Intro. [Recording date: March 3, 2025.]
Russ Roberts: Today is March 3, 2025, and before we dive into today’s discussion, I’d like to reveal the results of our annual poll showcasing your favorite episodes from 2024. A heartfelt thank you to everyone who participated in the voting.
Here are the Top Ten:
- And tied for 1st place, two episodes:
Once again, thank you for your votes! We’ll provide links to all the episodes that made the Top 10. Also, don’t forget to check out our Favorites category to enjoy past years’ listener favorites.
Now, onto today’s guest, author and educational consultant, Daisy Christodoulou. Daisy last joined us in February 2025 to discuss Coase, the rules of the game, and the costs of perfection. Her Substack is titled: No More Marking. For our audience outside the UK, ‘marking’ is equivalent to ‘grading’ in the U.S.—and frankly, I have no idea what they call it elsewhere. Today, we will explore feedback in education and the potential role of AI [artificial intelligence] in providing that feedback. Welcome back to EconTalk, Daisy!
Daisy Christodoulou: It’s great to be here again, Russ.
2:49
Russ Roberts: We’ll structure our conversation around some of your recent essays featured in your Substack, No More Marking. Just to clarify, is it correct that marking refers to grading in the UK?
Daisy Christodoulou: Yes, absolutely. In a technical sense, marking in the UK involves assigning a number or grade to a student’s work. However, I should mention that not everyone agrees with that definition, so perhaps we’re misnamed, but it’s certainly a catchy term.
Russ Roberts: Let’s begin by discussing feedback. You describe it as functioning like a thermometer or a thermostat. Could you elaborate on that analogy?
Daisy Christodoulou: Certainly! In the realm of education, feedback is a topic that resonates with both teachers and students, emphasizing the necessity of knowing how one is performing in order to improve.
However, feedback extends beyond education, rooted in concepts from cybernetics and information technology that focus on how systems adjust based on input. To illustrate, consider the distinction between a thermometer and a thermostat: a thermometer merely measures temperature, while a thermostat actively adjusts the environment based on that measurement.
In educational terms, feedback can either be a simple measurement—like assigning a grade—or it can provide actionable insights aimed at helping students reach their goals. For example, if a student’s temperature reading is 18 degrees Celsius and the desired goal is 20, the thermostat kicks in to raise the temperature. Similarly, the aim of educational feedback is to help students progress towards fluency in reading, insightful inference, and coherent writing, closing the gap between their current abilities and those goals.
6:22
Russ Roberts: Indeed, the challenge arises when a student receives a grade of C. Some may feel satisfied, thinking, ‘Great, I did better than expected!’ while others might feel compelled to improve. Yet, a simple letter grade doesn’t reveal how to enhance their performance; it merely indicates room for growth.
Moreover, it doesn’t pinpoint specific weaknesses. A student could be excelling in some areas while struggling in others, yet their overall grade remains a C. Thus, the advice to simply ‘try harder’ often lacks substance.
I’d like to differentiate between two feedback types: one related to writing, particularly essays, and the other concerning content knowledge, which can range from factual accuracy to applying learned concepts in new contexts. For instance, I recall a humorous piece of feedback I received as a teacher when a student rated my course a 1 out of 5, claiming, ‘This course is unfair. Professor Roberts expects us to apply the material to things we’ve never seen before.’ Ironically, that was precisely the course’s objective.
Do you agree that these two feedback categories—essay-related suggestions and content knowledge evaluation—differ significantly?
Daisy Christodoulou: Yes, I understand your perspective. Each type of feedback could manifest as comments written at the bottom of a student’s work. I focus on the technical accuracy of writing, such as sentence structure and grammar. For example, when I taught a play called “An Inspector Calls,” students often confused characters Gerald and Eric. While that may not have been the worst mistake, it could lead to confusion and hinder understanding, illustrating the importance of clear feedback.
Then, as you mentioned, there are higher-order skills that elevate writing quality, which are crucial for comprehensive feedback. Is that what you meant?
Russ Roberts: Yes, I’m reflecting on the context of feedback I’ve encountered on my children’s papers and my own work. Often, comments like ‘Awkward’ or ‘Confusing’ appear in the margins, leaving students wondering, ‘What now?’ This contrasts with clearer feedback that identifies specific errors, such as confusing two different characters, prompting students to re-read and learn.
Daisy Christodoulou: Exactly. I see the distinction you’re making, but I would argue that written feedback may not be optimal for any of these categories. In fact, I believe written feedback can often be ineffective.
11:01
Russ Roberts: Let’s delve into essay feedback within educational contexts. You’re involved in numerous fascinating projects aimed at scaling grading and feedback. Picture a student submitting a mediocre essay. There are countless reasons it could fall flat: incorrect tone, factual errors, lack of flair—resulting in a grade that reflects mediocrity. The student is then left wondering, ‘How can I improve?’
Some feedback may suggest, ‘Try rewriting that awkward section,’ but if I opt not to provide written feedback, how do I assist the student?
Daisy Christodoulou: That’s a great question. Before we dive into that, let’s consider the age groups involved. My experience primarily lies with students aged 11 to 18, but I also work with younger students. I’ve even authored peer-reviewed papers where I received written feedback from reviewers. I believe the points I’m making apply across age ranges, with some nuances for different developmental stages.
The central argument I’d like to present is that prose is often ineffective for prompting action or improvement. This ties back to Michael Polanyi’s theory of tacit knowledge, which posits that there are actions we can perform but cannot articulate clearly. For instance, riding a bike is a skill that can’t be fully learned through reading alone; practical experience is necessary.
Russ Roberts: That’s a compelling point. Please, continue.
Daisy Christodoulou: To illustrate, Dylan Wiliam, a prominent figure in education and assessment, shares an example from a middle school classroom where students received detailed written feedback on a science project. One student was told, ‘You need to make your scientific inquiries more systematic.’ When Wiliam asked what that meant, the student replied, ‘I don’t know. If I had known how to be more systematic, I would have done that the first time.’
Russ Roberts: That’s quite revealing.
Daisy Christodoulou: Indeed, it’s a rather flippant response, but it highlights a critical issue. I’ve spent considerable time as a teacher providing feedback like that, often because it was deemed best practice.
Unfortunately, this approach can lead to what Wiliam calls ‘True But Useless’ feedback—like telling a comedian to tell funnier jokes without providing actionable steps for improvement.
Russ Roberts: I can see the value in that analogy. What do you suggest instead?
Daisy Christodoulou: This leads to my next point: to bridge the gap between current performance and goals, we need to move beyond written feedback. Wiliam describes the need for a ‘recipe for action,’ emphasizing that sports often provide clearer pathways to improvement than academic contexts.
For instance, if a student needs to enhance their inferential skills, understanding that this requires a broader vocabulary and background knowledge is essential. A practical approach might involve lessons on vocabulary expansion rather than merely redoing the writing assignment.
Russ Roberts: It seems you’re advocating for a more holistic approach to education.
Daisy Christodoulou: Exactly! In sports, when training for a marathon, you wouldn’t simply run marathons in every training session. Instead, you’d develop a training plan that includes various activities, such as strength training and conditioning, which may not resemble the act of running itself but are crucial for improving performance.
Similarly, in academia, achieving better writing skills can involve activities that don’t directly involve writing but contribute to overall improvement.
Russ Roberts: I find that perspective intriguing; however, I hold a slightly different view. The challenge with sports analogies is the measurable nature of improvement, which is often elusive in academic settings. A personal example I find enlightening concerns baseball: if a player struggles to hit, a coach may simply advise, ‘Keep your eye on the ball.’ While this sounds straightforward, it’s a complex task that requires practice to master.
Daisy Christodoulou: That’s an excellent example, indeed!
Russ Roberts: Thank you! Another analogy is in soccer, where young players often hear, ‘Keep your head up.’ Yet, many are so focused on improving their ball control that they’re unable to do so. Immediate feedback like ‘keep your head up’ can be counterproductive if they don’t have the necessary skills.
Daisy Christodoulou: Right! Instead of vague advice, targeted drills to improve specific skills are essential.
Russ Roberts: Precisely. I often joke about how parents yell ‘Don’t bunch up!’ from the sidelines during youth games, but it’s largely ineffective. However, when it comes to academic feedback, I find that encouraging students simply to ‘read more and write more’ can be quite sufficient. Practice fosters improvement in writing; I know I became a better writer by writing regularly.
While I agree with you that structured guidance is essential, I think the key lies in fostering critical thinking. At Shalem College, we emphasize reading complex texts and engaging in discussions that challenge students to think deeply. This process, while still somewhat mysterious, undoubtedly leads to cognitive growth.
Daisy Christodoulou: I concur with some of your points, but I see differences based on age and ability levels. I deeply value reading and writing, but I caution against oversimplifying the solution for younger students as merely ‘read more’ or ‘write more.’
For example, teachers often note that many students need to build their writing endurance. Pressuring them to write excessively too soon can lead to the development of poor habits, which can be difficult to correct later.
Russ Roberts: That’s a fair point.
Daisy Christodoulou: Exactly. If a student pours their heart into a piece of writing but struggles with technical errors, it can hinder engagement with feedback, creating a significant challenge. Thus, addressing foundational skills before advancing is crucial.
While I appreciate the concept of ‘learning by doing,’ we must distinguish between productive and unproductive activities. For instance, while reading about marathons is helpful, it’s not a substitute for actual training. Similarly, we must ensure that students are engaging in meaningful activities that contribute to their overall skill development.
Russ Roberts: I agree with that distinction completely.
Daisy Christodoulou: Therefore, the responsibility of educators and curriculum designers is to develop effective models of progression. However, missteps can occur, and not all activities that diverge from the end goal will be beneficial.
For instance, if a student only reads about marathons without any physical training, they won’t be prepared for the race. It’s essential to balance knowledge acquisition with practical application.
