Open-Ended Response Coding Worksheet

Each nonblank input line becomes one response. Keep one answer per line; this tool does not parse quoted CSV or multiline records. It supports up to 200 responses and 20 unique code names. Duplicate response text remains separate because two people may give the same answer.

Define codes as Name | Definition, one per line. For example, Access | Difficulty finding or opening the service. Read every response in context and apply all relevant codes. Repeated mentions of the same theme in one response count once for that code.

A code percentage uses all loaded responses as the denominator, including unreviewed and no-code responses. With three responses, a code assigned to two is 66.7%. Because one response can receive several codes, percentages can total more than 100%. Review coverage is reported separately.

Mark each reviewed response explicitly, including answers that do not fit any code. A reviewed answer with no assignment differs from an answer not yet examined. Changing the source answers or code definitions requires applying a new worksheet, which clears old assignments.

Before reporting, inspect uncoded answers and have another reviewer apply the same definitions to a sample. Resolve disagreements with examples and update the codebook. This tool does not assess inter-rater reliability, discover hidden themes or establish population prevalence. Export contains code definitions, counts and denominators; it excludes raw answers.