a quantitative synthesis was conducted to estimate the quasi-experimental and experimental effects of teachers’ questioning strategies on academic achievement. Fourteen studies, which contrasted effects of predominantly higher cognitive questions and predominantly factual questions, were examined and were compared with the results of the earlier reviews. Higher cognitive questioning strategies have a small positive median effect on learning measures but not as large as has been suggested by the previous meta-analysis (Redfield & Rousseau 1981). Moderate and even large effects may exist but remain to be demonstrated.