3 ms·

Does anybody have any background on why US Universities seem to universally focus on continuous summative assessment rather than formative assessment followed b

Does anybody have any background on why US Universities seem to universally focus on continuous summative assessment rather than formative assessment followed by summative assessment at the conclusion of a course? Surely the knowledge a student can demonstrate at the end of the course is what they should be measured on? The only justification I can think of is some patronising attempt to force attendance. Here in the UK there are courses I went to because I wanted to learn about the subject but chose not to answer questions on in the exam because I couldn't quite get a solid grip on. And there are courses I didn't intend to master that I ended up realising I was more suited to so revised/studied to a high standard for the exam. Why should I be forced to learn in a very specific way if I can reach the same level of knowledge in my preferred learning style?

kqr 5y ago
I have taken classes that combine both. Work during the course count as points on the final exam and/or let you skip parts of the final exam. I think it's because different people have different learning styles.
pavon 5y ago
In the schools I've attended it was mostly because they didn't want to weigh final exams too heavily (due to test anxiety), and putting too much weight on a final project doesn't scale well when every class you are taking does so - too much crunch time. Over the past few decades most have moved to allowing you to drop your N lowest homework scores which helps a lot with students entering the course with varying preparedness levels.
schimmy_changa 5y ago
I've experienced both styles (studied abroad from the US to the UK) and the final exam counting as everything in the course tended to encourage cramming and a lack of learning during the rest of the course. Learning happens well when you are able to revisit and review content over time (see Anki repetition patterns). Most people need some structure to help them with this! Some people do not, and for those people they will be annoyed but will still learn the material over time! So this is sort of a lowest-common-denominator situation I think.
sjg007 5y ago
If everything is contingent on a final exam you find a lot of folks fail. You need the continuous feedback. That doesn't mean you have to grade assignments, but rather provide that feedback..
asdff 5y ago
Having some sort of regular feedback keeps students focused on whats happening in class and gives feedback to the educator if they need to do a better job explaining something in the next lecture. A lot of students have a bad time pacing out their studies or keeping up with lectures on their own, as if they have nothing in the near term to study for they just won't bother to spend time on that class over others with more pressing deadlines. Then when its deadline time and shit is hitting the fan, then they will flood the inboxes and office hours asking things that could have been clarified weeks ago when the topics were covered in class. Having assignments broken out in regular intervals is pedantic, but there's often not really a better way to get these sorts of students engaged without a threat to their GPA basically.
bsder 5y ago
> Does anybody have any background on why US Universities seem to universally focus on continuous summative assessment rather than formative assessment followed by summative assessment at the conclusion of a course? Because a single summative assessment is worse? Honestly, a single 3 hour exam is a horrible way to assess your mastery of practially any material. There are few "good" ways to assess large classes of students for mastery. Small classess of students have a lot more flexibility. However, you can also get the problem that if the professor simply doesn't like you, you have no recourse--hopefully that class wasn't a requirement for your major. I have been on both sides of this argument (lecturer and student). I had a Computer Arithmetic and an Analog Filters classes that were ferocious about grading--you had to be 100% correct or you got zero. That is really rough--if you blew an exam you dropped. When lecturing in CS, I tended to be a little different--exams got a lot of partial credit, homework got some partial credit, but the 5 or 6 projects were 100% or nothing. Of course, the difference there was that I had automated testing running and you could resubmit your project as often as you wanted over about 2 weeks. This was probably the best compromise I've come up with. It drives the students without pounding them, and it limits the destruction if you're just kind of "off" on a day when there is an exam.

kqr 5y ago

I have taken classes that combine both. Work during the course count as points on the final exam and/or let you skip parts of the final exam. I think it's because different people have different learning styles.

pavon 5y ago

In the schools I've attended it was mostly because they didn't want to weigh final exams too heavily (due to test anxiety), and putting too much weight on a final project doesn't scale well when every class you are taking does so - too much crunch time. Over the past few decades most have moved to allowing you to drop your N lowest homework scores which helps a lot with students entering the course with varying preparedness levels.

schimmy_changa 5y ago

I've experienced both styles (studied abroad from the US to the UK) and the final exam counting as everything in the course tended to encourage cramming and a lack of learning during the rest of the course. Learning happens well when you are able to revisit and review content over time (see Anki repetition patterns). Most people need some structure to help them with this! Some people do not, and for those people they will be annoyed but will still learn the material over time! So this is sort of a lowest-common-denominator situation I think.

sjg007 5y ago

If everything is contingent on a final exam you find a lot of folks fail. You need the continuous feedback. That doesn't mean you have to grade assignments, but rather provide that feedback..

asdff 5y ago

Having some sort of regular feedback keeps students focused on whats happening in class and gives feedback to the educator if they need to do a better job explaining something in the next lecture. A lot of students have a bad time pacing out their studies or keeping up with lectures on their own, as if they have nothing in the near term to study for they just won't bother to spend time on that class over others with more pressing deadlines. Then when its deadline time and shit is hitting the fan, then they will flood the inboxes and office hours asking things that could have been clarified weeks ago when the topics were covered in class. Having assignments broken out in regular intervals is pedantic, but there's often not really a better way to get these sorts of students engaged without a threat to their GPA basically.

bsder 5y ago

> Does anybody have any background on why US Universities seem to universally focus on continuous summative assessment rather than formative assessment followed by summative assessment at the conclusion of a course? Because a single summative assessment is worse? Honestly, a single 3 hour exam is a horrible way to assess your mastery of practially any material. There are few "good" ways to assess large classes of students for mastery. Small classess of students have a lot more flexibility. However, you can also get the problem that if the professor simply doesn't like you, you have no recourse--hopefully that class wasn't a requirement for your major. I have been on both sides of this argument (lecturer and student). I had a Computer Arithmetic and an Analog Filters classes that were ferocious about grading--you had to be 100% correct or you got zero. That is really rough--if you blew an exam you dropped. When lecturing in CS, I tended to be a little different--exams got a lot of partial credit, homework got some partial credit, but the 5 or 6 projects were 100% or nothing. Of course, the difference there was that I had automated testing running and you could resubmit your project as often as you wanted over about 2 weeks. This was probably the best compromise I've come up with. It drives the students without pounding them, and it limits the destruction if you're just kind of "off" on a day when there is an exam.