SurveyTools
surveytools is a new tidyverse compiant package for survey analysis, comprising of miscellaneous functions used often for survey analysis at Dalberg. We have used these tools internally to analyse multiple surveys and dataframes, as well as use as a base to build additional tools and GUI over. Please reach out to us for questions regarding use and example code.
Functions
We provide two key functions that can be used for survey analysis:
- Summariser:
summariser_base forms the work horse of the package. It takes in the following arguments:
| Parameter | Value |
|---|---|
| Data | (Required) Data to be summarised |
| Variable | (Required) Name of atleast one variable to be summarised by. Does not support vectors of values, but supports all tidyverse select formats |
| Grouping variables | Name, or tidyverse selection, of any variable to group the results by. Supports multiple groups concurrently |
| Stat | One of "mean", "median", or "total". Defines the stat to be summarised over |
| Survey Design | Required for analysis of surveys. Provides the structure of the survey (i.e. pps, etc.) Please see the `survey` package for full list of available options that can be passed in the list |
| Compare significance to | Changes the default total against which all significance resuts are tested. Please provide in a "group::response" structure |
| Default level of significance | 0.95. Can be modified |
The function is pipping compliant -> the output of the function is always a dataframe.
- Crosstabs:
cross_tab is our second primary function. It allows you to convert the tidy data recieved from the summarise_base function, into a more commonly discernable dataframe