Completed coding sheets and agreed ordinal profile for a two-sided documentary readiness instrument, Greece and four reference states
Description
This dataset holds the completed coding record behind the public-side ordinal instrument used in the article, which reconstructs nuclear readiness as a two-sided problem and reads the documentary preconditions of trust across Greece, Italy, Estonia, Poland and the United Arab Emirates. The instrument scores twenty five cells, five documentary dimensions for each of the five cases, on a zero to three band rule. The workbook has five tabs. A Read me tab records provenance and scope. The Coder A and Coder B tabs give the anchor read for each cell and that coder band, one cell per row, and Coder A also carries the baseline and strict coding conventions cell by cell, with the band each convention returns and the reason for the difference where they differ, so that the agreement between the two conventions can be followed at the level of the individual cell. The Divergences tab lists the seven cells where the two coders differ and how each was reconciled. The Agreed profile tab gives the reconciled band distribution, the mean, the sample standard deviation, and the two agreement tallies for each country. The Band columns hold the coder entries as returned and are not altered, and each coding tab also carries an Agreed band column holding the reconciled score so that the profile tab can be traced to a single column. Every figure in the Agreed profile tab is stored as a value and not as a formula, so the file opens with the numbers it reports and no recalculation is needed. The sources behind every anchor are public documents cited in the article with their retrieval dates. No personal or sensitive data are included and no new measurements were generated. The scoring rule, the intersection mapping and the search log are in the Supplementary Material of the article, which is deposited with the journal and not with this record.
Files
Steps to reproduce
1. Open the workbook in any spreadsheet application. The Agreed profile values are stored, so they appear without recalculation. 2. The agreed band of the twenty five cells is in the Agreed band column of either coding tab. It is the reconciled score of the two coders and not a third convention, so it follows Coder B in all 25 cells and Coder A in 18 of the 25. Against the strict column of Coder A it departs in nine cells, and the two conventions agree with each other in 20 of 25, the figure reported in Section 3.3. 3. Reproduce the country profiles from the five agreed bands. Greece gives 0.80 and 0.84, Italy 1.40 and 0.55, Estonia 1.20 and 0.45, Poland 2.60 and 0.55, and the United Arab Emirates 2.00 and 1.41, each as mean and standard deviation. These match the final column of Table 4. 4. Reproduce the convention agreement. Count the cells where the baseline and strict columns of Coder A return the same band, which gives 2, 4, 4, 5 and 5 by country and 20 of 25, with a mean absolute band shift of 0.20 and a quadratic weighted agreement of 0.881. The divergences are G1, G3, G5 and I3 on the 1 to 2 boundary and E5 on the 2 to 3 boundary. 5. Reproduce the coder agreement from the cells where the Coder A band equals the Coder B band, which gives 4, 3, 1, 5 and 5 by country and 18 of 25, an exact agreement of 72 percent with a mean absolute band shift of 0.28 and a quadratic weighted agreement of 0.851. 6. Reproduce the coefficients from the two band columns. Between the coders, Cohen kappa returns 0.623 and the ordinal Krippendorff alpha 0.816. Between the conventions, the unweighted kappa returns 0.702 and the ordinal alpha 0.845. Quadratic weighting and the interval metric define the same distance, so each pair reads as one measurement under two definitions. 7. Trace any number to its cell through the Cell identifier, where G1 to G5 are Greek, I1 to I5 Italian, E1 to E5 Estonian, P1 to P5 Polish and U1 to U5 Emirati.
Institutions
- National and Kapodistrian University of AthensAttica, Athens