And also, couldn't you account for response bias (via increased SEs) by including information about race (for example) in the random effects structure? Curious what #surveyscience people think about this...

I'm learning new things about survey methodology, and as I learn more about design-based analysis approaches, I can't help but make connections to random-effects models.

Couldn't you build a model that takes into account different strata (for example) by including a random intercept of strata?

I'm integrating the "page order" data into the raw data and it seems simple enough. but obviously a button option to include in the data would be a dream!!!

Oh cool! I can figure it out analytically...but yes, something like this would be extremely nice!

You have the only other #surveymonkey hashtag on this site :P Do you happen to know if there is a way to export an indicator of completeness with the raw data in survey monkey?

Is there a way to export an indicator of survey completeness in SurveyMonkey? There is in Qualtrics.... so surely there is a non-obvious way in SM as well? #surveymonkey #surveyresearch

What is it about R that inspires love, devotion, and creativity in us? This talk is a journey through the many things that make R gloriously weird.
Dr. Kelly Bodwin @kellybodwin.bsky.social
Associate Professor of Statistics and Data Science, Cal Poly

www.youtube.com/watch?v=KOQB...
#RStats

yeah unforch that's what I'm doing - thanks for the help though!! :)

Someone must have snuck me into a starter pack 😆

omg fun!!!

This is fantastic... unfortunately I think I need to make a data dictionary because my column headers are long survey questions... but I will keep this for future reference!

Ugh unfortunately not! Or I haven't found it yet... I loved that feature in qualtrics!!

This is exactly the approach I was thinking I'd have to do, so this is super helpful to see/follow! TY!! Not google forms, but survey monkey instead.

Thanks! I was hoping for something a little more complex/possibly doesn't exist. I'm working with survey data and the entire question is printed in the column name, which isn't ideal. I may have to come up with some new colnames manually :(

Is there an R package that creates shorter versions of your column header names? Dealing with some survey data and trying to avoid renaming things by hand... TY!!

Has anyone had any experience downloading data from Semantic Scholar using an API key? If so, I'd love to chat with you! Please RT!!

Figure showing various levels of statistical power when the number of participants, effect size and dropout rate varies.

Here is a brief example on how to do a power analysis for a longitudinal multi-level study when you want to vary the input parameters and easily produce visual output of all variations.
#rstats #openscience

pgmj.github.io/powerviz.html

The Tenure Hack I live by! "It is a lazy afternoon and you are in your boat fishing. You have several lines in the water, cast in various directions. Half asleep, you hear only the sounds of the water, gently tapping the sides of the boat. And then it happens. A pole bends. 1/