Fwd: Re: Re: Fwd: Re: Re: Fwd: Regarding help for multiple regression

27 views
Skip to first unread message

bansalsuman198918

unread,
Jan 6, 2015, 3:43:15 AM1/6/15
to dataanalys...@googlegroups.com
Dear Sonia 
I obtained my data from various sources such as annual report of companies , ownership and financial data and corporate governance reports are obtained from CMIE Prowess data base , some data obtained from money control .com, and Quant partners. Com and some historical data obtained from bseindia.com also.
Regards
suman


Sent from Samsung Mobile


-------- Original message --------
From: Sonia Jindal
Date:06/01/2015 13:31 (GMT+05:30)
Subject: Re: Re: Fwd: Re: Re: Fwd: Regarding help for multiple regression

Dear Suman madam can u tell me what is the sources of ur data 

On Mon, Jan 5, 2015 at 11:47 PM, bansalsuman198918 <bansalsu...@gmail.com> wrote:

Respected sir
Good After Noon
Sir if we have metric DV and and two or more independent variables and among all independent variable one is categorical and all others are metric . And we want to control for the effect of some variables which are metric than which tool should I apply.
My dependent variables is RoE
Independent variables : promoters ownership ,  board size ,  proportion of independent directors these are metric independent variable . 
Position of chairman is a categorical independent variable .
Firm size , debt equity ratio , firm age , growth of sales are control variables of my study . 
Regards
Suman

Sent from Samsung Mobile


-------- Original message --------
From: Neeraj Kaushik
Date:06/01/2015 10:13 (GMT+05:30)
To: dataanalysistraining
Subject: Re: Fwd: Re: Re: Fwd: Regarding help for multiple regression

One addition here in 2 way ANOVA example:

If we wish to check the effect of gender on Sales then we'll use 2 indep sample t-test
If we wish to check the effect of exp on Sales then we'll use 1 way ANOVA

However if we wish to check the combined effect of Gender & Exp on Sales then we will use 2 way ANOVA
Here Null hypothesis will be that there are no significant differences among the sales of 6 groups:
Men with low exp
Women with low exp
Men with medium exp
Women with medium exp
Men with long exp
Women with long exp

Plz see we're looking at combined effect of Gender & Exp on Sales here, hence we'll use 2 way ANOVA

For the question of How to work on 2 way ANOVA, plz search for the post "Beyond ANOVA" on group

Best wishes
Neeraj




On Tue, Jan 6, 2015 at 10:03 AM, Neeraj Kaushik <kaushi...@gmail.com> wrote:
Dear Suman

Seema is correct in her advice. You shd use Multiple Regression.
I wish to clear there the concept of ANOCA (Analysis of Co-variance).

See when we've metric DV and one or more metric IDV then we shd use Regression (Simple or Multiple)
Ex: We wish to check whether Sales depends on Advt and No. of salesmen
Here Sales is metric DV
Advt is mtric IDV No of salesmen is metric IDV
So we'll use Multiple Regression here
----------------------------------------
However if there is one metric DV and one or more categorical IDVs then we use n-way ANOVA (n= No. of categorical IDV )
Ex: We wish to check whether Sales is related with Experience
Here Sales is metric DV
Experience is categorical IDV which >2 categories (Low exp, medium exp and Long exp)
so we shd 1- way ANOVA as there is 1 IDV

If we wish to check whether Sales is related with Experience and Gender
Here Sales is metric DV
Experience is categorical IDV
Gender is categorical IDV
So we shd use 2-way ANOVA here as there are 2 non-metric (categorical) IDV
----------------------------------------
Now if there are one metric IDV and rest all IDVs are categorical then we use a version of ANOVA which is called as n-way ANCOVA (n=No of categorical IDVs)
This metric IDV is called as Co-variate and is treated like a Control variable.

Ex: Whether Sales is related with Experience and Advt
Here Sales is metric DV
Experience is categorical IDV which >2 categories (Low exp, medium exp and Long exp)
Advt is metric IDV

Now there are 2 ways to analyze it:
If emphasis is on Advt and we wish to see this relation for the avrious exp categories then we shd use Regression with dummy vars

However if the emphasis is on the Exp and we wish to take Advt as control avr then we shd use 1-way ANOVA
where Advt is be called as Co-variate

Plz feel free to ask for ur doubts now.

Best wishes
Neeraj



On Mon, Jan 5, 2015 at 3:53 PM, Seema Malik <seema.m...@gmail.com> wrote:

Your dep and independent variables are metric in nature. So go for multiple regression.

On 5 Jan 2015 15:03, "bansalsuman198918" <bansalsu...@gmail.com> wrote:
Respected sir 
My dependent variable is Roe
Independent variables are promoter ownerships , FII ownership , banks& financial institutions ownership , public ownership , Mutual funds ownership 
And control variables are 
Growth of sales 
Age of companies
Size of companies 
Debt equity ratio of companies
Sir kindly help me which tool should I apply.
Regards
Suman


Sent from Samsung Mobile


-------- Original message --------
From: Neeraj Kaushik
Date:05/01/2015 14:47 (GMT+05:30)
To: dataanalysistraining
Subject: Re: Re: Fwd: Regarding help for multiple regression

Dear Suman
Plz aware what exactly u wish t do here?
What are ur DV & IDVs?
Best wishes
Neeraj

On Mon, Jan 5, 2015 at 1:19 PM, bansalsuman198918 <bansalsu...@gmail.com> wrote:

Good after noon Everybody
Kindly help me with regards to my data I am attaching please help me which tool should I apply Multiple Regression or ANCOVA  . Kindly help me or also help me how to apply this particular tool in SPSS
Regards
Suman
Department of Commerce
MD university 
ROHTAK
Sent from Samsung Mobile


-------- Original message --------
From: suman bansal
Date:02/01/2015 19:41 (GMT+05:30)
Subject: Re: Fwd: Regarding help for multiple regression

Thank you  ma"am for replying 
Here i am sending my data . kindly suggest me how to apply multiple regression. 
Regards
Suman

On Fri, Jan 2, 2015 at 5:09 AM, Seema Malik <seema.m...@gmail.com> wrote:

First of all please tell about your data. If data is panel and only one dependent variable is there with many independent variables , multiple regression will be suitable.

On 2 Jan 2015 17:59, "bansalsuman198918" <bansalsu...@gmail.com> wrote:



Sent from Samsung Mobile


-------- Original message --------
From: bansalsuman198918
Date:02/01/2015 17:55 (GMT+05:30)
Subject: Regarding help for multiple regression

Respected sir

Good evening
Sir I am working on a research paper and my variables of interest are as follows
Dependent variable = RoE
Independent variables = promoter's ownership , FII ownership,  banking and financial services institution ownership, public ownership , mutual fund ownership
Along with these variables there are certain control variables also such as growth rate of sales
Age of firm
Debt: equity ratio
Size of firm measured in total assets of companies 
Sir I want to know which test I have to apply  mutilate regression or ANCOVA and how it is applied. 
Regards 
Suman
Department of commerce
MR University 
Rohtak
ent from Samsung Mobile 

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.


--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistrai...@googlegroups.com.
To post to this group, send email to dataanalys...@googlegroups.com.
Visit this group at http://groups.google.com/group/dataanalysistraining.
For more options, visit https://groups.google.com/d/optout.

Sonia Jindal

unread,
Jan 6, 2015, 3:54:22 AM1/6/15
to dataanalys...@googlegroups.com
thank you

bansalsuman198918

unread,
Jan 6, 2015, 7:13:40 AM1/6/15
to dataanalys...@googlegroups.com
Respected sir
Thank you vry much for replying . Sir I want to ask one more question is hirercheal regression is different from step wise regression ?  If yes than please tell me how to apply it in SPSS .
Regards 
Suman


Sent from Samsung Mobile


-------- Original message --------
From: Neeraj Kaushik
Date:06/01/2015 15:11 (GMT+05:30)
To: dataanalysistraining
Subject: Re: Re: Fwd: Re: Re: Fwd: Regarding help for multiple regression

Dear Suman
If you get only 1 categorical IDV then u can use Dummy for that.
Moreover if u wish to take some var(s) as control then u shd use Hierarchical Regression (can be done in SPSS).
Here we take all var(s) in model 1-by-1. So its different from ENTER or Stepwise method.
Best wishes
Neeraj
 

Radha garg

unread,
Jan 6, 2015, 9:36:27 AM1/6/15
to dataanalys...@googlegroups.com
Dear Suman,

There is no harm in asking the question. But, it is always better to first do some research at our end and then ask for query in any forum. Because, we all are the researchers here and we should do some searches for our own query at-least. It also help us to understand the concept in more detail. If you just search your query in google, you will get the answer in first 2-3 links. 

Anyway, as regard to your query, The answer is Yes. Hierarchical regression is different from step wise regression. 

A stepwise variable selection procedure in which variables are sequentially entered into the model. The first variable considered for entry into the equation is the one with the largest positive or negative correlation with the dependent variable. This variable is entered into the equation only if it satisfies the criterion for entry. If the first variable is entered, the independent variable not in the equation that has the largest partial correlation is considered next. The procedure stops when there are no variables that meet the entry criterion.

Whereas, in hierarchical regression we decide which variable to enter at what stage, basing your decision on purpose and logic of the research.

The concept of hierarchical regression is already very well explained by Neeraj sir. Kindly refer to this link mentioned below:-


Best Regards
Radha Garg
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.
To post to this group, send email to dataanalysistraining@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.
To post to this group, send email to dataanalysistraining@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.
To post to this group, send email to dataanalysistraining@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.
To post to this group, send email to dataanalysistraining@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.
To post to this group, send email to dataanalysistraining@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.
To post to this group, send email to dataanalysistraining@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.
To post to this group, send email to dataanalysistraining@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.
To post to this group, send email to dataanalysistraining@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "DataAnalysis" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataanalysistraining+unsub...@googlegroups.com.
To post to this group, send email to dataanalysistraining@googlegroups.com.

Neeraj Kaushik

unread,
Jan 6, 2015, 9:38:02 AM1/6/15
to dataanalysistraining
Reply all
Reply to author
Forward
0 new messages