Consumers’ Perception of In-Vitro Meat in New Zealand Using the Theory of Planned Behaviour Model

: The purpose of this study was to investigate the perception of in-vitro meat (IVM) among New Zealand consumers and to understand their purchase and consumption behaviour using the Theory of Planned Behaviour framework developed in this study. An online survey questionnaire was created using the Qualtrics software to understand the perception of IVM, based on the conceptual framework. Participants ( n = 206) were recruited in this survey, and the data collected were subjected to PLS-PM analysis. The conceptual framework was tested for validity, and Goodness of ﬁt (GoF). The internal validity was assessed using Cronbach’s alpha, KMO value, inter-item correlation values ( β -coefﬁcients) and p -values. The ﬁndings suggest that variables such as environment and sustainability, health and safety, as well as current purchase and consumption behaviour have a strong relationship and a robust effect on IVM purchase and consumption behaviour. Consumers’ cultural beliefs had minimal inﬂuence on IVM purchase likelihood. Results in this study also indicated that most New Zealand consumers had neutral opinions in terms of engaging with IVM.


Introduction
In-vitro meat is a type of meat that is produced using animal cells, under laboratory conditions [1]. In-vitro meat (IVM) works on the principle of cellular agriculture where stem cells are extracted from donor animals through biopsy. A recent review by Post etal. (2020) reported on the scientific, sustainability and regulatory challenges of IVM [2]). The extracted cells usually belong to embryonic stem cells (ESC) [3], adult stem cells (ASC) [4,5], mesenchymal stem cells (MSC) [6,7] or induced pluripotent stem cells (iPSC's) [4]. These cells are cultured either by scaffolds or self-organising techniques [4] or are cultured in a sterile bioreactor. The extracted cells, under favourable conditions, undergo cell proliferation and differentiation to form myofibers, which eventually form muscle tissues. These muscle tissues combine to form skeletal muscles, which can be harvested as edible meat.
There has been extensive research on consumer perception of IVM in the past few years. Most studies on IVM perception have focused on consumers in the USA [8][9][10][11] and European countries [12,13]. There have however been a few studies on IVM with consumers from China and India [10], New Zealand [14], Australia [15], and Brazil and the Dominican Republic [16]. Consumer perception on IVM was generally found to be positive among European consumers. European consumers were more interested in engaging with IVM, with 52% of the Netherlands consumers definitely willing to try IVM [17], 68% of the United Kingdom consumers willing to eat IVM [18], and 44% of Italian consumers willing to buy IVM [19]. Acceptance of IVM by Europeans consumers [12,17,[19][20][21][22] was due to increased IVM familiarity brought about by the unveiling of the first lab-grown burger in London, 2013. A study that investigated the perception of IVM in the USA using an online Hypothesis 1. Consumers' views on the environment and sustainability have a significant effect on IVM purchase and consumption behaviour.

Hypothesis 2.
Consumers' views on health and safety have a significant effect on IVM purchase and consumption behaviour. Hypothesis 3. Consumers' cultural beliefs have a significant effect on IVM purchase and consumption behaviour.

Hypothesis 4.
Consumers' current purchase and consumption behaviour have a significant effect on IVM purchase and consumption behaviour.

Ethics Statement
This study was approved by the Auckland University of Technology Ethics Committee (AUTEC 19/68). Participants provided written and informed consent prior to commencement of the study.

Research Design
In order to explore the perception of IVM by New Zealand consumers, quantitative research was conducted using an online survey as it is a research tool that is widely used in consumer science studies, and a useful platform for collecting large amounts of anonymous responses. This study also attempts to measure behavioural intentions such as willingness to try (WTT) and willingness to buy (WTB), since the commercial success of IVM, as a sustainable alternative, lies in the WTT, WTB, and willingness to eat (WTE) IVM. These approaches were used with the USA, European and Asian consumers to determine their WTT, WTB and WTE IVM [8,10,19,32].

Online Survey Recruitment
Consumer views were collected through an online survey in New Zealand, and the survey was administered using the Qualtrics Software (Provo, UT, USA). The survey was open to all New Zealanders aged between 18 to 65 years old, irrespective of their eating habits. The survey was administered in English, from March to September 2019. Most of the participants were recruited in-person at public spaces located in central Auckland, whilst the rest were recruited online through an advertisement posted on social networking sites like Facebook, Instagram and LinkedIn.
Participation in the research was entirely anonymous, voluntary, and confidential. Participants were under no obligation to take part in the study and had the freedom to withdraw at any stage without any further questions. The participants were given a coffee voucher as a token of appreciation. The eligibility criteria to take part in the study were respondents over the age of 18 and they must be living in New Zealand. A total of 265 participants were recruited, but once incomplete surveys were omitted, 206 completed surveys were obtained for further analyses. Participants were provided with the following information prior to administering the survey: "With the increase in population, the demand for food has also increased. FAO experts suggest that by 2050, there will be a 100% increase in the demand for food. The demand for meat in particular will be about 73% higher. Rearing of ruminants such as cattle, sheep and goat are carried out on green pastures that are cleared by deforestation. Animal breeding has been employed to further improve and increase calf production (95% of cattle and 5% of beef cattle).
In the traditional meat industry animals are caged, castrated, confined to cages, and treated with antibiotics, pesticides and growth hormones, prior to being slaughtered. Most animal welfare associations are concerned about the well-being of animals and seek to prevent unnecessary animal suffering and death. The negative environmental effects associated with meat production include pollution through use of fossil fuel, animal methane, effluent waste, and water and land consumption.
In order to address the growing environmental and ethical concerns amidst the negativity associated with production and consumption of meat, the production and perception of in vitro meat (IVM) has been increasingly researched. IVM production involves the culturing of stem cells from farm animals in bioreactors by employing advanced tissue engineering techniques. IVM advantages include being environmentally friendly, requiring lower energy consumption, lowering greenhouse gas emission, lowering land and water consumption, and resulting in low carbon footprint. In addition, IVM is high in protein, low in unhealthy fats, highly sustainable, environmentally friendly, ethical and animal friendly. Hence, IVM is quickly becoming the best sustainable alternative to conventional meat."

Online Survey Procedure
The questionnaire was mostly accessed either by scanning the QR code that contains an anonymous link generated by Qualtrics. The QR code was shared through links on Facebook, Instagram, and LinkedIn. Participants used a checkbox to indicate their consent to take part in the study, before answering the survey questionnaire. Before commencement of the survey, all participants were briefed about the study and were given some information (the information pack contained general information on annual worldwide demand for food, livestock farming practices, animal welfare, and advantages of consuming IVM) on in-vitro meat to educate, rather than influence consumers. Provision of information has been proven to improve the consumer perception [13,22,33] and also result in increased WTT IVM [17,22].
Participants were first asked demographical questions, followed by questions on eating behaviour and meat consumption. In addition, questions on participants' familiarity, as well as their willingness to try/buy IVM if commercially available, were also included. Participants were asked to rate their answers on a 5-point Likert scale. Furthermore, the questionnaire also included questions on other aspects of IVM such as marketing, regulatory, ethical, and religious views that indirectly affect the perception.
The data from the survey were automatically saved on the Qualtrics website and were extracted on completion of data collection. The collected raw data were analysed using XLSTAT (version 2018.5) (Addinsoft, New York, NY, USA) and R Studio (version 1.1463) operating with R-version 3.6.1.

Data Analysis
The conceptual framework model as shown in Figure 1 used in this study was developed to understand the effect of New Zealand consumer perception of IVM purchase and consumption behaviour. As stated in the hypotheses, consumers' views on environment and sustainability, health and safety, cultural beliefs and consumers' current purchase and consumption behaviour of meat play an important role in purchase intention of IVM. Consequently, most studies in the IVM literature examined the consumer perception of IVM and its effects on purchase likelihood [8,[10][11][12][13]19,22,25,[32][33][34]. However, the framing of the research objectives, hypothesis and variables examined vary with each study.
Previous studies investigating consumer perception of IVM used other regression models such as logistic regression, mixed logic and multiple regression models [8,10,16,22,34] and other classical statistical approaches such as ANOVA [11,16]. PLS-PM can be classified as a correlational structural equation modelling, which allows the estimation of complex cause and effect modelling through the use of latent and manifest variables. In this study, PLS-PM was used since it was a more suitable method for studying the complex multivariate relationships between latent and observed variables [35]. The conceptual model was validated and measured using Goodness of Fit statistics, and provided the overall model fit for PLS-PM. The internal validity of the variables was assessed using Cronbach's α and Kaiser Meyer Olkin (KMO) values. The Cronbach's alpha value measures the internal consistency and scale reliability, whereas KMO values measure the sampling adequacy for each variable and for the model. The above statistical tests are essential for justifying results and confirming the conceptual model. However, at this point, no inter-item correlation was performed due to the numerous variables involved in this study.
Since PLS-PM does not hold any distribution assumption [36], the resampling task was performed to validate and obtain further data on the variability of parameter estimates [37]. Accuracy of the PLS-PM model was obtained by bootstrapping. This is a non-parametric procedure that measures the significance of the PLS results such as coefficients, Cronbach's alpha, HTMT and the R 2 value [38,39]. The R 2 value of the conceptual model in this study was found to be 0.777. PLS-PM analysis was carried out using the PLSPM package version 0.4.9 [35] by R Studio (RStudio, Boston, MA, USA) (version 1.1463) operating with R-3.5.1.

Sociodemographic Characteristics
Sociodemographic characteristics of the sample population involved in this study are summarized in Table 1. The demographic factors included gender, age, qualification, income, nationality, religious, eating habits and frequency of meat consumption. Most of the respondents belonged to the 18-25 years age group (73.79%), with more women participants (61.17%), meat-eaters/non-vegetarians (69%), and participants who consumed meat most of the time (48.54%). The mean and standard deviations of variables in the conceptual model are shown in Table 2. The mean values of most variables are in the range of between 2 and 3 on a 5-point scale, which correspond to poor perceptions and purchase intentions. Poor perceptions and purchase intentions have been attributed to consumer reluctance towards novel food products, especially if it is genetically modified due to techno scepticism, food neophobia [10,16], and disgust due to unnaturalness [11,40].

Reliability
The factors of the conceptual model were measured for reliability using the Cronbach's alpha value, which tested the subscales for coherent dimensions [41]. Alpha values between 0.5 to 0.7 and 0.7 to 1 are considered as moderately reliable and highly reliable, respectively [42]. Values in this study ranged between 0.57 and 0.86 (Table 2), which indicates an exceptionally good consistency [40,43].

Sampling Adequacy
The Kaiser Meyer Olkin (KMO) values were used as a measure of sample adequacy for each factor in the conceptual model [44]. Research showed that a KMO value greater than 0.5 was considered a good sample size [45]. Results from this study ( Table 2) showed KMO values between 0.50 to 0.90, meaning that it was acceptable [37].

Partial Least Squares Path Modelling (PLS-PM): Structural Model Assessment
PLS-PM analysis was performed to examine the relationship of variables investigated in this study. The results (Table 3) showed that all variables had positive β-coefficients. Furthermore, all variables in the conceptual model had significant p-values (p < 0.05), which indicated that all variables had a significant effect on purchase and consumption behaviour. In this study, the IVM purchase and consumption behaviour variable measure consumers' willingness to try (WTT), willingness to buy (WTB), and willingness to eat (WTE). Both IVM purchase and consumption behaviour were influenced by their views on IVM and other variables (such as environment and sustainability, health and safety, cultural beliefs and current purchase and consumption behaviour). As a result, these variables had a significant and robust relationship.
Typically, most studies examine the effect of perception on IVM purchase and consumption, but different terms were used (like absolute opposition [9] and acceptance [10,11]), and different aspects of purchase and consumption behaviour were investigated (like willingness to eat (WTE) [8,9] and willingness to purchase (WTP) [19]). A similar approach was observed in other studies [8,10,11,16,19,46]. Table 4 shows that 44% of New Zealanders were neutral about trying IVM over traditional meat (Mean = 3.31, SD = 1.05) even though IVM had a health-friendly nutrition profile (Mean = 3.42, SD = 1.05). In addition, 32% of New Zealand consumers were neither willing nor unwilling about buying IVM over traditional meat (Mean = 3.35, SD = 0.98). As seen in Table 4, the means for most questions were close to 3 on a 5-point scale, which implies that consumers had neutral opinions. Neutrality in opinion may be due to the nascent IVM technology itself. Since IVM technology is relatively new and unknown to consumers, neutral opinions are given.
Results showed that 50% of NZ consumers would purchase IVM regularly (Mean = 2.97, SD = 1.04) compared to traditional meat. Furthermore, 53% of NZ consumers would not buy IVM (Mean = 2.4, SD = 1.04) if more highly priced than traditional meat. From the results in Table 4, it is evident that New Zealand consumers have an interest in engaging with IVM, and price is one of the key barriers for their purchase behaviour. New Zealand consumers are one of the highest meat consumers in the world [47,48]. As a result, their diet is mainly meat-based and, in such cases, there will be higher meat attachment and higher inclination to try meat alternatives [10]. Thus, New Zealanders were equally interested in trying or buying IVM.
Most New Zealand consumers (36%) were unaware of IVM and only 23% were slightly aware of IVM (Mean = 2.27, SD = 1.24), indicating that the familiarity of IVM among consumers was very low. Familiarity is created when there is media exposure to IVM technology. Unfortunately, IVM awareness in NZ is not as high compared to European (UK, Belgium, France, Italy and Netherland) and other Western countries (USA and Canada). Thus, New Zealanders were less interested in trying or buying IVM. This may account for poor IVM purchase and consumption intentions amongst NZ consumers. Perception of IVM has been shown to improve when consumers (n = 525) were provided with more information on IVM, and as a result, even increased purchase and consumption behaviour (willingness to try-WTT = 54% and willingness to buy-WTB = 44.2%) [19].

Environment and Sustainability
Environmental attitude towards IVM depends on factors related to general environmental and animal welfare awareness, as well as consumers' openness towards sustainable alternatives. Overall effect of the environmental and sustainability variable on IVM purchase was significant (Table 2). There was a positive correlation between the variables (β = 0.100). In other words, a one unit increase in consumers' perceptions on environment and sustainability will lead to increase in IVM purchase intentions among consumers. Hence, the environment and sustainability variable was a significant factor that influenced IVM purchase intentions in this study. The means for questions under the environment and sustainability variable (Table 5) showed varied results. It is evident from the results that although 72% of NZ consumers are environmentally conscious (Mean = 3.83, SD = 1.01) and 70% are pro-animal welfare (Mean = 3.84, SD = 1.11), not many consumers are aware of the negative repercussions of the traditional meat industry (Mean = 2.94, SD = 1.29). As a result, most consumers do not understand the importance of sustainable alternatives and have decreased purchase intentions towards IVM, as seen in Table 5.
On the contrary, a study [9] showed that US consumers (n = 673, Mean = 1.97, 1-much more-5-much less) who were environmentally conscious agreed that IVM was an environmentally friendly alternative, and subsequently resulted in higher purchase and consumption intentions. Similarly another study [19], showed that environmentally conscious Italian consumers (n = 525) perceived IVM favourably that resulted in higher purchase intentions (44.2% said yes to willingness to buy IVM). The environmental and sustainability variable also measures consumer's general awareness and their willingness towards sustainable alternative consumption, such as plant-based meat and cell-based meats. The underlying question here was to check for food neophobia among consumers towards novel food products [9,10,49,50]. The means of questions in Table 6 showed that New Zealand consumers are not entirely open to sustainable alternatives. A study with US consumers [9,51] reported that consumers had moral absolutism (Moral absolutism is the belief that the morality or immorality of an action can be judged according to fixed standards of right and wrong) and disgust towards genetically modified (GM) foods. In addition, moral absolutism was found to be more evident among US consumers compared to European consumers [51]. The reduced openness to sustainable alternatives in this study could also be linked to subtle absolutism. Furthermore, reduced openness towards sustainable alternatives can also be due to eating behaviour. For example, consumers with high meat consumption have minimal interest in engaging with sustainable alternatives [14], as is the case with New Zealand consumers. Additionally, openness to sustainable alternatives, and consumer reluctance towards IVM is also due to other factors like unnaturalness [8], disgust, fear of perceived risks [13] and food neophobia [9,10]. The results in Table 5 show that consumers were not familiar with sustainable alternatives available in the market. In addition, NZ consumers were less familiar with IVM, cell-based meats, and other plant-based protein. Similarly, only 13% (n = 180) of Belgian consumers [22] and 14% (n = 1296) Netherlands consumers were familiar with IVM [17]. In addition, US consumers (57.3%), Chinese consumers (35.5%) and Indian consumers (25.5%) were not at all familiar with IVM (Bryant et al., 2019c). Generally, lack of familiarity with IVM among consumers is one of the primary concerns for poor purchase intentions, as previously explained.

Health and Safety
Consumer opinions on the health and safety aspects of IVM mainly focus on food safety and perceived health concerns, such as cancer and other food-related diseases. These opinions collectively form the health and safety variable, which indirectly affects IVM purchase intentions. In our study, health and safety variables are vital as they measure consumer concerns, which influence consumer perceptions of IVM. Consumers are wary about the perceived health and safety risks of IVM [13] due to unnaturalness, scientific distrust and unfamiliarity [13].
Results in Table 6 show that IVM health and safety (Mean = 2.72, SD = 0.95) had a significant effect on IVM purchase and consumption behaviour. There was a positive correlation (β = 0.316) between the factors, which suggest that perceived health and safety concerns on IVM are likely to influence IVM purchase intentions. Findings on IVM in terms of health and safety perception have also been reported in other studies [8,14,16,19,40]. Some studies showed that consumers who were not in favour of IVM due to health and safety reasons led to consumers having decreased purchase intentions [12,14].
Results in Table 6 showed that 52% of New Zealand consumers would probably not buy IVM (Mean = 2.43, SD = 0.99) and 38% were neutral about buying IVM, even if it was manufactured by government-approved agencies. Furthermore, 39% of New Zealand consumers had health and safety concerns about IVM, whereas 45% of consumers were neutral about IVM health and safety (Mean = 3.27, SD = 0.94). On the other hand, 12% of NZ consumers perceived IVM as probably having food safety risk and 41% of consumer had neutral opinions on IVM food safety risk (Mean = 2.54, SD = 0.88). These findings could be due to the nature of IVM, as it is an artificial/man-made food product using animal cells. Health and safety concerns were one of the common reasons for rejection of IVM, and this concern is widely reported in previous IVM literature [8,11,13,16,19,52]. Health and safety concerns appear to be related to the perceived unnaturalness of tissue engineering. Due to the involvement of genetic modification, IVM is susceptible to disgust, concerns on unnaturalness and fear of perceived risks [13]. IVM is often perceived as being 'unnatural' as it is often confused with genetically modified foods due to both involving biotechnology methods for production. IVM is in fact non-GMO when produced from unmodified cells extracted from animals through biopsy. However, genetic modification might be applied to increase the production efficiency of IVM in the future.

Religious and Cultural Beliefs
The religious and cultural beliefs variable measures the IVM purchase intention based on the consumers' cultural identity and beliefs. The underlying question here was to check for cultural implications on IVM as a sustainable meat alternative. The effect of religious and cultural beliefs (Mean = 2.05, SD = 1.08) on IVM purchase and consumption behaviour was significant. There was a positive correlation between the variables (β = 0.516), which suggest that these beliefs are likely to influence IVM purchase intentions.
The results in Table 7 showed that NZ consumers (Mean = 2.4, SD = 1.08) would not consume IVM, even if their religious beliefs permitted (Mean = 2.4, SD = 1.09), their religious leaders promoted IVM (Mean = 2.16, SD = 1.08), or if IVM was available as halal or kosher meat (Mean = 1.89, SD = 1.05). It is most likely that the majority of NZ consumers are giving negative answers to indicate that this is not a relevant consideration to them. Only 43.69% of the respondents identified themselves as being religious in this study (Table 1). In some religions, the intellectual level of the animal to be slaughtered for meat purposes are considered, such as "only those animals with lower intellectual capacity and pain sensation are to be slaughtered" whereas, in-vitro meat (IVM) does not require any animal slaughter at all. Do you think in-vitro meat (IVM) is a better meat alternative than conventional meat?

Current Purchase Behaviour and Consumption
The current purchase behaviour and consumption variable measures the IVM purchase intentions based on consumers current purchase behaviour and meat-eating behaviour. The underlying idea here is to test the effect of current eating and purchase behaviour on IVM purchase and consumption behaviour. The effect of current purchase behaviour and consumption (Mean = 2.39, SD = 1.14) on IVM purchase and consumption behaviour was significant. There was a positive correlation between the variables (β = 0.097), which suggests that current purchase behaviour and consumption are likely to influence IVM purchase intentions. Table 8 show that 70% of NZ consumers had higher meat intake (n = 206, Mean = 3.69, SD = 1.10), with poultry (71%) being the most preferred choice of meat. These results showed that NZ consumers had a higher meat attachment and would not give up meat for the welfare of animals (Mean = 2.78, SD = 1.27). However, 32% of NZ consumers were uncertain about buying traditional meat due to its carbon footprint (Mean = 3.14, SD = 1.18) but were willing to buy plant-based meat alternatives compared to traditional meat products (Mean = 4.08, SD = 1.18), mainly due to its nutritional profile. These findings demonstrated that although NZ consumers are fond of meat, they were willing to engage with plant-based meat alternatives but were reluctant to try IVM. The main barriers towards IVM among NZ consumers can be due to several reasons, such as the nature of IVM, differences in eating habits, higher meat attachment and general food neophobia as previously discussed.

Conclusions
This study sets out to understand the perceptions of IVM by New Zealand consumers, particularly with respect to IVM purchase and consumption behaviour. The general results indicated that if the overall perception of IVM improved, then the purchase likelihood and consumption behaviour would improve as well. Although perception plays an important role in purchase intention, other variables such as consumers' current purchase and consumption behaviour, environment and sustainability awareness and consumers opinions on health and safety of IVM, also influence consumers' purchase intentions. However, the results also revealed that consumers' religious and cultural beliefs had minimal influence on IVM purchase likelihood. It is important to note that there was no pre-screen or checks for the consumers' religious beliefs. Hence, in this instance, the concept of halal and kosher may not be relevant for non-Muslim or non-Jewish groups.
Results in this study indicated that most New Zealand consumer have minimal awareness of IVM, as a plurality of consumers were neutral in terms of engaging with IVM. In addition, New Zealand consumers are hesitant to engage with IVM due to lack of familiarity and thus NZ consumers need to be educated about IVM and its potential benefits. Awareness of IVM can bring about a change in its perception, and in some cases even lead to consumer acceptance.
These findings add to a growing body of literature on the perception of IVM among New Zealanders. However, there may be some possible limitations in this study. The first limitation of this study is the use of convenience sampling, which may have resulted in partly biased findings. Thus, it would be beneficial to understand the perception of the general population in future studies. Furthermore, this study did not test consumers for food neophobia. Further work can investigate whether consumer reluctance towards IVM is due to food neophobia or if it is merely due to extensive meat intake. Additionally, it would be interesting for future studies to provide a blinded (without knowing its origin) meat sample for consumers to further understand consumers' perception and preferences.