I want know the regression to 1 independent and multiple dependents variables with single test like correlation test, if they are any methods pls suggest me
Hi Munji,
Your problem falls in 3rd category, so it is multivariate regression.
Analysis in SPSS:
Multivariate regression is done in SPSS using the GLM-multivariate.option. Here is a nice tutorial on the same:
https://statistics.laerd.com/spss-tutorials/one-way-manova-using-spss-statistics.php#procedure
IBM support doc can also be useful: http://www-01.ibm.com/support/docview.wss?uid=swg21476743
Analysis in R:
If you prefer using R, the this link may be useful:
http://www.psych.yorku.ca/lab/psy6140/lectures/MultivariateRegression2x2.pdf
R package mvtnorm can be used for this purpose.
With usual regression you cannot do this as you need one variable on the left side of your equation. There are some data reduction techniques which reduce the dependent variables to one quantity and calculate the correlation between the redcued quantities e.g. canonical correlation and partal least suqares regression. If your model is more complex you could also consider simultaneoues euqation modeling.
Best regards,
Florian
If you have multiple dependent variables you can use multivariate regression (multiple regression is multiple predictors)
see my posting repeated here
I am currently writing a chapter on multilevel multivariate models and here is an extract....
"
Introduction
his chapter is concerned with the analysis of data when there is more than one response which we want to analyse simultaneously. To be clear, this is multivariate regression and not multiple regression. This approach was originated by Goldstein (1989) in the context of the analysis of development as growth curves.[1] Multiple response variables represent repeated measurements of distinctive but not unrelated outcome variables. For example the responses could be four health related behaviours - drinking, smoking exercise and diet - measured on individuals who are nested in communities and our analysis could be used to estimate how correlated are the behaviours at the individual level and community level (taking account of other characteristics). The multivariate model could also be used to answer question such as whether there are unhealthy communities as well as unhealthy individuals. Other straightforward examples where the multivariate approach has been used include the simultaneous analysis of diastolic and systolic blood pressure[2] and the analysis of marks derived form coursework and an unseen examination for students nested in a school.[3]
The multivariate model is the underlying building block of more complex models. Examples include Steele et al’s (2007) simultaneous equation modelling of the two- way relationship between school educational resources and pupil attainment; and Steele’s et al’s (2006) multilevel factor analysis of women’s status in Bangladesh which considered two dimensions of social independence and decision-making power, and how the factor structure differed at the woman and district level.[4] The Gallery at the Centre for Multilevel Modelling can be used to find more such studies.[5]
This chapter is concerned with the introduction of the basic model although we will also sketch the potential for more complex variables at the end. Substantively in the present chapter we will use data from the European Social Survey when there are two continuous responses: one is respondents’ answer to the question “what is the start of old age”; the other is “when does youth end”. We will use a multilevel version of the multivariate model in which the two responses are seen as nested within an individual who are nested with the 29 countries of the ESS.
..............................
The chapter concludes with an overview of how the model can be extended; this is particularly important as this model is fundamental to a surprisingly large number of other models including the multinomial model for multiple discrete outcomes, multilevel factor analysis, a very flexible model for analysing repeated measures, and as the basis for multiple imputation of missing values in a valid manner.
[1] Goldstein, H. (1989) Models for multilevel response variables with an application to growth curves in Bock, R.D. (ed.) Multilevel analysis of educational data, New York, Academic Press, although the ideas are also present in Goldstein H. (1986).Multilevel mixed linear model analysis using Iterative Generalised Least Squares, Biometrika 73 (1) 43-56.
[2] Duncan, C, Jones, K and Moon, G (1995) Blood pressure, age and gender, in Woodhouse, G (ed). A guide to MLn for new users, Institute of Education, London.
[3] Rasbash, J., Steele, F., Browne, W.J. and Goldstein, H. (2009) A User’s Guide to MLwiN, v2.10. Centre for Multilevel Modelling, University of Bristol, Chapter 14.
[4] Steele, F., Vignoles, A. and Jenkins, A., (2007) The effect of school resources on student attainment: a multilevel simultaneous equation modelling approach Journal of the Royal Statistical Society, Series A, 170(3), 801-824, 2007 ; iSteele, F and Goldstein, H (2006) A multilevel factor model for mixed binary and ordinal indicators of women’s status, Sociological Methods & Research 35: 137-153.
[5] http://www.bristol.ac.uk/cmm/research/gallery/"
end of extract.
Three further examples
1 simultaneous modeling of logits of happiness, wellbeing and health in different waves of the European Survey in this paper
https://www.researchgate.net/publication/270103441_Social_Policy_and_Economic_Development_in_Europe_Investigating_the_Macro_Determinants_of_Individual_Health_and_Well-being_in_a_Multilevel_Multivariate_Analysis_of_Thirty_European_Nations
2 Simultaneous modeling of quality (do you smoke?) and quantity (how many smoke?); the MLwin Software and especially the Stat-JR software can handle responses with different outcomes measured on different scales eg one response is a logit or probit; the other is continuous ; see Craig Duncan (1997) Multivariate mixed model PLACES AND PEOPLE: MULTI LEVEL MODELLING IN GEOGRAPHICAL RESEARCH. edited by .P. Westert and R.N. Verhoeff. KNAG/Netherlands Geographical Studies. Utrecht,
3 matrix sampling designs - where you have asked core questions of everybody but you ask detailed questions of random subsets of people but you model all the outcomes simultaneously to 'borrow strength' - this can be very useful and is used in the PISA study of educational attainment see https://eo3.commpartners.com/users/aera/session.php?id=10651
The key point about the random effects multilevel version of the multivariate model is that you can having missing values so that not everybody has to be measured on all outcomes. This is based on the relatively forgiving MAR - missing at random- assumption.
Taken from
Can you share any examples of the use of multivariate regression? - ResearchGate. Available from: https://www.researchgate.net/post/Can_you_share_any_examples_of_the_use_of_multivariate_regression [accessed Mar 15, 2016].
Conference Paper Social Policy and Economic Development in Europe: Investigat...
Hi Munji,
Your problem falls in 3rd category, so it is multivariate regression.
Analysis in SPSS:
Multivariate regression is done in SPSS using the GLM-multivariate.option. Here is a nice tutorial on the same:
https://statistics.laerd.com/spss-tutorials/one-way-manova-using-spss-statistics.php#procedure
IBM support doc can also be useful: http://www-01.ibm.com/support/docview.wss?uid=swg21476743
Analysis in R:
If you prefer using R, the this link may be useful:
http://www.psych.yorku.ca/lab/psy6140/lectures/MultivariateRegression2x2.pdf
R package mvtnorm can be used for this purpose.
The multiple dependent variables with one predictor mean multivariate regression, and that suitable for your case, the easiest software for this purpose you can use SPSS
Good Luck
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression