De-identified Evaluation Data and Analytical Materials for Structured Information Security Policy Review
Description
This dataset accompanies the manuscript “Structured Deliberation for Information Security Policy Review: Design and Initial Evaluation.” It contains de-identified data and analytical materials from two evaluations of the proposed information security policy-review workflow: (1) ratings provided by nine information security and IT governance professionals across 21 five-point Likert items, and (2) ratings provided by nine organizational workshop participants across seven five-point Likert items. The workbook also includes the research instruments, the de-identified deliberation guide, a participant-by-theme binary matrix, recurrence counts for 28 qualitative themes, a thematic codebook, ten analytical categories mapped to six framework functions, and formula-linked calculations for descriptive statistics, Cronbach’s alpha, item-level agreement proportions, and the mean item agreement proportion for the expert appraisal. The deliberation guide and research instruments were originally administered in Spanish; English translations are included for accessibility. No direct personal identifiers are included. Timestamps, exact organizational roles, respondent-linked sectors, raw qualitative narratives, free-text responses, and any crosswalk linking pseudonymous participant codes to identities are withheld because the small samples and role-specific organizational context could enable deductive identification. Aggregate organizational-role groups and expert-sector distributions are retained. The dataset supports recalculation of the reported quantitative summaries and thematic recurrence counts and permits inspection of the de-identified qualitative coding structure. The participant-by-theme matrix uses arbitrary participant codes that are not linked to identities or post-workshop survey codes. There are no missing values in the expert or post-workshop Likert datasets. Version 2 updates terminology and documentation to align the deposited materials with the revised manuscript. Item-level proportions of ratings of 4 or 5 are described as agreement proportions rather than content-validity indices, and the former S-CVI/Ave terminology is replaced by mean item agreement proportion. No participant responses, qualitative coding values, recurrence counts, reliability estimates, descriptive statistics, or other empirical results were changed.
Files
Steps to reproduce
1.Download and open the XLSX file in Microsoft Excel or compatible spreadsheet software. 2.Consult the README worksheet for the dataset structure, sample descriptions, de-identification decisions, worksheet contents, and Version 2 terminology changes. 3.The Expert Data worksheet contains the ratings of nine information security and IT governance professionals across 21 five-point Likert items. The Workshop Data worksheet contains the ratings of nine organizational participants across seven post-workshop five-point Likert items. 4.The Calculations worksheet derives item means, sample variances, participant total scores, item-level agreement proportions, the mean and minimum item agreement proportions for the expert appraisal, overall item means, and Cronbach’s alpha. 5.The Summary worksheet is linked to the Calculations worksheet and reproduces the quantitative results reported in the manuscript. 6.The Thematic Matrix worksheet contains the de-identified participant-by-theme binary coding matrix used to calculate thematic recurrence. Arbitrary participant codes are used and are not linked to participant identities or post-workshop survey codes. 7.The Theme Codebook and Analytical Codebook worksheets document the 28 consolidated themes, their recurrence counts, and their mapping to ten analytical categories and six framework functions. 8.The Deliberation Guide, Expert Instrument, and Workshop Instrument worksheets document the research instruments used in the study. Spanish-language materials and English translations are provided where applicable. 9.Raw qualitative narratives, exact organizational roles, respondent-linked sectors, timestamps, free-text responses, and any identity crosswalks are not included because the small samples and organizational context could enable deductive identification. 10.The dataset can be used to recalculate the reported descriptive statistics, reliability estimates, agreement proportions, and thematic recurrence counts. Version 2 changes terminology and documentation only; no participant responses, coding values, recurrence counts, or empirical results were modified.