Conference:
ECER 2008
Format:
Paper
Session Information
PRE_MOCK, Mock Viva (Part 1)
Mock Viva
Time:
2008-09-09
14:00-15:00
Room:
C1 22 Margaretha Huitfeldts Auditorium
Chair:
Shosh Leshem
Discussant:
Vernon Trafford
Contribution
ntroduction
The overall aim of this thesis is to explore the dimensionality of grades in order to better understand and explain what grades actually measure. The starting point has been research which has indicated the existence of a variety of interpretations of grades; that grades are interpreted as a measure influenced by student cognitive as well as non-cognitive factors. A substantial body of research has indicated that grades are influenced by subjectivity and that factors other than achievement such as different student characteristics, teachers’ grading practices and systematic differences within and between schools, exert an influence on the awarding of grades (Brookhart, 1991, 1993, 1994; Cliffordson, in press; Hidi et al. 2004; Pilcher, 1994; Wentzel, 1991). Research has also shown that gender and family background exert an influence on grades (Rosén, 1998; Murphy, 2000).
However, the educational and grading system presumes that grades are a reliable and valid measure of student academic knowledge and that grades are comparable over schools and time, which legitimate grades to function as an instrument for selection to the next level in the educational system. This puts high demands on the grading system to secure the legal rights of individuals.
The purposes have been:
• To investigate the dimensionality and structure of grades in order to understand, but also attempt to explain what grades actually measure.
• To explore systematic differences in grades within and between schools
• To explain differences in grades related to gender and parental educational
Method
Theory and method
Three large-scale studies have been conducted which each highlights and focus on different aspects concerning the validity of grades. The conceptual framework is foremost concerning theories of validity issues such as the unified validity approach argued by Messick (1989) and the interpretive and the validity arguments as Kane discusses (2006). The relation between cognitive and non-cognitive factors such as students’ attitudes, interest, motivation, family values and gender seem to be of importance for understanding achievement and grade differences (Gustafsson & Balke, 1993). Grades are a summarization, a score based upon both quantitative and qualitative judgements of different performances for example on tests and performances in the classroom.
Data is based upon one dataset which comes from The Gothenburg Educational Longitudinal Database (GOLD) which contains register and questionnaire data. In all, 99,070 students born in 1987 and who left compulsory school in 2003 were included in the analysis. In sum, 1246 schools were included.
In order to investigate and explain variance in grades, multivariate multi-level structural equation modelling (SEM) has been used. The advantage of SEM modelling is that this technique admits for relations between underlying constructs or latent variables. The choice of multi-level technique was due to the fact that phenomena in educational settings have a hierarchical structure where students belong to a classroom which belongs to a school which belongs to a neighbourhood (Gustafsson & Stahl, 2005). Another important advantage with multi-level analysis is the capacity to include incomplete cases in the analysis (Hox, 2002).
Expected Outcomes
Result and conclusions
The result from the analysis show that it is possible to separate the structure of grades into two dimensions; a subject-specific dimension and a common grade dimension. The subject-specific dimension was related to national tests and subject grades in each subject (Swedish, English and mathematics) and hypothesized to reflect cognitive abilities while the common grade dimension was related only to the subject grades in Swedish, English and mathematics and hypothesized to reflect variance in grades common for the subjects. The subject-specific dimension explain a large amount of variance in grades but the common grade dimension, which cuts across the different subjects (Swedish, English and mathematics) also explained variance in grades in all three subjects. The result shows that there are gender differences in the common grade dimension of .147 on the individual level which means that girls are favoured in the grade setting practice in terms of that their grades are more influenced by non-cognitive aspects. On the school level, parental educational background explain a substantial part of the variance in the subject-specific dimension while parental educational background has a significant and substantial negative relation to the common grade dimension. One explanation may be that on schools with a large proportion of parents with lower education, the students’ benefits in the common grade dimension and grades which may be due to a compensatory grading practice.
Student characteristics such as general interest for learning, adjustment in school, parental engagement have also been shown to be a non-cognitive characteristic of importance for grades. The gender effect of .147 in the common grade dimension is totally mediated by student general interest in learning which means that girls seem to be more interested in school work and therefore receive relatively higher grades.
On the school level, the result shows that there are systematic differences in grades between schools in that the influence of non-cognitive factors on grades are higher for schools with a large proportion of students with parents with lower education and for schools located in the suburbs while the teacher-staff characteristics such as average age, experiences and certification do not seem to influence the common grade dimension in grades.
References
Search the ECER Programme
- Search for keywords and phrases in "Text Search"
- Restrict in which part of the abstracts to search in "Where to search"
- Search for authors and in the respective field.
- For planning your conference attendance you may want to use the conference app, which will be issued some weeks before the conference
- If you are a session chair, best look up your chairing duties in the conference system (Conftool) or the app.