Katharine Bailey

Professor Larry Hedges ‘education should be evidence-informed’ #TBED2019

#TBED2019 Prof Larry Hedges. "When social science indicators become high stakes we can expect them to be corrupted"

#TBED2019 prof Larry Hedges - purely statistical issues have had the lion's share of attention - oartly cos easy to do. But a new role of expert reviewer is emerging. Not all ed researchers understand what doing good reviewing involves. New institutions too.

Larry Hedges remind us that not all good researchers can do good systematic reviews. #TBED2019

The amazing Larry Hedges sharing his experience at #TBED2019

#TBED2019 @dylanwiliam lists four important questions for "critical consumers" to ask of education research: - Does it solve a problem we have? - How much improvement will we get? - How much will it cost? - Will it work here?

.@dylanwiliam refers to the issue of the ‘file drawer problem’: it’s easier to get experiments that have a statistically significant result published, but the statistical power of most social studies experiments is low; so only ‘lucky’ experiments tend to get published! #TBED2019

Things reported as effects may just be associations e.g. self-reported grades. Smart kids know they're smart. Dylan Wiliam @ #TBED2019

#TBED2019 @dylanwiliam great overview of impact of closeness of assessment to approach to T&L. Incentive to use sensitive ones for researchers - but this challenges application potential & in real world sensitive assessments v hard to design well

“Effect size can be influenced by the outcome measures used... assessments that test things that are very close to what has been taught in an intervention will tend to result in larger effect sizes than those that are more remote, like standardized test” @dylanwiliam #TBED2019

Another issue with meta-analysis... Interventions vary in their: Duration Intensity Collateral effects e.g. assignment of teachers [teacher quality] Dylan William #TBED2019

💥The Big Evidence Debate is happening TOMORROW💥 Why use meta-analysis in #education? If you couldn't make it in person, follow #TBED2019 or tune in to our live stream. @dylanwiliam @ProfCoe @ProfKime @PhilippaCcuree @joeybagstock @DamsonEd @profstig @chemDrK @Lem_Exeter

An exciting day ahead with lots of interesting talks...lets get started ☺ #TBED2019

Enter the ballot for a place at The Big Evidence Debate next year #TBED2019 ⁦@CEMatDurham⁩ thebigevidencedebate.com

*NEW POST* this month I’ve written a post for @CEMatDurham. The Level Illusion cem.org/blog/the-level…

This is brilliant from @jpembroke on the story of Levels That Just Will Not Die

'Mastery' - a bonus point for the bright kids! A brilliant blog from @jpembroke The Level Illusion – CEM Blog shar.es/a1VPtH

@CEMatDurham @jpembroke A good blog James and CEM! The impact is little or low because the vast majority of ‘tracking data’ is pure fiction. Pseudo science at its worst - the pretence of a mathematically precise line from an unwarranted assumption to a forgone conclusion!

@jpembroke @CEMatDurham As ever, great blog. Totally agree with what you are saying. First task in my new headship was to scrap existing complicated tracking system, put in a simple one and use @CEMatDurham Incas+ assessments to cross reference judgments.