Sorry for the late reply. You're right! My mistake. I continue the mistake at 8:43 but the following plot does have the right variables. The formula is corrected when the results are put in at 10:35. Sadly, I cannot edit the video to fix it (that I know of) without removing this one & reuploading.
By 0 and 1 I assume you are referring to the 'am' variable? If I recall that data correctly, it was already coded 0 and 1; however, it did not have to be. Dichotomous variables can be coded as any 2 values, e.g., 34 and 89, and the math will compensate in the resulting coefficients to produce the same end results. Having the binary 0 and 1 is just far more intuitive and easier to interpret.
i've seen all the videos on interaction terms and this is the only one that has helped me understand!! King!!
Your videos are the best!! They are way easier to understand than every other source online, and my stats professors.
Thanks! Glad you're enjoying them!
Useful explanataion, Thanks Sir
I think 7:26 should say disp instead of hp in params --> "b0 + b1(disp) + b2(am) + b3(disp*am)"
Sorry for the late reply. You're right! My mistake. I continue the mistake at 8:43 but the following plot does have the right variables. The formula is corrected when the results are put in at 10:35. Sadly, I cannot edit the video to fix it (that I know of) without removing this one & reuploading.
this video is amazing
Does it matter what you set to 0 vs. 1. I don't think it does, but I'm not sure why.
By 0 and 1 I assume you are referring to the 'am' variable? If I recall that data correctly, it was already coded 0 and 1; however, it did not have to be. Dichotomous variables can be coded as any 2 values, e.g., 34 and 89, and the math will compensate in the resulting coefficients to produce the same end results. Having the binary 0 and 1 is just far more intuitive and easier to interpret.
@@ShawnJanzen Thanks! That's exactly what I thought would happen. This was a really great video!
@@mrschaucerssquire thanks! Glad you found it useful.