Dear Acted - X assignment results (open letter)

[avanbuiten]

Dear Acted

Would you consider releasing figures (updated weekly or monthly) on individual series X assignment results? (e.g. X1, X2, etc).

This could be done with a bar chart, where bar widths are 4% wide (e.g. 45 -49%, 50 - 54%, etc) and height of the bar is determined by the proportion of candidates falling into that category. Of course, candidates identities would remain anonomous.

I think this would be a good idea because it then helps students to see how they are doing compared to others. Relative exam performance is a significant factor in determining likelihood of passing an actuarial exam. It is also useful for identifying where a candidate is comparatively weak in one area (and needs to work more), and overly-performing in another area (so need not spend so much time on in future).

I am sure you must already keep these statistics for your own use, so why not release them for our use?

I look forward to your reply (or anyone else who wishes to comment).

Regards,

avanbuiten

 

[Sauny Bean]

I think this is a great idea.

I have been put off doing the X assignments in the past because of potential low scores, but if I knew that I was still doing okay compared to everyone else, this would boost my confidence.

I guess meaningful statistics would only be available several weeks after you'd done an assignment though. Also, perhaps those submitting might be more likely to obtain solutions from candidates not submitting X Assignments, particularly if they are required to do the X assignments by the employer, so there may be some distorted results.

 

[avanbuiten]

There may be some slight distortion but I would hope the majority of candidates would be honest when they tick the "was it done under exam conditions?" box : Yes, Nearly, No.

The results could then be subdivided by these 3 answers, but still also show us a summary result without distinguishing between them. Like all statistics in real life, there are likely to be a few inconsistencies. I'd be happy with about 90% confidence.

I also think meaningful statistics could be available within one to two weeks of submitiing an assignment, if you are one of the earlier submitters, however, the results would require updating on a weekly (or monthly) basis in my opinion.

 

[OneTwoThreeFour]

There may be some slight distortion but I would hope the majority of candidates would be honest when they tick the "was it done under exam conditions?" box : Yes, Nearly, No.

Does everyone actually record their true conditions for sitting assignments? (possibly this should be a separate thread - sorry!)

Have to admit I don't

I've found that if I'm truthful about the conditions the feedback I receive is pretty close to useless...

...for example in the past for an X1 assignment it's taken me four hours using the notes and I've recorded this accurately as taking 4 hours and "no" for exam conditions. The entirety of my feedback was - "Good attempt - try to do this under exam conditions and in the allocated time for the next assignment" - which doesn't really help. If I just put down 3 hours and "nearly" I tend to get some decent feedback.

Diverging even further from the point I can't really see the worth of sitting papers under exam conditions for the early assignments anyway. Should ActEd really be encouraging people to be trying to learn the lists etc necessary to sit a paper under exam conditions several months before the exams?

Attempting to return to the point: sounds like a good idea to me. But as SB says they couldn't really publish them until a few weeks after the final deadline (and that would exclude some people using marking vouchers). I guess they could publish last sittings results if the assignment hadn't changed much?

 

[avanbuiten]

I always do the assignments under exam conditions. Sometimes I think I know something well, but when I struggle in the assignment it helps my identify any weak areas.

It also makes the transition from doing assignments to sitting past papers less painful.

As far as this 'out of date' things go, I just don't buy it. Most people will send in assignments well before the final deadline date, otherwise they are really going to struggle to finish the course on time (for CA1 anyway). However, I do accept that the later the statistics are published the more credability they'll have which is why I think weekly updating will be a good idea.

Even if only 30 had sent their assignments back at the first publication, I think I could still get a relatively good indicator of my comparative performance. In fact I learn something just from asking the other two people (same exam) at work!

Maybe I should actually send my open letter to Acted by email, guess they might find it a little strange me communicating in this manner?

 

[Admin]

Not strange and it's nice to get some opinions on this topic. (Of course if you definitely want us to read something then yes please email us directly.)

I'm sure we can do something. We produce something fairly similar for markers to help them set the grade at the right level and we also state the average score for each assignment on the Company Reports that go to Contacts/Mentors. We'll ask the IT team to extract the relevant data.

Publishing the previous session's results would probably be the easiest thing to do although we'll also look into issuing more up-to-date results, but they tend to be fairly consistent from session to session. Splitting down between attempt conditions is probably a bit much (and in fact not possible at the moment as this is not recorded on the database - it's just used to guide the markers).

I've just scanned a random sample of scripts looking at the exam condition question, and the split between yes/nearly/no is roughly equal. Before looking I would have said that the majority circle nearly for assignments and yes for mocks. We're all different and prefer to use the assignments in our own way (as eveidenced by two conflicting views above). The key is to pass the real thing (stating the obvious).

Finally, the majority of students (but not all) who use marking do hand in their assignments on or just before the final deadline. I think many wait until results before committing (although this isn't possible in CA1 or in the summer) and many simply aren't as advanced in their studies as some might think.

We'll let you know when something has been sorted out.

 

[avanbuiten]

That's great.

We will wait with eager anticipation.

 

[Admin]

Sample graphs

We've produced graphs for the CA1 marks that students were awarded over the summer. The results are not particularly credible for CA1 because very few students used the marking service in CA1 over the summer.

Three samples attached. Do you have a preference? I expect that it's number 3 but thought we ought to check before progressing.


Update: also now added a graph for this session's results to date. Data for X1-X5 probably more credible. The X6-X10 graphs are based on only a few scripts.

 

[avanbuiten]

Of the 3 options I prefered option 3 best.
It seemed much easier to understand visually - I really liked it, well done!

Good choice using CA1

 

[bystander]

Agreed. 3 is far clearer and easier to compare assignment to assignment. If for some reason you want to print it and only have b&w it would still be legible.

Look forward to the 'launch'.

 

[LouiseF]

People get 80-100% on some assignments???

Oh no. Am in trouble.

I got a 73% once and was so chuffed I thought I deserved new shoes. Some people out there must be doing a lot of shopping.

Must get the study cap on.

 

[avanbuiten]

People get 80-100% on some assignments???

Oh no. Am in trouble.

I got a 73% once and was so chuffed I thought I deserved new shoes. Some people out there must be doing a lot of shopping.

Must get the study cap on.

There will always be some people who get marks in this range, but I was surprised that in some instances it was more than 5%.

My guess is people may be "cheating", although to them it's not counted as cheating, because that's just how they work best.

Alternatively, I'm wrong and there is a lot of people regularly getting over 80%