Author: Mark Rieke
License: MIT
Note: As of November 3rd, 2023, this project is no longer under active maintenance. If you are a Memorial Hermann employee and would like to take over as the maintainer, please open an issue.
Overview
heRmann is a collection of functions for the Consumer Experience Analytics (CXA) team at Memorial Hermann Health System. The package provides convenience functions for interacting with Qualtrics’ API and performing common survey-specific data manipulation tasks.
Installation
This package is intentionally not available on CRAN. You can install the most up-to-date version from github with the devtools package.
devtools::install_github("markjrieke/heRmann")Usage
A typical workflow involves extracting survey responses from Qualtrics and prepping for analysis. Below is an example using fake data for the Inpatient, Pediatric Inpatient, and Inpatient Rehab surveys. To import real responses, however, we can pass a vector of survey names as they appear in the Qualtrics UI to fetch_surveys().
library(heRmann)
surveys <-
c("Live - Inpatient",
"Live - Pediatric Inpatient",
"Live - Adult Inpatient Rehab") |>
fetch_surveys()
surveys#> # A tibble: 3 × 2
#> survey_name responses
#> <chr> <list>
#> 1 Live - Inpatient <tibble [6 × 3]>
#> 2 Live - Pediatric Inpatient <tibble [1 × 3]>
#> 3 Live - Adult Inpatient Rehab <tibble [2 × 3]>The survey names in the UI, however, are lengthy and annoying. Further, because of Qualtrics’ HCAHPS sampling engine, the Inpatient survey contains duplicate responses from the Pediatric Inpatient and Inpatient Rehab surveys. fix_survey_names() and deduplicate_ip() address these issues, respectively.
surveys <-
surveys |>
fix_survey_names() |>
deduplicate_ip()
surveys
#> # A tibble: 3 × 2
#> survey_name responses
#> <chr> <list>
#> 1 Pedi Inpatient <tibble [1 × 3]>
#> 2 Inpatient Rehab <tibble [2 × 3]>
#> 3 Inpatient <tibble [3 × 3]>Finally, unnesting reveals that the default recodes for campus are similarly lengthy and annoying. fix_campus_names() reformats to a more intuitive description:
surveys <-
surveys |>
tidyr::unnest(responses)
surveys
#> # A tibble: 6 × 4
#> survey_name UNIQUE_ID campus response
#> <chr> <chr> <chr> <chr>
#> 1 Pedi Inpatient 12345 Children's Memorial Hermann Hospital Promoter
#> 2 Inpatient Rehab 23456 Memorial Hermann - Texas Medical Center Passive
#> 3 Inpatient Rehab 34567 Memorial Hermann - Texas Medical Center Promoter
#> 4 Inpatient 45678 Memorial Hermann - Texas Medical Center Passive
#> 5 Inpatient 56789 Memorial Hermann - Texas Medical Center Detractor
#> 6 Inpatient 67890 Memorial Hermann - Texas Medical Center Promoter
surveys |>
fix_campus_names()
#> # A tibble: 6 × 4
#> survey_name UNIQUE_ID campus response
#> <chr> <chr> <chr> <chr>
#> 1 Pedi Inpatient 12345 CMHH Promoter
#> 2 Inpatient Rehab 23456 TMC Passive
#> 3 Inpatient Rehab 34567 TMC Promoter
#> 4 Inpatient 45678 TMC Passive
#> 5 Inpatient 56789 TMC Detractor
#> 6 Inpatient 67890 TMC Promoter