Видео

Great class!

Were you not aware of FIML when making this video, or do you think MI is just a better method? My understanding is FIML is ideal - it doesn't require any extra work and is also less biased because there's no model assumption.

Thank you, you explain robust SE excellently. You are a master teacher, thank you. Your application is questionable. Many authors, such as David Freedman defend a slightly misspecified models as still interpretable if the evidence of an effect is still clear (effect estimate is larger than the robust SE).. It is a matter of context. Whether to go for formal perfection (model SE=robust SE) or not is often just a matter of taste. You are a perfectionist. That is a complement, by the way. The larger problem with your application is the use of the wacky Box-Cox transformation. The Box-Cox is widely ridiculed for producing totally uninterpretable results. (The formula is a block box and what does a Box-Cox transformed outcome mean, anyway?). A better solution perhaps would have been a generalized linear model, using a Gamma distribute outcome, for instance, to take account for the non-normalty of the outcome. .So you chose formal perfection (insist on equality of standard errors) but then implement it in a wacky way (Box Cox rather than GLM). Your approach seems somewhat strange to this PhD statistician.... I'd be grateful to hear a response.... Thanks again! P.S> There are even more modern machine learning approaches like Random Forest that could be used on your application...

You are wrong. A histogram is a good judge of a distribution being normal, by definition. You are a poor speaker. You have deceived yourself. I suppose what you meant to ask was, is a histogram a good judge of whether the source distribution is normal, as a producer distribution that produced the final curve. You have done nothing in the question to verbally distinguish the generating distribution from the final distribution as a collection of points. You were so anxious to create a trick question that you didn't outline the question correctly. -5 points for you

Insofar as you were worried, you should know that I laughed out loud at your product placement joke 16 minutes in.

Professor, I just came across this video and found it incredibly insightful, even 3 years after its publication. However, I have a question about a point you raised at 57:52. You mentioned that in the ideal social science experiment, ΔSx and ΔSu shouldn't be zero. From my understanding, though, in such experiments, it's generally crucial to aim for ΔSx = 0 and ΔSu = 0. This would mean ensuring equal variability (ΔSx) in the treatment and control groups, and no disparity in unmeasured variables (ΔSu) between them. The goal, as I understand it, is to isolate the treatment as the sole differing factor between the groups, allowing us to attribute any differences in outcomes specifically to the treatment. Could you please shed some light on this?

Very intuitive explanation, Many thanks

Muito bom.

Thank you for the good lecture :) Your video is so very helpful for me to understand a matching method.

14:00 what is about 20:10 Who take this course? 23:50 reading 26:27 teach you 35:19 fundamentals 35:40 practical statistics 36:12 quantitative 38:40. Undergraduate write for journals? 43:20 peer reviewed 43:40 Model is an abstractoon 47:00

Dear Professor, what is the code of this course in Perusal?

Excellent Video! Love the fact that you relate your explanations to real life problems

Excellent lectures, thanks very much!

Prof King, could you please tell us what statistical tool was used to visualise the data (e.g. the T and C groups)? Thanks!

I am attempting to identify the appropriate code for calculating clustered standard errors following the execution of a regression using the multinom() function in r. I attempted the following code but consistently encountered an error: Calculate the cluster-robust variance-covariance matrix vcov_clustered <- vcovCL(reg1, cluster = mtcars$cyl) Error in UseMethod("estfun") : no applicable method for 'estfun' applied to an object of class "c('multinom', 'nnet')"

Thats all well and good, but how do I use MDM in SPSS?

Extremely helpful lecture on RSE.

Hello, i have a question. if i have a countinuous variable, do i need to reshape to a categorical before running cem? thank you

Is there a book you published with the content of this course?

Sería hermoso que alguien subtitulara esto

Thank you so much! This was exactly what I needed.

Thank you so much to you and your team for these amazing resources! Are some non-harvard students interested in starting a study group on Perusall?I am a bit confused if non-harvard students can actually create a study group on perusall btw, and which code to use (I didn't manage to use the codes "6AFEYRZH4P" and "PCDKPTWZ39"). Thank you!

Thank you! Great talk. When working with say 20 variables, could I do PCA first and if the first two components explained 98% of variability, then run the matching using the 2 components instead?

may GOD double or triple the lives of people like you Prof Gary King for making the world better

Absolutely fantastic!

Your video is a blessing

Warm greetings and thanks so much! 🌻🌻🌻 In two weeks final exam in statistics (Master / Psychology)... Now there is hope! Greetings from Germany 🙃🌲

What a clear and succinct explanation! Thank you Prof Gary.

Best wishes. Prof.

love your articulation !

Doesn't adding the sampling distribution concept to the likelihood inference framework eliminates the sampling invariance advantage over Neyman-Pearson framework?

If you would like to create a study group let me know I'll begin watching the lectures

I just don't understand why all the videos have the audio skewed always to the left speaker, kind of annoying but the content is very good

hi, sir, this video is really helpful! thank you so much!!! :) may i know if the pdf version is available?

Thankyou, prof!

Thank you very much for this lesson Professor King.

Thank you Prof G.K

Excellent. Love the energy and content explanations- question. Where can I find the model frontier website/ algorithm. Which software is it coded on. Many thanks. Loved it!!

Wish all stats professors were this clear and engaging. Thanks, this helped a lot.

Giving a thumbs up IN SPITE OF the misery the KKV book "Designing Social Inquiry" has visited upon grad students. ;-) As a poli sci research methods instructor myself, I'm very grateful for this. I feel like I need refreshers regularly! Thanks, Dr. King.

Thank you so much, Professor King!

I want to give this a thousand likes!

dude there's nothing more inspiring and energizing than seeing a teacher who loves teaching.

love the product placement

I'm trying to study Political Science using Harvard's curriculum. I noticed that, unfortunately, most course materials are gated. Your course materials and all the extra educational resources posted on your website are one of the few exceptions. Thank you for your generosity and for your hard work. You're enriching more lives than you know. ❤

Really great explanation. Thanks a lot

Thank you.

Prof King, this was an extremely informative lecture, delivered with great elegance. Thank you from a medical student in India.

Thank you so much ! I fully understand PSM now!

I did matching to trim my control set using PSM and I saw my matches, it was really bad. This made me look at other methods and I landed here. Thanks for sharing this. This is really helpful.