# The Physics Problem and Standards-Based Grading

Inspired by the Standards-Based Grading Borg, I am slowly putting together the picture of my own potential SBG implementation. One thing which I still have yet to sort out is how does the typical physics problem fit into a Standards-Based Grading implementation? Or to ask a slightly different question..

How does one assess a typical physics problem in a Standards-Based Grading implementation?

Note: I have not read any of the SBG literature and my exposure to SBG comes entirely from the SBG blog Borg so my questions arise from the implementations with which I am familiar.

Second note: This post is meant to start a conversation that will hopefully help me sort out how I can make Physics problems and SBG ideals happily co-exist in a course. I am not tied to the exams or any other part of my colleagues’ sections of the same intro courses that I teach, but in the end I have to show that my students are able to solve the same kind of problems that show up on my colleagues exams.

### The Physics problem

I would argue, that in the typical university physics course, the physics problem is the most commonly used assessment item. They are assigned for homework, they show up on exams, students do tons of end-of-the-chapter problems to study for exams, and most college instructors use them as the primary method to teach new concepts. My intro-level courses are far from traditional from the point of view of how class time is spent (pre-lecture reading assignments, clicker questions using peer instruction, strong conceptual emphasis, lots of group work), but I still use a traditional assessment system. My students work problems on their homework assignments, in their quizzes and on their exams. Approximately 1/3 of the marks on my exams come from problems.

I am being overly generous and calling what is usually just an exercise a problem. A “physics problem” is something which requires actual problem solving on the part of the person attempting it and not just some algorithmic pattern matching based on previously seen or completed exercises. But let’s not get hung up on this distinction. Let’s just say that a physics problem is something which requires some or all of the following skills:

• Identifying relevant physics concepts and correctly applying those concepts to the situation discussed in the problem statement;
• Building a model;
• Keeping track of units;
• Procedural/mathematical execution;
• Respecting significant figure rules;
• Making and justifying estimates;
• Checking the reasonableness of answers;
• Translation of representation (for example between graphical, words, formulas, motion diagrams and/or pictorial);
• Writing up clear, coherent and systematic solutions.

I’m sure folks could come up with many others, but those are the skills that my brainstorm yielded.

### Assessing a Physics problem in SBG

Let’s say that for various reasons (administration, the rest of my department, transfer status of the course to larger universities, etc), that I must have my students tackle numerous Physics problems during a course and these are problems which consistently require as many of the above-listed skills as possible. How do I assess these problems in SBG?

My colleagues would argue that most of those above-mentioned skills used when solving a problem are very important skills for a person continuing on in physics and that we should “teach those skills” to our students. I could write-up the above mentioned skills as standards. I could then have one of two primary strategies: assess most of these standards individually or assess multiple standards on each problem.

The problem with the individually assessed standards is that they are all part of the problem-solving package and assessing each of them individually doesn’t assess the coherent whole of problem solving.

With assessing multiple standards on each problem, not every standard can be present at the same level. And for some of these standards it seems as if trying to assess them from a tiny piece of an individual problem would be the same as assessing how accurate of a basketball shot somebody has based on seeing them shoot only a single free-throw.

### Incorporating Physics problems into an SBG-assessed course

Now let’s say that the above is still true, I must have my students tackle Physics problems. But, for whatever reason, I can’t/don’t/won’t come up with a good way to assess them. But I want to be true to SBG and not have to just tack on 10% for homework problems or just give them an exam at the end that has problems and is worth a certain fraction of their mark while all their SBG-assessed standards are worth the remaining fraction of the their marks. I just want their grade to be some sensible reflection of their scores from the assessed standards.

How do I incorporate Physics problems into my course in this case?

### A couple more quick questions

• If I am using problems to assess multiple problem-solving standards at a time, how do they earn the “exceeds expectations” levels (4/4 or 10/10) on their standards?
• The common practice in SBG seems to be to make the standard being tested nice and explicit. But having a standard like “identify relevant physics concepts” means that you have to avoid making the conceptual standards explicit with a problem. Is that good, bad, or does it matter?

1. bretbenesh says:

Hi Joss,

I completely forgot about this when you mentioned something similar in the comments of my weblog, and it gives a possible answer for your question (I am going to copy this over to my weblog for completeness, but you do not have to read it twice):

I made “problems” a separate “topic/category/standard.” Of course, I called them “proofs,” as I am a mathematics teacher. But I think that the idea is the same. So really, I had four different types of assessments in my course:

Exams (the boring, usual exams; nothing SBG about them)
Quizzes (this is the SBG part of the course)
Proofs (this plays the role of your “problems”)
Project (they had to do one, but it is not relevant to their discussion)

So I had the students do their synthesis work on the “proofs” portion. Here are the logistics: I assigned one per week. The students submitted them, and I gave them feedback. There were two grades: “Acceptable,” and “Incomplete.” An “Incomplete” means that they can resubmit again for no penalty. Repeat until they get an “Acceptable.”

Here is how they affected their final grades: my general grading practice has been “the more evidence you provide me, the higher the grade.” In very simplified terms, this means that if a student demonstrated that they could do a standard five times, they got an A. Four times got them a B, and three times got them a C (this is VERY oversimplified).

The “Proofs” worked similarly. If they successfully solved all of the proofs, they got an A for that part. If they are only missing one, they get a B, etc.

In the end, I made a judgement call for the final grade. If everything was at an A level except for one B, then they got an A. There was no rigid formula (I was clear about this in the syllabus).

So that was my solution. I am interested to hear what other people came up with.
Bret

2. Joss Ives says:

For reference, I have been asking Brett questions about his SBG implementation over at this post.

Thanks for the info Bret. So if a student handed in a proof, but there were enough issues that you would consider it “Incomplete” (let’s say it would earn below an 8/10 or 9/10 on a regular old-fashioned marking scale), would you let them fix/correct it and resubmit to earn the “Acceptable”? This seems like a brilliantly simple solution that shifts the student execution of a problem/proof into the spirit of SBG where they have opportunities that are similar to being reassessed.

Like you, I have to have a final exam for political reasons. and have to have some problems on this exam. I wonder how to fit those exam problems into a system similar to yours.

3. Sounds like you are trying to differentiate between higher-order problem solving skills and lower-level skills. I have my skills ranked as core, intermediate, and advanced and those rankings also factor into determining a final grade. See: http://www.scribd.com/doc/48673856 and http://www.scribd.com/doc/48673857

• Joss Ives says:

Thanks Frank. I remember looking at your grading policy in the past, but that was before I was seriously considering my own implementation so I am now looking at it in a new light. I have questions! (Always with the questions)

1. How does it factor into their grade if they get some expert-level scores on your standards?

2. It seems like you give problems on your quizzes. For a problem that assesses a multi-model standard such as “I can use more than one model (BFPM, CVPM) plus experimental force laws to solve a problem for multiple unknowns” do you also assess standards such as “I can draw a labeled force diagram…” or any concept-application standards on the same quiz problem? If a student doesn’t bother to draw a FBD (on a problem that should have one), but does a good job on the problem, do they go back to “no evidence” for the FBD standard?

• Hi Joss. (1) As of right now, I don’t treat Expert any differently than Proficient. One reason is that some core standards can really only be evaluated on a Yes/No basis, there is no “expert” response. However, I am on the verge of modifying the grade scale slightly so that students would need to do some exemplary/expert work to go above 95. I’m thinking screencasts/pencasts, etc. Something that requires more work and explaining than a quiz question. Or making up their own TIPER-style problem (see my “reassessment experiment” post).

(2) A question for the “I can use more than one model…” would be like a multipart textbook question without the intermediate steps spelled out. Obviously a lot of substeps (and their standards) would be present, but I typically do not assess those additional standards — I assess them explicitly as substeps for another problem. I want to see that they can put all the pieces together themselves.

I hope all that made sense!

• Joss Ives says:

Hey Frank. Your “reassessment experiment” post piqued my interest so I commented over there.

This is sort of a follow-up to your answer to my question 2. What do you do, if on a problem, the student leaves their answer with 12 sig figs and no units? From the standards that you link to above, none of your standards deal specifically with these things.

• You asked, “What do you do, if on a problem, the student leaves their answer with 12 sig figs and no units?”

I do not have a separate standard for sig figs and/or units. I’m not a blow hard for sig figs in HS (I seriously believe their brains aren’t able to comprehend them yet) but most kids are great with giving me 3 or 4 digit answers.

Units, on the other hand, are a big deal and indicate a lack of understanding of the concept tested if the units are wrong or missing. So if the standard was about acceleration and the units were missing or given as “m/s”, then that is “developing” for that acceleration standard.

• kellyoshea says:

Re #2: I consider the diagram an integral part of solving a problem. Each of my units usually has an objective like “3.8 B CAPM I can use kinematics concepts to solve problems involving objects with changing velocity.” that is about solving problems with that model (always a B-level objective). To show mastery on this objective (in my class), they have to have everything correct, including the work, the diagram, and the answer. On the exam, just having the first two correct would be enough (I will mark them down for calculator-user error on a test and make them reassess because I don’t want them settling for mostly-good problem solving skills, but summative assessments are different). In any case, the answer is the least important part.

But I wouldn’t give a 0 for “2.1 A BFPM I can draw a properly labeled free body diagram showing all forces acting on an object.” if they didn’t draw the FBD because this isn’t a new piece of evidence.

• Joss Ives says:

Welcome Kelly and thanks for contributing.

I am seeing how both you and Frank are using “solve a problem” standards to assess the individual pieces of the process as a whole for a given model/concept. This is much less of a headache than trying to evaluate multiple standards at once.

In the case of the FBD standard, I see how you wouldn’t give a 0 if they simply didn’t do it as part of a problem. I imagine in many cases you would then not give them mastery on the “solve a problem” standard that was being tested since, as you mention, you (and most of us) consider drawing the diagrams as integral to the problem-solving process.

4. Mylene says:

Interesting questions. I am brand new at this so please realize that I am at the “bouncing ideas off of people” stage … don’t have nearly enough data to be close to “recommending”. Like Bret, my assessments are binary. Like Frank I group them into categories, which I call “level 1-5”. You can’t get credit for level 3 until you finish level 2. The requirement for synthesis increases at each level. If Standards A and B are both “level 2”, then Standard C, “determine whether to use A or B” would be at level 3. The levels map proportionally to a grade (a 3 in every unit is a 60%).

To deal with noise in the measurement, I give a quiz each week, on anything I feel I need more data about. These quizzes are mandatory, and if you skip the quiz or get questions wrong, your scores for those standards are reset to incomplete (I would prefer to use a decaying average but using the most recent score allows students to easily calculate their grade themselves, which I find pedagogically useful). You can reassess anything you like on Wednesday afternoons, if you request it in advance (GoogleDoc forms have saved my bacon on this one).

To “exceed expectations”: I hand out a list of standards for each unit. The list ends at 4/5. If you want to get a 5/5, you have to propose your own problem. It must involve developing a strategy (or applying one in a way not explicitly taught in class) and combining ideas from more than one unit. Occasionally I put a 5/5 question on a quiz; it is an actual problem, not an exercise (often I use it to introduce the next unit and get students attempting problems they don’t know how to do yet. So far they seem to find this fun.)

As for “making standards explicit”, I don’t think there’s anything ambiguous about a standard that says “choose which concepts apply.” I’ve certainly got a few of those. It’s awfully helpful for troubleshooting, though, if those concepts have already been assessed individually and explicitly.

Looking forward to hearing what you come up with.

• Joss Ives says:

Hi Mylene. It’s always nice to have more canucks running around doing good.

I’m curious about your complete the lower-level before being able to assess the higher-level system. How do you deal with situations where somebody misses the quiz where you assess (from your examples) Standards A and B and then the next week on a quiz you assess Standard C and they do the work that would earn them a “Complete” on Standard C because they determined that A was the way to go and even did the work that would earn them a “Complete” on Standard A?

I like your system for the 5/5 standards. Have you had any students propose their own problem?

One of the half-formed thoughts in my head has to do with how to deal with the “choose which concepts apply” standards. I imagine that, more often than not, the assessment would involve not only choosing which concepts apply, but actually applying those concepts. Here you (royal you) wouldn’t make explicit the standard(s) for the relevant concept that applies, but do you (again royal you) also assess those non-explicit relevant concept standards as part of the problem?

Vague and confusing paragraph alert -> They successfully choose (and justify) their use of A over B to earn “Complete” on C, but do a terrible job of actually executing A. Is standard A reset or do you not actually assess A as part of the problem and they instead earn “Incomplete” on C since their inability to properly apply A means that they can’t really do C (despite having chosen and justified it)?

• Mylene says:

Hi Joss,

On cumulative skills and synthesis:
If I give a quiz question where C requires you to use A, then that question counts for both standards (and the question says so in the margin… or maybe it says “C, A and/or B”). If it’s completely right, they get two skills checked off. If it’s partly right and I can tell which part they get, I’ll credit that part. If it’s partly right and I can’t tell which part they get, they lose both parts. The moral is, if their reasoning (or explaining!) is so muddled that I can’t even tell which part they understand, then they probably need more practice all round.

But ultimately I am trying to tell the difference between people who can “do” and people who can “choose” and people who can do both. That means I have to control some variables. If C requires you to choose between approach A and B, the problem probably says “choose between A and B. Justify your choice.” If they want to demonstrate later that they can do A and B, they have lots of ways to do that.

>I like your system for the 5/5 standards. Have you had any students propose their own problem?

Yes, and this is the part that I’m most excited about. I would say that nearly all of my students have at least one 5/5. They are really excited to have official recognition of their desire to mess around and build stuff and learn stuff. Topics have included everything from circuit-bending to regenerative braking. Last week a student left the shop during class, removed the relay from his trunk latch, and used it to build a crude boost chopper to light a 120V bulb from a 12V source (there are better ways to light that lamp, and I teach choppers and relays next year, but it was the problem-solving that was quality). Sometimes I approve their proposal on the condition that they present their results to the class. It gets them checking the web and their textbook for things I’m not going to teach. Nothing else I’ve done has ever had this much success in getting people to read the textbook.

>One of the half-formed thoughts in my head has to do with how to deal with the “choose which concepts apply” standards. I imagine that, more often than not, the assessment would involve not only choosing which concepts apply, but actually applying those concepts.

I agree that that’s tricky. It makes me wonder, what benefit do you get from requiring that both standards be assessed together? Would it be possible to derive that benefit some other way?

*grin* I don’t think it was confusing. I think I’ve given my approaches above…

One last point: if you choose to assess A and C separately, that doesn’t mean you never require students to do A and C together. Maybe it’s an in-class exercise that leads to a group discussion. That way they have some experience doing the whole process, you have some insight into what they need to practise, but the thorny “how do I set a clear grading policy” problem becomes less important.

Thanks for the food for thought — interesting discussion.

• Joss Ives says:

Hi Mylene,

I’m impressed with your student’s desire to tackle the 5/5s. I’m always trying to find that right combination of personal challenge and difficulty to have the students really rise to the occasion and impress me.

“It makes me wonder, what benefit do you get from requiring that both standards be assessed together? Would it be possible to derive that benefit some other way?”

I think that focusing on developing learning goals for my courses has also resulted in me creating quiz and exam questions that address multiple learning goals at once. I seem to now have the “assess the most stuff with fewest questions” mentality. Of course, when I put my thinking-about-SBG hat on, I think that it would often be very difficult to properly assess multiple standards using the same question.

Currently, everything that I work on with my students in class is working on stuff which will help them do their best on exams. If I think that it is an important or useful thing that they can do fermi problems, write a paragraph explaining their understanding of how the physics they know applies to a certain situation, or draw motion diagrams, then we will work on it in class, they practice it as part of their homework and questions asking them to do these things will show up on the exams. In an SBG implementation, I would need to work on my salesmanship to be able to convince them that getting them to do A and C together is worthwhile to do in class even if they won’t specifically be asked to do it as part of an assessment. I don’t doubt that I can sell it, but my salesmanship is still very much a developing skill. I have only taught a total of 9 courses so far and each round I find that I am a lot better at generating student buy-in for how I run my courses.

Thanks for joining in on this discussion!

5. I’ve been starting to put some thought into sbg at the intro physics level. My test run this semester with just 9 students in an advanced course is teaching me a lot but I know not all of it will transfer to a much bigger class.

One example is the amount of material we cover in an intro course. For this semester we’re only covering 9 chapters and I only have 2-4 standards per chapter. This amounts to roughly one standard for every “lecture” period. This semester I’ve already got two or three students who’ve fallen quite a ways behind. In an intro course we usually cover 18 chapters and I’d like to have, again, something like 2-4 standards per chapter. I’m not sure they or I could keep up with that rate.

Of course that still doesn’t deal with your excellent point about students needing to be able to do physics problems. This semester a student can show me they can model a complete physical system for a single standard. In intro physics that would be like the highest level problems at the end of a chapter. Students would have to realize that they should bang out a bunch of the easier ones to practice for such an assessment. I know from experience that students never believe lines like that.

One thing that has me thinking are the goal-less problems that Kelly O’Shea has blogged about (http://kellyoshea.wordpress.com/2011/03/05/goal-less-problems/). I think those would work well in class and could make interesting standards assessments.

Thanks, Joss, for this opportunity to brainstorm some future planning. The conversation really helps!

• Joss Ives says:

Hey Andy. These conversations about SBG and close-to-home implementations are really informative and have allowed for a ton of reflection on my part. I’m super glad I got involved in the conversation. Despite being a person that has a very solid set of explicit learning goals for each of my courses (which is going to really help when I attempt an SBG implementation), all of these conversations are helping me rethink assessments in my own traditionally-graded courses as well.

Looking over my own intro physics courses, I have close to 100 learning goals in each term, which works out to an insane rate of 8ish standards per week if they were converted straight over. With learning goals in a traditionally-graded course, you risk overwhelming the students with too many specific ones, but at the same time want to be as specific as possible so that they understand the types of questions you will ask. With standards, it seems like you want to take about a half-step back and wave the generalization wand to make it so that a given standard can be assessed multiple ways. The loss of specificity benefits them in that they can demonstrate mastery of a standard in multiple ways. It will/would certainly be an interesting task to rework my existing 100 learning goals to 30-40 standards that still capture the outcomes of the original learning goals.

I too have been thinking about Kelly O’Shea’s goal-less problems in terms of a way to assess “choose which concepts apply” or “build an appropriate model” types of standards. On the other hand, I’m not sure that these goal-less problems are ideal for “choose which concepts/models are most appropriate” types of assessments since it is often the goal that dictates the “most appropriate” part. You can tackle most mechanics problems using dynamics or energy, but often one makes the problem much simpler than the other. The goal-less problems can be used to have them show all sorts of interesting things when dynamics or energy concepts are applied, but don’t seem to lend themselves to this decision-making part of problem solving. Of course I have never actually tried using them so I could be completely mistaken. I certainly agree that they have the potential to be a good tool to extract and assess certain pieces of the problem solving process.

• Kelly O'Shea says:

I think the goal-less problems take on two different characters. In the practice realm, students are using them as sandboxes to play around and start finding the edges of the new tools and concepts that they are creating. In the assessment realm, students use them as performance opportunities to demonstrate mastery as best they can. What I’ve seen from my most skilled students (in both my “honors” and “regular” classes) is that they are purposeful about trying to do everything they can think to do with the given information. If there is a particular goal that I have in mind that students don’t try tackling in their work, I think it is because: (a) they ran out of time, (b) they did not know enough physics to have come up with that goal, or (c) that wasn’t actually a very interesting goal, so they didn’t consider it. Less skilled students, for the most part, seem to try to do the same, but with less success (for obvious reasons).

I think the main aspect of a regular question that you would miss out on by giving a goal-less problem is seeing who can give the most concise and time-efficient solution. And maybe that is a good thing to assess, but I’m not very concerned with that in a first-year physics class. Is being able to come up with an answer to a physics question the quickest a valuable skill? Probably the only kids who think so are the ones who have already taken a whole year of physics at a previous school (the same ones who think only the answer matters and fight over showing any work).

Anyway, the thing you pick up is being able to see what skills a student has no matter what their skill level. With a regular problem, you often can see what the best students can do, but students who can’t answer it might not show you evidence of anything that they know. Remove the intimidating goal, and they can (and as far as I can tell, will) go as far as they can in analyzing a problem. And you can be excited with them that they are finally drawing kinematics graphs in a correct and consistent way, then start to push them toward the next step.

Anyway, I have now rambled on quite a bit, and I’m not sure that I was even addressing what you actually said in your comment above. So probably best for me to stop here!

• Joss Ives says:

Another benefit of the goal-less that just occurred to me (for SBG and standard grading) is related to having quizzes based on homework (practice or graded). I think a common type of way to have a quiz question based on a homework problem is to repeat the question with different numbers or a slightly different setup or to make a question that is as close as possible in spirit to the original question. I think this tends to reward algorithm memorization as opposed to assessing conceptual understanding and problem solving. On the other hand, a goal-less problem based on a regular homework problem can take a situation with which the student is already familiar and ask them to expand on what they did in the homework or to take it in completely different directions.

“Is being able to come up with an answer to a physics question the quickest a valuable skill?”

I’m thinking of this as more of a synthesis or novice vs. expert framework issue. Do they see all of the tools at their disposal that they can bring to bear on the problem? Can they weigh the strengths and weakness of different approaches? These are pretty high-level skills that I don’t think are very easy to properly assess. I’m really trying to get a handle on how the highest-level skills we could expect intro students to have might be assessed (but haven’t made any progress yet!). I feel like I can vaguely see a path to assess these skills through regular problems that is not available from the goal-less problems, a path that requires the goal to focus the decision-making discussed above. But I don’t feel like I have ever actually done a good job of assessing these skills through regular problems so this whole discussion is based on abstracts…

“Anyway, the thing you pick up is being able to see what skills a student has no matter what their skill level. With a regular problem, you often can see what the best students can do, but students who can’t answer it might not show you evidence of anything that they know.”

This is the truest of true statements and is one of the things I find most frustrating with problems on exams. If the student can’t put together the pieces well enough to make any headway on a problem, you have no idea what they know or don’t for the concepts at hand.

6. I have a problem similar to the physics problem one, but worse. All the assessments in my courses for about the past 10 years have been “projects”, generally computer programs to design and write, week-long library research papers, or quarter-long reports on the student’s own research (which in turn often involves writing computer programs).

I spent a fair amount of last summer trying to figure out how to apply SBG and finally gave up. What I want to do in these graduate-level and senior courses is to assess the students’ ability to synthesize concepts and produce coherent reports of research. There were too many deeply entangled threads for me to separate things out into distinct standards. I could provide detailed feedback and allow redoing work until it was up to snuff (I’ve been doing that for years), but I couldn’t disaggregate.

Part of the problem is that at the level I’m interested in evaluating the students, the smallest problems I can come up with take a week to do and about 4 weeks to design. This means that I can’t come up with multiple assessments: it has taken me years to build up a stock of enough problems.

• Mylene says:

Hi GSWP, after having hashed this out at length on your blog, I realize I don’t know the answer to this: aren’t your multiple resubmissions in fact multiple assessments? Just a thought.

• I don’t think of them as multiple assessments, as the students are resubmitting programs or papers on exactly the same assignment after having gotten feedback. There is no independence between submissions, as I would expect from multiple assessments.

(To get pedantic, what you want is conditional independence, conditioned on the thing you are trying to measure. Resubmission of the same assignment doesn’t come close to that.)

• Joss Ives says:

@gswp – Be warned, spit-balling is about to go down.

I teach an upper-division physics lab course (commonly known as Advanced Lab) which features many of the same issues you are discussing in terms of trying to move to an SBG implementation. I have the students write journal articles based on 2 of the 3 large labs that they complete. I have them keep a proper lab logbook (which is graded) and they also do an oral presentation for one of their labs at the end. I let them resubmit their journal articles as many times as they like until they score over 75% using this rubric, after which they can resubmit one last time. Their mark for their paper is then 25%/75% for their first/final submission so that they don’t hand in the dog’s breakfast for their first version of the article.

So my rubric has 6 or 7 categories which could really transformed into maybe twice as many student-centric standards if one would like to. Part of the point is to get them to have written, in the end, a journal-quality article. So the revision process makes much more sense than trying to individually assess any standards that I could pull out of the rubric. But I think that this process is somewhat in the spirit of SBG. Even if it is not true reassessment, students are able to have their paper reassessed for an improved grade. There are reasonably well-defined criteria that can be used to assess their mastery of various aspects of writing a quality journal article.

I think there are a bunch of things that happen in this class for which standards could be created and slightly more SBGish standards brought to bear.

With the student presentations, I would say that there are three possible types of reassessment available. They could revise and redo their original presentation (not true reassessment as you point out), they could ask to have their question and answer part of their presentation reassessed (basically a second oral assessment based on their lab), or they could do a new presentation based on a different lab. For all three of these cases I would be happy to have the reassessment score replace the original.

The one thing which I did not work into my previous assessment scheme was how to assess their hand-on time in the lab: how good their design and execution of experiments was, their understanding of what they were up to as they went along, their ability to communicate to me what they were up to or how they were stuck. My thought is that next year when I teach this course again I will have the groups present to me snapshots of where they are at and what they have accomplished in the same way that a grad or co-op student would do at weekly meetings for the research group with which they are working. I could develop a mitt-full of standards that could be assessed through this type of interaction and the natural progress of each meeting would probably lend itself to assessing a few of these standards. If students want to be reassessed, they can spend some time sorting out some of the issues related to standards on which they performed poorly and then request a supplemental meting or even make those things the focus of the next scheduled meeting.

Back to your quandary. I think this model of a weekly “research group” meeting with students based on their projects might be a way to assess many standards outside of the finished product of the project that they are working on. I don’t think you could use a pure “last assessment is your score” type of system here because it will be very easy to catch them at a time when they are quite confused when they have moved onto a new project, but perhaps “research group meeting”-assessed standards for each project could have their own scores with reassessment available within the meeting cycle of that project.

7. This topic seems like a great idea for an organizing theme for our initial physicsed chat on Wednesday. What do you think?

8. Joss Ives says:

I just came across Jason Buell’s “Playing the Whole Game” post that discusses using 10-20 topic-level standards instead of 60+ narrow standards to be able to assess as higher proficiencies of the same standard some of the higher-order skill demonstrations such as the interplay between different skills and the “pick the best tool for the job” type of tasks.

9. […] The Physics Problem and Standards-Based Grading—Although it’s not what I’m looking for, it does consider the question, “How does one assess solving problems?” […]