Fig 1.
Illustration of the three cat species in this study.
Name, scientific name, shoulder height and body weight of caracal, cheetah and lion.
Fig 2.
Example of an image set used in the survey.
(A) control image of the species against a white background, (B) species being petted by a human, (C) species being walked by a human, (D) species in the wild. Faces have been blurred for publication purposes.
Fig 3.
Percentage of respondents who would like to be photographed with wild cat species in relation to the type of image shown.
Compared to the control image, respondents who were shown an image of a wild cat species being petted by a human were significantly more likely to have a desire to be photographed with wild cats.
Table 1.
Test statistics and effect sizes for variables in the model to assess the respondents’ intention to be photographed with wild cat species.
Parameter estimate (β), standard error (SE), Wald statistics (Wald), degrees of freedom (df), level of significance (p), cumulative odds ratio (Exp (β)) and 95% confidence interval of the cumulative odds ratio (95% CI Exp (β)) for variables in the model. The answer options no, maybe, yes were coded as 1, 2, 3.
Fig 4.
Percentage of respondents who perceived wild cat species as suitable to be photographed with tourists in relation to the type of image shown.
Compared to the control image, respondents who were shown an image of a wild cat species being petted by a human were significantly more likely to perceive wild cats as suitable to be photographed with tourists.
Table 2.
Test statistics and effect sizes for variables in the model to assess the respondents’ perception of the suitability of wild cat species to be photographed with tourists.
Parameter estimate (β), standard error (SE), Wald statistics (Wald), degrees of freedom (df), level of significance (p), cumulative odds ratio (Exp (β)) and 95% confidence interval of the cumulative odds ratio (95% CI Exp (β)) for variables in the model. The answer options unsuitable, not suitable but also not unsuitable, suitable were coded as 1, 2, 3.
Table 3.
Test statistics and effect sizes for variables in the model to assess the respondents’ intention to walk with wild cat species.
Parameter estimate (β), standard error (SE), Wald statistics (Wald), degrees of freedom (df), level of significance (p), cumulative odds ratio (Exp (β)) and 95% confidence interval of the cumulative odds ratio (95% CI Exp (β)) for variables in the model. The answer options no, maybe, yes were coded as 1, 2, 3.
Fig 5.
Percentage of respondents who perceived wild cat species as suitable to walk with tourists in relation to the type of image shown.
Compared to the control image, respondents who were shown an image of a wild cat species being petted or walked by a human were significantly more likely to perceive wild cat species as suitable to walk with tourists.
Table 4.
Test statistics and effect sizes for variables in the model to assess the respondents’ perception of the suitability of wild cat species to walk with tourists.
Parameter estimate (β), standard error (SE), Wald statistics (Wald), degrees of freedom (df), level of significance (p), cumulative odds ratio (Exp (β)) and 95% confidence interval of the cumulative odds ratio (95% CI Exp (β)) for variables in the model. The answer options unsuitable, not suitable but also not unsuitable, suitable were coded as 1, 2, 3.
Fig 6.
Percentage of respondents who would like to pet a wild cat species in relation to the type of image shown.
Compared to the control image, respondents who were shown an image of a wild cat species being petted by a human were significantly more likely to have a desire to pet wild cat species.
Table 5.
Test statistics and effect sizes for variables in the model to assess the respondents’ intention to pet a wild cat species.
Parameter estimate (β), standard error (SE), Wald statistics (Wald), degrees of freedom (df), level of significance (p), cumulative odds ratio (Exp (β)) and 95% confidence interval of the cumulative odds ratio (95% CI Exp (β)) for variables in the model. The answer options no, maybe, yes were coded as 1, 2, 3.
Fig 7.
Percentage of respondents who perceived wild cat species as suitable to be petted by tourists in relation to the type of image shown.
Compared to the control image, respondents who were shown an image of a wild cat species being petted by a human were significantly more likely to perceive wild cat species as suitable to be petted by tourists.
Table 6.
Test statistics and effect sizes for variables in the model to assess the respondents’ perception of the suitability of wild cat species to be petted by tourists.
Parameter estimate (β), standard error (SE), Wald statistics (Wald), degrees of freedom (df), level of significance (p), cumulative odds ratio (Exp (β)) and 95% confidence interval of the cumulative odds ratio (95% CI Exp (β)) for variables in the model. The answer options unsuitable, not suitable but also not unsuitable, suitable were coded as 1, 2, 3.
Table 7.
Test statistics and effect sizes for variables in the model to assess whether respondents perceived a wild cat species as a potentially nice pet.
Parameter estimate (β), standard error (SE), Wald statistics (Wald), degrees of freedom (df), level of significance (p), odds ratio (Exp (β)) and 95% confidence interval of the odds ratio (95% CI Exp (β)) for variables in the model. The answer option no (not a nice pet) was used as the reference category.
Fig 8.
Percentage of respondents who perceived wild cat species as sweet, not sweet but also not dangerous (neutral) or dangerous in relation to the type of image shown.
Compared to the control image, respondents who were shown an image of a wild cat species being petted by a human were significantly less likely to perceive wild cat species as dangerous.
Table 8.
Test statistics and effect sizes for variables in the model to assess the respondents’ perception of the danger wild cat species represent.
Parameter estimate (β), standard error (SE), Wald statistics (Wald), degrees of freedom (df), level of significance (p), odds ratio (Exp (β)) and 95% confidence interval of the odds ratio (95% CI Exp (β)) for variables in the model. The answer option dangerous was used as the reference category.
Fig 9.
Percentage of respondents who perceived wild cat species as sweet or dangerous in relation to their intention to be photographed with, walk with or pet wild cats and their perception of the suitability of wild cats for such tourist activities.
Respondents who expressed a desire to be photographed with, walk with or pet wild cats and/or respondents who thought wild cats are suitable for such activities with tourists more often perceived wild cats as sweet than those who did not.
Table 9.
Test statistics and effect sizes for variables in the model to assess the respondents’ perception of the conservation status of wild cat species.
Parameter estimate (β), standard error (SE), Wald statistics (Wald), degrees of freedom (df), level of significance (p), cumulative odds ratio (Exp (β)) and 95% confidence interval of the cumulative odds ratio (95% CI Exp (β)) for variables in the model. The answer options not endangered, endangered, very endangered were coded as 1, 2, 3.