wildetudor
06-01-2011, 05:25 PM
Hello everyone,
I have 50 musical rhythmic patterns, each characterised by two measures: rhythmic tension and joint entropy. In a repeated measures design, I collected participant ratings for each of these rhythms, in terms of the rhythms making them (i) want to move and (ii) feel pleasure.
I tested the ratings and the residuals for normal distribution, as well as their log, square root and reciprocal transformations, and the results showed a non-normal distribution in every case. My hypothesis (which scatter plots of the data seem to support) is that there is a quadratic relationship between the ratings and the independent variables.
Thus it would seem that I need to perform a nonparametric multiple nonlinear (quadratic) regression with two predictors and two dependent variables. I've tried every other way of doing this in SPSS that I could think of, but am really struggling. Can anyone help me with a model?
Any instructions on how to implement this in SPSS in particular would be much appreciated! Many thanks in advance.
Due to being pressed for time, apologies for cross-posting this on another forum (http://www.talkstats.com/showthread.php/18266-Multiple-nonlinear-regression-%28non-parametric%29-with-2-predictors-and-2-DVs?p=52960#post52960).
I have 50 musical rhythmic patterns, each characterised by two measures: rhythmic tension and joint entropy. In a repeated measures design, I collected participant ratings for each of these rhythms, in terms of the rhythms making them (i) want to move and (ii) feel pleasure.
I tested the ratings and the residuals for normal distribution, as well as their log, square root and reciprocal transformations, and the results showed a non-normal distribution in every case. My hypothesis (which scatter plots of the data seem to support) is that there is a quadratic relationship between the ratings and the independent variables.
Thus it would seem that I need to perform a nonparametric multiple nonlinear (quadratic) regression with two predictors and two dependent variables. I've tried every other way of doing this in SPSS that I could think of, but am really struggling. Can anyone help me with a model?
Any instructions on how to implement this in SPSS in particular would be much appreciated! Many thanks in advance.
Due to being pressed for time, apologies for cross-posting this on another forum (http://www.talkstats.com/showthread.php/18266-Multiple-nonlinear-regression-%28non-parametric%29-with-2-predictors-and-2-DVs?p=52960#post52960).