Popular in Professional

Business

The Allegory of the Favela: The Multifaceted Effects of Socioeconomic Mobility

University of Pennsylvania Leandro S. Pongeluppe

Architecture

Connecting Cities and Nature

University of Virginia Timothy Beatley

Human Rights Law

Human Rights and Corruption: Problems and Potential of Individualising a Systemic Problem

Max Planck Institute Anne Peters

Finance

Chinese Exchange Rate Policy: Lessons for Global Investors

University of California San Diego Michael Melvin

Law

A statistical approach to law school citation rankings

University of Virginia Joshua Fischman

Education

Firm Performance Pay as Insurance against Promotion Risk

Stockholm School of Economics Alvin Chen

Putting learners at the heart of the feedback process

View more insights in Professional - Education - Pedagogy

View more figures from UCL Institute of Education

View the page of Dylan Wiliam

Read the Study

Next in Education

Posthumanist new materialist pathways for reimagining music education research

Pamela Burnard | University of Cambridge

The Future of Great Universities

Jonathan R. Cole | Columbia University

Dylan Wiliam discusses research on feedback, its impact on task performance, peer feedback and why we need to provide feedback that improves the learner, and not only the work they submit for assessment.

Image courtesy of interviewee. November 14, 2023

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

This video discusses the concept of feedback, tracing its historical origins and its implications in both engineering and psychology. It highlights Norbert Wiener’s role during World War II in developing feedback mechanisms for anti-aircraft systems, leading to the feedback loop concept. The distinction between feedback in engineering (where it influences future actions) and in psychology (often simply informational) is emphasized.

Research by Kluger and DeNisi is cited, showing that while feedback generally improves performance, it can also hinder it in a significant percentage of cases. This video dichotomy raises concerns about the effectiveness of feedback strategies in educational contexts, where feedback should focus on enhancing the learner's understanding rather than just improving the output.

This video further explores the importance of timing and type of feedback, suggesting that delayed or evaluative feedback can sometimes be more beneficial than immediate, purely positive, or descriptive feedback. The relationship between the teacher and student, as well as the emotional context surrounding feedback, plays a crucial role in how feedback is received and acted upon.

It suggests that feedback should aim to empower students to take responsibility for their learning, moving towards self-assessment and self-regulation. This video underscores that effective feedback must foster trust and support, ultimately helping students recognize that effort can lead to improvement. Additionally, peer and self-assessment are addressed as tools for fostering learner independence, but caution is advised regarding their accuracy and the quality of feedback exchanged.

In conclusion, the essence of meaningful feedback lies in its application and the context within which it is delivered, emphasizing the need for nurturing relationships that motivate and guide students toward self-enhancement in their learning journeys.

Language:

Export as PDF

Export as PDF
Here are five key points from the video on feedback:

1. **Historical Convideo of Feedback**: The term "feedback" originated from engineering convideos, particularly from Norbert Wiener during World War II when he worked on anti-aircraft systems. He proposed the idea of using information about past performance (like the explosion of a shell) to improve future performance, thus establishing the concept of a feedback loop.

2. **Divergent Definitions in Psychology and Engineering**: In psychology, feedback is often defined broadly as any information to individuals about their performance. However, this contrasts with engineers who view feedback as actionable information that should lead to performance adjustments (e.g., a thermostat heating the room based on temperature readings).

3. **Research Findings on Feedback Effectiveness**: A significant review by Kluger and DeNisi found that while feedback generally improves performance, it may also worsen it in about 38% of cases. This highlights the complexity of feedback effects and warns against assuming that more feedback is always beneficial.

4. **Dependence on Feedback**: Feedback can lead to increased dependency, where students rely on external validation rather than developing internal motivation and self-regulation. Effective feedback should empower learners rather than just improving their immediate performance on tasks.

5. **Importance of Relationships and Convideo**: The effectiveness of feedback is moderated by the relationship between the student and the teacher and the convideo in which feedback is given. Understanding student beliefs about their abilities and fostering a supportive environment can enhance the constructive use of feedback. Teachers should aim to minimize emotional reactions to feedback to help students engage meaningfully with the improvement suggestions.

Overall, the discourse emphasizes the nuanced nature of feedback in educational settings and the importance of implementing it thoughtfully to truly benefit learners.

Export as PDF
Here are some key quotes from the video about feedback:

1. **Origin of Feedback:**
- “The first mention of feedback is in... mechanics of production... but I think the first use in what we think of feedback today was Norbert Wiener... developing range finders for anti-aircraft guns.”

2. **Feedback Loop Concept:**
- “The idea is the information collected about the explosion of the shell is used to improve the next time you fire it.”

3. **Challenges in Definitions:**
- “Psychologists define feedback as any information fed back to an individual about their task performance... that's not something engineers will recognize.”

4. **Effectiveness of Feedback:**
- “On average feedback did improve performance, but in 38% of the cases... feedback actually made things worse.”

5. **Risks of Feedback:**
- “In almost two out of every five cases, you're doing better off shutting up than giving the feedback.”

6. **Relationship to Performance:**
- “Even if feedback does make things better, it's not necessarily a good thing if the feedback makes the students more dependent on the feedback.”

7. **Focus on Learning:**
- “Too often feedback is about improving the work rather than improving the learner.”

8. **Usefulness of Feedback:**
- “Feedback is basically one-to-one tuition... we have to close the loop.”

9. **Emotional Response to Feedback:**
- “It's actually very difficult to use feedback intelligently if that emotion is triggered.”

10. **Impact of Mindset:**
- “If students believe that smart is not something you just are, but something you can get... then feedback is welcome.”

11. **Peer Assessment Considerations:**
- “If you choose to use peer assessment as a way of helping students improve, then you are still responsible for the quality of feedback they give each other.”

12. **Self-Assessment Goals:**
- “The goal of feedback is to empower learners to take more responsibility for their own learning.”

These quotes encapsulate the discussion about the nature of feedback, its effectiveness, the emotional dynamics involved, and the importance of fostering a growth mindset in learners.

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.145 --> 00:04.565
The first mention of feedback is in, me mechanics

00:04.585 --> 00:07.485
of production, with things like shuttles and,

00:08.065 --> 00:10.325
and me mechanical applications.

00:10.385 --> 00:14.685
But I think the first use in what we think of feedback today

00:15.465 --> 00:18.205
was Norbert Wiener, who during the second World War,

00:18.225 --> 00:22.165
was developing range finders for anti-aircraft, guns.

00:22.825 --> 00:25.645
And there the important thing is that your chances

00:25.645 --> 00:29.605
of hitting, an aircraft directly are very small.

00:29.905 --> 00:33.565
So what you do is you actually put a timer on the shell so

00:33.565 --> 00:36.285
that it explodes in the vicinity of the airplane

00:36.905 --> 00:39.645
and that gives you much more chance of actually causing,

00:39.865 --> 00:41.045
damage to that airplane.

00:41.665 --> 00:43.765
And so the crucial thing was he realized that

00:43.765 --> 00:45.605
what you needed to do was to have a system

00:46.155 --> 00:49.165
whereby you took into account where the shell exploded

00:49.745 --> 00:54.165
and you fed that back to make your next shot, both in terms

00:54.185 --> 00:55.205
of distance

00:55.225 --> 00:57.925
and in terms of, of accuracy,

00:58.195 --> 00:59.445
more likely to hit the airplane.

00:59.465 --> 01:01.845
So that that became the idea of a feedback loop.

01:02.265 --> 01:06.325
The idea is the information collected about the explosion

01:06.325 --> 01:08.085
of the, of the shell is used

01:08.085 --> 01:10.285
to improve the next time you fire it.

01:10.905 --> 01:12.205
So he used that term feedback

01:12.475 --> 01:14.685
because he was talking about the direction

01:14.805 --> 01:16.965
of information flow back to the system.

01:18.105 --> 01:21.405
The problem was that when this was, picked up

01:21.405 --> 01:22.765
by psychology in the sixties

01:22.905 --> 01:27.605
and seventies, psychologists def

01:27.625 --> 01:30.645
define feedback as any information fed back to

01:31.185 --> 01:34.285
an individual about their task performance.

01:34.345 --> 01:39.085
So your typing speed is 45 words per minute would be counted

01:39.185 --> 01:40.925
as feedback by psychologists.

01:41.465 --> 01:43.245
That's not something engineers will recognize

01:43.765 --> 01:45.685
'cause to an engineer unless the information is actually

01:45.755 --> 01:48.925
used like in the room thermostat, not just to tell you

01:48.925 --> 01:51.125
that the room is too cold, but then also to heat it.

01:51.835 --> 01:53.365
They wouldn't regard it as feedback.

01:53.985 --> 01:56.685
And so that's where I think the, the problem started is

01:56.685 --> 01:59.205
that in psychology, the term feedback was used

01:59.205 --> 02:00.405
to describe any process

02:00.755 --> 02:02.525
that we tell students about where they are.

02:03.065 --> 02:05.565
And just telling a student, you know, you need to do better

02:05.665 --> 02:09.325
or C minus or 43, I don't think is very helpful,

02:09.865 --> 02:12.045
but it still does count as feedback according

02:12.045 --> 02:14.445
to many definitions that are used in psychology today.

02:14.785 --> 02:16.565
The answer is it depends

02:16.665 --> 02:18.845
and it's one of those annoying psychology answers.

02:19.465 --> 02:22.685
The most important result I think in the last 50 years

02:23.305 --> 02:26.285
was the work of two psychologists at Rutgers University

02:27.165 --> 02:28.345
at Kluger and Danisi

02:28.885 --> 02:31.545
who reviewed about 3000 research

02:31.545 --> 02:32.825
reports on the effects of feedback.

02:33.745 --> 02:35.115
They went back to 1905

02:36.135 --> 02:38.435
and they found that on average feedback did improve

02:38.435 --> 02:41.315
performance, but in 38% of the cases

02:41.425 --> 02:44.235
that they actually found were trustworthy, the reliable

02:44.915 --> 02:49.075
research reports in 38% feedback actually made things worse.

02:49.495 --> 02:51.475
In almost two out of every five cases,

02:51.975 --> 02:53.235
you're doing better off shutting

02:53.255 --> 02:54.475
up than giving the feedback.

02:55.615 --> 02:57.435
And so this is why it's very dangerous

02:57.435 --> 02:58.715
to say that feedback has an

02:58.715 --> 02:59.765
Average effect size

03:00.195 --> 03:01.325
because it could be

03:01.395 --> 03:03.005
that you actually put into place

03:03.005 --> 03:04.165
the feedback that makes things worse.

03:04.585 --> 03:06.325
So we don't necessarily want more feedback.

03:07.065 --> 03:08.725
The other point that Kluger and Denise made,

03:08.725 --> 03:11.365
which people still ignore even though their article is

03:11.365 --> 03:14.525
coming up for its 30th anniversary, is they said,

03:14.835 --> 03:17.165
even if the feedback does make things better,

03:18.075 --> 03:19.645
it's not necessarily a good thing

03:20.625 --> 03:22.245
if the feedback makes the students

03:22.315 --> 03:23.685
more dependent on the feedback.

03:24.385 --> 03:25.685
So you can actually give feedback

03:27.315 --> 03:29.165
that makes students performance better,

03:29.545 --> 03:33.545
but it needs to be maintained to maintain that improvement.

03:33.965 --> 03:36.305
And so what they showed, what they argued was we need

03:36.305 --> 03:37.545
to be looking at what kinds

03:37.545 --> 03:39.585
of things does it prompt in the recipient.

03:40.565 --> 03:43.065
And my biggest concern about feedback is

03:43.065 --> 03:46.465
that too often feedback is about improving the work rather

03:46.465 --> 03:47.665
than improving the learner.

03:48.405 --> 03:49.745
How do you get a better grade

03:49.845 --> 03:51.345
or score on this piece of work?

03:51.885 --> 03:53.305
If the feedback is too detailed?

03:53.795 --> 03:55.745
It'll tell the student exactly what they need to do

03:55.745 --> 03:57.545
to get better, grades or scores,

03:58.085 --> 04:00.465
but it doesn't necessarily help them learn

04:00.985 --> 04:02.625
'cause the feedback is telling them what

04:02.625 --> 04:03.825
to do to get a better grade.

04:04.485 --> 04:07.985
The difficulty here is that people read books on feedback

04:07.985 --> 04:09.185
for teachers and they tell

04:09.185 --> 04:11.545
that feedback should be immediate, that it should be

04:12.185 --> 04:13.545
positive, that it should be

04:13.545 --> 04:15.105
descriptive rather than evaluative.

04:15.885 --> 04:18.505
And all those things are, you know, often good ideas,

04:19.165 --> 04:21.665
but the research doesn't support such clear conclusions.

04:21.925 --> 04:23.425
So there's quite a lot of research

04:23.425 --> 04:24.905
that shows particularly in higher education

04:25.215 --> 04:26.945
that delayed feedback can be more

04:26.945 --> 04:28.025
effective than immediate feedback.

04:28.765 --> 04:32.465
The reasons aren't that clear yet, but my hunch is it's

04:32.625 --> 04:35.995
because delayed feedback acts as a kind

04:36.455 --> 04:38.035
of delayed res study.

04:38.355 --> 04:39.915
Students are coming back to this when they

04:40.235 --> 04:41.475
actually had forgotten about it.

04:41.775 --> 04:43.315
And the research of Robert Bjork

04:43.315 --> 04:44.515
and others on memory shows

04:44.515 --> 04:46.475
that when you actually uddy something

04:46.745 --> 04:50.755
that you did a while ago, that UD actually has more

04:50.775 --> 04:53.595
of impact when you actually find it hard

04:53.595 --> 04:54.915
to remember what you did the previous time.

04:55.775 --> 04:58.195
So the fact is, delayed feedback can be more effective.

04:58.775 --> 05:01.475
The fact is that evaluative feedback can be very effective.

05:01.505 --> 05:03.195
Telling a student at the rate you are going,

05:03.375 --> 05:04.715
you're not gonna get the grade

05:04.715 --> 05:06.435
that you want might be helpful.

05:07.415 --> 05:10.715
The problem is you can't really separate out all those

05:10.995 --> 05:13.115
variables from the social situation in which

05:13.115 --> 05:14.195
the feedback takes place.

05:14.995 --> 05:18.675
I say to teachers, if your feedback is getting you more of

05:18.675 --> 05:21.075
what you want from your students, it's good feedback.

05:21.615 --> 05:22.835
If it's getting you less of what you

05:22.835 --> 05:24.615
want, it's bad feedback.

05:24.875 --> 05:26.735
And sometimes, you know, if you have good relationship

05:26.735 --> 05:30.135
with a student, you can say you can do better than this.

05:30.405 --> 05:31.615
This is not your best work.

05:32.075 --> 05:33.535
And people will throw up their hands and heart

05:33.535 --> 05:34.855
and say, you can't give feedback like that.

05:35.515 --> 05:38.095
But in a particular relationship, in a particular context,

05:38.645 --> 05:40.215
that might be the most effective thing.

05:40.575 --> 05:42.615
'cause you know, that student didn't try,

05:42.845 --> 05:43.975
they just phoned it in.

05:44.205 --> 05:45.815
They were trying it on waiting for you

05:45.815 --> 05:48.135
to give them some feedback about how to get a better grade

05:48.675 --> 05:51.215
and you just call them on it Too often we give students

05:51.495 --> 05:53.375
feedback, but we don't make sure that it's used.

05:54.485 --> 05:57.085
Feedback is basically one-to-one tuition, I don't know, know

05:57.085 --> 05:58.525
of anybody who can give feedback on two

05:58.525 --> 05:59.605
pieces of work at the same time.

06:00.505 --> 06:02.405
And yet we don't take advantage of that.

06:02.505 --> 06:04.605
So in my view, we have to close the loop.

06:05.015 --> 06:07.965
Going back to Norwood Wiener's point from the 1940s,

06:08.455 --> 06:09.925
let's close the loop of feedback.

06:10.175 --> 06:13.885
Let's give our students time in class in the sessions

06:14.105 --> 06:15.645
to respond to our feedback.

06:16.505 --> 06:17.765
If it's worthwhile taking time

06:17.765 --> 06:19.645
to write individual comments on students' work,

06:20.275 --> 06:23.165
it's worthwhile taking class time for them to respond.

06:23.505 --> 06:25.205
The crucial thing about feedback is

06:25.915 --> 06:27.045
what students do with it.

06:27.625 --> 06:30.925
And that's moderated by a whole range of factors,

06:31.035 --> 06:32.525
including the relationship.

06:32.585 --> 06:34.685
Do they trust the person who give them the feedback?

06:34.805 --> 06:36.005
Do they think they know what they're talking about?

06:37.245 --> 06:41.915
But the trouble is that feedback is often given in a context

06:41.965 --> 06:43.275
where you really wanna do well.

06:43.295 --> 06:45.795
You want to grade, you want a particular kind of outcome.

06:46.415 --> 06:47.475
And so it's inevitable

06:47.905 --> 06:50.795
that there will be an emotional aspect to any feedback,

06:51.345 --> 06:53.195
even if it's qualitative, even if just comments.

06:53.735 --> 06:55.075
If you think the piece of work is great

06:55.075 --> 06:56.515
and somebody tells you it's not very good,

06:56.945 --> 07:00.055
your first reaction is either, you know,

07:00.395 --> 07:01.535
I'm not as good as I thought I was.

07:01.635 --> 07:02.495
Or the person giving

07:02.495 --> 07:03.815
me feedback doesn't know what they're talking about.

07:05.065 --> 07:06.485
So it's actually very difficult

07:06.985 --> 07:10.165
to use feedback intelligently if that emotion is triggered.

07:10.625 --> 07:13.485
So in my view, although emotions are an inevitable part

07:13.485 --> 07:15.605
of this, the teacher's job is to try

07:15.605 --> 07:18.645
to minimize the emotional reaction to feedback

07:19.505 --> 07:23.365
and to maximize the engagement, the cognitive engagement

07:23.515 --> 07:24.725
with what the feedback is saying

07:24.725 --> 07:25.925
about what they need to do to improve.

07:26.155 --> 07:28.525
There's a big debate about whether grades are harmful

07:29.625 --> 07:31.285
and there are some studies that show

07:31.365 --> 07:32.645
that when you give students grades

07:32.705 --> 07:33.885
and comments at the same time,

07:34.145 --> 07:36.805
the first thing they look at do is look at the grades.

07:37.465 --> 07:38.725
And the second thing they do is look at

07:38.845 --> 07:39.885
somebody else's grade, what'd you get?

07:41.545 --> 07:43.045
So it's often quite hard

07:43.045 --> 07:44.445
to get the students to look at the comments.

07:45.405 --> 07:47.335
Some people say, well, why not give the comments today

07:47.475 --> 07:51.085
and give the grade tomorrow at least for 24 hours.

07:51.085 --> 07:53.245
They're gonna be reading the comments, presumably

07:53.305 --> 07:55.005
to find out what the, the grade is gonna be.

07:55.005 --> 07:56.045
But at least they're reading the comments.

07:56.745 --> 08:00.725
But my view is, well, why not give them a week

08:00.825 --> 08:04.515
to resubmit the piece of work in the light of the comments

08:05.015 --> 08:06.155
and then give them a grade?

08:06.735 --> 08:09.715
So I don't think the evidence supports strong

08:09.875 --> 08:11.355
conclusions like grades are harmful.

08:12.035 --> 08:14.275
I think what the research does support is this idea

08:14.275 --> 08:16.995
that it's very hard to tell students where they are

08:17.535 --> 08:19.715
and how to get better at the same time.

08:20.455 --> 08:21.715
So do one or the other.

08:23.005 --> 08:25.025
If the students are allowed to resubmit for extra credit

08:25.085 --> 08:26.145
or a higher grade,

08:26.295 --> 08:28.065
then give them advice about how to do that.

08:28.125 --> 08:29.545
But don't make it too specific,

08:29.655 --> 08:31.825
otherwise you're improving the work and not the learner.

08:32.575 --> 08:34.955
But at the end of the course, just give them a grade

08:35.415 --> 08:36.715
and don't bother justifying it.

08:37.305 --> 08:38.435
It's wa it's a waste of time.

08:38.815 --> 08:40.835
And of course in higher education, that's what we see.

08:41.175 --> 08:42.395
We see the most of the comments

08:42.395 --> 08:44.555
that are being given on grade, on student work.

08:44.785 --> 08:46.595
It's not really designed to help the student,

08:47.025 --> 08:49.835
it's the teacher justifying the grade they awarded.

08:50.375 --> 08:52.315
And that's quite helpful for audit purposes.

08:52.745 --> 08:54.995
It's not particularly helpful for learning purposes.

08:55.265 --> 08:57.485
The only good feedback is that which is used by learners.

08:58.345 --> 09:02.245
And so what people like Naomi Winston have done is

09:02.245 --> 09:04.365
to look at how do we get students ready

09:04.365 --> 09:05.565
to receive the feedback?

09:06.305 --> 09:08.245
And I think although there's a big debate about

09:08.545 --> 09:11.525
how big these effects are, I think there's evidence

09:11.525 --> 09:14.525
that having what Carol Dweck calls a growth

09:14.525 --> 09:16.245
mindset can be helpful.

09:16.905 --> 09:19.605
Now, I don't wanna get into the debate about whether

09:19.605 --> 09:20.725
this is a real thing or not.

09:21.265 --> 09:22.445
So I'm gonna sidestep this.

09:22.705 --> 09:24.965
I'm gonna say as long as students believe

09:24.965 --> 09:26.765
that smart is not something you just are,

09:26.985 --> 09:28.445
but smart is something you can get,

09:29.295 --> 09:33.185
that effort produces ability, then feedback is welcome

09:33.915 --> 09:35.325
because you know, you can get smarter

09:35.745 --> 09:37.685
and the feedback tells you how to get smarter quicker.

09:38.905 --> 09:40.965
If you believe that you only have a certain amount

09:40.965 --> 09:42.005
of ability that is fixed,

09:42.765 --> 09:44.695
then feedback is potentially unwelcome.

09:44.855 --> 09:46.135
'cause it might show you you're not as smart

09:46.135 --> 09:49.335
as you thought you were, and if the feedback is not the

09:49.495 --> 09:51.775
feedback you want, you might reject the feedback by saying

09:51.775 --> 09:53.175
that the person doesn't know what they're talking about.

09:53.835 --> 09:55.655
Or you might conclude that you're not really good at this

09:55.655 --> 09:57.255
subject and you lower your aspirations.

09:58.355 --> 10:00.775
So I think this idea that effort creates ability

10:01.995 --> 10:04.415
is really at the heart of helping students make sense

10:04.415 --> 10:07.575
of feedback because then they know they can get smarter

10:08.275 --> 10:09.455
and attending to the feedback

10:09.575 --> 10:10.775
will help you get smarter quicker.

10:11.855 --> 10:14.195
And ultimately it depends on relationships

10:14.505 --> 10:17.195
because every teacher knows that the same feedback given

10:17.215 --> 10:19.235
to one student will make that student try harder.

10:19.855 --> 10:21.435
And to a diff very similar student

10:22.375 --> 10:23.665
will make that other student give up.

10:24.085 --> 10:26.745
So you have to know when to push and now when to back off

10:27.685 --> 10:29.825
and your students have to believe in you that you know

10:29.825 --> 10:31.705
what you're talking about and you have their

10:31.705 --> 10:32.745
best interests at heart.

10:33.605 --> 10:35.985
So the messy conclusion from all this research

10:36.995 --> 10:38.535
is it's about relationships.

10:38.835 --> 10:39.975
And we can't package that.

10:40.035 --> 10:43.175
We can't tell teachers exactly what to do,

10:43.435 --> 10:45.855
but it's a very nice piece of work by David Jager

10:45.855 --> 10:49.375
and his colleagues where they just give half of a group

10:49.375 --> 10:53.295
of students, a, a post-it note to go

10:53.295 --> 10:55.775
with their feedback.

10:56.075 --> 10:57.375
So all the students got feedback,

10:57.485 --> 10:59.015
half the students got a post-IT note

10:59.015 --> 11:00.575
saying, I've given you feedback.

11:00.575 --> 11:01.735
So you have, I've given you comments.

11:01.735 --> 11:04.975
So you have feedback on your paper. Half the students got a

11:04.975 --> 11:08.215
post-it note that said, I've given you critical feedback on

11:08.215 --> 11:10.495
this work because I have very high standards

11:10.835 --> 11:12.095
and I know you can reach them.

11:13.545 --> 11:16.315
So the only difference was the post-IT note

11:16.465 --> 11:17.635
that came with the feedback.

11:18.695 --> 11:20.665
They then followed these students through,

11:21.395 --> 11:23.105
these were eighth grade students.

11:23.405 --> 11:26.935
So year nine in the UK context, they were more likely

11:27.595 --> 11:32.115
to enroll in a college at the end, at the age of 18,

11:32.505 --> 11:35.675
just a post-it note improve their educational achievement,

11:35.805 --> 11:37.715
their attitudes and trust towards school.

11:38.415 --> 11:40.515
It reduced the disciplinary problems in the following years

11:40.935 --> 11:42.955
and increased enrollment in college.

11:43.695 --> 11:47.355
So to me, this is evidence, this is quite low hanging fruit.

11:48.695 --> 11:50.075
We just need our students to understand,

11:50.415 --> 11:52.795
I'm giving you feedback because I care about you.

11:53.505 --> 11:55.285
I'm not doing it to be mean. I'm doing it

11:55.285 --> 11:59.435
because I want you to improve to reach the highest levels

11:59.435 --> 12:00.595
of achievement that you're capable of.

12:01.385 --> 12:04.845
So the difficulty with peer assessment is that most

12:04.845 --> 12:08.485
of the research has focused on whether peers can give each

12:08.485 --> 12:11.405
other grades that are as accurate as teachers grades.

12:12.185 --> 12:14.645
And that's quite a useful line of research.

12:14.665 --> 12:18.245
And the evidence is that they're not too bad For me,

12:18.305 --> 12:21.845
the more important aspect is whether peers can give each

12:21.845 --> 12:23.245
other feedback that is useful

12:23.465 --> 12:25.565
to the recipient in improving their learning.

12:26.105 --> 12:27.165
And the evidence is that they can.

12:28.305 --> 12:30.525
But there's a real problem here, which is

12:30.525 --> 12:32.085
that peer assessment is often inaccurate.

12:33.115 --> 12:36.455
Now there's a phenomenon called the Dunning Kruger effect.

12:37.085 --> 12:39.655
There's some debate about how real in effect this is,

12:39.915 --> 12:43.255
but it does appear that the lower levels of achievement

12:43.285 --> 12:45.855
that you have, the more likely you are to be confident

12:45.855 --> 12:47.455
that you're correct when you're actually incorrect.

12:48.485 --> 12:49.735
Well, whether that's true

12:49.755 --> 12:53.255
or not, we definitely see students who are confident

12:53.255 --> 12:56.345
that they know something, giving confident advice

12:56.345 --> 12:58.465
to their peers, which is actually inaccurate.

12:59.565 --> 13:03.385
So I say to teachers, if you choose to use peer assessment

13:03.845 --> 13:07.825
as a way of helping students improve,

13:08.335 --> 13:10.665
then you are still responsible for the quality

13:10.765 --> 13:11.985
of feedback they give each other.

13:12.165 --> 13:14.025
You need to be monitoring those conversations in your

13:14.025 --> 13:15.745
classroom or online

13:16.015 --> 13:17.785
because there's a real danger

13:17.785 --> 13:19.505
that they will give each other misleading information.

13:20.495 --> 13:22.755
As for peer as, as sorry, as for self-assessment,

13:23.315 --> 13:25.435
I think this is ultimately the main goal.

13:25.995 --> 13:29.715
I think all feedback should work towards its own redundancy.

13:30.215 --> 13:32.395
The goal of feedback is to empower learners

13:32.695 --> 13:34.795
to take more responsibility for their own learning.

13:35.655 --> 13:38.115
So I think for me, yes, peer assessment,

13:38.115 --> 13:40.595
self-assessment is often not very accurate to begin with.

13:41.095 --> 13:45.155
But our goal should be to empower our students

13:45.575 --> 13:47.875
to become more effective in managing their own learning,

13:48.485 --> 13:50.955
being critical, knowing when they understand something,

13:51.205 --> 13:52.915
being tough on themselves when they don't.

13:53.795 --> 13:56.775
That's I think the, the, the really powerful idea here.

13:57.155 --> 13:59.935
The idea that feedback should help students become in the

13:59.935 --> 14:02.855
psychological jargon, self-regulating learners.

Faculti AI can make mistakes

Research

Policy

Analysis