I think one important part of statistics a lot of people overlook is simply all of the assumptions being made behind the scenes. There are so many cases where you may use a particular test or model (even simple linear regression) where there are assumptions being made about sample size, normality, equal variance of residuals, etc., where you won’t get any warning or error in your code but the results may be untrustworthy because one or more assumptions are being violated. Always know what assumptions are being made when you implement certain methods, and to what degree these methods are robust to violations! It could mean the difference of being able to fit a desired model on the data directly, or needing to manipulate the data/methods drastically to be able to trust the results. Just because you see statistical significance does not mean something is significant, it means it is significant given all these other facts about the data hold true
Additional Tip: When studying a graph in addition to looking at the shape; always look at what is on the X and Y axis. For example, on a bell curve, what is on the Y-axis? Why?
Great video as always! I am just going through all of your videos. I am doing lots of regression analysis as a postdoctoral researcher and hearing that hypothesis testing is not a big thing in DS makes me a little sad haha. I find this part to be fun and I thought it would be useful in case I want to leave academia. For us, it is all about the p-value and how substantial the results are.
This is really helpful. I learned from many sources and really got stuck for not knowing which I “should” know. Lots of intermediate and advanced concepts that I learned are not really applicable in my scope of work, then I almost forgot them all :)) It is the best summary video of statistics for data analysts I have ever watched :) Thank you a lot
That’s great to hear! Yes I can totally relate to that, so much information and I’d feel paralyzed to decide what to learn first. Thank you so much for your supporting comment. All the best 🍀
OMG i actually understand what you are saying. I have just started my DA journey..week 3. Past 2 weeks, i have been swimming in the sea of statistics, just trying not to drown ^^. Please post more videos on statistics, tips on data cleaning, critical thinking, data presentation and other related topics to DA and DS. TQ heaps!!
A very comprehensive video for statistics for data analysis...I myself been trying to get into the data analysis field but as you said there is a lot of misguidances....you know what you are doing..much appreciated
Hey Vishal, thank you for watching and commenting! Indeed I have been there and found it hard to navigate myself, there are too many resources and things to learn. But the most important thing is to get the first few solid basic building blocks, then you can expand from there. All the best 👋
@@Thuvu5Thank you for your reply...Given the experience you have in this field I will be really glad I can take a few pointers from you... only if there was a way to connect with you.
@@vishalrana9594 Hey Vishal, sure! It'd be great to connect. You can find my email address and the link to my Linkedin on the "About" page of my YT channel. Talk soon!
I have just started learning data science and I watching many videos related to it. UA-cam recommend one of your videos and now have watched 3-4 videos of yours. Reall good resource. Keep doing the good work.
Thanks for the video!!. UA-cam started recommending you today and then I am here for 3 videos in a row (loved the one about book recommendations). Please keep making more videos!
Thank you for this concise video on what to actually learn and for what purposes. There’s an overwhelming wealth of info out there and discriminating ends up being really confusing
I guess for Data Science, I will say yes you need a core knowledge of Statistics and Maths with big data applications but for data analysts, you can manage with Little bit of Statistics knowledge and it might also depend on which position you are working or feel comfortable to work with. For data analysts some things can be considerable but when you say Data Scientist then it's a heavy word you need to know about everything from Data Engineering to Data Analytics to Data Architect. But again here comes the twist after realising Gen AI everything becomes easier and faster so not now but in the coming future it will mainly impact the role of Data Analysts and what options we have either transit or go for higher roles like Data Scientist/ AI.
Hi, firstly thank u so much to explain the basics, was so much confused like what to learn n how much to learn for DS, and after watching your video now its absolutely clear with the fundamentals too. Really very good n clear contents you upload regarding DATA SCIENCE. 😊👌
I will respectfully disagree. Yes, you do not need to know much about Statistics to find a job in data analysis/science. But I have seen some works that really made my hair stand. From Kaplan-Meier curves estimated from small data sets with lots of censoring to within-subject covariance being ignored. And this DOES impact decision-making in harmful ways, without the user of DS methods ever being warned. One should never, and I mean it literally, use a method without fully understanding the math behind it. Of course, the mistake's impact will differ from situation to situation, but really avoid it like the plague. For example, it is only by doing the mathematics behind the bootstrap methods, which are not really advanced but really boring and painful to follow, that one can notice that it cannot be used as is in time series context. A mistake that I did in my undergrad days and exposed a client to a larger financial risk than anticipated. Again, with no warnings.
Hello Thu Vu, I just discovered your channel three days ago, and it's quite strange how your channel has found its way to my list of favourite Data analytics channels. You really doing a great job! I love this channel and I look forward to savouring all your contents 😋. I really do hope you get more subscribers and more views. Cos, why not?!
One point I didn’t quite get the impression had enough backing was the near uselessness of the Central Limit Theorem. All A/B testing in the largest tech companies is done entirely with the help of the latter. Statistics is the cornerstone of inference. No inference - no decisions. At least that’s what I’ve encountered too many a time to blindly believe that the CLT is not the most fundamental thing ultimately underlying all of the decision-making.
HEY 1 year later i am enrolled in a university in germany in a BS Business Information systems. I am starting a class in statistics and landed AGAIN on this vid. its still too much information hahaha XD
I am sorry to comment on your video that meaning that you do not master statistics throughlly does not make you elligible to call them "obsolete" . Datascience, AI, Machine learning etc whether you like it or not are statistical works in progress. Every thing that is used to extract lnformation from data is the realm of statistics. Like any other field statistics is making use of technology. This does not make datascience, AI, machine learning etc out of the realm of statistics. So do not make people confussed by making baseless conclusions
I listed some nice resources in this video m.ua-cam.com/video/XAoZbtIsgU0/v-deo.html. Jbstatistics channel also has nice explanation videos for statistics as well. I’d recommend you to learn the most important and basic topics first and build it up over time, otherwise it would be a lot to take in. I tried to learn everything once and felt so overwhelmed and gave up 😅. So yeah.. just my experience
I shared some statistics resources in this video ua-cam.com/video/XAoZbtIsgU0/v-deo.html. Jbstatistics channel is also a great place to learn more statistics as well, his playlists are pretty comprehensive!
Dear please answer me is statistic for data science is the same as statistic for data analyst. I am a newbie trying to learn data analyst and need your reply I trust you
@@ashutoshnayak6242 Indeed descriptive statistics is what data analysts use most of the time. But if you also do predictive analysis and more complex modeling, then more advanced statistics will be required.
I think one important part of statistics a lot of people overlook is simply all of the assumptions being made behind the scenes. There are so many cases where you may use a particular test or model (even simple linear regression) where there are assumptions being made about sample size, normality, equal variance of residuals, etc., where you won’t get any warning or error in your code but the results may be untrustworthy because one or more assumptions are being violated. Always know what assumptions are being made when you implement certain methods, and to what degree these methods are robust to violations! It could mean the difference of being able to fit a desired model on the data directly, or needing to manipulate the data/methods drastically to be able to trust the results. Just because you see statistical significance does not mean something is significant, it means it is significant given all these other facts about the data hold true
Additional Tip: When studying a graph in addition to looking at the shape; always look at what is on the X and Y axis. For example, on a bell curve, what is on the Y-axis? Why?
Great tip, Jim! It's essential to look at the axes, indeed. Thanks for sharing! 🙌🏽
Great video as always! I am just going through all of your videos. I am doing lots of regression analysis as a postdoctoral researcher and hearing that hypothesis testing is not a big thing in DS makes me a little sad haha. I find this part to be fun and I thought it would be useful in case I want to leave academia. For us, it is all about the p-value and how substantial the results are.
This is really helpful. I learned from many sources and really got stuck for not knowing which I “should” know. Lots of intermediate and advanced concepts that I learned are not really applicable in my scope of work, then I almost forgot them all :)) It is the best summary video of statistics for data analysts I have ever watched :) Thank you a lot
That’s great to hear! Yes I can totally relate to that, so much information and I’d feel paralyzed to decide what to learn first. Thank you so much for your supporting comment. All the best 🍀
OMG i actually understand what you are saying. I have just started my DA journey..week 3. Past 2 weeks, i have been swimming in the sea of statistics, just trying not to drown ^^. Please post more videos on statistics, tips on data cleaning, critical thinking, data presentation and other related topics to DA and DS. TQ heaps!!
A very comprehensive video for statistics for data analysis...I myself been trying to get into the data analysis field but as you said there is a lot of misguidances....you know what you are doing..much appreciated
Hey Vishal, thank you for watching and commenting! Indeed I have been there and found it hard to navigate myself, there are too many resources and things to learn. But the most important thing is to get the first few solid basic building blocks, then you can expand from there. All the best 👋
@@Thuvu5Thank you for your reply...Given the experience you have in this field I will be really glad I can take a few pointers from you... only if there was a way to connect with you.
@@vishalrana9594 Hey Vishal, sure! It'd be great to connect. You can find my email address and the link to my Linkedin on the "About" page of my YT channel. Talk soon!
The most anticipated video. Thank you for this really awesome and useful video..👍
I have just started learning data science and I watching many videos related to it. UA-cam recommend one of your videos and now have watched 3-4 videos of yours. Reall good resource. Keep doing the good work.
Thanks for the video!!. UA-cam started recommending you today and then I am here for 3 videos in a row (loved the one about book recommendations). Please keep making more videos!
Aw that’s great to hear Lucas! Thank you so much 🤗
First time in my life I actualy understand what those statistical concepts actually mean. Your explanation is gold!
Aw that’s amazing! Thank you so much for this 🙌🙌
Thank you for this concise video on what to actually learn and for what purposes.
There’s an overwhelming wealth of info out there and discriminating ends up being really confusing
Wow! Such a great content video. Many thanks for the detailed explanation. Cheers
I guess for Data Science, I will say yes you need a core knowledge of Statistics and Maths with big data applications but for data analysts, you can manage with Little bit of Statistics knowledge and it might also depend on which position you are working or feel comfortable to work with.
For data analysts some things can be considerable but when you say Data Scientist then it's a heavy word you need to know about everything from Data Engineering to Data Analytics to Data Architect.
But again here comes the twist after realising Gen AI everything becomes easier and faster so not now but in the coming future it will mainly impact the role of Data Analysts and what options we have either transit or go for higher roles like Data Scientist/ AI.
This is a great video! Please post more statistics videos. Thanks for the great overview.
Thanks chị very much for very useful videos in Data science, these are helping me a lot to pursue this field.
Mam your giving a clear and very informative video in just 10mins, please post in detail with implementation and examples pls..
🤩Wow Thanks for this content. Waiting for more data analysis topic videos
Thank you so much ❤️. You kept it simple and taught very well.
Thank you Thu Vu.. as always
I think hypothsis testing is very important in statistic area, but most people don't know it, so you will not see them use.
Hi, firstly thank u so much to explain the basics, was so much confused like what to learn n how much to learn for DS, and after watching your video now its absolutely clear with the fundamentals too. Really very good n clear contents you upload regarding DATA SCIENCE. 😊👌
I will respectfully disagree. Yes, you do not need to know much about Statistics to find a job in data analysis/science. But I have seen some works that really made my hair stand. From Kaplan-Meier curves estimated from small data sets with lots of censoring to within-subject covariance being ignored. And this DOES impact decision-making in harmful ways, without the user of DS methods ever being warned. One should never, and I mean it literally, use a method without fully understanding the math behind it. Of course, the mistake's impact will differ from situation to situation, but really avoid it like the plague. For example, it is only by doing the mathematics behind the bootstrap methods, which are not really advanced but really boring and painful to follow, that one can notice that it cannot be used as is in time series context. A mistake that I did in my undergrad days and exposed a client to a larger financial risk than anticipated. Again, with no warnings.
Thank you with the helpful and relevant information
Hello Thu Vu,
I just discovered your channel three days ago, and it's quite strange how your channel has found its way to my list of favourite Data analytics channels. You really doing a great job!
I love this channel and I look forward to savouring all your contents 😋. I really do hope you get more subscribers and more views. Cos, why not?!
Yaaay! So glad you found my channel 🙌🤓 thank you for watching!
Thanks for this video! It is perfect, I learn a lot with you. You just gain a new follower :D
Hey Ana 👋 I’m so glad you liked the video! And thanks for subscribing 💛
Great video. You deserve more views
Aw thanks a lot! 😁
Bought you a cup of coffee this morning for sharing this video with me.
One point I didn’t quite get the impression had enough backing was the near uselessness of the Central Limit Theorem.
All A/B testing in the largest tech companies is done entirely with the help of the latter. Statistics is the cornerstone of inference. No inference - no decisions. At least that’s what I’ve encountered too many a time to blindly believe that the CLT is not the most fundamental thing ultimately underlying all of the decision-making.
wonderful video... you covered whole statistics...
Thank you :)
Very useful video, Thanks
thank you so much for this nice video
Really Nice list, could you also suggest some resources to study these topics (from an applied data science point of view).
Good 👍😊Work
oops! i just started google DA course and I somehow landed on this vid.. too much informatiooooooooon!!!!!!!!!! my head is spinning
HEY 1 year later i am enrolled in a university in germany in a BS Business Information systems. I am starting a class in statistics and landed AGAIN on this vid. its still too much information hahaha XD
I was really a nice and informative video ❤can you share some books name related to statistics and probability??😊
Thanks!
that is really helpful
I am sorry to comment on your video that meaning that you do not master statistics throughlly does not make you elligible to call them "obsolete" . Datascience, AI, Machine learning etc whether you like it or not are statistical works in progress. Every thing that is used to extract lnformation from data is the realm of statistics. Like any other field statistics is making use of technology. This does not make datascience, AI, machine learning etc out of the realm of statistics. So do not make people confussed by making baseless conclusions
Thank you! Is there an online course/article about practical stats knowledge that you recommend?
I listed some nice resources in this video m.ua-cam.com/video/XAoZbtIsgU0/v-deo.html. Jbstatistics channel also has nice explanation videos for statistics as well. I’d recommend you to learn the most important and basic topics first and build it up over time, otherwise it would be a lot to take in. I tried to learn everything once and felt so overwhelmed and gave up 😅. So yeah.. just my experience
Hello Thu Vu! Are there any great statistic books you recommand for an entry level data analyst? Thank You!!
hi mam, where these statistics methods can learn can you suggest some youtube resources to learn
Very useful...
Thanks
So it’s juast a fancy way of calling statistic
Any online statistics course for Data Analysis that you can recommend? Thanks
I shared some statistics resources in this video ua-cam.com/video/XAoZbtIsgU0/v-deo.html. Jbstatistics channel is also a great place to learn more statistics as well, his playlists are pretty comprehensive!
Dear please answer me is statistic for data science is the same as statistic for data analyst. I am a newbie trying to learn data analyst and need your reply I trust you
where is part 2 ?
Very nice vidéo
Great to hear! I appreciated it
Very Nice explanation. Can I get your linkedin profile if you don't mind?
Ay güey mi mente
Do a data analyst need to know statistics or not! Please reply.
Of course!!
@@Thuvu5 i think more descriptive statistics is required 🤔 for the data analyst jobs?
@@ashutoshnayak6242 Indeed descriptive statistics is what data analysts use most of the time. But if you also do predictive analysis and more complex modeling, then more advanced statistics will be required.
@@Thuvu5 thank you for your response 🙂🙂
All of this stuff is high school level tho...
Bạn người Việt à? có nói được tiếng Việt không?
Is it me or you have an unnaturally cute voice?
This is my normal voice 😂🙈. Thanks, I’ll take it as a compliment haha
@@Thuvu5 Then your normal voice is the sweetest one I have ever heard! Also, you are doing a great job helping people like me. Thanks a lot.
@@arifulislamkhan3254 Aw thank you so much! This is the first time someone ever said that to me ☺️. Glad to hear you found my videos useful!
@@Thuvu5 My pleasure! Hope to see you as one of the top UA-camrs in this niche 🙂.
@@arifulislamkhan3254 Aw thank you so much Ariful!! I will try my best 😅
Cute