Thank you! This video was very understandable. I especially liked the practical examples towards the end so I could see if I would calculate it right with you.
there is a mistake at around the 3.58 minute of the video. For log lin, a 1 UNIT (not percent) change in X brings betahat * 100 percent change in Y. The practical example interpretation you gave later in the video however is correct. Also there is a mistake in lin log. A 1 percent change in x brings a betahat/100 UNIT (not percent) change in Y. Again, at the beginning of video it there is the mistake, but at the end of the video your interpretation is right.
To my knowledge, It cannot be done. In logit, the dependent variable is binary, 1 or 0. It is not possible to take the log of '0'. Hence, the log-linear model in Logit may not be viable. Sorry for the late reply.
Many thanks for the great explanation! I have a question. How to intrepretate coefficients in log-log and log-linear models when both variables are transformed in first differences?
The Linear-Log model is wrong because the change in Y is always in units and not percent. This means that a 1 percent change in X brings a change in Y by β/100 units of measurement and NOT percent.
![Summer Walker - Heart Of A Woman [Official Lyric Video]](http://i.ytimg.com/vi/SFddctbZzw8/mqdefault.jpg)
Very good explanation, I finally got to understand this. Thanks!
Thank you! This video was very understandable. I especially liked the practical examples towards the end so I could see if I would calculate it right with you.
I will try to find the particular file and share it with you
there is a mistake at around the 3.58 minute of the video. For log lin, a 1 UNIT (not percent) change in X brings betahat * 100 percent change in Y. The practical example interpretation you gave later in the video however is correct.
Also there is a mistake in lin log. A 1 percent change in x brings a betahat/100 UNIT (not percent) change in Y. Again, at the beginning of video it there is the mistake, but at the end of the video your interpretation is right.
Thanks sir. I was confused
very clear explanation. it helped a lot, thank you
How do I do a log linear model (LOGIT)? I cannot find any information about it on youtube.
To my knowledge, It cannot be done. In logit, the dependent variable is binary, 1 or 0. It is not possible to take the log of '0'. Hence, the log-linear model in Logit may not be viable. Sorry for the late reply.
Many thanks for the great explanation! I have a question. How to intrepretate coefficients in log-log and log-linear models when both variables are transformed in first differences?
No difference in interpretation.
The Linear-Log model is wrong because the change in Y is always in units and not percent. This means that a 1 percent change in X brings a change in Y by β/100 units of measurement and NOT percent.
Thank you
hey is there any chance you can help get a better undertstanding of these types of problems
thx for the video sir, but in our class we have b0 + b1x. So alpha in your video is b0?
b0 is intercept which is generally denoted by aplha in classroom. But there is no hard and fast rule.