Skip to main content

clusteR Documentation

6 Analysis

clusteR provides tools to perform data analysis. Once data has been cleaned, you are ready to analyze. Data management tasks, like cleaning and weighting, are covered in an earlier section.

6.1 view_topline

This is the primary function to calculate and view survey results. It automatically combines columns, calculates the frequency of each answer, and applies weights to survey responses, then generates a report of the results.

view_topline only requires the codebook argument, which should be a string representing a file path to a key for the survey questions. The file must be tab-delimited and include a QN column containing the question numbers corresponding to columns in the cleaned data and a type column specifying the question type. The type may be:

  • mc: Multiple-choice, select one answer per question
  • ms: Multiple-select, select as many answers as apply per question
  • no: Numeric, provide a numerical answer to a question
  • ft: Free-text, provide a free-text answer to a question (which is usually categorized and coded like a multiple-select question during cleaning)

Without specifying other arguments, view_topline generates a simple PDF report based on the most recent cleaned data without weighting. This behavior can be altered with the following:

  • data can be set to a different file path to use different data
  • make can be changed to “csv” to produce a tidy spreadsheet
  • template can be changed to a path to a .Rmd file to customize the report produced; a starter template is available with custom_topline
  • wt can be set to apply survey weights, which must be a numeric variable in the survey data
  • exclude can be set to a character vector of questions to exclude from weighting, such as demographic variables; this does not exclude questions from the report, which can be accomplished by excluding them from the codebook file

6.2 view_breakdown

This function breaks down survey results by a stratification variable, typically a demographic variable. It automatically combines columns, calculates the frequency of each answer, and applies weights to survey responses, then generates a report of the results.

The only additional argument in view_breakdown compared to view_topline is strata, which allows you to specify a the name of a stratification variable as a string. Results tables in the report will have an additional column on the left representing the demographic variable. This variable must already exist in the survey data.

A starter custom template is available with custom_breakdown.