Dear all,
I am using GSEM model to testify the factors of perceived flexibility and satisfaction. Both flexibility and satisfaction are measured by 5-point Likert scale. (i.e. 1=very inflexible/dissatisfied, 5=very flexbile/satisfied).
I want to use perceived flexibility as mediators of the effects of sociodemographics on satisfaction (sociodemographics→flexibility→satisfaction). So I use ordered logit model for flexibility.
And in previous descriptive and Anova analysis, it is found that more inflexible (flexible=1/2) and more flexible activities (flexible=4/5) have higher satisfaction level than neutral ones ((flexible=3), which is like 'U'-shaped.
So when I conduct the ordered logit models on satisfaction, I want to use flexible=3 as base, to see the effects of more inflexible (flexible=1/2) and more flexible activities (flexible=4/5) on satisfaction.
Here are my syntax. Is it okay to do so?
gsem ( Satisfaction <- 1.flex 2.flex 4.flex 5.flex gender age40 age40to60 employed , ologit) (flex <- gender age40 age40to60 employed, ologit) ( timeflex_3cat <- gender age40 age40to60 employed , ologit)
I am using GSEM model to testify the factors of perceived flexibility and satisfaction. Both flexibility and satisfaction are measured by 5-point Likert scale. (i.e. 1=very inflexible/dissatisfied, 5=very flexbile/satisfied).
I want to use perceived flexibility as mediators of the effects of sociodemographics on satisfaction (sociodemographics→flexibility→satisfaction). So I use ordered logit model for flexibility.
And in previous descriptive and Anova analysis, it is found that more inflexible (flexible=1/2) and more flexible activities (flexible=4/5) have higher satisfaction level than neutral ones ((flexible=3), which is like 'U'-shaped.
So when I conduct the ordered logit models on satisfaction, I want to use flexible=3 as base, to see the effects of more inflexible (flexible=1/2) and more flexible activities (flexible=4/5) on satisfaction.
Here are my syntax. Is it okay to do so?
gsem ( Satisfaction <- 1.flex 2.flex 4.flex 5.flex gender age40 age40to60 employed , ologit) (flex <- gender age40 age40to60 employed, ologit) ( timeflex_3cat <- gender age40 age40to60 employed , ologit)
