I have 3 research questions to answer. RQ 1 & 2 ask if there is a predictive relationship between mindset and gpa and SES and gpa. RQ 3 ask if mindset predicts gpa while moderating for SES. All are continuous variable. Can one moderated multiple regression answer all three RQ questions? Will I have to run a linear and multiple regression?
I don't think so.because the models are asking different questions. If you would write everything out it would be helpful. Best wishes David Booth
So I need to have two seperate models. A linear regression for the two questions addressing if both mindset/SES predict the dependent variable of GPA Respectively? And a moderated multiple regression for the third question regarding if SES moderates An effect on mindset predicting GPA? Is that what you mean by write it out?
Hi Adrian,
so GPA is your dependent variable and both predictors (mindset and SES) are interacting predictors of GPA. If this is correct, you estimate a model with a product term of SES and mindset.
All the best
Holger
Mr. Holger, I’m not sure I understand what that means. “Product term” Will you please clarify?
Hi Adrian,
simple as it sounds. You simply multiply SES and mindset and include this product as a further variable into the regression model. Hence the equation is (MS = mindset)
Y_i = b0 + b1*SES_i + b2*MS_i + b3*SESxMS_i + e_i
Google for "moderated regression" and you will find tons of ressources.
One note: To better interpret the form of the interaction/moderation, you should mean-center SES and MS before multiplying them which makes b1 and b2 being the effect of SES [MS] at an average level of MS [SES]. This helps especially with graphical display of the interaction, see http://www.jeremydawson.co.uk/slopes.htm
Some say that this also reduces the inherent multicolinearity created by including the product while this is a myth according to others:
Echambadi, R., & Hess, J. D. (2007). Mean-centering does not alleviate collinearity problems in moderated multiple regression models. Marketing Science, 26(3), 438-445.
HTH
Holger
I wish i was as advance as you with understanding this, but I am new to statistics. However, I will be utilizing SPSS to calculate each of the regression equations. How can I make this adjustment in the software?
Which adjustment? The calculation of the product term? This is simple but if you forgive me, I would delegate this to you :) Watch a few tutorial. Everything needs time, software, too. But if you start from the beginning, better start with R (tidyverse). It will pay off!
As in all things, I’m under a deadline and need help. (Which is why I’m here asking for help from those who are much better at this than myself). Any help is deeply appreciated.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Graphs