- Видео 21
- Просмотров 106 151
Shonda Kuiper
Добавлен 21 июн 2013
Using Games to Train Students how to Make Decisions with Data
This video is focused towards statistic educators. However, these materials can be easily modified for any students interested in understanding common challenges when conducting research in any discipline. More details are available at: stat2labs.sites.grinnell.edu/
Просмотров: 95
Видео
Racer Data: How cleaning data can influence p-values
Просмотров 4614 года назад
This video provides an example of how how different researchers, using the exactly the same data, may come to very different conclusions. This data can be easily collected through an online race car game. More information, including instructor resources are available at stat2labs.sites.grinnell.edu/racer.html.
Play the Racer Game
Просмотров 794 года назад
This video describes how the free online Racer Game can be used to teach students how make better decisions with data. Players can race a variety of race cars on multiple tracks. Each car body, engine and set of tires impacts the speed and acceleration of the cars. Players can choose any options they like, but they will develop better strategies (i.e. improve their chances of winning) if they t...
Play the Greenhouse game.
Просмотров 2964 года назад
This is one of multiple activities on our website, stat2games.sites.grinnell.edu/, where data visualizations and statistical modeling can help you win the game.
C7b_Inference for Logistic Regression
Просмотров 2664 года назад
This video addresses several core ideas that should be considered before drawing conclusions from a logistic regression model.
Introduction to the bootstrap method
Просмотров 4,6 тыс.4 года назад
This video explains the reasoning behind the bootstrap method and then provides visualizations to demonstrate a few examples.
John Oliver P-Hacking
Просмотров 21 тыс.5 лет назад
This provides several clips from one of my favorite discussions on p-hacking ruclips.net/video/0Rnq1NpHdmw/видео.html
C3a: Introduction to Multiple Regression: t-test and R-sq
Просмотров 2 тыс.9 лет назад
This video provides an example where the t-test provides evidence that Mileage is important in predicting Price of a used car. However the R-squared value indicates that Mileage alone is not sufficient in getting an accurate estimate of price. In future presentations we build on this example to construct a multiple regression. It was created to supplement the material within the Practicing Stat...
C2: Normal Probability Plots
Просмотров 1,3 тыс.9 лет назад
This video demonstrates the basics of creating and interpreting a Normal Probability Plot. It was created to supplement the material within the Practicing Statistics Textbook: web.grinnell.edu/individuals/kuipers/stat2labs/Chapter2.html
C2b: Comparing the two-sample t-test, ANOVA, and regression
Просмотров 3,9 тыс.9 лет назад
This video compares the relationship between the underlying statistical models used in the t-test, ANOVA, and regression. This is the second video to supplement the Chapter 2 material within the Practicing Statistics Textbook: web.grinnell.edu/individuals/kuipers/stat2labs/Chapter2.html
C2a: Building Statistical Models
Просмотров 1,4 тыс.9 лет назад
This video introduces how to build a statistical model using a very simplistic example. In future presentations, we will build on this example to construct more complex statistical models used in regression and ANOVA. It was created to supplement the material within the Practicing Statistics Textbook: web.grinnell.edu/individuals/kuipers/stat2labs/Chapter2.html
C1 How to Use Minitab to Conduct a Randomization Test
Просмотров 6 тыс.10 лет назад
We demonstrate how to use Minitab to conduct a randomization test. The data and code are freely available online. This example is based upon our previous video: Introduction to Randomization Tests (Schistosomiasis Example) which is used in the Practicing Statistics Textbook: web.grinnell.edu/individuals/kuipers/stat2labs/Chapter1.html
C1 R: Using R to Conduct a Randomization Test
Просмотров 17 тыс.10 лет назад
We demonstrate how to use R to conduct a randomization test. The data and code are freely available online. This example is based upon our previous video: Introduction to Randomization Tests (Schistosomiasis Example) which is used in the Practicing Statistics Textbook: web.grinnell.edu/individuals/kuipers/stat2labs/Chapter1.html
C1b: Introduction to Randomization Tests
Просмотров 9 тыс.10 лет назад
C1b: Introduction to Randomization Tests
Comparison between CART Analysis and Logistic Regression Models
Просмотров 7 тыс.10 лет назад
Comparison between CART Analysis and Logistic Regression Models
How will I purchase this software for scrt?
R software is completely free! To learn how to install use this video: ruclips.net/video/eD07NznguA4/видео.html. After you have installed R, here are several of our tutorials on how to get started with the software: www.stat2labs.sites.grinnell.edu/RTutorials.html
There is a slight confusion. In 4:45 you said there are 10,000 different cases. Since you are drawing a random sample each time there is no guarantee that all the 10,000 different cases are different.
10,000 different cases (or simulations) were run. However, we expect several of the simulations to have the same results. I am not claiming that there are 10,000 different results. In fact, it is easy to prove mathematically several results will be the same. The histogram, at time 5:02, plots the results and shows that we observed many results close to 0.
See: Peter Doshi
Santa Clase 0 seconds ago .369 Reply Santa Clase Santa Clase 3 minutes ago sharing🥰🥰🥰😇😇😇Thanks for sharing
.369
sharing🥰🥰🥰😇😇😇Thanks for sharing
I always thought there was a standard deviation that could be used to determine the probability of results for more groups of the same types of items.
Yes, the variability of the residuals is key to calculating probabilities. We focus on comparing the variability within groups to the variability between groups. I don't address this directly, but you can see more in the next video: studio.ruclips.net/user/videojQkK0XMrAdM/edit.
Great content!
Best explanation of Re Randomization of many videos in RUclips. very concise and to the point. Thanks
funny how quickly this stooge forgot all about this during Common Cold-19. researchers aren't the only ones influenced by 'funding'.
Thank you Madam. Keep posting.
99% of the people in the world have no clue what is true information and how to investigate if it is true or false. 99% of the almost 8 billion on this planet don't actually know what is a scientific result and what makes it different than a personal opinion or a religion view or any source of information. That is why it is relatively easy to scam them with propaganda and religion and cults!
Thank you so much for this! I love this episode, but hesitated showing it to my research methods students until I found your compilation of clips.
P-hacking is being used and abused for racial justice
Lol skonda
Thanx for the clear explanation!
that was very helpful.
Thank you so much! This is just what I need for the assignment!
Where can I download this book?
The book is not free, but there are many free resources here: stat2labs.sites.grinnell.edu/Textbook.html The book can be purchased here: www.pearson.com/us/higher-education/program/Kuiper-Practicing-Statistics-Guided-Investigations-for-the-Second-Course/PGM116126.html
Very clear!! thanks a lot
Very nice, so brief and eye-opening lesson.
Thanks. Great Explanation. Stay blessed.
sharing🥰🥰🥰😇😇😇Thanks for sharing
Great Thanks
Excellent video. Great thanks
How we will perform randomization with three groups? when we are doing it for ANOVA!
There are several possible approaches when there are three or more groups. I would take the following steps: 1) using the observed sample data, calculate the test statistic by subtracting the largest mean from the smallest mean, 2) combine all the data into one vector {x <- c(data1, data2, data3)}, 3) conduct 10,000 iterations where data is randomly assigned into three groups, and 4) in each iteration take the max(mean1, mean2, mean3) minus the min(mean1, mean2, mean3). Finally the p-value would represent the number of times (out of 10,000), that our iterations provided a value as least as large as our observed difference (i.e. at least as large as our test statistic).
Thank you for your nice explanation. Maybe you have time to answer me that why you, only you paid attention to R square, level and not to significance of the parameters in your example? I would like to know that percentage data are quantitative or qualitative variable?
Thank you very helpful with effective explanation.
Thanks a lot! Can you please tell me where did the 7.6 in "p.value <- sum(results >= 7.6)/reps" come from?
In the observed sample data (provided on page 2 of the textbook and on the very first screen of this video) you see that the average for the control was 12 and the average for the treatment was 4.4, The p-value in a randomization test is use to calculate the probability of observing a difference as large as the one observed in our sample (12 - 4.4 = 7.6), assuming that the null hypothesis is true.
Dr. Kuiper, thank you for the videos, they are very helpful especially in assisting my online students. Will you be adding any more videos for the remainder of your textbook (chapters 9 and 11 would be great)? :)
Very good explanation of the concept.
Thank you. This video has really simplified the concept.