Diagnosing Students' Meta-Modelling Knowledge. Gathering Validity Evidence during Test Development.

Autor/in	Gogolin, Sarah
Titel	Diagnosing Students' Meta-Modelling Knowledge. Gathering Validity Evidence during Test Development.
Quelle	Berlin: Freie Universität Berlin (2017), 311 S. PDF als Volltext kostenfreie Datei (1); PDF als Volltext kostenfreie Datei (2) Link als defekt meldenVerfügbarkeit Dissertation, Freie Universität Berlin, 2017.
Sprache	englisch
Dokumenttyp	online; Monographie
URN	urn:nbn:de:kobv:188-fudissthesis000000105919-9
Schlagwörter	Diagnose; Testentwicklung; Biologie; Diagnose; Dissertation; Modellieren; Schüler; Student; Studentin; Schülerin; Gültigkeit; Schüler; Schülerin; Modellbildung; Biologie; Modellieren; Dissertation; Analogiemodell; Gültigkeit; Student; Studentin
Abstract	Models, on one hand, aid students in their understanding, on the other, enable researchers to investigate biological phenomena (Oh & Oh, 2011). From an educational point of view, the process of thinking in and about models as a form of scientific practice can be seen as one of the essential learning goals for science students (e. g., Germany: KMK, 2005; USA: NGSS Lead States, 2013). Existing instruments (e. g., Grosslight et al., 1991; Grünkorn, 2014) that are used to assess students' understanding of models are suitable for educational research but, due to their complexity, cannot be employed by teachers when seeking to obtain direct feedback on their students. In the present research project, we developed a forced choice task diagnostic instrument based on the 'model of model competence' (Krell et al., 2016), which was to allow for a valid and efficient diagnosis of students' meta-modelling knowledge in the aspects 'nature of models' and 'purpose of models'. In order to provide evidence for the validity of the proposed score interpretation with the instrument, the constructed forced choice tasks were put to the test within the first global research question (RQGI): In what way do evidence and theory support the interpretation of the test scores for the intended use of diagnosing students' meta-modelling knowledge? We gathered evidence for validity based on test content, on students' response processes, on relations to other variables and on internal structure (AERA et al., 2014). Most evidence supports the valid interpretation of the proposed scores of meta-modelling knowledge. Nevertheless, we have to advise that score interpretation is sufficiently supported only for grades ten to twelve. Furthermore, there is a demand for more research concerning construct-irrelevant variance caused by information in the task stem as well as construct underrepresentation due to a possible lack of differenciation across contexts. The second global research question was set to provide information about students' meta-modelling knowledge (RQGII): How frequently are the three levels of understanding within the aspects 'nature of models' and 'purpose of models' represented among students? For the aspect 'nature of models', the students mainly express perspectives on level II indicating that they understand models as idealised representations of an original according to the 'model of model competence' (Krell et al., 2016). For the aspect 'purpose of models', on the other side, the students mainly express perspectives on level I indicating that they understand the purpose of models to be the description of phenomena. In addition to these results, we assume students' meta-modelling knowledge to be context-specific as the frequency of the preferred levels altered across contexts. (Orig.).
Erfasst von	Deutsche Nationalbibliothek, Frankfurt am Main
Update	2018/2