Girls do better than boys?
My results do support my hypothesis that girls do better than boys in KS3; this is shown when I calculated the mean of the mid-point multiplied by the frequency divided the cumulative frequency of all the pupils in the sample I took. The mean score of the girls was 16.267 and the mean score for the boys was 16.217 I also calculated the mode to find the most common, the girls had a score of 17 and the boys a score of 15, this was also a very strong support for my hypothesis.
Additionally I drew out two box plots that I was able to draw out from my cumulative frequency diagram, and from this I was able to calculate the median and this showed a support for my hypothesis with the median for the girls being higher than the boys, 16.5 for the girls and 16 for the boys. Although these calculations were a strong support the inter quartile range didn’t back-up my hypothesis with the boys score ranging from the lowest score out of the sample taken to the highest score form the sample, this gave the result of the inter quartile range for the girls being 4.5 and for the boys 5.5.
I drew out a few diagrams such as the box plot (already mentioned), a cumulative frequency diagram and a histogram.
The C/F diagram showed very close scores in between the girls and the boys, this was shown by the way the ‘s’ shaped lines overlapped each other with no real anomaly results showing clearly. The histogram yet again has shown a positive result in connection with my hypothesis; there was a constructive amount of scores between the marks 17- 20 where the frequency density was higher than 5 and reached all the way to 11, on the other hand the boys ranged from a frequency density of 2 to 6 in between the marks 17- 20.
Market Bosworth do better than Winstanley?
My results do support my hypothesis that Market Bosworth has better KS3 results than Winstanley; this has been shown when I calculated the mean of the mid-point multiplied by the frequency divided by the number of pupils taken in the sample I took. Bosworth had shown to have a higher mean than Winstanley, Bosworth with a mean of 17.067 and Winstanley with a mean of 14.5. I also calculated the mode of the two schools and these also supported my hypothesis strongly, Bosworth with a mode of 17 and Winstanley 13 to 14.
Furthermore I drew out a box plot that I created from my cumulative frequency diagram and yet again my hypothesis was supported by the out come of the results, my box plots enabled me to find the median and the inter quartile range for both schools. The median was 17.5 for Bosworth and 14.5 for Winstanley and the inter quartile range that I calculated from the diagram by taking away the upper quartile from the lower quartile and I got the results, Bosworth 5 and Winstanley 3.5.
I also drew out the diagrams as followed: box plot (as already mentioned), cumulative frequency and histograms.
The C/F lines that I drew out on my graph didn’t overlap but grew further apart in the middle range of results and slowly closed up to open out towards the end, this shows that the results weren’t close and there were no outstanding anomaly results to be seen. My histograms once more has shown a optimistic look upon my hypothesis, with Bosworth’s marks ranging from 6 to 23 and Winstanley’s ranging from 6 to 19 there was a significant amount of higher marks from Bosworth.
The highest frequency density is Bosworth with a frequency density of 10 at 17 marks and for Winstanley it’s a frequency density of 9 at 16 marks.
Conclusion of the two hypotheses.
Overall my research went well, the only thing I found a challenge was counting up the sample of pupils I took because it was difficult to display the data in a group frequency table.
The only real problem I came across was tallying up the marks to put in the group frequency table.
If I were to do this again I would have used a stratified sample to make it fairer and to give a fair and more accurate out come overall.
Also we used secondary evidence, so next time I would used primary, this way I know for certain that nobody had changed any of the marks. I also might used pupils actual score from the tests they did instead of using the level they got because they might have a really high 6 or a really low 6 for example and by using the actual marks from the tests the results could have been very much more accurate.
I think I would have thought more carefully about how I would deal with marks classed as ‘B’,’N’ and ‘E’ to make it simpler or I could even leave them out.
I could also improve my method by using a larger sample to give a wider range of results with yet again a better out come of more accurate result being produced at the end.