Absolutely great idea! I also like the design choice of making it its own function that aligns with natural language instead of just adding a parameter to the filter function, which would more likely be missed by users.

Thanks a lot for making me aware! What a timing :)

I guess it would be nice to have a function "filter_out" that only DROPS values that evaluate to TRUE in its argument, keeping NAs in the data.frame. The alternative, as always, is that you handle NAs explicitly in your function but it's just easy to forget them :)

I cannot state this enough: The filter function cannot know if the logical values it received stem from a comparison of equvialence (e.g., x == 1) or of inequivalence (e.g., x != 2). It just receives a vector of logical values (including NAs). Therefore, it only makes sense to drop NA values.

Surprisingly, both 2 and NA get dropped, leaving only a tibble with the first entry of x == 1. The reason behind this is that NA != 2 gets evaluated as NA and not as TRUE and NAs are dropped in filter (which makes sense).

It was very insightful to see where the statistical tools that we use every day come from and why they are the way they are!

🧠♻️ IGOR Symposium
"Open Science Initiatives in Biopsychological Research"
🗓️ Fri, 20.06 | 🕥 10:30–12:00
📍 0.004

@ocklenburg.bsky.social C. Artemenko
@mariame.bsky.social @alinakoppold.bsky.social
@caggursoy.bsky.social @ufangyang.bsky.social
@mreutter.bsky.social @mklingelhoeferj.bsky.social [3/6]

Navigation on the slides is done by using the arrow keys. You can also click the three horizontal lines on the bottom left to open the outline and jump to any section!

For your best experience, the slides are created using RMarkdown/quarto. Thus, the R code on each slide is evaluated and the result is directly shown. You can also easily copy code components from the slides into your R scripts using the copy button on the top right of each code chunk.

Of course, all data, analysis scripts, and further materials are openly available🔓 #openscience
osf.io/4gz7f/

Here is the histogram of social anxiety scores. Almost half of the scores are above the optimal remission cutoff by von Glischinski et al. (2018).