The Green Consumption Behavior Process Mechanism of New Energy Vehicles Driven by Big Data—From a Metacognitive Perspective

Abstract

Green consumption behavior is the embodiment of pro-environmental behavior, which is of great value to curb carbon emissions. However, the existing research on the model construction and quantitative analysis of the psychological process of green consumption behavior needs to be further explored. Therefore, on the basis of green consumption behavior and metacognitive theory, this study constructs a conceptual model of a psychological process with a psychological control source, green consumption attitude, three aspects of metacognition, and green consumption behavior and puts forward the hypothesis of an action mechanism. This study combines text mining technology and expert knowledge to establish a user review mining dictionary and mines the variables in the quantitative conceptual model through word embedding to test empirically the mechanism hypothesis. The results show that psychological control source has a significant impact on green consumption behavior, and green consumption attitude plays a partial mediating role between them. Metacognitive knowledge plays a moderating role between the psychological control source and green consumption behavior; metacognitive experience plays a moderating role between the psychological control source and green consumption attitude. Metacognitive monitoring plays a moderating role between green consumption attitude and green consumption behavior. In view of the above research results, we put forward the following countermeasures and suggestions: For organizations, it is necessary to identify green consumption groups, attach importance to green consumption experience, perform well in green marketing, and improve the competitiveness of green products; for decision makers, it is necessary to control strictly the industry standards of the green product market and perform well not only in the quality supervision of green products but also in the post-market construction of green products.

1. Introduction

In recent years, global carbon emissions have been rising. According to the Carbon Monitor project, global carbon dioxide emissions climbed by 3.01% in 2022, hitting a record high of 36.07 billion tons, according to global real-time carbon statistics. The Carbon Monitor is an international initiative leveraging the efforts of ten research groups to provide, for the first time, regularly updated, science-based estimates of daily CO₂ emissions in different countries. The website is: www.carbonmonitor.org.cn (accessed on 1 April 2023). Of these emissions, 6.47 billion tons, or 17.93%, came from ground transportation (shown in Figure 1). Therefore, promoting the creation of green products and green consumption of new energy vehicles has become an important issue in the current carbon emission reduction work. A new energy vehicle refers to the use of unconventional vehicle fuel [1] as a source of power (or the use of conventional vehicle fuel and the use of new vehicle power plant), integrated vehicle power control and drive of advanced technology, or the formation of advanced technology, with new technology and new structure of the car. It is particularly important to measure the growth prospects of new energy vehicles, especially electric vehicles, because they can effectively reduce the negative impact of the greenhouse effect [2]. Green consumption of new energy vehicles is a new type of consumption behavior and process characterized by reducing carbon emissions [3]. Global carbon emissions can be reduced greatly by the altruistic conduct of pro-social and pro-environment customers [4,5]. It is highly valued by many countries and is of great significance for reducing global carbon emissions. In order to fully support the industry-wide transformation of new energy vehicles, developed countries and organizations led by the United States and the European Union have encouraged the transformation of automobile electrification [6] and adopted a number of policies, rules, and regulations. Although the sales of new energy vehicles are increasing year by year, there is still a big gap compared with traditional fuel vehicles [7]. Consumers continue to have a variety of concerns when making decisions of whether to buy new energy vehicles, including consumption concepts, value orientations, and product competitiveness, which have an impact on the practice of green consumption behavior. Therefore, exploring the process of green consumption behavior of new energy vehicles and analyzing the influencing factors of promoting or inhibiting green consumption behavior have become the focus of current research.

Although existing scholars have studied green consumption behavior and have achieved some results in the green consumption behavior mechanism, influencing factors, and green purchase results, there are still deficiencies.

Firstly, consumer motivation and attitude are often the focus of research on the mechanism behind green consumption behavior. Most studies [8,9,10] take green consumption intention as the outcome variable rather than constructing a complete process model from consumption motivation to consumption behavior. According to the research, there is a “gap” between consumers’ willingness to practice green consumption and their actual behavior [11]; the phenomenon is known as “ high willingness but less action” [12], which has less correlation and analysis with consumers’ actual green consumption behavior. Therefore, when designing the mechanism model of green consumption behavior, it is necessary to construct a complete decision model from green consumption motivation to green consumption behavior on the basis of the actual purchase situation of green products and to conduct quantitative analyses.

Secondly, some studies have examined the variables that affect consumers’ decision to purchase environmentally friendly products; these include subjective norms [13] and value perception [14]. However, the potential psychological and cognitive processes of consumers have not been fully explored. Metacognition is the awareness of one’s own behavior, which is manifested as self-consciousness, self-control, and self-monitoring of one’s own behavior. It is the embodiment of a person’s autonomy, which has a more thorough interpretation of the psychological mechanism behind individual actions. At present, metacognitive theory is widely used in education [15] and learning [16]. However, there are few studies on the application of green consumption behavior [17]. Therefore, the metacognitive theory can be applied to the process mechanism of green consumption behavior to further study the psychological process in green consumption behavior.

Thirdly, on the basis of the questionnaire research, the psychological process behind green consumption behavior is quantitatively analyzed [18,19,20]. However, there are problems with this research approach, such as small data samples, inflexible questionnaire content, low recovery rate, and low efficiency. Although some researchers have used big data mining technology [19,21] to investigate consumer review mining, the analysis of reviews is based mainly on machine recognition, lacking theoretical depth based on green consumption behavior and psychological recognition and having insufficient mining and recognition of the psychological process of behavior. Therefore, this study combines consumer behavior and metacognitive theory to establish a mining dictionary, conducts text mining on the network evaluation of new energy vehicle consumers, and analyzes the mechanism of action between each psychological process and the actual green consumption behavior results.

Fourthly, in the study of new energy vehicle consumption, the academic circles carry out research mainly from the perspectives of policy [22] and environmental characteristics [2], and they pay less attention to individual consumers. In terms of research methods, the questionnaire survey is the main method [23], but it has the limitation of data sample size; in the research using text mining methods [24], there is a lack of consumption models as theoretical support mechanisms.

In view of the above research deficiencies, this study aims to promote the green consumption behavior of new energy vehicles. On the basis of metacognition theory, a psychological process mechanism model of green consumption behavior is constructed, and consumer comments are analyzed. The psychological control source is taken as the explanatory variable, the green consumption attitude as the mediating variable, the green consumption behavior as the result variable, and metacognition as the moderating variable. The study presented in this manuscript systematically analyzes the psychological process mechanism of green consumption behavior, explores the psychological process factors that promote and inhibit green consumption behavior, and puts forward countermeasures and suggestions for enterprises and governments.

The main research contributions of this paper are as follows: (1) A conceptual model of the psychological process of green consumption behavior is constructed. From the perspective of metacognitive theory, this paper explores the explanatory role of psychological control source as consumers’ intrinsic motivation, the mediating role of attitude between motivation and behavior, and the moderating role of metacognitive knowledge, metacognitive experience, and metacognitive monitoring in metacognition. (2) The quantitative analysis method of the psychological process of green consumption behavior is improved. Taking the sales data on various models of new energy vehicles as the result variables, the text mining method is used to quantitatively analyze the influence of the psychological process variables of green consumption behavior. (3) On the basis of the theory of green consumption behavior and metacognition, a user review mining dictionary is established to mine deeply and analyze the consumer reviews of new energy vehicles and to improve the knowledge base of green consumer review mining. (4) The practical countermeasures to promote green consumption behavior are improved, and the improvement countermeasures are put forward.

The remaining sections are structured as follows: Section 2 proposes a hypothesis model based on green consumption behavior theory and metacognitive theory. Section 3 introduces the research design framework based on green consumption behavior theory and metacognitive theory, as well as methods and data collection. Section 4 presents the analysis of the results of the structural training, preliminary test, utility prediction, and cause diagnosis in a detailed discussion. Section 5 summarizes the conclusions and provides the implications for enterprises and the government as well as the limitations of the study.

2. Theoretical Analysis and Development of Hypotheses

2.1. Theoretical Background

2.1.1. Theory of Green Consumption Behavior

Green consumption behavior is the behavior of consumers to purchase and use products that are environmentally friendly, recyclable, and can alleviate ecological problems [25]. At present, the academic community has explored the psychological mechanism of consumers’ green consumption behavior, taking the Theory of Planned Behavior (TPB), Value–Belief–Norm Theory (VBN), Behavioral Reasoning Theory (BRT), and other theoretical models as the basic framework and expanding the research [26,27,28] to explain the psychological process mechanism of consumers’ green consumption behavior, including mainly green consumption motivation, green consumption attitude, and green consumption behavior.

Among them, green consumption motivation is the concrete manifestation of the psychological control source. Psychological control source is the individual’s generalized expectation of the relationship between behavior and event outcomes [29], which can be divided into internal control and external control. If the person believes that the event depends on their own behavior or relatively persistent personal characteristics, it is called the internal control belief type; if this person believes that outcomes depend on certain factors, such as luck, opportunity, and fate, and are complex and unpredictable due to external influences, they are called external control belief types. In this study, the internal psychological control source is regarded as the consumers’ intrinsic consumption motivation and as an explanatory variable in the green consumption behavior mechanism model.

Green consumption attitude refers to the psychological tendency of consumers toward green consumption [10], which is an internal psychological state. In the study of social psychology, attitude is often the intermediary variable between motivation and behavior. According to the theory of interpersonal behavior, individuals’ beliefs and evaluations of behavioral results affect their attitudes.

2.1.2. Metacognitive Theory

The concept of metacognition was proposed by American developmental psychologist Flavell in the 1970s [30]. The process of metacognition actually guides and regulates our cognitive process, which is the process of controlling and executing the selection of effective cognitive strategies. Its essence is self-consciousness and self-control of cognitive activities. In the application of metacognitive theory in the field of consumption [31], the behavior process of consumers is adjusted according to their psychological motivation and consumption attitude, and the cognition is continually updated and strengthened. Metacognition plays a regulatory role in this psychological process, and it controls and regulates various psychological stages of green consumption according to consumers’ willingness.

Specifically, the concept of metacognition includes three aspects: first, metacognitive knowledge, which refers to the general term of an individual’s cognitive activities [32], processes, results, and related knowledge of oneself or others; the second is metacognitive experience, which is a cognitive or emotional experience accompanied by cognitive activities [33]; and the third is metacognitive monitoring, which means that individuals actively monitor their cognitive activities in the cognitive process [34] and adjust them accordingly to achieve the intended goal.

Therefore, this study introduces metacognitive theory on the basis of green consumption behavior theory and investigates the regulatory role of metacognitive knowledge, metacognitive experience, and metacognitive monitoring in each stage of green consumption behavior.

2.2. Development of Hypotheses

2.2.1. The Impact of Psychological Control Source on Green Consumption Behavior

Internal psychological control source has a positive effect on pro-environmental behavior. Internal controllers believe that actions can change the environment, so they are more willing to make efforts for these behaviors. According to related research, there is a favorable relationship between psychological control source and pro-environmental behavior. Environmentally responsible behavior is positively influenced by internal psychological control source [35], whereas pro-environmental behavior and willingness to pay for environmentally friendly products are significantly impacted by environmental psychological control source [36]. At the same time, there are also studies [37] that use psychosocial factors to explain the purchase behavior of new energy vehicles and achieve research results. This study suggests Hypothesis 1 in light of the analysis above.

H1. 

The psychological control source has a significant impact on green consumption behavior.

2.2.2. The Mediating Role of Green Consumption Attitude

The degree to which consumers practice green purchase behavior and adopt green consumption attitudes is crucial. Studies [38] have shown that there is a link between public views and attitudes towards new energy vehicles and charging facilities. Psychological control sources have an impact on consumers’ attitudes toward environmentally friendly consumption and other pro-environmental activities [39]. Additionally, in the context of green consumption, psychological control sources have an impact on customers’ green consumption attitudes. Customers develop a stronger attitude toward green consumption practices the more they think they can affect the outcomes of their behavior through their own actions. This is because they believe that buying green and environmentally friendly products can have an impact on the environment [40]. The internal psychological control source indirectly influences consumer attitudes toward green consumption before directly influencing the occurrence of green consumption behavior in order to increase consumers’ psychological propensity for green consumption and other pro-environment behaviors [41]. Therefore, this study proposes the following assumptions and examines whether the variable of green consumption attitude plays a complete or partial mediating role.

H2. 

The psychological control source has a significant impact on green consumption attitude.

H3. 

The green consumption attitude has a significant impact on green consumption behavior.

H4. 

The green consumption attitude plays a mediating role between the psychological control source and green consumption behavior.

2.2.3. The Regulatory Role of Metacognition in Psychological Process

Metacognitive knowledge is knowledge about ‘cognition’, including consumers’ knowledge about cognitive subjects, the nature of cognitive tasks, and cognitive strategy knowledge [42]. Some studies [7] have shown that there is a tacit knowledge transfer that affects the purchase intention in the purchase behavior of new energy vehicles. For consumers with an internal psychological control source, in the process of applying the psychological control source to green consumption behavior, their cognitive knowledge is used to continually recognize and adjust during the psychological decision-making process [43], while metacognitive knowledge provides knowledge for the consumer’s cognitive process. Therefore, this study proposes Hypothesis 5 and hypothesizes the direction of metacognitive knowledge regulation.

H5. 

Metacognitive knowledge plays a regulatory role between the psychological control source and green consumption behavior.

Metacognitive experience has both cognitive and emotional components, reflecting an individual’s cognitive and emotional experience of the cognitive activity itself during cognitive activities [44]. Studies [45] have shown that while improving both the production and production capacity of new energy vehicles, it is necessary to promote the user experience and intelligence of new energy vehicles so that new energy vehicles can better adapt to the market. In the process of the influence of the psychological control source on green consumption attitudes, the effect is adjusted on the basis of the experience in the psychological process and the cognitive experience of behavior [46,47]. Therefore, this study proposes Hypothesis 6 and hypothesizes the direction of metacognitive experience regulation.

H6. 

Metacognitive experience plays a regulatory role between the psychological control source and green consumption attitudes.

Metacognitive monitoring is the regulation and processing of cognitive processes by individuals who integrate various factors, taking ongoing cognitive activities as conscious objects and continually actively and consciously monitoring, controlling, and regulating them [48]. Studies [49] have shown that in the research process of energy-saving and new energy vehicles, it is necessary to monitor the vehicle in real time, master the driving conditions and operating conditions of the vehicle, and provide technical support for the construction of energy-saving and new energy vehicle driving conditions, improvement, and optimization of control strategies. In the process of transforming consumers’ green consumption attitudes into actual green consumption behaviors, it is necessary to plan and predict and to control the degree of actual consumption behavior [50,51]. Therefore, this study proposes Hypothesis 7 and hypothesizes the direction of metacognitive experience regulation.

H7. 

Metacognitive monitoring plays a regulatory role between green consumption attitudes and green consumption behaviors.

Therefore, on the basis of the theory of green consumption behavior, this study takes the psychological control source as the explanatory variable, green consumption behavior as the result variable, and introduces green consumption attitude as the intermediary variable. The impact mechanism model of green consumption behavior constructed in this study and based on the above analysis is shown in Figure 2.

3. Methodology

3.1. Research Design

3.1.1. Overall Process

Regarding the research process method, we adopt the research method of hypothesis analysis based on network information and text mining [52,53,54], which is widely recognized and used in academic circles. The research approach is shown in Figure 3, which includes the following main steps:

Data collection and preparation are handled first. The online reviews of new energy vehicles on the forum (www.autohome.com.cn (accessed on accessed on 1 April 2023)) are obtained, including model, price, year, user satisfaction, dissatisfaction, purchase location, and other information. Then, the sales data of new energy vehicles are matched on the basis of the model information, and the data are pre-processed.

After that, the pre-processed sample data are statistically analyzed. The TextRank keyword extraction method is used to obtain the word frequency distribution of the comment data, and LDA topic recognition is used to obtain the keywords in the comments. The overall sentiment evaluation of the user comments is performed, and the Word2vec model is trained using the comment data.

Then, the variables of the theoretical model are quantified, and correlation analysis and hypothesis testing are performed. On the basis of the word frequency distribution of the text and the topic recognition keywords, the word vector segmentation is set according to the characteristics of each variable in the theoretical model, and the variables are quantified by the trained model. On the basis of green consumption behavior and metacognitive theory, a user review mining dictionary is established.

Next, the hypothesis correlation analysis is conducted and tested. The Pearson correlation coefficients are calculated for each variable in the theoretical model to determine correlation. Linear regression analysis is used to verify the mediating effect and moderating effect in the hypotheses, and then the hypotheses are verified.

Finally, we form the conclusion and discuss the motivation. The research conclusions are summarized and decision-making suggestions for enterprises and governments are provided.

3.1.2. Keyword Extraction Algorithm

We use TextRank as a keyword extraction algorithm for dictionary mining. This method is an unsupervised summary extraction method based on the PageRank algorithm. It transforms text analysis into a network graph model and evaluates the importance of nodes by looking at the weight of each node in the network graph [55]. Each word in the text should be regarded as a node. If two sentences are similar, there is an undirected weighted edge between the two words. The TextRank algorithm can divide the text into multiple units, use the word nodes to create a connection graph, use the similarity to calculate the TextRank value of the word segmentation through loop iteration, find the vector representation for each word after the word segmentation, calculate the similarity between words, store them in the matrix, and then calculate and sort them on the basis of the similarity matrix and network diagram. Lastly, the top n word segmentation is the final summary result. The core formula of the TextRank algorithm is as follows, where ${\omega }_{ji}$ is used to indicate that the edge connection between two nodes has different degrees of importance.

(1)$WS\left(V_i\right)=\left(1-d\right)+d\times {\sum }_{v_J\in In\left(V_i\right)}\frac{{\omega }_{ji}}{{\sum }_{v_J\in Out\left(V_j\right)}{\omega }_{jk}}WS\left(V_j\right)$

3.1.3. Word-Embedding Mining Model

In this article, the Word2vec model is used for word-embedding mining. This model is a bilinear model for calculating word vectors developed by Google’s MIKOLOV team; Tomas Mikolov is a scholar who produces many high-quality papers, from RNNLM to Word2Vec to the recently popular FastText. This model functions on the basis of the neural network language model and the Log _ A deep learning algorithm, which are often used to calculate text similarity. Word2Vec is one of the language models; it learns semantic knowledge from a large number of text predictions in an unsupervised manner and is widely used in natural language processing. Word2Vec is also a tool for generating word vectors [56], which are closely related to language models. Using word-embedding technology, a continuous vector space with a much smaller dimension is created by embedding a high-dimensional space with a dimension of the total number of words. Each word or phrase has a real number field vector assigned to it. The central idea of this model is to express the semantic information of words by converting each word into a word vector.

In this study, the user comments are first segmented, and then based on topic recognition and word frequency analysis combined with expert opinions, 12 words are selected as a set of subject word vectors for each variable in the theoretical model (except for the result variable). We calculate the word vector distance between the word vector of the user’s comment and the word vector of the given keyword:

(2)$S\left(x,y\right)=\mathit{cos}\left(x,y\right)=\frac{{\sum }_{k=1}^t{x}_k\times y_k}{\sqrt{{\sum }_{k=1}^t{x}_k^2}\times \sqrt{{\sum }_{k=1}^t{y}_k^2}}$

The coefficient matrix $words\_similarity$ is obtained:

(3)$words\_similarity=\left[\begin{array}{cccc}{S}_{11}& S_{12}& \dots & S_{1n}\\ S_{21}& S_{22}& \dots & S_{22}\\ \dots & \dots & \dots & \dots \\ S_{m1}& S_{m2}& \dots & S_{mn}\end{array}\right]$

and its binomial is calculated as a variable $v_i$:

(4)$v_i=‖words\_similarity‖$

3.1.4. LDA Topic Extraction and Emotional Evaluation

Latent Dirichlet Allocation (LDA) is a generative probability model for discrete data (such as text corpus) sets [57]. The LDA model is widely used in social media, image processing, text classification and clustering, and community methods. The LDA topic model determines the topic set by estimating the topic probability distribution of the document to mine the text topic for text classification and clustering. Using the LDA topic model word distribution matrix and topic distribution vector, short texts can be classified successfully. Document, topic, and word levels constitute the LDA model. With the help of the probability distribution of words, this method can extract topic models from documents to reflect the potential topics of writing. This study uses performance and coherence assessment factors to establish the ideal number of topics in the review text. The value of perplexity usually decreases as the number of possible topics increases. The stronger the generation ability of the topic model, the lower the perplexity value. The consistency index uses the most frequent terms in each topic to determine semantic similarity. The more consistent the model, the easier it is to explain. In addition, it uses unsupervised training techniques and is suitable for analyzing massive text corpus. Finally, the LDAvis package is used to visualize the output of the LDA model. pyLDAvis is an open-source Python library that helps analyze and create highly interactive visualizations of clusters created by LDA.

Baidu AI is an artificial intelligence technology developed by Baidu, which includes speech recognition, natural language processing, image recognition, and other fields. We connect emotional evaluation with Baidu AI’s emotional tendency analysis API, quantify users’ positive and negative emotional tendencies, and summarize the overall emotional polarity of comments.

3.2. Sample and Data Collection

The data in this study include basic feature information, user reviews, and corresponding sales data of new energy vehicles.

The collection of basic characteristics information and user comment data of new energy vehicles is obtained from the user comments of Autohome about new energy vehicles. There are two main reasons for selecting Autohome as a sample data selection website: for one, Autohome is a well-known large-scale automobile website in China, and a large number of car owners provide high-quality user comments on the website; secondly, Autohome is a one-stop automobile service platform, integrating content, tools, and marketing, which gives it a leading advantage in content richness.

In addition, in terms of sales data on various models of new energy vehicles, we query the sales data of related models from Marklines (www.marklines.com (accessed on 1 April 2023)). MarkLines is a professional automotive data website that provides automotive production and sales data as well as predictive automotive market delivery plans. Through the centralized collection, integration, and analysis of data information, a database is constructed and data services are provided.

After data pre-processing, 22,789 sets of valid data are obtained, including vehicle models, cities, years, sales volume, user reviews, vehicle sales, of which there are 188 models. User reviews are from 286 different cities in China from 2014 to 2021, involving nearly 100 new energy vehicle brands, such as Tesla, BYD, Weilai, and Xiaopeng.

3.3. Variable Setting Measurement

3.3.1. Measurement

The outcome variable of this study is the annual sales of new energy vehicles of each model, as a measure of green consumption behavior. The relevant data are obtained from Marklines.

3.3.2. Explanatory Variables, Mediating Variables, and Regulatory Variables

The explanatory variables of this study are the psychological control source and green consumption attitude, the mediating variable is green consumption attitude, and the moderating variables are metacognitive knowledge, metacognitive experience and metacognitive monitoring. According to the results of a subsequent word frequency analysis and LDA topic recognition combined with expert suggestions, a word vector segmentation is designed, and these variables are quantified using the Word2vec model.

3.3.3. Comment Feature Variables

For the characteristics of comments, we focus mainly on the emotional tendency of comments. Using the emotional tendency analysis of the Baidu AI platform, the positive and negative emotional tendencies of comments are quantified, and the public opinion orientation of users’ purchase behavior of new energy vehicles is analyzed.

3.3.4. LDA Topic Extraction and Sentiment Evaluation

This study considers the possible impact of other factors on the model and adds three control variables: year, price, and scenario policy factors.

Judging from the sales of new energy vehicles, in recent years, the total market value of new energy vehicles has continued to expand, and sales have increased year by year. The year has a significant impact on the sales of new energy vehicles, so data crawling is carried out using the comment years in the data.

The price of an automobile determines the product level and consumer group preference of the automobile style, to a certain extent. It is found that the discount of green consumer products can affect consumers’ willingness to consume green [58], reflecting consumers’ emphasis on price in green consumption. The price data, using the market price of the model, is expressed in CNY 10,000 as a unit.

The situational policy factors reflect the government’s management measures and direction guidance for the new energy vehicle market. Relevant research shows that green procurement policies can promote green procurement behavior [58,59]. The Word2vec model is used for quantification, and the segmentation is shown in Table 1.

3.3.5. Variable Name Description

Because the variable names used in this study are long, we use abbreviations to improve the readability of the variables; Table 2 define the abbreviations of the variable names.

4. Empirical Analysis and Results

4.1. Descriptive Results

For the comment data, we use the TextRank keyword extraction method for Chinese word segmentation and statistical analysis, generate a tree diagram, and visually mark the most frequently used keywords, as shown in Figure 4.

LDA topic recognition and pyLDAvis visualization are further used to extract 6 subject words and the first 12 keywords that best reflected each topic. On the basis of the above word frequency analysis and LDA topic recognition results, combined with expert knowledge to screen and supplement, a user comment mining dictionary is finally established. Expert knowledge refers to the information we gathered by consulting with practitioners and researchers in the field of green consumption when making text mining dictionaries, which enables us to collect vocabulary more accurately and reasonably, in line with industry practices and traditions. The results are shown in Table 3.

Using the above dictionary as the word-embedding analysis model of the word vector segmentation, text mining is performed on each variable module in this research model. The statistical description results are shown in Table 4.

4.2. Robustness Checks

After quantifying the variables in the model, we combine correlation analysis and regression analysis to test the hypotheses proposed in the study.

4.2.1. Correlation Analysis

This study uses IBM SPSS Statistics 28.0 to analyze and calculate the Pearson correlation coefficient between variables.

In the field of natural sciences, the Pearson correlation coefficient is widely used to measure the degree of correlation between two variables [60] and is defined as the quotient of the covariance and standard deviation between two variables.

(5)${\rho }_{X,Y}=\frac{cov\left(X,Y\right)}{{\delta }_X{\delta }_Y}$

The Pearson correlation coefficients between variables are shown in Table 5.

On the basis of the above research, some of the proposed hypotheses can be tested.

The Pearson correlation coefficient between psychological control source and sales is 0.029, which is significant at the level of 1%. Therefore, it is assumed that H1 is verified.

The Pearson correlation coefficient between the psychological control source and green consumption attitude is 0.015, which is significant at the level of 5%. Therefore, it is assumed that H2 is verified.

The Pearson correlation coefficient between green consumption attitude and sales is −0.124, which is significant at the 1% level. Therefore, it is supposed that H3 is verified.

4.2.2. Regression Results

This study constructs a multiple regression model to test the hypotheses. The specific model is

(6)$\begin{array}{c}QS={\beta }_0+{\beta }_{1}Pcs+{\beta }_{2}Gca+{\beta }_{3}Mk+{\beta }_{4}Mm+{\beta }_{5}Pcf+{\beta }_{6}Price\\ +{\beta }_{7}Year+{\beta }_{8}Pcs\ast Mk+{\beta }_{9}Gca\ast Mm+{\epsilon }_1\end{array}$

(7)$\mathrm{G}\mathrm{c}\mathrm{a}={\beta }_{10}+{\beta }_{11}Pcs+{\beta }_{12}Me+{\beta }_{13}Pcs\ast Me+{\epsilon }_2$

where QS is the sales volume of each vehicle model, Pcs is the psychological control source, Gca is the green consumption attitude, Mk is metacognitive knowledge, Me is metacognitive experience, Mm is metacognitive monitoring, Pcm is the policy situational factor, Price is the price, Year is the year, ${\beta }_0$ and ${\beta }_{10}$ are constants, ${\beta }_1$~${\beta }_9$ and ${\beta }_{11}$~${\beta }_{13}$ are the regression coefficients, and ${\epsilon }_1$ and ${\epsilon }_2$ are the error terms.

It is worth noting that because the control variables are designed mainly for sales, when discussing the moderating effect of metacognitive experience, the influence of control variables is not considered because the outcome variable is green consumption attitude.

Next, we test each hypothesis one by one. For H4, we use the bootstrap method to test the mediating effect. The mediating effect shows a certain ‘chain’ feature; that is, there is an ‘intermediate transmission’ of the Z variable in the process of X to Y, and the Z variable has a significant mediating effect. In this study, the role of the dependent variable is linear, and its conduction process emphasizes how it plays a role through the existing path. When the independent variable has an impact effect, it has divergent characteristics. This not only directly affects the target but also affects the impact that other indicators have on the target. In addition, the intermediary role of the characteristics is needed for the impact effect to be revealed. Bootstrap is a resampling technique in statistical learning, widely used to test the mediating effect [61]. The test method of bootstrap sampling is to determine whether the 95% confidence interval of the product term (a * b) of the regression coefficient a and the regression coefficient b contains 0. If the 95% confidence interval does not include the number 0, it indicates that there is a mediating effect; if the 95% confidence interval contains the number 0, there is no mediating effect. The test results are shown in the table. In order to facilitate the operation of the program, this study first defines the following: Z—green consumption attitude, X—psychological control source, and Y—sales volume. The process3.4 program is used to obtain the following table (shown in Table 6).

From the table, it can be seen that at a 95% confidence level, the confidence interval does not include 0. The green consumption attitude Z has a significant mediating effect on the psychological control source X and sales volume Y, and it is a partial mediating effect. Therefore, the H4 hypothesis is verified.

Regarding Hypotheses H5-H7, both are tests for regulatory effects. This study uses the hierarchical regression method and IBM SPSS Statistics 28.0 for testing. The specific implementation steps of the hierarchical regression method are as follows:

Model 1 is a benchmark model with sales volume as the outcome variable, and only the control variables are added;

Model 2 adds the psychological control source and metacognitive knowledge to Model 1, serving as a control for the testing of regulatory relationships;

Model 3 adds an interaction term between the psychological control source and metacognitive knowledge on the basis of Model 2 to test the regulatory effect of metacognitive knowledge on the relationship between the psychological control source and sales volume;

Model 4 incorporates green consumption attitudes and metacognitive monitoring on the basis of Model 1, serving as a control for the testing of regulatory relationships;

Model 5 adds an interaction term between green consumption attitude and metacognitive monitoring on the basis of Model 4 to test the regulatory effect of metacognitive monitoring on the relationship between green consumption attitude and sales;

Model 6 adds explanatory and regulatory variables to Model 1 to serve as a control for the stability test of the model;

Model 7 adds interaction terms to Model 6, which constitutes the entire model, with sales volume as the result variable;

Model 8 is a model with the result variable of green consumption attitude, incorporating the psychological control source and metacognitive experiences;

Model 9 adds an interaction term between the psychological control source and metacognitive experience on the basis of Model 8 to test the regulatory effect of metacognitive experience on the relationship between the psychological control source and green consumption attitude.

The summary of the model results is shown in Table 7, where the interaction items are abbreviated with the first capital letter of the word. The changes in $R^2$ of each model in Table 7 (the size and significance of $∆R^2$ for each variable), as well as the sign and significance of the regression coefficients of each variable are used to verify the hypotheses.

In Model 3, according to $∆R^2$ at the 5% level, the explanatory ability is significantly improved compared with Model 2. The regression coefficient of the interaction between the psychological control source and metacognitive knowledge is significantly negative at the 5% level, ${\beta }_8=-\mathrm{36,379.035},$ which indicates that metacognitive knowledge has a negative regulatory effect on the relationship between the psychological control source and sales volume. H5 is confirmed.

In Model 5, according to $∆R^2$ at the 5% level, the explanatory ability is significantly improved compared with Model 4. The regression coefficient of the interaction between green consumption attitude and metacognitive monitoring is significantly positive at the 5% level, ${\beta }_9=\mathrm{52,690.179}$, which indicates that metacognitive monitoring has a positive regulatory effect on the relationship between green consumption attitude and sales. H7 is confirmed.

In Model 9, according to $∆R^2$ compared with Model 8, the explanatory power is significantly improved., The regression coefficient of the interaction term is significantly positive at the 0.1% level, ${\beta }_9=0.410$, which indicates that metacognitive experience has a positive regulatory effect on the relationship between psychological control source and green consumption attitudes. H6 is validated.

In Model 7, according to $∆R^2$ at the 1% level, the explanatory ability is significantly improved compared with Model 6. Compared with Model 2, Model 3, Model 4, and Model 5, Model 7 includes all regulatory variables and interaction terms. In the complex situation where all regulatory effects exist, further analysis of the overall data verifies the stability of Hypotheses H5 and H7 and the model as a whole.

4.3. Robustness Checks

In order to test the robustness of the empirical results, this study selects the sample data for 2021 from the original sample data, a total of 11,460 groups, and conducts regression analysis to explore whether the conclusions drawn by the model in different periods are reliable. The results are presented in Table A1 and Table A2 of Appendix A. According to the robustness results, at the 95% confidence level, green consumption attitude has a significant partial mediating effect on the psychological control source in sales Y, metacognitive knowledge has a negative moderating effect on the relationship between the psychological control source and sales, and metacognitive monitoring has a positive moderating effect on the relationship between green consumption attitude and sales. At the 99.9% confidence level, metacognitive experience has a positive moderating effect on the relationship between the psychological control source and green consumption attitude. The above test results are consistent with the research conclusions of the original data samples, indicating that the green consumption behavior model constructed in this study has certain stability.

5. Conclusion and Implications

5.1. Discussion and Conclusions

In recent years, the green consumption of new energy vehicles has attracted great interest from practitioners and researchers. However, there is a lack of in-depth research on the model framework and psychological factors. On the basis of the theory of green consumption behavior and metacognitive theory, this study analyzes the psychological process factors that affect the green behavior of new energy vehicles to promote the diffusion of green behavior of new energy vehicles in consumer groups. On the basis of the conceptual model of the psychological process of green consumption behavior, a series of psychological process mechanism hypotheses are proposed. Through user comment topic mining and expert knowledge, a consumer comment mining dictionary is established. Through word-embedding mining, the quantification of each module variable of the conceptual model is completed, and the hypothesis tests are carried out with the sales data of each model to determine the influence of each psychological process module variable on green consumption behavior. The sales data on 188 cars and 22,789 user network comment data comprise the dataset used for the hypothesis testing.

The psychological control source has a significant impact on green purchase behavior and can use green consumption attitude as an intermediary to affect significantly green consumption behavior. In previous studies, psychological factors, such as consumers’ psychological distance and risk perception of air pollution [62], consumers’ perception of fuel consumption [63], and psychological needs [64], have been fully studied, reflecting the significant influence and important role of psychological factors in consumer behavior. The research conclusion of the psychological control source in this study is consistent with the research direction and has been expanded. Internally controlled consumers believe that they can alleviate environmental problems through their own actions, thus generating intrinsic consumption motivation, and they are more willing to purchase green products. However, the results of the impact of green consumption attitudes on behavior are different from previous studies [38] and may also be due to subjective deviations caused by survey data.

Metacognitive knowledge plays a moderating role before psychological control source and green consumption behavior. Although existing research shows that environmental knowledge has a positive effect on the consumption of new energy vehicles [65], on the basis of the difference between metacognition and cognition itself, it can be seen that the effects of the two are also different. In the process of the psychological control source positively affecting green consumption behavior, metacognitive knowledge can significantly inhibit this effect and inhibit the generation of green consumption behavior.

In the process of the psychological control source positively affecting green consumption attitude, metacognitive experience can significantly promote this effect and promote the positive tendency of green consumption attitude. According to the existing research on experience and consumption intention [66], good consumption process experience can indeed have a positive effect on consumers’ attitude and intention.

In the process of green consumption attitude negatively affecting green consumption behavior, metacognitive monitoring can significantly inhibit this effect and promote the occurrence of green consumption behavior. A large number of studies [67,68,69] on new energy vehicles have discussed the important role of monitoring in energy consumption, which also reflects that consumers evaluate and select the parameters and properties of green products in green consumption.

In general, this study successfully introduces metacognitive theory; further optimizes and improves the green consumption model, with psychological factors as explanatory variables; and cites the research method paradigm of network information and text mining to achieve data quantification and analysis.

5.2. Implications

5.2.1. Implications for Organizations

Enterprises should give priority to identifying different green consumer groups, subdivide them on the basis of consumer personality traits, gather consumers with internal control characteristics, and combine other influencing factors of green consumption behavior to conduct market positioning and optimize product development and design.

Organizations need to improve green consumption attitudes to promote green consumption behavior. By strengthening the user satisfaction survey, we can effectively understand consumers’ satisfaction with new energy vehicle products and improve the dissatisfaction factors in time.

Enterprises and organizations must continually refine products, promote technological innovation, and improve the competitiveness of green products in the market to keep up with the growing green consumption awareness of consumers in terms of space, power, appearance design, and charging.

Organizations need to pay attention to green consumption experience and perform well in green experience marketing. Through in-depth research on the green product attributes of new energy vehicles, a successful green product experience marketing strategy is created to stimulate consumers’ visual, auditory, tactile, and other senses, so that they can fully participate in the attributes of green products.

The organization needs to perform well in terms of after-sales service, quality assurance, and ease of use, and further enhance the positive effect of green consumption metacognitive monitoring on green consumption purchase behavior.

5.2.2. Implications for Policy-Makers

By promoting green travel, the government can make people aware of the carbon emissions of transportation energy, improve consumers’ awareness of green consumption, increase their motivation for green travel, and guide internal control consumers to reduce carbon emissions through the purchase behavior of new energy vehicles.

By encouraging public opinion, the government can change people’s attitudes towards environmentally friendly consumption. In order to encourage the healthy development of the green consumer market, it is necessary to encourage enterprises to improve the social status of products and market-oriented competitive advantages and effectively influence the public’s perception of green consumption.

The administrative department can enhance the supporting elements and links of policy subsidies. From the purchase subsidy link to the support for enterprise R & D, they can strengthen the performance management of financial support, encourage enterprises to improve the technical level and product quality, and encourage enterprises to enhance the social reputation and market-oriented competitive advantage of products.

The government should optimize the green product standards, quality assurance, and post-market policies of new energy vehicles to improve the metacognitive monitoring of green consumption. We must strictly control the industry standards of the green product market, perform well in the quality supervision of green products, and urge green product enterprises to perform well in post-market construction.

5.2.3. Implications for Future

In today’s world, environmental pollution has become one of the focuses of global attention, and automobile emissions are one of the main causes of air pollution. Therefore, the development of new energy vehicles has attracted the attention of governments and consumers. Policy support, energy innovation, technical support, and market demand have become the key words of green consumption development of new energy vehicles. In the future research on the consumption of new energy vehicles, the research on consumers will receive increasingly more attention. Using network information and text mining technology to paint a picture of consumers and their decision-making mechanisms will become one of the directions of green consumption behavior research.

5.3. Limitations

Firstly, each variable in this study is quantified mainly by the word-embedding Word2vec text similarity. The quantitative method is relatively simple, and more quantitative methods need to be explored, including polarity quantification of attitude variables and statistical analysis of the number of comments on each model. Secondly, the review data used in this study come mainly from a car forum platform in China. Future work can collect car review data from different countries for diversity analysis, analyze spatial diversity, and obtain more popular research conclusions, using additional sources such as AutoTrader and Jalopnik. Finally, spatial heterogeneity can be further analyzed and discussed. The development of green consumption of new energy vehicles in China is different from than in other countries and regions in the world. We can further discuss the differences in psychological process factors that cause green consumption behavior.

Author Contributions

Conceptualization, Q.L. and J.C.; methodology, J.C.; validation, J.C.; formal analysis, J.C. and Q.L; investigation, J.C.; writing—original draft preparation, J.C.; writing—review and editing, Q.L.; supervision, Q.L. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. Global carbon dioxide emissions.

Figure 2. Theoretical model of research hypotheses.

Figure 3. Research framework.

Figure 4. Statistically generated word frequency tree.

Table 1

Policy situational factor segmentation.

Theory Model Module	Implication	Keywords
Situational policy factors	The consumption environment in which individuals are engaged in consumption activities includes material and social factors.	Policy, double integral, subsidies, pilot, production access, loans, charging pile, limit line, limited purchase, right of way, acquisition tax, car, and ship tax

Table 2

Variable name abbreviation description.

Variable	Abbreviation
Psychological control source	PCS
Green consumption attitude	GCA
Metacognitive knowledge	MK
Metacognitive experience	ME
Metacognitive monitoring	MM
Policy situational factors	PSF

Table 3

User comment mining dictionary based on green consumer behavior and metacognition.

Theory Model Module	Implication	Keywords
Psychological control source	Intrinsic consumer motivation, including economic motivation and convenience motivation.	New energy, domestic, technology, green,future, fashion, manufacturer, electriclicense, promotion, concessions, save money
Green consumption attitude	Individual subjective evaluation and the resulting behavioral tendencies.	Support, hope, approval, like,expect, wish, love, believe,favor, affirmation, satisfaction, optimism
Metacognitive knowledge	Metacognitive knowledge is the knowledge about ‘cognition’, which explains what cognition is, including consumers’ knowledge about cognitive subjects, the nature of cognitive tasks, and cognitive strategies.	Endurance, energy consumption, appearance, interior, space, power, consumption, wind evaluation, environmental protection, energy saving, configuration, seat, mileage
Metacognitive experience	Cognitive experience and emotional experience generated by individuals in cognitive activities.	Comfortable, relaxed, satisfied, easyconvenient, happy, good-looking, excitedexcited, surprised, forceful, stable
Metacognitive monitoring	To plan, control, and adjust the ongoing cognitive activities and to evaluate and manage the results and risks of cognitive activities.	Cost-effective, test drive, after-sales, insurance, rental, lottery, maintenance, development, investment, cost, security, advantage

Table 4

Results for Descriptive Statistics.

Variable	Median	Average	Standard Deviation	Min	Max
PCS	0.256	0.266	0.164	0.000	0.852
GCA	0.349	0.342	0.146	0.000	0.766
MK	0.366	0.360	0.162	0.000	0.814
ME	0.384	0.373	0.138	0.000	0.770
MM	0.207	0.236	0.167	0.000	0.820
PSF	0.151	0.164	0.114	0.000	0.807
Price	16.580	21.520	17.810	1.900	1469.000
Year	2021.000	2020.130	1.150	2014.000	2021.000
Positivity Tendency	0.998	0.093	0.228	0.000	1.000
Negative Tendency	0.002	0.907	0.228	0.000	1.000
Quantity Sold	18,758.000	31,238.190	59,329.330	0.000	426,482.000

Table 5

Pearson correlation coefficients.

Variable	PCS	GCA	MK	ME	MM	PSF	Price	Year	PositivityTendency	Negative Tendency	Quantity Sold
PCS	1
GCA	0.015 *	1
MK	−0.019 **	0.658 **	1
ME	−0.061 **	0.277 **	0.273 **	1
MM	0.017 *	−0.081 **	0.294 **	0.143 **	1
PSF	−0.017 *	0.195 **	0.095 **	−0.211 **	−0.484 **	1
Price	−0.006	0.173 **	−0.111 **	−0.108 **	−0.178 **	0.246 **	1
Year	−0.091 **	−0.195 **	0.006	−0.264 **	0.092 **	0.068 **	−0.041 **	1
Positivity Tendency	0.186 **	−0.002	0.174 **	0.022 **	0.179 **	−0.077 **	−0.037 **	0.160 **	1
Negative Tendency	−0.029 **	0.124 **	0.227 **	0.140 **	0.320 **	−0.170 **	−0.024 **	0.053 **	0.205 **	1
Quantity Sold	0.029 **	−0.124 **	−0.227 **	−0.140 **	−0.320 **	0.170 **	0.024 **	−0.053 **	−0.205 **	−1.000 **	1

* at the 5% level (double-tailed), the correlation is significant. ** at the 1% level (double-tailed), the correlation is significant.

Table 6

Bootstrap test of mediating effect of Z.

Direct effect of X on Y	Effect	SE	t	p	LLCI	ULCI
	17,608.762	3172.946	5.550	0.0000	11,389.571	23,827.953
Direct effect of X on Y	Effect	Boot SE	Boot LLCI	Boot ULCI
	−12,264.883	2197.798	−16,654.986	−8088.271

Table 7

Results for hierarchical regression.

Variable	Quantity Sold							GCA
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7	Model 8	Model 9
PCS		7479.623 **	21,339.860 **			30,125.657 ***	499,45.290 ***	0.609 ***	0.469 ***
GCA				−24,465.333 ***	−32,864.123 ***	−34,953.644 ***	−43,420.052 ***
MK		−41,710.040 ***	−32,830.824 ***			−39,953.137 ***	−27,419.860 ***
ME								0.370 ***	0.277 ***
MM				6513.418 **	5873.312 *	−4207.731 †	−4414.833 †
$\mathrm{P}\mathrm{C}\mathrm{S}\ast \mathrm{M}\mathrm{K}$			−36,379.035 *				−50,433.865 **
$\mathrm{P}\mathrm{C}\mathrm{S}\ast \mathrm{M}\mathrm{E}$									0.410 ***
$\mathrm{P}\mathrm{C}\ast \mathrm{M}\mathrm{M}$					52,690.179 *		45,583.025 *
Price	−414.212 ***	−506.380***	−505.539 ***	−425.123 ***	−422.585 ***	−465.436 ***	−460.986 ***
Year	10,656.793 ***	10,984.571 ***	10,951.092 ***	11,277.986 ***	11,302.453 ***	11,580.414 ***	11,581.010 ***
PSF	−1841.376	−10,621.610 **	−11,769.259 ***	−7532.284 *	−24,781.234 **	−19,018.882 ***	−35,929.867 ***
Constant	−21,487,636.965 ***	−22,133,354.199 ***	−22,068,682.603 ***	−22,734,536.817 ***	−22,781,023.213 ***	−23,330,240.542 ***	−23,332,877.365 ***	0.042 ***	0.074 ***
Sample Size	22,789	22,789	22,789	22,789	22,789	22,789	22,789	22,789	22,789
$F$	398.197	295.442 ***	247.072 ***	256.498 ***	214.648 ***	224.477 ***	176.217 ***	14, 232.960	9632.080 ***
$R^2$	0.050	0.061	0.061	0.053	0.054	0.065	0.065	0.555	0.559***
$∆R^2$	0.050	0.011 ***	0.000 *	0.003 ***	0.000 *	0.015 ***	0.001 **	0.555	0.004 ***

*** Significant at the 0.1% level, ** significant at the 1% level, * significant at the 5% level, ^† significant at the 10% level.

Appendix A

Table A1

Bootstrap test of mediating effect of Z.

Direct effect of X on Y	Effect	SE	t	p	LLCI	ULCI
	60782.5864	6242.8795	9.7363	0.000	48,545.474	73,019.700
Direct effect of X on Y	Effect	Boot SE	Boot LLCI	Boot ULCI
	−48,404.38	4823.117	−57,85.018	−38,914.664

Table A2

Results for hierarchical regression.

Variable	Quantity Sold							GCA
	Model 1	Model 2	Model 3	Model 4	Model 5	Model6	Model 7	Model 8	Model 9
PCS		5704.691	30,438.746 *			36,782.899 ***	70,450.650 ***	0.645 ***	0.548 ***
GCA				−44,667.205 ***	−63,507.457 ***	−46,210.172 ***	−62,960.268 ***
MK		−85,761.305 ***	−70,047.426 ***			−83,361.518 ***	−62,346.090 ***
ME								0.363 ***	0.301 ***
MM				16,256.798 ***	14,922.265 ***	−6958.152	−7627.910
$\mathrm{P}\mathrm{C}\mathrm{S}\ast \mathrm{M}\mathrm{K}$			−64,710.853 *				−84,322.323 **
$\mathrm{P}\mathrm{C}\mathrm{S}\ast \mathrm{M}\mathrm{E}$									0.266 ***
$\mathrm{P}\mathrm{C}\ast \mathrm{M}\mathrm{M}$					117,685.927 *		89,233.751 †
Price	−1115.932 ***	−1465.968 ***	−1462.923 ***	−1153.786 ***	−1144.035 ***	−1378.714 ***	−1362.611 ***
Year	−14,557.459 *	−31,971.401 ***	−34,233.574 ***	−29,197.470 ***	−69,485.093 ***	−43,039.390 ***	−77,213.657 ***
PSF	72,938.033 ***	113,705.952 ***	108,401.551 ***	88,508.150 ***	95,329.415 ***	122,525.796 ***	121,016.511 ***	0.047 ***	0.071 ***
Constant	11,460	11,460	11,460	11,460	11,460	11,460	11,460	11,460	11,460
Sample Size	248.571	213.763 ***	171.943 ***	144.263 ***	116.720 ***	150.393 ***	114.365 ***	7370.308	4941.689 ***
$F$	0.042	0.069	0.070	0.048	0.048	0.073	0.074	0.563	0.564
$R^2$	0.042	0.028 ***	0.000 *	0.006 ***	0.001 *	0.031 ***	0.001 *	0.563	0.001 ***

*** Significant at the 0.1% level, ** significant at the 1% level, * significant at the 5% level, ^† significant at the 10% level.

References

1. Stockkamp, C.; Schaefer, J.; Millemann, J.A.A.; Heidenreich, S. Identifying Factors Associated with Consumers’ Adoption of e-Mobility-A Systematic Literature Review. Sustainability; 2021; 13, 10975. [DOI: https://dx.doi.org/10.3390/su131910975]

2. Wang, S. Exploring the Sustainability of China’s New Energy Vehicle Development: Fresh Evidence from Population Symbiosis. Sustainability; 2022; 14, 10796. [DOI: https://dx.doi.org/10.3390/su141710796]

3. Young, W.; Hwang, K.; McDonald, S.; Oates, C.J. Sustainable Consumption: Green Consumer Behaviour when Purchasing Products. Sustain. Dev.; 2010; 18, pp. 20-31. [DOI: https://dx.doi.org/10.1002/sd.394]

4. Gaker, D.; Vautin, D.; Vij, A.; Walker, J.L. The power and value of green in promoting sustainable transport behavior. Environ. Res. Lett.; 2011; 6, 034010. [DOI: https://dx.doi.org/10.1088/1748-9326/6/3/034010]

5. Griskevicius, V.; Tybur, J.M.; Van den Bergh, B. Going Green to Be Seen: Status, Reputation, and Conspicuous Conservation. J. Personal. Soc. Psychol.; 2010; 98, pp. 392-404. [DOI: https://dx.doi.org/10.1037/a0017346]

6. Ortar, N.; Ryghaug, M. Should All Cars Be Electric by 2025? The Electric Car Debate in Europe. Sustainability; 2019; 11, 1868. [DOI: https://dx.doi.org/10.3390/su11071868]

7. Xu, N.; Xu, Y. Research on Tacit Knowledge Dissemination of Automobile Consumers’ Low-Carbon Purchase Intention. Sustainability; 2022; 14, 10097. [DOI: https://dx.doi.org/10.3390/su141610097]

8. Yadav, R.; Pathak, G.S. Young consumers’ intention towards buying green products in a developing nation: Extending the theory of planned behavior. J. Clean. Prod.; 2016; 135, pp. 732-739. [DOI: https://dx.doi.org/10.1016/j.jclepro.2016.06.120]

9. Wang, S.; Fan, J.; Zhao, D.; Yang, S.; Fu, Y. Predicting consumers’ intention to adopt hybrid electric vehicles: Using an extended version of the theory of planned behavior model. Transportation; 2016; 43, pp. 123-143. [DOI: https://dx.doi.org/10.1007/s11116-014-9567-9]

10. Hartmann, P.; Apaolaza-Ibanez, V. Consumer attitude and purchase intention toward green energy brands: The roles of psychological benefits and environmental concern. J. Bus. Res.; 2012; 65, pp. 1254-1263. [DOI: https://dx.doi.org/10.1016/j.jbusres.2011.11.001]

11. De Pelsmacker, P.; Driesen, L.; Rayp, G. Do Consumers Care about Ethics? Willingness to Pay for Fair-Trade Coffee. J. Consum. Aff.; 2005; 39, pp. 363-385. [DOI: https://dx.doi.org/10.1111/j.1745-6606.2005.00019.x]

12. Hassan, L.M.; Shiu, E.; Shaw, D. Who Says There is an Intention–Behaviour Gap? Assessing the Empirical Evidence of an Intention–Behaviour Gap in Ethical Consumption. J. Bus. Ethics; 2016; 136, pp. 219-236. [DOI: https://dx.doi.org/10.1007/s10551-014-2440-0]

13. Paul, J.; Modi, A.; Patel, J. Predicting green product consumption using theory of planned behavior and reasoned action. J. Retail. Consum. Serv.; 2016; 29, pp. 123-134. [DOI: https://dx.doi.org/10.1016/j.jretconser.2015.11.006]

14. Biswas, A.; Roy, M. Leveraging factors for sustained green consumption behavior based on consumption value perceptions: Testing the structural model. J. Clean. Prod.; 2015; 95, pp. 332-340. [DOI: https://dx.doi.org/10.1016/j.jclepro.2015.02.042]

15. Broadbent, J.; Poon, W.L. Self-regulated learning strategies & academic achievement in online higher education learning environments: A systematic review. Internet High. Educ.; 2015; 27, pp. 1-13. [DOI: https://dx.doi.org/10.1016/j.iheduc.2015.04.007]

16. Panadero, E. A Review of Self-regulated Learning: Six Models and Four Directions for Research. Front. Psychol.; 2017; 8, 422. [DOI: https://dx.doi.org/10.3389/fpsyg.2017.00422]

17. Luttrell, A.; Teeny, J.D.; Petty, R.E. Morality matters in the marketplace: The role of moral metacognition in consumer purchasing. Soc. Cogn.; 2021; 39, pp. 328-351. [DOI: https://dx.doi.org/10.1521/soco.2021.39.3.328]

18. Yadav, R. Altruistic or egoistic: Which value promotes organic food consumption among young consumers? A study in the context of a developing nation. J. Retail. Consum. Serv.; 2016; 33, pp. 92-97. [DOI: https://dx.doi.org/10.1016/j.jretconser.2016.08.008]

19. Suki, N.M. Green product purchase intention: Impact of green brands, attitude, and knowledge. Br. Food J.; 2016; 118, pp. 2893-2910. [DOI: https://dx.doi.org/10.1108/BFJ-06-2016-0295]

20. Ham, M.; Jeger, M.; Ivkovic, A.F. The role of subjective norms in forming the intention to purchase green food. Econ. Res.-Ekon. Istraz.; 2015; 28, pp. 738-748. [DOI: https://dx.doi.org/10.1080/1331677X.2015.1083875]

21. Yu, Y.; Li, X.; Jai, T.-M. The impact of green experience on customer satisfaction: Evidence from TripAdvisor. Int. J. Contemp. Hosp. Manag.; 2017; 29, pp. 1340-1361. [DOI: https://dx.doi.org/10.1108/IJCHM-07-2015-0371]

22. Liu, J.; Chen, T.; Hu, B. Consumer acceptance under hydrogen energy promotion policy: Evidence from Yangtze River Delta. Int. J. Hydrog. Energy; 2023; 48, pp. 11104-11112. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2022.07.081]

23. Guo, Q.; Liu, Y.; Cai, L. An experimental study on the potential purchase behavior of Chinese consumers of new energy hybrid electric vehicles. Front. Environ. Sci.; 2023; 11, 1159846. [DOI: https://dx.doi.org/10.3389/fenvs.2023.1159846]

24. Wu, Z.; He, Q.; Li, J.; Bi, G.; Antwi-Afari, M.F. Public attitudes and sentiments towards new energy vehicles in China: A text mining approach. Renew. Sustain. Energy Rev.; 2023; 178, 113242. [DOI: https://dx.doi.org/10.1016/j.rser.2023.113242]

25. Mostafa, M.M. A hierarchical analysis of the green consciousness of the Egyptian consumer. Psychol. Mark.; 2007; 24, pp. 445-473. [DOI: https://dx.doi.org/10.1002/mar.20168]

26. Kamalanon, P.; Chen, J.-S.; Le, T.-T.-Y. “Why Do We Buy Green Products?” An Extended Theory of the Planned Behavior Model for Green Product Purchase Behavior. Sustainability; 2022; 14, 689. [DOI: https://dx.doi.org/10.3390/su14020689]

27. Zhang, L.; Fan, Y.; Zhang, W.; Zhang, S. Extending the Theory of Planned Behavior to Explain the Effects of Cognitive Factors across Different Kinds of Green Products. Sustainability; 2019; 11, 4222. [DOI: https://dx.doi.org/10.3390/su11154222]

28. Maichum, K.; Parichatnon, S.; Peng, K.-C. Application of the Extended Theory of Planned Behavior Model to Investigate Purchase Intention of Green Products among Thai Consumers. Sustainability; 2016; 8, 1077. [DOI: https://dx.doi.org/10.3390/su8101077]

29. Kormanik, M.; Rocco, T. Internal Versus External Control of Reinforcement: A Review of the Locus of Control Construct. Hum. Resour. Dev. Rev.; 2009; 8, pp. 463-483. [DOI: https://dx.doi.org/10.1177/1534484309342080]

30. Flavell, J.H. Metacognition and cognitive monitoring: A new area of cognitive–developmental inquiry. Am. Psychol.; 1979; 34, pp. 906-911. [DOI: https://dx.doi.org/10.1037/0003-066X.34.10.906]

31. Min, B. Interplay of consumer expectation and processing fluency in perception of product innovativeness and product evaluation. Eur. J. Mark.; 2023; 57, pp. 283-324. [DOI: https://dx.doi.org/10.1108/EJM-08-2021-0603]

32. Alter, A.L.; Oppenheimer, D.M. Uniting the Tribes of Fluency to Form a Metacognitive Nation. Personal. Soc. Psychol. Rev.; 2009; 13, pp. 219-235. [DOI: https://dx.doi.org/10.1177/1088868309341564] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19638628]

33. Eraut, M. Non-formal learning and tacit knowledge in professional work. Br. J. Educ. Psychol.; 2000; 70, Pt 1, pp. 113-136. [DOI: https://dx.doi.org/10.1348/000709900158001] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10765570]

34. Beauregard, M.; Levesque, J.; Bourgouin, P. Neural correlates of conscious self-regulation of emotion. J. Neurosci.; 2001; 21, RC165. [DOI: https://dx.doi.org/10.1523/JNEUROSCI.21-18-j0001.2001] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11549754]

35. Hines, J.M. Analysis and Synthesis of Research on Responsible Environmental Behavior; Southern Illinois Univ.: Carbondale, IL, USA, 1984.

36. Trivedi, R.H.; Patel, J.D.; Savalia, J.R. Pro-environmental behaviour, locus of control and willingness to pay for environmental friendly products. Mark. Intell. Plan.; 2015; 33, pp. 67-89. [DOI: https://dx.doi.org/10.1108/MIP-03-2012-0028]

37. Du, H.; Liu, D.; Sovacool, B.K.; Wang, Y.; Ma, S.; Li, R.Y.M. Who buys New Energy Vehicles in China? Assessing social-psychological predictors of purchasing awareness, intention, and policy. Transp. Res. Part F-Traffic Psychol. Behav.; 2018; 58, pp. 56-69. [DOI: https://dx.doi.org/10.1016/j.trf.2018.05.008]

38. Tan, R.P.; Lin, B.Q. Are people willing to support the construction of charging facilities in China?. Energy Policy; 2020; 143, 111604. [DOI: https://dx.doi.org/10.1016/j.enpol.2020.111604]

39. Wang, H.H.; Han, X.; Jiang, Y.; Wu, G. Revealed consumers’ preferences for fresh produce attributes in Chinese online markets: A case of domestic and imported apples. PLoS ONE; 2022; 17, e0270257. [DOI: https://dx.doi.org/10.1371/journal.pone.0270257]

40. Zawadzki, S.J.; Steg, L.; Bouman, T. Meta-analytic evidence for a robust and positive association between individuals’ pro-environmental behaviors and their subjective wellbeing. Environ. Res. Lett.; 2020; 15, 123007. [DOI: https://dx.doi.org/10.1088/1748-9326/abc4ae]

41. Bissing-Olson, M.J.; Iyer, A.; Fielding, K.S.; Zacher, H. Relationships between daily affect and pro-environmental behavior at work: The moderating role of pro-environmental attitude. J. Organ. Behav.; 2013; 34, pp. 156-175. [DOI: https://dx.doi.org/10.1002/job.1788]

42. Soodla, P.; Jogi, A.-L.; Kikas, E. Relationships between teachers’ metacognitive knowledge and students’ metacognitive knowledge and reading achievement. Eur. J. Psychol. Educ.; 2017; 32, pp. 201-218. [DOI: https://dx.doi.org/10.1007/s10212-016-0293-x]

43. Choi, D.; Johnson, K.K.P. Influences of environmental and hedonic motivations on intention to purchase green products: An extension of the theory of planned behavior. Sustain. Prod. Consum.; 2019; 18, pp. 145-155. [DOI: https://dx.doi.org/10.1016/j.spc.2019.02.001]

44. Sun, Q.; Zhang, L.J.; Carter, S. Investigating Students’ Metacognitive Experiences: Insights From the English as a Foreign Language Learners’ Writing Metacognitive Experiences Questionnaire (EFLLWMEQ). Front. Psychol.; 2021; 12, 744842. [DOI: https://dx.doi.org/10.3389/fpsyg.2021.744842] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34526945]

45. Tan, S.; Zhong, L. IOP. Research on the survival and development of new energy vehicles in China. Proceedings of the 2nd International Workshop on Renewable Energy and Development (IWRED); Guilin, China, 20–22 April 2018.

46. Lin, Y.-H. Determinants of Green Purchase Intention: The Roles of Green Enjoyment, Green Intrinsic Motivation, and Green Brand Love. Sustainability; 2023; 15, 132. [DOI: https://dx.doi.org/10.3390/su15010132]

47. Sharma, N.; Paco, A. Moral disengagement: A guilt free mechanism for non-green buying behavior. J. Clean. Prod.; 2021; 297, 126649. [DOI: https://dx.doi.org/10.1016/j.jclepro.2021.126649]

48. Sarac, S.; Onder, A.; Karakelle, S. The Relations Among General Intelligence, Metacognition and Text Learning Performance. Egit. Ve Bilim-Educ. Sci.; 2014; 39, pp. 40-53.

49. Lian, J.; Li, L.H.; Zhou, Y.F. Research on Information Collection System of Energy Efficient and New Energy Vehicles. Inf.-Int. Interdiscip. J.; 2012; 15, pp. 4661-4666.

50. Chu, M.; Anders, S.; Deng, Q.; Contador, C.A.; Cisternas, F.; Caine, C.; Zhu, Y.; Yang, S.; Hu, B.; Liu, Z. et al. The future of sustainable food consumption in China. Food Energy Secur.; 2022; 12, e405. [DOI: https://dx.doi.org/10.1002/fes3.405]

51. Xu, Y.; Zhang, W.; Bao, H.; Zhang, S.; Xiang, Y. A SEM-Neural Network Approach to Predict Customers’ Intention to Purchase Battery Electric Vehicles in China’s Zhejiang Province. Sustainability; 2019; 11, 3164. [DOI: https://dx.doi.org/10.3390/su11113164]

52. Jensen, L.J.; Saric, J.; Bork, P. Literature mining for the biologist: From information retrieval to biological discovery. Nat. Rev. Genet.; 2006; 7, pp. 119-129. [DOI: https://dx.doi.org/10.1038/nrg1768]

53. Pinero, J.; Bravo, A.; Queralt-Rosinach, N.; Gutierrez-Sacristan, A.; Deu-Pons, J.; Centeno, E.; Garcia-Garcia, J.; Sanz, F.; Furlong, L.I. DisGeNET: A comprehensive platform integrating information on human disease-associated genes and variants. Nucleic Acids Res.; 2017; 45, pp. D833-D839. [DOI: https://dx.doi.org/10.1093/nar/gkw943] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27924018]

54. Srinivasan, P. Text mining: Generating hypotheses from MEDLINE. J. Am. Soc. Inf. Sci. Technol.; 2004; 55, pp. 396-413. [DOI: https://dx.doi.org/10.1002/asi.10389]

55. Qiu, Q.; Xie, Z.; Wu, L.; Li, W. Geoscience keyphrase extraction algorithm using enhanced word embedding. Expert Syst. Appl.; 2019; 125, pp. 157-169. [DOI: https://dx.doi.org/10.1016/j.eswa.2019.02.001]

56. Levy, O.; Goldberg, Y. Neural Word Embedding as Implicit Matrix Factorization. Proceedings of the 28th Conference on Neural Information Processing Systems (NIPS); Montreal, QC, Canada, 8–13 December; 2015.

57. Blei, D.M.; Ng, A.Y.; Jordan, M.I. Latent Dirichlet allocation. J. Mach. Learn. Res.; 2003; 3, pp. 993-1022. [DOI: https://dx.doi.org/10.1162/jmlr.2003.3.4-5.993]

58. Isojarvi, J.; Aspara, J. Consumers’ behavioural responses to price promotions of organic products: An introspective pre-study and an online field experiment. Eur. J. Mark.; 2023; [DOI: https://dx.doi.org/10.1108/EJM-04-2022-0321]

59. Miyamoto, T.; Yajima, N.; Tsukahara, T.; Arimura, T.H. Advancement of Green Public Purchasing by Category: Do Municipality Green Purchasing Policies Have Any Role in Japan?. Sustainability; 2020; 12, 8979. [DOI: https://dx.doi.org/10.3390/su12218979]

60. Schober, P.; Boer, C.; Schwarte, L.A. Correlation Coefficients: Appropriate Use and Interpretation. Anesth. Analg.; 2018; 126, pp. 1763-1768. [DOI: https://dx.doi.org/10.1213/ANE.0000000000002864] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29481436]

61. Liu, Q.; Ding, T.; Li, Y.; Gao, Z.; Guo, Y. Function Mechanism of Intellectual Property Capability on Relay Innovation Based on CWLBGSO-DAG-Bootstrap SEM: Mediating Effect of Knowledge Matching and Moderating Effect of Relationship Learning. Comput. Intell. Neurosci.; 2022; 2022, 4357917. [DOI: https://dx.doi.org/10.1155/2022/4357917]

62. Liu, W.; Zeng, L.; Wang, Q. Psychological Distance Toward Air Pollution and Purchase Intention for New Energy Vehicles: An Investigation in China. Front. Psychol.; 2021; 12, 569115. [DOI: https://dx.doi.org/10.3389/fpsyg.2021.569115]

63. Muslim, N.H.; Keyvanfar, A.; Shafaghat, A.; Abdullahi, M.a.M.; Khorami, M. Green Driver: Travel Behaviors Revisited on Fuel Saving and Less Emission. Sustainability; 2018; 10, 325. [DOI: https://dx.doi.org/10.3390/su10020325]

64. Zhang, X.; Wang, K.; Hao, Y.; Fan, J.-L.; Wei, Y.-M. The impact of government policy on preference for NEVs: The evidence from China. Energy Policy; 2013; 61, pp. 382-393. [DOI: https://dx.doi.org/10.1016/j.enpol.2013.06.114]

65. Zheng, S.; Liu, H.; Guan, W.; Yang, Y.; Li, J.; Fahad, S.; Li, B. Identifying Intention-Based Factors Influencing Consumers’ Willingness to Pay for Electric Vehicles: A Sustainable Consumption Paradigm. Sustainability; 2022; 14, 16831. [DOI: https://dx.doi.org/10.3390/su142416831]

66. Hinnuber, F.; Szarucki, M.; Szopik-Depczynska, K. The Effects of a First-Time Experience on the Evaluation of Battery Electric Vehicles by Potential Consumers. Sustainability; 2019; 11, 7034. [DOI: https://dx.doi.org/10.3390/su11247034]

67. Connor, W.D.; Wang, Y.; Malikopoulos, A.A.; Advani, S.G.; Prasad, A.K. Impact of Connectivity on Energy Consumption and Battery Life for Electric Vehicles. Ieee Trans. Intell. Veh.; 2021; 6, pp. 14-23. [DOI: https://dx.doi.org/10.1109/TIV.2020.3032642]

68. Zhang, H.; Zhao, F.; Hao, H.; Liu, Z. Effect of Chinese Corporate Average Fuel Consumption and New Energy Vehicle Dual-Credit Regulation on Passenger Cars Average Fuel Consumption Analysis. Int. J. Environ. Res. Public Health; 2021; 18, 7218. [DOI: https://dx.doi.org/10.3390/ijerph18147218]

69. Zhao, F.; Hao, H.; Liu, Z. Technology strategy to meet China’s 5 L/100 km fuel consumption target for passenger vehicles in 2020. Clean Technol. Environ. Policy; 2016; 18, pp. 7-15. [DOI: https://dx.doi.org/10.1007/s10098-015-1019-5]