Investigation Report: Theoretical basis and occurrence of internet fraud victimisation: Based on two systems in decision-making and reasoning (DOI: 10.3389/fpsyg.2023.1087463)

Academic fraud report · Geng Detector

Summary

Verdict: QUESTIONABLE (suspected methodological and editorial carelessness rather than outright fraud). A 2023 systematic review published in Frontiers in Psychology by Shang, Wang, Tian, Zhou, Ma, and Liu contains clear arithmetic inconsistencies in its PRISMA-style article screening description: 223 remaining articles mutate to 233, yet 9 + 233 is reported as yielding a 232-article database, contradicting basic addition. The figure 232 is only consistent with the original 223 + 9 = 232 statement, indicating a typo. Additionally, Table 2 displays unformatted, copy-pasted regression output, including a stray equals sign before p-values. No image-manipulation concerns arise because all figures are author-constructed diagrams with no photographic content. Overall, evidence points to sloppy writing and inadequate proofreading rather than data fabrication. Confidence is moderate; claims should be verified against the published PDF and, if needed, author correspondence. No figures or data fabrication was alleged in this report.

Verdict

🟡 Questionable / suspected editorial negligence. Findings concern textual and tabular inconsistencies in a 2023 Frontiers in Psychology systematic review. No evidence of image manipulation or data fabrication was alleged.

Key findings

Arithmetic inconsistency in the PRISMA-style screening narrative: 223 articles suddenly become 233, yet 9 + 233 is reported to produce a 232-article database (9 + 233 ≠ 232). The intended calculation appears to be 223 + 9 = 232.
Table 2 contains unformatted, copy-pasted regression output, with a stray "=" symbol before p-values ("=0.66 < 0.001") and β values from two independent studies concatenated into a single cell.
No image-based concerns: Figures 1 and 2 are author-constructed diagrams (flowchart and theoretical model); pixel-level analysis is not applicable.

Evidence highlights

Location: Materials and Methods → Article screening section, pages 3–4.
Stated: "For the remaining 223 articles... 9 articles were not included in these 233 articles... We included these 9 articles and 233 articles... a database of 232 articles was built."
Internal contradiction: 9 + 233 = 242, not 232. Cross-checking later exclusion of 215 articles (232 − 215 = 17) suggests the original number should have been 223, not 233.
Location: Table 2, page 5 (Path analysis of heuristic-systematic processing).
Cell content: "Study 1: −0.04, Study 2: −0.17" for β; "=0.66 < 0.001" for p (Vishwanath et al., 2016).
Indicates raw statistical software output pasted without reformatting.
Location: Figures 1 and 2.
Figure captions explicitly state the model drawings come from the author's construction; no photographic or blot content is present.

Notes

DOI: 10.3389/fpsyg.2023.1087463
All quoted numbers (223, 233, 9, 232, 215, 17) are reproduced exactly as they appear in the report.
The identified issues are consistent with low-quality copy-editing and author-side proofreading lapses rather than intentional fraud. They nevertheless fall below the standard expected of a peer-reviewed SCI/SSCI publication.
Image reuse and Photoshop-based checks were not applicable because the article contains only schematic diagrams.
Recommended follow-up: contact the authors for clarification, file a comment on PubPeer, and consider notifying the editorial office if the authors do not respond.