R合并Likert点

Question

我有一个10分的利克特刻度表，看起来是这样的：

    > head(dsidata$Mood_2)
[1] Great Great Bad   Good  Great So-so
Levels: Couldn't be worse Very bad Bad Meh So-so Okay Good Very Good Great Really great

我想要‘合并’类别，使因素的数量减少到5个。选择选项“不能更糟”或“非常糟糕”的受访者将被分配到新的“非常负面”的类别，直到这10个因素减少到5个。

我们非常感谢您对此的任何帮助。

Answer 1

我肯定你的变量被归类为一个因素。所以，就用这种方法合并一些因素吧。

set.seed(42)  # for reproducibility

# mimicking your data
df1 <- data.frame(var=rdunif(100, 10))
df1$var <- factor(df1$var, 1:10, labels=c("Couldn't be worse", "Very bad", 
                                        "Bad", "Meh", "So-so", "Okay",
                                        "Good", "Very Good", "Great", 
                                        "Really great"))

df1$var1 <- df1$var

# merging levels
levels(df1$var1) <- list("Very negative"=c("Couldn't be worse", "Very bad"),
                        "negative" = c("Bad", "Meh"),
                        "neutral" = c("So-so", "Okay"),
                        "positive"=c("Good", "Very Good"),
                        "very positive"=c("Great", "Really great"))

产量

> tail(df1, 10)
                  var          var1
91           Very bad Very negative
92          Very Good      positive
93           Very bad Very negative
94           Very bad Very negative
95  Couldn't be worse Very negative
96  Couldn't be worse Very negative
97               Okay       neutral
98           Very bad Very negative
99          Very Good      positive
100         Very Good      positive

Answer 2

它取决于您的数据集，但您可以：

dsidata$Mood_2_newscale <- ifelse(dsidata$Mood_2 %in% c("Couldn't be worse","Very bad"),
"Very negative",
dsidata$Mood_2)