"I’ve been studying statistics for over 40 years & I still don’t understand it. The ease with which non-statisticians master it is staggering." -- Stephen John Senn
And if you say that statement to those who most need to hear it, it'll go over their heads and they'll think you understand less about statistics than they originally thought you did. :-) You can't teach intelligence; you can only nurture it if it already exists.
Well i wanna thank you for your stats advice and help throughout one of my biggest challenges in my life! I'm out here doing my thesis on pollinator visitations and it's all non-parametric and there are so many glmm's with various families and repeated measures and it's all a mess and even my advisors struggle with the stats and especially doing them in R. So basically your help has saved my sanity trying to analyze all this data and now I'm hopefully graduating in Spring! Keep on keeping on, appreciate you!
I feel you. I have known researchers at a Max Planck Institute telling off one of my students when telling them that their analysis was incorrect. Same old story: t tests are for everything 😅. Do not dispair! I am learning from you and passing the knowledge to my medical students. Keep doing what you are doing!!! I love it ❤
P.S. 2: and as counter you should do one about the amazing things about being an statistician (or at least, being able to understand a bit of statistics, like in my case 😊)
I went to grad school later in life with a bit more experience than most of my cohort. Was taking a class taught by our brilliant department chair. He asked where we saw ourselves in 5 years. Everyone had these lofty answers, mine was more or less, "I want to be in a place where my analysis is respected and not treated as something on a checklist". He gave me sly smile. It is tough to find such a job.
I am not a statistician, hence why I follow your videos to learn from someone who actually has the expertise and can guide me to get better at it. I hope you realize how invaluable your videos are. They are both incredibly entertaining and extremely useful. I teach applied statistics (among other things) to my students in psychology, and we're trying to shift the positions to bring them to using linear models. Thank you for your honesty with this video, and I am definitely glad that you are a statistician. It's hard to find someone like you who actually can share the motivation to do statistics.
"It is largely because of lack of knowledge of what statistics is that the person untrained in it trusts himself with a tool quite as dangerous as any he may pick out from the whole armamentarium of scientific methodology." --Edwin B. Wilson (1927), quoted in Stephen M. Stigler, The Seven Pillars of Statistical Wisdom.
Thank you for posting this. I am a retired clinician who has been doing unfunded quantitative analyses with published datasets for several years. I have no formal statistical training and have taught myself how to do some things in R. Your videos have often helped me considerably to better understand what I should be doing. I have been lucky in that a few statisticians in my University department will provide advice sometimes on an ad hoc basis. In my interactions with them, I have tried to stick to 2 rules - talk to them from the beginning (designing the research question) and never do anything against their advice. This seems to have worked out for me, as I have been fortunate enough to get a few papers accepted in decent journals. It is about respect for others and for the data. The more I learn, the more I realize how little I really know. Courage, mon brave! Best wishes, Louis
As a first-year graduate student studying biostatistics, I've found your videos to be the most helpful and digestible resource. They have transformed statistics from a source of instant defeat and frustration into a curious adventure. I am pleased to see you posting again and I hope you find comfort in the supportive online community that values your expertise and anticipates your contributions. I will be sure to think more about my statisticians.
I feel you. I love learnkng about statistics but scenarios like this always happen to me. I like well done things. And people always tend to go with the easy stuff only
I also relate to min. 7 Currently in this situation with my PhD supervisor who doesn't use mixed models and thinks they are fancy but he's worked 30 years without so, not needed
Dude, you have no idea the impact you created… only my close circle, we have 17 people, when we have one of the brain frts, we just start watching your channels to figure things out :D You explain greatly. Also, fudge the unethical people! You might be alone in the department, but there are people you have not met yet who still appreciate your time immensely and appreciate your help greatly!!!! hope it will cheer you up a bit :))))
I appreciate it! I'm not that disheartened and didn't intend it to come across as me being super depressed. It's just one of the disadvantages of being a statistician.
I’ve seen a couple of your videos, and I can say as an educational developer at a technical university, that I can relate so much! People say to me and my colleagues: “I have been lecturing about thermodynamics for 20 years, are you saying I don’t know anything about learning?!?” And we’ll be like “If you’re just talking for 2 hours, then…yes! Then they’ll say: “But my students do really well in the exam; they get good grades, and the course evaluations are great!” And we’ll be like: “Does the exam test for conceptual understanding? And btw are you aware that course evaluations don’t measure learning?” At this point they have stopped listening. We are constantly telling people things they don’t want to know! 🤷♀️
I really like mathematics and am currently doing my masters. To make sure I have a backup plan in case I don’t end up in academia, I’m taking some courses in statistics and programming. I’m glad I found this channel!
The point about linear models thumped so hard for me. As a grad student and as an undergrad, I simply followed what I was taught, in all its mind-boggling complexity! Frustrating that there's a better way, and a LOT of unjustified resistance to change...like with most things.
What's funny (and sorry for repeated comments) is that I never had stats classes in my Bachelor bc I studied in a really old school uni in France that taught only psychoanalysis. So I caught up with stats by myself through books and the internet for my Master (at a more scientific uni). I had missed all the Anova&co psych curriculum and learned modelling instead, because the only place where you're told that you should pick Anova&co over modelling is at uni. Or at least, I encountered this mindset neither online nor in books.
Hi, I have recently stumbled upon your great channel and have been binging your videos. To preface this, I am not anyone special. I am currently doing my Masters in Psych/Neuroscience and stats and stats communication is my side job. With that being said, I vehemently disagree with the perspective that everything should be taught as a linear model. Much of what we do in Psychology is linear modelling, but I found that it is simpler to teach things like the t-Test, the F-Test or the ANOVA independetly and then bring them all together into the little facets of the GLM. I usually frustrates me to no avail when people come to me for help and their professors have decided to just teach everything under the guise of a linear model. For them it just seems completly overwhelming. In principle, I get your argument. It would be better if people had a full understanding of the GLM or even the GLIM but purely didactically speaking this is just too much. Let me make an example. I can teach someone the basic of ANOVA in around 15 minutes. They do not need any special knowledge for my explanation. If they understand, I can teach them about the F-Test, which gives me a vehicle to repeat some things about the chi(squared) and the F distribution. If I want to later integrate it into the GLM, they will have heard the logic of ANOVA thrice (for one factor, for multiple factors and for ANCOVA) and the transition into the GLM is easier. If I teach the ANOVA purely through the GLM lense I need to do so much more preperation. My pupils would have to have a somewhat firm grasp about dummy coding, the basics of regression anslysis and the different kinds of linear model parameters. Then on top of this I have to explain the logic of ANOVA, but not the native logic of ANOVA. Instead I have to explain ANOVA in GLM which on it's own is (imo) a bit more difficult to explain than pure ANOVA. I could, of course, just explain ANOVA and then explain an ANOVA in GLM, but this is just the same thing that I initially outlines, just crammed into one session with more effort and less time spent on what ANOVA really does. Maybe as an aside, I think ANOVA is a brilliant procedure and gives a good introduction into quantifying questions that are not straight forward. However, in my experience, if ANOVA is being taught from the GLM perspective students get the impression that ANOVA is part of the GLM and falls to the same assumptions, which is just not true. ANOVA os exceedingly robust to almost all violations, which cannot be said of the GLM. I
Agree to disagree. "I can teach someone the basic of ANOVA in around 15 minutes." Yes, but then you have to take another 15 minutes to explain a one-sample t. Then another 15 for an independent t. Then another for a related t. And so on and so on. I can teach someone the basics of GLM in 15 minutes, then they never have to learn another procedure. Instead, they build on what I've already taught them. "My pupils would have to have a somewhat firm grasp about dummy coding." I wouldn't consider that a prerequisite. I do teach dummy coding, but tell them, "YOU don't have to do this. The computer will do this for you. But I'm showing you what's happening in the background so you know it's not magic." "if ANOVA is being taught from the GLM perspective students get the impression that ANOVA is part of the GLM and falls to the same assumptions, which is just not true. ANOVA os exceedingly robust to almost all violations, which cannot be said of the GLM." I very much disagree. ANOVA is just a different way to rearrange the math, so they're equivalent. The assumptions and robustness are identical. I've got a bunch of videos on this and a textbooks. You're welcome to check it out to see how it works.
Thank you for sharing, and please know you help so many with your willingness to be honest and express how you feel surrounding your work! Personally, I'm a student in a similar position as you were-I like a lot and have no idea what to do once I graduate. The honesty and willingness to express I mentioned earlier are a rarity and a genuine help to me navigating all this. Although I've been fortunate enough to have gotten great opportunities (work at WHO, research assistantships), I still seem to miss people with your willingness to express. I thought I'd comment this given that you wondered whether you should upload this video at all!
Thank you for all your videos! I'm finishing my masters in psychology and I'm planning to become a PhD student in quantitative methods, so I really enjoy your content:) In case your will record a podcast one day I would love to listen!!
Thank you for making this video. As a grad student in stats and data science, I have always wonder how other people view statistics and modelling, and what the ugly side could be.
I am in my final year of undergraduate and I plan to do post graduate in statistics. You are an inspiration and one of my favourite Statistic RUclipsrs. This video is very interesting to hear your perspective from your past work experience, I suppose it is like any job and a lot of it boils down to office politics. Sometimes it can be hard to make decisions you don't believe in or are not comfortable doing simply to entertain your boss. The ethics section is certainly something I have not been taught and most likely won't be which could be leading to biases in the future. I did have a few questions. 1: During undergraduate/graduate school how did you study for your classes? (I don't see many guides on how to study statistics compared to CS or med school etc). 2: What did you mean by teaching just linear models compared to X amount of other ones? Especially for biostatistics would multivariate not be a large area of interest? Please forgive my incompetence and thanks for your time if you read this.
#1-I'm not an expert on studying (and I was, in some ways, a lazy grad student :). #2-I'm referring to teaching about models versus teaching about tests. See this video: ruclips.net/video/p8W4LcdGq6o/видео.htmlsi=6VSHKRk016VuPVJU
Keep doing these videos and stick to your guns! There are a lot of people in STEM who have a lot of issues with linear algebra and still do statistics and give advice.
Your stories seem to happen here and there. It's good to know that reasearch projects need the proper statistical method, not the simplest. Your channel is a lighthouse
Humans are the best and worst part about psychology! 🙂 I'm sorry that you're having a stressful time, but please know that many of us are very grateful that you take the time to explain these concepts. Though some people don't want to improve, many of us do, and we recognize you--and your tools!--for enabling us to do better. From one psychologist (organizational) to another, I hope things improve.
@@QuantPsych very interesting! We had to minor in quant and that was all I needed to know that at my best, I can be a competent user of statistical tools but never able to build them 😅
The problem is... You're in the minority, despite being right. It's like Galileo saying heliocentrism. But everybody else thinks otherwise. Most people are morally bankrupt, are not in the pursuit of truth and research. If more people were in support of your model, viewpoint, you'd get people to agree by sheer force of number. My eyes have been opened since reading accidentally papers on pvalues and stumbling on your channel. Thanks for existing and for your passion and integrity. I'm looking for a way out and it's right on this channel. I have not gone through all the material yet but I'm wondering how I can do this in my field (future bioinfo). But I'm also not sure if I could also go against the status quo either. So it's either stand tall or follow everyone to be accepted, it's very much like social groups, except in science...and I have a hard time to imagine doing compromise on truth. It's like you were accidentally too good at stats. So now you're stuck in this limbo between truth and practicality. And I know, even as a under/graduate student that in biology we're just so ignorant about stats, but now I'm ignorant and in the other side of stats, barely better but I see the wrong stuff. Insane, I exactly thought the same thing about research. Don't be a researcher, this job sucks. I just can't see myself doing something else, I want to solve problems and pursue truth in science. And my other potential interest are even harder and even less profitable (yes it's musician 😂).
Great Video.. I'm a rookie in statistics, just completed my Masters in Stats, coming from a different background. This might be a great heads up to what I CAN face in the industry.
Hey! Quick question, if it's bad practise to dichotomize continuous variables (which totally makes sense), then how does one estimate say for example, the effect of weather shocks on agricultural produce? This is something I have been working on, where I take within entity deviations from the long run average and then dichotimize it based on an officially documented cutoff. But if there is no cutoff, how do I define what is a shock in my case?
Enjoyed the video, you can do a top 10 crimes against statistics, with discussion. Your valuable experience on where we as non-statisticians typically mess up would be very valuable.
It just makes me sad about certain areas of research. I came into grad school passionate about understanding human nature through psych and will be graduating from my program skeptical of the entire field (minus the quant psych area, of course)
Don't spare quant from your skepticism :) I had a very similar experience. That's why I went into biostats, only to find out biologists/medical researchers are way worse than psychologists.
The best kinds of researchers to consult with are either those who know nothing about statistics or admit no knowledge. Those who think they know something are much more difficult to deal with because almost every word that comes out of their mouth is incorrect. They have zero understanding of how much one has to work with and study this stuff for it to even begin to make sense. If they knew how much quantitative folks struggle with this stuff themselves, they wouldn't even try to talk stats. I've found that the smarter the researcher, the more they DON'T try to talk stats with you. The "less intelligent" folks often lack the humility to recognize they know nothing. Why? Because they think stats has something to do with SPSS.
Thank you for this video!! Can you go into the differences between linear regression vs. the “decision tree” of tests in more detail? Is it a matter of pros and cons of methods or just old vs new techniques? One obvious thing that comes to my mind is one can’t account for repeated sampling with a t-test or ANOVA, right?
I'm working on three separate courses, actually (one for each). I'm hoping to launch mixed models in the next few weeks, then I'll work on the others :)
Ahahah I had a large group of 11 to 27 years olds in my Master's thesis, it was actually a lot of work to get enough people of all ages, and then colleagues (with more authority than my student self) asked me to split them into 2 groups, adolescents vs adults. So sad xd
i am a "pure" math phd that left academia and work as a data analyst currently. i have had this goal of working as a statistician for some place like the CDC, census bureau, bureau of labor stats -- is it really that bad?
I warn researchers sometimes that their model is severely mis-specified, their parameter estimates are biased, and their model is grossly misleading to the point that it may have been more ethical to not run the model at all or they should at minimum report the weaknesses of their approach (and the statistical method they used) in their discussion. They don't care. Not one bit. They don't care about what's ethical in the use of statistics. If it gets published (reviewed by reviewers who don't care one bit either and don't know any better), that's all that matters. Publication mill factory. Just list it on the CV and everyone is happy. And if you bring up vital statistical or philosophy of science concerns about their research even in a very friendly way, you're just being "theoretical" according to them because they have no chance of understanding the argument even if you water it down significantly. Lonely? Very. You just have to walk away. Trying to teach a clinician that because they used a "causal model" does not by itself imply X "causes" Y in their data is just an exercise in frustration. Philosophy of science 101, and these researchers with 30-page CVs filled with garbage can't grasp it and just think you're being a stickler or difficult. They have never once thought about the limitations of statistical modeling in their lives because they are too busy filling their CVs with SPSS-generated BS and playing the publication game.
People with high on agreeableness and high conscientiousness MUST know what they want otherwise they will get taken advantage of. If you're going to be high on conscientiousness, it's best to be assertive because there are already few of you anyway. I am sure you've heard of Pareto- 80% of productivity comes from 20% of people. It's best for EVERYONE that people high on conscientiousness are low on agreeableness because human progress will unfold more rapidly.
"I’ve been studying statistics for over 40 years & I still don’t understand it. The ease with which non-statisticians master it is staggering." -- Stephen John Senn
Ha! I love that :)
And if you say that statement to those who most need to hear it, it'll go over their heads and they'll think you understand less about statistics than they originally thought you did. :-) You can't teach intelligence; you can only nurture it if it already exists.
"It is difficult to get a man to understand something when his salary depends upon his not understanding it." - Upton Sinclair
I have found when someone wants you to keep them honest, what they really mean is they want you to tell them what they are doing is right.
That sounds about right :)
Well i wanna thank you for your stats advice and help throughout one of my biggest challenges in my life! I'm out here doing my thesis on pollinator visitations and it's all non-parametric and there are so many glmm's with various families and repeated measures and it's all a mess and even my advisors struggle with the stats and especially doing them in R. So basically your help has saved my sanity trying to analyze all this data and now I'm hopefully graduating in Spring! Keep on keeping on, appreciate you!
Yay! Congrats!
I feel you. I have known researchers at a Max Planck Institute telling off one of my students when telling them that their analysis was incorrect. Same old story: t tests are for everything 😅. Do not dispair! I am learning from you and passing the knowledge to my medical students. Keep doing what you are doing!!! I love it ❤
P.S. please tell us about all the crazy request scientists ask you to do that drives you nuts 😂😊
I will!
P.S. 2: and as counter you should do one about the amazing things about being an statistician (or at least, being able to understand a bit of statistics, like in my case 😊)
I went to grad school later in life with a bit more experience than most of my cohort. Was taking a class taught by our brilliant department chair. He asked where we saw ourselves in 5 years. Everyone had these lofty answers, mine was more or less, "I want to be in a place where my analysis is respected and not treated as something on a checklist". He gave me sly smile. It is tough to find such a job.
I LOVE your answer :)
I am not a statistician, hence why I follow your videos to learn from someone who actually has the expertise and can guide me to get better at it. I hope you realize how invaluable your videos are. They are both incredibly entertaining and extremely useful. I teach applied statistics (among other things) to my students in psychology, and we're trying to shift the positions to bring them to using linear models. Thank you for your honesty with this video, and I am definitely glad that you are a statistician. It's hard to find someone like you who actually can share the motivation to do statistics.
Thank you!
i am not in anyway involved with doing statistics. i just love hearing a man be real about things.
Lovin this. Maybe your best video of all.
"It is largely because of lack of knowledge of what statistics is that the person untrained in it trusts himself with a tool quite as dangerous as any he may pick out from the whole armamentarium of scientific methodology."
--Edwin B. Wilson (1927), quoted in Stephen M. Stigler, The Seven Pillars of Statistical Wisdom.
Thank you for posting this. I am a retired clinician who has been doing unfunded quantitative analyses with published datasets for several years. I have no formal statistical training and have taught myself how to do some things in R. Your videos have often helped me considerably to better understand what I should be doing. I have been lucky in that a few statisticians in my University department will provide advice sometimes on an ad hoc basis. In my interactions with them, I have tried to stick to 2 rules - talk to them from the beginning (designing the research question) and never do anything against their advice. This seems to have worked out for me, as I have been fortunate enough to get a few papers accepted in decent journals. It is about respect for others and for the data. The more I learn, the more I realize how little I really know. Courage, mon brave! Best wishes, Louis
I wish everyone followed that advice :)
I feel you too. And I'm not even a statistician, but seeing colleagues doing wrong things, or telling wrong things to students drives me nuts.
Totally.
As a first-year graduate student studying biostatistics, I've found your videos to be the most helpful and digestible resource. They have transformed statistics from a source of instant defeat and frustration into a curious adventure. I am pleased to see you posting again and I hope you find comfort in the supportive online community that values your expertise and anticipates your contributions. I will be sure to think more about my statisticians.
Thank you! It's so awesome to hear the videos are helping others.
I have a feeling this is going to get shared - a lot. Completely resonate
I feel you. I love learnkng about statistics but scenarios like this always happen to me. I like well done things. And people always tend to go with the easy stuff only
I also relate to min. 7
Currently in this situation with my PhD supervisor who doesn't use mixed models and thinks they are fancy but he's worked 30 years without so, not needed
SO frustrating!
Dude, you have no idea the impact you created… only my close circle, we have 17 people, when we have one of the brain frts, we just start watching your channels to figure things out :D You explain greatly. Also, fudge the unethical people! You might be alone in the department, but there are people you have not met yet who still appreciate your time immensely and appreciate your help greatly!!!! hope it will cheer you up a bit :))))
I appreciate it! I'm not that disheartened and didn't intend it to come across as me being super depressed. It's just one of the disadvantages of being a statistician.
I’ve seen a couple of your videos, and I can say as an educational developer at a technical university, that I can relate so much! People say to me and my colleagues: “I have been lecturing about thermodynamics for 20 years, are you saying I don’t know anything about learning?!?” And we’ll be like “If you’re just talking for 2 hours, then…yes! Then they’ll say: “But my students do really well in the exam; they get good grades, and the course evaluations are great!” And we’ll be like: “Does the exam test for conceptual understanding? And btw are you aware that course evaluations don’t measure learning?” At this point they have stopped listening. We are constantly telling people things they don’t want to know! 🤷♀️
I really like mathematics and am currently doing my masters. To make sure I have a backup plan in case I don’t end up in academia, I’m taking some courses in statistics and programming. I’m glad I found this channel!
The point about linear models thumped so hard for me. As a grad student and as an undergrad, I simply followed what I was taught, in all its mind-boggling complexity!
Frustrating that there's a better way, and a LOT of unjustified resistance to change...like with most things.
What's funny (and sorry for repeated comments) is that I never had stats classes in my Bachelor bc I studied in a really old school uni in France that taught only psychoanalysis. So I caught up with stats by myself through books and the internet for my Master (at a more scientific uni). I had missed all the Anova&co psych curriculum and learned modelling instead, because the only place where you're told that you should pick Anova&co over modelling is at uni. Or at least, I encountered this mindset neither online nor in books.
I wish people in academia knew that! They seem to think it's the only way it's taught. But it's not!
Hi, I have recently stumbled upon your great channel and have been binging your videos.
To preface this, I am not anyone special. I am currently doing my Masters in Psych/Neuroscience and stats and stats communication is my side job.
With that being said, I vehemently disagree with the perspective that everything should be taught as a linear model. Much of what we do in Psychology is linear modelling, but I found that it is simpler to teach things like the t-Test, the F-Test or the ANOVA independetly and then bring them all together into the little facets of the GLM.
I usually frustrates me to no avail when people come to me for help and their professors have decided to just teach everything under the guise of a linear model. For them it just seems completly overwhelming.
In principle, I get your argument. It would be better if people had a full understanding of the GLM or even the GLIM but purely didactically speaking this is just too much.
Let me make an example. I can teach someone the basic of ANOVA in around 15 minutes. They do not need any special knowledge for my explanation. If they understand, I can teach them about the F-Test, which gives me a vehicle to repeat some things about the chi(squared) and the F distribution.
If I want to later integrate it into the GLM, they will have heard the logic of ANOVA thrice (for one factor, for multiple factors and for ANCOVA) and the transition into the GLM is easier.
If I teach the ANOVA purely through the GLM lense I need to do so much more preperation. My pupils would have to have a somewhat firm grasp about dummy coding, the basics of regression anslysis and the different kinds of linear model parameters.
Then on top of this I have to explain the logic of ANOVA, but not the native logic of ANOVA. Instead I have to explain ANOVA in GLM which on it's own is (imo) a bit more difficult to explain than pure ANOVA.
I could, of course, just explain ANOVA and then explain an ANOVA in GLM, but this is just the same thing that I initially outlines, just crammed into one session with more effort and less time spent on what ANOVA really does.
Maybe as an aside, I think ANOVA is a brilliant procedure and gives a good introduction into quantifying questions that are not straight forward. However, in my experience, if ANOVA is being taught from the GLM perspective students get the impression that ANOVA is part of the GLM and falls to the same assumptions, which is just not true. ANOVA os exceedingly robust to almost all violations, which cannot be said of the GLM.
I
Agree to disagree.
"I can teach someone the basic of ANOVA in around 15 minutes." Yes, but then you have to take another 15 minutes to explain a one-sample t. Then another 15 for an independent t. Then another for a related t. And so on and so on. I can teach someone the basics of GLM in 15 minutes, then they never have to learn another procedure. Instead, they build on what I've already taught them.
"My pupils would have to have a somewhat firm grasp about dummy coding." I wouldn't consider that a prerequisite. I do teach dummy coding, but tell them, "YOU don't have to do this. The computer will do this for you. But I'm showing you what's happening in the background so you know it's not magic."
"if ANOVA is being taught from the GLM perspective students get the impression that ANOVA is part of the GLM and falls to the same assumptions, which is just not true. ANOVA os exceedingly robust to almost all violations, which cannot be said of the GLM." I very much disagree. ANOVA is just a different way to rearrange the math, so they're equivalent. The assumptions and robustness are identical.
I've got a bunch of videos on this and a textbooks. You're welcome to check it out to see how it works.
Thank you for sharing, and please know you help so many with your willingness to be honest and express how you feel surrounding your work!
Personally, I'm a student in a similar position as you were-I like a lot and have no idea what to do once I graduate. The honesty and willingness to express I mentioned earlier are a rarity and a genuine help to me navigating all this. Although I've been fortunate enough to have gotten great opportunities (work at WHO, research assistantships), I still seem to miss people with your willingness to express.
I thought I'd comment this given that you wondered whether you should upload this video at all!
That's great to hear. :)
Thank you for all your videos! I'm finishing my masters in psychology and I'm planning to become a PhD student in quantitative methods, so I really enjoy your content:) In case your will record a podcast one day I would love to listen!!
Thank you for making this video. As a grad student in stats and data science, I have always wonder how other people view statistics and modelling, and what the ugly side could be.
A very nice video! Thank you for sharing
I am in my final year of undergraduate and I plan to do post graduate in statistics. You are an inspiration and one of my favourite Statistic RUclipsrs. This video is very interesting to hear your perspective from your past work experience, I suppose it is like any job and a lot of it boils down to office politics. Sometimes it can be hard to make decisions you don't believe in or are not comfortable doing simply to entertain your boss. The ethics section is certainly something I have not been taught and most likely won't be which could be leading to biases in the future. I did have a few questions.
1: During undergraduate/graduate school how did you study for your classes? (I don't see many guides on how to study statistics compared to CS or med school etc).
2: What did you mean by teaching just linear models compared to X amount of other ones? Especially for biostatistics would multivariate not be a large area of interest?
Please forgive my incompetence and thanks for your time if you read this.
#1-I'm not an expert on studying (and I was, in some ways, a lazy grad student :).
#2-I'm referring to teaching about models versus teaching about tests. See this video: ruclips.net/video/p8W4LcdGq6o/видео.htmlsi=6VSHKRk016VuPVJU
Keep doing these videos and stick to your guns! There are a lot of people in STEM who have a lot of issues with linear algebra and still do statistics and give advice.
agreed
Thanks a lot for sharing. I am in the same boat, and not alone.
I'm trying to become just like you fr
Your stories seem to happen here and there. It's good to know that reasearch projects need the proper statistical method, not the simplest. Your channel is a lighthouse
Humans are the best and worst part about psychology! 🙂 I'm sorry that you're having a stressful time, but please know that many of us are very grateful that you take the time to explain these concepts. Though some people don't want to improve, many of us do, and we recognize you--and your tools!--for enabling us to do better. From one psychologist (organizational) to another, I hope things improve.
Thanks! (I actually started grad school to do IO psych, then changed to quant :))
@@QuantPsych very interesting! We had to minor in quant and that was all I needed to know that at my best, I can be a competent user of statistical tools but never able to build them 😅
The problem is... You're in the minority, despite being right. It's like Galileo saying heliocentrism. But everybody else thinks otherwise.
Most people are morally bankrupt, are not in the pursuit of truth and research.
If more people were in support of your model, viewpoint, you'd get people to agree by sheer force of number.
My eyes have been opened since reading accidentally papers on pvalues and stumbling on your channel.
Thanks for existing and for your passion and integrity. I'm looking for a way out and it's right on this channel.
I have not gone through all the material yet but I'm wondering how I can do this in my field (future bioinfo). But I'm also not sure if I could also go against the status quo either. So it's either stand tall or follow everyone to be accepted, it's very much like social groups, except in science...and I have a hard time to imagine doing compromise on truth.
It's like you were accidentally too good at stats. So now you're stuck in this limbo between truth and practicality. And I know, even as a under/graduate student that in biology we're just so ignorant about stats, but now I'm ignorant and in the other side of stats, barely better but I see the wrong stuff.
Insane, I exactly thought the same thing about research. Don't be a researcher, this job sucks. I just can't see myself doing something else, I want to solve problems and pursue truth in science. And my other potential interest are even harder and even less profitable (yes it's musician 😂).
Great Video.. I'm a rookie in statistics, just completed my Masters in Stats, coming from a different background. This might be a great heads up to what I CAN face in the industry.
Hopefully you won't!
Hey! Quick question, if it's bad practise to dichotomize continuous variables (which totally makes sense), then how does one estimate say for example, the effect of weather shocks on agricultural produce? This is something I have been working on, where I take within entity deviations from the long run average and then dichotimize it based on an officially documented cutoff. But if there is no cutoff, how do I define what is a shock in my case?
Enjoyed the video, you can do a top 10 crimes against statistics, with discussion. Your valuable experience on where we as non-statisticians typically mess up would be very valuable.
Great idea!
It just makes me sad about certain areas of research. I came into grad school passionate about understanding human nature through psych and will be graduating from my program skeptical of the entire field (minus the quant psych area, of course)
Don't spare quant from your skepticism :) I had a very similar experience. That's why I went into biostats, only to find out biologists/medical researchers are way worse than psychologists.
The best kinds of researchers to consult with are either those who know nothing about statistics or admit no knowledge. Those who think they know something are much more difficult to deal with because almost every word that comes out of their mouth is incorrect. They have zero understanding of how much one has to work with and study this stuff for it to even begin to make sense. If they knew how much quantitative folks struggle with this stuff themselves, they wouldn't even try to talk stats. I've found that the smarter the researcher, the more they DON'T try to talk stats with you. The "less intelligent" folks often lack the humility to recognize they know nothing. Why? Because they think stats has something to do with SPSS.
Thank you for this video!! Can you go into the differences between linear regression vs. the “decision tree” of tests in more detail? Is it a matter of pros and cons of methods or just old vs new techniques? One obvious thing that comes to my mind is one can’t account for repeated sampling with a t-test or ANOVA, right?
I think this video will address that: ruclips.net/video/KwVl_K_TLxo/видео.html
@@QuantPsych Doesn't get much better than that! Thank you! 😅
hey are you still making a hotchpotch course on random forest / mixed models / generalized linear models? I'd buy that instantly Prof. Sir Statsman
I'm working on three separate courses, actually (one for each). I'm hoping to launch mixed models in the next few weeks, then I'll work on the others :)
Ahahah I had a large group of 11 to 27 years olds in my Master's thesis, it was actually a lot of work to get enough people of all ages, and then colleagues (with more authority than my student self) asked me to split them into 2 groups, adolescents vs adults. So sad xd
I need to make a video about median splits and how terrible they are. Then you can send it to them!
Hi Dustin. The undergrad and grad curriculum playlist links in the description aren’t working for me. Can you check the links?
Oops! I just fixed it.
i am a "pure" math phd that left academia and work as a data analyst currently. i have had this goal of working as a statistician for some place like the CDC, census bureau, bureau of labor stats -- is it really that bad?
I’m a ”pure” math masters student. I’m pretty sure you could make this kind of video for just about any profession.
I warn researchers sometimes that their model is severely mis-specified, their parameter estimates are biased, and their model is grossly misleading to the point that it may have been more ethical to not run the model at all or they should at minimum report the weaknesses of their approach (and the statistical method they used) in their discussion. They don't care. Not one bit. They don't care about what's ethical in the use of statistics. If it gets published (reviewed by reviewers who don't care one bit either and don't know any better), that's all that matters. Publication mill factory. Just list it on the CV and everyone is happy. And if you bring up vital statistical or philosophy of science concerns about their research even in a very friendly way, you're just being "theoretical" according to them because they have no chance of understanding the argument even if you water it down significantly. Lonely? Very. You just have to walk away. Trying to teach a clinician that because they used a "causal model" does not by itself imply X "causes" Y in their data is just an exercise in frustration. Philosophy of science 101, and these researchers with 30-page CVs filled with garbage can't grasp it and just think you're being a stickler or difficult. They have never once thought about the limitations of statistical modeling in their lives because they are too busy filling their CVs with SPSS-generated BS and playing the publication game.
Well said!
I’d be proud of this video if I had made it :)
Min 3'20": how tall Is tall? .. LOL... Everyday Situation of people asking for dicotomization
OK brendan fraser
?
And the Oscar goes to...
Maybe in my next life :)
35 and alone
Me? Or you? (I'm 39 :)
❤
People with high on agreeableness and high conscientiousness MUST know what they want otherwise they will get taken advantage of. If you're going to be high on conscientiousness, it's best to be assertive because there are already few of you anyway. I am sure you've heard of Pareto- 80% of productivity comes from 20% of people. It's best for EVERYONE that people high on conscientiousness are low on agreeableness because human progress will unfold more rapidly.