• https://www.sciencedirect.com/science/article/pii/S0268401221001092/pdfft?md5=4b3c03a17eae5921e34599477f714588&pid=1-s2.0-S0268401221001092-main.pdf

Main idea

• The univaersality of the 'exaggeration' assumption (fake reviews tend to be more exaggerated than authentic ones) doesn't work well in extreme ends of hotel category—luxury and budget.
• With siganling theory, luxury (or budget) hotels have own signals for expected experience. The rating 'posivtive vs negative' on them has prevailing expectation before actual assessment. The most important aspects for fake reviews are confirmatory contexts between signals and expectation for positive ones or disconfirmatory contexts for negative ones.
• These facts make the exaggeration of fake review is not more rather than authentic ones.

Definition

• Luxury hotel (L) vs Budget hotel (B)
• Positive review (P) vs Negative review (N)
• Writing (W) vs Reading (R)
• Authentic review (A) vs Fake review (F)

Model & Conclusion

Model

• Confirmatory contexts
  • Exaggeration in fake reviews will not exceed in authentic reviews on dimensions as follows:
    • (LPWF = LPWA) or (BNWF = BNWA)
    • (LPRF = LPRA) or (BNRF = BNRA)
• Disconfirmatory contexts
  • Exaggeration in fake reviews will not exceed in authentic reviews on dimensions as follows:
    • (LNWF = LNWA) or (BPWF = BPWA)
    • (LNRF = LNRA) or (BPRF = BPRA)
• Comparing confirmatory and disconfirmatory contexts
  • Exaggeration will fall short in confirmatory contexts rather than in disconfirmatory one
    • LNW* > LPW*
    • LNR* > LPR *
    • BPW* > BNW*
    • BPR* > BPR*

Conclusion

• LPW* > LPW* is partially supported from data
• (BPW* > BNW*) and (BPR* > BPR*) are rejected from data