Investigating Enablers, Barriers, and Consequences of Psychological Safety in Software Engineering

This repository contains the replication package for the paper "Investigating Enablers, Barriers, and Consequences of Psychological Safety in Software Engineering", submitted to SBES 2026. Overview This study investigates the enablers, barriers, and consequences of psychological safety (PS) in software engineering (SE) through semi-structured interviews with 12 software professionals. Data were analyzed using thematic analysis (Clarke & Braun, 2013), with codes organized across four levels of influence (Leadership, Team, Organizational, Individual). Contents The package consists of a single spreadsheet file (Data_and_Codebook-InterviewPS.xlsx) with five sheets: 1. Enablers (RQ1) Coded excerpts supporting RQ1 (factors that enable PS). Each row corresponds to one coded excerpt. Column Description Participant Participant ID (P1–P12), as described in Table 4 of the paper Level Level of influence (Leadership, Team, Organizational, Individual) Enablers Enabler code assigned to the excerpt Quote English translation of the excerpt original quote in Portuguese Original excerpt as stated by the participant, in Portuguese (the language in which interviews were conducted) 2. Barriers (RQ2) Coded excerpts supporting RQ2 (barriers that hinder PS), with the same column structure as the Enablers sheet (Barriers replaces Enablers as the code column). 3. Consequences (RQ3) Coded excerpts supporting RQ3 (consequences of PS and its absence). Column Description Participant Participant ID (P1–P12) Type Valence of the consequence (Positive / Negative) Level Level of impact (Organizational, Individual, Innovation) Consequences Consequence code assigned to the excerpt Quote English translation of the excerpt original quote in Portuguese Original excerpt in Portuguese 4. Codebook The full codebook used in the thematic analysis, organized in three sections (Enablers, Barriers, Consequences). 5. Script The semi-structured interview script used in data collection, including the informed consent form and the full set of interview questions (mapped to RQ1–RQ3, as described in Table 1 of the paper). Translation process All interviews were conducted in Portuguese, the native language of the participants, to reduce linguistic and cultural bias. Excerpts reported in the paper are presented in English for accessibility to an international audience. To support verification of translation fidelity, the original Portuguese excerpt is provided alongside its English translation for every coded excerpt in the Enablers, Barriers, and Consequences sheets. Translations were produced and manually reviewed by the research team. Data collection and ethics Interviews were conducted online (40–60 minutes), audio-recorded with participants' consent, transcribed using Tactiq, and manually reviewed for accuracy. All transcripts were anonymized prior to analysis. This study was approved by the Research Ethics Committee (protocol number withheld from this preprint for blind review and will be disclosed in the camera-ready version), following the ethical guidelines established by the National Health Council Resolution.