#statschat #statsdon #stats #rant
While not totally accurate, do you think if we start saying "if you're wrong, the p-value is random" would improve understanding of why they provide little info?

#statschat #statsdon #statistics #statistics

Do you know what Type M error and Type S error are?

(see Gelman & Carlin, 2014)

#statschat #statsdon #stats

What is the best way to "correct" for optional stopping (data peeking)?

I understand if I plan for a sequential design how to control the alpha, but what if someone comes to me saying "Oh, I looked at the data last week and the p-value was still big, I'm putting in for more funding to get more people"?

How do you correct the conditional probability?

I caved and updated my #rstudio version
aaaaand I've lost all my addins :( what are your favourite #r #rstudio addins? #coding #statsdon
I had a word count one...and more but I forgot what 😭

p-values higher than your alpha (say, 5%) are not to be taken as "evidence of no effect". Everyone with me so far?

OK, now explain why this is exactly how almost all mediation analyses, multivariate regression models, SEM, fMRI papers, etc. use them! 🤷‍♂️

#statsdon #statschat

Fantastic paper describes the bias that may arise *from* #regularisation in a multi level model.

https://www.cambridge.org/core/journals/political-analysis/article/understanding-choosing-and-unifying-multilevel-and-fixed-effect-approaches/8101D49CFD3B129F5753FC878F416980

Simple correction the authors propose only works if there is *no* treatment-confounder feedback(see: https://go-bayes.github.io/b-causal/posts/confounder-treatment-feedback/conf-out-feedback.html)

#causalinference #statsdon

Is there a way to copy references from word into R markdown? I am updating some work I did a while back, but I wrote it in word, using Zotero for citations. Now I am using R to write it, and I'd rather not have to manually update each reference. #r #rstats #statsdon #coding

#statschat #rstats #statsdon
Why don't ppl ever apply correlation corrections when they report results for single studies? e.g. attenuation and range restriction.

Is there some fundemental issue with these approaches, or is it just that most don't know they exist?

(I have a similar pet peeve with ppl never using tests for difference in correlations, like Steiger's Z, and just claiming difference)

I need to improve my coding. Everything I write is a Frankenstein of things that worked last year, & other people's genius. That's OK but it's V long & messy.

Are there any blogs/classes/groups that review code, or where I can learn to write nicer code? #rstats #statsdon #WomenWhoCode

Follow up ggplot question:
I'm trying to understand why I need to specify color in the ggplot(aes) & again manually add it:

df%>%
ggplot(aes(x, y, colour=factor(z), shape=factor(z)) +
geom_point()+
scale_color_manual(values = my_col) +
scale_shape_manual(
values = c(15, 17, 18))

If I don't put " ggplot(aes( colour=factor(z)" the manual colour doesn't work. If I don't add manual color, I can't specify the colours.

#rstats #statsdon #dataviz #ggplot

📣 ggplot enthusiasts. Can someone help me to understand what the difference is between the aes() in the ggplot() function, and the aes() in the geom_line() function? 🙏
e.g:
ggplot(df, aes(x, y, colour = factor(colour))) +
geom_line(aes(group = 1), size = 2) 📊
#rstats #ggplot #statsdon #datavis

Is "semi-parametric discrete choice" a thing? I'm thinking of e.g. a logit with a random forest component. Maybe this is like a control function approach to discrete choice?

References appreciated 🙏

#EconTwitter #statsdon

#rstats question:

If I have an imputed dataset of 5 lists, is there a way to find the average for a variable over all lists?

E.g, what is the average age in the imputed dataset. At the moment I would just do mean(df[[1]]$age) for each 5, and manually average those. Seems long. #statsdon

Slowly realising that the #Bayesian model I've built is really specific to my data, which means I decide what to report & how to report it. I keep looking for inspiration on reporting the results, but I need to do the thinking on what makes sense for the intended audience.

When we are taught stats, its usually a 'prescription'. 'Compare two group means, do a t test, report the p value, df and test stat'. I'm scared of my model 😂 I want the prescription back!! #datascience #rstats #statsdon

As I learn more about statistics, I wish there was more material out there on how to better understand your teams, operations, and workloads. The hurdle to learn what little I know about stats was daunting for me. Is there a “stats for sysadmins/devops” book out there?

#statsdon #math #statistics #devops #sysamin

@kaz_yos I saw #epidon and #statsdon

#rheumatodon #epitodon #statsdon may sound better. The last one seems to exist

#rheumatodon #epitodon #statsdon may sound better. The last one seems to exist

A common statistican sin is to calculate the mean of an ordinal scale and use the mean for further analysis.

Does anyone know of a good and easily understandable explanation of why this is wrong, and what to do instead?

(I should probably add a CW to this, because it could trigger trauma in statisticians.)

#Stats #Statsodon #Statsdon #Statistics

A common statistican sin is to calculate the mean of an ordinal scale and use the mean for further analysis.

Does anyone know of a good and easily understandable explanation of why this is wrong, and what to do instead?

(I should probably add a CW to this, because it could trigger trauma in statisticians.)

#Stats #Statsodon #Statsdon #Statistics