SMart teaching, intelligent approach, all with sense of humour; thank you for not considering us dafts! You teaching 100% worths my time, have been watching your videos several times and it does help me a lot!
When I was in my math pedagogy courses, I learned that if I can’t explain something simply, I don’t really understand it yet. Thank you for doing what I have never seen another stats prof do (explain it simply).
Super super helpful! Ready to go through the series of your glm videos. T-test, ANOVA, and all the other so-called "statistical tests" have been bothering me for so long. The equation: "Outcome= intercept + slope*predictor +e " is just simplistic and useful!! Thank you so much for making these videos.
Hilarious title! EDIT: This video is equally hilarious. EDIT 2.0: I never expected to have a laughing fit while learning statistics. Thanks for the videos!
super interesting and informative！ Thanks for this great video!
2 ปีที่แล้ว
After watching this video I'm literally super excited to watch the rest of the videos. I think you should be doing telemarketing or something like that.
Thank you for giving the formulas to connect the GLM with the decision tree BS. The GLM makes more sense because (for more of us than the alternative), H0: we have prior learning that we can connect GLM to. The other stats methods? Yeah… it mostly just makes the case for more stats profs to scaffold their instruction.
I just have a kind of existential question: What are polynomials really doing when it comes to real life facts.. like how do they model stuff, what does the following equation: 4*(#of_tv_hours)^2 + 2*(#of_tv_hours) +7 have to do with real life, with what's actually happening in reality( like when someone watches the tv for one hour, 100 cells in his brain are damaged..etc that could be modelled perfectly with this but otherwise, why use that??) I know it doesn't have to be accurate but approximate enough.. so.. why does it make it approximate enough?! is everything in life related together with some polynomial out there? What's so sacred about polynomials? (polynomials are very obvious in physics/dynamics,i.e: throwing a ball upwards make it make a shape of parabola, the rule of physics are built on polynomials. But how do relations between some facts in life are related to this... and then if we could model it, how can we reduce that relationship modeled by a polynomial into a linear equations... where did the quadratic and cubic relationships go!!?? @QuantPsych
These videos should be shared to lots of researchers. Do you have a twitter account of something like that ? Would help sharing. BTW, very nice as usual !
Vincent Cabibel thanks! I have two Twitter accounts: @dustinfife and @jedistats, but I rarely use them. The second one, though, is managed by one of my research students.
A linear model, an ANOVA, and a T-test enter to a bar and the bartender ask “are you waiting for someone or just be by yourself?” 😂 Great videos, do you have anyone explaining gls(), for instance when your data have linear relationships but not homogeneity of variation? (Note, I don’t want to transform variables because the complication of understanding slopes transformed)
Hi, sorry I'm late to the party. Great video, but I have a question: Isn't measuring data in interval or ratio scale an asumption we make? Would depression score but interval or just ordinal scale? What about IQ? Thanks for your videos.
Technically we make that assumptions, but it turns out that violating that assumption isn't all that problematic. If I weren't lazy, I'd give you a reference, but I am lazy :)
@@QuantPsych Thanks for the answer. Thinking about it it is probably hidden somewhere in the linearity asumption. If the predictor scale is not linear then predicting with a linear model is probably not the right approach... Great videos btw.
I didn't know statistics could make me happy but THIS LITERALLY CURED MY DEPRESSION
You may have undiagnosed ADHD
One of the best statistics teacher here on youtube! Thank you!
SMart teaching, intelligent approach, all with sense of humour; thank you for not considering us dafts! You teaching 100% worths my time, have been watching your videos several times and it does help me a lot!
When I was in my math pedagogy courses, I learned that if I can’t explain something simply, I don’t really understand it yet. Thank you for doing what I have never seen another stats prof do (explain it simply).
Super super helpful! Ready to go through the series of your glm videos. T-test, ANOVA, and all the other so-called "statistical tests" have been bothering me for so long. The equation: "Outcome= intercept + slope*predictor +e " is just simplistic and useful!! Thank you so much for making these videos.
Hilarious title!
EDIT: This video is equally hilarious.
EDIT 2.0: I never expected to have a laughing fit while learning statistics. Thanks for the videos!
“You can suck it reviewers “
I can totally relate to that 🤪!
Thanks for the videos!
Awesome prof!! Your explanations are simply awesome!!!
I'm preparing for my stat exam, and think that I'm so dumb to understand all those concepts. but your videos help a lot rn. thank u !!!
Very nicely explained .. amazing way of teaching ! Thanks
super interesting and informative！ Thanks for this great video!
After watching this video I'm literally super excited to watch the rest of the videos. I think you should be doing telemarketing or something like that.
I worked at a call center once and hated it more than any other job I've had :)
It worked. Indeed, it worked. I'm a subscriber now!
Thank you for giving the formulas to connect the GLM with the decision tree BS. The GLM makes more sense because (for more of us than the alternative), H0: we have prior learning that we can connect GLM to. The other stats methods? Yeah… it mostly just makes the case for more stats profs to scaffold their instruction.
It's hilarious!! You are doing great using just your own face and a couch!
🤯 nice
Thank you so much for these amazing videos!! :)
Thank you for your amazing comment :)
At 5:47, why is the model concluded as not fitting the data?
I just have a kind of existential question: What are polynomials really doing when it comes to real life facts.. like how do they model stuff, what does the following equation:
4*(#of_tv_hours)^2 + 2*(#of_tv_hours) +7
have to do with real life, with what's actually happening in reality( like when someone watches the tv for one hour, 100 cells in his brain are damaged..etc that could be modelled perfectly with this but otherwise, why use that??) I know it doesn't have to be accurate but approximate enough.. so.. why does it make it approximate enough?!
is everything in life related together with some polynomial out there? What's so sacred about polynomials? (polynomials are very obvious in physics/dynamics,i.e: throwing a ball upwards make it make a shape of parabola, the rule of physics are built on polynomials. But how do relations between some facts in life are related to this...
and then if we could model it, how can we reduce that relationship modeled by a polynomial into a linear equations... where did the quadratic and cubic relationships go!!??
@QuantPsych
These videos should be shared to lots of researchers. Do you have a twitter account of something like that ? Would help sharing. BTW, very nice as usual !
Vincent Cabibel thanks! I have two Twitter accounts: @dustinfife and @jedistats, but I rarely use them. The second one, though, is managed by one of my research students.
Why the cyanide and happiness theme?
A linear model, an ANOVA, and a T-test enter to a bar and the bartender ask “are you waiting for someone or just be by yourself?” 😂
Great videos, do you have anyone explaining gls(), for instance when your data have linear relationships but not homogeneity of variation? (Note, I don’t want to transform variables because the complication of understanding slopes transformed)
Ha! Love the joke.
I don't have videos on what....at least not yet.
GLM = generalized linear model != general linear model
correct.
what software did you show in this video, sir? Looks like a good and easy software to use
please reply here if you know
@@manojkumarpandiri8031 It's SPSS
Its jasp , a free version of spss if im not wrong,
Late to the party, but it's JASP and it rules.
Hi, sorry I'm late to the party. Great video, but I have a question: Isn't measuring data in interval or ratio scale an asumption we make? Would depression score but interval or just ordinal scale? What about IQ? Thanks for your videos.
Technically we make that assumptions, but it turns out that violating that assumption isn't all that problematic. If I weren't lazy, I'd give you a reference, but I am lazy :)
@@QuantPsych Thanks for the answer. Thinking about it it is probably hidden somewhere in the linearity asumption. If the predictor scale is not linear then predicting with a linear model is probably not the right approach... Great videos btw.