Review labelled data

Package: labelled

---

Function: look_for()

---

1. Review all labelled information in a data dictionary format

Let’s create some labelled data

d10 <- d10 %>%
  labelled::set_variable_labels(
    Var1 = "Name",
    Var2 = "Interest in Homework",
    Var3 = "Lunch Status") %>%
  labelled::set_value_labels(
    Var2 = c("not interested" = 1, "mildly interested" = 2, "very  interested" = 3), 
    Var3 = c("none" = 1, "reduced-price" = 2, "free" = 3)) %>%
  labelled::set_na_values(Var2 = -999, Var3 = -999)

View the labelled data (d10)

# A tibble: 3 x 3
  Var1                      Var2        Var3
  <chr>               <hvn_lbl_>  <hvn_lbl_>
1 a        1 [not interested]       3 [free]
2 b     -999 (NA)                   1 [none]
3 c        2 [mildly interested] -999 (NA)  

Review labelled information

• Note: The argument details=TRUE is added to get na_values for variables

• Note: The function labelled::generate_dictionary() is another function equivalent to labelled::look_for(). You can use either function.

labelled::look_for(d10, details = "full") %>%
  labelled::lookfor_to_long_format() %>%
  labelled::convert_list_columns_to_character() %>%
  dplyr::select(variable, label, value_labels, na_values)

# A tibble: 7 x 4
  variable label                value_labels          na_values
  <chr>    <chr>                <chr>                 <chr>    
1 Var1     Name                 <NA>                  ""       
2 Var2     Interest in Homework [1] not interested    "-999"   
3 Var2     Interest in Homework [2] mildly interested "-999"   
4 Var2     Interest in Homework [3] very  interested  "-999"   
5 Var3     Lunch Status         [1] none              "-999"   
6 Var3     Lunch Status         [2] reduced-price     "-999"   
7 Var3     Lunch Status         [3] free              "-999"   

---

Function: var_label()

---

1. Review just the variable labels

View the data (d10)

# A tibble: 3 x 3
  Var1                      Var2        Var3
  <chr>               <hvn_lbl_>  <hvn_lbl_>
1 a        1 [not interested]       3 [free]
2 b     -999 (NA)                   1 [none]
3 c        2 [mildly interested] -999 (NA)  

Review variable labels

d10 %>%
  labelled::var_label()

$Var1
[1] "Name"

$Var2
[1] "Interest in Homework"

$Var3
[1] "Lunch Status"

---

Function: val_labels()

---

1. Review just the value labels

View the data (d10)

# A tibble: 3 x 3
  Var1                      Var2        Var3
  <chr>               <hvn_lbl_>  <hvn_lbl_>
1 a        1 [not interested]       3 [free]
2 b     -999 (NA)                   1 [none]
3 c        2 [mildly interested] -999 (NA)  

Review value labels

d10 %>%
  labelled::val_labels()

$Var1
NULL

$Var2
   not interested mildly interested  very  interested 
                1                 2                 3 

$Var3
         none reduced-price          free 
            1             2             3 

---

Function: na_values()

---

1. Review just the NA value labels

View the data (d10)

# A tibble: 3 x 3
  Var1                      Var2        Var3
  <chr>               <hvn_lbl_>  <hvn_lbl_>
1 a        1 [not interested]       3 [free]
2 b     -999 (NA)                   1 [none]
3 c        2 [mildly interested] -999 (NA)  

Review NA value labels

d10 %>% 
  labelled::na_values()

$Var1
NULL

$Var2
[1] -999

$Var3
[1] -999

Package: sjPlot

---

Function: view_df()

---

1. Review all labelled information in a data dictionary format

View the data (d10)

# A tibble: 3 x 3
  Var1                      Var2        Var3
  <chr>               <hvn_lbl_>  <hvn_lbl_>
1 a        1 [not interested]       3 [free]
2 b     -999 (NA)                   1 [none]
3 c        2 [mildly interested] -999 (NA)  

Review labelled information

• Note: You will notice that we do not get NA labels here like we do using the labelled::look_for() function. We also do not get the variable class. Instead we can ask for the variable type by adding the argument show.type=TRUE, which shows us that Var2 for instance is numeric but does not let us know that the class is haven_labelled_spss.

d10 %>%
  sjPlot::view_df(show.type = TRUE)

Data frame: .

ID	Name	Type	Label	Values	Value Labels
1	Var1	character	Name		<output omitted>
2	Var2	numeric	Interest in Homework	1 2 3	not interested mildly interested very interested
3	Var3	numeric	Lunch Status	1 2 3	none reduced-price free

Return to Label Data