Abstract:

While past pro-social research has focused on charitable donations to humans, we know little about charitable assistance for stray animals. However, investigating the factors that promote human’s assistance to stray animals is of great practical importance. Currently, an increasing number of organizations and platforms are involving themselves in rescuing stray animals. When these organizations and platforms present animal rescue information, the information ad contains both animal type and spatial distance. Therefore, to address this research gap, this paper aims to study how animal type and spatial distance jointly influence consumers’ willingness to rescue stray animals, as well as the mechanisms and boundary conditions.

We propose a novel “far dog, near cat” effect. Specifically, we predict that under the near spatial distance, rescuing cat (versus rescuing dog) increases consumers’ rescuing willingness, whereas under the far spatial distance, rescuing dog (versus rescuing cat) increases consumers’ rescuing willingness. To test this effect, we conducted a total of nine experiments (N = 2848), including one implicit association test, one field experiment, one laboratory experiment, and six online experiments in different scenarios. We determined sample size of each experiment using G*power calculator.

Overall, this study found that cats were more compatible with proximal spatial distance, while dogs were more compatible with distal spatial distance (Experiment 1a, 1b). Therefore, presenting a stray cat (vs. a stray dog) in the proximal spatial-distance ad triggered consumers’ higher willingness to rescue the animal, while presenting a stray dog (vs. a stray cat) in the distal spatial-distance ad triggered consumers’ higher willingness to rescue the animal (Experiments 2-5). We further found that processing fluency mediated the “far dog, near cat” effect (Experiments 4-5). In addition, we found that the “far dog, near cat” effect was moderated by consumers’ thinking styles such that the "far dog, near cat" effect was evident when consumers adopted affective thinking style and disappeared when consumers adopted cognitive thinking style (Experiment 6).

Study 1a recruited 188 university students (76.1% female, M_age = 23.31 years, SD = 2.43 years) and conducted a 2 (animal type: cat vs. dog) × 2 (spatial distance: near vs. far) repeated-measures ANOVA with average response time as the dependent variable. The results showed a significant interaction effect, F(1, 186) = 16.08, p < 0.001, η²_p = 0.080. Further simple effects analysis revealed that participants responded faster to words indicating near spatial distance when viewing cats compared to dogs (M_cat = 819.86 ms, SD = 182.05 vs. M_dog = 908.99 ms, SD = 279.29), F(1, 186) = 6.72, p = 0.010, η²_p = 0.035. Additionally, participants responded faster to words indicating far spatial distance when viewing dogs compared to cats (M_cat = 961.97 ms, SD = 362.12 vs. M_dog = 862.05 ms, SD = 294.54), F(1, 186) = 4.31, p = 0.039, η²_p = 0.023. A similar analysis with average accuracy rate as the dependent variable also showed a significant interaction effect, F(1, 186) = 4.75, p = 0.031, η²_p = 0.025. Further simple effects analysis showed that compared to seeing cats, participants had a higher accuracy rate for far spatial distance words when seeing dogs (M_cat = 0.94 (94%), SD = 0.12 vs. M_dog = 0.98 (98%), SD = 0.07), F(1, 186) = 4.68, p = 0.032, η²_p = 0.025; whereas the accuracy rate for near spatial distance words was higher when seeing cats compared to dogs, but this difference was not significant (M_cat = 0.98 (98%), SD = 0.07 vs. M_dog = 0.97 (97%), SD = 0.08), F(1, 186) = 0.58, p = 0.448.

Study 1b involved 200 participants from the Credamo survey platform (M_age = 30.14 years, SD = 8.96 years, 68.5% female). The results indicated that participants who saw cat-related content chose images with closer spatial distances compared to those who saw dog-related content (M_cat= 1.24, SD = 0.90 vs. M_dog= 5.23, SD = 4.31), F(1, 198)= 82.26, p < 0.001, η²_p = 0.294. Similarly, the chosen images of participants in the cat group showed closer spatial distances between people and animals compared to those chosen by the dog group (M_cat= 1.63, SD = 0.96 vs. M_dog= 3.17, SD = 1.45), F(1, 198)= 78.41, p < 0.001, η²_p = 0.284.。

Study 2 had 310 participants (M_age = 20.51 years, SD = 1.81 years, 53.2% female) and conducted a 2 (animal type: stray cat vs. stray dog) × 2 (spatial distance: near vs. far) ANOVA, revealing a significant interaction effect on the willingness to participate in animal rescue activities, F(1, 306) = 8.57, p < 0.001, η²_p = 0.027. Further simple effects analysis indicated that at a close spatial distance, participants were significantly more willing to spend time on animal rescue activities for stray cats than for stray dogs (M_dog = 52.12, SD = 34.93 vs. M_cat = 63.28, SD = 38.87), F(1, 306) = 4.25, p = 0.040, η²_p = 0.014. Conversely, at a far spatial distance, participants' willingness to spend time on animal rescue activities was significantly higher for stray dogs than for stray cats (M_dog = 63.12, SD = 31.94 vs. M_cat = 51.04, SD = 32.65), F(1, 306) = 4.33, p = 0.038, η²_p = 0.014. When willingness to purchase animal food was used as the dependent variable, a 2 (animal type: cats vs. dogs) × 2 (spatial distance: close vs. far) factorial between-subjects analysis of variance showed a significant interaction effect, F (1, 306) = 10.88, p = 0.001, η²_p = 0.034. Further simple effects analysis revealed that at a close spatial distance, participants' willingness to purchase food was significantly higher for stray cats than for stray dogs (M_dog = 4.39, SD = 1.24 vs. M_cat = 4.79, SD = 1.14), F(1, 306) = 4.45, p = 0.036, η²_p = 0.014. However, at a far spatial distance, participants' willingness to purchase food was significantly higher for stray dogs than for stray cats (M_dog = 4.89, SD = 1.15 vs. M_cat = 4.37, SD = 1.39), F(1, 306) = 6.47, p = 0.011, η²_p = 0.021.

Study 3 recruited 317 participants (M_age = 31.45 years, SD = 8.48 years; 56.5% female) and used willingness to help as the dependent variable in a 2 (type of animal: stray cats vs. stray dogs) × 2 (spatial distance: near vs. far) factorial between-subjects ANOVA. The results indicated a significant interaction effect, F(1, 313) = 33.29, p < 0.001, η²_p = 0.096. Further simple effects analysis revealed that at a near spatial distance, participants' willingness to help stray cats was significantly higher than their willingness to help stray dogs (M_dog = 5.33, SD = 1.00 vs. M_cat = 5.79, SD = 0.98), F(1, 313) = 7.30, p = 0.007, η²_p = 0.023. Conversely, at a far spatial distance, participants' willingness to help stray dogs was significantly higher than their willingness to help stray cats (M_dog = 5.79, SD = 1.09 vs. M_cat = 4.87, SD = 1.16), F(1, 313) = 7.30, p < 0.001, η²_p = 0.087.

As a supplementary experiment to Experiment 3, 306 participants took part in study 3S on the Credamo platform (M_age = 30.95 years, SD = 8.01 years, 71.2% female). This experiment also focused on the willingness to adopt as the dependent variable, using the same 2 (type of animal: stray cats vs. stray dogs) × 2 (spatial distance: near vs. far) between-subjects factorial design. The results showed a significant interaction effect, F(1, 302) = 13.48, p < 0.001, η²_p = 0.043. Further analysis of simple effects indicated that at a near spatial distance, participants' willingness to adopt stray cats was significantly higher than their willingness to adopt stray dogs (M_dog = 5.67, SD = 0.84 vs. M_cat = 5.97, SD = 0.63), F(1, 302) = 4.13, p = 0.043, η²_p = 0.013. However, at a far spatial distance, participants' willingness to adopt stray dogs was significantly higher than their willingness to adopt stray cats (M_dog = 6.08, SD = 0.61 vs. M_cat = 5.65, SD = 1.20), F(1, 302) = 10.17, p = 0.002, η²_p = 0.033.

Study 4 was a pre-registered study involving 300 college students (M_age = 22.94 years, SD = 3.34 years, 64.7% female). It utilized a 2 (type of animal: stray cats vs. stray dogs) × 2 (spatial distance: near vs. far) between-subjects factorial design, with donation amount as the dependent variable. The results showed a significant interaction effect between the type of animal and spatial distance, F(1, 296) = 9.51, p = 0.002, η²_p = 0.031. Further simple effects analysis revealed that at a close spatial distance, participants donated significantly more to stray cats than to stray dogs (M_dog = 45.70, SD = 28.43 vs. M_cat = 56.93, SD = 34.50), F(1, 296) = 4.79, p = 0.029, η²_p = 0.016. Conversely, at a far spatial distance, donations were significantly higher for stray dogs than for stray cats (M_dog = 52.47, SD = 32.15 vs. M_cat = 41.54, SD = 29.12), F(1, 296) = 4.72, p = 0.031, η²_p = 0.016. Additionally, using donation amount as the dependent variable, type of animal as the independent variable, spatial distance as a moderator, and processing fluency as a mediator, a moderated mediation analysis was conducted using PROCESS (Model 8, 5,000 bootstraps; Hayes, 2018). The results indicated that processing fluency mediated the interaction between animal type and spatial distance on the donation amount (indirect effect = -3.64, SE = 1.79, 95% CI = [-7.6767, -0.6922], not including 0). Further analysis showed that the indirect effect of processing fluency was significant at a close spatial distance (indirect effect = 1.86, SE = 1.03, 95% CI = [0.2001, 4.2033], not including 0) and also significant at a far spatial distance (indirect effect = -1.78, SE = 1.13, 95% CI = [-4.4359, -0.0556], not including 0).

Study 4S, serving as a supplementary study to study 4, had 280 participants (M_age = 41.84 years, SD = 15.64 years, 58.9% female). The study conducted a 2 (animal type: stray cat vs. stray dog) × 2 (spatial distance: near vs. far) analysis of variance with the intention to help as the dependent variable. The results showed a significant interaction F (1, 276) = 9.42, p = 0.002, η²_p = 0.033. Further analysis of simple effects revealed that at a closer spatial distance, participants were more willing to assist stray cats over stray dogs (M_dog = 4.32, SD = 1.47 vs. M_cat = 4.93, SD = 1.64), F (1, 276) = 3.87, p = 0.050, η²_p = 0.014; whereas at a farther spatial distance, the preference shifted towards helping stray dogs rather than cats (M_dog = 4.94, SD= 1.58 vs. M_cat = 4.30, SD = 1.84), F (1, 280) = 5.87, p = 0.016, η²_p = 0.021. Additionally, with the willingness to help as the dependent variable, type of animal as the independent variable, spatial distance as a moderating variable, and processing fluency as a mediating variable, a moderated mediation analysis was conducted using PROCESS (Model 8, 5,000 bootstraps; Hayes, 2018). The results showed that processing fluency mediates the effect of the interaction between type of animal and spatial distance on the willingness to help (indirect effect = 0.13, SE = 0.08, 95% CI = [0.0036, 0.3206], not including 0).

Study 5 involved 308 participants (M_age = 30.95 years, SD = 10.53 years, 57.1% female) using adoption intention as the dependent variable, and conducted a 2 (animal type: stray cats vs. stray dogs) × 2 (spatial distance: near vs. far) between-subjects factorial ANOVA. The results showed a significant interaction effect, F(1, 304) = 51.49, p < 0.001, η²_p = 0.145. Further simple effects analysis revealed that at a close spatial distance, participants' willingness to adopt stray cats was significantly higher than for stray dogs (M_dog = 5.44, SD = 1.20 vs. M_cat = 6.25, SD = 0.65), F(1, 304) = 30.26, p < 0.001, η²_p = 0.091; whereas at a far spatial distance, participants' willingness to adopt stray dogs was significantly higher than for stray cats (M_dog = 6.25, SD = 0.65 vs. M_cat = 5.46, SD = 0.98), F(1, 304) = 21.68, p < 0.001, η²_p = 0.067. Additionally, using adoption willingness as the dependent variable, type of animal as the independent variable, processing fluency as the mediator, and spatial distance as the moderator, a moderated mediation analysis was conducted using PROCESS (Model 8, 5,000 bootstraps; Hayes, 2018). The results indicated that processing fluency mediated the interaction between animal type and spatial distance on adoption willingness (indirect effect = -0.24, SE = 0.10, 95% CI = [-0.4701, -0.0745], not including 0). The mediating effect of processing fluency was significant at a close spatial distance (indirect effect = 0.10, SE = 0.06, 95% CI = [0.0027, 0.2320], not including 0) and also significant at a far spatial distance (indirect effect = -0.14, SE = 0.07, 95% CI = [-0.2992, -0.0266], not including 0).

Study 6 aimed to examine the moderation of thinking styles, recruiting a total of 639 participants (M_age = 31.29 years, SD = 8.77 years, 65.9% female). The study used adoption intention as the dependent variable to conduct a 2 (animal type: stray cats vs. stray dogs) × 2 (spatial distance: near vs. far) × 2 (thinking style: affective vs. cognitive) three-factor between-subjects ANOVA. The results indicated a significant three-way interaction, F(1, 631) = 23.26, p < 0.001, η²_p = 0.036. Further simple effect analyses revealed that, after priming a cognitive thinking style, the interaction between animal type and spatial distance was not significant, F(1, 631) = 1.56, p = 0.180. When the spatial distance was near, participants' adoption intentions for stray cats and dogs were similar (M_dog = 5.66, SD = 1.28 vs. M_cat = 5.92, SD = 0.80), F(1, 631) = 2.98, p = 0.085. Similarly, when the spatial distance was far, participants' adoption intentions were also similar for both animal types (M_dog = 5.60, SD = 1.02 vs. M_cat = 5.58, SD = 0.92), F(1, 631) = 0.03, p = 0.865. After priming an affective thinking style, the interaction between animal type and spatial distance was significant, F(1, 631) = 57.50, p < 0.001, η²_p = 0.083. Specifically, when the spatial distance was near, participants' adoption intention for stray cats was significantly higher than for stray dogs (M_dog = 5.60, SD = 0.90 vs. M_cat = 6.25, SD = 0.56), F(1, 631) = 19.02, p < 0.001, η²_p = 0.029. Conversely, when the spatial distance was far, participants' adoption intention for stray dogs was significantly higher than for stray cats (M_dog = 6.30, SD = 0.39 vs. M_cat = 5.24, SD = 1.22), F(1, 631) = 51.37, p < 0.001, η²_p = 0.075. Additionally, with adoption intention as the dependent variable, animal type as the independent variable, processing fluency as the mediating variable, spatial distance as the first-level moderator, and thinking style as the second-level moderator, a moderated mediation analysis was conducted using PROCESS (Model 12, 5000 bootstraps; Hayes, 2018). The results showed that processing fluency mediated the effect of the interaction among animal type, spatial distance, and thinking style on adoption intention (indirect effect = -0.16, SE = 0.08, 95% CI = [-0.3334, -0.0157], not including 0). Under a cognitive thinking style: at near spatial distances, the mediating effect of processing fluency was not significant (indirect effect = -0.01, SE = 0.03, 95% CI = [-0.0638, 0.0712], including 0); at far spatial distances, the mediating effect was also not significant (indirect effect = -0.001, SE = 0.03, 95% CI = [-0.0680, 0.0668], including 0). Under an affective thinking style: at near spatial distances, the mediating effect of processing fluency was significant (indirect effect = 0.08, SE = 0.04, 95% CI = [0.0056, 0.1642], not including 0); at far spatial distances, the mediating effect was also significant (indirect effect = -0.08, SE = 0.05, 95% CI = [-0.1879, -0.0007], not including 0).

This paper has significant theoretical contributions and practical implications. Theoretically, this study focuses on stray animals as a novel object of charitable donations and builds the implicit linkage between animal type and spatial distance. Also, this study identifies the “far dog, near cat” effect in stray animal rescue, adding to past pro-social literature in general and donation literature in particular. Practically, the “far dog, near cat” effect we identified in this paper can guide charitable organizations how to present animal-rescue information appropriately.

Key words: The “far dog, near cat” effect, animal rescue, processing fluency