Olaf Köller

Does interest matter? The relationship between academic interest and achievement in mathematics

A total of n = 602 students (59.5% female) from academically selected schools in Germany were tested at three time points--end of Grade 7, end of Grade 10, and middle of Grade 12-in order to investigate the relationships between academic interest and achievement in mathematics. In addition, sex differences in achievement, interest, and course selection were analyzed. At the end of Grade 10, students opted for either a basic or an advanced mathematics course. Data analyses revealed sex differences in favor of boys in mathematics achievement, interest, and opting for an advanced mathematics course. Further analyses by means of structural equation modeling show that interest had no significant effect on learning from Grade 7 to Grade 10, but did affect course selection-that is, highly interested students were more likely to choose an advanced course. Furthermore, interest at the end of Grade 10 had a direct and an indirect effect (via course selection) on achievement in upper secondary school. In addition, results suggest that, at least from Grade 7 to Grade 10, achievement affected interest-that is, high achievers expressed more interest than low achievers. The findings underline the importance of interest for academic choices and for self-regulated learning when the instructional setting is less structured.

Classroom climate and contextual effects : conceptual and methodological issues in the evaluation of group-level effects

Classroom context and climate are inherently classroom-level (L2) constructs, but applied researchers sometimes—inappropriately—represent them by student-level (L1) responses in single-level models rather than more appropriate multilevel models. Here we focus on important conceptual issues (distinctions between climate and contextual variables; use of classroom L2 rather than student-level L1 measures) and more appropriate multilevel models. To illustrate these issues, we consider the effects of two L2 classroom climate variables and one L2 classroom contextual variable on two L1 student-level outcomes for 2261 students in 128 classes. Through this example, we illustrate how to apply evolving doubly latent multilevel models to (a) evaluate the factor structure of L1 and L2 constructs based on multiple indicators of classroom climate and context measures, (b) control measurement error at L1 and L2, (c) control sampling error in the aggregation of L1 responses to form L2 constructs (the average of student-level responses to form classroom-level constructs), and (d) provide guidelines for appropriate analysis of classroom climate as an L2 construct. [Supplementary materials are available for this article. Go to the publishers online edition of Educational Psychologist for the following free supplemental resources: Substantive basis of the present investigation and more detailed description of the methodology.]

Reunification of East and West German School Systems: Longitudinal Multilevel Modeling Study of the Big-Fish-Little-Pond Effect on Academic Self-Concept

Longitudinal data from large cohorts of seventh grade (n = 2,778) East and West German students were collected at the start of the reunification of the school systems to evaluate how this remarkable social experiment affects self-concept formation. Multilevel modeling demonstrated a negative “big-fish-little-pond effect” (BFLPE); attending classes where class-average math achievement was higher led to lower math self-concepts. West German students attended schools that were highly stratified in relation to ability before and after the reunification, whereas East German students first attended selective schools after the reunification. Consistent with theoretical predictions based on this difference, the negative BFLPE—the negative effect of class-average achievement—was more negative in West German schools at the start of the reunification. This difference, however, was smaller by the middle of the year and had disappeared by the end of the first post-reunification school year. Whereas East and West German results both support the negative BFLPE, their differences supported theoretical predictions, extended theory, and demonstrated how changes in school policy influence the formation of academic self-concept.

Dimensional comparisons : An experimental approach to the internal/external frame of reference model

According to the internal/external frame of reference model (H. W. Marsh, 1986, 1990a), students not only use social comparisons to evaluate their performance (external frame of reference) but they also use dimensional comparisons (internal frame of reference), comparing their own achievement in one subject with that in other subjects. Three experimental studies were conducted to investigate the psychological processes underlying the effects of achievement in one domain on self-perceived competence in another. In Study 1 (N = 36), high achievement in one domain led to lower self-perceived competence in the other domain. Study 2 (N = 45) showed inverse effects on self-perceived competence when achievement feedback included explicit dimensional comparison information about students achievement in both tasks. In Study 3 (N = 90), dimensional comparison effects were shown even when additional social comparison information was presented.

The Reciprocal Internal/External Frame of Reference Model An Integration of Models of Relations Between Academic Achievement and Self-Concept

The reciprocal internal/external frame of reference model (RI/EM) combines the internal/external frame of reference model and the reciprocal effects model. The RI/EM predicts positive effects of mathematics and verbal achievement and academic self-concepts (ASC) on subsequent mathematics and verbal achievements and ASCs within domains and negative effects of mathematics and verbal achievements and ASCs on subsequent achievements and ASCs across domains. Although ample support was provided for the I/E model by cross-sectional data and for the REM within a single domain, there has been almost no research on the longitudinal generalizability of the reciprocal cross-domain effects. Using three waves of data collection from Grade 5 to Grade 8 with N = 1,508 students, analyses supported the validity of the RI/EM, revealing positive longitudinal effects of grades and ASCs on subsequent grades and ASCs within domains and negative effects of grades on subsequent ASCs across domains. There were also small negative effects of ASCs on subsequent grades across domains.

Kurswahlen von Mädchen und Jungen im Fach Mathematik: Zur Rolle von fachspezifischem Selbstkonzept und Interesse

Zusammenfassung: Daten von insgesamt N = 934 Gymnasiasten am Ende der Klassenstufe 10 und Mitte der Klassenstufe 12 wurden zur Vorhersage des Kurswahlverhaltens in der Oberstufe ausgewertet. Als Fach wurde Mathematik gewahlt. Basierend auf dem erweiterten Erwartungs-mal-Wert-Modell von Eccles (1983, 1994) wurden Fachleistungen (standardisierte Tests), Noten, das Fahigkeitsselbstkonzept und das Interesse an Mathematik am Ende der Klassenstufe 10 erhoben, die Kurswahlen dann in der 12. Klasse. Logistische Regressionsanalysen belegen die pradiktive Kraft von Noten, Interesse und Selbstkonzept im Fach Mathematik fur die Leistungskurswahl, wobei die gefundenen Regressionsgewichte fur Jungen und Madchen identisch sind. Es zeigt sich weiterhin, das die haufigeren Leistungskurswahlen der Jungen in Mathematik zu einem erheblichen Teil uber akademische Selbstkonzepte und Interessen vermittelt sind.

An extension to the internal/external frame of reference model to two verbal and numerical domains

Beside interindividual social comparisons, intraindividual dimensional comparisons in which students compare their achievements in one subject with their achievements in other subjects have an impact on their academic self-concepts. The internal/external frame of reference (I/E) model by Marsh (1986) assumes that dimensional comparisons lead to negative paths from achievement in one subject (e.g., math) to self-concept in another subject (e.g., English). In the present study, the I/E model was extended to two verbal domains (German as the native language and English as a foreign language) and two numerical domains (mathematics and physics). Grades and domain-specific academic self-concepts ofN=1440 students from 63 classes were assessed. In support of the extended I/E model, (a) math, physics, German, and English achievement were positively correlated, as were; (b) self-conceptswithin the verbal and numerical domains, while; (c) self-conceptsbetween the verbal and the numerical domains were almost uncorrelated; (d) positive paths were received from math, physics, German, and English achievement on the corresponding self-concepts; (e) negative paths were found from achievement in one domain to self-concept in the other; (f) positive paths were found from math (physics) achievement to physics (math) self-concept. Finally, (g), almost no effects were found within the verbal domain, i.e., from English (German) achievement to German (English) self-concept. Therefore, there is some support for the I/E model between domains; whereas effects from achievements on self-concepts within the domains were not negative.RésuméContrairement aux comparaisons sociales, les comparaisons dimensionnelles dans lesquelles les élèves comparent leurs résultats dans un sujet avec les résultats dans d’autres sujets ont un impact sur leur concepts de soi académiques.Le «Modèle du système de référence intérieur/extérieur (I/E)» de Marsh (1986) suppose que les comparaisons dimensionnelles causent des sentiers négatifs de la réussite dans un sujet (par exemple, des mathématiques) à la concept de soi dans un autre sujet (par exemple, anglais). Dans la présente recherche, le modèle d’I/E a été élargi à deux domaines verbaux (l’allemand — la langue maternelle — et l’anglais) et à deux domaines numériques (les mathématiques et la physique). Les notes et les concepts de soi académiques qui sont typiques pour un domaine deN=1440 élèves ont été évalués. Grâce au modèle d’I/E élargi, (a) les résultats dans des mathématiques, la physique, l’allemand et l’anglais ont été corrélés de façon aussi positive que les concepts de soi; (b) dans les domaines verbaux et numériques, tandis que les concepts de soi; (c) entre les domaines verbaux et numériques ont été presque non corrélés; (d) un sentier positif a été reçu concernant la réussite dans des mathématiques, le physique, l’allemand et l’anglais et leurs concepts de soi correspondants; (e) les sentiers négatifs ont été trouvés de la réussite dans un domaine au concept de soi dans l’autre; (f) les sentiers positifs ont été trouvés de la réussite dans des mathématiques (la physique) à concept de soi dans la physique (des mathématiques). Finalement, (g), presque aucuns effets n’ont été trouvés dans le domaine verbal, c ’est-à-dire, de réussite en anglais (en allemande) au concept de soi en allemand (en anglais).

Der Einfluß der Leistungsstärke von Schulen auf das fachspezifische Selbstkonzept der Begabung und das Interesse

Zusammenfassung. Daten von N = 1939 Gymnasiasten/innen im letzten Jahr der Sekundarstufe II, die an der dritten internationalen Mathematik- und Naturwissenschaftsstudie (TIMSS) teilgenommen hatten, wurden zur Uberprufung des big-fish-little-pond-effects (BFLPE) ausgewertet. Der BFLPE beschreibt das Phanomen, das sich das mittlere Leistungsniveau einer Klasse oder Schule bei Konstanthaltung der individuellen Leistungsfahigkeit negativ auf das Begabungsselbstkonzept auswirkt (Marsh, 1986, 1990). In der vorliegenden Untersuchung wurde zusatzlich zur individuellen und mittleren Mathematik-Testleistung das Kursniveau (Grund- vs. Leistungskurs) als unabhangige Variable erhoben. Auf seiten der abhangigen Variablen wurde neben dem fachspezifischen Begabungskonzept auch das Interesse an Mathematik als Indikator fur intrinsische Lernmotivation erhoben. Mehrebenenanalysen auf der Basis des hierarchical linear modelings (Bryk & Raudenbush, 1992) replizieren den BFLPE fur das Selbstkonzept. Es zeigt sich auch ein BFLP...

The reciprocal internal/external frame of reference model using grades and test scores.

BACKGROUNDnThe reciprocal I/E model (RI/EM) combines the internal/external frame of reference model (I/EM) with the reciprocal effects model (REM). The RI/EM extends the I/EM longitudinally and the REM across domains. The model predicts that, within domains, mathematics and verbal achievement (VACH) and academic self-concept have positive effects on subsequent mathematics and VACH and academic self-concept within domains but have negative effects across domains.nnnAIMSnThe main purpose is to validate the RI/EM and extend it using objective achievement indicators and grades.nnnSAMPLEnTwo waves of data collection from grade 5 to grade 9 with N = 1,045 secondary school students were used.nnnMETHODSnTest scores, grades, and self-concept data were obtained. The main analyses were conducted using hierarchical linear modelling.nnnRESULTSnThe positive longitudinal effects of grades and test scores on subsequent grades, test scores and academic self-concept within domains and the negative effects of grades and test scores on subsequent academic self-concept across domains supported the RI/EM. The effects of academic self-concept on subsequent grades and test scores across domains were near zero when prior achievement indicators were controlled for. Overall, the results using school grades as achievement measures were replicated using standardized achievement test scores.nnnCONCLUSIONSnThe results serve to highlight the importance of the combination of common theories, which are mostly investigated individually, to enhance our understanding of the complexity of within- and across-domain relations between academic self-concepts and achievement using grades as well as test scores.

Stability of Diagnoses in Various Psychiatric Disorders: A Study of Long-Term Course

The frequency of diagnostic shift was investigated in 267 patients diagnosed with various psychiatric disorders according to the 9th revision of the International Classification of Diseases (ICD-9). Forty-six patients fulfilled the diagnostic criteria for schizophrenic psychosis, 71 for affective psychosis, 66 for neurotic disorder, 24 for personality disorder, and 40 for psychogenic reaction. The remaining 20 patients were diagnosed with heterogenous disorders. The mean follow-up time was 12.5 years. Every episode of inpatient treatment was diagnosed cross-sectionally. Patients with an initial episode of schizophrenic psychosis showed by far the greatest stability (93%). A high stability was found in patients diagnosed with neurotic disorders (79%). Forty-two patients with an index diagnosis of affective psychosis suffered more than one type of episode during the course of disease. The lowest stability was noted in patients with a psychogenic reaction (10%). Eleven percent of patients with a final diagnosis of schizophrenia had an index diagnosis of psychogenic reaction. Our findings demonstrate that the course of disease should be considered in making the final diagnosis.