Fixed Effect vs. Random Effects Models - Common Mistakes in Meta-Analysis and How To Avoid Them 

This presentation is pure gold! Thank you for sharing it with the rest of us unfortunate to have not attended this session

This is the best explanation of RE models and FE model in meta-analysis that I have ever heard.

Finally understood it! Thanks a lot for sharing this, and for the reference to your paper for the mathematical differences.

这个视频不断的让我想到了某国的大部分教授们，总是在不断地强调结果是不是显著，怎么巧妙的处理就可以让结果看着更好看……这种模式下学了很多甚至变得“很厉害”成为了“大佬”一头雾水的开始分享怎么做一篇“优秀”文章的经验，sucks。Anyway，这真的是一篇perfect lecture，highest respect for this vedio and the professor！

Thank you very much! I found the paper you referred to before watching this video. After watching it I understand most of it.

Beautiful explanation. I'm going to sort the paper I am working on right away.

Now I can see what exactly are fixed-effect and random-effects models. Thank you!!

Thank you for taking the time to make this helpful video!

Brilliantly clear presentation. I will be sure to use the word 'universe' in my study!

What an absolutely outstanding presentation! Wonderful! Thank you for sharing this.

Enjoyable and clarifying!! Thank you!!

Your insight is profound. Thank you.

Fantastic tute as usual. Thanks alot Michael

I applaud this explanation! Thank you dr. Borenstein. Random effects models rule. The fixed effect model is simple, clear, and wrong.

Thank you, Dr. Borenstein. Excellent video!!

Thank you so much! Very helpful!

Thank you so much. Very clear and concise. Look forward to reading the book / using the software. Best Regards, Babak Khoshnood

We owe a lot to Dr.Borenstein. Thanks doctor, i wish all the best to you.

good speaker. so clear.

Thank you for your lecture. very easy to understand

Great Lecture Dr. Borensterin!!!

As an MD and clinical researcher for 5 years, i observed that some MDs/physicians really have no/little idea on how to run statistical tests for research (which is understandable since statistics is really not part of the MD curriculum) so they are wholly/partially dependent on their statisticians. There are times also when the statisticians have the right idea / executed the tests correctly but the clinician interpreted the methods and/or results differently. 😅😅😅

Let me add three more to the list: eyeballing funnel plots to determine whether they're asymmetric, treating Egger's test as though it detected publication bias, using trim/fill or PET PEESE on highly heterogeneous studies

great video. thank you very much

Thank you so much!!!!!!!!!!

Thankss a lot for sharing this amazing lecture, I learned a lot from this lecture.

Thank you for your lecture

Thank you, your video is very informative

You should mension the method of Peto and the Fixed EffectS model sooner in the video! As a disclaimer. Now I've watched the entire video (and learned a lot!) and only found out at the end that the study I am interested in actually uses the fixed effectS model and Peto OR's.

he mentioned he'd been talking about giving an appropriate weight to each study in another module, does anyone know which module it is?

thanks sir

Hi Sir, if Hausman test indicates that fixed model is more appropriate than random effect model, and if in that case, in data time period (T) > cross section units (N), which FEM is to be chosen: time (T) FEM or Cross section (N) FEM?

Should fixed effect model instead of random effects model be used at 23:21? Is that right?

Sir, can you please share the relevant reference for the above

Good

nice

So it’s wrong to chose a fixed effect model based on heterogeneity less than 50

❤

It seems that the narrator is reading a written script, because I can hear he's turning the pages. This makes the flow a little fast. I really need 100% focus.

Fake and Fabricated! Bad presentation, and wrong and way off any reasonable analysis. Sad, I was excited