Unit for research and development
THE SOCIAL SERVICES ADMINISTRATION
SE - 106 64 Stockholm, Sweden
PH.D Ann-Christin Kjellman
Paper presented at the European Conference on Educational Research, University of Hamburg, 17-20 September 2003.
Analysis of grade scores from the 8th grade spring semester 2002 in the eight middle schools in Stockholm and 8 control schools.
Sub- Study 2
Projekt: Experimental Educational Program in mathematics and science at Eight (8) Public Schools in Stockholm, from 7th up to 9th grade.
The following study is an analysis of grade scores from the 8th grade spring semester 2002 in the eight middle schools in Stockholm that are participating in the ‘Experimental Educational Program In Mathematics and Science from 7th to 9th Grade’ and 8 control schools, totaling 1780 students. Comparisons have been made between: the eight advanced Math/Science classes, between the control classes and other classes, between classes in the same school, and between boys and girls.
Main results
-The grade average differences between the advanced Math/Science classes in the eight Math/Science Schools are relatively large.
-In most cases, to attend an advanced Math/Science class seems to have a positive effect on scores in Math and Science, independent of which of the eight Math/Science schools one attends, compared to other classes at the same school and the control schools
-There is a relatively equal high grade-level in Science class in all eight schools, while the differences between Math classes in the schools are large.
-A larger part of the students in the Math/Science classes have achieved high grades in Math and Science compared with other classes at the advanced Math/Science schools and with the control schools.
-Girls in the advanced Math/Science classes have, with few exceptions, a higher grade point average in Math and Science than boys.
-Even if the girls in the Math/Science classes have a higher grade point average than the boys in these classes, the differences between the genders are less in the Math/Science classes, compared with control classes and other classes.
-There are few significant connections between how students responded to the questionnaire in grade 8, (why they had chosen Math/Science classes and their experience attending these classes) and their grades.
-The significant connections that do exist between the questionnaire responses and the students’ grades are small.
Main purpose
The main purpose of the following study is to analyze grade scores in grade 8 and to reach an understanding of whether attending advanced Math and Science classes that began Autumn semester 2000 has had a positive effect on grades in Math and Science.
Background
In the sub-report 1[1] it became apparent that the eight testing schools had very different experiences with Math and Science classes when the project began. It was also apparent that the schools had very different preliminary conditions depending on the social and educational composition in the schools’ district[2]. In addition, schools had different acceptance procedures dependent upon the number of applicants. The schools had different goals for their advanced classes and a large difference between the amount of hours that students studied in their subjects. All in all these differences have resulted in difficulties when comparing the schools with each other in regard to grade results.
To be able to say anything about the students’ knowledge development within the advanced subject areas, the best strategy would have been to compare all the students in Stockholm’s elementary schools in grade 8 during the spring semester 2002 with the grades in the eight advanced Math/Science classes. However these grades were not available as data files. Thus, due to limited economic and time resources, we chose 8 additional schools amongst the fifty junior high schools in Stockholm that would make up the control group.
To make an accurate comparison it was necessary to partner each advanced school to a control school that was as similar as possible. The schools that would be included as control schools are Aspuddensskola, Bäckahagenskola, Gärdesskolan, Hagsätraskolan, Husbyskolan, Johan Skytte skolan, Mälarhöjdensskola and Österholmskolan.
The entire student body that was to be examined consisted of 1780 students, 828 students in the test schools and 952 in the control schools. The 404 students that are regarded as ‘missing’ in regard to grade scores in Swedish do not have Swedish as a native language and are therefore enrolled in Swedish as a Second Language (Swedish 2).
In accordance with the evaluation focus, the analysis of the results will primarily be concentrated on Math and Science subjects. A comparison between the schools in Swedish and English is also being done, though the analysis of these subjects is much smaller.[3]
Previous comparisons of average grades in the Math/Science schools and control schools show that the grade averages in the 16 schools are roughly the same. Thus the schools could be used as comparison schools according to their purpose. The percentage of students that have not received a passing grade is comparatively larger, however, in the Math/Science schools than in the control schools. Nearly one-fourth of the students in other classes in the advanced Math/Science schools have not reached a grade of C in Math and Science. This is likely due to the fact that a larger percentage of students in certain advanced Math/Science schools, compared with the control schools, take Swedish 2 instead of Swedish.
Results
The results show that the control classes are more similar to the advanced Math/Science classes in regard to grade scores than the other classes, which show that the eight control schools have found adequate control classes. In both the advanced Math/Science classes, and the control classes, there is a larger percentage of students who have A´s in all four presented subjects and a smaller percentage of students who have not received a C, compared with other classes.
In a comparison between grade averages in the four nuclear subjects it is obvious that the differences between the three groups, the advanced Math/Science classes, the control classes and other classes in the advanced Math/Science schools, are largest in Math and Science. The percentage of students that have received an A in these subjects are largest in the Math/Science classes. In English it is the control classes that have the largest percentage of students with an A. The group of other classes has the smallest percentage of students that received an A in all four nuclear subjects. We can also state that the percentage of students that have not received a C, especially in English, is significantly larger in the other classes than in the Math/Science classes and the control classes. That the differences between the classes are not as large in Swedish as in English is probably due to the large part of students who take Swedish 2 and are therefore not included in the statistics for Swedish.
We also find that the grade average differences between the Math/Science classes in the eight advanced schools are quite significant. The differences are larger in Math than in other subjects. The grades in Science are significantly more even and higher in all the Math/Science classes in the eight advanced Math/Science schools on the whole. That the differences in Science between the schools is not as large as the difference between Math may be due to that there were more schools in the introductory phase that had more experience with Science than with Math.
We also find that in the majority of Math/Science schools there are significant grade differences between the Math/Science classes and other classes at the schools to the advantage of the Math/Science classes.
Even if there are large grade differences between the schools, it seems that in most cases there is a positive effect on the grades scores for students attending an advanced Math/Science class, compared with other classes at the school and the control schools. That the students in the Math/Science classes at the majority of the schools spend more hours studying Math and Science than other students may be a positive factor in this scenario. It is, however, not only in these subjects that the Math/Science classes have a higher grade-average. The study shows that classes with a high grade-average in the Math/Science classes in general also have a higher grade-average in other basic subjects.
These differences between classes within the advanced Math/Science schools cannot be explained by social groupings within the school districts. We also find large grade differences between the advanced Math/Science classes and other classes, not just in more socially vulnerable schools such as Solbergaskolan, but also in schools like Engelbrektskolan, Matteusskolan and Sturebyskolan where the parents in the schools district have a high education level[4]. A more in-depth analysis of individual schools would be required to explain these differences in grade averages between classes within the same school.
Gender differnces
We also find that there is a larger percentage of girls than boys, that have the highest grades. There is also a smaller percentage of girls than boys who have not received a passing grade. We also find that there is a difference between the control classes, the advanced Math/Science classes and the other classes. In the control classes it is the boys who have the highest grade-average, whereas the girls average grade is highest in both the advanced Math/Science classes and the other classes. This pattern raises new questions about work methodology and traditions at the different schools. At this point of the study, however, it is difficult to predict why the grade pattern in the control classes differs from the other classes.
Another interesting difference between the classes is that even if the girls in the Math/Science classes have a higher grade average than the boys in these classes, the differences between the average grades between the genders is large, both in the control classes and the other classes, than in the Math/Science classes. Especially the schools that have the highest grade-average the difference between girls and boys average grades is so small that they are hardly worth mentioning
The results that are most surprising are that girls, compared with boys, have a higher grade-average and a larger percentage of A’s in Science in all the three control groups. Earlier studies have shown that it is in the Science-based subjects that the differences are the largest, to the boys’ advantage[5].
Correlation between grade scores and questionnaire responses
We have also analyzed a possible correlation between grade scores and questionnaire responses from the advanced Math/Science classes in grade 8. We find only six relevant correlations between grades and questionnaire responses, where four are positive and two negative. The correlations we find however are not especially large.
That there is not a significant connection between high grades in Math and the choice of Math class based on interest is somewhat surprising. There is however, connection between a high grade and students who responded that the school has good methodology teaching Math. That the student responded that the school had good teaching methodology in Science has no significant connection with higher grades.
That the student is satisfied with their possibility to develop knowledge at their own pace has a positive connection with their grade scores. We also find a significant connection between high grades and the response that doing well on Math tests and being a high achiever is important. All connections are rather weak.
We also find a significant negative connection between grades and that the student responded that parent opinion had a large effect on the choice of a Math/Science class. This connection is probably indirect. More analysis would be needed to be able to state what the connection is. In sub report 1 it is apparent that students from schools in areas with high immigrant populations have rather low grade-averages compared with other advanced Math/Science classes. To a high Degree immigrant students also stated that parent opinion had a large effect on the choice of a Math/Science class. This may explain the negative connection between grades and parent opinion in choice of Math/Science class. Another possible explanation is that student’s whose choice was not self-motivated, but rather due to parents’ wishes, may not have the motivation required to receive higher grades. There is also a significant negative connection between students choosing an advanced class because they wanted to remain with peers from grade 6, that is students who did not make a self-motivated choice, but rather followed their friends who had interest and high ambitions for school.
A more concrete answer to different possible explanations for the results we found in sub report 2 will hopefully be answered by the next student questionnaire that was sent out to the eight advanced Math/Science classes and the eight control classes in the first week of May 2003. In this questionnaire we ask for example, which program they plan to study in high school and what possibilities they have already had in middle school to read high school Math classes.
Bibliography
Dimenäs, J. (2001) Innehåll och interaktion. Om elevers lärande i naturvetenskaplig undervisning. AKTA UNIVERSITATIS GOTHOBURGENSIS
Eriksson, R (1994) Spelar valet av skola någon roll? Effekter av grundskola och omgivning på övergång till gymnasiet. Ur Erikson, R & Johnsson, J.O (red). Sorteringen i skolan. Stockholm. Carlssons Bokförlag
Kjellman, A-C (2003) Profilklasser I matematik och naturvetenskap. En enkätundersökning med elever i år 8 i Stockholm. Delrapport 1. Individ, omvärld och lärande/Forskning nr 14. htp://
PM in English: Experimental Educational Program in mathematics and science at Eight (8) Public Schools in Stockholm, from 7th up to 9th grade.
Sub- Study 1:
Kjellman, A-C (2003) Jämförelse av betyg i ma/no- klasser och betyg i jämförelseklasser år 8, vårterminen 2002. Delrapport 2. Utbildningsförvaltningens rapportserie 2003. socilatjänstförvaltningen FoU-enheten. 2003:11.
Lindahl, M (2000). Summer Learning and the Effect of Schooling: Evidence from Sweden. Studies of Causal Effects in Empirical Labor Economics. Swedish Institute for Socilal Research. 43. Stockholms Universitet
Ramstedt, K. (1996) Elektriska flickor och mekaniska pojkar. Om gruppskillnader på prov – en metodutveckling och studie av skillnader mellan flickor och pojkar på centrala prov i fysik. Pedagogiska institutionen Umeå universitet.
SOU 1993:85 Ursprung och utbildning. Social snedrekrytering till högre studier. Stockholm. Utbildningsdepartementet
Staberg, E-M (1992) Olika världar, skilda värderingar. Hur flickor och pojkar möter högstadiets fysik, kemi och teknik. Pedagogiska institutionen Umeå universitet.
1
[1] Kjellman, A-C (2003)
[2] Information about social compositions in the schools district builds on statistics from USK 1996
[3] See, Subreport 2 (2003). Comparisons of grades grade 8 Spring semester 02 in Math/Science classes compared with control classes/schools. FoU- unit in Stockholm. 2003:11
[4] In areas with many high educated, the students in general have a higher grade average and the transition to secondary education is more widespread than in areas with less educated people. (SOU 1993:85, s. 270, Eriksson, 1994, s. 151) Lindahl (1999)
[5] Staberg, E-M (1992) Ramstedt (1996, s. 359), Dimenäs, (2000)