No description available.

Hear about lessons learned from a 10-year practice of using exploration curriculum and service learning to advance middle school students’ interest in and achievement of science learning.

No description available.

No description available.

The purpose of this dissertation research was to explore how engagement in action research influenced high school science teachers’ beliefs and classroom practices related to student voice and participation. This included exploring any affordances or obstacles they experienced with attempting to increase student voice and participation in their classrooms. I facilitated individual action research projects with the teachers, and they were able to discuss individual action research plans and share ideas with colleagues. Four high school science teachers began this research. While all four teachers were selected as cases, only two of the teachers designed and followed through with their action research plans. The teachers were provided with a hierarchy of student participation and examples of each of the four levels of the hierarchy in order to guide and design their action research. Collected data included transcripts from the action research group meetings, classroom observations, teacher journals, and interviews with teachers. A combination of direct interpretation and thematic coding was used to analyze the data. Direct interpretation involved writing narratives based on what I saw and interpreting events as I experienced them. The teachers either maintained their positive beliefs or formed new beliefs related to the inclusion of student voice. In addition, both teachers successfully used action research as a way to overcome obstacles and increase student voice and participation in their classrooms. This study has implications for teachers who want to use action research to better understand their teaching related to student voice, administrators and teacher educators who want to provide PD opportunities and support to meet individual teacher needs, and researchers who are interested in studying student voice.

Först ges en beskrivning av undervsiningsfilosofin Investigative Science Learning Environment (ISLE) och research-based assessment instruments (RBAIerna) CLASS och ECLASS, följt av en översättning till svenska av RBAIerna. En analys skedde sedan av hur svaren till påståendena i RBAIerna (pre- vs post-) förväntas ändras om en ISLE-baserad undervisning ges. I analysen försökte jag koppla påståendena till någon eller några av de sex vetenskapliga förmågorna och deras tillhörande bedömningsmatriser, vilka nu är en integrerad komponent i ISLE filosofin, men som ursprungligen utvecklades av Rutgers Physics and Astronomy Education (PAER) group. Jag använde även litteratur om ISLE och textboken "College Physics: Explore and Apply", vilken används i ISLEbaserade kurser, för att analysera påståendena. Resultatet blev att en majoritet av alla påståenden - 72\% av CLASS- och 77\% av ECLASSpåståendena - bör besvaras mer expertlikt efter en ISLE-baserad undervisning än innan. Ett påstående i CLASS, vilket motsvarar 3\% av påståendena i enkäten bedömdes även besvaras mindre expertlikt och resterande kunde jag inte, utifrån min analysmetod förutsäga om de skulle besvaras mindre, lika eller mer expertlikt. Efter analysen och undersökningar av tidigare studier om studenters prestationer på CLASS- och ECLASSenkäterna, formulerades en förutsägelse att en ISLE-baserad undervisning bör leda till att elever svarar mer expertlikt än om de undervisas traditionellt. En pilotstudie av den svenska översättningen genomfördes sedan på elever inom olika utbildningsnivåer mellan gymnasie- och doktorandnivå. Resultatet analyserades och jämfördes med tidigare forskning i andra länder, även om både skillnader och likheter kunde hittas var det svårt att dra några generella slutsatser på grund av få svarande i min studie. / A description of the teaching philosophy Investigative Science Learning Environment (ISLE) and the research-based assessment instruments (RBAIs) CLASS and ECLASS are given, followed by a translation of the RBAIs into Swedish. An analysis is then made of how the answers to the RBAIs (pre- vs. post-) are expected to change if students take an ISLE-based physics course. In the analysis, I tried to connect the statements to one or more of the six scientific abilities and their associated rubrics, which are now a component of the ISLE philosophy but are originally developed by the Rutgers Physics and Astronomy Education (PAER) group. I did also use literature about ISLE and the textbook “College Physics: Explore and Apply” which is used in ISLE based courses, to analyze the statements. I found that that the majority of the statements - 72 \% in CLASS and 77 \% in ECLASS – are expected to be answered in a more expert-like way after an ISLE-based course. One statement in CLASS, which corresponds to 3 \% of the statements in the survey, is expected to be answered in a less expert-like way. For the rest of the statements I could not predict, using my method of analysis, if they would be answered less, equally or more expert-like.  After performing the analysis and examining the existing research literature on student performance on CLASS and ECLASS surveys, a prediction that ISLE-based teaching should lead to more expert-like answers than traditional teaching is formulated. Results from a recent empirical study done in the US, published after my analysis was complete, appear to be in agreement with the prediction that follows from my analysis. A pilot study of the Swedish translations of CLASS and ECLASS was also performed on students at different educational levels, ranging from upper secondary to doctoral level. The results were analyzed and compared to previous research in other countries. Although both differences and similarities could be found, it was not possible to draw any general conclusions due to the small number of respondents in my study.

No description available.

No description available.

This study investigated the effects of grade level and gender on elementary school students&rsquo / motivation towards science learning. This study was carried out during 2007-2008 fall semester at 12 public elementary schools in Yenimahalle districts of Ankara. A total of 2231 students (1121 boys, 1093 girls, 17 did not report their gender) participated in the study. In terms of grade level, 1164 were enrolled in 6th grade level and 1055 in 8th grade (12 did not report their grade level). Data were collected through Students&rsquo / Motivation Towards Science Learning Questionnaire (SMTSL). Two-way Multivariate Analysis of Variance (MANOVA) was conducted in order to identify the effects of grade level and gender on six dimensions of motivation (i.e. self efficacy, active learning strategies, learning environment stimulation, science learning value, achievement goal and performance goal) towards science learning. Two-way MANOVA results showed that grade level and gender had a significant effect on the collective dependent variables. Follow-up univariate analysis indicated that there was a significant difference between 6th and 8th grade students&rsquo / motivation towards science learning regarding mean scores on each motivational variable. In addition, mean scores on each motivational variable, except leaning environment stimulation, were significantly different for boys and girls, in favors of girls. Result demonstrated that students&rsquo / motivation towards science learning declined as the grade level increased and girls had a higher motivation towards science learning than boys.

Nous présentons dans cette thèse une nouvelle méthode pour la conception de moteurs de reconnaissance personnalisables et auto-évolutifs. La contribution majeure de cette thèse consiste à proposer une approche incrémentale pour l'apprentissage de classifieurs basés sur les systèmes d'inférence floue de type Takagi-Sugeno d'ordre 1. Cette approche comprend, d'une part, une adaptation des paramètres linéaires associés aux conclusions des règles en utilisant la méthode des moindres carrés récursive, et, d'autre part, un apprentissage incrémental des prémisses de ces règles afin de modifier les fonctions d'appartenance suivant l'évolution de la densité des données dans l'espace de classification. La méthode proposée, Evolve++, résout les problèmes d'instabilité d'apprentissage incrémental de ce type de systèmes grâce à un paradigme global d'apprentissage où les prémisses et les conclusions sont apprises en synergie et non de façon indépendante. La performance de ce système a été démontrée sur des bancs d'essai connus, en mettant en évidence notamment sa capacité d'apprentissage à la volée de nouvelles classes. Dans le contexte applicatif de la reconnaissance de gestes manuscrits, ce système permet de s'adapter en continue aux styles d'écriture (personnalisation des symboles) et aux nouveaux besoins des utilisateurs (introduction à la volée des nouveaux symboles). Dans ce domaine, une autre contribution a été d'accélérer l'apprentissage de nouveaux symboles par la synthèse automatique de données artificielles. La technique de synthèse repose sur la théorie Sigma-lognormal qui propose un nouvel espace de représentation des tracés manuscrits basé sur un modèle neuromusculaire du mécanisme d'écriture. L'application de déformations sur le profil Sigma-lognormal permet d'obtenir des tracés manuscrits synthétiques qui sont réalistes et proches de la déformation humaine. L'utilisation de ces tracés synthétiques dans notre système accélère l'apprentissage et améliore de façon significative sa performance globale.