Identification of the main individual-level determinants of public understanding and acceptance of FCH technologies (D1.3)

Ref. Ares(2024)3876903 - 30/05/2024 2 D1.3 Identification of the main individual-level determinants of public understanding and acceptance of FCH technologies DELIVERABLE TYPE Report MONTH AND DATE OF DELIVERABLE Month 12, 31/05/2024 WORK PACKAGE WP 1 LEADER IMI DISSEMINATION LEVEL Public AUTHORS Dr Aaron Jensen Dr Eric Jensen Dr Sarah Spooner Lali van Zuydam Daniela Martin PROGRAMME HORIZON EUROPE GRANT AGREEMENT 101111933 START June 2023 DURATION 24 months Contributors NAME ORGANISATION Aaron Jensen, Eric Jensen, Sarah Spooner, Lali van Zuydam, Daniela Martin IMI Peer Reviews NAME ORGANISATION Marianna Franchino ENVI Maria Panadero CNH2 Revision History The information and views set out in this report are those of the author(s) and do not necessarily reflect the official opinion of the European Union, neither the European Union Institutions and Bodies nor any person acting on their behalf. 3 VERSION DATA REVIEWER MODIFICATIONS Index of Contents 1. Introduction 1 2. Methodology 1 2.1 Methods: Survey questions examined......................................................................................................1 2.2 Methods: Statistical tests used and how they are interpreted.......................................................3 3. Secondary analysis of public perceptions and reactions to hydrogen 7 3.1 Overall hydrogen awareness........................................................................................................................7 3.2 Overall views of hydrogen energy and technologies..........................................................................9 3.3 Other General Trends...................................................................................................................................11 4. Discussion 13 4.1 Regional and country differences............................................................................................................13 4.2 Environmental concern................................................................................................................................14 4.3 Social media......................................................................................................................................................14 5. Conclusion and recommendations 15 6. Appendix A 16 7. Appendix B 21 8. References 23 Index of Tables Table 1. Questions from the Gallup survey representing public awareness and attitudes toward hydrogen technology 9 Table 2.Models for individual-level determinants of hydrogen technology attitudes and related variables 11 Table 3. Ordinal Logistical Regression with Odds Ratios for responses to the question 'Negative view of hydrogen's environmental impact' 12 Table 4. Binomial GLMs of public awareness of hydrogen 23 Table 5. Binomial GLMs of public perception towards Hydrogen 26 Table 6. A.1: Model statistics of all models when compared by AIC 28 Index of Figures Figure 1. Awareness and familiarity of hydrogen technology 14 Figure 2. Views of hydrogen as a sustainable energy source 15 Figure 3. Views of hydrogen as a safe energy sources 16 Figure 4. Views of hydrogen as polluting 17 4 Partners short names ENVI Parco Scientifico Tecnologico Per L’ambiente Environment Park Torino Spa IMI Institute for Methods Innovation IME Fundacion IMDEA Energia APRE Agenzia per la Promozione della Ricerca Europea CNH2 Centro Nacional Del Hidrógeno RIGP Regionalna Izba Gospodarcza Pomorza CLUSTER TWEED Cluster Tweed BH2C Balkanski Vodoroden Klaster Acronyms FCH fuel cell and hydrogen FDR False Discovery Rate GLM Generalised Linear Model OR Odds Ratio 5 Executive Summary This deliverable uncovers key factors influencing public awareness and acceptance of hydrogen and fuel cell (FCH) technologies within the European Union. This secondary analysis uses data from the Public Opinion Survey (Gallup International 2023). Here, we analyse a range of relevant demographic and behavioural variables to establish a baseline understanding of individual-level determinants of public opinion on hydrogen technologies. The analysis focuses on a selection of eight key survey questions representing public awareness and attitudes toward hydrogen technology. Various statistical tests, including Generalised Linear Models (GLMs), were employed to assess predictor variables that may help explain variation in these awareness and attitude variables, such as country, gender and social media use. Key findings: Awareness and perception: ●There is high general awareness of hydrogen technologies across the EU, with relatively modest average variation across countries. ●Gender plays a critical role, with women generally less aware of hydrogen technologies but more informed about specific applications than men. ●Social media ownership is positively correlated with hydrogen awareness. However, extensive social media use does not significantly improve knowledge of hydrogen applications. Regional and country differences: ●Significant variations in hydrogen technology awareness and attitudes were found across different countries and EU regions. ●Countries such as Cyprus, Ireland, Italy, Malta, and Portugal show higher positive perceptions of hydrogen as a sustainable energy source, while Estonia, Slovenia, and Sweden exhibit more negative views. Environmental concern: ●Greater environmental concern correlates with positive perceptions of hydrogen as a solution for reducing fossil fuel energy dependency because of its perceived lower environmental impact. ●However, greater environmental concern corresponds to lower awareness of hydrogen's transport, heating, and industrial impact reduction applications. Influence of social media: ●Having multiple social media platform profiles is associated with slightly greater hydrogen technology awareness. ●Increased social media use correlates with decreased awareness of hydrogen applications, suggesting a need for targeted educational content or engagement in the social media domain. 6 1. Introduction This report is a deliverable for the HYPOP project to understand public opinion on hydrogen technology, focusing on understanding and acceptance. It contains a secondary data analysis from a previously conducted Public Opinion Survey on hydrogen technology in EU27 countries (Gallup International, 2023). This deliverable, led by The Institute for Methods Innovation (IMI), builds from the baseline in the HYPOP D1.2 ‘State-of-the-art analysis of public perceptions and reactions to hydrogen and fuel cell technologies’ while extending the analysis with more advanced statistical modelling. This report enhances the robustness of the original research, informing the development of public engagement workshops planned for WP3. 2. Methodology This analysis draws on the same data as the Gallup International (2023) Awareness of Hydrogen Technologies Survey Report. The public opinion survey was launched to assess European citizens’ attitudes towards and level of knowledge of hydrogen technologies and determine a baseline for monitoring changes in public opinion over time. The online survey research, with supplementary telephone interviews, was conducted in Autumn 2022, with a representative sample of 1000 citizens aged 15 years and above, in each EU Member State. The samples in all the countries were stratified by gender, age, administrative region, and type of locality. The survey explored various issues, including knowledge and awareness of general energy, particularly hydrogen energy. The Gallup International (2023) report did not conduct or share a detailed analysis of hydrogen attitudes disaggregated by country and gender. Here, we have conducted a deeper analysis of the data from this previously conducted survey and present disaggregated country-level results regarding public perceptions of hydrogen technology by gender and age to provide a more complete picture of key patterns that could have implications for public engagement. 2.1 Methods: Survey questions examined The Gallup survey collected information on respondents’ profile information, including stratified data on age (year of birth and age bracket), education (amount of education received, age at leaving education), income (how easy bill paying is, whether the situation has improved in the past 5 years), gender, country, region, locality (rural, town or city), social media presence (use and ownership) and environmental concerns. Comparing these profiles to awareness and attitude responses allows assessment of the influence of each factor. However, the Gallup report did not offer a detailed analysis of predictive factors that could influence the varying degrees of hydrogen awareness, familiarity, and attitudinal patterns. The secondary analysis IMI has conducted for this report was limited by the predictor variables available in the Gallup dataset, and it is possible that other untested factors contribute to public opinions toward hydrogen. This analysis has focused on eight outcome-related questions from the Gallup survey that represented public awareness and attitudes toward H2 technology. The goal was to identify predictive factors that have an influence on responses to awareness/familiarity attitudes questions (Table 1). Seven of the awareness/attitude questions 1 used an agree-disagree response scale, while one question (A03) used a numerical scale from 0 to 10 (0 = not damaging to the environment to 10 = extremely damaging). Table 1. Questions from the Gallup survey representing public awareness and attitudes toward hydrogen technology Outcome variable name Full question Unique ID in the dataset Hydrogen energy awareness Have you seen, read or heard anything about each of the following energy sources? (Hydrogen) A02_4 Negative view of hydrogen's environmental impact According to what you know, could you tell us to what extent each of the following sources has an impact or not on the environment? (Hydrogen) A03 Hydrogen mobility awareness Have you heard before of each of the following Hydrogen applications? A04 Hydrogen building heating awareness Have you heard before of each of the following Hydrogen applications? A04 Awareness of hydrogen's environmental impact for industry Have you heard before of each of the following Hydrogen applications? A04 Awareness of hydrogen as a solution for reducing energy dependence To what extent do you agree or disagree with each of the following statements regarding Hydrogen energy and technologies B07_2 Awareness of hydrogen as a sustainable energy source To what extent do you agree or disagree with each of the following statements regarding Hydrogen energy and technologies B07_3 Awareness of pollution relating to hydrogen To what extent do you agree or disagree with each of the following statements regarding Hydrogen energy and technologies B07_5 2 2.2 Methods: Statistical tests used and how they are interpreted This section details the statistical analysis approach used to evaluate the predictive power of a range of independent variables on dependent variables including awareness, familiarity and attitudes about hydrogen technology. The independent variables examined here included demographics such as age, country and gender, as well as potentially related environmental concerns and behavioural factors such as social media use. Statistical analysis was performed using R (Version 4.3.3) statistical software. Data were checked for skewness and heteroscedasticity before analysis to ensure normal distributions. Since most responses were categorical, factor data transformations using log or square functions were impossible. Instead, the data were grouped to make categories more similar and comparable. For example, ‘male’ and ‘other’ were combined to compare with ‘female’. Likert scale responses (agree-disagree) were coded as 0 (disagree) or 1 (agree). Scores for social media use and ownership were calculated based on 1) the number of different social media channels owned (Social Media Ownership Score) and 2) the frequency of social media channel usage each week (Social Media Use Score). Internet access was initially considered as a predictor variable. Still, it was excluded because all participants had some degree of internet access at home, work, an internet cafe, or through a mobile device. The Environmental Concern score was based on a composite score created from responses to each element of the question A01: ‘To what extent do you feel that each of the following represents an important issue facing [Country] at the moment? This included three items: greenhouse gas emissions,air pollution, and water scarcity. Scores were reverse coded to ensure that a lower score indicated less concern and a higher score indicated greater environmental concern. The Institute for Methods Innovation research team conducted binomial Generalised Linear Models (GLMs) for outcomes that followed a binomial distribution (e.g., agree/disagree responses) to compare responses to predictor variables. GLMs compared respondents' attitudes and perceptions against predictor variables to calculate the probability of a response given specific predictor variables (i.e., beliefs). Each GLM assessed and identified a range of variables using statistical significance (p-values). The percentage of deviance explained (%D) measured how much variability was described by the model. The alpha False Discovery Rate (αFDR) was used to control the proportion of false positives1(Type I errors) when multiple comparisons are made. This is a way of managing the risk that when numerous statistical tests are conducted with the same set of variables, the risk of a false positive is elevated. Variables in the model were considered statistically significant if their p-values smaller were below this αFDR-adjusted threshold. The model's effect size was shown by %D value; a higher %D indicated the factor was more likely to predict a response. For instance, a 100%D model would explain all factors influencing someone's beliefs. If a 100%D model had three factors (X, Y, and Z), X might explain 60%D, Y 1A false positive is when a statistically significant effect is detected by the analysis but in reality does not exist. 3 Hydrogen as safe as the use of any other energy source While there was a general association with hydrogen being as safe as any other energy source (Figure 3), GLMS revealed differences between countries regarding their likelihood of supporting this statement. Figure 3. Views of hydrogen as a safe energy sources Respondents from Cyprus (p = 0.0230), Denmark (p = 0.0006), Greece (p = 0.0142), Ireland (p ≤ 0.0001), Italy (p = 0.0038), Poland (p = 0.0009), Portugal (p ≤ 0.0001) and Spain (p = 0.0256) were more likely to believe hydrogen is as safe as any other energy source. In other words, being from one of these countries was linked with increased positive views about hydrogen safety. In contrast, those from Czechia (p = 0.0019), Latvia (p = 0.0003), Slovakia (p = 0.0017), and Slovenia (p = 0.0021) were more likely to view hydrogen as less safe than other energy sources. This meant that being from one of these countries was associated with decreased positive views of hydrogen safety. Respondents’ country explained 0.94% of the model (Model M64: AIC=24602, %D = 2.84). This suggested that while the country was a very small but statistically significant factor in predicting views on hydrogen safety. Hydrogen is as polluting as fossil fuels There was a strong overall belief that hydrogen was less polluting than diesel or gasoline (Figure 4). GLMs indicated 17 countries as significant positive predictors of believing hydrogen fuels were as polluting as diesel (p ≤ 0.0001 - 0.0012). Being from a rural area also influenced this belief (p = 0.002), but it explained an almost negligible amount of deviance (0.03 %D). This finding suggested that whether someone lived in a rural or urban area had little impact on their perception of hydrogen pollution compared to fossil fuels. 10 Figure 4. Views of hydrogen as polluting The respondents' countries were identified as influencing the belief that hydrogen was less polluting than diesel or gasoline. However, the country factor explained less than 1% of the model, with the percentage of deviance explained (%D) being only (0.94 %D). This meant that while the respondents' country was significant, it was unlikely to be the driving factor behind this belief. Overall, the findings indicated that while there was a strong belief in hydrogen being less polluting than diesel or gasoline, the respondents' country and whether they lived in a rural or urban area had little impact on shaping this belief. The minimal influence of these factors suggested that other elements, such as awareness campaigns, media representation, or general knowledge about hydrogen technologies, played a more substantial role in forming these perceptions. 3.3 Other General Trends Gender and age as predictors Respondents' gender and age were identified as predictors in several models. Being female (p ≥ 0.0001 - 0.0001) and being aged between 15 and 24 years (p ≤ 0.0001) were factors associated with viewing hydrogen as a good solution, sustainable, and a safe energy source. However, these factors explained only a small portion of the model deviance (0.16 - 1.75%D). In simpler terms, women and younger people were more likely to see hydrogen positively, but this influence was minimal. Being female (p ≤ 0.001, t = 42.274), having a lower level of education (<15 years (p ≤ 0.001, t = 4.373), being in the youngest age group (15-19 years (p ≤ 0.001, t = 6.378)) were associated with a higher likelihood of seeing hydrogen as harmful to the environment. Conversely, individuals in the second to youngest age group (19–24 years (p ≥0.001, t = -11.663)), those in professional positions (p = 0.004, t = -2.865 to -2.916), or business owners (p = 0.004, t = -2.888) were less likely to view hydrogen as having detrimental environmental impacts. 11 Economic situation and income Economic situation and income were only included as factors in two models (M42 and M53). Individuals who found paying bills very difficult (p = 0.008), somewhat difficult (p ≤ 0.0001), or somewhat easy (p ≤ 0.0001) were more likely to view hydrogen as just as polluting as diesel/gasoline. The ease of bill paying is explained by 0.84% of the model. People who felt their financial situation had not changed over the past five years were more likely to view hydrogen as sustainable (p < 0.0001, 0.16%D) than those who had seen improvements or negative changes. Environmental concern score A high Environmental Concern score was associated with viewing hydrogen as a good solution for reducing a country's energy dependency (p ≥0.0001, 0.98%D). This meant that people more concerned about the environment tended to see hydrogen as a favourable option for energy independence, which explained less than 1% of the model's variation. Regional differences in perceptions Respondents' regions were also identified as factors predicting whether hydrogen was seen as impacting the environment. Individuals from four regions of Cyprus (p < 0.001 - 0.002, t = 3.076 - 3.877), five regions of Denmark (p = 0.001 - 0.002, t=3.095 - 3.353), five regions of Sweden (p < 0.001 - 0.004, t = 2.894 - 3.808) and one region in Greece (p = 0.001, t = 3.246) and Lithuania (p ≥0.001, t = 4.063) were more likely to believe hydrogen had a detrimental impact on the environment. An increase in respondents from these regions resulted in a decrease in positive views on hydrogen's environmental impact. Overall findings Although the best-fit models were chosen, none explained more than 11% of the deviance. This indicates that the responses to awareness and opinions about hydrogen were influenced by factors beyond the demographic variables (country, region, age, gender, education, income, and occupation) included in the Gallup survey. 12 4. Discussion 4.1 Regional and country differences While regional and country differences had limited significance, they still showed some patterns in respondents' opinions. The region of Limassol in Cyprus was significant in increased awareness of hydrogen in cars and transport and an increased negative perception of how damaging hydrogen could be for the environment. The University of Limassol is leading research into the Hydrogen Economy for Cyprus, and it may be that factors such as association with the university contributed to Limassol’s slightly increased awareness of hydrogen. However, since survey data from Cyprus was not identified in the evidence synthesis (D1.2), information on Cypriot beliefs and regional trends remained limited. Several regions within Cyprus, Denmark, Greece, Lithuania, and Sweden were all identified as more likely to believe that hydrogen is damaging to the environment. Respondents from Lithuania, Denmark, and Sweden were also more likely to view hydrogen as polluting as diesel or gasoline. Although these predictors only represented a small effect size, there could be room for educating people from these regions about the positive environmental impacts of hydrogen. Links between respondents' countries and their views of hydrogen safety were identified. Respondents from Bulgaria, Czechia, Latvia, Slovakia, and Slovenia all had a more negative perception of hydrogen's safety. These countries are all within the EU13, and targeting information regarding safety in these areas may improve perceptions. Several trends emerged when comparing the findings from factor analysis with the evidence synthesis. Safety concerns were common throughout the EU27 and notably in the EU13 (Delaney 2021; Diamants et al. 2012; Huan et al. 2023; Ingaldi & Klimecka-Tatar 2020; Zachariah-Wolff et al. 2006; Zimmer et al. 2012). When education was provided, these concerns appeared to be reduced, and support for hydrogen increased (Alanne, 2018). Support for hydrogen remained relatively weak, and a single incident could influence opinion changes (Heinz et al. 2008). The factor analysis conducted in this study showed that while country and region were significant predictors, they only explained a limited amount. This was supported by the evidence synthesis (D1.2), as all countries showed mixed or cautious support of hydrogen technologies. Other factors may be much more influential than the respondents' country, including whether specific education regarding hydrogen had been received and cultural beliefs outside of country boundaries (Achterberg, 2014). Offering public education programmes about the uses of hydrogen, particularly on safety and environmental implications, would be a critical next step. Several studies identified in the literature supported this finding (Delaney 2021; Gryz & Kacmarczyk 2021). If prioritisation were necessary, 13 targeting education around the EU13 countries and hydrogen safety would be a good next step. 4.2 Environmental concern An individual's Environmental Concern score had mixed impacts on attitudes and opinions. A high Environmental Concern score was associated with lower awareness of hydrogen for use in transport, heating, and reducing industrial environmental impacts. This reduced awareness may reflect broader trends as hydrogen is still underrepresented in the literature compared to more established renewable energy sources such as solar, wind, and nuclear power. This was supported by the evidence synthesis (D1.2), which found only 30 studies investigating public opinion and hydrogen technologies. Within the identified studies, there was uncertainty about hydrogen as a fuel. For example, two studies found that only 20% of respondents would consider buying a hydrogen-powered car compared to much higher support for other green technologies, such as e-vehicles (Apostolou, & Welcher, 2021; Moula et al., 2017). In contrast to awareness, higher Environmental Concern scores were associated with positive perceptions of hydrogen as a solution for reducing energy dependency. These scores were also linked to views that hydrogen had a lower environmental impact. This suggested that individuals might be aware of hydrogen as a sustainable energy source but uncertain about its use cases. The existing literature suggested mixed trust and awareness of hydrogen as a sustainable energy solution. Countries like the Netherlands initially supported hydrogen but later became more uncertain about its environmental impact (Achterberg, 2014). Other studies showed that respondents only support hydrogen if it is sourced renewably (Parente et al., 2024). Increasing understanding of hydrogen and its environmental implications, both positive and negative, would help shape public opinion and alleviate environmental concerns. Given that those already interested in environmental issues were less aware of hydrogen options than other energy sources, they might be a good group to target initially in awareness campaigns. 4.3 Social media Social media ownership and use had a minor predictive role in hydrogen attitudes and awareness. A higher number of social media profiles owned was slightly associated with improved support for hydrogen as a sustainable energy source. In fact, social media use was associated with a lowered awareness of hydrogen use in transport, heating, and industry. Social media tends to have a selection bias, meaning users may focus on reading information that interests them and see ads and posts that link to their interests. Increased social media use was unlikely to improve awareness unless individuals were already interested in hydrogen. Evaluating social media trends (D1.4) provides more detail about the public's response to social media in relation to hydrogen technologies. 14 5. Conclusion and recommendations While some factors (respondents’ region, country, gender, social media use, and environmental concern score) had a predictive impact on public beliefs and awareness of hydrogen technologies, these only had minimal predictive power in terms of model variation. Based on evidence synthesis from other public opinion surveys conducted within the EU (D1.2), exposure to specific education about hydrogen and its uses is likely a major factor in predicting opinions. Providing more extensive public education about hydrogen, its uses, and environmental impacts, as well as addressing safety concerns, is the most important step in increasing public awareness and acceptance of hydrogen. Recommendations Targeted education campaigns: ●Develop and implement education programs tailored to specific regions and demographics to improve hydrogen awareness and acceptance. ●Prioritise educational efforts in countries with lower awareness and more negative perceptions, particularly within the EU13 nations. Leveraging social media: ●Enhance social media's role in spreading accurate information about hydrogen technologies, focusing on engaging content that addresses safety, environmental benefits, and practical applications. ●Target social media campaigns at demographics showing lower awareness to bridge the knowledge gap. Focus on safety and environmental impact: ●Address safety concerns through comprehensive information campaigns highlighting hydrogen’s safety compared to other energy sources. ●Emphasise hydrogen’s environmental benefits in public messaging to align with the increasing environmental concerns of the populace. 15 6. Appendix A Table 4. Binomial GLMs of public awareness of hydrogen A04_1: Have you heard before of Hydrogen as a fuel for transport (cars, buses, trucks…)? A04_2: Have you heard before of Hydrogen for heating houses or buildings? A04_3:Have you heard before of Hydrogen used in some industries to reduce their impact on the environment? A02_4 Have you seen, read or heard anything about Hydrogen? M7 M1 M15 M23 M32 Region Region 30 Lemesos/Limas sol, Cyprus (+) p = 0.0002 %D = 3.26 Region 30 Lemesos/Limass ol, Cyprus (+) P = 0.0001 %D = 1.46 Region 153 Friesland, Netherlands (-) p = 0.0016 %D = 2.0 Country - Belgium (-) p = <0.0001 Bulgaria (-) p = 0.0006 Croatia (-) <0.0001 Cyprus (-) p = <0.0001 Czechia (-) p = <0.0001 Denmark (-) p = <0.0001 Estonia (-) p = <0.0001 Finland (-) p = <0.0001 France (-) p = <0.0001 Greece (-) p = <0.0001 Hungary (-) p = <0.0001 Ireland (-) p = <0.0001 16 Country (cont.) Latvia (-) p = <0.0001 Lithuania (-) p = <0.0001 Luxembourg (-) p = <0.0001 Malta (-) p = <0.0001 Poland (-) p = 0.006 Portugal (-) p = 0.0024 Romania (-) p = 0.0004 Slovenia (-) p = <0.0001 Spain (-) p = <0.0001 Sweden (-) p = <0.0001 %D = 3.03 Town/City/Rural - - Year Born - - - Age - - Gender Female (+) p = 0.0001, %D = 4.4 Female (+) p = <0.0001, %D = 4.39 Female (+) p = <0.0001, %D = 0.88 Female (+) p = <0.0001, %D = 1.88 Female (-) p = <0.0001, %D = 2.62 Age left education - Years of Education received 15-19 years (+) p = <0.0001 Refuse to answer (+) p = <0.0001, %D = 0.4 15-19 years (+) p= <0.0001 Refuse to answer (+) p = 0.0002, %D = 0.35 Still studying (+) p = <0.0001, %D = 0.28 Still studying (+) p = <0.0001, %D = 0.18 17 Occupation O7 Professional (-): p = <0.0001 O9: Business owner (-): p = 0.0002 O10: Employed professional (-): p = <0.0001 O11: General manager (-): p = <0.0001 O12: Middle management (-): p = <0.0001 O13: Employed office worker (-): p = <0.0001 %D = 0.4 O7: Professional (-): p = <0.0001 O9: Business owner (-): p = 0.0002 O10: Employed professional (-): p = <0.0001 O11: General manager (-): p = <0.0001 O12: Middle management (-): p = <0.0001 %D = 0.39 O6: Fisherman (-) p = <0.0001 O7: Professional (-) p = <0.0001 O9: Business owner (-) p = <0.0001 O11: General manager (-) p = <0.0001 %D = 0.55 O9: Business owner (-) p = 0.0009 O11: General manager (-) p = <0.0001 %D = 0.33 Social Media Use (-) p = <0.0001, %D = 0.4 (-) p = <0.0001, %D = 0.39 (-) p = <0.0001, %D = 1 (-) p = <0.001, %D = 1.11 Social Media Owned (+) p = <0.0001, %D = 0.3 Environmental concern (-) p = <0.0001, %D = 0.8 (-) p = <0.0001, %D = 0.82 (-) p = <0.0001, %D = 0.95 (-) p = 0.0001, %D = 1.05 Model statistics AIC: 32111 %D = 10.3 αFDR 0.0016 AIC: 32111 %D = 10.29 αFDR 0.0016 AIC: 30084 %D = 8.02 αFDR 0.0017 AIC = 33512 %D = 7.10 αFDR = 0.00116 AIC: 21672 %D = 5.9 αFDR = 0.036 18 Table 5. Binomial GLMs of public perception towards Hydrogen B07_2: Hydrogen is a good solution for reducing the energy dependence of [Country] B07_3: Hydrogen is a sustainable energy source B07_5: Hydrogen is as polluting as Diesel or gasoline. B07_6: The use of Hydrogen is as safe as the use of any other energy source M33 M42 M53 M64 Region - Country Cyprus (+) p = 0.0039 Estonia (-) p = <0.0001 Ireland (+) p =0.0100 Italy (+) p = 0.0070 Malta (+) p = 0.0030 Portugal (+) p = 0.0002 Slovenia (-) p = <0.0001 Sweden (-) p = 0.0015 %D = 1.14 Belgium (+) p = <0.0001 Bulgaria (+) p = 0.0012 Croatia (+) p=<0.0001 Czechia (+) p=<0.0001 Denmark (+) p=<0.0001 Estonia (+) p=<0.0001 Finland (+) p=0.0001 France (+) p=<0.0001 Ireland (+) p=<0.0001 Italy (+) P=0.0012 Latvia (+) P=<0.0001 Lithuania (+) P=<0.0001 Netherlands (+) p=<0.0001 Poland (+) p=<0.0001 Slovakia (+) p=0.00023 Slovenia (+) p=<0.0001 Sweden (+) P=<0.0001 %D = 0.94 Bulgaria (-) p = 0.0001 Cyprus (+) p = 0.0230 Czechia (-) p = 0.0019 Denmark (+) p = 0.0006 Greece (+) p = 0.0142 Ireland (+) p = <0.0001 Italy (+) p = 0.0038 Latvia (-) p = 0.0003 Poland (+) p = 0.0009 Portugal (+) p = <0.0001 Slovakia (-) p = 0.0017 Slovenia (-) p = 0.0021 Spain (+) p = 0.0256 %D = 0.94 Town/City - Rural (+) p= 0.002 %D = 0.03 Year Born - - Age 15 – 24 years (+) p =<0.0001 25 – 39 years (-) p= <0.0001 %D = 1.3 15 – 24 years (+) p = <0.0001 25 – 39 years (-) p = <0.0001 %D = 0.63 Gender Female (+) p = <0.0001 %D = 0.16 Female (+) p=0.0001 %D = 1.75 Female (+) p =<0.0001 %D = 0.64 19