Using Qualitative Analysis Software to Advance Rigor and Transparency In Evidence Synthesis In Social Work: A Methodological Case Study In the Genai Context

Background and Purpose Rigorous, auditable, and reproducible evidence synthesis is foundational to social work knowledge-building and evidence-based practice — yet the methodological quality of how reviews are conducted remains less examined. As AI tools reshape research practice, their limitations in transparency and accountability demand deliberate methodological responses to ensure synthesis processes remain fully reconstructable. Widely used screening platforms such as Covidence efficiently manage inclusion and exclusion workflows but offer limited capacity for integrated thematic coding and source-level verification — functions essential to rigorous evidence synthesis. This study examines how qualitative analysis software, such as MAXQDA, can serve as a unified synthesis platform spanning screening and coding, using a completed review of AI applications in social work as the methodological case.

Methods A completed review of 127 studies identified through a Web of Science search of JCR-indexed social work journals (2000–2025) served as the methodological case. The original review employed Covidence for screening and a 32-variable extraction instrument. In this methodological study, the identical corpus and codebook were imported into MAXQDA to replicate and extend the synthesis workflow. The platform was used for: (a) document import with bibliographic metadata preserved; (b) title/abstract and full-text screening using document sets and color-coding with memo-linked decision records; and (c) systematic thematic coding across five domains (practice, education, research, policy, and ethics). Workflow differences between platforms were documented systematically across two primary dimensions: screening auditability and coding transparency.

Results Replicating the synthesis workflow in MAXQDA revealed two structural differences with direct implications for rigor and transparency. First, the platform's screening workflow generated memo-linked decision records for each document, enabling full reconstruction of inclusion/exclusion reasoning — a function beyond the scope of Covidence's screening-focused design. Second, thematic coding within the integrated environment allowed simultaneous quantitative and qualitative analysis of the same corpus without data export, and each coded segment remained directly linked to its source text — allowing direct verification of coding decisions against original texts at any stage of the analysis, a capacity that Covidence is not designed to provide. Together, these findings demonstrate that integrated qualitative analysis platforms can substantially strengthen the auditability and analytical transparency of evidence synthesis workflows.

Conclusions and Implications As methodological accountability becomes an increasingly urgent standard in social work research, the choice of synthesis tools carries direct consequences for rigor and reproducibility — particularly when 80% of studies provide no replication materials. This study demonstrates that qualitative analysis software, exemplified by MAXQDA, offers a unified synthesis environment that preserves the full chain of analytical reasoning from raw corpus to interpretive claims, directly addressing concerns about auditability and reproducibility. Beyond procedural rigor, such integrated environments support theory-building by enabling researchers to move iteratively between data and emerging conceptual frameworks — a capacity that extraction-focused platforms are not designed to facilitate. Future work will explore comparable platforms — including NVivo, ATLAS.ti, and open-source alternatives such as Taguette and QualCoder — to assess whether these workflow advantages generalize across integrated qualitative analysis environments.