Mark, this is just a quick answer because I haven't had time to read the above document as yet. (Just got back from a conference on Science and Math Education in Penang.)

When I was a school teacher, I used to report on students' academic achievement as well as improvement in their skills eg. SPS (science process skills), and improvements in their behavior or soft-skills. When parents came to school to get the report cards; I spent more time talking about the 2nd and 3rd, because the first was already well-expressed, written as grades or raw marks. Throughout the years, I observed that soft-skills contributed a great deal to my students' success in college and beyond that. I hope I have understood the question, and I will get back to you. Thanks.

Hi Miranda The intent of the question is to examine different ways to report to students on Heir academic on achievement performance specifically eg one way is to compare their performance to their cohort statistically. Another might be to compare them to expected Standards for age or grade. Another might be to compare their performance to a fixed set of standards through which they progress. Another might be to make a list of competencies or 'things they can do'. etc

Patrick, what you are suggesting seems to be verbal descriptors of achievement as specific skills learned or weaknesses overcome. Is that correct?

Sorry Mark. I have been away a long time. The 1st week, I was attending a conference in Penang, where I presented 2 papers; and during the week after that I was involved in running our church ABC, annual bible camp. But it was a great joy, the theme was 'Paul, the apostle, on he run'. What kind of 'refugee food' do you think we ate?

Back to the question, I think that the ways you suggested are fine. But in my country the preferred mode is school-based assessment at the moment, that is rather restricted.

Well, Patrick has added that the students themselves should report on their own progress, come up with their own swot analysis. This will show our students clearly that they are really responsible for their own learning, although the teachers are always there to facilitate the process. A good idea, right?

An important question Miranda 'What DID you eat?'

My current preference, as is probably self evident from my questions and answers in other places, is for reexamining the paradigm of 'assessment of learning' as is required in most systems by bureaucrats and politicians. I am not sure myself if it is as critical as is usually made out to be. However, assuming that assessment OF learning is critically important to schooling as a system, what should it look like. I posit that the major characteristic should be that it focussed the audience of the report onto progression and personal improvement, and for that to happen the criteria need to be designed as a continuum that is fixed throughout a particular period (say grade 1 to 10 as the most critical to establishing desired attitudes to learning)

Hi Mark,

Your questioning is extremely relevant and should be asked more frequently. In most educational systems 'numbers' makes sense to decision-makers. This may be pass rates, throughput rates, class averages etc. The question, however, is whether these scores are indications of development. If I present a class everyday and exclusively and repeatedly give students practice questions to complete and give them the answers, will they pass? Most probably! Will they develop? Most probably not!

Maybe assessment of learning should include measurements of other developmental attributes important for student development, i.e. presentation skills, problem solving, independent learning, self-confidence...

Do you think recording and analysing individual behaviour during learning sessions at different points of the learning cycle may be beneficial? I mean, will you be able to observe development? I have thought of doing this. For example, would you consider that a student who hardly participated during group activities at the start of a semester, but joyously contribute to group processes by the end of the semester, have developed inter-personally?

-And then: What would you consider achievement? What would it look like? Then we can figure out how to measure it... Right!?

Tobias, I am inclined to think that the bigger personal attributes such as problem solving should be ALL that are measured. What we currently measure are small evidences of learning. I refer to these as the 'toolkit 'that the learner builds as they progress. It includes facts, how to, etc so to solve a problem, a learner accesses their toolkit in various ways to meet their needs. Some use memory. Others use research skills. Others use first principles. All methods would be relevant to solving the problem, and valuing one method over another is useless and maybe inequitable. In the end it is how well the problem is solved that is what matters. Was the solution valid,

ethical, useful?

Not to be excessively agreeable, but I have to agree :) Is the question then: how do we measure these graduate attributes? And do these attributes include qualities such as problem solving, creativity, independent learning, self-directed learning, critical thinking etc.

Tobias, I finally absorbed the last part of the response. 'How do we measure etc', maybe by giving meaningful/authentic problem based tasks to be completed over a period of time, that require a cognitive demand commensurate with the course demands. The lecturer/tutor collaborates with the students to solve the problem/complete the task. That way he/she knows the thought processes and approach of each student (annotating student work for later assessment is complicated but can be done). I mandate that students send me regular updates that I analyse, comment on etc. If uncertain, a student conference normally clears the mind. A series of questions that students respond to a-la RG could work for some courses.

With reference the the valued attributes, I find it difficult to separate the different types of high level cognition you mention, so my strategy is to annotate work with reference to each when I am certain, and a generic higher order thinking reference when uncertain. Then I use these to make my final judgement. Low level aspects such as grammar, punctuation, correct referencing style etc do count, but not as much as the higher order thinking. I write down my criteria, then describe what i would expect to see for each criterion for each standard. I don't use any algorithm. In Queensland, Australia, 'most like' judgements on criteria have been shown to be successful for the last 30 years. In other words the student work is judged to be 'most like' a particular standard and can be justified if need be by reference to the standard descriptor.