Review Papers Against Their Own Results

 Quick Explanation Copied

 Long Explanation

What the Paper Reports — and Whether the Numbers Hold Together

The study searches the Web of Science Core Collection (June 23, 2025) with a stated Boolean query, applies PRISMA-style screening, and analyzes 629 articles from 188 sources using Bibliometrix/Biblioshiny and VOSviewer Methods and PRISMA pipeline. Descriptive outputs are internally consistent where cross-checkable: 629 articles ÷ 188 sources ≈ 3.3 articles/source, and the top-10 journals account for 214 articles (214/629 ≈ 34.0%, vs. the paper's claim of ~30%) Leading journal rankings.

Critical red flag — the PRISMA accounting does not close. The text states that deduplication excluded 629 records, screening eliminated 99, and full-text eligibility removed 728. If the final set is 629 articles, the identification stage would need to have yielded 629+629+99+728 = 2,085 records; the paper never reports this initial count, and the reuse of "629" for both exclusions and the final set is either a coincidence or a reporting error. No PRISMA flow numbers are given for identification, so the flow diagram's arithmetic cannot be verified — a direct violation of PRISMA's transparency purpose Inconsistent PRISMA exclusion counts.

Substantive Findings

Geographic analysis shows the US leading (125 articles, 18,267 total citations) and China second (110 articles, 3,102 citations, 32.7% multi-country publication rate); Germany and Canada show the highest collaboration shares (41.4% and 46.2% MCP) Country contribution and citation data. Note the citation asymmetry: US citations per article (18,267/125 ≈ 146) are ~5.6× China's (3,102/110 ≈ 28), which the paper attributes to dominance without addressing citation-window or self-citation confounds. Keyword co-occurrence (threshold ≥7) yielded 228 keywords clustering around purchase intention, social commerce, trust/risk, and data-driven methods Keyword co-occurrence clusters.

Interpretive overreach: the paper's claim that Western scholars use TAM/UTAUT while Asian research draws on collectivist frameworks is asserted from a handful of conceptual citations (e.g., Hofstede), not from any within-dataset regional analysis — a correlation asserted at field level without corresponding evidence Regional theoretical framing claim.

Known and Unknown Limitations

The paper itself contains no limitations section. Identifiable blind spots (BGPT assessment): single-database coverage (Scopus, no publications in the social-science citation index (SSCI), EBSCO excluded); English-only restriction excluding non-English scholarship; WoS citation-indexing lag depressing 2024–2025 citation counts (explaining why 2025-output journals like Cogent have TC=14 vs. IJEC's 4,464 — an age confound the ranking never adjusts for); and no qualitative synthesis of actual behavioral mechanisms. Growth attribution to COVID-19 is plausible but untested against the retrieved dataset's own time series, and the narrative around the 2020–2021 peak is truncated mid-sentence ("peaking at 85 articles in …"), suggesting an incomplete draft Truncated production-growth narrative.

Bottom line: a useful descriptive map with a fully reproducible search query and clear tooling, but marred by an unverifiable PRISMA flow, an unexplained exclusion-count coincidence, missing identification-stage numbers, no limitations discussion, and an incomplete manuscript sentence. What would change this assessment: publication of the raw WoS export and full PRISMA counts reconciling 2,085 identified → 629 included.

 BGPT Paper Review

Study Novelty

Bibliometric mappings of e-commerce consumer behavior are common; the 2000–2025 scope and PRISMA framing add incremental, not fundamental, novelty.

Scientific Quality

Internally contradictory PRISMA exclusion counts, unreported identification totals, a truncated results sentence, ~30% vs. 34% journal-share discrepancy, and no limitations section are significant rigor failures despite a reproducible search strategy.

Study Generality

The mapping covers a broad interdisciplinary domain but is bounded by one database, English language, and business/management framing.

Study Usefulness

Journal, country, and keyword-cluster rankings offer practical orientation for researchers entering the field; limited causal or mechanistic insight.

Study Reproducibility

The Boolean query, date, and tools are specified, but missing identification counts, unverifiable PRISMA arithmetic, and no data/code availability prevent full replication.

Explanatory Depth

Purely descriptive statistics; regional-paradigm and COVID-19 interpretations are asserted without within-dataset testing.