Children’s Visual Scanning of Textual Documents: Effects of Document Organization,Search Goals,and Metatextual Knowledge

This study examines children’s strategies when scanning a document to answer a specific question. More specifically, we wanted to know whether they make use of organizers (i.e., headings) when searching and whether strategic search is related to their knowledge of reading strategies. Twenty-six French fifth graders were asked to search single-page documents presented on the screen of an eye tracker in order to respond to questions requiring the location of information in either a single paragraph (Location questions) or multiple paragraphs (Comparison questions). Location questions were easier and faster to answer than Comparison questions. The presence of headers led to more selective reading strategies but did not significantly speed up the search. Strong individual differences were observed in children’s scanning strategies: Some systematically fixated headers, whereas others did not. These differences were not significantly related to the participants’ decoding or comprehension skills but rather to their knowledge of reading strategies.     

Objectives, agreements and matching in science-industry collaborations: reassembling the pieces of the puzzle

Science-industry collaborations have been the subject of a considerable attention in the last few years. The paper argues that, however, the existing studies are still restricted to a partial view of the phenomenon. In this respect, our study departs from the specialised literature by taking into consideration, for a given set of heterogeneous collaborations, information from both the academics’ and the firms’ side, relative to their characteristics, their aims and the collaboration settings. To do so, we constructed a database of 46 collaborations from original data we collected, in several European countries and in the US, within the SESI-TSER project. The first outcome of our study is a typology of science-industry collaborations built on a formal procedure (a multi-correspondence analysis followed by an ascendant hierarchical classification) exhibiting five coherent types of collaborations that we describe precisely. Highlighting the crucial importance of research agendas, we thus propose a simple rationale for the matching process. The latter leads us to propose an “assortative” matching hypothesis generating two polar configurations which we test and confirm. Finally, we examine the policy implications of the findings.     

How can students’ academic performance in statistics be improved? Testing the influence of social and temporal-self comparison feedback in a web-based training environment

The aim of this study was to examine how social or temporal-self comparison feedback, delivered in real-time in a web-based training environment, could influence the academic performance of students in a statistics examination. First-year psychology students were given the opportunity to train for a statistics examination during a semester by doing online exercises in a web-based training environment. Once connected, students received in real-time either social comparison feedback (their score was compared to the mean score of all first-year students) or temporal-self comparison feedback (their score was compared, week by week, with their own previous score). The fact that students were free to connect to the web-based training environment heightened self-regulation differences such as academic procrastination, which was considered as a moderating variable in this study. Because academic performance was measured, the students’ background in mathematics and statistics was also controlled. Irrespective of the students’ background, the results reveal a positive influence of social comparison feedback on statistics exam performance, but only among students who did not delay doing exercises in the web-based training environment. By contrast, temporal-self comparison feedback did not have any effect on performance. Some recommendations for optimizing the efficacy of web-based training environments can be proposed, taking into account both social comparison feedback and academic procrastination.     

DEVELOPMENT, VALIDATION, AND USE OF A GREEK-LANGUAGE QUESTIONNAIRE FOR ASSESSING LEARNING ENVIRONMENTS IN GRADE 10 CHEMISTRY CLASSES

This study describes the development and validation of a Greek-language instrument that can be used to assess grade 10 students' perceptions of their chemistry classroom environment as a means of showing differences between chemistry learning environments in Greece (Attica) and Cyprus. The development of the instrument was based on available learning environment questionnaires. The questionnaire was administered to 1,394 students from 49 chemistry classes in Attica, and the resulting data were analyzed to explore the reliability and the validity of the new instrument. The validated questionnaire was administered to 225 students from 15 classes in urban areas of Cyprus. The data analyses supported the questionnaire's internal consistency, discriminant validity, and ability to differentiate between classrooms. Effect sizes and independent samples t test analyses revealed differences between the two samples. Cypriot students viewed their chemistry classroom environment more favorably than did the Attica students. A possible cause for this difference could be the knowledge-centered aspect of the grade 10 chemistry curriculum in Greece compared to the corresponding tool-instrumental knowledge curriculum in Cyprus.     

Randomization‐based statistical inference: A resampling and simulation infrastructure

Statistical inference involves drawing scientifically‐based conclusions describing natural processes or observable phenomena from datasets with intrinsic random variation. We designed, implemented, and validated a new portable randomization‐based statistical inference infrastructure ( http://socr.umich.edu/HTML5/Resampling_Webapp ) that blends research‐driven data analytics and interactive learning, and provides a backend computational library for managing large amounts of simulated or user‐provided data. We designed, implemented and validated a new portable randomization‐based statistical inference infrastructure ( http://socr.umich.edu/HTML5/Resampling_Webapp ) that blends research‐driven data analytics and interactive learning, and provides a backend computational library for managing large amounts of simulated or user‐provided data. The core of this framework is a modern randomization webapp, which may be invoked on any device supporting a JavaScript‐enabled web browser. We demonstrate the use of these resources to analyse proportion, mean and other statistics using simulated (virtual experiments) and observed (e.g. Acute Myocardial Infarction, Job Rankings) data. Finally, we draw parallels between parametric inference methods and their distribution‐free alternatives. The Randomization and Resampling webapp can be used for data analytics, as well as for formal, in‐class and informal, out‐of‐the‐classroom learning and teaching of different scientific concepts. Such concepts include sampling, random variation, computational statistical inference and data‐driven analytics. The entire scientific community may utilize, test, expand, modify or embed these resources (data, source‐code, learning activity, webapp) without any restrictions.