Thread by @PepsMccrea

How does time spent in direction instruction vs peer interaction vs practice/assessment impact learning? And does the answer differ by subject?

@profsimonb et al analysed the performance and data from 250+ teachers to answer these Qs.

Here's what they found:

↓

@profsimonb We know that teacher choices affect student learning and lives.

But we know less about exactly *how* the 'macro' moves of the classroom predict learning.

@profsimonb Burgess et al examined observational and GSSE attainment data for 250 maths & English teachers across 32 schools. They bucketed teaching into:

• Direct instruction
• Student peer interaction
• Personalised instruction
• Practice and assessment

And hunted for correlations 🧐

@profsimonb They found that:

→ When maths teachers allocated more time than average to 'practice & assessment', students tended to learn more
→ When English teachers allocated more time to whole class discussion and group work, students tended to learn more

Okay. But how much more?

@profsimonb The gains were similar for both subjects → 0.08SD, which is equivalent to about 2-3 extra weeks of lessons per year, or the difference between having a teacher in their 5th year rather than 1st year.

(which translates to about a 1.6% boost in lifetime student earnings)

@profsimonb Hold on, surely it depends on how well the teachers *implemented* these activities?

Good question...

The researchers explored controlling for this using an observation instrument (the 'Danielson Framework') and found that the effects still held.

Mostly...

@profsimonb There *were* some signs that students of more skilled teachers tend to learn more under a balanced diet.

Ie. Students of skilled maths teachers benefit from relatively more direct instruction; students of skilled English teachers benefit from relatively less peer interaction.

@profsimonb REFLECTIONS

First up, teaching is super complex. It's rendered in a fairly crude way here.

So, we need to be super careful not to over-apply these findings.

Eg. Let's avoid the lethal mutation where maths teachers only do practice and english teachers only do discussion 🧬

@profsimonb Furthermore, I don't think it shatters the assumption that the best teaching is responsive—to the thing being taught, the prior knowledge of the class, and the expertise of the teacher.

@profsimonb That said, it raises some interesting questions for us a as a profession, eg:

→ How intentional are we when it comes to allocating time the macro-acts of teaching?
→ How should our conceptualisation and support for effective teaching differ based on the subject being taught?

@profsimonb CAVEATS

1/ This study design allows us to draw mostly correlational—rather than causal—inferences.

There may well be other underlying factors which better explain what's going on here.

@profsimonb 2/ The findings rest on the validity of the tools used, and observation instruments have a number of limitations.

They tend to overemphasise the observable aspects of teaching (Danielson is better than most for this) and their validity is generally low (Gill et al, 2016).

@profsimonb BONUS

In contrast to that last caveat...

The authors argue that teacher effectiveness *can* be gauged by untrained peer observations, something that has been questioned by studies in the past (eg. Strong et al, 2011).

I'm honestly not sure what to think about this yet 🤪

@profsimonb *HUGE* thanks to @profsimonb et al for doing studies like this and making it all freely available.

I'd highly recommend folks check it out.

And if you have any thoughts, I'd love to hear them (this is fascinating stuff).

👊

@profsimonb @profbeckyallen your instincts were strong on this one