survey - @gtsummary converting double to categorical why

Question

I am working on a COVID-19 survey. The main goal was to ask a question and score were given depending on the answer ranging from 0-30, but when I performed the analysis the package resume my variables as if they were categorical. I want the mean score for each variable.

Here is the data frame:

> head(PRA001)
# A tibble: 6 × 6
  ID    A quelle fréquence v…¹ Distance_2m Eviter_contact_sociaux Porter_masque_facial
  <chr>                  <int>       <int>                  <int>                <int>
1 ORC0…                     20          20                     10                   10
2 ORC0…                      0          10                      0                   20
3 ORC0…                      0           0                      0                    0
4 ORC0…                     20          20                     20                   20
5 ORC0…                      0           0                      0                    0
6 ORC0…                      0           0                      0                    0

summary_stats <- PRA001%>%select(-ID)%>%tbl_summary() %>%
  add_stat_label(label = everything() ~ "Mean (SD)")

I was expecting to have the mean score for each variable before and after (from the time point variable) the intervention then the mean difference but this what I got.

Answer 1

When there are fewer variables, gtsummary treats it as categorical. I am not sure what the exact threshold for that is. I am sure someone may tell me.

You can use the type argument to make sure things are analysed as continuous variables. (you seem to have tried using the label argument).

summary_stats <- head(PRA001)%>% 
select(-ID)%>%
tbl_summary(., 
type = list(everything() ~ "continuous"))

You can also make a quick adjustment to the variable labels:

summary_stats[["table_body"]][["label"]]<- summary_stats[["table_body"]][["label"]]%>%gsub("_", " ", .)%>%str_to_title()