Quality_Control_Array_Writer: Perform automatic quality checks on the filtered surveys THIS...

In bpresentati/surveyR: surveyR: Get things done with LimeSurvey!

Perform automatic quality checks on the filtered surveys THIS VERSION DOES NOT HAVE AN UP-TO-DATE DOCUMENTATION! This is misleading!

Description

This function takes a survey list (filtered) and performs quality control checks on it. The QC revolves around four main aspects:

1. Survey timing. The bottom 2.5% of times are automatically filtered out. The bottom 2.5%-5% are flagged, and need to be combined with one of the other three criteria to classify the complete as BQ.

2. Variance in the driver questions: For the America's Best Employers Suite, we have a battery of "driver questions" that are 1-5 Likert type arrays. These all have the prefix "qd" in their variable codes and are found in the first survey section. The variance of these questions is tested, and the bottom 5% is flagged.

3. "Q34" - Please answer this question with "5" if you are paying attention- otherwise the response is flagged.

4. "Plausibility check" - Does the average driver response (good or bad) generally agree with the overall level of recommendation (q5.1.)? Otherwise the response is flagged.

Usage

Quality_Control_Array_Writer(
  Filtered_Survey_List,
  NPS_Name = "q5.1.",
  HighNPS = 7,
  LowDriver = 2,
  LowNPS = 3,
  HighDriver = 4,
  QCQ_Value = 5,
  QCQ_Name = "qdxQCA.x001.",
  Variance_Percentile = 0.05,
  Driver_Identifier = "^qd",
  key_name = "key"
)

Arguments

Filtered_Survey_List	Contains the entire filtered survey list
NPS_Name	<optional, default is "q5.1."> The name of the variable that contains the NPS (or "most important" score, that can be compared against a battery of drivers). This is the "big" NPS question (or something similar, evaluated on a ten point scales and that can be compared with the drivers)
HighNPS	This is the "high" direct evaluation value for the plausibility check (default is 7)
LowDriver	This is the "low" mean value for the question battery for the plausibility check (default is 3)
LowNPS	This is the "low" direct evaluation value for the plausibility check (default is 2)
HighDriver	This is the "high" mean value for the question battery for the plausibility check (default is 4)
QCQ_Value	This is the value that the respondent should click on for the Quality Check Question, so that it will NOT be flagged as bad quality.
QCQ_Name	<optional, default "qdxQCA.x001."> - this is the identifier of the quality control question
Variance_Percentile	<optional, default 0.05> this is the percentile under which the variance of the drivers is considered a flag for BQ
Driver_Identifier	<optional, default "^qd"> - this is whatever uniquely identifies the driver questions. You can use regex here.
key_name	default "key" - this is the unique ID name for each response

Value

The filtered survey list, but with a column in each list element that is called "BQ" to flag all of the responses that fail the quality control test.

Examples

Quality_Control_Array_Writer(Filtered_Survey_List, NPS_Name = "q5.1.", HighNPS = 7, LowDriver = 2, LowNPS = 3, HighDriver = 4, QCQ_Value = 5, QCQ_Name = "qdxQCA.x001.", Variance_Percentile = 0.05, Driver_Identifier = "^qd", key_name = "key")

bpresentati/surveyR documentation built on March 19, 2022, 3:40 a.m.