Load and Clean CEnTR*IMPACT Survey Data — load_survey

Reads a CEnTR*IMPACT survey CSV file, detects the export format (Google Forms or Qualtrics), applies the appropriate header translation and cleaning, and returns analysis-ready data frames.

Usage

load_survey_data(file, survey = c("main", "alignment"))

Arguments

file: Character. Path to the CSV file to load.
survey: Character. Which survey type to load. One of "main" or "alignment".

Value

For survey = "main": a named list with elements:

indicators: long-format indicators data frame for visualize_indicators().
dynamics: long-format dynamics data frame for analyze_dynamics().
cascade: wide-format cascade parameters data frame for analyze_cascade().
source: the detected source format, "google" or "qualtrics".

For survey = "alignment": a long-format data frame for analyze_alignment().

Details

Survey types:

survey = "main" reads the primary CEnTR*IMPACT survey containing indicators, project dynamics, and cascade network parameters. The function detects the export format automatically:

Google Forms: identified by the presence of a Timestamp column. Verbose question-text headers are translated to compact variable names, dynamics_selection (absent from the form) is inserted as NA, and cascade probability fields are rescaled from the 1–10 integer scale to \([0, 1]\).
Qualtrics: identified by the presence of Qualtrics system columns (StartDate, ResponseId, etc.). System meta columns, descriptor rows, and _selection columns are removed. The " Applicability" suffix is stripped from rating values and the "Not Applicabie" typo is corrected.

Returns a named list with three analysis-ready data frames: indicators, dynamics, and cascade.

survey = "alignment" reads the alignment survey containing researcher and community partner ratings across eight partnership factors. Qualtrics system columns and descriptor rows are removed when present. Returns a single long-format data frame ready for analyze_alignment().

Multiple rows: For both survey types and both formats, if more than one data row is present after cleaning, only the last row is kept. This handles Qualtrics exports that include preview or test responses before the final submission.

Examples

if (FALSE) { # \dontrun{
# Load the main survey (auto-detects Google Forms or Qualtrics format)
survey <- load_survey_data("data/mhfa_main.csv", survey = "main")

# Access the ready-to-use sub-frames
indicators_data <- survey$indicators
dynamics_data   <- survey$dynamics
cascade_data    <- survey$cascade

# Load the alignment survey
alignment_data <- load_survey_data("data/mhfa_alignment.csv", survey = "alignment")
} # }