Upload your Barrier Analysis data (CSV or Excel), define Doers and Non-Doers according to responses given for particular BA questions (the field names in the column headers of your spreadsheet), and get responses ranked by Estimated Relative Risk — replicating the automated BA Tabulation Table.
1Upload your data file
One row per respondent, one column per question/response, with column headers in the first row. CSV, .xlsx, .xls and .xlsm accepted (only the data is read — macros are ignored).
2Define a Doer
How to type values: enter the value exactly as it appears in your spreadsheet cells, without quotation marks — type ALL, not "ALL". Matching ignores capitalization and extra spaces, so all, All and ALL are all treated the same.
Combining AND and OR: within a group, every condition must be true (AND). A respondent is a Doer if they match any one group (OR). Example — "#1 = Yes AND #3 = No AND (#4 = Sometimes OR #5 = Always)" becomes Group 1: #1 = Yes, #3 = No, #4 = Sometimes; Group 2: #1 = Yes, #3 = No, #5 = Always.
3Define a Non-Doer
Respondents who are neither Doers nor Non-Doers are excluded from the analysis (as on the tabulation sheet).
4Analysis settings
Results
How this is calculated: for each response a 2×2 table is built — Doers giving / not giving the response, and Non-Doers giving / not giving. The p-value tests whether Doers and Non-Doers differ (default: N−1 chi-square, switching to Fisher's exact when an expected count falls below 1; other tests, including the ABATT-matching Wald test on the log odds ratio, are selectable above). The Estimated Relative Risk (ERR) is the prevalence-adjusted relative risk of being a Doer: using your estimated prevalence P, it compares P(Doer | gave the response) with P(Doer | did not), which is the appropriate adjustment for BA's fixed Doer / Non-Doer sampling design. ERR > 1 (green) = response associated with Doers (potential enabler); ERR < 1 (red) = associated with Non-Doers (potential barrier). The Interpretation column restates each statistically significant finding in plain language (always as a number greater than 1). The Odds Ratio is shown for reference. Columns used to classify Doers/Non-Doers are excluded automatically.