如何在4组中找到不变性？

Question

我使用此代码查找性别、年龄组之间的差异：

    GeneralModel.fit = cfa(model = GeneralModel,
                           data=Invariance,
                           meanstructure = TRUE)

summary(GeneralModel.fit,
        standardized=TRUE,
        rsquare=TRUE,
        fit.measure = TRUE)

parameterestimates(GeneralModel.fit, standardized=TRUE)
fitted(GeneralModel.fit)
residuals(GeneralModel.fit)
fitmeasures(GeneralModel.fit)
modificationindices(GeneralModel.fit)

semPaths(GeneralModel.fit,
         whatLabels = "std",
         layout = "tree")

Teenagers = subset(Invariance, Group == "0")

Teenagers.fit = cfa(GeneralModel,
                     data=Teenagers,
                     meanstructure=TRUE)
summary(Teenagers.fit,
        standardized=TRUE,
        rsquare=TRUE,
        fit.measure=TRUE)

parameterestimates(Teenagers.fit, standardized = TRUE)
fitted(Teenagers.fit)
residuals(Teenagers.fit)
fitmeasures(Teenagers.fit)

semPaths(Teenagers.fit,
         whatLabels = "std",
         layout = "tree")

CollegesStudents = subset(Invariance, Group == "1")
CollegeStudents
CollegeStudents.fit = cfa(GeneralModel,
                     data=CollegeStudents,
                     meanstructure=TRUE)
summary(CollegeStudents.fit,
        standardized=TRUE,
        rsquare=TRUE,
        fit.measure=TRUE)

parameterestimates(CollegeStudents.fit, standardized = TRUE)
fitted(CollegeStudents.fit)
residuals(CollegeStudents.fit)
fitmeasures(CollegeStudents.fit)
IM <- measurementInvariance(model=GeneralModel, data=Invariance, group="Group") 

对于性别差异，我使用了同样的方法(当然，名字也不同)。

现在的问题是：

如何改变这一脚本，以找到不变的四组(青少年妇女，青少年男子，大学妇女，大学男子)？在创建子集时，是否必须使用代码Group1="Sex"和Group2="Age Group"来指定，例如：

Group2 == 0

• teen

• 青少年女性存在：Group1 == 0，Group1 == 0男性：Group1 == 1，Group2 == 0，

• 学院女性存在：Group1 == 0, Group2 == 1

• and大学男性是Group1 == 1, Group2 == 1？

F 218

我不知道我是否用这句话回答了我自己的问题，所以我想确认这是否错误，以及是否有更好的办法来做我想做的事。我提前感谢你。

Answer 1

考虑在用户定义的方法中泛化您的进程，然后使用by运行传递给方法的各种级别的子集。由于某些子集可能不返回任何行，所以在tryCatch中包装方法可以在出错时安全地返回NULL (空结果)：

find_diffs <- function(df) {
    tryCatch({
        model_fit = cfa(model = GeneralModel,
                        data = df
                        meanstructure = TRUE)

        summary(model_fit,
                standardized=TRUE,
                rsquare=TRUE,
                fit.measure = TRUE)

        # NAMED LIST OF RESULTS
        results <- list(estimates = parameterestimates(model_fit, standardized=TRUE),
                        fit = fitted(model_fit),
                        residuals = residuals(model_fit),
                        fitmeasures = fitmeasures(model_fit),
                        mod_indices = modificationindices(model_fit)
                   )
        }, error = function(e) return(NULL)
    )
}

by NO

GeneralModel_fit <- find_diffs(Invariance)
# ALL RESULTS
GeneralModel_fit

# SELECT RESULTS
GeneralModel_fit$estimates
GeneralModel_fit$fit
GeneralModel_fit$residuals
...

by ONE GROUP

sex_fit_list <- by(Invariance, Invariance$Group1, find_diffs)

# WOMEN
sex_fit_list$`0`    
# MEN
sex_fit_list$`1`


age_fit_list <- by(Invariance, Invariance$Group2, find_diffs)

# TEENAGERS
age_fit_list$`0`    
# COLLEGE STUDENTS
age_fit_list$`1`

by 二组

sex_and_age_fit_list <- by(Invariance, Invariance[,c("Group1", "Group2")], find_diffs)

# TEEN GIRLS
sex_and_age_fit_list[[1]]

# COLLEGE WOMEN
sex_and_age_fit_list[[2]]

# TEEN BOYS
sex_and_age_fit_list[[3]]

# COLLEGE MEN
sex_and_age_fit_list[[4]]