ETA: I'll leave what I've typed below, but actually it looks as though I should eat some of my words, because looking back at
this thread that I started when DS was 5, I am reminded that at that stage we did have similar issues with ALEKS. FWIW, our resolution was that DS did ALEKS up to Level 6, then did ALEKS Chemistry, taking a long time over it (just over a year, IIRR) then went straight on to ALEKS Algebra 1, and didn't feel the bump - he did Alg 1 in a few days in fact, as a "let's use up the subscription we've paid for" exercise. (He had, of course, learned a bit of maths elsewhere in between.) So if part of the issue is that you're feeling your DS should do all the courses in sequence - I don't think you've said which it is that he's doing just now - let me urge you to be a devil and consider alternatives :-) Gaps can be positively helpful, sometimes.
Some probably are, but there's no way for me to just find and plug the holes. Holes certainly exist due to mismatch in the scope and sequence with what he's done before (Singapore and assorted problem solving), but it's not only time-consuming but prone to error to make assumptions based on abstracts of the scope and sequence of each.
In fact, the algorithms that ALEKS implements do IME do a pretty good job of accounting for scope and sequence mismatches, if you let them.
Because the main reason for the interim assessements, as opposed to the regular ones, is to cement what's been recently learned. These are the great bulk of questions on them, perhaps the entirety-- they're more quizzes than level-setting assessments like the initial bigger one.
I think this description is misleading, perhaps indicating that you've been misled. The way the underlying theory works, and it appears to me to be implemented pretty straight, is essentially that the system selects each question based on what will give it maximum information about what the student knows, or to put it another way, it selects a question where the calculated probability that the student will get the question correct is as close to 50% as possible. During a course it uses prior information about what the student has done in learning mode (which of course it doesn't have in the initial assessment) but other than that, the process is identical in all kinds of assessment.
What it doesn't do, AFAIK, is to adapt to individual students, or allow customisation of the prior probability that a student will answer correctly on the most recently added topic. It could be that for our children, the most informative question might actually be on a "you are ready to learn next" topic rather than on a "recently added to pie" topic that might be the result of the calculation with default parameters.
However, this doesn't normally matter too much, since the number of recently added topics will, with the default assessment schedule, always be fewer than the number of questions in an assessment so there is room to ask on all recently added topics and still have at least a little space to explore beyond. Come to think of it, if you are cancelling assessments, then the assessments you do let it do are indeed likely to get filled up with questions on topics that are on the pie, just because it limits it to 25-30 questions and if there are that many new topics, those are likely to win. Also, if your DS is doing scattered topics from all over the pie, the algorithm will work less well than if he does a long strand in a session e.g. sticks to one slice of the pie. To give a simple example, if he does 5 topics each of which depends on the previous one, then what the algorithm will (most likely) do is to ask him a question on the last one; if he gets that right, the probability it assigns to his being able to answer questions on the earlier 4 goes up so it probably won't ask him on those, leaving more spare questions for other topics.
But really, an assessment will only ever ask one question per topic, at a maximum, unless the child is making mistakes. There seems to be an inconsistency between you saying that that's too much review and you wondering whether he's getting enough practice - which is your main concern?
Have you actually watched him doing the assessments? Is it possible that, being 6, he's actually making slips that are causing the system to ask him repeated questions on the same topics? At that age, I did sometimes sit next to DS and go "Ahem" at crucial moments to avoid this, though I did this sparingly and more often gave him enough rope to hang himself, as it were...
