In addition to the other data collection methods, I chose to analyse the modular examination results for a group of Year 10 students, one group (n=29) taught using SOLO methods, and a larger group (n=82) who were not.
To try to ensure that, as much as possible, the groups were comparable, I chose students who had started school with either a level 4 or 5 from their Key Stage 2 English tests. The students were all from one of the two parallel top sets. Students who did not have a Key Stage 2 level were excluded from the analysis.
All students in the groups selected took GCSE English Literature exam module A663 (Prose from Different Cultures) in the summer of Year 10 and studied the same text – Of Mice and Men. The results were analysed using descriptive statistics to gain an overview and identify areas where the data warranted a closer look. In areas which appeared to show a difference, a chi-square test was applied to test significance; a significance threshold level of p<0.5 was set to ensure that any significance was meaningful.
Although every attempt was made to make the analysis as unbiased as possible, for example choosing to focus on a module which was externally marked, it is important to remember that looking:
At results before and after a new intervention is rolled out…can be very misleading, as other factors may have changed at the same time. (Goldacre, 2013:9)
In addition, as the group selection was not randomized, or carried out over a longer period of time, the results may not be replicable, although I feel that they may provide some indications for areas which would be worth investigating further.
The non-SOLO group shows normal distribution with a modal grade of a B. The SOLO group also shows a relatively normal distribution, which is less steep than the non-SOLO group. The tail at the higher end of the SOLO group does not drop off to the extreme of the non-SOLO group. The modal grade for the SOLO group is one grade higher, an A grade.
Comparing the results of the combined level 4 and 5 students between the SOLO and Non-SOLO groups, using a chi-square test, suggests that the probability of this distribution happening by chance is 4%; a significant result. These results suggest that, based on this limited study, SOLO may have a positive impact on exam achievement.
To identify whether this impact can be pinpointed, the data was explored in subsets according to level and gender.
Exploring the subset data, it is apparent that, in this sample, the SOLO group females did not achieve significantly different grades to the non-SOLO group.
However, it is when comparing the achievement of students starting school on a level 5 and male groups that a very highly significant difference is evident.
Comparing the results of the students entering school with a level 5, using a chi-square test, suggests that the probability of this distribution happening by chance is 0%; a very highly significant result.
As suggested by the graph there is a significant difference between the achievements of the two male groups. Looking at this distribution, using a chi-square test suggests that this distribution (p=0.0000) is very highly significant.
Although there are limitations to this particular aspect of the study, the results suggest that SOLO techniques may have a measurable impact on student exam results. Therefore, it would certainly be worth further, structured research.