How badly can a meta-analysis go wrong? (Very)
How long can the authors ignore serious problems? (5+ years)
What ways can you check an existing meta-analysis for errors? (Let me show you!)
New preprint: https://t.co/VkhUn3Fklr
We're thrilled to present you our latest special issue on the scientific reform movement!
In response to the replication crisis and fraud cases in the behavioral and social sciences, a reform movement has arisen that pleads for a different culture and new practices... 1/2
At the last conference I was at, I stood up and told a presenter that I had rejected their paper for XYZ reason. He said he hadn’t received the editor’s decision yet, so I told him to get ready. It silenced a whole room of academics. Look. There’s nothing wrong with honesty.
@andrewheiss @rbjanis5 Hey this sounds like magic! Do you happen to have any resources that can take me from start to finish in making something similar?
@askryanair I was denied boarding bc of overbooking in July & missed a wedding. After 4 months, customer support told me they have no evidence that happened. I then filled out the EU form, posted to your headquarters. Still no response.
When can I expect my compensation?
@Ryanair I was denied boarding bc of overbooking in July & missed a wedding. After months of waiting, customer support told me they have no evidence that happened. I then filled out the EU form, posted to your headquarters. Still no response.
When can I expect my compensation?
Yesterday we had the honor to receive the Open and Responsible Science award of @erasmusuni and it was such a joy 🥳☺️
Thanks to our whole team who worked tirelessly (and for almost 3 years!) on #OpenScience♥️
Check them out👇 (cute 🐱 pic coming!)
@nbballou@rstatstweet To put words in the mouth of the infamous proponent of equivalence testing @lakens - a p-value can help you control your error rate for a one-tailed or two-tailed hypothesis. The one-tailed hypothesis test in particular is quite intuitive to implement with p-values
@rappa753@levikul09 This looks neat! I'm now thinking it looks like a coefficient (some say tie-fighter) plot minus the confidence intervals and with an extra lollipop line. Wouldn't an audience that cares about correlation coeffs care more about uncertainty estimates?
I like how it looks though!