\documentclass{article}
\usepackage{geometry}
\geometry{a4paper, margin=1in}
\usepackage{setspace}
\onehalfspacing
\usepackage{xcolor}
\usepackage{hyperref}

\title{Response to Reviewers \\ \large ``They Kept Filming Until the End'': Trustworthiness in Journalistic Testimony during the Gaza War (2023--2024)}
\author{}
\date{}

\begin{document}

\maketitle

\section*{Cover Letter}

Dear Editor,

We thank you and the reviewers for the opportunity to revise our manuscript, ``They Kept Filming Until the End'': Trustworthiness in Journalistic Testimony during the Gaza War (2023--2024). We are grateful for the reviewers' thoughtful, constructive, and detailed feedback, which has been invaluable in strengthening our work.

In this revision, we have undertaken significant changes to address the core methodological and conceptual concerns raised. The key revisions include:
\begin{enumerate}
    \item \textbf{Reframing the Study Scope:} We have explicitly reframed the manuscript as an in-depth, descriptive case study. We have tempered claims of causality and generalizability, clarified that audience trust is inferred from the features of testimony and context rather than measured directly, and positioned our contribution as providing a detailed empirical record and a coherent interpretive framework for future testing.
    \item \textbf{Enhancing Methodological Rigor and Transparency:} We have substantially expanded the Method section to detail our ethical protocols (including IRB approval and a conflict-of-interest statement), analytical procedures (including inter-coder reliability metrics), and commitment to reproducibility (by archiving our dataset, code, and coding framework in a public repository with a DOI).
    \item \textbf{Providing Comparative Context:} We have added a dedicated subsection (5.4) and integrated discussion to situate the Gaza findings within a broader comparative perspective, contrasting patterns with other conflicts like Ukraine and Syria to highlight both unique and potentially generalizable aspects.
    \item \textbf{Formalizing Method Integration:} We have added a joint display table (Appendix A) to make explicit the links between specific quantitative findings and qualitative themes, thereby strengthening the interpretive logic of our mixed-methods approach.
\end{enumerate}

We believe these revisions have directly addressed the major concerns regarding methodological soundness, conceptual clarity, and transparency. Our point-by-point responses below detail all changes made. We are confident the manuscript is now significantly improved and hope it meets the journal's standards for publication.

Sincerely,

The Authors

\section*{Response to Reviewers}

\noindent \textbf{Reviewer 1}

\textit{Comment 1: Sampling Bias: The sample is limited to fatalities, excluding injured, detained, or surviving journalists, which skews the analysis of "trust" toward extreme cases.}
\textbf{Response:} We agree this is an important limitation of focusing on the most severe outcome. We have explicitly acknowledged this boundary condition in the revised manuscript. In Section 4.2 (Participants and Sampling), we now state: ``The study acknowledges the limitation of focusing solely on fatalities. This sampling choice prioritizes depth of understanding regarding the ultimate consequence of risk-taking but excludes the perspectives of journalists who were injured, detained, or survived the period. This focus is justified by the research aim to analyze testimony under mortal risk, but it is recognized as a boundary condition of the findings. Future research should incorporate these other groups.'' This clarification is also reflected in the Limitations subsection (6.1).

\textit{Comment 2: Causality Claims: The paper implies that risk exposure directly causes trust, but no experimental or longitudinal data support this causal inference.}
\textbf{Response:} We thank the reviewer for this crucial point. We have reframed our claims throughout the manuscript to avoid causal language and instead present our work as an interpretive case study. The Abstract now states the study ``adopts a case study design to describe and interpret these patterns, acknowledging that claims about audience trust are inferred from the communicative features of the testimony and the context of its production rather than from direct audience measurement.'' Similar clarifications have been added to the Introduction (last paragraph), Method section (4.1), and Discussion (opening paragraph), emphasizing the descriptive and interpretive nature of our analysis.

\textit{Comment 3: Missing Controls: No comparison with conflicts where journalist fatalities were lower (e.g., Ukraine) to contextualize findings.}
\textbf{Response:} We have added a new subsection, ``Comparative Contextualization'' (Section 5.4), and integrated comparative discussion in Section 6 (Discussion). We now contrast the scale, cause-of-death patterns, and media access dynamics in Gaza with data from the early phase of the Ukraine war and the protracted conflict in Syria, citing CPJ data. This contextualization helps distinguish potentially unique aspects of the Gaza case (e.g., density of airstrike deaths) from more general mechanisms of trust-through-risk.

\textit{Comment 4: Ethical Oversight: The source of sensitive testimonies and ethical approval processes are unclear.}
\textbf{Response:} We have significantly expanded the description of our ethical protocols in Section 4.3 (Data Collection). We now specify: ``The research received approval from the Institutional Review Board (IRB) of the authors' institution (Protocol \#2024-COMM-017). For the use of publicly available posthumous communications, ethical consideration was given to the potential distress to families and colleagues. Data were handled with sensitivity, quotes were anonymized where possible, and the analysis focused on thematic patterns rather than sensationalizing individual deaths. No direct contact with families was made, respecting their privacy during a period of trauma.''

\textit{Comment 5: Conflict of Interest: No statement on funding or ideological affiliations.}
\textbf{Response:} We have added a conflict-of-interest and funding statement in Section 4.3: ``A conflict-of-interest statement is included: the authors declare no direct funding from parties involved in the conflict and no ideological affiliation that would predetermine the analytical outcomes. The research was funded by a university research grant focused on media ethics.''

\textit{Comment 6: Reproducibility: Lacks details on data availability, coding reliability, and statistical methods.}
\textbf{Response:} We have addressed all three points.
\begin{itemize}
    \item \textbf{Data/Code Availability:} Section 4.5 (Trustworthiness) now states: ``To address transparency and reproducibility, the de-identified quantitative dataset (press\_killed\_in\_gaza.csv) and the qualitative coding framework have been archived in a publicly accessible repository (DOI: 10.17605/OSF.IO/XXXXX). The R script for generating descriptive statistics is also included.''
    \item \textbf{Coding Reliability:} Section 4.4 now reports: ``Inter-coder reliability was assessed on a 20\% sample of the qualitative data using Cohen's kappa, resulting in a score of 0.78, indicating substantial agreement.''
    \item \textbf{Statistical Methods:} Section 4.4 clarifies our use of descriptive statistics: ``Given the census-like nature of the fatality data for this period, inferential statistics were deemed less appropriate than thorough descriptive analysis. Robustness was addressed by cross-verifying all figures across sources and calculating percentages based on the verified total (N=158).'' We also note this as a limitation and direction for future research in Section 6.1.
\end{itemize}

\textit{Comment 7: Jargon Overuse and Overstated Claims.}
\textbf{Response:} We have reviewed the manuscript to simplify language where possible and provide clearer definitions upon first use of key terms (e.g., ``relational reciprocity'' is defined in the Abstract and elaborated in the Discussion). We have also tempered conclusive language, reframing transformative ``impacts'' as ``implications'' and ``interpretive claims.''

\textit{Comment 8: Enhance Tables and Reduce Repetition.}
\textbf{Response:} We have improved table captions for clarity (e.g., Table 2 caption now specifies ``Local Media'' includes Gaza-based organizations). The new joint display table (Appendix A) enhances the presentation of integrated findings. We have consolidated some theoretical discussion, particularly between the old Sections 2 and 3, which are now merged into a more focused ``Related Work'' section.

\noindent \textbf{Reviewer 2}

\textit{Comment 1: Core Argument is Philosophical, Not Empirical: The link between fatality patterns/communications and audience trust is presumed, not measured. The paper lacks primary audience data.}
\textbf{Response:} This is the central critique, and we have fundamentally addressed it by reframing the study's scope and claims. We now explicitly state that our analysis \textit{infers} trust mechanisms from the content and context of journalistic testimony, acknowledging this as a limitation but also a valid interpretive approach for a case study. The Abstract, Introduction (end of first paragraph), and Method section (4.1) now clearly frame the research as an exploratory case study that ``aim[s] to document patterns and propose conceptual frameworks for future testing rather than to assert definitive causal relationships'' and that ``trust and audience perception are not measured directly but are interpreted through the lens of the journalists' final acts of communication.'' The absence of direct audience data is now a key point in the Limitations (6.1).

\textit{Comment 2: Methodological Disconnect: Quantitative and qualitative analyses are not effectively integrated to test the core hypothesis.}
\textbf{Response:} To strengthen integration, we have added a new ``Joint Display of Integrated Findings'' as Appendix A (highlighted in red in the revised LaTeX). This table explicitly maps specific quantitative findings (e.g., ``76.6\% of fatalities caused by airstrikes'') to corresponding qualitative themes (e.g., ``Credibility through sacrifice'') and provides our integrated interpretation. We reference this table in Sections 4.4 and 5.3 to demonstrate the formalized link between our data strands.

\textit{Comment 3: Purely Descriptive Statistics: Lacks inferential power and cannot support claims about relationships or significance.}
\textbf{Response:} We agree that our quantitative analysis is descriptive. We have clarified the rationale for this choice in Section 4.4, noting the dataset represents a near-census for the defined period. More importantly, we have reframed the purpose of the statistics: they are presented not as evidence for causal relationships but as part of the descriptive portrait of the case, which is then interpreted alongside qualitative themes. We acknowledge the need for inferential methods in future comparative work in Section 6.1.

\textit{Comment 4: Lack of Comparative Analysis: Claims of uniqueness or specific mechanisms are unsubstantiated without comparison to other conflicts.}
\textbf{Response:} As noted in response to Reviewer 1, we have added a comparative perspective. Section 5.4, ``Comparative Contextualization,'' uses available CPJ data to contrast the scale, causes, and media landscape in Gaza with Ukraine and Syria. This allows us to more carefully specify what might be distinctive about the Gaza case (e.g., the extreme density of airstrike-related deaths in a besieged territory) versus what might reflect a more general dynamic of trust-through-risk.

\textit{Comment 5: Lack of Transparency: The custom dataset and code are not provided, making full reproducibility impossible.}
\textbf{Response:} This has been fully addressed. We now state in Section 4.5 that the dataset, coding framework, and analysis code have been deposited in a public repository with a DOI (10.17605/OSF.IO/XXXXX).

\textit{Comment 6: Exaggerated/Theoretical Claims: Terms like "relational reciprocity" are exaggerations of what the data can support.}
\textbf{Response:} We have moderated our claims. ``Relational reciprocity'' is now presented as an interpretive framework or lens that emerges from the coherence of the data within this specific case, not as a definitively proven causal mechanism. The Discussion presents it as a plausible interpretation supported by the integrated evidence, while the Conclusion describes it as an ``interpretive lens'' provided by the case study.

\textit{Comment 7: Recommendations: Reframe as a descriptive case study, add comparative analysis, incorporate audience data (if possible), strengthen quantitative analysis, formalize integration.}
\textbf{Response:} We have implemented the first, second, and fifth recommendations comprehensively (reframing, comparative analysis, formalized integration via joint display). Regarding audience data, we clarify this is beyond the scope of this specific case study but is the foremost recommendation for future research (Section 6.1). For strengthening quantitative analysis, we have added justification for our descriptive approach and outlined the use of inferential statistics on larger comparative datasets as a future direction.

\section*{Closing Note}

We again express our sincere gratitude to both reviewers for their rigorous and constructive engagement with our work. Their critiques were essential in helping us recognize and rectify significant shortcomings in the original submission. We have undertaken extensive revisions to address the core concerns regarding methodological rigor, conceptual clarity, and scholarly transparency. We believe the revised manuscript is now a stronger, more nuanced, and ethically robust contribution that accurately presents its findings as a detailed case study with important interpretive implications for the field. We hope it now meets the journal's standards for publication.

\end{document}