Lecture 13 - introduction to effect sizes for meta-analysis | Hard-Boiled Synthesis (Fall 2020)

Welcome to Hard-Boiled Synthesis (Fall 2020)!
This course aims to introduce two key research synthesis practices, systematic reviews and meta-analysis, by completing an entire research project from start to end.

PROJECT: synthesize studies testing catnip as a repellent to mosquitoes!

Lecture 13 contents:

1. Description of Hedge’s d effect sizes and how it relates to other common effect sizes (e.g., correlation coefficient, Odd’s ratio)
2. issues with log response ratio used in ecology
3. how effect sizes are typically rooted with a linear model foundation
3. origin of common effect sizes

[LajeunesseLab]

Hi all, thanks for watching!

The full lecture series of Hard-Boiled synthesis can be found here:

[kalalakanyanda]

Have watched all the series and am so grateful to you, am in the field of medicine but the same applies to all fields.

[bhavnagrover]

hi, can't thank u enough for such a great series of lectures. I was trying to learn meta analysis for a long time, but all I got was theory theory and theory. This is the first time I really feel like, " I Can Do It". Thanks a million.

[anamashraf]

Thanks a million Prof. for explaining "stat tings" in a comprehendible way. You have done a great job. Kindly deliver lectures on stats for ecologists.

[القرآنالكريم-شيخنورينمحمدصديق]

Keep going sir you are doing great work ♥️

[dianaobregoncorredor]

I am working on a meta-analysis and around half of my studies don't present any variance metric, especially the old ones, that is why I was thinking of using the log ratio. Also, considering that I am comparing outcomes from a very specific experiment, so I don't have data from regressions. Do you think in this case is appropriate to use log ratio? I don't want to lose all that data. Thanks