About Vivian Expertise I can answer questions on probability, distributions, statistical inference, statistical estimation, hypothesis testing, analysis of categorical data, linear regression, generalized linear regression, ANOVA, and linear mixed models. I cannot answer questions on stochastic processes.
Experience I have worked as a research assistant at the University of Michigan, Ann Arbor for two years.
Organizations American Statistical Association
Education/Credentials University of Michigan, Ann Arbor
Master of Science
I have the following variables that I would like to examine but am struggling to work out which stats technique to use (I have SPSS 16)
I am looking at the relationship between work (demands, control, & support) , stress, and depression.
All variables have been measured on a continuous scale however depression has a peak of respondants at 0 (ie no depression) followed by a strong positive skew/kurtosis. The pattern for stress is similar-but not as strong (and skweness/kurtosis is not too bad).
I would want to examine the main effects and interactions of demand, control, support on the relationship between stress versus depression.
The interactions that I would want to look at would be for example: demand (workload) x control (workload) x support (co-worker) aswell as : demand (self) x control (workload) x support (co-worker) etc etc.
I want to do these intxns in order to answer the questions of which types of control buffer which types of demands , and which types of support buffer which types of demands.
Because of the potential need to change my DVs of depression and perhaps into categorical variables, how would logistic regression fare? And which version? Mutlinomial? Ordinal? Can the DV be split into three categories? Can it tolerate continuous IVs as my output using multinomial regression gives a Wald statistic for each number on my continuous scale, as opposed to one Wald statistic for each continous variable.
Note that I have also only used one DV at a time.
I hope this makes sense. Any guidance would be extremely helpful.
Suzi
Answer Hi Suzi
Because of the potential need to change my DVs of depression and perhaps into categorical variables, how would logistic regression fare? And which version? Mutlinomial? Ordinal? Can the DV be split into three categories?
>> You may use logistic regression. You should use ‘ordinal’ if you have more than 2 categories since the order makes sense. Sure, you can split it into three categories when it is reasonable to consider no depression, low depression, and high depression. By the way, what does DV stand for? Domain of values?
Can it tolerate continuous IVs as my output using multinomial regression gives a Wald statistic for each number on my continuous scale, as opposed to one Wald statistic for each continous variable.
>> I do not understand this question well. Could you explain it? What dose IV stands for? IV stands for interval?
