Writing skills support in multilingual secondary schools. The effect of profiled revision tasks on the written production of 6th grade students in the L1s German and Turkish and the L2 German

Introduction

Despite the importance of literacy education, especially for students raised speaking heritage languages in addition to German, little is known about writing competencies in both languages of students as they begin secondary school in Germany. Drawing on insights from multilingualism and literacy research, the goal of SimO was to understand better how different writing settings in majority language German classes can support writing skills in the first language German, in the second language German and – for students participating in Turkish heritage language classes – in the Turkish language. The collaborative project thus examined both the effects of the differently profiled writing settings on writing skills in German and the potential for interlingual transfer into the heritage language Turkish.

What was investigated and how?

To this end, 322 6th grade students in 15 classes from three different schools participated in an intervention study. Each student took part in one of four different writing settings over the course of one month. The writing settings involved either the presentation of (1) topic knowledge (this became the control group), (2) topic knowledge and task schemata (language functions), (3) topic knowledge and language-dependent text schemata (language forms), or (4) topic knowledge, task schemata, and language-dependent text schemata.

The study extended over five months, and included a pretest, four intervention tasks, and two follow-up tests in German, as well as seven control tasks in Turkish. In all cases the writing arrangement consisted of a revision task in which students were asked to improve on a poorly written description of a superhero or super villain. All tasks had the same structure but used different characters to prevent students from reiterating previous texts. Students composed a total of seven different character descriptions in German class and – for those who took part in Turkish classes – seven subsequent descriptions during the same weeks in Turkish.

The analysed data included all 2166 German texts of the 322 participants , as well as 607 Turkish texts written by the 91 students who also took part in Turkish class. Supplementary data were gathered on students’ reading abilities in German (using the standardised FLVT test) and in Turkish (using an adaptation of the TELC test for Turkish), students’ classroom grades, and diverse individual information, including reading and writing preferences, interests, and bilingual and biliteral competencies. The written texts were analysed according to three measures: (1) text length (number of orthographic words), (2) analytic rating of text quality, developed specifically for the SimO project, and (3) holistic rating of text quality.

Results

Results showed that, first, there were no differences between students who speak solely German at home with their parents, students who speak mostly a heritage language at home, and students who speak a combination of both. Thus, earlier studies showing differences between these groups were not supported by the SimO study.

Second, intralingual intervention effects were evident. Students participating in writing tasks involving task schemata (groups two and four above) profited most from the intervention, whilst students receiving only topic knowledge support or form-focused support without schematic information also improved, but not as much as those who received task schemata information.

Third, interlingual intervention effects were also evident, provided students profited from the intervention in German and participated in interventions which focused on task schemata only (group two above). Thus, even those students who profited in German from a form-based intervention, or from a schema and form-based intervention could not transfer this knowledge to Turkish, whilst students who profited from an intervention involving (only) language functions could, and subsequently did, produce better texts.

What does this mean for educational practice?

The study showed that students’ writing can be improved through language-focused instruction, especially when information on language function  is included in the writing tasks; simply providing language forms, however, does not result in a marked increase in text quality. Furthermore, the concentration on language function has an added, interlingual benefit: students who focus on task schemata not only benefit in the focus language, but can also transfer this newly gained knowledge to another language, even without further intervention.