lm formula: the difference between `as.factor()` and `as.as.character()`

I was wondering why factor(gpid) doesn't work like as.character(gpid) in my manova() call below?

data <- read.csv("https://raw.githubusercontent.com/rnorouzian/v/main/df.csv")

#   gpid anx socskls assert
#1     1   5       3      3
#2     1   5       4      3
#3     1   4       5      4

summary(manova(cbind(anx,socskls,assert) ~ factor(gpid), data = data))

#          Df   Pillai approx F num Df den Df Pr(>F)
#gpid       1 0.073734   0.7695      3     29 0.5205

But when I change gpid to character the output is what it should be:

summary(manova(cbind(anx,socskls,assert) ~ as.character(gpid), data = data))

#                    Df  Pillai approx F num Df den Df   Pr(>F)   
# as.character(gpid)  2 0.62209   4.3642      6     58 0.001058 **

from Recent Questions - Stack Overflow https://ift.tt/3b2X8hP
https://ift.tt/eA8V8J