Thread by @DipeshGopal

Going through @dylanwiliam's plenary and pulling out the nuggets!

1⃣ There are no perfect solutions to anything. There are trade-offs. You need to decide which trade-offs you are willing to accept from new/old solutions. (Chesterton's fence)

#AMEE2021 #FOAMed

2⃣ Time spent on one thing means you miss out time on another thing.

Which things are worthy of inclusion and exclusion in the curriculum?

#AMEE2021

3⃣ Teaching is a wicked learning environment. Where students learn things is NOT where they apply them.

This makes working out what is effective teaching, a nightmare!

From Hogarth et al, 2015: journals.sagepub.com/doi/abs/10.117…

#AMEE2021

4⃣a

Teaching motivated students gets great feedback. Try playing @stokecity away on a cold windy night! 😂 bjorklab.psych.ucla.edu/wp-content/upl…

Great lesson performance does not always translate to long-term memory...must be a small amount of struggle. journals.uchicago.edu/doi/abs/10.108…

#AMEE2021

4⃣b

What is applicable in one study from #MedEd might not be applicable to another study - class size reduction in Tennessee Star study ... schools included were not representative edsource.org/wp-content/upl…

#AMEE2021

5⃣a

Be careful when making generalisations in meta-analyses (meta-analyses came from #MedEd)

Publication bias is a huge problem here and may exaggerate the true effect.

#AMEE2021

5⃣b

True effect size, taking multi-centre, replicated studies was 1/3 of reported effect sizes:
nature.com/articles/s4156…

Ioannides (2005): much research is false journals.plos.org/plosmedicine/a…

"If you torture the data long enough, it will confess." (Coase, 1981)

#AMEE2021

5⃣c

In #MedEd, problems with meta-analyses:

- few causal studies
- few studies on intervention strength, e.g. numbers in intervention group
- variability in teaching groups often unaccounted
- do we use sensitive measures with effect sizes tandfonline.com/doi/abs/10.108…

#AMEE2021

5⃣d

#MedEd meta-analyses may not reliable for several reasons:

#AMEE2021

6⃣

Will the solution solve a problem?

What is required for solution to occur?

What is the measurable benefit?

Will the solution work here?

#AMEE2021

7⃣a

Short-term memory limited to 4-7 chunks for 20 seconds

But long-term memory is limitless

So the main purpose of education is to extend long-term memory.

Do our curricula increase long-term memory?

#AMEE2021

7⃣b

Very difficult tasks can be completed but excessively difficult task do not transform to long-term memory.

Cognitive load theory: tandfonline.com/doi/abs/10.310…

#AMEE2021

7⃣c

There are 3 types of cognitive load:

➡️ Intrinsic - the structure/complexity of the information

➡️ Extrinsic - the way the info is presented

➡️ Germane - the way new info integrated with previous learning

Confused? Watch this:

#AMEE2021

7⃣d

Ways to improve things:

⬇️ extrinsic load - simple instructions, remove distracters

⬆️ intrinsic load - start simple getting more complex AND repetition

↔️ germane load - some complex tasks for the high achievers

#AMEE2021

7⃣e

Spaced learning and practice testing are great strategies to improve memory.

Your memory is characterised by storage & retrieval strength.

We can remember old previously learned languages when shown options ... good storage but poor retrieval strength.

#AMEE2021

8⃣

What is formative assessment for?

It's not to tell you what to teach.

#AMEE2021

9⃣

The timing of feedback is important but complex.

Feedback should be more work for the recipient rather than the donor.

Feedback should be less necessary (redundant): we want people who are self-regulating and can give reflect and improve on their own weaknesses

#AMEE2021

🔟

Assessments allows us to draw conclusions but not all conclusions are valid from those assessments.

#AMEE2021

1⃣1⃣

You need reliability for valid conclusions.

But increasing reliability will decrease validity,

e.g. getting great info about all the work or high quality of some of the work neglecting the rest: floodlight metaphor

#AMEE2021

1⃣1⃣

Students - getting better at football is not just by playing football...

...Learning stuff that will come up on the test won't make you good at passing tests!

Further resources:

@P_A_Kirschner and @MirjamN's blog:
3starlearningexperiences.wordpress.com

#AMEE2021