I am in need of widespread quality dialogue around the purpose and development of rubrics as we prepare for the implementation of Version 9 (V9) of the Australian Curriculum.
Once upon a time my school had a rubric that was similar to the SACE (South Australian Certificate of Education) performance standards. It was suggested that this was inappropriate as it did not indicate assessment of the Australian Curriculum Achievement Standards.
We then moved to a rubric divided into the profiency strands of fluency, understanding, reasoning and problem-solving, and used qualifying language like 'consistently' and 'in un/familiar' situations. The relevant Achievement Standard sentence was attached to the rubric.
In planning for V9 we are redeveloping our rubric. It is proving to be a challenging thing to do as there is limited consensus among stakeholders as to what a middle school mathematics rubric in South Australia should look like. Some common discussion points are that a rubric should:
- Show what skills/knowledge are to be demonstrated
- Clearly define the difference between the grade bands
- Be wordy enough to maintain clarity, but not so verbose as to be unusable by teachers
- Be student/parent friendly so that students are able to determine exactly what is required to achieve a desired grade.
We have been provided with a performance standard from our system administrators which uses the Achievement Standard 'as is' for a C-grade band. The other grade bands are determined using cognitive verbs relating to hierarchies of Bloom's Taxonomy. Some dilemmas:
- Should we have a separate rubric for every assessment, or every line of the Achievement Standard? (teachers generally agree this is too much work)
- Is it more appropriate to have a SACE style rubric that is generic and can be used for any assessment?
- Is Bloom's taxonomy the generally accepted way to differentiate grade bands? i.e. is an A level achievement all about synthesis and analysis? Can these higher levels of Bloom's not occur at a C level? And can the 'lower' levels of Bloom's not occur at an A level?
- If not, what is the best way to differentiate the grade bands? Is it about consistency of accuracy? Is it about operating above your allocated year level?
I asked ChatGPT for advice, using the prompt, "What makes a good rubric in middle school maths?" Here is the response:
Clarity and specificity: The rubric should clearly state the learning goals and objectives of the assessment. It should also provide specific criteria that define what success looks like at each level of achievement. Clear and specific language in the rubric will help students understand what they need to do to succeed.
Consistency: A good rubric should be consistent across all assessments. This means that the criteria and expectations should be the same for all students and all assignments. This will help ensure that grades are fair and consistent, and that students understand what is expected of them.
Alignment with curriculum: The rubric should be aligned with the middle school math curriculum. This means that the criteria should reflect the skills and concepts that students are expected to master at this level. The rubric should be reviewed and revised regularly to ensure that it stays aligned with the curriculum.
Differentiation: A good rubric should differentiate between levels of achievement. This means that it should have clear descriptors for each level of achievement, such as "needs improvement," "proficient," and "excellent." This will help teachers identify areas where students need additional support and also help students understand what they need to do to improve.
Feedback: The rubric should provide opportunities for feedback. This can include written comments, or opportunities for students to discuss their performance with their teacher. Feedback should be specific and constructive, highlighting areas where the student has excelled and areas where improvement is needed.
It seems we are on track with some of our discernment, however I find it hard to marry "Consistency" and "Alignment with Curriculum" in the Australian Maths Curriculum context. Given the Achievement Standards are the mathematical content, which we are only grading a small part of in each assessment, it seems challenging to create a rubric that can both be used both across all assessments and align with the curriulum. What am I missing?
I am at the point in my consultation where I would really like to get some comments from other schools and how they view all of this. It would be great to get some system-wide consistency, but I feel we are yet to reach a point where we are clear of the purpose and design criteria of rubrics. Please join the conversation :)
I am glad others are attempting to produce quality rubrics that don’t necessarily mimic the SACE performance standards.
ReplyDeleteI’ve always said that Maths is the strangest faculty for rubric writing.
I really like Blooms Taxonomy for domains such as English and History. It’s aligned to Australian Curriculum too, which is helpful. However, Analysis, Evaluation and Synthesis doesn’t exactly define the A & B level in many mathematical contexts.
Have you experimented with SOLO Taxonomy for Maths? I think in some contexts it could be quite useful. Especially in rubrics for investigations. However, may not work for a generic rubric. Could significantly narrow the assessment if teachers felt they needed to use the 5 levels of SOLO for everything.
Good luck!
Thanks for your response Adrian. I hadn't even thought to look into other taxonomies, will certainly be throwing SOLO and perhaps others into our future conversations. Thank you for your support!
ReplyDeletenice
ReplyDeletenice
ReplyDeletenice
ReplyDeleteHi, I have been working on this for a long time at both secondary and tertiary levels. I worked in Curriculum Services for several years and was responsible not just for content but also the development of rubrics. I also ran teacher workshops on exactly this concern. See the Sass Education website (which is under revision at the moment). I have very strong feelings about this topic and the common sticking points relate to consistency of language, using measurable verbs, understanding the incremental levels A-E, aligning the task and the rubric, and not assessing different skills at different levels - it happens all the time! Is a generic or global rubric possible? Yes, but it needs to be clearly defined. All this sounds very obvious but it is not so easy to achieve. The other key point is that a rubric is only as good as the task that is being assessed and many times I have seen students being assessed on skills, knowledge and understandings that are not in the task itself! Good luck... what you are trying to achieve IS possible.
ReplyDeleteThank you so much for your comment, valuable insights. I will check out the SASS website 😀
ReplyDeleteI know it is a long way down the track but I happened across this again and wanted to make a couple of comments to your questions. I HAVE CREATED LITERALLY HUNDREDS OF RUBRICS AND HAVE A MODEL FOR APPROACHING THIS THAT WORKS WELL IN PRACTICE.
ReplyDeleteWe have been provided with a performance standard from our system administrators which uses the Achievement Standard 'as is' for a C-grade band. The other grade bands are determined using cognitive verbs relating to hierarchies of Bloom's Taxonomy. THIS IS TRICKY – THE EXPLANATION IS LONG BUT IT NEEDS CAREFUL TEASING OUT OF THE STANDARDS AND IDENTIFYING ONE AT A TIME WHAT IS RELEVANT AND APPROPRIATE. Some dilemmas:
• Should we have a separate rubric for every assessment, or every line of the Achievement Standard? (teachers generally agree this is too much work) I DON’T AGREE – THAT IS THE NATURE OF TEACHING. AND THE RUBRICS ARE USED AGAIN OR REFINED IN THE FUTURE. IT REQUIRES A BIT OF HARD WORK UP FRONT OR FIND A GOOD SOURCE AND THE HARD WORK IS DONE FOR YOU.
• Is it more appropriate to have a SACE style rubric that is generic and can be used for any assessment? I ENTIRELY DISAGREE. THERE IS A LOT OF INFORMATION ABOUT HOLISTIC RUBRICS AND THEY ARE MOSTLY PROBLEMATIC AND DO NOT PROVIDE A VERY ACCURATE REPORT ON STUDENT WORK.
• Is Bloom's taxonomy the generally accepted way to differentiate grade bands? i.e. is an A level achievement all about synthesis and analysis? (NO) Can these higher levels of Bloom's not occur at a C level? YES THEY CAN, JUST AT A DIFFERENT QUALITY. ONE OF THE FUNDAMENTAL RULES ABOUT WRITING RUBRICS IS THAT THE SAME SKILL SHOULD BE ASSESSED AT ALL LEVELS. And can the 'lower' levels of Bloom's not occur at an A level? ABSOUTELY. THEY UNNDERPIN THE HIGHER LEVEL RESPONSES SO OF COURSE THEY CAN.
• If not, what is the best way to differentiate the grade bands? Is it about consistency of accuracy? Is it about operating above your allocated year level? LANGUAGE IS THE KEY
• AND BTW CHAT GPT CAN ONLY REPRODUCE THE INFORMATION OUT THERE AND IT IS OFTEN INCORRECT. I WOULD NEVER RELY ON AI. I SPENT ENDLESS HOURS TRYING TO TRAIN AI TO RETURN THE KIND OF RESPONSES I WANTED AND IT WAS ALL BUT IMPOSSIBLE. BUT OF COURSE THIS RELIES ON YOU KNOWING EXACTLY WHAT IT IS YOU WANT AI TO RETURN TO YOU. IF YOU DON’T, THEN THE ERRORS ARE PERPETUATED.
I WORK ON THESE KINDS OF QUESTIONS DAILY - STILL. AND I DELIVER WORKSHOPS TO TEACHERS AROUND THIS AND I FIND THE SAME MISCONCEPTIONS COME UP ALL THE TIME. IT IS NATURAL UNTIL YOU KNOW OTHERWISE, THEN IT MAKES PERFECT SENSE.
GOOD LUCK. I HOPE YOU MADE GOOD PROGRESS ON THIS.