eng Rovedar Small Animal Advances 2821-2363 2026-09-30 5 3 30 41 10.58803/saa.v5i3.60 62 Paa Kobina Turkson 0 https://orcid.org/0000-0002-5319-2021 David Gyau Sarpong 0 https://orcid.org/0009-0006-9125-0751 School of Veterinary Medicine, University of Ghana, Legon, Accra, Ghana Introduction: Rabies is a major health issue in Ghana; therefore, its control and prevention rely on the knowledge, attitudes, practices, and perceptions (KAPP) of pet keepers. The present study aimed to identify key KAPP factors for rabies among pet owners using principal component analysis (PCA) and to develop predictive models for KAPP. Materials and methods: This study surveyed 315 pet keepers at a veterinary hospital in Accra, Ghana, using a KAPP questionnaire about rabies from August 2024 to August 2025. The questionnaire measured knowledge (11 items), attitudes (4 items), practices (5 items), and perceptions (13 items) of rabies. Respondents were then categorized as having high or low knowledge, positive or negative attitudes, good or poor practices, and favorable or unfavorable perceptions, depending on whether their scores fell above or below the established threshold. The PCA grouped correlated questionnaire items into broader, independent themes called components. Bartlett's and KMO analyses confirmed data suitability for PCA, Cronbach's alpha assessed questionnaire reliability, binary logistic regression analyzed the predictors of KAPP outcomes, and Cox and Snell and Nagelkerke R-squared evaluated model explanatory power and fit. Results: For knowledge, 15 indicators (11 knowledge questions plus rabies-related demographic variables) were entered into PCA, and the analysis grouped them into 6 distinct themes that together explained 54% of the total variation in responses. Eight attitude indicators (including four attitude questions and rabies-related demographics) were entered into PCA, which grouped them into four themes explaining 63% of response variation. For practices, 9 indicators were reduced to 4 components, explaining 60% of the total variance, and for perceptions, 17 indicators were reduced to 7 components, explaining 60% of the total variance. The Cox and Snell values (0.71-0.75) indicated that the models explained 75% of the variation in KAPP. The Nagelkerke values (0.96-1.00) indicated a strong model fit, meaning the predictors accounted for nearly all the variability in the outcomes. Conclusion: The present study identified important indicators needing prioritization to improve KAPP of pet owners and to help enhance rabies prevention and control in Accra, Ghana. https://saa.rovedar.com/index.php/SAA/article/view/60 Attitude Knowledge Perceptions Practices Principal component analysis Rabies