The purpose of this study is to examine how the teaching of non-fiction texts looks like in grade 3 in three different classes and how the students in these classes are being prepared for the subsample that tests the ability to write a non-fiction text in the national test in Swedish and Swedish as a second language. I have, based on the purpose, issued three question that form the basis of this study. They are: What tasks does the students in these classes implement in Swedish lessons dealing with non-fiction? What models are used in Swedish teaching in these three classes dealing with non-fiction? How are the students being prepared for the subsample that tests the ability to write a non-fiction text in the national test in Swedish and Swedish as a second language? The methods used to carry out this study are observation and interview. The purpose of using these methods are to gain knowledge and to be able to explain a social phenomenon, in this case how the teaching of non-fiction texts looks like in grade 3. The theoretical concepts that are relevant for the implementation of this study are the writing discourses of Genre and Process. The result is that all of the teachers that featured in this study prepared their student for the national test in Swedish by implementing similar tasks in their teaching, which the students face in the national test. The students read non-fiction texts, selected relevant words from the texts and wrote their own non-fiction text during the lessons I observed. I detected two separate models used in the teaching, which are Genre pedagogy and Process writing. None of the teachers in this study used the models fully, but used some parts of them in their teaching.

The main purpose of this study is to analyze texts written by pupils during the national test in the subject Swedish for grade 9 in order to identify and illustrate how they have met the instructions for the writing task including genre. The texts have been analyzed using Palmér & Östlund-Stjärnegårdh’s (2005) model for student text analysis. This model has been supplemented by studying word and sentence length, as well as indexes for level of readability (LIX) and word variation (OVIX) in order to illustrate possible differences in a gender perspective. The results of this study are in line with previous research, as the girls’ texts are longer and more descriptive, while the boys’ texts are shorter and show a wider range of word variation. The pupils have succeeded in fulfilling the genre and have to a large extent met the task instructions, producing communicative and engaging writing. However, the results of the analysis would have further improved and developed the texts.

The Swedish debate regarding the reform of the third person pronoun plural has been examined with the NLS theory and with the perspective of literacy as not only a mental process but also as a social practice. The debate was initiated in autumn of 2016 by an article written by the teacher Henrik Birkebo in which he argued for a reform. To examine the legitimacy of the main arguments in the debate in relation to pupils’ actual use of the Swedish alternatives for the third person pronoun plural de/dem and dom in secondary school, 307 answers to the National Tests in Swedish and Swedish as a Second Language from 2015 has been quantified in a corpus. The result shows that the argument for a reform due to the hardship for pupils to make out the difference between de and dem sees literacy mainly as a mental process. This is not shown in the corpus. A minority cannot use de and dem as the norm requires, and this minority consists of pupils with high and low grades. The argument to not too hastily make a reform sees literacy as a social practice and as a performance in a social context. This is more likely to be true according to the results of the analysis of the corpus. 82 % of the pupils want to use de/dem which shows that this is seen as a more respected literacy than the use of dom, which only 6 % of the pupils use in the corpus.