I got my grades back from this year, and I’m very interested to look not at the scores, but at the standard deviation of those scores. Out of 100 percent, 3.279 is a very narrow variance, which would show that my weighted mean is a fair representation of my scores.
Blue: Subjects where I scores 63 & 64%
Green: The most fantastic module where we made giant posters and presented them in a poster conference, then had online discussions about the content of said posters, which was graded. Followed by a multiple choice exam.
Orange: A coursework assignment, followed by a final exam, with multiple choice questions and short essay questions.
Yellow: Completely unimaginative final exam, with multiple choice questions and short essay questions.
And it falls into the 2:1 classification category. Good.
This summer, 70% of students across the nation (and my sister is one of them) will graduate with a 2:1 degree or higher. Those who don’t will either be sent to an old poly to do a masters degree, or to do a PGCE, or to work in the HR department at Tesco. Because we’ve created a stigma to go along with our degree classifications, that if you don’t get your 2:1 you’ve basically failed, the universities are doing everything they can to ensure that the 70% statistic holds.
My solid 65.25% could mean that the grading system is an accurate measure, confirmed by test-retest reliability across time. But what exactly is 62.25% is a score of? Recall? How much of a neuro-psychologist I am? Or a developmental psychologist I am? Or a psycholinguist? Or a researcher? Or a computer scientist?
Wait?! Computer scientist? That has very little to do with psychology. Yet my different modules are also quite unrelated to one another, they talk about different paradigms, different topics, different research methods, some more exciting, some less. Yet even so, my grades are reliably steady. I reckon I could get a 2:1 in absolutely anything. I believe in me!
That’s why I’ve used colour coding in the table above, to try connect the different teaching methods and spot trends in the data. And again, there’s a very random spread among the different levels of the independent variable. For the ‘coursework and typical final exam’ subjects, I got 70%, with subsequent scores regressing to the mean, and for the very ordinary ‘typical exam’ subjects I got my lowest score, as well as one in line with the mean. Most interestingly, the subject which I enjoyed the most, Personality and Individual Differences* was right on the mode average and just one percent above the median. Also very interesting is that the two research methods modules, (RMIII & RMIV), both measured identically, scored a time consistent 60%, despite me thinking I was improving.
This tells us that these results do not conclusively argue for the class results being an accurate measure merely of memory under exam conditions. However they could argue that the system is designed to produce lots and lots of 2:1’s, since across a whole array of different subjects, teaching methods and assessment methods, the results are steady.
I don’t think that the current system allows for the genuine expression of student’s novel thought and intellectual talents at undergraduate level. There’s enough hoop jumping and invalid measures to ensure that students who are at least obedient will get their ordinary degree, without demonstrating much beyond content memorisation.
*Personality and Individual Differences was assessed very uniquely, with class groups making posters, and then subsequent graded online discussion using social media.