Crunch core survey data
Once SQL data has been downloaded, you’ll want to generate funder-level ratings from core survey questions. These ratings will be pulled into the JSON file for a specific funder later on. At crepe interface, type option 3
to start the process. With this step, the following files will be generated under *pr/core/agg
:
allmean.csv
: Using the respondent data filexl.csv
and the list of likert variables in*pr/core/var/likert.csv
, crepe will generate all funder-level ratings in this file.allprop.csv
: Using the respondent data filexl.csv
and the variable dictionariesbar.csv
andstack.csv
in*pr/core/var
, crepe will generate all funder-level categorical proportions, used for bar charts, stacked bar charts, and toggle tables, in this file.allcount.csv
: Using all variable dictionaries in*pr/core/var
and the respondent data filexl.csv
, crepe will generate counts of responses by funder by question. This file is used to filter out funder data that has fewer than five responses.meanstats.csv
: This file contains all comparative data, e.g., quartile ratings at 0%, 25%, 50%, 75%, and 100% that can be fed into the percentile charts for Likert questions.meanprop.csv
: This file contains all comparative data, e.g., average and median proportions of all funders that can be fed into bar charts, stacked bar charts and tables for categorical questionsscmeanstats.csv
: This file contains all standard cohort data, e.g., median rating of all Health Conversion Funders, that can be fed into the percentile charts for Likert questions.scpropstats.csv
: This file contains all comparative data, e.g., average and median proportions of all Health Conversion Funders, that can be fed into bar charts, stacked bar charts and tables for categorical questions.