Yesterday, I wrote:
We need assessment systems that work for the student first and everyone else second.
Assessments support evaluation, criticism and ranking (Wolff). That’s what it does and, in many cases, that also constitutes a lot of why we do it. But who are we doing it for?
I’ve reflected on the dual nature of evaluation, showing a student her or his level of progress and mastery while also telling us how well the learning environment is working. In my argument to reduce numerical grades to something meaningful, I’ve asked what the actual requirement is for our students, how we measure mastery and how we can build systems to provide this.
But who are the student’s grades actually for?
In terms of ranking, grades allow people who are not the student to place the students in some order. By doing this, we can award awards to students who are in the awarding an award band (repeated word use deliberate). We can restrict our job interviews to students who are summa cum laude or valedictorian or Dean’s Merit Award Winner. Certain groups of students, not all, like to define their progress through comparison so there is a degree of self-ranking but, for the most part, ranking is something that happens to students.
Criticism, in terms of providing constructive, timely feedback to assist the student, is weakly linked to any grading system. Giving someone a Fail grade isn’t a critique as it contains no clear identification of the problems. The clear identification of problems may not constitute a fail. Often these correlate but it’s weak. A student’s grades are not going to provide useful critique to the student by themselves. These grades are to allow us to work out if the student has met our assessment mechanisms to a point where they can count this course as a pre-requisite or can be awarded a degree. (Award!)
Evaluation is, as noted, useful to us and the student but a grade by itself does not contain enough record of process to be useful in evaluating how mastery goals were met and how the learning environment succeeded or failed. Competency, when applied systematically, does have a well-defined meaning. A passing grade does not although there is an implied competency and there is a loose correlation with achievement.
Grades allow us to look at all of a student’s work as if this one impression is a reflection of the student’s involvement, engagement, study, mistakes, triumphs, hopes and dreams. They are additions to a record from which we attempt to reconstruct a living, whole being.
Grades are the fossils of evaluation.
Grades provide a mechanism for us, in a proxy role as academic archaeologist, to classify students into different groups, in an attempt to project colour into grey stone, to try and understand the ecosystem that such a creature would live in, and to identify how successful this species was.
As someone who has been a student several times in my life, I’m aware that I have a fossil record that is not traditional for an academic. I was lucky to be able to place a new imprint in the record, to obscure my history as a much less successful species, and could then build upon it until I became an ACADEMIC TYRANNOSAURUS.
But I’m lucky. I’m privileged. I had a level of schooling and parental influence that provided me with an excellent vocabulary and high social mobility. I live in a safe city. I have a supportive partner. And, more importantly, at a crucial moment in my life, someone who knew me told me about an opportunity that I was able to pursue despite the grades that I had set in stone. A chance came my way that I never would have thought of because I had internalised my grades as my worth.
Let’s look at the fossil record of Nick.
My original GPA fossil, encompassing everything that went wrong and right in my first degree, was 2.9. On a scale of 7, which is how we measure it, that’s well below a pass average. I’m sharing that because I want you to put that fact together with what happened next. Four years later, I started a Masters program that I finished with a GPA of 6.4. A few years after the masters, I decided to go and study wine making. That degree was 6.43. Then I received a PhD, with commendation, that is equivalent to GPA 7. (We don’t actually use GPA in research degrees. Hmmm.) If my grade record alone lobbed onto your desk you would see the desiccated and dead snapshot of how I (failed to) engage with the University system. A lot of that is on me but, amazingly, it appears that much better things were possible. That original grade record stopped me from getting interviews. Stopped me from getting jobs. When I was finally able to demonstrate the skills that I had, which weren’t bad, I was able to get work. Then I had the opportunity to rewrite my historical record.
Yes, this is personal for me. But it’s not about me because I wasn’t trapped by this. I was lucky as well as privileged. I can’t emphasise that enough. The fact that you are reading this is due to luck. That’s not a good enough mechanism.
Too many students don’t have this opportunity. That impression in the wet mud of their school life will harden into a stone straitjacket from which they may never escape. The way we measure and record grades has far too much potential to work against students and the correlation with actual ability is there but it’s not strong and it’s not always reliable.
The student you are about to send out with a GPA of 2.9 may be competent and they are, most definitely, more than that number.
The recording of grades is a high-loss storage record of the student’s learning and pathway to mastery. It allows us to conceal achievement and failure alike in the accumulation of mathematical aggregates that proxy for competence but correlate weakly.
We need assessment systems that work for the student first and everyone else second.
From the previous post, I asked how many times a student has to perform a certain task, and to which standard, that we become confident that they can reliably perform the task. In the Vocational Education and Training world this is referred to as competence and this is defined (here, from the Western Australian documentation) as:
In VET, individuals are considered competent when they are able to consistently apply their knowledge and skills to the standard of performance required in the workplace.
How do we know if someone has reached that level of competency?
We know whether an individual is competent after they have completed an assessment that verifies that all aspects of the unit of competency are held and can be applied in an industry context.
The programs involved are made up of units that span the essential knowledge and are assessed through direct observation, indirect measurements (such as examination) and in talking to employers or getting references. (And we have to be careful that we are directly measuring what we think we are!)Hang on. Examinations are an indirect measurement? Yes, of course they are here, we’re looking for the ability to apply this and that requires doing rather than talking about what you would do. Your ability to perform the task in direct observation is related to how you can present that knowledge in another frame but it’s not going to be 1:1 because we’re looking at issues of different modes and mediation.
But it’s not enough just to do these tasks as you like, the specification is quite clear in this:
It can be demonstrated consistently over time, and covers a sufficient range of experiences (including those in simulated or institutional environments).
I’m sure that some of you are now howling that many of the things that we teach at University are not just something that you do, there’s a deeper mode of thinking or something innately non-Vocational about what is going on.
And, for some of you, that’s true. Any of you who are asking students to do anything in the bottom range of Bloom’s taxonomy… I’m not convinced. Right now, many assessments of concepts that we like to think of as abstract are so heavily grounded in the necessities of assessment that they become equivalent to competency-based training outcomes.
The goal may be to understand Dijkstra’s algorithm but the task is to write a piece of code that solves the algorithm for certain inputs, under certain conditions. This is, implicitly, a programming competency task and one that must be achieved before you can demonstrate any ability to show your understanding of the algorithm. But the evaluator’s perspective of Dijkstra is mediated through your programming ability, which means that this assessment is a direct measure of programming ability in language X but an indirect measure of Dijkstra. Your ability to apply Dijkstra’s algorithm would, in a competency-based frame, be located in a variety of work-related activities that could verify your ability to perform the task reliably.
All of my statistical arguments on certainty from the last post come back to a simple concept: do I have the confidence that the student can reliably perform the task under evaluation? But we add to this the following: Am I carrying out enough direct observation of the task in question to be able to make a reliable claim on this as an evaluator?
There is obvious tension, at modern Universities, between what we see as educational and what we see as vocational. Given that some of what we do falls into “workplace skills” in a real sense, although we may wish to be snooty about the workplace, why are we not using the established approaches that allow us to actually say “This student can function as an X when they leave here?”
If we want to say that we are concerned with a more abstract education, perhaps we should be teaching, assessing and talking about our students very, very differently. Especially to employers.
I’m at the Australasian Computer Science Week at the moment and I’m dividing my time between attending amazing talks, asking difficult questions, catching up with friends and colleagues and doing my own usual work in the cracks. I’ve talked to a lot of people about my ideas on assessment (and beauty) and, as always, the responses have been thoughtful, challenging and helpful.
I think I know what the basis of my problem with assessment is, taking into account all of the roles that it can take. In an earlier post, I discussed Wolff’s classification of assessment tasks into criticism, evaluation and ranking. I’ve also made earlier (grumpy) notes about ranking systems and their arbitrary nature. One of the interesting talks I attended yesterday talked about the fragility and questionable accuracy of post-University exit surveys, which are used extensively in formal and informal rankings of Universities, yet don’t actually seem to meet many of the statistical or sensible guidelines for efficacy we already have.
But let’s put aside ranking for a moment and return to criticism and evaluation. I’ve already argued (successfully I hope) for a separation of feedback and grades from the criticism perspective. While they are often tied to each other, they can be separated and the feedback can still be useful. Now let’s focus on evaluation.
Remind me why we’re evaluating our students? Well, we’re looking to see if they can perform the task, apply the skill or knowledge, and reach some defined standard. So we’re evaluating our students to guide their learning. We’re also evaluating our students to indirectly measure the efficacy of our learning environment and us as educators. (Otherwise, why is it that there are ‘triggers’ in grading patterns to bring more scrutiny on a course if everyone fails?) We’re also, often accidentally, carrying out an assessment of the innate success of each class and socio-economic grouping present in our class, among other things, but let’s drill down to evaluating the student and evaluating the learning environment. Time for another thought experiment.
Thought Experiment 2
There are twenty tasks aligned with a particularly learning outcome. It’s an important task and we evaluate it in different ways but the core knowledge or skill is the same. Each of these tasks can receive a ‘grade’ of 0, 0.5 or 1. 0 means unsuccessful, 0.5 is acceptable, 1 is excellent. Student A attempts all tasks and is acceptable in 19, unsuccessful in 1. Student B attempts the first 10 tasks, receives excellent in all of them and stops. Student C sets up a pattern of excellent,unsuccessful, excellent, unsuccessful.. and so on to receive 10 “Excellent”s and 10 “unsuccessful”s. When we form an aggregate grade, A receives 47.5%, B receives 50% and C also receives 50%. Which of these students is the most likely to successfully complete the task?
This framing allows us to look at the evaluation of the student in a meaningful way. “Who will pass the course?” is not the question we should be asking, it’s “Who will be able to reliably demonstrate mastery of the skills or knowledge that we are imparting.” Passing the course has a naturally discrete attention focus: focus on n assignments and m exams and pass. Continual demonstration of mastery is a different goal. This framing also allows us to examine the learning environment because, without looking at the design, I can’t tell you if B and C’s behaviour is problematic or not.
A has undertaken the most tasks to an acceptable level but an artefact of grading (or bad luck) has dropped the mark below 50%, which would be a fail (aggregate less than acceptable) in many systems. B has performed excellently on every task attempted but, being aware of the marking scheme, optimising and strategic behaviour allows this student to walk away. (Many students who perform at this level wouldn’t, I’m aware, but we’re looking at the implications of this.) C has a troublesome pattern that provides the same outcome as B but with half the success rate.
Before we answer the original question (which is most likely to succeed), I can nominate C as the most likely to struggle because C has the most “unsuccessful”s. From a simple probabilistic argument, 10/20 success is worse than 19/20. It’s a bit tricker comparing 10/10 and 10/20 (because of confidence intervals) but 10/20 has an Adjusted Wald range of +/- 20% and 10/10 is -14%, so the highest possible ‘real’ measure for C is 14/20 and the lowest possible ‘real’ measure for B is (scaled) 15/20, so they don’t overlap and we can say that B appears to be more successful than C as well.
From a learning design perspective, do our evaluation artefacts have an implicit design that explains C’s pattern? Is there a difference we’re not seeing? Taking apart any ranking of likeliness to pass our evaluatory framework, C’s pattern is so unusual (high success/lack of any progress) that we learn something immediately from the pattern, whether it’s that C is struggling or that we need to review mechanisms we thought to be equivalent!
But who is more likely to succeed out of A and B? 19/20 and 10/10 are barely distinguishable in statistical terms! The question for us now is how many evaluations of a given skill or knowledge mastery are required for us to be confident of competence. This totally breaks the discrete cramming for exams and focus on assignment model because all of our science is built on the notion that evidence is accumulated through observation and the analysis of what occurred, in order to be able to construct models to predict future behaviour. In this case, our goal is to see if our students are competent.
I can never be 100% sure that my students will be able to perform a task but what is the level I’m happy with? How many times do I have to evaluate them at a skill so that I can say that x successes in y attempts constitutes a reliable outcome?
If we say that a student has to reliably succeed 90% of the time, we face the problem that just testing them ten times isn’t enough for us to be sure that they’re hitting 90%.
But the level of performance we need to be confident is quite daunting. By looking at some statistics, we can see that if we provide a student with 150 opportunities to demonstrate knowledge and they succeed at this 143 times, then it is very likely that their real success level is at least 90%.
If we say that competency is measured by a success rate that is greater than 75%, a student who achieves 10/10 has immediately met that but even succeeding at 9/9 doesn’t meet that level.
What this tells us (and reminds us) is that our learning environment design is incredibly important and it must start from a clear articulation of what success actually means, what our goals are and how we will know when our students have reached that point.
There is a grade separation between A and B but it’s artificial. I noted that it was hard to distinguish A and B statistically but there is one important difference in the lower bound of their confidence interval. A is less than 75%, B is slightly above.
Now we have to deal with the fact that A and B were both competent (if not the same) for the first ten tests and A was actually more competent than B until the 20th failed test. This has enormous implications for we structure evaluation, how many successful repetitions define success and how many ‘failures’ we can tolerate and still say that A and B are competent.
Confused? I hope not but I hope that this is making you think about evaluation in ways that you may not have done so before.
There was a time before graphics dominated the way that you worked with computers and, back then, after punchcards and before Mac/Windows, the most common way of working with a computer was to use the Command Line Interface (CLI). Many of you will have seen this, here’s Terminal from the Mac OS X, showing a piece of Python code inside an editor.
Rather than use a rich Integrated Development Environment, where text is highlighted and all sorts of clever things are done for me, I would run some sort of program editor from the command line, write my code, close that editor and then see what worked.
At my University, we almost always taught Computer Science using command line tools, rather than rich development environments such as Eclipse or the Visual Studio tools. Why? The reasoning was that the CLI developed skills required to write code, compile it, debug it and run it, without training students into IDE-provided shortcuts. The CLI was the approach that would work anywhere. That knowledge was, as we saw it, fundamental.
But, remember that Processing example? We clearly saw where the error was. This is what a similar error looks like for the Java programming language in a CLI environment.
Same message (and now usefully on the right line because 21st Century) but it is totally divorced from the program itself. That message has to give me a line number (5) in the original program because it has no other way to point to the problem.
And here’s the problem. The cognitive load increases once we separate code and errors. Despite those Processing errors looking like the soft option, everything we know about load tells us that students will find fixing their problems easier if they don’t have to mentally or physically switch between code and error output.
Everything I said about CLIs is still true but that’s only a real consideration if my students go out into the workplace and need some CLI skills. And, today, just about every workplace has graphics based IDEs for production software. (Networking is often an exception but we’ll skip over that. Networking is special.)
The best approach for students learning to code is that we don’t make things any harder than we need to. The CLI approach is something I would like students to be able to do but my first task is to get them interested in programming. Then I have to make their first experiences authentic and effective, and hopefully pleasant and rewarding.
I have thought about this for years and I started out as staunchly CLI. But as time goes by, I really have to wonder whether a tiny advantage for a small number of graduates is worth additional load for every new programmer.
And I don’t think it is worth it. It’s not fair. It’s the opposite of equitable. And it goes against the research that we have on cognitive load and student workflows in these kinds of systems. We already know of enough load problems in graphics based environments if we make the screens large enough, without any flicking from one application to another!
You don’t have to accept my evaluation model to see this because it’s a matter of common sense that forcing someone to unnecessarily switch tasks to learn a new skill is going to make it harder. Asking someone to remember something complicated in order to use it later is not as easy as someone being able to see it when and where they need to use it.
The world has changed. CLIs still exist but graphical user interfaces (GUIs) now rule. Any of my students who needs to be a crack programmer in a text window of 80×24 will manage it, even if I teach with all IDEs for the entire degree, because all of the IDEs are made up of small windows. Students can either debug and read error messages or they can’t – a good IDE helps you but it doesn’t write or fix the code for you, in any deep way. It just helps you to write code faster, without having to wait and switch context to find silly mistakes that you could have fixed in a split second in an IDE.
When it comes to teaching programming, I’m not a CLI guy anymore.
Earlier, I split the evaluation resources of a course into:
- E1 (the lecturer and course designer),
- E2 (human work that can be based on rubrics, including peer assessment and casual markers),
- E3 (complicated automated evaluation mechanisms)
- E4 (simple automated evaluation mechanisms, often for acceptance testing)
E1 and E2 everyone tends to understand, because the culture of Prof+TA is widespread, as is the concept of peer assessment. In a Computing Course, we can define E3 as complex marking scripts that perform amazing actions in response to input (or even carry out formal analysis if we’re being really keen), with E4 as simple file checks, program compilation and dumb scripts that jam in a set of data and see what comes out.
But let’s get back to my first year, first exposure, programming class. What I want is hands-on, concrete, active participation and constructive activity and lots of it. To support that, I want the best and most immediate feedback I can provide. Now I can try to fill a room with tutors, or do a lot of peer work, but there will come times when I want to provide some sort of automated feedback.
Given how inexperienced these students are, it could be a quite a lot to expect them to get their code together and then submit it to a separate evaluation system, then interpret the results. (Remember I noted earlier on how code tracing correlates with code ability.)
Thus, the best way to get that automated feedback is probably working with the student in place. And that brings us to the Integrated Development Environment (IDE). An IDE is an application that provides facilities to computer programmers and helps them to develop software. They can be very complicated and rich (Eclipse), simple (Processing) or aimed at pedagogical support (Scratch, BlueJ, Greenfoot et al) but they are usually made up of a place in which you can assemble code (typing or dragging) and a set of buttons or tools to make things happen. These are usually quite abstract for early programmers, built on notional machines rather than requiring a detailed knowledge of hardware.
Even simple IDEs will tell you things that provide immediate feedback. We know how these environments can have positive reception, with some demonstrated benefits, although I recommend reading Sorva et al’s “A Review of Generic Program Visualization Systems for Introductory Programming Education” to see the open research questions. In particular, people employing IDEs in teaching often worry about the time to teach the environment (as well as the language), software visualisations, concern about time on task, lack of integration and the often short lifespan of many of the simpler IDEs that are focused on pedagogical outcomes. Even for well-established systems such as BlueJ, there’s always concern over whether the investment of time in learning it is going to pay off.
In academia, time is our currency.
But let me make an aesthetic argument for IDEs, based on the feedback that I’ve already put into my beautiful model. We want to maximise feedback in a useful way for early programmers. Early programmers are still learning the language, still learning how to spell words, how to punctuate, and are building up to a grammatical understanding. An IDE can provide immediate feedback as to what the computer ‘thinks’ is going on with the program and this can help the junior programmer make immediate changes. (Some IDEs have graphical representations for object systems but we won’t discuss these any further here as the time to introduce objects is a subject of debate.)
Now there’s a lot of discussion over the readability of computer error messages but let me show you an example. What’s gone wrong in this program?
See where that little red line is, just on the end of the first line? Down the bottom there’s a message that says “missing a semicolon”. In the Processing language, almost all lines end with a “;” so that section of code should read:
Did you get that? That missing semicolon problem has been an issue for years because many systems report the semicolon missing on the next line, due to the way that compilers work. Here, Processing is clearly saying: Oi! Put a semi-colon on the red squiggle.
I’m an old programmer, who currently programs in Java, C++ and Processing, so typing “;” at the end of a line is second nature to me. But it’s an easy mistake for a new programmer to make because, between all of the ( and the ) and the , and the numbers and the size and the rect… what do I do with the “;”?
The Processing IDE is functioning in at least an E4 mode: simple acceptance testing that won’t let anything happen until you fix that particular problem. It’s even giving you feedback as to what’s wrong. Now this isn’t to say that it’s great but it’s certainly better than a student sitting there with her hand up for 20 minutes waiting for a tutor to have the time to come over and say “Oh, you’re missing a semicolon.”
We don’t want shotgun coding, where random fixes and bashed-in attempts are made desperately to solve a problem. We want students to get used to getting feedback on how they’re going and using this to improve what they do.
Because of Processing’s highly visual mode, I think it’s closer to E3 (complex scripting) in many ways because it can tell you if it doesn’t understand what you’re trying to do at all. Beyond just not doing something, it can clearly tell you what’s wrong.
But what if it works and then the student puts something up on the screen, a graphic of some sort and it’s not quite right? Then the student has started to become their own E2, evaluating what has happened in response to the code and using human insight to address the shortfall and make changes. Not as an expert but, with support and encouragement, a developing expertise.
Feedback is good. Immediacy is good. Student involvement is good. Code tracing is linked to coding ability. A well-designed IDE can be simple and engage the student to an extent that is potentially as high as E2, although it won’t be as rich, without using any other human evaluation resources. Even if there is no other benefit, the aesthetic argument is giving us a very strong nudge to adopt an appropriate IDE.
Maybe it’s time to hang up the command line and live in a world where IDEs can help us to get things done faster, support our students better and make our formal human evaluation resources go further.
What do you think?
If we want to give feedback, then the time it takes to give feedback is going to determine how often we can do it. If the core of our evaluation is feedback, rather than some low-Bloom’s quiz-based approach giving a score of some sort, then we have to set our timelines to allow us to:
- Get the work when we are ready to work on it
- Undertake evaluation to the required level
- Return that feedback
- Do this at such a time that our students can learn from it and potentially use it immediately, to reinforce the learning
A commenter asked me how I actually ran large-scale assessment. The largest class I’ve run detailed feedback/evaluation on was 360 students with a weekly submission of a free-text (and graphics) solution to a puzzle. The goal was to have the feedback back within a week – prior to the next lecture where the solution would be delivered.
I love a challenge.
This scale is, obviously, impossible for one person to achieve reliably (we estimated it as at least forty hours of work). Instead, we allocated a marking team to this task, coordinated by the lead educator. (E1 and E2 model again. There was, initially, no automated capacity for this at the time although we added some later.)
Coordinating a team takes time. Even when you start with a rubric, free text answers can turn up answer themes that you didn’t anticipate and we would often carry our simple checks to make sure that things were working. But, looking at the marking time I was billed for (a good measure), I could run an entire cycle of this in three days, including briefing time, testing, marking, and oversight. But this is with a trained team, a big enough team, good conceptual design and a senior educator who’s happy to take a more executive role.
In this case, we didn’t give the students a chance to refactor their work but, if we had, we could have done this with a release 3 days after submission. To ensure that we then completed the work again by the ‘solution release’ deadline, we would have had to set the next submission deadline to only 24 hours after the feedback was released. This sounds short but, if we assume that some work has been done, then refactoring and reworking should take less time.
But then we have to think about the cost. By running two evaluation cycles we are providing early feedback but we have doubled our cost for human markers (a real concern for just about everyone these days).
My solution was to divide the work into two components. The first was quiz-based and could be automatically and immediately assessed by the Learning Management System, delivering a mark at a fixed deadline. The second part was looked at by humans. Thus, students received immediate feedback on part of the problem straight away (or a related problem) while they were waiting for humans.
But I’d be the first to admit that I hadn’t linked this properly, according to my new model. It does give us insight for a staged hybrid model where we buffer our human feedback by using either smart or dumb automated assessment component to highlight key areas and, better still, we can bring these forward to help guide time management.
I’m not unhappy with that early attempt at large-scale human feedback as the students were receiving some excellent evaluation and feedback and it was timely and valuable. It also gave me a lot of valuable information about design and about what can work, as well as how to manage marking teams.
I also realised that some courses could never be assessed the way that they claimed unless they had more people on task or only delivered at a time when the result wasn’t usable anymore.
How much time should we give students to rework things? I’d suggest that allowing a couple of days takes into account the life beyond Uni that many students have. That means that we can do a cycle in a week if we can keep our human evaluation stages under 2 days. Then, without any automated marking, we get 2 days (E1 or E2) + 2 days (student) + 2 days (second evaluation, possibly E2) + 1 day (final readjustment) and then we should start to see some of the best work that our students can produce.
Assuming, of course, that all of us can drop everything to slot into this. For me, this motivates a cycle closer to two to three weeks to allow for everything else that both groups are doing. But that then limits us to fewer than five big assessment items for a twelve week course!
What’s better? Twelve assessment items that are “submit and done” or four that are “refine and reinforce to best practice”? Is this even a question we can ask? I know which one is aesthetically pleasing, in terms of all of the educational aesthetics we’ve discussed so far but is this enough for an educator to be able to stand up to a superior and say “We’re not going to do X because it just doesn’t make any sense!”
What do you think?