Reducing Speciesism: An Intervention to Change People’s Attitudes and Behavioral Intentions

ABSTRACT

Humans mistreating, exploiting, and abusing animals is a common problem in our society and is related to speciesism. Speciesism refers to when humans assign different worth to beings based on their species. A way used by animal activists to reduce harmful behaviors toward animals involves showing people illustrations of situations in which animals are harmed but where the roles of animals and humans are reversed. It is unclear, however, whether this intervention is successful in reducing speciesist attitudes and behaviors toward animals. As the intervention may increase awareness of the treatment of animals, perspective-taking, and feelings associated with injustice we expected it to reduce speciesism compared with a control condition. In two studies (n_Study1 = 231 and n_Study2 = 399), participants either watched or did not watch a video showing situations in which humans take the role of nonhuman animals and vice versa. Participants’ speciesist attitudes, behavioral intentions, and behavioral responses were measured. Studies 1 and 2 showed that participants in the intervention condition more strongly intended to reduce their (in)directly harmful behavior toward animals than those in the control condition. Feelings associated with injustice mediated this effect (Study 2). There were no effects on speciesist attitude or behavioral responses. The intervention shows promise as people intended to change their behavior. Reasons for why the intervention changed intentions but not attitudes or behavioral responses are discussed.

KEYWORDS:

• Behavior change
• human–animal interaction
• prejudice
• speciesism

“I need to revise my way of living urgently” – anonymous participant

Why do we eat pigs, wear cows, and experiment on dogs when most people claim to love animals? Generally, people condemn animal suffering, yet they still harm animals either directly or indirectly (Rothgerber & Rosenfeld, 2021). Humans use animals for consumption, medical studies, and entertainment. The abuse, exploitation, and mistreatment of animals relate to speciesism (Bègue, 2020; Sollund, 2011). Ryder (2017) define speciesism as “a prejudice or attitude of bias in favor of the interests of members of one’s own species and against those of members of other species” (p. 3). This paper investigates the effectiveness of an intervention, wherein the roles of humans and animals are reversed, on people’s speciesist attitudes and behavioral intentions that are directly or indirectly harmful to animals.

To what extent do existing interventions focusing on creating awareness of how animals are treated have the potential to reduce the harm done to animals? Animal activists use various ways to create awareness of the common practices toward animals (e.g., Buddle et al., 2017; Lu et al., 2013). They may contribute to changing people’s speciesist attitudes and behaviors. However, it is unclear to what extent many of these interventions are successful. One way animal activists aim to create awareness is by sharing videos with thought-provoking art showing a parallel universe wherein the roles of humans and animals are reversed. These illustrations may attract more people to watch than real videos of animal exploitation because the thought-provoking art may be less disturbing and more interesting to watch.

Although numerous experimental studies have investigated the effects of interventions on animal-product consumption (see Bianchi et al., 2018; Grundy et al., 2022; Harguess et al., 2020; Vos et al., 2021), according to our knowledge no studies have focused on reducing speciesist attitudes and behaviors toward animals in general. A few of the studies that investigated the specific speciesist behavior of meat intake also used images of animals in their intervention. Tian et al. (2016) showed that participants were less willing to eat beef when they received a recipe together with a photo of a cow compared with participants who only received the recipe. Zickfeld et al. (2018) displayed pictures of animals in hypothetical meat advertisements. They showed that this increased participants’ empathy for animals and reduced their willingness to eat meat compared with a no-picture control condition. These studies, however, did not focus on testing an existing intervention used by animal activists. Moreover, these studies did not investigate speciesist attitudes or behavior toward animals in general. The current intervention not only focused on animal-product consumption but also on other industries in which animals are used as well.

An intervention video showing a parallel universe could be effective for three reasons. The first is that it may increase awareness by showing how animals are being treated by humans. People have limited knowledge about this (Mayfield et al., 2007; Worsley et al., 2015). Raising awareness by informing people can reduce avoidance behavior and willful ignorance toward how animals are being treated. These two mechanisms prevent people from changing their harmful behaviors toward animals (Rothgerber & Rosenfeld, 2021).

A second reason why the intervention could be effective is that the video may facilitate taking the animals’ viewpoint. Showing the reversed roles of animals and humans leads people to have to switch mentally, which is an aspect of perspective-taking (Healey & Grossman, 2018). Increased empathic feelings for animals via perspective-taking could reduce speciesism (e.g., Zickfeld et al., 2018). Furthermore, perspective-taking has been shown to decrease prejudice (Matera et al., 2021; Shih et al., 2009). As speciesism and other prejudices, such as racism and sexism, are related (Caviola et al., 2018; Jackson, 2019), an intervention focusing on reversing roles between humans and animals could reduce discriminatory behavior toward animals as well. Speciesism and other prejudices share the same psychological mechanism: people’s desire for achieving and maintaining group dominance, termed social dominance orientation (SDO; Caviola et al., 2018; Dhont et al., 2016).

Finally, speciesist attitudes and behaviors may be changed by people experiencing emotional reactions toward observing unjustified suffering. The intervention could be effective because learning about suffering while perspective-taking leads people to experience feelings associated with injustice, such as being angered and alarmed (Stephan & Finlay, 1999). These feelings associated with injustice play an important role in reducing prejudice (Dovidio et al., 2004). Taken together, as the intervention may increase awareness of animal suffering, perspective-taking of the animals, and feelings associated with injustice, it was hypothesized that the intervention would reduce speciesist attitudes and behavioral intentions compared with no intervention. To investigate this, participants either watched or did not watch a video showing humans and animals in reversed roles. Then, speciesist attitudes and behavioral intentions regarding (in)directly harming animals (e.g., consuming animal products, going to the circus) were measured. Also, participants were asked whether they would sign a petition to support animal rights.

Ethical Review

This research was approved by the BMS Ethics Committee of the University of Twente, Netherlands (approval numbers 211440 and 220655). We report all manipulations, measures, and exclusions (if any) in these studies. The materials and data from the studies are available via the Open Science Framework (https://osf.io/q6dvm/).

Study 1

Methods

Participants and Design

To our knowledge, no study using a speciesism intervention has been done before. The study by Zickfeld et al. (2018, Study 1), which is similar to ours, suggests a medium effect size is required. A sensitivity analysis using G*Power 3.1.9.7, with d = 0.50, 95% power, and α = 0.05, showed that 105 participants per group (210 total) are needed for a t-test.

Two hundred and forty-one students participated for credit points. Those who did not finish the survey or finished it more than once were excluded from the analyses, leaving 231 participants (51 males, 175 females, 5 other). Their ages ranged between 17 and 31 years (M = 20.36, SD = 1.94). The independent variable was intervention (yes vs no). The dependent variables were speciesist attitude, behavioral intentions, and behavioral response.

Procedure and Materials

Participants conducted the study online. If they provided informed consent – it was mentioned that some content may cause distress and/or discomfort – the study started. The questionnaire began with questions about demographics (age and gender).

Manipulation

Participants were either assigned to the intervention or the control condition. In the intervention condition, a video was presented which showed a parallel world wherein the roles of humans and animals were reversed (see https://youtu.be/0zzTaPutHww). The video started with a warning that the images could be disturbing to some and that participants were free to stop at any time. Ten participants did not watch the whole video. Excluding these participants from the analyses yielded comparable results, so they were included.

The video presented illustrations of situations in which animals are harmed, but where the humans took the role of the animals and vice versa. As most existing videos vary between 2 and 4 min, we aimed for a similar length. Participants were shown static drawn images of seven categories of different situations (entertainment, experiments, farming, fashion, held home, hunting, and offered as products). For each category, 3 illustrations were shown, resulting in 21 illustrations in total, which were accompanied by sad music. Thirteen illustrations were in black-and-white, while eight were in color. Each illustration was shown for 8 s, which gave participants sufficient time to process them. In total, the video ran for 2:48 min. The illustrations were drawn by artists; they were asked for permission to use their art for this research. An example of an illustration is a crocodile walking around with a purse made of human skin with a human face on it. In the control condition, no video with illustrations was shown. They were not given any other task.

Dependent Variables

After participants either watched or did not watch the video, the dependent variables were measured. First, speciesist attitude was measured using the Speciesism Scale (Caviola et al., 2018). This scale is empirically validated and has high internal consistency and test–retest reliability. It consists of six items (e.g., “Morally, animals always count for less than humans”), with answers given on a 7-point scale that ranges from “strongly disagree” to “strongly agree.”

Next, participants were asked to what extent they intended to change their behavior regarding (in)directly harmful behavior toward animals. For each of the seven categories of the intervention, two items were included, except for the category “offered as products” as it was akin to the other categories (e.g., “I intend to consume meat … ” and “Regarding the cruelty-free cosmetic products I intend to buy … ”). As not all people have pets and/or hunt and the categories regarding pets and hunting focused more on attitudes than on behavioral intentions, these items were not included in the final scale. The results were comparable with these items included. The response scale ranged from 1 to 4: “more than I currently do,” “just as much as I currently do,” “less than I currently do,” “stop doing that altogether,” and a fifth answer possibility, “I do not do that and I would stick to that.” The fifth possibility was added to discriminate participants who couldn’t show a behavioral change toward less speciesism as they already didn’t engage in such behaviors from those who could show a behavioral change toward less speciesism but chose not to.

Then, a behavioral response was measured. Participants were asked to read a fictitious animal rights petition that promoted giving animals the same rights as humans. They were asked whether they wanted to sign it or not. Participants were then asked a control question: “I watched the video attentively” (intervention condition only). This was measured on a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). Five participants indicated to not have watched the video attentively (which is a score below 5). Excluding these participants from the analyses resulted in similar findings, so they were included. Participants were also asked their type of diet (i.e., omnivore, vegetarian, etc.). Finally, they were debriefed.

Results

For attitude and behavioral intentions, outliers were identified. These outliers did not distort the results and were therefore included in the analyses. The number of outliers and tests regarding homogeneity of variances and normality are reported in the online supplemental materials. For all means and standard deviations, see Table 1. For correlations between the variables, see Table 2 and the online supplemental materials.

Table 1. Mean scores and standard deviations for speciesist attitude and behavioral intentions (Study 1).

	Mean (SD)
	Intervention	Control
Speciesist attitudes	2.35 (0.90)	2.48 (0.88)
Behavioral intentions	4.04 (0.54)	3.84 (0.62)

Table 2. Correlations between speciesist attitude, behavioral intentions, and behavioral response (Study 1, n = 231).

	Speciesist attitude	Behavioral intentions	Behavioral response
Speciesist attitude	–
Behavioral intentions	r = –0.63, p < 0.001	–
Behavioral response	rs = –0.41, p < 0.001	rs = 0.37, p < 0.001	–

Note: For attitude, a higher score means more speciesist attitude. For behavioral intentions, a higher score means more reduction of hurtful behaviors. For behavioral response, a higher score means more likely to sign a petition to support animal rights. Pearson correlations (r) were calculated for all relationships, except when behavioral response was included, in which case the nonparametric Spearman’s rho (r_s) was calculated.

Attitude

One item was recoded so that for all items a higher score meant more speciesism. The Cronbach’s alpha was 0.74. An analysis of variance (ANOVA), with intervention (yes vs no) as the independent variable and the mean of the attitude items as the dependent variable, showed no significant difference between the intervention condition (M = 2.35, SD = 0.90) and the control condition (M = 2.48, SD = 0.88) (F_{(1, 229)} = 2.87, p = 0.271, η_p²= 0.005, d = 0.15).

Behavioral Intentions

The Cronbach’s alpha for all the behavioral intention items was 0.72. An ANOVA, with intervention (yes vs no) as the independent variable and the mean of these items as the dependent variable, indicated that participants in the intervention condition more strongly intended to reduce harming animals (M = 3.77, SD = 0.61) than participants in the control condition (M = 3.59, SD = 0.63) (F_{(1, 229)} = 4.85, p = 0.029, η_p²= 0.021, d = 0.29).

To investigate whether participants in the intervention condition intended to change their behavior toward less speciesism as compared with their current behavior, we conducted a one-sample t-test within this condition, testing the mean against the value 2 (which indicated participants intended to show the behavior just as much as they currently did). The analyses showed that for both the intervention (t₍₁₁₄₎ = 31.18, p < 0.001) and the control conditions (t₍₁₁₅₎ = 27.28, p < 0.001) participants intended to reduce their harmful behavior toward animals.

Behavioral Response

A chi-square test, with intervention (yes vs no) as the independent variable and the petition item as the dependent variable (yes vs no), showed no significant association between the intervention and signing the petition, (χ²₍₁₎ = 0.194, p = 0.659): 79.1% of the participants signed the petition in the intervention condition and 76.7% did so in the control condition.

Study 2

The results of study 1 showed that participants in the intervention condition more strongly intended to reduce their harmful behaviors toward animals than those in the control condition. Participants’ speciesist attitude and behavioral response were unaffected. In study 2, we investigated whether the findings of study 1 could be replicated. Furthermore, we examined whether the mechanisms awareness of how animals are treated, perspective-taking, and feelings associated with justice, as mentioned in the introduction, would indeed play a role in the intervention reducing participants’ speciesist intentions.

One limitation of study 1 is that it may be easy to agree to signing a petition. Therefore, in study 2 we added a behavioral response for which agreeing requires more effort. Furthermore, study 1 might not have been powered sufficiently. So, a bigger sample was used in study 2. Finally, it is conceivable that when reporting their behavioral intentions, participants may answer in line with what they think the researcher expects (Kwak et al., 2021). To correct for the possibility that a social desirability bias could inflate the effects of our intervention (see Mathur et al., 2021), we included a social desirability scale.

Methods

Participants and Design

Based on study 1, we expected a small to medium effect size. A power analysis using G*Power 3.1.9.4, with d = 0.35, 95% power, and α = 0.05, showed that a t-test would require 186 participants per group (372 total). We attracted 399 Prolific participants (216 males, 172 females, 11 other), each of whom received £3.75 for participating. Their ages ranged between 18 and 70 years (M = 31.59, SD = 10.78). The independent variable was intervention (yes vs no). The dependent variables were speciesist attitude, behavioral intentions, and behavioral responses. Possible mediating variables were awareness, perspective-taking, and feelings associated with injustice.

Procedure and Materials

The procedure of study 2 was similar to that of study 1, with a few exceptions. First, the four items that were omitted from analyses on the behavioral intentions scale in study 1 were now included. Also, an item was added about participants’ intention to consume eggs. Second, as it may be relatively easy to agree with signing the petition, a behavioral response was added about whether participants would agree to answer five additional questions concerning petitions. If they accepted, they were presented with five filler items.

To investigate possible mediating mechanisms, perspective-taking, feelings associated with injustice, and awareness were measured. Perspective-taking was measured with three items (e.g., “I feel empathy for animals”). Note that we did not measure the extent to which participants took perspective of the animals as shown in the video. We asked them about their current experience with animals in general. Feelings associated with injustice were measured with four items (as in Dovidio et al., 2004; e.g., “I feel alarmed”). Awareness of the common practices of animals was measured with nine items (e.g., “I am aware of the common practices of how animals are being treated in the meat industry”). All items were measured using a 7-point Likert scale. For perspective-taking and awareness, the scale ranged from “strongly disagree” to “strongly agree”; for feelings associated with injustice, it ranged from “not at all” to “extremely.”

Finally, a 12-item version of the Social Desirability Scale of Marlowe-Crowne was included (e.g., “It is sometimes hard for me to go on with my work if I am not encouraged”; Reynolds, 1982; Cronbach’s α = 0.71). Note that Lanz et al. (2021) concluded that social desirability (SD) scales do not clearly measure a bias. Vesely and Klöckner (2020), however, did obtain a small but significant meta-analytic relation between SD scales and environmental behavior.

Results

Four participants indicated to not have watched the video until the end, of which two indicated to not have watched the video attentively. These participants were included in the findings as excluding them yielded comparable results. For attitudes, perspective-taking, and awareness, outliers were identified. These outliers also did not distort the results and were therefore included in the analyses. For the number of outliers and tests regarding homogeneity of variances and normality, see the online supplemental materials. All means and standard deviations are reported in Table 3. For all correlations, see Table 4 and the online supplemental materials.

Table 3. Mean scores and standard deviations for speciesist attitude, behavioral intentions, perspective-taking, injustice, and awareness (Study 2).

	Mean (SD)
	Intervention	Control
Speciesist attitude	2.77 (1.13)	2.58 (1.10)
Behavioral intentions	3.40 (0.70)	3.22 (0.74)
Perspective-taking	5.37 (1.08)	5.51 (1.05)
Injustice	3.43 (1.62)	2.87 (1.55)
Awareness	4.89 (1.16)	4.81 (1.19)

Table 4. Correlations between speciesist attitude, behavioral intentions, behavioral response (signing petition), answering additional questions, perspective-taking, injustice, and awareness (Study 2, n = 399).

	Speciesist attitude	Behavioral intentions	Signing petition	Answering additional questions	Perspective-taking	Injustice	Awareness
Speciesist attitude	–
Behavioral intentions	r = –0.44, p < 0.001	–
Signing petition	rs = –0.56, p < 0.001	rs = 0.39, p < 0.001	–
Answering additional questions	rs = –0.24, p < 0.001	rs = 0.14, p = 0.006	rs = 0.17, p = 0.001	–
Perspective-taking	r = –0.58, p < 0.001	r = 0.38, p < 0.001	rs = 0.45, p < 0.001	rs = 0.26, p < 0.001	–
Injustice	r = –0.30, p < 0.001	r = 0.25, p < 0.001	rs = 0.31, p < 0.001	rs = 0.04, p = 0.43	r = 0.35, p < 0.001	–
Awareness	r = –0.23, p < 0.001	r = 0.29, p < 0.001	rs = 0.24, p < 0.001	rs = 0.09, p = 0.06	r = 0.28, p < 0.001	r = 0.12, p = 0.02	–

Note: For attitude, a higher score means more speciesist attitude. For behavioral intentions, a higher score means more reduction of hurtful behaviors. For behavioral response, a higher score means more likely to sign a petition to support animal rights. Pearson correlations (r) were calculated for all relationships, except when behavioral responses (i.e., signing the petition and answering five additional questions) were included, in which case the nonparametric Spearman’s rho (r_s) was calculated.

Attitude

After recoding the reversed item, the Cronbach’s alpha for all the attitude items was 0.83. An ANOVA with intervention (yes vs no) as independent variable and the mean of the attitude items as dependent variable was not significant (F_{(1, 397)} = 2.83, p = 0.094, η_p²= 0.007, d = 0.17). If anything, participants in the intervention condition had a stronger speciesist attitude (M = 2.77, SD = 1.13) than those in the control condition (M = 2.58, SD = 1.10). When social desirability was added as a covariate, the result of the ANOVA was still nonsignificant (F_{(1, 396)} = 2.54, p = 0.111, η_p²= 0.006, d = 0.17). Social desirability accounted for 0.0086 (0.86%) of the variability of the speciesist attitude.

Behavioral Intentions

The Cronbach’s alpha for all the behavioral intention items was 0.84. An ANOVA with intervention (yes vs no) as the independent variable and the mean of these items as the dependent variable showed a significant effect of intervention (F_{(1, 397)} = 5.97, p = 0.015, η_p²= 0.016, d = 0.25). Participants in the intervention condition more strongly intended to reduce their harmful behavior toward animals (M = 3.40, SD = 0.70) compared with the control condition (M = 3.22, SD = 0.74). This effect remained significant when social desirability was included as a covariate (F_(1,396) = 6.46, p = 0.010, η_p²= 0.016, d = 0.25). Social desirability accounted for 0.007 (0.77%) of the variability of behavioral intentions, so it had almost no effect on intentions to change.

One-sample t-tests within each condition testing the condition mean against the value 2, indicating participants intend to show the behavior just as much as they currently do, showed that for both the intervention (t₍₁₉₈₎= 28.39, p < 0.001) and the control conditions (t₍₁₉₉₎ = 23.31, p < 0.001) participants intended to reduce their harmful behavior toward animals.

Behavioral Responses

Separate chi-square tests with intervention (yes vs no) as the independent variable showed no significant effect of the intervention on signing the petition (χ²₍₁₎ = 0.002, p = 0.964) or on answering additional petition items (χ²₍₁₎ = 0.001, p = 0.975). Most (55.3%) of the participants signed the petition in the intervention condition; 55.5% did so in the control condition. Most (71.36%) agreed to answer additional items in the intervention condition; 71.5% did so in the control condition.

Possible Mediators

The Cronbach’s alpha for all the perspective-taking items was 0.84. An ANOVA with intervention (yes vs no) as independent variable and the mean of these items as dependent variable revealed no significant difference of intervention (F_{(1, 397)} = 1.64, p = 0.201, η_p²= 0.004, d = 0.13) (intervention: M = 5.37, SD = 1.08; control: M = 5.51, SD = 1.05).

Cronbach’s alpha for all the injustice items was 0.91. An ANOVA with intervention (yes vs no) as independent variable and the mean of the injustice items as dependent variable showed that participants in the intervention condition experienced stronger feelings associated with injustice (M = 3.43, SD = 1.62) than the control group (M = 2.87, SD = 1.55) (F_{(1, 397)} = 12.56, p < 0.001, η_p²= 0.031, d = 0.36).

The Cronbach’s alpha for all the awareness items was 0.91. An ANOVA with intervention (yes vs no) as independent variable and the awareness items as dependent variable revealed no significant difference (F_{(1, 397)} = 0.51, p = 0.478, η_p²= 0.001, d = 0.070) (intervention: M = 4.89, SD = 1.16; control: M = 4.81, SD = 1.19).

Mediation

We tested whether feelings associated with injustice mediated the effect of the intervention on behavioral intentions. We used the bootstrapping mediation method of Preacher and Hayes (2004, 2008), with 5,000 bootstrap resamples and effect size 1. There was a significant effect of intervention on behavioral intentions (B = –0.176, SE = 0.72, t = –2.44, p = 0.015). This effect was reduced to nonsignificant when the mediator feelings associated with injustice was included (B = –0.117, SE = 0.71, t = –1.64, p = 0.102), while there was a significant effect between feelings associated with injustice and behavioral intentions (B = 0.105, SE = 0.22, t = 4.73, p < 0.001). Finally, there was a significant effect of intervention on the mediator feelings associated with injustice (B = –0.56, SE = 0.16, t = –3.54, p < 0.001). The bootstrap results for the indirect effect of the intervention on behavioral intentions via feelings associated with injustice was significant (z = –2.797, SE = 0.021, p = 0.005; 95%CI [–0.101; –0.018]).

General Discussion

The results of both studies showed that the intervention led participants to intend to reduce their harmful behavior toward animals more strongly compared with those in the control condition. When controlling for social desirability, this effect was still obtained. The results of study 2 demonstrated that feelings associated with injustice mediated this effect. The intervention did not lead to reduced speciesist attitudes or to more signed petitions for equal animal rights.

The effect of the intervention on behavioral intentions and on feelings associated with injustice can be regarded as small to moderate (ds = 0.25–0.29) according to Cohen (1988). Note, however, that these values should not be interpreted rigidly (e.g., Thompson, 2007) – small effect sizes can have large consequences (Lakens, 2013). When interpreting the means for behavioral intentions in the intervention condition, these reflect scores between harming animals less than they currently do (a score of 3) and stopping the harm altogether (a score of 4).

Interestingly, the results about behavioral intentions show that participants in the control condition also intended to reduce harming animals. This may seem surprising, but nowadays there is a strong focus on climate change, zoonoses, and animal welfare in the media. This may explain why most participants intended to change their behavior regardless of the condition they were in. Importantly, participants in the intervention condition experienced an even stronger intention to change their behavior compared with those in the control condition.

The present studies are the first to experimentally assess the effectiveness of an intervention toward changing people’s speciesist attitudes and behavioral intentions. Previous studies found that different types of interventions were successful in reducing people’s intentions to consume animal products (for overviews, see Bianchi et al., 2018; Grundy et al., 2022; Harguess et al., 2020; Vos et al., 2021). We add to these findings by showing that our broader intervention – focusing on multiple industries instead of only focusing on animal-product consumption – led participants to intend to reduce harmful behaviors toward animals more strongly in general.

What drove this effect? The results imply that it is not necessarily the role reversal that caused a stronger intention to reduce speciesist behaviors. We did not test this directly with a control condition without reversing roles as we were interested in whether this existing intervention as opposed to no intervention would affect speciesism. By measuring possible mediators, however, we could provide insight into why this intervention changed behavioral intentions. First, the intervention did not increase perspective-taking and awareness of the common treatment of animals. Perspective-taking and awareness, however, were related to speciesists attitudes and behavioral intentions. This means that, as could be theoretically expected, perspective-taking and awareness are important factors that could influence speciesism. The intervention, however, only elicited feelings associated with injustice, which mediated the effect of the intervention on behavioral intentions. This implies that any intervention eliciting feelings associated with injustice toward animals would result in similar findings. Animal rights activists often use information to elicit awareness in people. Our findings suggest that it is not sufficient to provide information only; there is also the need to elicit feelings associated with injustice.

The feelings measured are not merely associated with injustice but can be experienced in other situations as well. An alternative interpretation of these feelings is that participants may have felt angry by the video “pushing” them into behaviors they resist showing. This does not hold, however, as these feelings would not have mediated the effect of the intervention on reducing speciesist behavioral intentions. Also, these emotions would not have correlated with perspective-taking.

The present studies show that it is not easy to reduce people’s speciesist attitudes. Note that for study 2 the means of attitude were in the opposite direction of what was expected (p = 0.094). This could be interpreted as a boomerang effect (e.g., Kuang et al., 2020). However, as p-values between 0.05 and 0.10 are regarded as having low evidential value (e.g., Olsson-Collentine et al., 2019), and we did not obtain means opposite from what was expected in study 1, this could be a chance outcome. To successfully reduce speciesism, possibly multiple interventions are needed as attitudes do not change quickly over time (Rice et al., 2020; Tonsor & Olynk, 2011). Also, people may vary in what they respond to in an intervention. An intervention could therefore include multiple aspects (e.g., consequences for health, the environment, and the animals). Finally, the correlational analyses showed that the more people take the perspective of animals, the less their speciesist attitudes (r = –0.58). Future research could investigate whether an intervention explicitly asking people to take the perspective of animals would successfully reduce speciesist attitudes.

The results showed no effect on the behavioral response, whereas behavioral intentions were affected. A reason for this is that the behavioral response was quite different from the intentions. The behavioral response focused on the more activistic behavior of making a societal change by signing a petition for equal rights, whereas the behavioral intentions focused on behaviors participants could perform themselves. Even though signing a petition is more activistic, the majority of participants were willing to sign. It is possible that simply indicating yes or no on an online petition is not reflective of real behavior as it is very easy to do. Even though in study 2, we increased the effort participants had to make to agree, the majority still agreed. It is possible that the behavioral response is more strongly affected by other variables than the result of speciesism, such as those related to prosocial behavior (e.g., personality traits, cooperative values; Politi et al., 2021) and collective action (e.g., efficacy, identity; van Zomeren et al., 2008). Nevertheless, the correlational analyses showed that signing the petition was strongly related to speciesist attitude. Future research could investigate whether people’s behavior changes by a priori measuring their actual speciesist behavior and then repeating this measure after a period of time after the intervention.

Potentially, warning about possible distress when collecting informed consent may have led sensitive individuals to decide not to participate. So, individuals who may have been affected more strongly by the intervention may not be represented in our sample. Other than considering this possible selection bias which could have dampened the effects, we have no reason to believe the findings cannot be generalized to a broader population.

Interventions by animal activists are shared via different mediums (e.g., live, tv, radio, social media). We did not investigate the reach and effect of sharing this intervention in these different mediums as this was beyond the scope of our research. Our aim was to investigate whether this intervention, when being watched, would reduce speciesism. Previous research investigating the effects of showing animal welfare issues in the media shows inconsistent results (e.g., Rice et al., 2020; Sinclair et al., 2018; Tiplady et al., 2013; Tonsor & Olynk, 2011). As in other research fields, change may be dependent on several factors, including message valence, length of information, and communicator (e.g., Bergmann et al., 2022; Wang et al., 2021). Importantly, media attention on animal welfare at least increases community discussion (Rice et al., 2020; Tiplady et al., 2013). Multiple media exposures, however, may result in a cumulative effect leading to a significant behavioral change (Rice et al., 2020; Tonsor & Olynk, 2011).

Conclusion

In conclusion, the results show that watching the video, with the aim to increase awareness of the way animals are being treated by humans, did not affect speciesist attitudes or people’s willingness to sign a petition for animal rights compared with those who did not watch the video. It did, though, reduce people’s intentions to harm animals. The results also provide ideas for future interventions to reduce speciesism and prevent the harming of animals.

Disclosure Statement

No potential conflict of interest was reported by the authors.