Students' general English proficiency prior to CLIL. Empirical evidence for substantial differences between prospective CLIL and non-CLIL students in Germany.

Autor/in	Rumlich, Dominik
Titel	Students' general English proficiency prior to CLIL. Empirical evidence for substantial differences between prospective CLIL and non-CLIL students in Germany.
Quelle	Aus: Breidbach, Stephan (Hrsg.); Viebrock, Britta (Hrsg.): Content and language integrated learning (CLIL) in Europe. Research perspectives on policy and practice. Frankfurt am Main: P. Lang (2013) S. 181-201Verfügbarkeit
Beigaben	grafische Darstellungen
Sprache	englisch
Dokumenttyp	gedruckt; Sammelwerksbeitrag
ISBN	978-3-631-64400-3
Schlagwörter	Wissen; Erziehungsziel; Empirische Forschung; Empirische Untersuchung; Quasi-Experiment; Fremdsprachentest; Gymnasium; Lehrer; Lehramtsstudent; Lehrerausbildung; Experiment; Sprachstandsforschung; Unterrichtssprache; Bilingualer Unterricht; Mehrsprachigkeit; Englisch; Forschungsstand; Deutschland; Nordrhein-Westfalen
Abstract	DENOCS (Development of North Rhine-Westphalian CLIL Students) is a large-scale longitudinal, quasi-experimental study aimed at carrying out a detailed evaluation of the effects of CLIL over a time span of two years. The article at hand reports on findings from the first phase of the project and focuses on the general English proficiency of n=968 sixth-graders (M age=l 1.87) in 47 classes. For the main group of Gymnasium students, a Kruskal-Wallis test on the results of two C-tests shows that future CLIL students perform significantly better than both students from non-CLIL strands at the same school as well as students from schools where no CLIL is offered (H(2)=8.66, p(.000). These results are confirmed by Mann-Whitney U-tests for the individual group comparison; they illustrate that all of the groups are significantly different from one another with the CLIL students at the top, the regular students from non-CLIL schools in the middle and the non-CLIL students from the CLIL schools at the end of the spectrum; the effect sizes range from large to medium (1.20 ( Cohen's d ( 54). Likewise, the divergences among students at Realschule are highly statistically significant (N=88, U=233, z-5.82, p(.000) and even larger in size (N=88, d=1.7). The initial findings of this study corroborate the results of previous studies on the issues in question on a larger scale; furthermore, the data of a sub-sample of N=l 10 students, who were divided into CLIL and non-CLIL students only after the preparatory phase, indicates that teacher effects on students' performance are quite small and that individual learner characteristics play a considerable role in students' (foreign-language) development. These indications will be followed-up in the prospective stages of the project and, for the first time, allow a comprehensive evaluation of the emerging differences between CLIL and non-CLIL students. The overall findings entail that in evaluations of (German) CLIL programmes it is important to take into account the head start of the CLIL students in order to obtain a more realistic estimate of the benefits of CLIL. (DIPF/Orig.).
Erfasst von	DIPF \| Leibniz-Institut für Bildungsforschung und Bildungsinformation, Frankfurt am Main
Update	2015/2