Hi, so when looking at the ANOVA table, how can we know if Males are 1 or 0 ( in a case where we didn't write the code and just has the table....so for example, is the unlisted value=0?)
There is no way to tell, but in your example, men compared to women, the difference in means is identical except for the sign of the difference. It is even harder to tell if you have an independent variable with more categories, e.g. married, widowed, divorced, separated, and never married. But, if you estimate the marginal effects you can determine all of the raw effect sizes.
richard mawulawoe ahadzie Richard, I am not exactly certain what you want, but maybe the following will help. regress y x predict yhatin if e(sample), xb predict yhatout if !e(sample) The command -predict- is a model post-estimation command that can be used to create predicted values. You can learn more by typing -help predict- in the command window. When you estimate a model Stata creates a temporary marker of the sample used in the model called -e(sample)-. This marker is equal to 0 is an observation was not used in the model and 1 if it was. Finally, the exclamation mark "!" means "not" (see -help operator-). So, the first predict creates in-sample predicted values stored in the variable "yhatin" and the second creates out-sample predicted values stored in the variable "yhatout". Best, Alan
Dear Alan I have this regression and not sure about what I have done and you are an expert, please have a glance: Accruals it = α1 + α2((ΔREVit - ΔRECit )/Ait-1) + α3(PPEit/Ait-1) + α4(CFOit/Ait-1) + α5NEG_CFOit + α6((NEG_CFOit * CFOit)/Ait-1) + εit all variables are numeric except the last two which are: ΔCFOi,tNEG is an indicator variable taking the value of 1 if the change in cash flows from operations is negative and 0 otherwise. CFOi,t*CFONEGi,t is an interaction term which is defined as ΔCFOi,t multiplied by ∆CFONEGi,t. I did the following for the last two; 1: gen NEG_CFO =. replace NEG_CFO=1 if delta_CFO0 2: gen interaction2_CFO = NEG_CFO*(CFO/lagged_assets) and I run the regression Is that correct? Please your response is highly appreciated
If you are trying to run a regression with an interaction term, it is better in Stata to use factor notation. For example: sysuse auto, clear regress mpg c.weight c.weight#c.weight margins, at(weight=(1800(200)4800)) marginsplot, recast(line) recastci(rarea) see "help fvvarlist" for more information.
Thank you, Dr Alan.
May you be peaceful and successful throughout your life.
The way you explain things is very clear. This video is helpful for undergraduate students like me. Thank you for your video!
Than you!
You are making my life easy. That is very helpful. Thank you.
+Mihiretu Molla Thank you!
Very nice and detailed descriptions, speech tempo is easy to catch and the video framing is also wonderful. Thank you so much!!!
Thanks for the help! Great overview for multiple regression within stata
Thank you! I'm glad you found the video useful.
Thank you very much! This helped me a lot in my survey :) Greetings from Greece!
I'm glad the video helped.
Thank you, your i.sex explenation helped me a lot.
Great videos. Thank you for sharing with all.
thanl you really coz these videos are helping good
Thank you so much. Its too much helpful.
hi Mr Alan. please can you chow how to estimate the fixed-effect panel threshold model(Hansan 1999) ??
hey....am having a problem in keying in the commands for QUAIDS model in stata. Kindly assist
Impressive lecture than you very much learned something
Rajesh Shigdel I'm glad you found it useful!
Hi, so when looking at the ANOVA table, how can we know if Males are 1 or 0 ( in a case where we didn't write the code and just has the table....so for example, is the unlisted value=0?)
There is no way to tell, but in your example, men compared to women, the difference in means is identical except for the sign of the difference. It is even harder to tell if you have an independent variable with more categories, e.g. married, widowed, divorced, separated, and never married. But, if you estimate the marginal effects you can determine all of the raw effect sizes.
Great, thanks!
what is the difference between ## ad # command in stata?
This is so helpful!! thank you!
Thank you. I'm glad you found this useful.
How do I make In-sample and out-of-sample predictions in STATA
richard mawulawoe ahadzie
Richard, I am not exactly certain what you want, but maybe the following will help.
regress y x
predict yhatin if e(sample), xb
predict yhatout if !e(sample)
The command -predict- is a model post-estimation command that can be used to create predicted values. You can learn more by typing -help predict- in the command window.
When you estimate a model Stata creates a temporary marker of the sample used in the model called -e(sample)-. This marker is equal to 0 is an observation was not used in the model and 1 if it was.
Finally, the exclamation mark "!" means "not" (see -help operator-).
So, the first predict creates in-sample predicted values stored in the variable "yhatin" and the second creates out-sample predicted values stored in the variable "yhatout".
Best,
Alan
thanks Alan, will try and get back to you
What is that intro song?
The White Collar Holler by Stan Rogers, a fabulous Canadian folk singer who unfortunately is no longer alive.
Thank you very much for this video
I'm glad you found it helpful.
Dear Alan
I have this regression and not sure about what I have done
and you are an expert, please have a glance:
Accruals it = α1 + α2((ΔREVit - ΔRECit )/Ait-1) + α3(PPEit/Ait-1) + α4(CFOit/Ait-1) + α5NEG_CFOit + α6((NEG_CFOit * CFOit)/Ait-1) + εit
all variables are numeric except the last two which are:
ΔCFOi,tNEG is an indicator variable taking the value of 1 if the change in cash flows from operations is negative and 0 otherwise. CFOi,t*CFONEGi,t is an interaction term which is defined as ΔCFOi,t multiplied by ∆CFONEGi,t.
I did the following for the last two;
1:
gen NEG_CFO =.
replace NEG_CFO=1 if delta_CFO0
2:
gen interaction2_CFO = NEG_CFO*(CFO/lagged_assets)
and I run the regression
Is that correct?
Please your response is highly appreciated
No response?
If you are trying to run a regression with an interaction term, it is better in Stata to use factor notation. For example:
sysuse auto, clear
regress mpg c.weight c.weight#c.weight
margins, at(weight=(1800(200)4800))
marginsplot, recast(line) recastci(rarea)
see "help fvvarlist" for more information.
Alan Neustadtl Thank you very much for your help.
Thank you!! :)
I'm glad the video was useful!
8:40 dummy