Content uploaded by Christopher Bryant
Author content
All content in this area was uploaded by Christopher Bryant on May 17, 2020
Content may be subject to copyright.
- 1 -
Educated Choices Program
An Impact Evaluation of a Classroom
Intervention to Reduce Animal Product
Consumption
Christopher Bryant
& Courtney Dillard
January 2020
Note: The researchers were contracted by ECP to carry out this impact assessment, but
ECP had no input on the analysis and reporting of the data.
- 2 -
Executive Summary
In this comprehensive evaluation of Educated Choices Program’s educational
intervention, we report on our analysis of 95,241 student survey responses. We are
excited to share these findings for a number of reasons. First and foremost, our analysis
clearly demonstrates the effectiveness of this educational intervention in positively
impacting student attitudes, behavioral intentions and self-reported behaviors in regard
to their food choices. The scale of the dataset and the comprehensive nature of the
analyses conducted provides a strong basis for funding considerations for educational
interventions. This is particularly heartening because similar impacts have been difficult
to find in other consumer-facing advocacy interventions. As will be highlighted later in
the report, ECP’s model of intervention is both high quality and cost effective, allaying
some fears about the feasibility of deploying effective interventions of this nature on a
large scale.
Our key findings include:
Overall Effectiveness
1. Watching an ECP presentation reduced students’ intended and subsequent
self-reported animal product consumption.
2. Watching an ECP presentation increased students’ rates of vegetarianism and
veganism.
Motivations & Barriers
3. Health was the most commonly cited motivation, but those who cited animal
welfare as their main motivation were most likely to change their diets.
4. Taste was by far the most commonly cited barrier, and those who cited taste
as their main barrier were least likely to change their diets.
Additional Findings
5. A presentation on animal product alternatives increased willingness to
consume alternatives and decreased intentions to consume animal products.
6. Many students reported making other changes after seeing an ECP
presentation, including increasing vegetable consumption, trying to reduce
their impact on the environment, and sharing information about food-related
issues with others.
- 3 -
Strategic Observations
7. Presentations were effective for males and females, for all age groups, and for
students in all US states as well as Canada.
8. Presentations focused on animal ethics lead to fewest people intending to
reduce animal products, but most people intending to eliminate animal
products altogether.
- 4 -
Contents
Executive Summary 2
Contents 4
1. Introduction 5
2. Methods 7
2.1 Procedure 7
2.2 Materials 8
2.3 Participants 9
3. Findings 10
3.1. Overall program effectiveness 10
3.1.1 Intentions: treatment vs. control 11
3.1.2 Vegetarianism and veganism: repeat viewers vs. first time viewers 12
3.1.3 Long term changes: follow up surveys 13
3.2 Motivations and barriers 16
3.2.1 Motivations 17
3.2.2 Barriers 19
3.3 Future food technologies 22
3.3.1 Impact of ‘Future of Meat and Dairy’ presentation 23
3.3.2 Demographic differences 25
3.3.3. Barriers and motivations 27
3.4 Other program outcomes 29
3.5. Effectiveness of different presentations 30
3.6. Receptiveness of different students 33
3.6.1 Gender differences 34
3.6.2 Age differences 35
3.6.3 Countries and states 37
4. Discussion & Recommendations 39
4.1 Summary 39
4.2 Recommendations 40
4.3 Limitations 41
References 42
Appendices 44
Appendix A: Key question wording 44
Appendix B: Author biographies 49
- 5 -
1. Introduction
The Educated Choices Program (ECP) is a USA non-profit which aims to reduce animal
product consumption through educational classroom presentations. The organisation
develops and delivers high quality presentations to educate middle school, high school,
and college students in the USA and Canada about the impacts of their food choices, in
particular the ethical, environmental, and health impacts of animal agriculture. The
presentations are well-sourced, regularly updated, and delivered by educators who are
employed and trained specifically to deliver consistent presentations without bias or
judgement.
Animal Charity Evaluators (2014) has highlighted a paucity of evidence around the
effectiveness of classroom education, deeming it ‘a promising intervention and worth
further investigation.’ Existing studies on the effectiveness of similar programs are very
limited: many are outdated, they often have small sample sizes, focus exclusively on
elementary school age children, and have very limited follow-ups (Animal Charity
Evaluators, 2014; O’Brien, 2003). Most of these studies do not evaluate education
programs focused on the benefits of plant-based eating, and do not use animal product
consumption as an outcome measure, instead focusing on wild or companion animals
(O’Brien, 2003).
The research that has been done on educational interventions regarding farm animals is
largely limited to examining changes in knowledge and attitudes rather than food
choice. Aguirre and Orihuela (2010) found that at a 10 week program in rural Mexican
schools aimed at six-year-olds could lead to increased knowledge and some positive
changes in attitudes towards farm animals. Another study of a shorter intervention in
Mexico, a two week program, also resulted in increased knowledge and slight changes in
attitudes (Orihuela, Aguirre & Lakestani, 2015). Limited research in the United Kingdom
on an educational event related to chicken welfare demonstrated a short-term increase
in knowledge and minimally impacted attitudes (Jamieson et al., 2012). Finally, Samuels,
Meers and Normando (2016) found that, after participation in an 11 week program
student attitudes towards a variety of animal types, including farm animals, were more
positive than those in a control group.
The most relevant research, according to Animal Charity Evaluators (2014), is from a
2013 study conducted in collaboration with other organisations. The study compared
attitudes and behaviors of students who attended in-class presentations by several
animal advocacy organizations to those from the same schools who did not. The surveys
were sent out roughly two months after the presentations were given to measure longer
term impacts. This study did not find significant reductions in animal product
consumption for a group exposed to an educational intervention compared to a control
group, and in fact found that consumption of some products was higher in the treatment
- 6 -
group (Animal Charity Evaluators, 2013). However, as with previous studies, the sample
size was very limited (169 in the treatment group, 60 in the control group) which meant
it was relatively unlikely to detect an effect.
Crucially, an important consideration for effective animal advocates is cost
effectiveness; Animal Charity Evaluators (2013) concluded that classroom interventions
are unlikely to be particularly effective, partly because they are considerably more
expensive to deliver compared to other interventions such as leaflets and online
adverts. However, ECP is able to deliver their presentations at a cost of just $2.68 per
student (Educated Choices Program, 2019), meaning that they could be cost-effective if
they encourage many more students to reduce animal product consumption compared
to other interventions.
Furthermore, there is reason to think that classroom education may be more effective
than other consumer-facing interventions. A meta-analysis found that leafleting has
limited or no effects on animal product consumption (Animal Charity Evaluators, 2017).
Likewise, Animal Charity Evaluators’ (2016) review of online adverts argued that this
type of short impersonal interaction can be expected to have limited impacts. Similarly,
Bryant (2019) points to evidence suggesting that leaflets and online adverts have
limited impact on behaviour. A long classroom session with high quality interactive
material delivered by a trained educator can reasonably be expected to have more
impact on an individual than a leaflet or an online advert.
This impact evaluation sought to determine the effectiveness of ECP’s classroom
presentations for reducing animal product consumption. This evaluation’s major
strength is the size of the dataset compared to previous studies - data has been collected
from tens of thousands of students over the course of several years. We conduct a
variety of analyses to assess different program outcomes, and report findings and
recommendations.
- 7 -
2. Methods
2.1 Procedure
Students were exposed to a classroom education intervention which aimed to reduce
their consumption of animal products. Students saw one of five different presentations.
The first four of these were ECP’s primary presentations which had run for several
years. The fifth was a newer presentation developed to inform students about animal-
free food technology.
Table 1: Five different ECP presentations shown to students.
Presentation
Focus
The Environment and Modern Agriculture
The negative environmental impacts of
animal agriculture
Modern Animal Agriculture
The animal suffering involved in animal
agriculture
Healthful Eating
The negative health impacts of consuming
animal products
The Ethics of Eating
A combination of environmental, animal
welfare, and health issues
The Future of Meat and Dairy
Animal product replacements including
plant- and cell-based meat and dairy
After watching the presentation, students privately answered an online survey for the
chance to win a $100 Amazon voucher. Students were made aware that the prize would
be allocated in a random draw without regard to the content of survey responses.
Separately, a different set of students answered an online survey before being exposed
to the classroom presentation. This set of students would act as a control group, against
which the responses of the treatment group could be compared.
Finally, students who were over 18, had seen the presentation between 3-30 months
ago, and had consented to be contacted for a follow-up, were contacted via email to take
part in a follow-up survey for the chance to win a $500 Amazon voucher.
- 8 -
2.2 Materials
The online surveys were designed in Survey Monkey, and contained questions about:
a) Demographics (age, gender, country of residence)
b) Diet prior to the presentation
c) Intended future consumption of animal products
d) Barriers and motivations regarding adopting a plant-based diet
e) Opinions of cell-based meat
f) Educator knowledgeability and respectfulness
g) Novelty, importance, and reliability of the presentation subject matter
h) Previous exposure to ECP presentations
i) Additional information including school name, contact details, and written
feedback
The exact wording of key questions can be found in Appendix A. We analysed data from
four different datasets:
Dataset 1 includes 84,182 student responses to an in-class survey with records from
2017 until the end of October 2019.
Dataset 2 includes 7,634 responses to an in-class survey. Of these, 1,525 are from a
control group, and 6,109 are from groups who had seen ECP presentations since the
beginning of November 2019.
Dataset 3 includes 261 responses to a follow-up survey which was sent to students by
email between 3-30 months after they saw the presentation, including 190 matched to
the original responses of those students to the in-class survey.
Dataset 4 includes 3,164 responses to a survey about new food technologies, including
plant-based and cell-based meat and dairy. Of these, 595 are from a control group, and
2,569 are from groups who saw an ECP presentation about future food technologies.
- 9 -
2.3 Participants
Participants (total n = 91,816) were students at schools and colleges in the USA and
Canada. The overall age and gender makeup of the sample is shown in Graph 1:
- 10 -
3. Findings
3.1. Overall program effectiveness
Key takeaways:
1. Viewing an ECP presentation leads to a large significant increase
in intentions to reduce animal product consumption.
2. Viewing an ECP presentation leads to a significant increase in
vegetarianism and veganism.
3. Viewing an ECP presentation leads to a reduction in animal
product consumption in the long term behavior of most
students.
We assessed the overall effectiveness of ECP presentations on students who saw them
using three approaches. First, we compared the intended future diets of those who had
seen an ECP presentation to a control group who had not seen the presentation. Second,
we compared the rates of vegetarianism and veganism between students who had seen
an ECP presentation before and those who had not. Third, we conducted a follow-up
survey to assess how many students actually changed their diets, and how reliable their
intended changes were.
- 11 -
3.1.1 Intentions: treatment vs. control
Using Dataset 2, we compared the responses of 5,710 students who saw an ECP
presentation to a control group of 1,210 students who had never seen an ECP
presentation. Both groups answered the same question about their intended future
consumption of meat and animal products.
As shown in Graph 2, 69.4% of those who had seen an ECP presentation intended to
reduce their consumption of meat and other animal products, whilst 5.2% intended to
give them up altogether. In the control group, just 34.0% said they would reduce their
consumption of animal products, and 3.7% said they would give them up altogether.
The difference between the treatment and control groups was statistically significant
(𝝌2=630.227, p<0.001).
The large difference in intended animal product consumption between students who
had seen an ECP presentation and those who had not indicates that seeing an ECP
presentation caused many students to intend to reduce their animal product
consumption.
- 12 -
3.1.2 Vegetarianism and veganism: repeat viewers vs. first time viewers
In Dataset 1, some students (n = 14,387) answered a question about whether they had
seen an ECP presentation before. Those who had already seen an ECP presentation
answered the same survey again, and can be seen as a ‘natural follow-up’ group.
Observing the rates of vegetarianism and veganism in students who had seen an ECP
presentation before compared to those who had not enables us to assess the
effectiveness of the program from another angle.
As shown in Graph 3, 6.8% of those who had seen an ECP presentation before were
vegetarian, and 5.6% were vegan - this compares to just 5.9% vegetarians and 3.9%
vegans in the group who had not seen an ECP presentation before. This difference was
also statistically significant (𝝌2=13.067, p=0.001).
The difference in rates of vegetarianism and veganism between those who had already
seen an ECP presentation compared to those who had not indicates that seeing an ECP
presentation caused some students to become vegetarian or vegan.
- 13 -
3.1.3 Long term changes: follow up surveys
We sent out a follow-up survey to students who (a) were now over 18, (b) consented to
be contacted for a follow-up, and (c) had seen the ECP presentation between 3-30
months ago. Of the 3,778 students we contacted via email, 261 (6.9%) responded to the
survey (see Dataset 3). Of these, 224 were meat-eaters before seeing the presentation.
Of these 224 students, the majority agreed that seeing the ECP presentation had caused
them to decrease their consumption of animal products, as shown in Graph 4.
Additionally, we asked about students’ consumption of specific animal products
compared to before they saw the presentation.
- 14 -
As shown in Graph 5, 62.9% said they had reduced or eliminated their consumption of
beef, while 71.9% reduced or eliminated their consumption of pork. 40.2% said they
had reduced or eliminated their consumption of chicken and 46.0% said they had
reduced their consumption of fish - this is compared to just 10.7% and 12.1% who
increased their consumption of these foods respectively.
Importantly, this offers evidence that meat reducers did not tend to replace red meat
with more white meat. This is an important concern for animal advocates, because
whilst health and environmental arguments favour reducing red meat consumption,
white meat is associated with more animal suffering and therefore a replacement effect
could actually increase animal suffering (Bryant, 2019). No such replacement effect is
observed here.
Of the follow-up respondents, we could link 190 to their original in-class survey
responses using their email address; 172 of these ate meat before seeing an ECP
presentation. Comparing these responses allows us to observe whether reporting an
intention to reduce animal product consumption predicted an actual (self-reported)
decline in consumption.
- 15 -
As shown in Graph 6, a much higher proportion of those who intended to reduce their
animal product consumption actually did reduce their consumption compared to those
who did not intend to. An even higher proportion of those who intended to eliminate
their animal product consumption managed to cut down. This tells us that, while
intentions are not enough, they do at least predict self-reported behaviour to a
considerable extent.
- 16 -
3.2 Motivations and barriers
Key findings:
1. Health was the most common motivator, followed by animals,
followed by the environment.
2. However, selecting health as the main motivator was associated
with lower intentions to reduce animal product consumption.
3. Selecting animals as the main motivator was associated with
higher intentions to stop consuming animal products altogether.
4. Taste was by far the most common barrier, followed by finding
enough plant-based options and convincing parents.
5. Selecting taste as the main barrier was associated with lower
intentions to reduce animal product consumption.
- 17 -
3.2.1 Motivations
In Dataset 1, students were asked about their main motivations for wanting to follow a
plant-based diet.
As shown in Graph 7, health was the most common motivator amongst all students,
followed by animals, followed by the environment. We can also see that, for girls,
animals were almost as common a priority as health, whereas for boys, the environment
was more common a concern than animals. There were few notable differences in age
groups. Interestingly, respondents who were vegetarian most often said animals were
their main motivation, whilst meat-eaters more commonly chose health.
- 18 -
As shown in Graph 8, those who said animals were their main motivation were far more
likely to say they would not just reduce, but eliminate animal product consumption. On
the other hand, those who said health was their main motivation were far more likely to
say they would change nothing. These differences were statistically significant
(𝝌2=1611.82, p<0.001).
- 19 -
3.2.2 Barriers
In Dataset 1, students were asked about their most pressing barriers to eating a plant-
based diet.
As shown in Graph 9, the most common barrier to a plant based diet by far was
concerns about taste.; second was finding enough plant-based options, and third was
convincing parents. We can compare the prevalence of these most common barriers
across different groups:
- 20 -
Graph 10 shows that, although taste was the most commonly cited barrier for all
omnivorous groups, boys were relatively more likely to consider taste the primary
barrier, whilst girls were relatively more likely to consider practical factors like finding
plant-based options and convincing their parents. Interestingly, older students tended
to be less concerned about taste or convincing their parents, and more concerned about
finding plant-based options relative to younger students. Of those who were already
veg*n, the most common barrier was finding enough plant-based options - they were far
less concerned about taste than meat-eaters.
- 21 -
As shown in Graph 11, those who considered taste to be the most significant barrier
were much more likely to say they would not change their diet compared to those who
gave other common barriers. They were far less likely to say they would eliminate their
animal product consumption, also. It appears that those who gave practical barriers like
finding enough plant-based options and convincing their parents were more likely to
intend to reduce or eliminate their animal product consumption compared to those who
said taste was the primary barrier.
- 22 -
3.3 Future food technologies
Key findings:
1. Viewing a presentation about future foods including plant-based
and cell-based meat and dairy led to more positive attitudes
towards these foods.
2. This presentation also led to more negative attitudes towards
animal products, and an increased willingness to reduce animal
product consumption.
3. Plant-based alternatives were most appealing to girls and to
older students.
4. Food safety was considered the most important issue with
regards to cell-based meat.
- 23 -
3.3.1 Impact of ‘Future of Meat and Dairy’ presentation
In Dataset 4, students (n = 3,164) were asked about their views of new cell-based and
plant-based meat and dairy products. This included a treatment group (n = 2,569) who
had seen an ECP presentation about future food technologies, and a control group (n =
595) who had not seen the presentation. Participants rated their interest in trying
various foods and the importance of different food characteristics on a 1-5 scale.
First, we observed statistically significant (p < 0.05) differences between the treatment
and control groups with respect to their attitudes towards future food technologies, as
shown in Graph 12.
While the major differences were in attitudes towards plant-based and cell-based food
products, we also detected differences in attitudes towards conventional meat and dairy
also changed, which were significantly more negative in the treatment group compared
to control. For the control group, conventional meat and dairy were the most appealing
foods, followed by plant-based dairy. For the treatment group, cell-based and plant-
based meat were the most appealing foods, while conventional meat and dairy were the
least appealing.
We also wanted to test whether this presentation affected students’ intended
consumption of animal products. Whilst the presentation highlighted the major issues
with animal agriculture, it also presented alternative products (some of which are not
- 24 -
yet commercially available) as a viable solution, and could therefore reduce the urgency
to change.
As shown in Graph 13, viewing the presentation about future food technologies actually
increased intentions to reduce animal product consumption compared to a control
group. This was a statistically significant difference (𝝌2=578.18, p<0.001), and indicates
that viewing this presentation caused students to intend to reduce their animal product
consumption.
- 25 -
3.3.2 Demographic differences
We also observe demographic differences in the interest towards these products.
As shown in Graph 14, plant-based meat and dairy tended to be more appealing to girls
compared to boys, whereas the reverse was true for conventional meat and dairy.
Additionally, plant-based products appeared to be more appealing to older students,
whereas conventional animal products were more appealing to younger students.
- 26 -
Predictably, vegetarians and vegans were more interested in plant-based products and
less interested in conventional animal products compared to omnivores.
Additional data from Dataset 1 provides further insight onto some demographic trends
with respect to attitudes towards cell-based meat. Students (n=70,572) were asked
about their willingness to eat ‘clean meat’, a term sometimes used to refer to cell-based
meat (see Bryant & Barnett, 2019).
As shown in Graph 15, most students said they would be very or extremely interested in
eating cell-based meat. Older students were more likely to say they found cell-based
meat appealing. We also see higher interest amongst omnivores compared to veg*ns, a
trend which is consistent with survey data from elsewhere (Wilks & Phillips, 2017).
There is very little difference between the genders.
- 27 -
3.3.3. Barriers and motivations
Additionally, students were asked about the importance of various motivators, barriers
and conditions for trying cell-based meat specifically:
As shown in Graph 16, participants rated food safety as the most important concern by
far, followed by taste and texture. Health was rated as the most important motivator,
though all reasons were considered to be important. Interestingly, data from Dataset 1
provides an alternative view on the matter of motivations, as shown in Graph 17.
- 28 -
When participants who saw one of the main four presentations (i.e. not the Future of
Meat and Dairy presentation) were asked to choose one motivation for trying cell-based
meat, the largest choice for all groups was animals. This could indicate that viable
alternatives to animal products could open the door to stronger consideration of animal
welfare in general. Indeed, this idea is in line with experimental research which has
demonstrated that eating meat causes a people to consider animal suffering less
important (Loughnan, Haslam & Bastian, 2010).
- 29 -
3.4 Other program outcomes
In Dataset 3 (the follow up survey, n = 261), we also asked about a range of other
outcomes.
As shown in Graph 18, most students who responded to the follow-up survey said they
had engaged more in other positive behaviours such as eating more healthy plant-based
foods, sharing information about the impact of animal agriculture on animals, the
environment, and human health, and trying to reduce their environmental impacts in
other ways.
Respondents said they discussed things they had learned in the presentation with an
average of 15 other people. Interestingly, they were more likely to share information
about the environment and animal welfare than health.
- 30 -
3.5. Effectiveness of different presentations
Key finding: The animal-focused presentation resulted in the smallest
proportion intending to reduce animal product consumption, but it
also resulted in the highest proportion intending to give up animal
products altogether.
Using Dataset 1, we compared the future dietary intentions of omnivorous students
(n=79,241) who saw four different ECP presentations (see Table 1).
As shown in Graph 19, the outcomes for each of the four presentations were overall
fairly similar. However, compared to the other presentations, a notably lower
proportion of those who saw the Modern Animal Agriculture presentation said that they
would reduce their consumption of animal products, and a higher proportion said they
would change nothing. Interestingly, a higher proportion also said they would cut out
animal products altogether. This difference was statistically significant (𝝌2=185.05,
p<0.001). This likely reflects that animal suffering is a more ‘all or nothing’ issue
compared to the incremental approach one could take to reducing environmental
impact or caring for ones health.
It is worth noting that the Future of Meat and Dairy data is not included here, and that
this presentation appeared to be much more effective in reducing intended animal
product consumption compared to the other 4 presentations (see Graph 13). However,
- 31 -
the students who saw this presentation answered a different questionnaire, which may
have primed them differently, so responses are not necessarily comparable. It is also
possible that, in the Future of Meat and Dairy survey, more students interpreted the
question about future consumption of animal products in the context of evolving food
technology, and thought they would consume fewer animal products by default in the
future rather than intending to consciously reduce their animal product consumption.
This is less likely to have been the case in the main four presentations. In any case, it
seems this comparison is not as relevant as comparing the four primary ECP
presentations.
We also compared various other ratings given to the four presentations by all students
who saw them (n=84,182):
As shown in Graph 20, all of the presentations were rated very well in terms of their
importance, reliability, novelty of information, and the knowledgeability and
respectfulness of educators. Although by a small margin, Modern Animal Agriculture
was fairly consistently rated the poorest on average on most of these measures. This
might reflect the stronger dissonance associated with the subject matter compared to
the other presentations.
Notably, boys were especially put off by the discussion of animal ethics - they were
significantly less likely to reduce their meat consumption not only in the Modern Animal
Agriculture presentation, but also in the Ethics of Eating presentation compared to the
other two (𝝌2=93.853, p<0.001), suggesting that the addition of animal ethics to
otherwise persuasive material might undermine the effectiveness of the presentation
for some boys.
- 32 -
- 33 -
3.6. Receptiveness of different students
Key findings:
1. The impact of the ECP presentation on intentions to consume
animal products was not significantly different for male and
female students, nor for students of different age groups.
2. After viewing an ECP presentation, most students in both the
USA and Canada intended to reduce their consumption of animal
products.
3. After viewing an ECP presentation, most students in all US States
for which we could collect data intended to reduce their
consumption of animal products, including those in conservative
and agricultural states.
- 34 -
3.6.1 Gender differences
In the preceding analysis, the students were randomly allocated to the presentations,
and observing the differences in intended diet could tell us which presentations were
most effective. However, observing the difference in effectiveness for different
demographic groups is more complicated, because we know that future dietary
intentions are likely to be different for different demographic groups. For example, girls
may be a priori more likely than boys to intend to reduce their animal product
consumption. Therefore, to measure the impact of watching an ECP presentation on
these different groups, we must use a ‘difference in differences’ approach. We observe
the difference in intended diets for each demographic group between a treatment and
control group.
As shown in Graph 21, females were more likely than males to say they would reduce or
eliminate animal product consumption after viewing an ECP presentation (C vs. D).
However, females were also more likely than males to intend to reduce animal product
consumption in the control group (A vs. B). We can also see that, for both males and
females, the presentation substantially increased intentions to reduce animal product
consumption (A vs. C and B vs. D). The question we want to answer for strategic
purposes is whether the presentation made more of a difference for females (B vs. D)
than it did for males (A vs. C). We can see from simple subtractions that the differences
between control and treatment groups was about 2 percentage points larger for females
than for males
1
.
1
The difference in those intending to change nothing = (55.7 - 17.3) - (71.5 - 35) = 1.9.
The difference in those intending to eliminate animal product consumption = (6.6 - 4.2) - (3.2 - 2.8) = 2.
- 35 -
To test whether this difference was statistically significant, we used an ordinal logistic
regression model. The dependent variable was dietary intentions: responses are coded
as 0 (they do not intend to reduce their animal product consumption), 1 (they intend to
reduce their animal product consumption), or 2 (they intend to eliminate their animal
product consumption). The independent variables were the experimental condition
(control vs. treatment), gender (male vs. female), and an interaction term (condition x
gender). We already know that experimental condition and gender account for
substantial differences in intentions to reduce animal product consumption, so we
expect these to be significant predictors in the regression. If the regression also
identifies the interaction term as significant, we can say that the presentation had more
impact on one gender than the other.
The results confirmed that experimental condition (ß = 1.693, p < 0.001) and gender (ß
= 0.892, p < 0.001) both had significant effects on the outcome. However, the interaction
term was not significant (ß = 0.179, p = 0.193). This indicates that the ECP presentation
was not significantly more likely to change females’ behaviour compared to males.
3.6.2 Age differences
Similarly to gender differences, in order to assess whether the ECP presentation had a
different impact on different age groups, we must compare the treatment vs. control
groups in Dataset 2 using a difference in differences approach.
- 36 -
As shown in Graph 22, older students tended to have slightly higher intentions to
reduce animal product consumption compared to younger students in the control
group. There is a somewhat similar pattern in the treatment group, though this is less
clear.
To test whether these differences were statistically significant, we used a similar ordinal
logistic regression model. The dependent variable was dietary intentions: responses are
coded as 0 (they do not intend to reduce their animal product consumption), 1 (they
intend to reduce their animal product consumption), or 2 (they intend to eliminate their
animal product consumption). The independent variables were the experimental
condition (control vs. treatment), age group (12-13 vs. 14-15 vs. 16-17 vs. 18+), and an
interaction term (condition x age group). We already know that experimental condition
accounts for substantial differences in intentions to reduce animal product
consumption, so we expect this to be a significant predictor in the regression. If the
regression also identifies age group as significant, we can say that there are significant
differences in the intentions of different age groups. If the regression identifies the
interaction term as significant, we can say that the presentation had more impact on
some age groups than on others.
The results confirmed that experimental condition (ß = 1.369, p < 0.001) had a
significant outcome on dietary intention. However, there were no significant differences
between outcomes for different age groups (all ßs ≤ .325, all ps ≥ 0.051). Furthermore,
the interaction terms were also non-significant (all ßs ≤ .278, all ps ≥ 0.277). This
indicates that the ECP presentation had a similar impact on all age groups.
- 37 -
3.6.3 Countries and states
We can also look at the program outcomes for students in different geographical areas.
In Dataset 1, more recent respondents indicated the country they live in (since ECP
started in Canada). As shown in Graph 23, the students in the USA and Canada had
similar intentions after viewing a presentation. This difference was not statistically
significant (𝝌2=0.836, p=0.658)
2
.
We can also look at dietary intentions after viewing an ECP presentation across states.
As shown in Graph 24, ECP presentations led to significant intentions to reduce and
eliminate animal products in all US states where they were given, making it an effective
intervention in all regions of the country. It is particularly interesting to note that this
impact can be seen even in states which are generally considered to be more
consevative like Georgia and South Carolina, as well as in states highly associated with
animal agriculture like Texas and Indiana.
2
This analysis is different from the difference in differences approach above, since we do not have control
data for Canada or all US states.
- 38 -
- 39 -
4. Discussion & Recommendations
4.1 Summary
Overall, we found compelling evidence that ECP presentations are an effective
intervention to reduce consumption of animal products amongst adolescents and
young adults. The majority of students who saw an ECP presentation intended to
reduce or eliminate their animal product consumption. Those who saw ECP
presentations were significantly more likely to reduce their animal product
consumption compared to a control, and those who had seen an ECP presentation
before were significantly more likely to be vegetarian or vegan than those who had not.
Moreover, evidence from a follow-up survey indicates that most of those who intended
to reduce their animal product consumption followed through, and many increased
their consumption of plant-based alternatives.
Given that the available evidence suggests that other consumer-facing interventions
including leafleting and online advertisements are of limited impact, the results of this
study are hugely encouraging for animal advocates. Until now, there has been very
limited evidence on the effectiveness of classroom education interventions aimed at
reducing animal product consumption. Much of the available evidence is based on small
and limited samples, and often addresses general animal issues rather than animal
product consumption. The scale of the datasets used in this study have allowed us to
observe patterns previously unknown to advocates, and provide good evidence of the
effectiveness of classroom education interventions.
- 40 -
4.2 Recommendations
Based on these findings, we can make the following recommendations to improve the
effectiveness of ECP presentations overall.
1. Continue to reach all groups of students. We found evidence that the majority
of students who watched an ECP presentation intended to reduce their animal
product consumption, irrespective of age, gender, country and state. In fact,
analyses indicate that the presentations had similar impacts across genders and
age groups. Therefore, the optimal strategy to pursue in educational
interventions is reaching as many students as possible without discriminating
based on demographics.
2. Use the most impactful materials. There is reasonably good evidence that
fewer students will intend to reduce their animal product consumption in
presentations focused on animal suffering compared to presentations focused on
the environment or health. However, more students in this group intend to
eliminate animal product consumption altogether. Moreover, students who say
animals are their main motivation are more likely to reduce or eliminate animal
product consumption. Therefore, evidence on overall presentation effectiveness
is mixed: a focus on animals leads to more commitment from fewer students,
whereas a focus on other issues leads to less commitment from more students.
3. Address the major barriers. The three most common barriers to adopting a
plant-based diet are taste (by far the most common), finding enough plant-based
options, and convincing parents. Students who listed the latter two as their main
concern were more likely to reduce their animal product consumption than
those who said taste was most important, indicating that taste is the major
barrier. Advocacy groups could consider offering samples of high quality vegan
food to help overcome this barrier, although this is likely to be expensive, and
may be impractical in a classroom setting.
4. Highlight the major benefits. Health was the most common motivation given
for all groups, but was also associated with the lowest rate of intended action -
animal welfare was the motivation corresponding to the highest rate of intended
action. Girls are especially likely to be receptive to animal suffering messages,
whilst boys generally find this argument the least compelling. Since an animal
message corresponds to the highest rate of intended action, advocates should not
shy away from this message.
- 41 -
4.3 Limitations
There are some limitations to this evaluation which we must acknowledge.
Firstly, all of these results are based on self-reported behaviour. This is known to be less
than perfectly accurate, as participants might not accurately forecast/remember their
behaviour. This is likely to be compounded by the presence of social desirability bias -
the tendency to give answers that reflect positively on oneself, particularly with respect
to moralised issues.
Secondly, the sample in Dataset 3, the follow-up, was predictably skewed towards
those who were more positive about the ECP presentation - those who intended to
eliminate their animal product consumption were over-represented, while those who
intended to change nothing were under-represented. This could mean the results of the
follow-up were more positive than would be the case for all students.
Thirdly, all survey respondents were offered the opportunity to win an Amazon voucher
worth at least $100 - it is possible that some respondents interpreted this as being more
likely to be awarded to people who gave positive responses to the survey. That said, it is
standard practice to offer incentives in exchange for survey completion in research of
this kind. Additionally, there is no reason to think that this effect would be bigger in the
treatment group than in the control.
- 42 -
References
Aguirre, V., & Orihuela, A. (2010). Assessment of the impact of an animal welfare
educational course with first grade children in rural schools in the state of Morelos,
Mexico. Early Childhood Education Journal, 38(1), 27-31.
Animal Charity Evaluators. (2013). 2013 Humane Education Study. Accessed 30
December 2019 at https://animalcharityevaluators.org/advocacy-
interventions/interventions/humane-education/2013-humane-education-
study/#analysis
Animal Charity Evaluators. (2014, June). Archived Intervention Report: Humane
Education. Accessed 30 December 2019 at
https://animalcharityevaluators.org/advocacy-interventions/interventions/humane-
education/
Animal Charity Evaluators. (2016, August). Archived Intervention Report: Online Ads.
Accessed 06 January 2020 at https://animalcharityevaluators.org/advocacy-
interventions/interventions/online-ads/#report
Animal Charity Evaluators. (2017, November). Leafleting. Accessed 06 January 2020 at
https://animalcharityevaluators.org/advocacy-
interventions/interventions/leafleting/#report
Bryant, C. (2019, September 03). A Guide to Effective Animal Campaigning. Accessed 30
December 2019 at https://www.vegansociety.com/about-us/research/research-
news/project-update-guide-effective-animal-campaigning
Bryant, C. J., & Barnett, J. C. (2019). What's in a name? Consumer perceptions of in vitro
meat under different names. Appetite, 137, 104-113.
Educated Choices Program. (2019). Year-End Funder Report 2019.
Jamieson, J., Reiss, M. J., Allen, D., Asher, L., Wathes, C. M., & Abeyesinghe, S. M. (2012).
Measuring the success of a farm animal welfare education event. Animal Welfare-The
UFAW Journal, 21(1), 65.
Loughnan, S., Haslam, N., & Bastian, B. (2010). The role of meat consumption in the
denial of moral status and mind to meat animals. Appetite, 55(1), 156-159.
O’Brien, H. (2003). An Annotated Bibliography of Research Relevant to Humane
Education. The National Association for Humane and Environmental Education.
- 43 -
Accessed 30 December 2019 at
https://www.humanesociety.org/sites/default/files/docs/bibliography-humane-
research-volume2.pdf
Orihuela, A., Aguirre, V., & Lakestani, N. (2015). Farm Animal Welfare and Children: A
Preliminary Study Building an Attitude Scale and Evaluating an Intervention. Society and
Animals, 23(4), 363-378.
Samuels, W. E., Meers, L. L., & Normando, S. (2016). Improving upper elementary
students’ humane attitudes and prosocial behaviors through an in-class humane
education program. Anthrozoös, 29(4), 597-610.
Wilks, M., & Phillips, C. J. (2017). Attitudes to in vitro meat: A survey of potential
consumers in the United States. PloS one, 12(2), e0171904.
- 44 -
Appendices
Appendix A: Key question wording
A1: Dataset 1: Dietary intentions
The following question was used to assess dietary intentions in Dataset 1:
As a result of this presentation, you plan to (please read ALL the choices and choose
the BEST answer for your situation):
1. Eat less meat (but not less dairy/eggs)
2. Eat less dairy/eggs
3. Eat less meat AND dairy/eggs
4. Never eat meat again
5. Never eat dairy/eggs again
6. Never eat meat AND dairy/eggs again
7. Keep the same eating habits (eating meat AND dairy) you had prior to the
presentation because you did not find the information compelling enough to
change your eating habits
8. Continue eating dairy/eggs because you are happy with your current
VEGETARIAN diet
9. Continue eating only plant-based foods because you are happy with your
current PLANT-BASED diet
Participants who selected options 1-3 were coded as intending to reduce animal
product consumption. Participants who selected options 4-6 were coded as intending to
eliminate animal product consumption. Participants who selected option 7 were coded
as intending to make no changes. Participants who selected options 8-9 were excluded
from the analyses.
It has been pointed out that the inclusion of the phrase ‘because you did not find the
information compelling enough to change your eating habits’ for option 7 is an
unnecessary qualifier which could put people off choosing this option, since it is socially
awkward and potentially insulting to the presenter. Our analysis indicates that
removing this qualifier increased the selection of this option by 4.3 percentage points.
This was considered a small enough difference to keep the large amount of data
collected using the original question wording. Nonetheless, we changed the question
wording to exactly match the wording in the control group (see A7).
- 45 -
A2: Dataset 1: Current diets
The following question was used to assess current diets in Dataset 1:
Prior to this presentation, you regularly ate:
1. Meat and dairy/eggs (meat = beef, chicken, pork, fish, etc) (dairy milk,
cheese, yogurt, ice cream, etc)
2. Dairy/eggs but no meat
3. Plant-based foods only
A3: Dataset 1: Clean meat
The following question was used to assess attitudes towards cell-based meat in Dataset
1:
How interested are you in trying cell-based meat?
1. Not at all interested
2. Not very interested
3. Somewhat interested
4. Very interested
5. Extremely interested
A4: Dataset 1: Cell-based meat motivations
The following question was used to assess the most important benefits of cell-based
meat in Dataset 1:
Which of the following benefits of cell-based meat are most appealing to you?
1. Environment
2. Animal welfare
3. Health
A5: Dataset 1 & 2: Motivations
The following question was used to assess motivations in Datasets 1 and 2:
IF you were going to consider eating only a plant-based diet, which of the following
would be your most important REASON for doing so? (If you are already a plant-
based eater, what was your most important reason?):
- 46 -
1. The environment
2. Your health
3. The animals
A6: Dataset 1 & 2: Barriers
The following question was used to assess barriers in Datasets 1 and 2:
IF you were going to consider eating only a plant-based diet, which of the following
would be your greatest OBSTACLE to success? (If you are already a plant-based
eater, what was your greatest obstacle?):
1. Convincing my parents
2. Dealing with my friends
3. Finding enough plant-based food choices
4. Giving up the taste of meat and dairy
5. Giving up holiday traditions
6. No obstacles
7. Other (please specify): _________
A7: Dataset 2: Dietary intentions
The following question was used to assess dietary intentions in Dataset 2.
In the future, you plan to (please read ALL the choices and choose the BEST answer
for your situation):
1. Eat less meat (but not less dairy/eggs)
2. Eat less dairy/eggs
3. Eat less meat AND dairy/eggs
4. Never eat meat again
5. Never eat dairy/eggs again
6. Never eat meat AND dairy/eggs again
7. Keep the same eating habits (eating meat AND dairy)
8. Continue eating dairy/eggs because you are happy with your current
VEGETARIAN diet
9. Continue eating only plant-based foods because you are happy with your
current PLANT-BASED diet
This question was shown to students in the control group and students who had just
watched an ECP presentation (since November 2019).
- 47 -
A8: Dataset 3: Overall animal product consumption
The following question was used to assess the overall change in animal product
consumption in Dataset 3:
To what extent do you agree with the following statements? The ECP
presentation(s) I saw caused me to reduce my consumption of animal products.
1. Strongly disagree
2. Disagree
3. Neither agree nor disagree
4. Agree
5. Strongly agree
Participants who selected 1 or 2 were coded as disagreeing; participants who selected 4
or 5 were coded as agreeing; participants who selected 3 were coded as neither.
A9: Dataset 3: Specific animal product consumption
The following question was used to assess the change in consumption of specific animal
products in Dataset 3:
Compared to your behavior before the presentation, how often are you now eating
EACH of the following: [Beef, Chicken, Pork, Lamb, Fish, Eggs (including as an
ingredient in other foods), Dairy (e.g. cheese, milk, yogurt), Plant-based meat
alternatives, Plant-based dairy alternatives]
1. Not at all
2. Much less often
3. Slightly less often
4. About the same
5. Slightly more often
6. Much more often
Participants who selected 1 or 2 were coded as disagreeing; participants who selected 4
or 5 were coded as agreeing; participants who selected 3 were coded as neither.
A10: Dataset 4: Interest in trying animal product alternatives
The following question was used to assess students’ interest in consuming animal
product alternatives in Dataset 4:
- 48 -
How interested are you in eating/drinking each of the following? [Not at all
interested, not very interested, somewhat interested, very interested, extremely
interested]
1. Plant-based meat
2. Plant-based dairy alternatives
3. Cell-based meat
4. No-cow dairy
5. Conventional meat
6. Conventional dairy
A11: Dataset 4: Importance of motivations for cell-based meat
The following question was used to assess the importance of different motivations to try
cell-based meat in Dataset 4:
How important are each of the following factors in motivating you to try cell-based
meat? [Not at all important, not very important, somewhat important, very
important, extremely important]
1. Cost
2. Taste/texture
3. New process/technology
4. Naturalness
5. Food safety
A12: Dataset 4: Importance of barriers for cell-based meat
The following question was used to assess the importance of different barriers to try
cell-based meat in Dataset 4:
Please indicate the importance of each of these issues in any decision to try cell-
based meat? [Not at all important, not very important, somewhat important, very
important, extremely important]
1. Cost
2. Taste/texture
3. New process/technology
4. Naturalness
5. Food safety
- 49 -
Appendix B: Author biographies
Christopher Bryant
Christopher Bryant is a PhD candidate at the University of Bath in the United Kingdom.
His research has focused primarily on consumer perceptions of cell cultured meat, and
additionally on strategies for reducing animal product consumption generally. He is
aligned with effective altruism, and has worked on research projects for several pro-
animal charities.
Courtney Dillard
Courtney Dillard has a PhD in Communication Studies from the University of Texas-
Austin. Her research agenda has largely focused on social change and animal advocacy
efforts from 1980 to the present. She taught for close to two decades in the Civic
Communication & Media Department at Willamette University in Salem, Oregon before
recently accepting a position as the Social Change Researcher at Mercy for Animals
(MFA).
