Background 

This is the third and final report in our series describing the results of Faunalytics’ longitudinal study of new vegans and vegetarians (veg*ns). It focuses on the critical issue of barriers and supports facing people who start a new veg*n diet, as well as the effectiveness of various strategies. The barriers and strategies included in this study are described briefly in the tables below. For more detail, see the Method & Results tab.

Table 1. Barriers & Strategies Considered

Participants 

This study includes 222 members of the general public in the U.S. and Canada, all of whom had started transitioning to a vegan or vegetarian diet within the past two months.

The Level of Commitment section of the first report shows that more than 90% of the sample said they would probably or definitely continue their new diet change permanently. This sample should therefore be considered most representative of people who have already moved beyond a simple interest or desire to change into the stage where they are ready to actively work toward a veg*n goal. Stages of change are considered in more detail in the second report.

Key Findings

1. The worst barriers to diet change were feeling unhealthy, not seeing veg*nism as part of one’s identity, and believing society perceives veg*nism negatively. People with these particular barriers were more likely than others to abandon their attempt to go veg*n. Specifically, people who felt unhealthy on their veg*n diet were more than three times as likely to abandon it within the first six months (30% vs. 8%). People who did not see veg*nism as part of their personal identity were about twice as likely as others to abandon it (16% vs. 8%). And people who thought society perceives veg*nism negatively were about 1.5 times as likely as others to abandon their diet (13% vs. 8%). In addition to these top three, there were many additional barriers that made it more difficult for people to cut out animal products. These are listed in order of importance in Table 9 on the Method & Results tab.
2. Cost-reduction strategies were the most useful type of strategy across all barriers: for instance, researching low-cost products that fit one’s diet (e.g., tofu). Cost-reduction strategies were the only strategy type that appeared to consistently protect against diet abandonment. These strategies appeared to help people with cost concerns and other barriers cut out animal products and continue their transition to veg*nism.
3. Strategies to increase or maintain one’s motivation to continue the veg*n diet were also very helpful: for instance, learning about farmed animals or about social justice, health, or religious reasons for veg*nism. Even in this sample of participants who tended to be highly motivated from the beginning, using these motivational strategies was associated with cutting out more animal products and in some cases, protecting against diet abandonment. Notably, they were identified as helpful for combating barriers of low motivation and negative beliefs about society’s perceptions of veg*nism.
4. Health-promoting strategies were useful, but feeling unhealthy remained a challenging barrier. Using health-promoting strategies, such as talking to a medical professional about how to be healthy on a veg*n diet or researching it oneself, appeared to help people with a range of barriers cut out animal products and get closer to their veg*n goals. However, these strategies did not appear to protect against diet abandonment, which we identified above as a risk for people who were feeling unhealthy on their veg*n diet. This suggests that feeling unhealthy remains a difficult challenge to overcome, though using health strategies in combination with other strategies that reduce the risk of diet abandonment (cost and motivation strategies) may be protective.
5. Social strategies were helpful for people with one or more social barriers: For people who were experiencing low autonomy support (support from friends and family), negative influence from one’s culture, or a small network of other veg*ns, social strategies helped them cut out animal products and get closer to their consumption goals. In general, social strategies are about creating a supportive network for yourself by meeting new people and requesting support from important people: for instance, joining an online veg*n community or asking friends and family to be supportive.
6. Strategies targeting ability barriers were somewhat effective, but not for people who needed the help most. These strategies included attempts to improve access to veg*n food or ability to prepare it, such as researching products, switching grocery stores, or increasing one’s own cooking. Crucially, while these strategies were helpful to people with some unrelated barriers, they had no apparent effect on people who were actually having difficulty finding or preparing veg*n food or having low personal control over food. This finding highlights the fact that individual-level solutions to structural problems of access and ability are not simple and may not even exist.
7. Though barriers can be a challenge, many people are able to lessen or overcome them with time. For instance, at the beginning of the study, just 36% of participants identified strongly as veg*ns, and 46% believed society sees their diet positively. Six months later, 66% identified strongly as veg*ns and 67% believed society sees their diet positively. Similar results occurred across many of the other barriers as well.

Recommendations

1. Encourage all new veg*ns to set themselves up for success by thinking about strategies they can use when they face challenges. New veg*ns are more successful when they use multiple strategies regardless of what those strategies are, so don’t be shy about trying things! Previous research on goal pursuit has also shown that personal strategies people come up with themselves can be more effective than “expert” strategies, so we recommend encouraging people to try strategies that they think might work for them regardless of whether or not they appear on our list (Peetz & Davydenko, 2021).
2. Use Table 10 to suggest strategy types according to which barriers they help with. Some barriers are harder to combat than others, but knowing which types of strategies are associated with success for people with those barriers is the first step in overcoming them.
3. Acknowledge, validate, and continue to research challenging barriers that don’t respond easily to personal strategies. Feeling unhealthy was associated with greater likelihood of abandoning one’s diet, and although the health strategies we measured helped people get closer to their diet goals, they didn’t help with diet abandonment. Health issues can’t be dismissed or ignored, so we encourage additional research into what’s needed to help people who feel healthy on a veg*n diet. As always, we encourage advocates to meet people where they are on their veg*n journey and support any positive changes they are able to make, without judgment.
4. Advocate for equal access to affordable and healthy veg*n food. If a person or group lacks access to healthy and affordable food in general, the additional demands of a veg*n diet may be a challenging hurdle. Groups like the Food Empowerment Project have written extensively about the systematic lack of access to healthy and affordable food in many areas of the United States. We encourage vegan advocates to examine whether their region is affected by this structural issue and to lobby politicians and/or corporations for food justice: for instance, by creating stores providing affordable and healthy food in all neighborhoods, or by lobbying for plant-based options in food kitchens/pantries, other hunger relief programs, and publicly funded institutions. There are also helpful guides available about how to improve access to healthy food in underserved communities.

Applying These Findings

We understand that reports like this have a lot of information to consider and that acting on research can be challenging. Faunalytics is happy to offer pro bono support to advocates and nonprofit organizations who would like guidance applying these findings to their own work. Please visit our Office Hours or contact us for support.

Other Reports From This Study

The purpose of this study is to provide solid data for advocates about how to help new veg*ns maintain their change of lifestyle. This is the third report in a three-part series. Previously:

• The first report focused on overall levels of success and described the variety of ways that people transition to veg*nism.
• The second report looked at how people’s motivations and influences for starting the diet related to their success over the first six months.

This project has produced a huge amount of data, all of which will be posted on the Open Science Framework once we have completed our own analyses and publications. In the meantime, if you have additional research questions that you would like us to consider, please contact [email protected].

Research Team

The project authors are Jo Anderson (Faunalytics) and Marina Milyavskaya (Carleton University). However, this project was a massive undertaking and could not have happened without the support of multiple individuals and organizations.

We are very grateful to Faunalytics volunteers Renata Hlavová, Erin Galloway, Susan Macary, and Lindsay Frederick for their support and assistance with this work, as well as former Carleton student Marta Kolbuszewska and the dozens of animal advocates who helped with recruitment. We are also very thankful to Animal Charity Evaluators, the Social Science and Humanities Research Council (SSHRC), and VegFund for funding this research. Finally, we thank all of our survey respondents for their time and effort.

Method Overview

This project focused on the experiences of new vegans and vegetarians (for simplicity referred to collectively as veg*ns in this report) in the U.S. and Canada. Participants were asked to complete a survey when they signed up to participate, as well as six follow-up surveys that were sent monthly over the next six months.

Participants’ demographics were quite representative of the general population, but we also weighted the descriptive results to be even closer to the U.S. population. For more details about the study method, see the first report.

Representativeness, Weighting, And Attrition

The sample for this study was 222 participants, which a pre-registered power analysis showed is more than sufficient to detect significant effects in the regression analyses that we used to investigate our main research questions. While the sample is smaller than you may be used to seeing in many Faunalytics studies, larger samples are generally for studies where one of the main goals is to estimate population statistics. A sample of about 1,000 people provides a 3.1% margin of error, while this current study has a margin of error of 6.6%. While this wouldn’t be great if estimating population statistics were our main goal, a smaller sample is necessary for our key research questions, as noted above. You can read more about margin of error in the Research Advice section of our website.

To ensure that this sample is as representative of new veg*ns as possible, we followed a pre-registered plan of comparing them against the larger sample (n = 11,399) of veg*ns from Faunalytics’ 2014 study. We were pleased to find that the current sample matched most of those demographics well already, but to maximize the representativeness, we weighted the descriptive results to match.

Overall, 65% of participants completed the entire study. We examined the characteristics of people who left the study and found no evidence of significant differences between the people who did and did not complete it (differential attrition). This examination is described in detail in the first report.

Analysis Method

The predictive analyses in this report were preregistered on the Open Science Framework. However, while we had indicated that we would use participants’ barriers at the beginning of the study (baseline; BL) to predict their success at the final follow-up (Follow-Up 6; FU6), this approach excluded too many participants for some of the statistical models to run, due to people who had just started their diets at baseline being unable to report on their barriers.

To avoid excluding participants who didn’t answer the barrier questions at baseline, we used their responses from the first follow-up survey instead when available.

Terminology

Throughout this report, if you encounter research terms you don’t know, consider checking the Faunalytics Glossary for definitions and examples.

Results

This study’s pre-registration, survey instruments, analysis code, and data are available on the Open Science Framework.

Measures Of Success

The measures of dietary success that we used in this study are described in far more detail in the first report, but in short, we looked at how successful people were with their new diets in three ways:

1. Diet Maintenance Vs. Abandonment: Whether participants maintained or abandoned their new veg*n diet during the six months of the study.
2. Consumption Success: How close people’s actual diet was to their goal diet each month, in servings of animal products. (For instance, someone working toward a vegan goal intends to eat 0 servings of animal products per month, so if they are still eating 3 servings, their consumption success is worse than someone who has cut down to 1 serving per month.)
3. Felt Success: How successful people felt with their dietary goal each month, on a scale of 0 to 100.

Barriers And Supports For Successful Diet Change

This section describes key barriers and supports to participants’ dietary goals—things that help them achieve their goals or get in the way. Note that we consider most barriers and supports to be opposite ends of a spectrum rather than entirely separate ideas. For instance, identifying strongly as a vegan or vegetarian may be helpful to success while not identifying much with veg*nism may be a barrier.

This section also shows how participants’ barriers/supports changed over the course of the study, using latent growth modelling to examine whether there was a significant upward or downward trend (see Supplementary Materials for details). Bear in mind that the overall frequency of experiencing barriers was quite low in this sample of participants. It would be reasonable to expect more barriers/fewer supports in the general population.

Perceived Healthiness On Diet

Many lapsed veg*ns cite health concerns as reasons for abandoning their diets: everything from concerns about nutrient deficiencies to feeling fatigued to not noticing the benefits they had expected (Faunalytics. 2014). Anticipating that this could be a barrier, we asked participants “How have you felt overall?” to get at their perceptions of how healthy they were—though we aren’t able to say whether those perceptions map onto actual health issues. Response options ranged from 1, very unhealthy, to 5, very healthy, and responses of 1 or 2 were considered a barrier.

The figure below shows how perceptions of healthiness changed over time. Participants’ average feelings of healthiness did not significantly increase or decrease over the course of the first six months (common slope p > .19).

Figure 1. Perceived Healthiness Over Time

Strength Of Identification With Goal Diet

Do you think of your diet as a behavior (not eating meat or animal products) or as an identity (being vegetarian or vegan)? Chances are that many readers of this report are strongly identified with a diet or even a lifestyle of veganism, but when you first transitioned away from eating meat, that new identification may not have happened immediately.

This is important for several reasons. First, when a behavior is part of your identity, it is more likely to continue. For instance, research has shown that people who see themselves as health-conscious or “green” consumers are more likely to eat less animal fat and purchase organic products, respectively (Carfora et al., 2019; Sparks & Guthrie, 2006)—and importantly, this identity predicts behavior even taking into account key predictors like attitudes and social norms.

Similarly, Faunalytics’ 2014 study found that seeing veg*nism as part of one’s identity was much less common in people who gave up their veg*n diet than in people who maintained it: 58% of former veg*ns said their diet wasn’t part of their identity even while they were eating that way, versus just 11% of current veg*ns.

In this study, we used a visual measure of how much people see veg*nism as part of themselves, shown in the figure below. Similar tools have been reliably used to measure how a range of goals, thoughts, objects, and even people become part of a person’s self-concept (Hatvany, Burkley, & Curtis, 2017). The first two responses were considered to indicate a barrier.

Figure 2. Measure of Identification with Veg*nism

The figure below shows changes in strength of identification with the goal diet over time. Statistical modelling indicated a significant increase in participants’ average identification with veg*nism over time (common slope p < .001). In the figure, this is most noticeable in that the proportion of people who were strongly identified with veg*nism increased from 36% at the beginning of the study to 66% six months later.

Figure 3. Identification With Diet Over Time

Societal Perceptions Of Diet

Apart from how your friends and family think about your choices, many people may be influenced by how they think society at large perceives it. In this study, we measured how people think their diet is seen by society using a single-item measure: “How do you feel your goal diet is seen by society?” Response options ranged from 1, it is seen as very negative, to 5, it is seen as very positive, and responses of 1 or 2 were considered a barrier.

The figure below shows how beliefs about society’s perceptions change over time. Statistical modelling indicated a significant positive shift in how participants believed society sees veg*nism over time (common slope p < .001). You can see this clearly in the proportion of people who believed society sees their diet positively, which increased from 46% at the beginning of the study to 67% six months later.

Figure 4. How Veg*ns Think Their Diet Is Seen By Society Over Time

Autonomy Support

Social support is very important to making and sustaining a positive change. “Autonomy support” is the specific name given to the important feeling that one’s choices are supported by friends and family—in this case, that they support your decision to go veg*n (Williams et al., 2006). We measured this feeling using a version of a standardized, six-item scale that we adapted to the context of veg*n diet change (e.g., “I am able to be open with my family and friends about my dietary goals”). Response options for each statement ranged from 1, strongly disagree, to 5, strongly agree. These were then averaged together, with higher scores indicating more autonomy support. For the purpose of barrier analysis, we considered responses that were one standard deviation below average to be representative of a barrier (scores at or below 3.1).

The table below shows respondents’ level of autonomy support when they first started their new veg*n diets: the percentage who agreed or strongly agreed with each of the six items, followed by the average score (out of 5) representing how supported participants felt at the beginning. Higher scores indicate a stronger feeling of support.

Table 2. Autonomy Support When First Going Veg*n

The figure below shows how autonomy support changes over time. Overall, we found a slight but statistically significant increase in participants’ average level of autonomy support over time (common slope p < .05). In the figure, this is most noticeable between baseline and the first follow-up, when the proportion of people not experiencing low autonomy support increased from 82% to 88%.

Figure 5. Autonomy Support Over Time

Cultural Influence On Ease Of Following Diet

Culture is another aspect of one’s social environment that influences the ability to follow a diet. Not only does it play into the societal perceptions measured above, many cultures have a strong influence on what one is expected to eat. We measured this with a single item: “How has your culture affected your ability to follow your goal diet?” Response options ranged from 1, made it much more difficult to follow my goal diet, to 5, made it much easier to follow my goal diet, and responses of 1 or 2 were considered a barrier.

The number of respondents reporting a cultural influence was too low to meaningfully examine the change over time. Instead, the figure below shows the level of influence participants reported at the beginning of the study. Most participants felt that their culture did not influence the ease or difficulty of following a veg*n diet, though we suspect that this is at least partially due to the difficulty of having insight into the influence of one’s own culture.

Figure 6. Cultural Influence When First Going Veg*n

The table below shows a few notable comments from participants about cultural influence. These observations cannot be generalized to other people from the same cultures, but they provide examples of some of the cultural considerations new veg*ns may face.

Table 3. Selected Comments On Cultural Influence

Cost

The cost of veg*n food has been identified as a barrier to diet change in previous research (Faunalytics, 2014; Grassian, 2019), and people tend to think that veg*n diets are expensive (e.g., Bryant, 2019), though research has shown that that is not always the case (Vance, 2012). In this study, we measured cost with a single item: “How did the cost of following your current diet compare to your previous diet?” Response options ranged from 1, it cost a lot more to follow my goal diet, to 5, it cost a lot less to follow my goal diet, and responses of 1 or 2 were considered a barrier.

The figure below shows people’s perception of the relative cost of their veg*n diets over time. There was no significant change over the six months (common slope p > .25).

Figure 7. Relative Cost of Diet Over Time

Extent Of Habit Formation

Part of maintaining a new behavior pattern is turning it into a habit rather than something you have to plan out. We measured whether participants had formed a habit of veg*n eating using a validated, 4-item scale known as the Self-Report Behavioral Automaticity Index (SRBAI; Gardner et al., 2012). It measures how automatic a behavior is—in this case, the behavior of choosing veg*n food.

We measured the extent of habit formation with four statements (e.g., “I choose veg*n food automatically”). These were then averaged together, with higher scores indicating a stronger habit. For the purpose of barrier analysis, we considered responses that were one standard deviation below average to be representative of a barrier (scores at or below 2.6).

The table below shows habit formation scores when respondents first started their new veg*n diets: the percentage of participants who agreed or strongly agreed with each of the four items. At the bottom is the average score, which indicates how strong the habit of choosing veg*n food was. Higher scores indicate a stronger habit.

Table 4. Strength of Veg*n Habit Formation When First Going Veg*n

The average in the table above shows that participants in this study had fairly strong veg*n habits already, while the figure below shows that those habits tended to get stronger over the first six months. Statistical modeling indicated a significant increase in the average strength of participants’ veg*n habits over time (common slope p < .001). In the figure, this is evident from the changing proportion of people who were not experiencing a habit barrier, which went from 82% at the beginning of the study to 97% six months later.

Figure 8. Strength of Veg*n Habit Formation Over Time

Dietary Perfectionism

In 2014, we found that 43% of lapsed veg*ns reported that they had begun to feel it was too difficult to be “pure” with their diets (see Companion to the Initial Findings). This raises the question of whether perfectionism may have been an obstacle to success for some participants. Research has found that not all perfectionism is problematic (Sirois et al., 2010): Having high personal standards can be motivating, while perfectionism that comes with self-blame and rumination over failures is maladaptive. Perhaps related to the high personal standards form of perfectionism, some previous research has found that people who choose to pursue veganism—which is a more “perfect” diet in terms of animal product avoidance than reducetarianism—were more likely to succeed at it (Grassian, 2019).

In this study we asked people to indicate the extent of their agreement with the idea “I can’t feel satisfied unless I follow my dietary goal perfectly.” Response options ranged from 1, strongly disagree, to 5, strongly agree. We had originally assumed that too much perfectionism would be the barrier, making people more likely to abandon their diet, but results showed the opposite, as discussed below. Because some kinds of perfectionism can be problematic, we don’t consider a lack of perfectionism to be a barrier and don’t talk about it as such in this report. Nevertheless, responses of 1 or 2 were treated like a barrier in the barrier analyses.

The figure below shows how people’s dietary perfectionism changes over time. Overall, we found that participants’ average level of perfectionism increased over time (common slope p < .001). In the figure, this can be seen in the proportion of “perfectionist” people increasing from 61% to 73% over six months.

Figure 9. Dietary Perfectionism Over Time

Ability To Find Or Prepare Food

The availability of veg*n food options is a necessary piece of a person’s success, as highlighted in a lot of previous research (e.g., Faunalytics, 2014; Grassian, 2019). Although this concept has sometimes been referred to as “convenience” in the past—by us and others—we strongly recommend using other terms so as not to minimize real problems with food availability, especially when they are disproportionately encountered by poor and racialized individuals (see Food Empowerment Project for more). In addition, the preparation of nutritionally complete plant-based meals is a skill that most people lack (Corrin and Papadopoulos, 2017), so it can also pose challenges.

We measured participants’ self-reported ability to find or prepare food with the item: “How easy or difficult was it for you to find or prepare food to fit your goal diet?” Response options ranged from 1, very difficult, to 5, very easy, and responses of 1 or 2 were considered a barrier.

The figure below shows how people’s ability to find or prepare veg*n food changes over time. Statistical modelling indicated a significant increase in average ability over time (common slope p < .01). In the figure, this can be seen in that the proportion of people saying it was easy to find or prepare food grew from 68% to 80%, as well as in the shrinking of the “difficult” proportion from 21% to 10%.

Figure 10. Ability To Find Or Prepare Food Over Time

Shame/Pride In Diet

Shame and pride are social emotions, meaning that they are rooted in how we think other people see us. In this study, we measured these emotions as opposite ends of a single scale, with the question: “Regardless of how you think you should feel, how do you generally feel when you think about your goal diet?” Response options ranged from 1, very ashamed, to 5, very proud, and responses of 1 or 2 were considered a barrier.

The figure below shows feelings of pride versus shame over time for our participants. As you can see, while there was fluctuation in responses, there was no overall upward or downward trend (common slope p > .55).

Figure 11. Pride Or Shame Over Time

Personal Control Over Food

Some people have much more control over choosing what they eat than others do. For example, university students may rely on the food options in their cafeteria, and those who live with family may rely on family members for meal planning or preparation. Our 2014 research found evidence that lack of personal control over food posed difficulties for a substantial proportion of current and former veg*ns, so these individuals may have a harder time transitioning to veg*nism.

We asked participants to indicate the percentage of control they felt over the content of their meals. Response options ranged from 0% to 100%, we considered responses that were one standard deviation below average to indicate a barrier (scores at or below 71.4).

The figure below shows how personal control over food changes over time. On average, we found that participants reported significantly more over time (common slope p < .001). In the figure, this is most noticeable in that the proportion of people without a personal control barrier increased from 83% at the beginning of the study to 90% six months later.

Figure 12. Personal Control Over Time

Size Of Veg*n Network

As shown in Faunalytics’ 2014 study, having a support network of fellow vegans and vegetarians can be a big help to people who are just starting out. We asked participants how many people they know personally who are vegetarian or vegan. There was a wide range, with some people reporting 10 or more veg*n connections. Responses of 0 or 1 were considered a barrier for the purpose of barrier analysis, because a cut-off of one standard deviation below average includes scores at or below 1.1.

On average, our participants knew 4 other veg*ns and many knew 5 to 10 or even more. This suggests that much of our sample was relatively well-connected with a potential support network when they started—not something that is guaranteed for many new veg*ns.

The figure below shows how people’s veg*n networks changed over time.  Overall, we found that the average number of veg*ns participants knew increased significantly (common slope p < .001). In the figure, you can see that the proportion of people with a veg*n network barrier—those who knew 0-1 other veg*ns—decreased from 24% at the beginning of the study to 9% six months later.

Figure 13. Size of Veg*n Network Over Time

Progress On Health Goals

Because many people adopt veg*n diets for health reasons, it is also important to consider how those health goals are progressing. If they don’t feel that the diet is helping, why continue? For all participants who selected “health” as one of their motivations for pursuing a veg*n diet, we asked, “how has your diet affected your health goals?” Response options ranged from 1, it has interfered a lot with my health goals, to 5, it has helped a lot with my health goals, and responses of 1 or 2 were considered a barrier.

We found minimal evidence of an upward trend in average responses over time (weak model fit: CFI = .866, RMSEA: .077; marginally significant common slope: p < .06). Because most people said their veg*n diet was helping at least a little with their health goals, the figure below shows separate percentages for the people who said it helped “a little” versus “a lot” over the course of the study. As you can see, “helped a little” responses were overtaken in frequency by “helped a lot” responses over time, but this should not be taken as strong evidence of change.

Figure 14. Progress on Health Goals Over Time

Feelings Of Motivation

Feelings of motivation can wax and wane over time, and a lack of motivation has been previously identified as a reason for abandoning a veg*n diet (Faunalytics, 2014; Grassian, 2019). We asked participants “How motivated have you felt to follow or work toward your goal diet?” Response options ranged from 1, very unmotivated to continue, to 5, very motivated to continue, and responses of 1 or 2 were considered a barrier.

The figure below shows how feelings of motivation changed over time. There was little fluctuation in this highly motivated group of participants and we found very minimal evidence of an upward trend in average responses over time (weak model fit: CFI = .851, RMSEA: .101; marginally significant common slope: p < .08).

Figure 15. Feelings of Motivation Over Time

Cravings For Animal Products

Many new veg*ns experience cravings for meat or animal products. Faunalytics (2014) found that about a third of former veg*ns reported difficulties with cravings. To measure cravings, we asked participants, “How often have you had cravings for meat or animal products?” Response options ranged from 1, daily, to 5, never, and responses of 1 or 2 were considered a barrier.

The figure below shows how cravings fluctuated over time. Statistical modeling indicated a significant decrease in participants’ cravings over time (common slope p < .01). In the figure, this is most noticeable in that the proportion of people who experienced cravings less than once a week increased from 65% at the beginning of the study to 79% six months later.

Figure 16. Frequency of Cravings Over Time

Each month, we asked participants to indicate what they were craving. The table below shows the most common responses, combined across all time points. These suggest one way of identifying gaps or lacks in current plant-based offerings.

Table 5. Most Common Animal Product Cravings

Barriers/Supports & Success

The analyses in this section were conducted to examine, on the one hand, the associations between individual barriers/supports that participants experienced and, on the other, our three measures of success: diet maintenance versus abandonment, consumption success, and felt success. Full regression results are provided on the Supplementary Materials tab under Barriers/Supports & Success: Detailed Results.

Barriers To Diet Maintenance

Three barriers/supports were related to whether people continued their veg*n diets or not. Most notably, people who felt unhealthy on their veg*n diet were significantly more likely to abandon it within the first six months than those who felt healthier: Close to 92% of people who didn’t have a health barrier maintained their diets for at least six months, versus just 70% of people with this barrier. As you will see in a subsequent section, it did not matter whether the individual’s primary motivation was health.

Additionally, people who did not see veg*nism as part of their personal identity and people who thought society perceives veg*nism negatively were marginally more likely to abandon it within the six months.  These findings are shown in the table below.

Asterisks (*) indicate statistically significant predictors, while those with a dagger (†) are marginally significant. That means that although they didn’t reach the traditional level of significance (p < .05) after correcting for False Discovery Rate (FDR), they were close enough (p < .10) to include as possibly important.

Table 6. Barriers To Diet Maintenance

Barriers To Consumption Success

Consumption success—how close people were to their goal level of consumption—was significantly affected by five supports/barriers, and there was marginal evidence for the influence of another three, as shown in the table below.

People were the furthest from their goal level of consumption at Follow-Up 6 when they were unconcerned about following the diet perfectly, when their culture made going veg*n difficult, when they had low autonomy support, when they didn’t have a strong habit of choosing veg*n food, and when their veg*n diet was costing more than their previous diet.

Marginal predictors that may (with lower certainty) make it more difficult to reach one’s veg*n goal included feeling ashamed of one’s diet, having difficulty finding or preparing veg*n food, and feeling unhealthy on the diet.

Although dietary perfectionism (not feeling satisfied unless you follow your diet perfectly) was the most strongly associated with success, we do not recommend you think of the opposite as a barrier to be overcome because some types of perfectionism are harmful (Sirois et al., 2010). We had included this measure in the study expecting too much perfectionism to be a barrier rather than a support, so while it’s great that it appears to work for people who choose it, we recommend against telling people to be more perfectionistic in their approach to veg*nism. For a longer discussion of this issue, see the Conclusion section.

Table 7. Barriers To Consumption Success

Barriers To Felt Success

Participants also tended to feel more successful when they had supports rather than barriers. As shown in the table below, three of these associations were statistically significant after FDR correction (all ps > .05).

People felt the least successful when they had low personal control over their food, when they had difficulty finding or preparing veg*n food, and when being veg*n wasn’t a strong part of their identity. Interestingly, while the latter two feelings reflect real struggles with consumption success or diet abandonment, low personal control was not associated with either. In other words, having low personal control over your food choices—for instance, being reliant on a family member to shop and cook—made people feel significantly less successful but they didn’t appear to be any less successful than anyone else on more objective measures.

Table 8. Barriers To Felt Success

Relative Importance Of Supports/Barriers

Below, we have ranked the apparent importance of the supports/barriers that we measured in this study. This ranking, based on the results above, uses the assumption that abandoning one’s diet is the worst possible outcome, not reaching one’s goal level of consumption is less concerning but still a bad outcome, and feeling unsuccessful is unfortunate but the least problematic in terms of outcomes for animals.

Table 9. Importance of Supports/Barriers, From Most to Least

Bear in mind that just because some of these barriers were not associated with success in this study does not mean that individual people are never affected by them. As noted in our first report, these participants were, on average, more committed to their veg*n dietary goals than we would expect to see in the general population. Some of the lower-importance barriers for them could be more crucial for less committed individuals. More generally, findings like this can help us focus on common experiences but should never be used to dismiss or devalue individual experiences that differ from the norm.

Useful Strategies To Support Diet Change

Having now covered support and barriers individually, the question is how and when strategies to support diet change are helpful. This was a complex set of analyses, so we have described the full method on the Supplementary Materials tab, in the section entitled Useful Strategies To Support Diet Change: Detailed Results.

The table below shows which strategies were useful for people with different barriers, while the sections following it describe each type of strategy in more detail. For the full list of strategies and strategy types, see Table 20 in the Supplementary Materials.

Cell color indicates usefulness:

• Cells shaded in blue indicate that this type of strategy was helpful for people with this barrier. For example, social strategies were helpful for people with many types of barrier, including those who were feeling unhealthy on their new veg*n diet.
• Cells shaded in amber indicate that this type of strategy was not helpful for people with this barrier. For example, social strategies were not helpful for people who don’t identify strongly as veg*n.

Icon indicates what the strategy was useful for (if anything):

• A door () indicates that this result applied to diet maintenance vs. abandonment.
• A veggie burger () indicates that this result applied to consumption success (i.e., reducing servings of animal products).
• A head with heart () indicates that this result applied to participants’ felt success.

Methodological note: Results include both significant and marginally significant findings (ps < .10) after FDR correction. For simplicity, we have described both helpful main effects and interactions as helpful (blue), and both null effects and unhelpful interactions as not helpful (amber). Full details of these results are available on the Supplementary Materials tab.

Table 10. Effectiveness of Strategies by Barrier

Barriers	Strategy Types
	Lower cost	Increase motivation	Improve health effects of diet	Increase social support	Increase ability to follow diet	Deal with cravings
Feeling unhealthy on the veg*n diet
Low identification with veg*nism
Believing society perceives veg*nism negatively
Low autonomy support
Cultural influence making it more difficult to go veg*n
Cost
Weak habit formation around choosing veg*n food
Difficulty finding or preparing veg*n food
Low personal control over food
Small veg*n network
Low health goal progress
Low motivation
Frequent cravings for animal products

Strategies To Lower Cost

Cost strategies were among the most useful, given that they were helpful for people with a wide range of barriers and were the only strategy type that appeared to consistently protect against diet abandonment.

As shown in the table below, the most promising cost strategy was researching low-cost products that fit one’s diet (e.g., tofu). Using this strategy more frequently was significantly associated with reduction of animal product servings.

Table 11. Individual Cost Strategies

Strategies To Increase Motivation

Motivation strategies were also useful—even in this sample of highly motivated participants—in that they appeared to help people with a wide range of barriers and to protect against diet abandonment for those with some barriers, notably including the barriers of low motivation and negative beliefs about society’s perceptions of veg*nism.

As shown in the table below, there were many promising motivation strategies that were associated with getting closer to one’s veg*n goal. These included learning about farmed animals and seeing how poorly they’re treated, as well as learning about social justice, health, cost savings, and religious or spiritual reasons for following a veg*n diet.  The association of learning more about meat’s environmental impact with success did not attain statistical significance, though it’s possible that this is a quirk of our sample.

Table 12. Individual Motivation Strategies

Strategies To Improve Health Effects

Health strategies appear to be quite useful, in that they helped people with a range of barriers reduce their animal product consumption to get closer to their veg*n goals. Importantly, they were still helpful for individuals with some of the most challenging barriers. However, health strategies did not appear to protect against diet abandonment, which we identified as a risk for people who were feeling unhealthy on their veg*n diet. This suggests that feeling unhealthy remains a difficult challenge to overcome, though using health strategies in combination with other strategies that reduce the risk of diet abandonment (cost and motivation strategies) may be protective.

As shown in the table below, the most promising health strategies were informational: researching how to be healthy on your veg*n diet and talking to a medical professional about it were both associated with getting closer to one’s veg*n goal. This suggests that other informational strategies to learn more about particular health concerns for those who are feeling unhealthy could also be useful, though we did not test them in this study.

Table 13. Individual Health Strategies

Strategies To Increase Social Support

Social strategies appeared to help people with a wide range of barriers get closer to their veg*n goals, including those with one or more social barriers: experiencing low autonomy support (support from friends and family), negative influence from one’s culture, or a small network of other veg*ns.

There were several promising social strategies, as shown in the table below. In general, social strategies are about creating a supportive network for yourself by meeting new people and requesting support from important people.

Table 14. Individual Social Strategies

Strategies To Increase Ability To Follow Diet

Ability strategies were of limited usefulness. While they appeared to help people with some barriers get closer to their veg*n goals, those goals did not include the ability barriers they are intended to overcome: having difficulty finding or preparing veg*n food and having low personal control over one’s food choices.

For barriers they did help with, some ability strategies were more promising than others. As shown in the table below, finding specific veg*n products and where to buy them was useful. It was also helpful for some people to find a restaurant or dining hall with better options, and/or increase the cooking they do themselves.

Table 15. Individual Ability Strategies

Strategies To Deal With Cravings

Cravings strategies were of very limited usefulness for our participants. They had no significant effect on success for people with a cravings barrier and only helped for people with two social barriers that are likely better helped by social strategies.

Despite their limited usefulness, some cravings strategies were associated more strongly with reducing animal product consumption than others. Planning ahead, avoiding tempting situations, and using cognitive strategies during cravings were all helpful, while trying to fight the urge or using a plant-based substitute for the craved animal product did not appear to be.

Table 16. Individual Cravings Strategies

Using Multiple Strategies Is A Good Strategy!

The findings in this section are based on correlational data—that is, they show that using certain strategies tends to be associated with success. However, we can’t say for sure that strategy use caused that success, and the biggest reason is that people who use one type of strategy a lot also tend to use other ones. That can make it difficult to tell which one caused their success—and in fact, using a greater number of strategies was also associated with greater consumption success (p < .04), regardless of which specific strategies were used (i.e., controlling for them). On average, people who reached their veg*n goal within six months were using 29.7 strategies, while people who were 10 or more servings from their goal were using just 26.5 strategies. While this is a significant difference, it’s also worth noting that everyone in this study was using a large number of strategies, which likely goes hand in hand with the high level of commitment we’ve pointed out in each report. It may also be a big part of the reason for the overall high level of success we observe in this study.

Strategy Use Over Time

The figures below show how frequently each strategy was used over time. They are presented with different types in separate figures only for readability, to make it easier to see all of them.

Please note that the wording of some strategies has been simplified for these figures. The original wording is available in the above table.

Figure 17. Use Of Strategies To Lower Cost

Figure 18. Use Of Strategies To Increase Motivation

Figure 19. Use Of Strategies To Improve Health Effects Of Diet

Figure 20. Use Of Strategies To Increase Social Support

Figure 21. Use Of Strategies To Increase Ability To Follow Diet

Figure 22. Use Of Strategies To Deal With Cravings

No Evidence For Influence Of Initial Motivation 

In the second report, we noted that the most common primary motivations to go veg*n were health (42%), animal protection (20%), and environmental concern (18%), but that these general motivations did not have any effect on how successful people were with their diets.

In this report, we took the analysis further by examining two questions: Whether people with different initial motivations tend to experience different barriers, and whether we should be suggesting different strategies to people with different initial motivations. We found no evidence for either possibility, suggesting again that the reasons people pursue their veg*n diet are not that impactful on their success as long as they are committed to it.

Methodological note: For both questions, we adjusted p-values for FDR at the level of a given motivation: that is, across all analyses examining the motivation’s association with the 15 barriers in the first case and all analyses examining the motivation’s association with the 6 types of strategy in the second. Additional details are provided in the Supplementary Materials.

Conclusions

Barriers

This study examined a number of barriers to veg*n diet change that have been identified in previous research, with the goal of determining how they influence success over a moderate time period—the first six months of one’s new diet. 

We identified three barriers as the most problematic because people who experience them when first trying to go veg*n were more likely to abandon the attempt within the first six months. Those three were:

1. Feeling unhealthy on one’s veg*n diet,
2. Not seeing veg*nism as part of one’s identity, and 
3. Believing that society sees veg*nism negatively.

Additional problematic barriers were those associated with having more trouble reaching one’s goal level of consumption, which included:

1. Low autonomy support from friends and family, 
2. Negative cultural influence, 
3. Weak habit formation,
4. High cost,
5. Being ashamed of one’s diet, and
6. Difficulty finding or preparing food.

Dietary perfectionism—not feeling satisfied unless following one’s new diet perfectly—was also associated with consumption success, such that people who were more perfectionist tended to get closer to their goal. However, we do not refer to low perfectionism as a barrier because while it may work well for people who chose it freely in this correlational study, perfectionism has a dark side and recommending it as a strategy could have harmful consequences (Sirois et al., 2010). It should be tested experimentally before being considered a potential strategy.

For full details, see the section Barriers and Supports for Successful Diet Change.

Strategies 

Recommending strategies for diet maintenance can be as simple or as complicated as you like. At the simplest level, using more strategies and using them frequently is helpful: Just using strategies more often was predictive of consumption success. We also know from previous research into other types of goal pursuit that personal strategies people come up with themselves can be more effective than “expert” strategies (Peetz & Davydenko, 2021), so it’s a good idea to encourage people to try strategies that they think might work for them regardless of whether or not they appear on our list.

But it’s also possible to get a lot more specific and take account of an individual’s particular barriers, as outlined below.

Cost Strategies

Regardless of the barriers a person was experiencing, cost strategies were often associated with a lower likelihood of abandoning one’s veg*n diet. This suggests that even when people don’t identify cost as a concern or are dealing with other barriers, having affordable plant-based options available is important for diet maintenance. 

Cost strategies included four individual strategies, of which one was most promising for success: Researching low-cost products (e.g., tofu). This doesn’t mean that the others aren’t useful—they certainly may be, especially for some people or in combination with other strategies—but if you are looking for a particular cost strategy to recommend to someone, helping them find low-cost products is the best option.

Motivation Strategies

Strategies for increasing motivation were effective for people with a range of barriers and were sometimes associated with a lower likelihood of abandoning one’s veg*n diet, including for people who suffered from low motivation. These strategies appeared to help people with low motivation cut out animal products and make them less likely to abandon their diet.

For those considering motivation strategies to help with low motivation or for other reasons, any of them may help, but the following were the most promising:

• Learn more about animals that are used for food
• Learn more about world hunger or social justice reasons for following a veg*n diet
• Watch unpleasant or graphic images/video of farmed animals
• Learn more about religious/spiritual reasons for following a veg*n diet
• Learn more about health benefits of following a veg*n diet
• Learn more about saving money by following a veg*n diet

Health Strategies

Strategies for improving health effects were moderately effective. They helped people who had several different barriers, including people who were feeling unhealthy on their diet, get closer to their goal level of animal product consumption. However, these strategies did not appear to protect against diet abandonment, which is a key risk for people who feel unhealthy on their veg*n diet. This suggests that feeling unhealthy remains a difficult challenge to overcome, though using health strategies in combination with other strategies that reduce the risk of diet abandonment (cost and motivation strategies) may be protective. 

Of the strategies we considered, two were identified as more promising than the rest:

• Research how to be healthy on a veg*n diet
• Talk to a medical professional about your diet

For the latter, however, we encourage advocates to let people know that not all medical professionals are up to date on the health benefits of plant-based diets, despite a wealth of evidence and direct recommendations to physicians to advise them (Tuso, 2013).

Social Strategies

Social strategies were helpful for people who were experiencing most of the barriers we measured, making them the most flexible type of strategy. Most notably, they were helpful for people with the social barriers of low autonomy support (support from friends and family), negative influence from one’s culture, or having a  small veg*n network. Social strategies helped individuals with those barriers cut out animal products and get closer to their veg*n goals.

Unfortunately, however, social strategies were less effective for people who don’t identify strongly as a veg*n, suggesting that advocates may need to suggest other strategies and find ways to increase identification. This is somewhat surprising, as we might expect that spending more time around other veg*ns would increase that identification.

For those considering social strategies, using any could help, but the following were the most promising:

• Participate in an online community (e.g., Facebook group) for people with diets similar to yours
• Ask your family or friends to be supportive of your diet 
• Try to meet new people with diets similar to yours
• Avoid people who are unsupportive or critical of your diet
• Explain to your family or friends why this diet is important to you

Ability Strategies

Strategies for improving one’s ability to follow the diet were somewhat effective, helping people with a range of different barriers be successful. However, they had no apparent effect on people who were experiencing the ability-related barriers of difficulty finding or preparing veg*n food or having low personal control over food, indicating the challenge of overcoming these practical problems. Other research highlights the problem of systemic lack of access to healthy and affordable food in many areas (see Food Empowerment Project), and this finding further illustrates that individual-level solutions to these problems may not exist.

For individuals who do have choices available, the following ability strategies can be recommended as the most promising:

• Research products (e.g., meat alternatives) that fit your diet
• Switch to a restaurant, dining hall, etc., with better options for your diet
• Switch to a grocery store with better options for your diet
• Eat products that are designed as meat replacements (e.g., veggie burger, soy chick’n)
• Increase the amount of cooking you do yourself

Cravings Strategies

Strategies for dealing with cravings were less useful than the rest, but some individuals may feel that they need them and may find them helpful. For those who want to try them, we recommend several strategies that were individually associated with better consumption success:

• Plan a strategy for dealing with temptation if it occurs
• Avoid places or situations that might tempt you 
• Change the way you were thinking about a craving or a food you craved 
• Distract yourself from a craving 
• Plan meals in advance (e.g., before grocery shopping, going to a restaurant) 
• Remind yourself why you’re following this diet 

Caveats & Limitations

As with all studies, this one has some important caveats and limitations to bear in mind. In addition to the general limitations covered in the first report, there are some that are specifically worth thinking about with regard to the findings presented in this report.

Correlational Data

First, as we have previously noted, this study is not an experiment: We did not randomly assign people to use different strategies or experience different barriers (and we never will do the latter, as it would be unethical). This means that all reported findings are correlational. Because the study is longitudinal, with data collected over a six-month period, we are more able to draw conclusions about what caused what than we usually can with correlational data—something reported on the last survey could have been caused by something reported on the first survey, but not vice-versa—but there are still limits on interpretation. These differ for strategies and barriers, as described below.

Self-Selected Strategies

The correlational nature of the study, as described above, means that the strategies discussed in this report were self-selected by the participants—they chose how to approach their own veg*n transition and how to tackle any barriers they were experiencing.

While this tells us which strategies tend to be associated (correlated) with success, it is not possible to say for sure that they caused that success. The issue is that people who were more likely to succeed all along for other reasons may be more likely to use a particular strategy. This may seem far-fetched, but consider an example: Someone who finds going vegan easy because they never ate much meat anyway will be more successful than average. And because they don’t have to spend a lot of time thinking about things like how to stay motivated, find cheaper meat alternatives, or deal with cravings, they have more time to spend on “fun” strategies like meeting other vegans or looking up recipes. This is a form of advantage or privilege that would increase the association between their chosen strategies and success, but it isn’t the strategies that made these people successful.

Because of this limitation, it’s best to consider the findings of this study as indicative of strategies that are the most promising, but they are not proven to help everyone. For this reason, we plan to follow up this research with an experiment to get stronger casual data about some of the most promising strategies.

Not All “Barriers” Should Be Addressed: Don’t Recommend Perfectionism

When it comes to what we describe as barriers in this study, what we are referring to are pre-existing factors that were associated (correlated) with less success on one’s veg*n diet. In most cases, if you as an advocate have an opportunity to help people remove those barriers, it seems clearly advisable to do so. Lacking social support, finding veg*nism expensive, having low motivation—these are all barriers that it may be possible to overcome through direct solutions. Making new friends, finding cheaper products, and increasing one’s motivation are all good things. 

However, you may have noticed that throughout this report, we have included dietary perfectionism in sections referring to barriers/supports but have not included it in discussions of barriers the way we do for the others. That’s because of the correlational nature of the study coupled with the nature of perfectionism.

We found that dietary perfectionism—the tendency of people to say that they can’t be satisfied unless they follow their diet goal perfectly—was associated with greater consumption success, getting closer to one’s diet goal. This might make you think that we should recommend that people be more perfectionist in their approach to veg*nism. Instead, we caution against that. 

If someone is naturally inclined toward this approach, that’s their choice and it may help them, so don’t feel the need to steer them away from it. However, advising people that they should only be satisfied with perfection has several possible negative consequences, all supported by research: it could drive them toward harmful, self-blaming form of perfectionism, they may stop listening to you or give up on changing their diet, or it may play into negative stereotypes of vegans (Faunalytics, 2012; Faunalytics, 2019). Unless and until experimental research shows that directing people to use a perfectionistic approach to diet change is helpful and doesn’t cause harm to their mental health and/or the animal protection cause, the research does not support it as a strategy.

Limited Barriers In This Sample

The final caveat on the results in this report is that this study had a relatively privileged sample of participants: In addition to being highly committed to their diet change, as noted in every report, we also found that they reported limited barriers even at the beginning. 

While we don’t have the data to say for sure that these participants experienced fewer barriers than most people who try to go veg*n, it seems likely based on the barriers reported by former veg*ns in Faunalytics’ 2014 study. It is also plausible that this would occur in a longitudinal study because the commitment of the study itself is substantial—participants had to agree to complete seven surveys over six months, something that more committed individuals and those who had fewer barriers to juggle may be more likely to agree to. So what does this mean for the findings? There are two major implications. 

First, with limited barriers to analyze, our ability to find statistically significant results is lower, meaning that some associations that should have emerged with a more representative sample may not have. For example, maybe low motivation should have been a bigger barrier or cravings strategies should have been more helpful. We included marginally significant findings in our reporting to help address this issue, but it is still a possibility that only continued research can address.

Second, the descriptive details about how common different barriers and strategies are should not be assumed to generalize to all new veg*ns. It is very probable that the average person from the general population experiences more barriers and uses fewer strategies than the people in this study. These participants, who had the time and ability to participate in long-term research, should be considered the low-hanging fruit of new veg*ns, the easiest ones to support.

While these caveats are important to bear in mind, we don’t feel they undermine the value of the research. All research is subject to limitations, and the value of longitudinal studies is primarily in seeing how things evolve over time. The findings with respect to how barriers and strategies influence success are the most important focus of this study, and they point to many important recommendations for advocates and future researchers. As always, we advise you to read research through a critical lens and never take results as definitively proven, but we hope and believe that these findings will be useful to many people.

Barriers And Supports For Successful Diet Change: Details Of Over-Time Analyses

As noted in its introductory paragraph, the section Barriers And Supports For Successful Diet Change shows how participants’ barriers/supports changed over the course of the study with graphs and analyses. The graphs show the barriers/supports in useful categories, but the analyses themselves treated the data as continuous.

We used latent growth modeling with the lavaan package in R, adjusting for missing data with full information maximum likelihood modeling (FIML) to examine whether there was a linear trend in the barriers/supports experienced over the seven study time points. To do so, we conducted one analysis per barrier/support, using its numeric form. While not technically correct for ordinal variables, we chose this simplification over excluding all missing data, as lavaan’s latent growth model function is not yet set up for ordinal data.

We fixed the intercept weighting at each time point to 1, and the slope weightings to linear coefficients from 0 through 6. We examined the results for the fit of the overall model and the significance of the common slope. We considered the fit to be adequate for all models except those indicated in the report body. All CFI values not otherwise indicated were greater than .91, and all RMSEA values not otherwise indicated were less than .08.

Barriers/Supports And Success: Detailed Results

In the section Barriers/Supports & Success, we describe how barriers/supports were associated with diet maintenance versus abandonment, consumption success, and felt success. These were obtained using 15 bivariate regression analyses per dependent variable: one with each support/barrier as the predictor.

Results are shown in the tables below. P-values were corrected for False Discovery Rate (FDR) at the level of dependent variable, as shown in the last column.

Table 17. Diet Maintenance Vs. Abandonment Predicted By Each Support/Barrier

Table 18. Distance from Goal Predicted By Each Support/Barrier

Table 19. Felt Success Predicted By Each Support/Barrier

Useful Strategies To Support Diet Change: Detailed Results

In the section Useful Strategies To Support Diet Change, we describe which strategies appear to help with which barriers. The analyses were too complex for most of our readers to be able to action, so we simplified the results considerably in the body of this report.

A full description of the strategies, strategy types, and analyses is provided below.

Strategies & Strategy Types

On each survey, participants were asked about six types of strategy they might have used, with between 4 and 11 individual strategies (including an open-ended ‘other’ option) per type. They checked off any that they had done in the past month, or selected none of the above (not shown). All of these strategies we presented to participants are shown in the table below, in the type groupings used.

Table 20. All Individual Strategies By Type

Analysis Overview

We began by testing a pre-registered hypothesis that strategies matching a particular barrier would reduce the detrimental effect of that barrier on success, applying an FDR correction across the 9 estimates obtained per barrier-strategy combination.

After finding partial support for the matching hypothesis, we proceeded to exploratory analyses to examine all possible combinations of barriers and strategies. To be cautious in our inferences, we made FDR corrections across the 54 estimates obtained per strategy type in these exploratory analyses. See sections below for more details.

Ideally, we would have liked to conduct multivariate analyses to tease apart the effects of different types of strategy. Unfortunately, strategy types were correlated—people who used one type frequently tended to also use other types. While the correlations themselves were moderate, ranging from r = .36 to r = .60 (all ps < .0001), they produced substantial multicollinearity in the models, particularly once interaction terms were introduced (several VIFs > 5 or > 10). As a result, we pivoted to an approach in which we analyzed each strategy-barrier pair individually. While this means that some effects are likely driven by shared variance with other, untested predictors, partialled estimates from multivariate analyses would be too unstable to be trusted.

Defining Barrier Variables 

In this study, we measured supports/barriers primarily using bipolar ordinal scales. However, before beginning analyses we dichotomized the variables so that we could define people as having barriers or not. We used the two most negative scale points for ordinal variables (e.g., very unmotivated and unmotivated) and set a cut-off one standard deviation below the mean for continuous variables.

For the strategy tables in the body of the report, we used participants’ barriers at the beginning of the study. Specifically, we used their response from the baseline survey if it was available, but if not (because people who had just started their new diet were not asked about barriers at baseline) we used their response from the first follow-up survey. Cases with missing data on both of the first two surveys were left as missing.

Defining Strategy Variables

All individual strategy variables measured are listed in Table 20 above. As a reminder, in each survey, participants were asked about six types of strategy they might have used, with between 4 and 11 specific strategies (including an open-ended ‘other’ option) per type. They checked off any that they had done in the past month, or selected none of the above for that group.

To examine the impact of strategy use, we primarily relied on participants’ frequency of using a particular type of strategy across the study period. This was calculated as the number of times over the full six-month period that a person used any strategy of that type (e.g., any social strategy) divided by the number of surveys they completed (i.e., the maximum number of times they could have reported using the strategy).

Translating The Results Below Into Table 10

Multiple tables of complex results are reported in the two subsequent sections, but in the report body, strategies are simply described as either “helpful” (blue) or “not helpful” (amber). Strategies were considered helpful for a particular barrier if they showed one of two patterns with significance or marginal significance (p < .10 after FDR correction): an interaction such that people with the barrier benefited from using the strategy to a greater extent than people without the barrier (e.g., as observed for autonomy support and social strategies in pre-registered analyses; interaction p < .001; see Table 21); or a main effect of strategy use such that people who used it more benefited regardless of whether or not they had the barrier (e.g., as observed for societal perceptions of veg*nism and social strategies in pre-registered analyses; main effect p < .001; see Table 21).

Strategies were considered not helpful for a particular barrier if they showed one of two patterns with significance or marginal significance (p < .10 after FDR correction): an interaction such that people without the barrier benefited from using the strategy to a greater extent than people with the barrier (e.g., as observed for the effect of cost and cost strategies on consumption success in pre-registered analyses; interaction p < .001; see Table 21); or a null effect (e.g., as observed for personal control and ability strategies in pre-registered analyses; p = .22; see Table 21). In cases where both helpful and unhelpful effects were found (such as for cost and cost strategies), we reported the positive effect because none of the unhelpful effects were negatively associated for people with barriers, just null.

Where there is overlap in the pre-registered and exploratory analyses, the adjusted significance values from the pre-registered analyses were used for reporting, as that was the purpose of pre-registration.

Hypothesis Testing

We began with a pre-registered hypothesis that strategies matching a particular barrier would reduce the detrimental effect of that barrier on success. We adjusted p-values for FDR across the 9 estimates obtained per barrier-strategy combination.

Matching strategies and barriers were pre-defined as follows:

• Social strategies with low autonomy support, negative beliefs about societal perceptions, low pride in the diet, small veg*n network, and negative cultural influence,
• Ability strategies with personal control over food and ability to find or prepare food,
• Cost-reduction strategies with perceiving the diet as costly,
• Health strategies with low feelings of healthiness and low perceived progress on health goals,
• Motivation strategies with low motivation, and
• Craving-reduction strategies with frequent cravings.

Our first series of regression analyses was set up to examine our matching hypothesis. We conducted three regressions per barrier/strategy pair: one for each of consumption success, felt success, and diet maintenance vs. abandonment. The consumption success and felt success analyses were conducted using lavaan so that we could account for missing data using FIML estimation. Logistic regressions predicting diet maintenance versus abandonment were conducted using the mice package, with missing data imputed.

Within each of those three regression analyses, there were three estimates of interest: the main effect of barrier (effect-coded presence/absence), the main effect of strategy use frequency (calculated as described under Strategy Variables above and standardized), and the interaction (multiplication) of those two terms. We corrected for FDR across the 9 estimates of interest per barrier/strategy pair (3 regression models x 3 estimates of interest per model).

The findings for each set are shown in the table below.

Table 21. Regression Results for All Matching Strategy-Barrier Pairs

Exploratory Barrier-Strategy Analyses

After the above analyses, we used the same approach of building three regression models per strategy-barrier combination, one per outcome variable, but this time ran all possible strategy-barrier combinations (e.g., autonomy support barrier with social strategy use, autonomy support barrier with motivation strategy use, autonomy support barrier with social strategy use, autonomy support barrier with health strategy use, and so on).

Given the very large number of tests and estimates involved in this exploratory process, we applied the FDR correction across all estimates for a given barrier: 6 strategies x 9 estimates = 54 estimates per barrier. This is more conservative than the approach we used for the pre-registered analyses, enabling us to maintain a Type I error rate of 5% when considering “what works” for each barrier.

The findings are shown below. Notably, there were several barriers for which these results would suggest multiple strategies worked equally well, which we suspect can be attributed to the multicollinearity mentioned above. This is particularly noticeable for autonomy support and cultural influence. Given our pre-registered hypothesis for these barriers, we suspect that social strategies drive the positive effect on these barriers.

Table 22. FDR-Adjusted Results for All Strategies with Autonomy Support

Table 23. FDR-Adjusted Results for All Strategies with Societal Perceptions

Table 24. FDR-Adjusted Results for All Strategies with Network Size

Table 25. FDR-Adjusted Results for All Strategies with Dietary Perfectionism

Table 26. FDR-Adjusted Results for All Strategies with Strength of Identification

Table 27. FDR-Adjusted Results for All Strategies with Perceived Healthiness

Table 28. FDR-Adjusted Results for All Strategies with Motivation

Table 29. FDR-Adjusted Results for All Strategies with Cravings