Consumer acceptance of and willingness to pay for food nanotechnology: a systematic review

Consumer’s attitudes to, and acceptance of, emerging technologies and their applications, are important determinants of their successful implementation and commercialisation. Understanding the range of socio-psychological, cultural and affective factors which may influence consumer responses to applications of nanotechnology will help “fine-tune” the development of consumer products in line with their expectations and preferences. This is particularly true of applications in the food area, where consumer concerns about technologies applied to food production may be elevated. This research applied systematic review methodology to synthesise current knowledge regarding societal acceptance or rejection of nanotechnology applied to agri-food production. The objective was to aggregate knowledge derived from different research areas to gain an overall picture of consumer responses to nanotechnology applied to food production. Relevant electronic databases of peer-reviewed literature were searched from the earliest date available, for peer-reviewed papers which reported primary empirical data on consumer and expert acceptance of agri-food nanotechnology, using a formal systematic review protocol. Inclusion criteria for papers to be included in the review were: empirical peer-reviewed papers written in English; a population sample of adults aged 18 years and over used in the research; a research focus on consumer and expert acceptance of agri-food nanotechnology; and research on attitudes towards, and willingness to pay for, different applications of agri-food nanotechnology. Two researchers independently appraised the papers using NVivo 10 QSR software. Studies examining consumer and expert acceptance were thematically analysed, and key information was collated. The results were synthesised in order to identify trends in information relevant to consumer acceptance of nanotechnology applied to food production. Eight key themes were identified from the 32 papers which were extracted from the literature. These themes were applied to understand the determinants of consumer acceptance of agri-food nanotechnology. Nanotechnology is more likely to be accepted by consumers when applied to development of novel packaging with distinct benefits rather than when integrated directly into agri-food products. Trust and confidence in agri-food nanotechnology and its governance need to be fostered through transparent regulation and development of societally beneficial impacts to increase consumer acceptance. Electronic supplementary material The online version of this article (doi:10.1007/s11051-015-3270-4) contains supplementary material, which is available to authorised users.

electronic databases of peer-reviewed literature were searched from the earliest date available, for peerreviewed papers which reported primary empirical data on consumer and expert acceptance of agri-food nanotechnology, using a formal systematic review protocol. Inclusion criteria for papers to be included in the review were: empirical peer-reviewed papers written in English; a population sample of adults aged 18 years and over used in the research; a research focus on consumer and expert acceptance of agri-food nanotechnology; and research on attitudes towards, and willingness to pay for, different applications of agrifood nanotechnology. Two researchers independently appraised the papers using NVivo 10 QSR software. Studies examining consumer and expert acceptance were thematically analysed, and key information was collated. The results were synthesised in order to identify trends in information relevant to consumer acceptance of nanotechnology applied to food production. Eight key themes were identified from the 32 papers which were extracted from the literature. These themes were applied to understand the determinants of consumer acceptance of agri-food nanotechnology. Nanotechnology is more likely to be accepted by consumers when applied to development of novel packaging with distinct benefits rather than when integrated directly into agri-food products. Trust and confidence in agri-food nanotechnology and its governance need to be fostered through transparent regulation and development of societally beneficial impacts to increase consumer acceptance.
Keywords Nanotechnology Á Consumer Á Acceptance Á Expert opinion Á Systematic review Background There has been extensive debate about the potential societal responses to (different) applications of nanotechnology primarily because consumer's attitudes towards, and acceptance of, emerging technologies, and their applications are important determinants of their successful implementation and commercialisation, and without consumer acceptance the potential economic and social benefits of nanotechnology may not be realised (Burri and Bellucci 2008;Frewer et al. 2011;Gupta et al. 2011;Kim et al. 2014;Lowe et al. 1993;Macoubrie 2006;Pidgeon et al. 2011;Renn and Roco 2006;Roco 2003). Stakeholders (drawn from industry and policy communities) have identified applications in the agri-food sector as being the potentially most controversial as far as societal acceptance is concerned (Gupta et al. 2013;Matin et al. 2012). To some extent this reflects expert perceptions that the pattern of societal response to different applications of nanotechnology will be similar to those observed following the introduction of genetically modified (GM) foods (Gupta et al. 2015;Mehta 2004). To date however, there has been little evidence of consumer opposition to agri-food applications of nanotechnology, (George et al. 2014), nor has formalised opposition (for example, through activities linked to pressure groups) been as extensive as that associated with GM foods (Seifert and Plows 2014;van Broekhuizen and Reijnders 2011). It is also important to note that attitudes towards technology are unlikely to remain static in space and time, and the results of a single study are unlikely to reflect an aggregated analysis of multiple studies which use different methodologies, study populations, or applications, and which are embedded in different contexts. The aim of this study was to synthesise current knowledge regarding consumer and expert acceptance or rejection of nanotechnology applied to agri-food production, to identify emerging trends and patterns, and to assess gaps in knowledge.
Whilst there have been systematic reviews of the regulatory situation surrounding nanotechnology (Grobe 2008), to the best of the authors' knowledge, no systematic reviews of research investigating consumer attitudes, perceptions and acceptance of agrifood nanotechnology have been conducted or registered on the PROSPERO 1 (PROPSERO 2012) database, nor on the databases of the Centre for Reviews and Dissemination (Centre for Reviews and Dissemination 2012; Besley et al. 2008) The systematic reviews that have been conducted to date are in the general area of nanotechnology application (e.g. in relation to risk assessment) or have focused on specific food issues, such as vitamin D food fortification (Black et al. 2011). A systematic review of research into consumer's attitudes towards and acceptance of agri-food nanotechnology is timely and policy relevant, as simply considering attitudes to specific applications may not reflect general trends in attitudes and consumer priorities for development.
This review seeks to synthesise existing knowledge regarding consumer attitudes towards agri-food nanotechnology in order to provide policy makers, nanotechnology experts, and food manufacturers with robust and high-quality evidence concerning consumer acceptance of nanotechnology applied within the agri-food sector. The results can be applied to providing evidence which will assist key stakeholders in their decision making, facilitate fine-tuning of policies and enable an estimation of how consumers may react to future food products, in line with best practices in agri-food technology application (Cook and Fairweather 2007;Raley et al. 2015).

Methods
A protocol (see Supplementary Data 1) for the review was compiled in full before searching commenced, and there were no substantive variations from protocol during the course of the study. Reporting of the review follows the Preferred Reporting Items for Systematic Reviews (PRISMA checklist; Fig. 1) guidelines (see Supplementary Data 2: Moher et al. 2009).

Information sources
Seven electronic databases of peer-reviewed literature were searched from the earliest date available (indicated in brackets) to October 2015. These were CAB Abstracts (1910), EBM Reviews (1991), Embase (1980), Medline (1946, PsycINFO (1806), Scopus (1960 and Web of Science (1864). The search strategy combined relevant terms for 'nanotechnology', 'food' and 'consumer acceptance', and search strings were adapted as appropriate for each database. Examples of the search terms used are provided in Supplementary Data 3. Additionally, reference lists of all papers meeting the inclusion criteria were also reviewed, and citation searches of included papers were conducted using Web of Science. Endnote X6 was used to manage search results, with NVivo 10 QSR International software subsequently used for data analysis.

Eligibility criteria
Studies deemed eligible for inclusion were papers which reported primary empirical data on consumer and expert acceptance of agri-food nanotechnology. Only peer-reviewed papers, written in English, were included in this review in order to focus on highquality evidence on the acceptance of agri-food nanotechnology. The inclusion criteria are fully described in Table 1 and were established to answer the primary research question: How acceptable is nanotechnology to consumers and experts when applied to agri-food products?
Study selection, appraisal and synthesis Papers were screened by two independent researchers (ELG and BC) in a three-stage process in relation to the eligibility criteria. This was done at title, abstract and full text level. Any disagreements were resolved by face-to-face discussion. Due to reference lists and citation searches being conducted, some studies were included which contained the same population as previous studies (Brown et al. 2015;Yue et al. 2015b;Roosen et al. 2011). Where studies report the same data, they are only reported once in the result i.e. there are 32 papers but only 29 stand-alone studies.
Quality assessment of included studies was carried out independently by two researchers (ELG & BC) with the Critical Appraisal Skills Programme Qualitative Research Checklist (Critical Appraisal Skills Programme 2013) used to assess qualitative research.
To assess the quantitative papers, the survey research tool by Petticrew and Roberts (2006) was used. For the mixed methods papers, both tools were used for quality appraisal. Disagreements were resolved through discussion (ELG & BC). The studies examining consumer and expert acceptance are presented in a tabular summary for narrative synthesis (see Table 2). They are described in terms of their aims, methods and study participants along with a brief summary of their key findings. Due to the plethora of findings, inconsistency in reporting styles and complexity and mixed methods nature of the data, studies were deemed too heterogeneous for metaanalysis, a four-stage thematic analysis approach was taken (Braun and Clarke 2006).
The first stage involved reading through the papers line-by-line and highlighting relevant data (e.g. a word or a paragraph), to which a code was assigned. These codes were either sociologically constructed. This means that a code was given to the data by the researchers (ELG and BC), which was either a word, sentence or paragraph, and which best reflected the meaning within the data (e.g. safety, lack of testing, too expensive)-or an 'in vivo' code-a code which directly copies what was published in the text (Barnett-Page and Thomas 2009). The second stage of the coding process involved examining these initial codes to ensure all data had been thematically analysed (by ELG and BC). The third stage involved sorting the initial codes into broader categories. Here, the researchers (ELG and BC) reflected upon the array of codes and generated broader categories by merging some codes with others, creating new codes, or renaming or deleting existing codes. The fourth stage involved assigning several themes, which essentially grouped the initial codes into major themes that would help address the research questions. Memo notes were made on how and why these analytical codes were generated by one researcher (ELG), with two further researchers (BC and SK) verifying them. These themes are presented in Table 5, and are discussed in the next section. They are illustrated using representative quotations to illustrate each theme.

Results
Thirty two papers were included; six qualitative studies (Becker 2013;Brown et al. 2015;Brown and Kuzma 2013;Gupta et al. 2012;Gupta et al. 2015;Köhler and Som 2008), 23 quantitative studies (surveys and experiments) (Besley et al. 2008;Bieberstein et al. 2013;Capon et al. 2015;Casolani et al. 2015;Cobb and Macoubrie 2004;Conti et al. 2011;Cook and Fairweather 2007;Farshchi et al. 2011;Groves 2013;Gupta et al. 2013;Marette et al. 2009;Roosen et al. 2015;Roosen et al. 2011;Schnettler et al. 2013a;Schnettler et al. 2014;Schnettler et al. 2013b;Siegrist et al. 2007;Siegrist et al. 2009;Siegrist et al. 2008;Stampfli et al. 2010;Suhaimee et al. 2014;Yue et al. 2015a, b) and three mixed methods papers (Handford et al. 2015;Simons et al. 2009;Yawson and Kuzma 2010) (see Table 2). During sifting, 17 papers were excluded because they were unavailable from Newcastle University, the Internet or through inter-library loans, or they were unobtainable in English (Ahmadi and Ahmadi 2013;Cheng et al. 2009;Lin et al. 2011;Militaru and Ionescu 2013;Mir 2007;Rakia 1993;Rogers et al. 2013;Schiffeler 2014;Scholl 2013;Siegrist 2007;Stone 2009;Suerdem et al. 2013;Tanaka 1995;Teggatz 2013;Thoenes 1982;Thompson n.d.;Zimmer 2008), but which may have been potentially relevant. The qualitative empirical papers Participants are more likely to purchase low fat lamb or beef made using nanotechnology. The intentions to purchase these products were influenced by self-identity, attitude and subjective norms   (Handford et al. 2015),'commercializers' (individuals who make deliberate efforts to increase the presence of products on the market that employ nanotechnology or contain nanomaterials'' (Becker 2013); and one study surveyed consumers, academic, business and government stakeholders (Capon et al. 2015). Quality appraisal of the qualitative studies is shown in Table 3, and the quantitative studies in Table 4. For the qualitative studies, all six papers included a clear statement of the aims of the research and employed a qualitative methodology. The majority of studies had designs appropriate to the aims and objectives, used a suitable recruitment strategy, collected data in a way that was appropriate to the research topic and provided a clear statement of findings. However, the majority of studies did not consider the impact of the relationship between the researcher and the participants, and only two of them explicitly state how they had considered ethical issues. For the experimental studies, a lack of information reported in the papers meant that many study attributes were rated as 'unclear', most likely due to reporting restrictions in the respective journals. Finally, for one of the qualitative studies, information to demonstrate the rigour of the data analysis was not provided. All quantitative studies employed a methodological approach appropriate to the research topic and most undertook appropriate analyses, with the remaining four being unclear as to exactly how they analysed the data. However, for the majority of the studies it was not possible to determine whether a representative sample and objective measures (e.g. validated survey questions) had been used, with only studies, typically the experimental ones, using quota sampling to ensure samples were representative. Less than half of the studies justified their sample size or reported the response rate during recruitment. Finally, in terms of the quality of the papers, it may be that key methodological issues were not reported, rather than these being weak areas of study design, although this is potentially interpretable as evidence of bias. In the absence of validated quality appraisal tools, a best match was used.
The results below present the main themes that were identified from the thematic analysis (see Table 5). We indicate the relevant supplementary data boxes which are pertinent to each theme throughout the next section.
Theme 1: type and applications of agri-food nanotechnology Nanotechnology can be integrated into food products, can form part of the packaging of food, and/or can be used when processing food products. When considering these three types of application, overall, the majority of the studies (regardless of sample population) reported greater consumer acceptance of nanotechnology when it was applied to agri-food packaging and processing activities, compared to when it was integrated into agri-food products (see Supplementary Box 1).
Both consumer and expert opinion were divided on whether they found nanotechnology to be acceptable or unacceptable when used directly in foods as such. Experts appear to rate nanotechnology when      Unclear Unclear applied to food and food products to be more acceptable than do consumers, but that could be because many of these experts worked in the nanotechnology field hold some asymmetric information (i.e. greater knowledge and information about risk and benefit assessment which is not available to consumers).

Theme 2: benefits and risks of agri-food nanotechnology
Often agri-food-related nanotechnology was considered acceptable by experts when clear benefits could be identified. Experts considered benefits in relation to food freshness and safety, and wider environmental and food manufacturing advantages. In particular, if nanotechnology could prevent food spoilage and enhance the shelf-life of the food, and reduce the amount of packaging that would need to be used, it was viewed as acceptable. Additional wider applications of nanotechnology included using nanotechnologies to reduce food shortages, and to improve (reduce) calorie content of food. Ultimately, if the perceived benefits were thought to outweigh the perceived risks then nanotechnology applied to agri-food production was acceptable (see Supplementary Box 2a). The available evidence suggests that consumers view agri-food nanotechnology favourably, for example in comparison to other agri-food technology innovations recently introduced such as genetically modified (GM) foods. Moreover, if the technology results in cheaper consumer products, and when it could assist beneficial food modifications (such as improved taste and disease prevention), it was perceived as acceptable. As found in the expert studies, the consumer studies found that if the perceived benefits outweighed the perceived risks, then agri-food nanotechnology is more acceptable to consumers (see Supplementary Box 2b).
The 'commercializers' perceived agri-food nanotechnology to be societally acceptable, although this may be attributable to participant's professional roles in promoting such products (see Supplementary Box 2c). Ultimately, commercialisers viewed agrifood nanotechnology to be novel, to pose a low risk to individuals in terms of health impacts, and to be societally acceptable given that there are ''riskier'' technologies within the marketplace (although it was not clear to which 'riskier' technologies participants were referring in the published research).
However, both experts and consumers expressed concerns about the potential risks associated with using nanotechnology to produce food and food products. Experts perceived a greater risk associated with nanotechnology applied to the production of food products directly as compared to food packaging (see Supplementary Box 2d).
Experts and commercialisers noted that, even when nanotechnology was used in food packaging, there may be the potential for it to contaminate food with which it came into contact, increasing risks to consumers (see Supplementary Box 2e). The proximity of nanoparticles to the human body, and in particular ingestion of the particles, was viewed as high risk, and hence unacceptable by some experts.
Within the consumer studies, multiple concerns were raised. These included concerns about potential side effects, and beliefs that the technology could be misused; both of these concerns were underpinned by a fear of the unknown (see Supplementary Box 2f). Agri-food nanotechnology was also considered to be unacceptable because foods containing the technology are not perceived to be ''natural'' products. There was also a concern that nanotechnology is used for increasing profit, rather than for producing improved food products with discrete consumer benefits.

Theme 3: socio-demographic influences
The studies included in the review are heterogeneous in nature and so it is difficult to conclusively link opinions about agri-food nanotechnology to individual socio-demographic characteristics. However, there is some indication that certain population groups may be more accepting of agri-food nanotechnology than others (see Supplementary Box 3). In particular, white, male population groups perceive fewer risks to be associated with the application of nanotechnologies. In terms of expert opinion regarding perceived acceptance, Europeans and Australasians appeared to be less open to agri-food nanotechnology than other population groups. In addition, those who are traditional in their outlook may perceive greater risks to the use of agri-food nanotechnologies, compared to those who are open to new technologies. However, in most of these studies no explanation was provided to explain how and why these particular socio-demographic groups may influence levels of consumer acceptance of agri-food nanotechnology.
Theme 4: creating an informed and trusting consumer The available evidence suggests that consumer acceptance of agri-food nanotechnologies may increase if there is clarity regarding who takes responsibility for creating and regulating safe nanotechnology products, as well as regarding who provides information about safety to the general public (see Supplementary Box 4a). Although regulations regarding the protection of human health is an obvious requirement for the effective commercialisation of any agri-food technologies, participants indicated that (harmonised) regulations are also required to facilitate trade of food products developed using nanotechnology across countries (see Supplementary Box 4b). Whether or not information should be provided through product labels, to inform consumers that particular products have been produced using nanotechnology, was a more contentious issue. It is unclear how much information consumers should be provided with, nor who should be responsible for educating and informing consumers about agri-food nanotechnology (see Supplementary Box 4c). Underpinning consumer acceptance (or rejection) of foods made using nanotechnology was the issue of trust. There is evidence that a higher level of trust in the nanotechnology industry was linked to greater acceptance of the technology (see Supplementary Box 4d). Consumers place a greater degree of trust in nanotechnology when it was used in food packaging compared to when it is integrated into food products. Many studies indicated that consumers have limited knowledge about nanotechnology and how it can be applied to food products. For some consumers this may encourage early adoption of the technology, for others it can create concerns. Low levels of knowledge about nanotechnology may translate into a lower willingness to accept and purchase agri-food nanotechnology products because of a lack of understanding of how it is used in the food (see Supplementary Box 4e).
Commercialisers recognised that, in order to increase consumer acceptance of, and trust in, agrifood nanotechnology, rigorous testing of products may have to be undertaken by companies who use nanotechnology in their products (see Supplementary Box 4f). Being prepared for regulatory and labelling changes was deemed important, to help increase consumer confidence in agri-food nanotechnology, even if there was some scepticism about how well consumers would understand labelling of nanotechnology in agri-food products (see Supplementary Box 4g).
Theme 5: characteristics of food nanotechnology Acceptance of agri-food nanotechnology appears to be partly determined by the technology underpinning nanotechnology products, product characteristics and the cost of nanotechnology products (see Supplementary Box 5a). Those who preferred foods to be produced using ''natural'' processing methodologies, and who associated this with being healthy, perceived nanotechnology to be less acceptable, due to greater perceptions of risk. If agri-food nanotechnology brings tangible and concrete advantages to consumers (e.g. in relation to increased food security), then experts are more likely to rate the different applications as acceptable (see Supplementary Box 5b). Consumers were however, not willing to pay more for products developed using nanotechnology, independently of the benefits that will be delivered through its application.
Theme 6: link to historical agri-food technology concerns In some of the studies reviewed, consumers linked agri-food nanotechnology to GM foods. This may have lowered the acceptability of agri-food nanotechnology if GM foods are perceived negatively (see Supplementary Box 6). Where there was consumer uncertainty about the acceptability of agri-food nanotechnology, individuals utilised their existing ''reference points'' to assess the risks and benefits arising from the technology. As one of these reference points is potentially GM foods, this may have created lower consumer acceptance of agri-food nanotechnology.
Theme 7: marketing and commercialisation In order to encourage consumer purchases of agri-food nanotechnology products, the role of marketing and, in particular, branding is potentially an important topic of research. Highlighting the benefits to consumers via marketing communications was rated important, as was the development of a ''trustworthy brand''. These recommendations are not dissimilar to the role marketing plays for other types of products and services (see Supplementary Box 7a).
It was recognised that encouraging increased repeat purchases of agri-food nanotechnology would inspire confidence in other population groups and thus increase acceptance. Thus it was suggested that those consumers who view agri-food nanotechnology to be most acceptable may ''lead'' in terms of technology adoption, which may then open up the market for other agri-food nanotechnology products (see Supplementary Box 7b). It was also reported that food packaging should be commercialised ahead of foods produced using nanotechnology, as this would be more acceptable to consumers. Furthermore, informed expert opinion might usefully be utilised to facilitate the formation of consumer opinions regarding agri-food nanotechnology and its potential acceptability by consumers.
Theme 8: future applications of agri-food nanotechnology Most recommendations for future research focused on understanding the determinants of consumer acceptance of food nanotechnology in different cultures. Comparing expert and consumer opinion was considered an important research area, as there may be a mismatch between what experts would provide in terms of agri-food nanotechnology and what would be accepted by consumers (see Supplementary Box 8a). This applied to future developments as well as those currently well advanced in terms of their innovation trajectories.
When consumer characteristics were considered in the studies reviewed, there was a focus on demographic characteristics rather than wider psychographic characteristics. Thus, moving beyond the focus on socio-demographic characteristics and to consider other psychological and cultural determinants was also identified as important (see Supplementary Box 8b). For example, consumers with an internal ''health locus of control'' (who perceive that they are able to influence their own health status through their behaviours) may be more inclined to adopt consumer products with distinct health benefits (Poínhos et al. 2014).
Exploring the drivers of social negativity towards new technologies, as well as risk aversion in the context of agri-food nanotechnology, were identified as future research priorities (see Supplementary Box 8c). Furthermore, there was a call for consumer acceptance research to use real nanotechnology products, rather than hypothetical scenarios, in order to provide study participants with a real experience of such products. This could help to provide a more realistic evidence base regarding consumer acceptance of nanotechnology, although it is clearly dependent on both the product innovation trajectory and regulatory approval of such products, in particular if they were consumed by study participants, or in some other way come into physical contact with consumers.
Finally, other key issues were identified that might influence consumer acceptance of agri-food nanotechnology. These considerations also related to the themes identified above, particularly providing clear and detailed information, involving multiple stakeholders in the debate on nanotechnology and building consumer confidence and trust (see Supplementary Box 8d).

Statement of main findings
We believe that this is the first systematic review to explore empirical findings reporting on consumer and expert acceptance of nanotechnology applied to the agri-food sector. Included in this review are 32 empirical studies focused on consumer and expert opinions towards agri-food nanotechnology. The majority of these studies used a survey methodology to assess acceptance, although each survey asked very different questions of participants. In-depth empirical (i.e. qualitative research), or experimental research (for example, that which examined the impacts of information interventions on consumer attitudes) exploring consumer acceptance was limited, and it may be useful to follow this up in future research. The analysis of the research reported in the papers included in the review identified eight themes which appear relevant to understanding societal acceptance of agrifood nanotechnology. The consumer studies, and those involving expert assessment of consumer perceptions, suggested that the benefits and risks which consumers perceive to be associated with nanotechnology applied to food production and food products is likely to be an important determinant of consumer responses. In this respect, agri-food nanotechnology is likely to be accepted by consumers if the perceived benefits in some way outweigh the perceived risks and associated consumer concerns. In particular, nanotechnology was deemed more acceptable when it was used in food packaging and processing rather than as an integral part of food products themselves. It was also found that agri-food nanotechnology may be more acceptable if it results in cheaper, safer, consumer products, i.e., a tangible and desirable consumer benefit is delivered as a consequence of its application.
There is reasonable consistency in the literature regarding societal acceptance of agri-food applications of nanotechnology. Although consumers express some concerns about nanotechnology applied to food production per se, less concern is expressed about nanotechnology applied to innovative novel food packaging. However, the consumer rejection of nanotechnology applied to food production, anticipated by some stakeholders, and following consumer reaction to GM applied to food production in some parts of the world, has not been supported by the evidence identified in this review. Increased inputs by consumers into the product development process, when concrete and tangible consumer benefits are being incorporated into specific products, is required to ensure what is being developed is also what consumers want (Raley et al. In Press).
Our systematic review has also highlighted a major gap in the available literature which concerns research which utilises theoretical approaches to understanding societal acceptance of nanotechnology applied to agri-food production. Developing research which is theoretically informed is potentially advantageous insomuch as it may facilitate greater ability to predict consumer's requirements of nanotechnological innovation in the future. Utilising theoretically driven approaches will also enable more systematic comparison of research outcomes across studies (for example, between populations with different characteristics, with respect to societal acceptance of different applications, and analysis of trends on consumer acceptance with time), in particular if a common theoretical or methodological framework or approach is adopted. It is also notable that many of the studies included in the review identified further exploration of the drivers of social negativity towards new technologies, as well as social negativity and risk aversion as future research priorities. Given that one conclusion of this systematic review is that perceived benefit is a relevant and important determinant of consumer behaviour, it will also be important to understand drivers of acceptance and benefit acquisition. It would be useful if future research systematically integrated both risk and benefit perception analyses in the research design, not least because benefit information might usefully be applied to refining the product development trajectory in the future. Commercial success will depend on consumers perceiving tangible and concrete benefits to be associated with the application of nanotechnology to food products.

Strength and weaknesses of studies included in the review
The majority of the studies reviewed used quantitative survey methodologies. Often large-and sometimes nationally representative-samples were used. This facilitated comparative analysis of the acceptance of agri-food nanotechnology across different consumer segments but did not allow for exploration or in-depth analysis of why these views were held by consumers, given the method used to collect the data. Three studies utilised experimental methodologies (i.e. choice experiments) to explore consumer preferences for (hypothetical) food nanotechnology products. Consumer experience (whether positive or negative) of foods produced using nanotechnology may influence subsequent choice behaviours, and as such limit the generalisability of findings from studies using choice experiments.
In addition, the application of formal quality appraisal indicated that studies were poor at reporting sampling and analytical procedures, and often ethical approvals for research which utilised human participants. However, the studies assessed acceptance of agri-food nanotechnology across a wide range of stakeholders, including representative groups of consumers, experts and commercialisers, as well as reporting data from a cross-section of participants, from multiple countries and backgrounds. Therefore, whilst the findings of this review highlight acceptance of agri-food nanotechnologies from the perspective of multiple stakeholders, further research is required to see how the gap can be narrowed between expert/commercialiser opinions and consumer views, to ensure nanotechnologies are acceptable to consumers, whilst being commercially viable to those who produce such technologies.

Strengths and weaknesses of this review
We believe that this systematic review has captured the available empirical evidence exploring consumer and expert opinion towards agri-food nanotechnology. Similar findings are reported across the included papers, and so we are confident that we have reached data saturation (Francis et al. 2009) regarding consumer and expert acceptance of agri-food nanotechnology. In particular, this systematic review affords those interested in commercialising nanotechnology with a quick reference guide to consumer and expert opinions towards nanotechnologies when applied to agri-food products and production methods. This review synthesises the factors that both help and hinder food nanotechnology commercialisation and provides suggestions for future research, legislation of nanotechnology and consumer education. By synthesising all of the relevant literature in these areas, this systematic review allows those interested in the field to gain an oversight of these key issues much more quickly than would occur by reading individual papers. Aggregation of the literature in this systematic review allows readers an opportunity to identify key issues, areas of concern and future developments in the field that would not be obtainable by reading individual papers in a stand-alone context.
Whilst the authors are of the opinion that data saturation was reached, 17 papers were excluded because they were unobtainable in English and/or they were unavailable. Likewise, we have not reviewed the grey literature in this area, and so again, we may have missed relevant opinions that have not been published in English language peer-reviewed journals. Some of the papers refer to grey literature, such as the Eurobarometer (European Commission 2010), but they do not discuss themes that are wholly different to the results of our systematic review.
A further weakness is that we have been unable to undertake a quantitative meta-analysis given the heterogeneity of dependent variables across the included papers. However, it may be feasible to revisit this review at a future date to conduct a meta-analysis, once there are a greater number of published empirical studies in this area which report suitable data.

Implications for policy and practice
A consistent finding was that acceptance depends on the perceived benefits of nanotechnology outweighing the perceived risks, although there is less consistency in reporting what constitutes a ''desirable benefit'' in terms of consumer perceptions. Benefits may refer to generic factors like (cheaper) prices or benefits specific to different agri-food applications. Systematic analysis of what these preferred benefits are, and which consumers want them, is needed. Policy makers and other stakeholders should also be aware that much of the research indicated that, for agri-food nanotechnology to be accepted in the marketplace, consumer confidence and trust in nanotechnology, food manufacturers, regulators and nanotechnology experts, must be developed and maintained. This might be achieved, for example, through good technology governance practice, e.g. (see Bernstein et al. 2014;Marchant 2012), effective risk-benefit communication, (Binder et al. 2011;Frewer et al. 2015) and stakeholder and end-user involvement on technology development, in line with best practice in responsible Research and Innovation policies (de Bakker et al. 2014;von Schomberg 2013).
A focus on communicating the potential benefits and risks of nanotechnology, building on consumer concerns, and investigation of how food nanotechnology can be regulated in a way that inspires consumer confidence, will increase the likelihood of food nanotechnology purchases.

Conclusion
Nanotechnology is more likely to be accepted in food packaging rather than integrated into food products. Trust and confidence in agri-food nanotechnology needs to be fostered, to increase consumer acceptance. Providing information to consumers on the benefits of nanotechnology, and ensuring an informed public could help to reduce consumer concern and could inspire food nanotechnology purchases. However, research is needed to understand what consumers perceive as beneficial, as well as how they construe risks. Adopting theoretically underpinned approaches to understanding consumer perceptions and attitudes will facilitate comparative analysis across different groups of consumers, different food nanotechnology applications, and allow assessment of trends in consumer priorities and concerns with time.