Get one column of outcome questions for each model

get_cces_question(
  qcode,
  year,
  qID,
  dataframe = NULL,
  y_named_as = "response",
  data_dir = "data/input/cces",
  verbose = TRUE
)

Arguments

qcode: A string, the variable name of the outcome variable of interest, exactly as it appears in the CCES dataset. Together with year, uniquely identifies the outcome question data
year: A string for the year the question was asked, which defines the the dataset. If using a dataset from the cumulative, set to "cumulative". Together with qcode, uniquely identifies the outcome question data
qID: A string, the user's unique name for the question. For example, we use the syntax in our question table.
dataframe: An optional argument to pass the entire CCES dataframe in-environment. If left empty, this will supersede the arguments year and dataframe and use the provided datasets instead.
y_named_as: What to name the response variable. Defaults to "response", which is also the default assumption for cces_join_slim.
data_dir: A path for the directory where flat files for the CCES common content is stored. Currently the data must be of the form cces_{year}.rds (e.g. "cces_2016.rds") and it must exist wherever data_dir is. We recommend looping over get_cces_dataverse to download data first.
verbose: whether to print a message; defaults to TRUE

Value

A three-column dataframe with the columns

year: The year
case_id: The respondent identifier within year
qID: The question ID
response: The outcome question, of class factor

The object will also have an attribute called question, which will save the question identifier qID

Details

This transformation currently only supports Yes/No questions. For some common ordinal questions that can, with manual recodes, be set to a Yes/No question, there is some hard-coding of the recode in the question. In the future, this should be defined outside of the function in a taxonomy.

Examples


 # with sample data
 get_cces_question("pid3", dataframe = cc18_samp, year = 2018, qID = "pid3")
#> Using the dataframe provided 
#> # A tibble: 1,000 × 4
#>     year case_id   qID   response   
#>    <int> <chr>     <chr> <fct>      
#>  1  2018 415395741 pid3  Independent
#>  2  2018 414164923 pid3  Republican 
#>  3  2018 412379892 pid3  Democrat   
#>  4  2018 414203529 pid3  Republican 
#>  5  2018 412148048 pid3  Independent
#>  6  2018 412329835 pid3  Independent
#>  7  2018 417352072 pid3  Democrat   
#>  8  2018 414614677 pid3  Democrat   
#>  9  2018 416797006 pid3  Republican 
#> 10  2018 412962561 pid3  Independent
#> # ℹ 990 more rows

 # need data/input/cces/cces_2018.rds to run this
 if (FALSE) {
  get_cces_question(qcode = "CC18_326", year = "2018", qID = "TCJA")

  # A tibble: 60,000 x 4
  # year   case_id qID   response
  # <int>     <int> <chr> <fct>
  # 1  2018 123464282 TCJA  Support
  # 2  2018 170169205 TCJA  Support
  # 3  2018 175996005 TCJA  Support
  # 4  2018 176818556 TCJA  Oppose
  # 5  2018 202120533 TCJA  Oppose
  # 6  2018 226449148 TCJA  Oppose
  # 7  2018 238205342 TCJA  Oppose
  # 8  2018 238806466 TCJA  Support
  # 9  2018 267564481 TCJA  Support
  # … with 59,991 more rows
 }