Mark Ramos @mframos.bsky.social · 18/08/2026Half of 274 statistically significant claims in social science #research were replicated in the SCORE project. How good/bad that sounds depends on how you phrase the headline. How good it actually is can be considered using some math, #statistics intuition, & common sense. doi.org/10.1093/jrss... 110
Mark Ramos @mframos.bsky.social · 10/11/2025The idea of "Wisdom of the Crowd" may seem strange at first glance, until you realize that the historical foundations of the Gaussian distribution is based on observations of human measurement error. Then, the idea becomes quite... Normal. #Statsky #statistics #edusky 030
Mark Ramos @mframos.bsky.social · 27/08/2025For my fellow Dungeons and Dragons fans who may taking a #statistics course this semester. This will come in handy when the time comes to learn hypothesis testing. #DnD #D&D 040
Mark Ramos @mframos.bsky.social · 16/08/2025Just finished Eyes of Wakanda and can say that it soundly makes up for both What if...? Season 3 and Season 2. 000
Mark Ramos @mframos.bsky.social · 09/08/2025In this paper, I discuss an often misunderstood aspect of inference that can trip up even seasoned researchers. If you wonder what "practical significance" means in #statistics, I guarantee you that reading this will make you a better scientist. #statsky #StatsSky #science doi.org/10.1093/jrss... 360
Mark Ramos @mframos.bsky.social · 06/08/2025Had a lot of fun doing this talk at #JSM2025. The transformed and relative Lorenz curves are available for use in the glorenz R package. #statsky #StatsSky #datascience #dataliteracy #jsm cran.r-project.org/web/packages... 033
Mark Ramos @mframos.bsky.social · 05/08/2025Today I will be presenting our recently published paper on New Graphical Displays and Related Statistical Measures of Health Disparities Among Groups. Hope to see you there! (CC-210, 2pm-3:50pm). #JSM #JSM2025 #statsky #StatsSky #statistics. academic.oup.com/jrsssa/advan... 010
Mark Ramos @mframos.bsky.social · 04/08/2025#jsm2025 kicks into high gear! So many exciting sessions to watch. What session are you looking forward to? Let’s gooooo! #jsm #statsky #statssky 000
Mark Ramos @mframos.bsky.social · 04/08/2025Awesome estimathon game by #janestreet at #jsm. #jsm2025 #statssky. We were team two sided XD. 010
Mark Ramos @mframos.bsky.social · 03/08/2025#jsm2025 is here in Music City! #stats #statsky #jsm #statssky Anyone try the hot chicken yet? 100
Mark Ramos @mframos.bsky.social · 02/08/2025Off to Nashville for the Joint Statistical Meetings. #jsm #jsm2025 #statsky. Doing a talk on our paper about some new disaggregated displays and measures for comparing continuous quantities between groups. Wish me luck! 031
Mark Ramos @mframos.bsky.social · 26/07/2025No method is perfect; no assumptions hold exactly. We use #statistics to make sense of uncertainty as objectively as we can. Most importantly, proper statistical practice keeps us honest, especially when results don't conform to what we hoped for. #datascience #statsky #dataliteracy #science #edusky 020
Mark Ramos @mframos.bsky.social · 14/06/2025Statistical power is the chance of detecting an effect given that a specific one exists. Being transparent with how much power we can get for a study before spending resources on it gives confidence to funders that their money is not being spent on a coin toss. #academicsky #research #science #STEM 131
Mark Ramos @mframos.bsky.social · 26/04/2025In #ABAtherapy, children with #autism who accomplish a task at 80% mastery criterion do not necessarily have 80% mastery level. #BCBA and researchers can now use the miebl R-package to better contextualize criterion choices in discrete trial training. #ABA cran.r-project.org/web/packages... 110
Mark Ramos @mframos.bsky.social · 20/03/2025Every teacher needs to ask themselves: "Knowing that my students have easy access to gen #ai after class, how can I develop my pedagogy to maximize the value of this resource for them?" It won't be easy, but it can be glorious. #edusky #academicsky #EdTech #Educhat #K12 #EduSkyAI #Edchat 031
Mark Ramos @mframos.bsky.social · 14/03/2025I will force a brain freeze deck every time. It’s just too much fun. This won off the back of breach with LED and brain freeze in the yard. 010
Mark Ramos @mframos.bsky.social · 12/03/2025You can interrogate data ceaselessly until it confesses but remember that at that point, it will just tell you whatever it was that you wanted to hear. Learning #statistics is knowing better than to fool oneself. #datascience #edusky #statsky #dataliteracy #iteachmath art by @bonkerific.bsky.social 042
Mark Ramos @mframos.bsky.social · 22/02/2025This was in the 60's. Now kids have TI calculators & apps like Desmos as part of math class. A similar thing is happening now with #ai, but this time on writing instead of computing. As before, #education must adapt to prepare students for their future, which includes ai tools. #edusky #iteachmath 020
Mark Ramos @mframos.bsky.social · 18/02/2025On this day, the brilliant statistician responsible for many tools we still use now (ANOVA,MLE,Exact Tests, etc.) but most importantly for popularizing significance & p-values, was born. Sir Ronald Aylmer Fisher FRS (17 Feb 1890 – 29 Jul 1962). #statsky #datascience #statistics #pvalue #significance 020
Mark Ramos @mframos.bsky.social · 16/02/2025Imagine having one of them shoot a target for you. You're not allowed to see whether they hit the target or not but you are required to decide on whether or not you think the target was hit. What decision will you make? Why? What statistical concept did I just explain? #edusky #statsky #iteachmath 220
Mark Ramos @mframos.bsky.social · 07/02/2025My kid was telling me earlier how glad she was that she didn't have to deal with geometry anymore (she took it last year) and that advanced algebra is way more fun. We both like Marvel movies (me more than her), so I just sent this to her on discord: 010
Mark Ramos @mframos.bsky.social · 01/02/2025I wanna say that sharp rise between 2000 and 2020 is because of this guy. 010
Mark Ramos @mframos.bsky.social · 31/01/2025Handwashing wasn't #commonsense until the 1840's, when Ignaz Semmelweis showed it drastically reduced puerperal fever incidence among new mothers. Even then, it faced resistance, and took decades of more #science from Pasteur and Lister to be accepted by the public. Do you wash your hands? #edusky 010
Mark Ramos @mframos.bsky.social · 28/01/2025See, this is why ai is bad. The human version is so much better. 010
Mark Ramos @mframos.bsky.social · 25/01/2025Alarming indeed. Replication is the antibody of research. Even a single, well-designed and well-powered replication has a huge impact on the credibility of a new finding. 010
Mark Ramos @mframos.bsky.social · 24/01/2025That time when a study studied if studies on positive publication bias were subject to positive publication bias. 😆 (imo, results look lackluster with only 26 studies and I disagree with some aspects of the interpretation, but this idea is quite fun). #science #researchsky #academicsky #bias #edusky 021
Mark Ramos @mframos.bsky.social · 24/01/2025More people should realize that a ton of things that they regard as #commonsense (e.g. wash your hands, don't drink and drive, stay hydrated etc.) actually come from good, well-communicated science. 010
Mark Ramos @mframos.bsky.social · 23/01/2025Some fun with gganimate. Graph shows journal articles in a hypothetical world where 80% of null hypotheses studied are true, studies have good (80%) power, & publishers have no positive bias. Do you like this world? How close/far are we from it? #statsky #edusky #academicsky #researchsky #iteachmath 020
Mark Ramos @mframos.bsky.social · 21/01/2025One of the surest ways to waste your time when doing #science is to do sorely underpowered studies. If your power is 50%, that's like saying that if the effect is there, you are relying on a coin toss to detect it. #edusky #statsky #academicsky #datascience #phdchat #phdsky #STEM image for #genx 030
Mark Ramos @mframos.bsky.social · 18/01/2025Low p-values don't imply large effects. Large estimates are not necessarily great estimates. Context is everything. #edusky #statsky #academicsky #dataliteracy #datascience #statistics #iteachmath 060
Mark Ramos @mframos.bsky.social · 17/01/2025Interesting rabbit hole to fall into at 5am. P11 is on citation of co-authors while P12 citation of co-authors of co-authors. Lots more studies on this. Let me know if you also fall in. doi.org/10.1371/jour... #edusky #academicsky #academicchatter #researchsky #science #sciencesky #phdchat #phdsky 020
Mark Ramos @mframos.bsky.social · 16/01/2025This should not be the only thing retained by your students from your hypothesis testing lesson (although I suppose it's also not the worst possible scenario). #statistics #edusky #statsky #dataliteracy #datascience #iteachmath #academicsky Art by @bonkerific.bsky.social 271
Mark Ramos @mframos.bsky.social · 15/01/2025Fighting against publication bias is the responsibility of everyone in #science and policymakers should always approach decisions involving new research results with skepticism and much appetite for validation/replication. #statsky #academicchatter #academicsky #edusky #phdchat #datascience #STEM 071
Mark Ramos @mframos.bsky.social · 15/01/2025There are bad actors, but note that type 1 errors happen without cheating. Although yes, it would be nice if we could get the oracle to tell us what each of our type 1 error rates are. Those who have much higher rates than their published significance levels would be suspect. 010
Mark Ramos @mframos.bsky.social · 13/01/2025This image was made by #ai. The time has never been better to teach #datascience to younger students. The tools available now make visualizing abstract concepts much easier and #teachers should take full advantage of it. #edusky #statsky #dataliteracy #academichatter #iteachmath #STEM #academicsky 110
Mark Ramos @mframos.bsky.social · 11/01/2025How is everyone's day going? #edusky #statsky #phdchat #academicChatter #iteachmath #academicsky #statistics #datascience #science #researchlife 020
Mark Ramos @mframos.bsky.social · 11/01/2025Using p-values while ignoring/not setting significance levels is like playing basketball without a hoop. Art by @bonkerific.bsky.social #edusky #statsky #phdchat #academicChatter #iteachmath #academicsky #statistics #datascience #science #researchlife 081
Mark Ramos @mframos.bsky.social · 10/01/2025Still teach #statistics w/these tables? If so, teach your students what they mean, not just how to use them. Otherwise, the lesson is quickly lost and they end up remembering statistics as just “that subject with many tables.” #edusky #statsky #phdchat #academicChatter #iteachmath #academicsky 121
Mark Ramos @mframos.bsky.social · 09/01/2025Frequentist and #Bayesian statisticians may disagree on some things, but we are all interested in uncovering truth from data while fiercely protecting against false positive research outcomes that can have disastrous consequences. #edusky #statsky #phdchat #academicChatter #datascience #academicsky 020
Mark Ramos @mframos.bsky.social · 08/01/2025Slightly belated #HappyBirthday to Émile Borel. Credited (among other great stuff) with Borel's Law of Large Numbers: independent, identical repetitions of an experiment will approximate the probability of every outcome by their proportion #statsky #edusky #AcademicChatter #datascience #statistics 010
Mark Ramos @mframos.bsky.social · 08/01/2025I found this to be a source of cognitive dissonance to grad students. The idea that you can do everything "right" in #statistics analysis and yet be wrong. Risk of error in #science has always been there, stats just quantifies it for you. #edusky #AcademicChatter #phdchat #academicsky #iteachmath 180
Mark Ramos @mframos.bsky.social · 08/01/2025How do you remember your first #statistics course? Mine was decades ago but I vividly remember hating it. Half of it was tedious arithmetic; the other half was weird rules and tables. (bonus if yours had the lesson below) #academicChatter #STEM #edusky #statsky #academicsky #mathed #iteachmath 140
Mark Ramos @mframos.bsky.social · 07/01/2025We must always resist the temptation to use #statistics tools in ways they were not meant to be used. The p-value gets a bad rap not because it's bad, but because people behave badly around it. #edusky #statsky #academicChatter #phdchat #STEM #iteachmath #academicsky #humor #humour #researchLife 061
