Simple Linear Regression - t tests of parameters
Вставка
- Опубліковано 6 вер 2024
- In this video I discuss the distribution of the estimators for the slope, b1, and intercept, b0, and provide an explanation of the t-test for these simple linear regression parameters.
Below is the link to the data used in this video: drive.google.c...
When a teacher is able to communicate complex concepts that clearly, it means they have really assimilated and metabolised the nitty gritty of it. Thanks for the efforts in making such great videos!
Thanks so much for this comment! It means a lot to me that you found my content helpful!
@Stats4everyone please specify when do we use t test for Beta, and when to use ANOVA for beta.
Great question. Thank you for the comment. Here is a video I found online that does a good job discussing this: ua-cam.com/video/TFAVu2FQki8/v-deo.html. A t test can be used to test if beta equals a particular number ( for example in this video I tested if beta is not 5), whereas an ANOVA will always test if beta is not zero. For simple linear regression (only 1 covariate, x), when we want to test if beta1(the slope) is not zero, ANOVA and a t test will give exactly the same p-value conclusion. If we have more than 1 covariate (more than 1 x), then an ANOVA will test if at least 1 of the betas is not zero, whereas a t test will be useful to test if beta1 is not zero, and another t test for if beta2 is not zero, etc. Please let me know if you have any follow-up questions about this :)
I love your work.
you are my saviour 😭