Evaluation of Energy-Production Preferences Using ANP Methodology Based on a Comprehensive Residential Survey

Abstract

Public opinion significantly influences discussions on the selection of technologies for future energy systems. Our goal was to support the development of energy scenarios by evaluating decision-making criteria and energy-production alternatives at a national level. The importance of the criteria and the determination of energy-production preferences were based on a social survey conducted in European countries. However, public opinions on specific technologies are frequently based on limited knowledge and can vary depending on the context. In our study, we analyzed the comparative data sets from an international survey on energy sources using three approaches: direct ranking, a comparison using various criteria, and classifying the knowledge of technologies. The evaluated results not only provide an excellent opportunity for the development of energy scenarios but also enable policymakers to develop more effective communication strategies to inform the public about technological choices and raise awareness of the need for modifications to the energy system.

1. Introduction

Nowadays, the energy sector strives to meet significant challenges, with particular emphasis on efforts to reduce climate impacts. A primary goal of the European Union is to establish a carbon-neutral energy sector, considering the continuous increase in electricity energy demand [1,2]. This focus often overshadows aspects related to energy that only occasionally come to the forefront due to accidents or international conflicts, such as the safe operation and reliable and consistent supply of energy carriers. When developing energy strategies, it is crucial to define the objectives to be achieved, specifically how we aim to secure energy supplies for future generations. The primary goals are formulated using criteria that necessitate comprehensive analyses of various factors related to the energy sector. Numerous techniques exist in the literature for considering certain criteria together and formulating a unified objective, allowing us to reduce the monetized value, ecological or carbon footprint, or even living planet or sustainability indices [3,4,5,6]. The applicability of most techniques is limited by the rigid criteria system, while in other cases, the combination of criteria and distinguishing their importance pose challenges. One of the most commonly used frameworks for combining evaluation criteria is the Multi-Criteria Decision Analysis (MCDA) technique. The importance of criteria can be flexibly considered in MCDA models, even with criterion weights, providing opportunities to incorporate the preferences of individuals, nations, or any social groups.

In our work, we aimed to compare the importance of various criteria associated with the energy sector using one of the most widely used MCDA techniques: the Analytical Network Process (ANP) method. The application of this method has shown two advantages: first, it is not sensitive to the structure of the criteria system, and second, the ranking of technologies utilizing different energy sources is also determined. The evaluation of energy production technologies in our study was comprehensive, primarily focusing on energy sources. Perhaps the only exception is solar energy, where industrial and household-scale power plants were distinguished. Since our work aimed to support the creation of long-term energy scenarios, carbon-neutral technologies play a central role in our survey, while the extended timeframe justified considering fusion energy in the context of nuclear energy.

The literature contains many studies on energy decision making, considering various energy alternatives and criteria systems. An early review by Wang et al. [7] analyzed 26 evaluation criteria, highlighting that investment costs and CO₂ emissions are top priorities due to an increased emphasis on environmental protection. This review also discussed the common use of equal weighting in MCDA and the preference for the Analytical Hierarchy Process (AHP) for more rational, sustainable energy decisions. Afgan and Carvalho [8] focused on establishing criteria systems to evaluate the sustainability of energy systems, considering factors like environmental capacity and social and economic aspects. They confirmed that the sensitivity of the priority list to the criteria rating underscores the necessity for a thorough examination of the decision-making process before reaching a final decision. In the study by Streimikiene et al. [9], they established a broad indicator system covering different approaches to sustainability. The analysis demonstrated that future energy policy should focus on sustainable energy technologies, specifically water and solar thermal energy. Alexandru’s study [10] ranked electricity-generation technologies by their sustainability in industry, using weights determined by a survey of 62 academics. They provided a ranking of 13 energy alternatives and found that large hydroelectric projects are ranked as the most sustainable. At the same time, they emphasize that the importance weights may also vary among academics from different countries, highlighting the need for geographically specific analyses. Pyakurel and Marasimi [11] and Podyal [12] highlight the importance of formulating sustainability policies for middle- and low-income countries, showing that sustainability policies cannot be generalized. Shaaban et al. [13] explored future energy technologies in Egypt, assessing them across technical, economic, environmental, and social dimensions, with stakeholder preferences playing a key role in criteria evaluation. Höfer and Madlener [14] evaluated energy-transition scenarios, taking into account the perspectives of multiple stakeholders. They found significant differences in stakeholders’ opinions on a sustainable energy transition but were able to identify three main strands of opinion among them.

In addition to forming stakeholders’ opinions, it is also very important to understand the views of society, including the general public, as their support and increasing their knowledge are of fundamental importance. Rijnsoever and Farla [15] examined public preferences for specific energy technologies using a questionnaire distributed to households in the Netherlands. Through their study, they were able to identify the most important five criteria (risk of catastrophe, economic security, private costs and discomfort, spatial impact, and price) that respondents distinguished. Sovacool [16] conducted a study on the varying cultures of energy security using a survey available in eight languages. His research highlights the complexity of energy security as a cultural issue, which reinforces the importance of performing country-specific surveys. Demski et al. [17] investigated energy security concerns across Europe using data from nationally representative surveys. They found that concerns about energy security are higher in countries with lower economic and human well-being. However, a large, up-to-date, representative sample focusing on various criteria and covering Europe is scarcely available in this subject area.

The effectiveness of survey methodologies in assessing public opinions on emerging technologies has been acknowledged, though challenges persist in distinguishing between genuine and superficial opinions, especially when public familiarity with technology like fusion is limited [18,19]. Our research focuses on collecting informed opinions across various energy alternatives, including fusion.

This paper offers a detailed analysis of part of the survey research conducted in 2023 within the Socio-economic Studies (SES) Programme at EUROFUSION [20]. A comprehensive analysis of the outcomes of the survey is still underway, and a paper presenting a detailed discussion will be published soon by the team of sociologists of the EUROfusion SES group. This paper does not intend to focus on the social acceptability of fusion but rather to derive metrics from the public perceptions of electricity generation technologies and a few evaluation criteria resulting from the survey to be applied in future MCDA studies.

Following the Introduction chapter, in Section 2, we describe the methodology of data collection and evaluation. Subsequently, we present and evaluate the results examined using different methodologies, highlighting the differences between them. In summary, the results discussed in the paper are valuable in many cases on their own; however, primarily, determining the weights of the criteria for each country and evaluating the scores of various energy-production alternatives can form the basis of a decision-support process customized to each examined country. In the conclusion, we specifically address the drawbacks of the methodology and briefly outline our future plans.

2. Evaluation of Survey Data

Our study provides a basis for the development of long-term energy scenarios where, alongside economic considerations, reliability, safety, and environmental impact also play a role. Our primary goal is to quantify the importance of these criteria in relation to energy-generation technologies. There are many methods known for weighting decision criteria, among which pairwise comparison is the most widespread, but numerous other methods can also be found in the literature [7,21,22]. A common feature of these methods is that, in almost all cases, we need to engage broader societies or narrow professional circles to ensure reliable results. There are also innovative methods available that can be applied without surveys, where we infer the ranking of criteria from already realized decisions [23], but there are still uncertainties regarding their elaboration. Our work related to the preparation of energy scenarios was connected to a comprehensive European survey, where we had the opportunity to engage a representative social sample characteristic of each selected country. Since the framework of the survey was fixed, only a very limited set of criteria could be applied, and the development of the evaluation method also required significant compromise. Therefore, we replaced the pairwise-comparison methodology with direct evaluation for a small set of criteria and implemented the ranking of alternatives through direct evaluation based on these criteria. The applied methodology, the details of which have been previously published [24], enables the simple adaptation of the ANP method.

2.1. Survey

A self-administered questionnaire was used to assess public attitudes toward fusion energy in 21 European countries. Developed specifically for this purpose, the survey collected data through online panels. Data for this study were collected in the final quarter of 2023.

While the focus of this study is not on the comprehensive results of the questionnaire, we will discuss selected questions that are pertinent to our research objectives. Initial questions assessed prior attitudes, including perceptions of problems (energy trilemma), familiarity with, and attitudes, toward different energy options.

Participants were first asked to rate their level of concern on a 1 to 5 scale regarding the reliability, environmental impact, cost, and safety of electricity generation and supply in their country. Then, they were asked to rate seven energy technologies (residential photovoltaic systems, utility-scale solar plants, wind farms, hydropower stations, nuclear power, gas power plants, and fusion energy) in each of the four considered criteria (reliability, environmental impact, cost, and safety). Additionally, subjective familiarity with these energy technologies and initial support for them were also rated on a similar scale. The questions processed in our analysis are presented in the Supplementary Material (Table S1). The limited number of criteria used was determined using the scope constraints of the survey. However, instead of short phrases, we defined the concepts to be evaluated in concise sentences. The environmental impact criterion integrates the effects of climate change beyond direct harmful physiological impacts. Based on this, its general definition could be framed as “how harmful or harmless to the environment”. We considered it important to separate the concept of safety from the security of supply. Thus, safety was emphasized as “safe in terms of potential accidents and hazardous emissions or wastes”, while the latter was interpreted as “reliable and consistent supply”. There was no opportunity to elaborate on the economic aspects in detail, so the interpretation focused on a general approach: “how costly or cheap do you think it is to produce electricity”. In order to limit the number of options to be rated, only the technologies currently in operation in the EU and producing more than 5% of the total electricity generation were included. Coal power was excluded because of the current policy plans for a progressive phase-out. Fusion energy was instead among the possible alternatives, given the current renewed interest in the technology and the boost in research from private companies.

This study encompassed a total of 19,144 European citizens aged 18 and over, sampled from large national panels in November 2023. The number of respondents per country ranged between 909 and 914. Respondents were recruited from an online panel in each of the study countries. Panel members were incentivized to participate in surveys. To ensure sample representativeness, hard quotas were implemented for gender and soft quotas for age and education level. The sample size was calculated to achieve a margin of error of less than 3.5% with 95% confidence. While the sample closely resembles the general population in terms of gender distribution in all the study countries, it over-represents individuals in some age segments in some countries. To mitigate potential biases, a weighted analysis was conducted, resulting in minimal differences in the key findings. Table S2 summarizes the demographic characteristics of the non-weighted sample.

2.2. Knowledge-Level Correction

Since a respondent’s knowledge level about a particular technology can significantly influence the evaluation of an alternative, we placed special emphasis on considering this factor. Therefore, when designing a survey, accounting for the respondents’ knowledge of the subject is crucial to ensure the accuracy and relevance of the data collected. There are several strategies that can be used to effectively gauge and incorporate respondents’ knowledge levels, such as pre-survey screening questions, knowledge assessment questions, self-assessment of knowledge, tailored question paths, conducting pilot tests, etc.

In our study, we chose the strategy of self-assessment. When respondents provide a self-assessment of their knowledge, several methods can be applied to adjust for the knowledge level, such as weighting responses based on the knowledge level, segment analysis, filtering responses, cross-tabulating responses, adjusting statistical methods, and interpreting in context. Among these, we chose the weighting method.

In the questionnaire, respondents were asked to rate their own knowledge level for each alternative on a scale from 1 to 5. Taking this into account posed a significant challenge during the evaluation, as it is difficult to determine what minimum depth of knowledge or experience is required to utilize the answers. In our work, we did not define a threshold value for the minimally expected knowledge; rather, we created a knowledge level weighting factor with their help. This factor was determined by first normalizing the respondents’ knowledge level ratings for each alternative so that the sum of these values for each respondent equals one. By normalizing in this way, we aimed to give equal consideration to the opinions of all respondents. Subsequently, this knowledge level weighting factor was applied as an additional weight to the importance weights calculated using various methodologies. This ensures that the opinion of each respondent is represented with proportional weight in the final decision.

2.3. MCDA-ANP Analysis

The most important phase of this study was determining the priority rankings based on respondents’ answers. In general, for a coherent evaluation of criteria and alternatives, the ANP method [25] was chosen, using all answers related to criteria and alternative ranking while considering selected criteria. A coherent evaluation means that the ANP method determines the importance of criteria and alternatives by means of a single iterative algorithm. It is important to mention that the application of the ANP model primarily had a computational advantage since we did not exploit the method’s main advantage—its applicability to interconnected criteria systems—due to the small number and disjoint set of criteria.

During the evaluation, we filled the ANP method’s Super Decision Matrix with the scoring answers to all the questions presented above, except for the knowledge level value. This means that we considered the direct ranking (extent of concern) of the criteria, the extent of support for alternatives, and the evaluation of each alternative according to the criteria (perception associated with each criterion).

A key strength of the ANP method, compared to direct ranking, is that the direct evaluation of criteria influences the ranking of alternatives through the evaluation of each alternative according to the criteria. This can be interpreted in such a way that, for example, if an alternative is positively evaluated on a given criterion but performs poorly overall in the alternative ranking, this will affect the evaluation of that criterion and consequently lower its perceived importance.

The main results of an ANP model are the overall priorities of the alternatives and criteria, determining the relative importance of both variables.

In addition, the evaluation of the responses was examined using the AHP methodologies [26] alongside the direct ranking based on selected responses. However, in presenting the results, we primarily relied on the ANP method’s outcomes, which are more distinct and align better with the direct ranking results compared to the AHP model’s results. At the same time, we also address the presentation of the AHP method’s results for some special cases. During the analysis, we found it interesting to demonstrate the perception associated with each given alternative, which we could only evaluate using the AHP method. Both ANP and AHP methods are based on the comparison of criteria and alternatives, which we implemented through direct ranking in our work. This was possible because of the small number of criteria and alternatives considered throughout the questionnaire survey.

After evaluating the responses using the ANP, we applied corrections, taking into account the respondents’ stated level of knowledge for certain alternatives. The results obtained are discussed at a national level. During the presentation of results, we also emphasize the differences in criterion weights obtained from different methodologies at the national level. This is particularly interesting because, with direct ranking, we grade on the basis of criteria we consider important, whereas with predetermined criteria evaluation, only these factors are considered. Moreover, we discuss the impact of considering the level of knowledge.

3. Results and Discussion

3.1. Importance of Criteria by Country

3.1.1. Importance of Criteria by Country Using Direct Ranking

When evaluating the criteria assessed by direct ranking (on a scale from 1 to 5), we examine which criterion received the highest score from the entire population of the country (most important criterion) and discuss the most frequently given rating for each criterion (reliability, environmental impact, cost, and safety).

When comparing the total scores for each country, the criterion that received the highest score in all countries is the cost, making it the most important criterion. However, when we look at which rating is most frequently given for each criterion, we obtain the following results (Figure 1a–d). In Ukraine, all four criteria received a score of 5 in concern, indicating that these criteria are truly important to them. In Denmark, all criteria most frequently received a score of 1 or 2, meaning participants were not particularly concerned about the selected criteria in their country. In Scandinavian countries, it is observed that apart from cost, all other criteria received low levels of concern, indicating that, similar to the Danes, the public has a low level of concern about the selected criteria. The Baltic states consistently gave the same values. In Western European countries (France, Spain, Portugal), we received similar results (concern of 4), except for Portugal, where reliability received a lower score. The UK provided evaluations similar to those of Central European countries, with a generally lower score observed in the Czech Republic. In Eastern European countries, we see a mixed picture, typically rating criteria with medium levels of concern.

Of course, comprehensive conclusions can hardly be drawn from these results; it is much more important to focus on the differences between the perspectives articulated by the individual respondents.

3.1.2. Importance of Criteria by Country Using ANP Method

The direct evaluation of the criteria, detailed in the previous section, leaves open the possibility that cultural or technical differences in evaluation could cause discrepancies between countries. This method can only be used to examine the extent to which the respondent is concerned with each criterion. In contrast, the application of the ANP allows for a more complex examination, enabling the determination of the relative importance of each criterion. Considering this, it is important to examine the results on a standardized scale as well. The simplest way to achieve standardization is by normalizing the criterion values of each questionnaire, which is also one of the initial steps of the ANP method. Since the final outcome of the ANP method provides criterion weights normalized to a unified scale, we will rely on this moving forward. Therefore, the results of the ANP method quantify the subtle differences between the criterion values of individual respondents, averaged by country. We must emphasize that the significance of these values lies not primarily in their ranking but in their numerically determined value, which provides opportunities for further use, such as in the development of national energy scenarios. Multiple averaging and normalization naturally smooth out individual characteristic differences, so even a few percentage points of deviation can hold considerable significance here.

Our results based on ANP calculations (

Table 1. Country-specific importance weight for each criteria.

	Reliability	Environment	Cost	Safety
Austria	0.260	0.252	0.236	0.251
Belgium	0.261	0.252	0.231	0.256
Bulgaria	0.261	0.260	0.226	0.253
Czech Republic	0.257	0.255	0.238	0.250
Denmark	0.256	0.256	0.236	0.252
Finland	0.258	0.249	0.241	0.251
France	0.263	0.251	0.232	0.255
Germany	0.258	0.255	0.236	0.251
Greece	0.159	0.547	0.140	0.154
Italy	0.260	0.260	0.230	0.251
Latvia	0.266	0.250	0.235	0.249
Lithuania	0.266	0.248	0.237	0.249
Netherlands	0.261	0.252	0.238	0.249
Poland	0.261	0.255	0.232	0.252
Portugal	0.273	0.254	0.217	0.256
Romania	0.265	0.257	0.222	0.256
Slovenia	0.265	0.254	0.229	0.252
Spain	0.159	0.544	0.143	0.155
Sweeden	0.262	0.251	0.235	0.253
United Kingdom	0.263	0.257	0.225	0.255
Ukraine	0.271	0.254	0.225	0.249

) show that, for almost all countries, the criterion with the highest weight is reliability, except in Greece, Spain, and Denmark, where the environment criterion received the highest value. It is important to note that in several other countries, the environment and cost criteria received values very close to reliability.

Therefore, it can be concluded that when evaluating criteria using direct ranking, the cost received the highest level of concern for the population; when investigating data based on the evaluation of each alternative and criteria using ANP, the highest priority was found to be the criterion of reliability in general. At the same time, it is important to emphasize once again that, in many cases, only small differences are observed.

3.2. The Effect of Correction on Knowledge Level

One of the key elements of the questionnaire was that respondents had to rate their perceived knowledge of each energy-production alternative on a scale of 1 to 5. This was of crucial importance during data evaluation because we cannot give equal weight to the opinions of those who have little or no knowledge about a given alternative. Below, we briefly discuss the effect of correcting for knowledge level.

It can generally be stated that, in both direct ranking and ANP ranking methods, correcting for knowledge level refines the solution by reducing the number of ties (alternatives receiving the same score), thereby distinguishing the differences between each alternative for every respondent and in the national average as well. To highlight this effect, we examine the frequency of the first-ranked alternatives for each country based on the individual ANP results. It was found that in more than half of the countries examined, the sequence of the alternatives changes due to this correction (see the detailed results in Figure S1).

Due to the knowledge correction, PV production units overtake wind farms in five countries, while in Finland and Slovenia, the correction has the opposite effect. In several countries, it is also observed that the correction causes nuclear power plants to exchange places (i.e., surpass) with fusion alternatives, presumably because the latter type is typically less known to the majority of the population. In Finland, while the value of wind energy remains unchanged with the knowledge-level correction, hydro and solar energy, which were preferred without correction, fall behind, making wind the most popular alternative. Additionally, the value of nuclear energy increases with the correction. In Germany, the evaluation of gas power plants increases with the correction, causing fusion energy to fall to the last place instead of gas power plants.

In the cases of Greece and Spain, the correction for knowledge level reveals a slight decrease in the popularity of rooftop PV.

In Romania, we encounter an evaluation result where the correction for knowledge level increases the popularity of the gas alternative. In this country, the knowledge about gas power plants in the representative sample is significantly higher than the perceived knowledge about nuclear power plants, unlike in other countries. Therefore, the support for gas power plants increases with the correction of knowledge level.

In the following, all results are calculated with the knowledge-level correction applied.

3.3. Ranking of Energy-Production Alternatives

3.3.1. Direct Ranking of Energy-Production Alternatives

One of the questions in the survey aimed to assess the extent of support for the expansion of each alternative in the respondents’ country. Based on this, we can examine the aggregated national score for each alternative, allowing us to rank the support for each alternative at the national level. The results are discussed based on the most popular (first place ranked) alternatives, comparing the studied countries.

Out of the seven energy-production alternatives asked about in the questionnaire, only solar energy-based production was divided into two different alternatives in terms of size (commercial PV and PV panel). If we consider rooftop PV, the results generally show that it is the most popular alternative (ranking first in 13 countries), surpassed only by PV power plants in 5 countries (Ukraine, Greece, Germany, Slovenia, and Lithuania), by wind farms in 2 countries (UK and Denmark), and by hydro energy production in the Czech Republic. However, if we examine the second most popular alternative after rooftop PV, wind farms and hydro production generally come out on top, while traditional gas and nuclear power plants never make it to the first three places (Figure 2a).

3.3.2. Ranking of Energy-Production Alternatives Using ANP Method

The popularity ranking of the seven selected alternatives can be examined not only based on direct ranking but also using the ANP methodology, similar to the criteria ranking. This methodology determines the ranking of energy alternatives by considering the selected criteria. Based on our results, it can be established that we obtain a different order between the rankings only in the case of a few countries (Figure 2a,b). Changes occur only in the cases of Italy and Romania, where wind energy is ranked higher than hydropower in the ANP evaluation for both countries. If we consider rooftop PV, the overall picture becomes even more homogeneous, with only three countries where some form of PV is not the most popular: the Czech Republic, Denmark, and Finland (Figure 2c). Based on the figures, it is clear that the results of the direct ranking and the ANP method are in perfect sync. This was achieved by having the respondents first evaluate the alternatives in general and then, in a later stage, according to specific criteria. The high degree of agreement in the results provides significant feedback on the applicability of the methodology and confirms the consistency of the respondents in filling out the questionnaires.

As a result of the individual evaluations, we previously examined the most popular alternative (the one that received the most first-place rankings) in each country. However, the country-specific average evaluation score of each alternative obtained from the ANP method of individual responses is also interesting and yields slightly different results from the previous analysis. The calculated average evaluation score for each country and electricity-generation alternative is shown in

Table 2. Country-specific evaluation score for each electricity-generation alternative.

Country	Residential Solar Panel	Solar Plants	Wind Farms	Hydroelectricity	Gas Power Plants	Nuclear Power	Fusion Energy
Austria	0.1852	0.1853	0.1709	0.1943	0.0948	0.0798	0.0897
Belgium	0.1708	0.1603	0.1680	0.1328	0.1189	0.1298	0.1194
Bulgaria	0.1695	0.1666	0.1589	0.1442	0.1189	0.1212	0.1207
Czech Republic	0.1411	0.1640	0.1587	0.1691	0.1163	0.1359	0.1148
Denmark	0.1795	0.1727	0.1912	0.1531	0.0915	0.1109	0.1012
Finland	0.1692	0.1636	0.1712	0.1683	0.0961	0.1269	0.1047
France	0.1744	0.1720	0.1564	0.1557	0.1012	0.1258	0.1145
Germany	0.1720	0.1766	0.1639	0.1646	0.1091	0.1091	0.1048
Greece	0.5327	0.1052	0.0886	0.0882	0.0785	0.0532	0.0536
Italy	0.1955	0.1514	0.1716	0.1698	0.1100	0.1000	0.1017
Latvia	0.1909	0.1488	0.1756	0.1768	0.1074	0.1030	0.0975
Lithuania	0.1679	0.1887	0.1763	0.1594	0.1051	0.1123	0.0904
Netherlands	0.1998	0.1523	0.1864	0.1117	0.1169	0.1222	0.1106
Poland	0.1777	0.1691	0.1738	0.1360	0.1146	0.1256	0.1032
Portugal	0.1927	0.1646	0.1753	0.1626	0.1134	0.0944	0.0970
Romania	0.1704	0.1468	0.1644	0.1623	0.1421	0.1056	0.1085
Slovenia	0.1715	0.1746	0.1657	0.1615	0.1012	0.1194	0.1062
Spain	0.5389	0.0868	0.0975	0.0853	0.0667	0.0633	0.0614
Sweeden	0.1771	0.1504	0.1733	0.1885	0.0830	0.1287	0.0991
Ukraine	0.1824	0.1841	0.1760	0.1521	0.1029	0.1064	0.0961
United Kingdom	0.1785	0.1574	0.1869	0.1511	0.1083	0.1158	0.1021

.

As a result of averaging the evaluation score of alternatives from individual responses in each country, it can generally be stated that the direct ranking results differ significantly from the predefined ranking based on four criteria for only a few countries. The most popular alternative remains some form of PV in most countries. At the same time, in the case of Greece and Spain, the ANP results show that rooftop PV is exceptionally popular, whereas this was not as evident when respondents considered their own criteria during the direct ranking of alternatives (although it remains the most popular). This also means that for the residents of these countries, there are other factors that negatively impact the perception of rooftop PV, which were not evaluated in this study. Moreover, hydropower received higher importance weights in Austria and Sweden, while wind farms were presumed to be more important than PV in the United Kingdom. The bottom three alternatives consistently remained nuclear-based and gas-based power plants in all countries. In summary, it can be stated that the results of the two methods harmonize excellently in the overwhelming majority of cases. This is important because we conducted the evaluation of respondents not only with different questions but also based on different methodologies, which can lead to contradictions. However, the anomalies that appear upon closer examination suggest that the modest number of criteria used in the survey cannot be considered complete. Of course, it was not our aim, nor was it possible, to use a comprehensive set of criteria, but ultimately, the discrepancy can also be attributed to this.

In the following, we examine how deterministic the composition of a given country’s installed capacity is for different production types and the distribution of energy generation according to different alternatives. In most cases, we considered the electricity production data for each alternative aggregated for the year 2023, while the installed capacities were taken from the latest available 2024 data from the ENTSO-E database [27]. Since we do not have specific information on rooftop PV, for the sake of comparability, we treat the two different PV sizes as a combined total in the subsequent analysis.

In Figure 3, we illustrate the installed capacities for the examined alternatives in each country, the annual energy production by each alternative, and the distribution of the most popular alternatives as determined by the ANP and direct-ranking methods. These distributions are shown as percentages for comparability. As previously mentioned, one can observe significant deviations in Spain and Greece due to the results obtained for PV popularity using the ANP method. However, in the case of both countries, the composition of installed capacity is much more similar to the direct ranking results of the alternatives. It can be generally stated that in countries with a high installed wind capacity (e.g., Denmark, Finland, Lithuania, and Portugal), its perception is not as positive. In Sweden, even with a high proportion of wind energy production and low PV production, PV remains the most popular. The same can be said for countries with a high installed hydro capacity (e.g., Austria, Bulgaria, Portugal, Romania, and Slovenia), where solar power plants were found to be more popular. In countries with higher nuclear power plant capacity, it can generally be stated that the perception of this alternative is more positive than in countries with little or no nuclear share.

In France, despite a significant share of installed and produced nuclear capacity, solar energy-based production is clearly the most supported, while in the Czech Republic and Sweden, renewables share popularity instead of nuclear.

In Italy and the Netherlands, there is a high proportion of gas-based power plants, which is also reflected in production data; however, their support is very low. In Latvia, along with high gas usage, there is significant hydro capacity, while in the UK, high gas is paired with a significant wind capacity. However, in these countries, PV and wind are the most popular energy sources.

The greatest similarity between the ranking of supported alternatives and the actual energy-production alternative mix can be observed in Poland, Belgium, and Germany.

3.4. Ranking of Criteria Associated with Each Electricity-Generation Alternative

The criteria played a role in two question blocks during the survey: in the direct comparison of criteria and in the comparison of alternatives based on criteria. However, we also used the comparison of alternatives based on criteria from another perspective. Since the alternatives were evaluated on the same scale for every criterion, using AHP methodology, it allowed us to compare how a given alternative was evaluated according to each criterion; that is, to study the best ranking criteria among the alternatives. Thus, we can examine the ranking of the criteria for each alternative. Interestingly, the knowledge-level correction significantly altered the results. Without the knowledge-level correction, the cost was clearly the most favorable criterion for almost all alternatives in every country. However, with the correction, a completely different ranking of criteria emerged.

In eight countries (Belgium, France, Germany, Italy, Lithuania, Latvia, Poland, and Slovenia), cost remains the most favorable criterion, except for gas power plants, where the safety criterion receives the highest rating. Additionally, in Portugal, Romania, Ukraine, and Bulgaria, reliability is rated the most positively for gas power plants, but cost remains the criterion with the best evaluation for other alternatives.

In Austria, Germany, and the Czech Republic, safety is the most favorable criterion for all alternatives. In Finland, Sweden, and the United Kingdom, safety receives the best evaluation for all alternatives except for nuclear power plants.

In Spain and Greece, opinions are similar again, with different criteria deemed to be the best rated for different alternatives: reliability for gas power plants, environmental impact for PV plants, and cost for the other alternatives.

The summarized national results generally reflect the strengths of the energy-production alternatives well, aligning with the established expectations regarding them. Thus, the obtained results clearly indicate the thoughtful work of the survey participants.

3.5. Opinions on Fusion Energy

In the questionnaire, we placed special emphasis on assessing the public evaluation of fusion energy, and therefore, we discuss these results separately. The average knowledge level about fusion energy in each country is shown in Figure 4a. For every country, the most common knowledge level rating was 3, with national averages ranging between 2.3 and 3.1. Therefore, no extreme values were observed. It can be observed that the populations in Western European and Balkan countries perceive themselves as having the highest knowledge levels. In contrast, residents of Northern European countries rated their knowledge levels as lower.

When determining the importance weight of fusion energy, we did not observe significant differences (Figure 4b), although we obtained significantly lower values for Spain and Greece using the ANP method. These extreme cases are due to the fact that rooftop PV received such a high evaluation score that it lowered the importance weights of all other alternatives. If we disregard these two countries, Austria has the lowest importance weight for fusion energy, which might be related to the general public resistance to nuclear energy. However, this resistance is not reflected in our results for the German population.

The differences observed between the direct ranking and ANP methodologies indicate that in the cases of the Netherlands, Lithuania, and Ukraine, we obtained higher importance weights with the ANP method, which again raises the question of the completeness of the criteria system.

4. Conclusions

The aim of our work was to determine the ranking of a predefined set of criteria and energy-generation alternatives based on a recent European-level representative survey for each country. We evaluated the results using multiple methodologies, with special emphasis on considering the knowledge level of individual respondents. In summary, the results of the applied methods harmonize excellently in the overwhelming majority of cases. The obtained results clearly indicate the applicability of the MCDA-ANP method and the thoughtful and consistent work of the survey participants, which resulted in a reliable database. When ranking energy-production alternatives, renewable energy sources were clearly the most popular and supported across all studied European countries, with solar energy-based alternatives standing out significantly. However, when evaluating the selected four energy criteria (expanded trilemma), environmental considerations did not prevail, which we view as a topic for further research to explain the reasons behind this.

Regarding the regional distribution, we found that geographical and natural conditions often determine support levels (e.g., high PV support in southern countries). Additionally, the existing mix of installed electricity-generation alternatives is also influential; the public tends to be more accepting of available alternatives (e.g., a higher acceptance of nuclear power was found in countries where this alternative is dominant). Of course, there are exceptions, such as the case of only fossil fuel, where even in countries with significant gas consumption (Italy, the United Kingdom, and the Netherlands), the support for gas power plants received low importance. Based on the current data, it is not clear whether the discrepancy between usage and support is due to the fossil fuel itself or merely cannot ignore the current political issues and is related to the fact that this survey was conducted following the 2022–2023 gas crisis.

Knowledge-level correction proved to be important, as it clearly influences the rankings of energy-production alternatives. Summarizing the entire sample, it can be generally stated that familiarity with fusion energy was significantly lower compared to renewable sources like solar and wind, with the highest familiarity reported for residential solar panels.

Due to space constraints, our survey does not cover all of Europe, with only a few countries surveyed in the Central European region. Additionally, a broader set of criteria would have provided a more nuanced picture. For similar reasons, we limited the differentiation of alternatives by size and technology, only splitting PV into two parts. It would have been beneficial to apply this differentiation to other energy sources as well. In addition to evaluating the survey results from a social science perspective, the future goal is to determine country-specific long-term energy scenarios based on the database generated by the survey, as well as the criteria weights obtained through the methodology presented above. The country-specific modeling results based on a representative sample can contribute to a credible energy policy that takes local social needs into account.