Understanding Electric Vehicle Adoption in Türkiye: Analyzing User Motivations Through the Technology Acceptance Model

Abstract

The popularity of electric vehicles offers the opportunity to analyze decision-making processes by examining user behavior. Determining the motivation of the user to use an innovation will guide decision-makers in supporting the innovation in question. This study investigates the factors electric car users in Türkiye consider based on the Technology Acceptance Model. A questionnaire was used to measure Perceived Ease of Use, Perceived Usefulness, and Intention to Use with the external factors of Subjective Norm, Compatibility, and Image. The relationships were analyzed with PLS-SEM established with the participation of 414 electric vehicle users. Subjective Norms and Image directly impact Perceived Usefulness, Perceived Ease of Use, and Intention to Use. It has been determined that Compatibility has a direct effect on Ease of Use and an indirect effect on Usefulness and Intention. According to this study, in which most people are dissatisfied with charging and range issues, the opinion of the social environment and family is the most important external factor affecting intention. Our findings suggest improving the charging station network and technology, as well as implementing informative activities related to the features of electric vehicles, in order to contribute to users’ adoption of electric vehicles.

1. Introduction

The Global Electric Vehicle Outlook 2024 report, prepared by the International Energy Agency (IEA), highlights key developments in the electric vehicle (EV) industry and the rising adoption of EVs [1]. The report projects that 17 million electric vehicles will be sold in 2024, meaning that one in every five cars sold globally will be electric. Most of these sales are concentrated in China (60%), followed by Europe (25%) and the United States (10%). The IEA’s forecast suggests that by 2030, one in every three cars in China and one in every five cars in Europe and the US will be electric. The number of electric vehicles on the road is set to grow significantly, with over ten million currently in operation (Figure 1).

Electric vehicles (EVs) refer to vehicles that convert electrical energy into kinetic energy and are categorized into three main types: Hybrid Electric Vehicles (HEVs), Plug-in Hybrid Electric Vehicles (PHEVs), and Battery Electric Vehicles (BEVs). HEVs and PHEVs have an internal combustion engine and an electric motor. The critical difference is that PHEVs can be recharged via external power sources, unlike HEVs, which rely solely on regenerative braking for charging. On the other hand, BEVs run entirely on electricity, with rechargeable batteries serving as their only power source.

Electric vehicles offer several benefits, such as low emissions, reduced impact on air quality, purchasing incentives, and lower operating costs. EVs generally have lower running costs than fossil fuel-powered vehicles, as electricity is cheaper than gasoline or diesel. Additionally, EVs require less maintenance since their engines have fewer moving parts and do not need frequent services like oil changes. Many countries provide incentives to promote the use of EVs, including subsidies, tax breaks, and investments in charging infrastructure. According to Tax Benefits and Incentives Report of Acea [2], in Germany, funding applications have no longer been accepted since 31 December 2023. France offers partial tax exemptions and subsidies based on emission levels, with up to EUR 5000 in scrappage incentives for low-emission vehicle purchases. In Italy, fully electric or hybrid vehicles are tax-exempt for five years, with subsidies changing between EUR 5000–EUR 11,000 for low-emission models. In Spain, low-emission vehicles are tax-exempt, and electric vehicles can receive up to EUR 7000 in subsidies with scrappage incentives [2]. Compared with these examples in EU countries, it is evident that there is no direct purchase support or scrappage incentive program for fully electric and hybrid vehicles in Türkiye. Türkiye has announced an additional 40% customs duty on China-origin BEV (Battery Electric Vehicle) vehicles. However, fully electric vehicles benefit from tax support through lower special consumption tax (ÖTV) brackets compared with their internal combustion engine counterparts. There is no similar support for hybrid electric vehicles. The motor vehicle tax (MTV) for fully electric vehicles is 25% of that for their internal combustion counterparts. EVs also showcase cutting-edge technologies, with ongoing advancements in battery efficiency, energy management systems, and autonomous driving features. Alongside these innovations, EVs offer improved comfort and driving performance, which appeal to potential buyers.

When the studies investigating the factors affecting the intention to use electric cars are examined, the most common factors considered are Environmental Concerns, Subjective Norms, Charging, Attitude, Range, and Economic Characteristics (

Table 1. Summary of research studies on intention to use electric car.

(Authors, Year) Country	Attributes Investigated	Method Used	Main Findings
Heffner et al., 2007 [3] USA	Environmental Concerns; Ethics; Reduce Support to Oil Producers; Embrace New Technology	Semi-structured, ethnographic interviews	The symbolic meaning of owning an HEV differs across households (e.g., protecting the family’s future, technological interests, avoiding fuel dependency)
Ozaki and Sevastyanova, 2011 [4] UK	Financial Benefits; Social Norms; Norms of Groups	Exploratory Factor Analyses	Knowledge about hybrid cars would help overcome fears about technical and practical performance.
Egbue and Long, 2012 [5] USA	Range; Cost; Charging Infrasturcture; Realiability; Safety; Experience; Interest; Sustainability	Chi-square test	Attitudes, knowledge, and perceptions on EVs differ across gender, age, and education levels. Sustainability and environmental benefits follow cost and performance as effective factors.
Hackbarth and Madlener, 2013 [6] Germany	Purchase Price; Fuel cost; CO2 emissions; Recharging Time; Policy Incentives	Mixed Logit Model	Young, well-educated, environmentally aware car buyers are the most sensitive to EV adoption.
Junquera et al., 2016 [7] Spain	Charging Times; Price; Range; Age	Logistic Regression Analysis	Willingness to buy EV decreases as the price and the charging time increase.
Schmalfuß et al., 2017 [8] Germany	Direct Experience	Theory of Planned Behavior	Findings reveal the effects of experience on attributes and attitudes but no effect on purchase intention.
Wang et al., 2017 [9] China	Incentives; Information Policies; Convenience; Environmental Concern	Partial Least Squares Structural Equation Modeling (PLS-SEM)	Convenience policy measures (e.g., permissions for some privileges) are found to be the most effective factor in EV purchases.
Bigerna and Micheli, 2018 [10] Italy	Climate Awareness; Fuel Economy; Car Use Frequency; Bus Users; Range	Fuzzy Set Comparative Analysis	Public Transportation users have more intention to use EVs. Attitudes related to EVs differ across age.
Huang and Ge, 2019 [11] China	Subjective Norm; Attitude; Perceived Behavior Control; Cognitive Status; Product Perception; Incentives.	Theory of Planned Behavior	Significant differences among demographic variables are determined. Subjective Norms and incentives do not affect purchasing intention.
Simsekoglu and Nayum, 2019 [12] Norway	Perceive Accident Risk; Knowledge; Perceived Car Attributes; Subjective Norm	Theory of Planned Behavior	Economic, environmental attributes, Subjective Norms, and perceived behavioral control positively relate to the intention to buy EVs.
Sovacool et al., 2019 [13] China	Cost; Policy Supports; Performance; Charging; Knowledge	Statistical and Multivariate Analysis, Principal Component Analysis.	Performance of EVs, perceived benefits of driving an EV, and promotions are found to be related to adopting EVs.
Tu and Yang, 2019 [14] China	Attitude Toward Behavior (Perceived Usefulness, Perceived Ease of Use, Compatibility, Personal Innovativeness); Subjective Norm (Interpersonal and External Influence); Self Control Ability (Self Efficacy; Facilitating Condition; Perceived Behavior Control)	Framework based on Theory of Planned Behavior (TPB), Technology Acceptance Model (TAM), and Innovation Diffusion Theory (IDT),	Resources required to purchase EV, and environmental and national benefits positively affect purchase intention. Opinions of others do not have any effect.
Yang et al., 2020 [15] China	Knowledge; Brand Trust; Perceived Risk	Structural Equation Modeling	Consumer attitudes are found to be the most influential factor in purchasing EVs.
Curtale et al., 2021 [16] The Netherlands	Social Influence; Performance Expectancy; Effort Expectancy; Anxiety-free Experience; Trust; Personal Attitude; Car Ownership	Unified Theory of Acceptance and Use of Technology (UTAUT)	Social influence is the most important factor affecting behavioral intention.
Dutta and Hwang, 2021 [17] Taiwan	Attitude; Subjective Norm; Perceived Behavioral Control; Vehicle Performance; Purchasing Price; Charging Facility; Maintenance and Battery Cost	Theory of Planned Behaviour, SEM, CFA	Users in Taiwan are concerned about the greenhouse effects on the environment.
Emanović et al., 2022 [18] Croatia	Economic Factors; Battery; Charging Stations	Statistical and Visual Data Analysis	Range and range anxiety is the most significant barrier, followed by economic concerns.
Kocagöz and İğde, 2022 [19] Türkiye	EV Attribute Evaluation, Perceived Price, Environmental Concerns	Correlation and Regression Analysis, T-test, Anova	Perceived Price Value and environmental concerns have strong effects on EV purchase intention. Differences are determined across gender and marital status.
N. Wang et al., 2022 [20] China	Vehicle–Grid Integration; Charging Schedule; Initial Trust; Social Value; Social Influence	Technology Acceptance Model	Initial Trust, Social Value, and Social Influence have positive indirect effects, while perceived risk has a direct and negative effect. Users’ acceptance differed among marital status and age, but showed no difference among gender, education, and income levels.
Buranelli De Oliveira et al., 2022 [21] Brazil	Attitude; Emotions; Subjective Norm; Perceived Behavioral Control; Complexity; Relative Advantage; Compatibility	Structural Equation Modeling	Perceived Relative Advantage and Compatibility have a positive effect on intention through attitudes. Attitudes and emotions are the most influential factors in the intention to use.
Yeğin and Ikram, 2022 [22] Türkiye	Environmental Concerns; Green Trust (environmental effects of EVS); Subjective Norm; Attitude; Perceived Behavioral Control	Structural Equation Modeling based on the Theory of Planned Behavior	Environmental concern and green trust are the leading components of the intention to use EVs.
Arora and Singh, 2024 [23] India	Operational benefits; Environmental Concerns; Attitude; Social Norms; Personal Norms; Trialability	Principal Component Analysis	The significant factors in EV purchase intention are operational benefits, trialability, and positive attitude.
Buhmann et al., 2024 [24] Spain	Attitude; Perceived Behavioral Control; Subjective Norm; Moral Norm	Extended Theory of Planned Behaviour	Environmental concerns, pricing strategies, and incentives could influence BEV adoption.

).

Product quality is a crucial factor influencing consumers’ purchasing decisions. Several factors shape purchasing decisions regarding electric vehicles (EVs), including product attributes, price, performance, comfort, quietness, ease of driving, financial savings, and environmental considerations. A study by Ozaki and Sevastyanova [4] surveyed 1263 participants in the United Kingdom to identify factors influencing the adoption of EVs and hybrid vehicles. These factors were grouped based on their impact on comfort, quietness, ease of driving, and vehicle performance [4]. In Italy, studies have shown that range anxiety—concerns about a vehicle’s driving range—is the most significant factor affecting EV purchases. However, the limited availability of charging infrastructure remains a barrier to widespread adoption [2,25].

In Norway, research by Bjerkan et al. [26] found that financial savings motivate consumers to choose EVs. A survey by the Norwegian EV Association revealed that most buyers cited financial incentives, such as tax exemptions on electric vehicles, as their primary reason for purchasing EVs. These incentives are perceived as critical in influencing consumer choices [26]. Additionally, factors such as the lower purchase price and reduced fuel costs of electric vehicles play a significant role in consumer decision-making. EVs generally offer lower operating costs, encouraging buyers to choose them over traditional vehicles [3,6].

A study involving 481 users found that attitudes toward electric vehicles (EVs) can vary based on gender, age, and education levels [10]. Bigerna and Misheli [10] further explored attitudes across different age groups, identifying environmental concerns as a critical factor influencing EV adoption. In contrast with Egbue and Long’s [5] study, their research concluded that there is no need for separate policies for different age groups but noted that public transportation users are more inclined to consider EVs [5]. In a study conducted in Croatia, the primary barrier to EV adoption was a lack of technical knowledge. Participants also highlighted factors such as the charging network, battery life, and vehicle design as crucial considerations [18].

Challenges remain, such as long charging times, limited infrastructure, and shorter driving ranges. Rapid progress in battery technology and charging station availability are addressing these issues. Consumers increasingly expect convenient and fast charging options, whether through public fast-charging stations, home charging units, or an expanding network of charging points. These improvements in charging accessibility play a significant role in the growing adoption of electric vehicles for everyday use.

In Türkiye, BEV sales have seen a significant increase of 251% between the first quarter of 2022 and the first quarter of 2023. In 2023, 22,640 electric vehicles were sold in Türkiye, accounting for 12.9% of total vehicle sales. Similar sales proportions were observed in the U.S. (14.6%) and Australia (13.0%), while the average in Europe was 48.1% and in China, it was 28.8%. The average market share across the examined markets is 26.8%. The EU’s international trade of used electric vehicles is growing rapidly, with the market increasing 70% in both 2021 and 2022. The majority of the trade is within the EU, while countries like Norway, the UK, and Türkiye are also key trading partners. Belgium, Germany, the Netherlands, and Spain are the largest exporters [27]. As of August 2024, a report from the Turkish Statistical Institute shows there were 137,009 electric and 316,742 hybrid vehicles in Türkiye, representing 0.9% and 2% of the total vehicle population, respectively [28]. While research on EV adoption in Türkiye is limited, several studies have addressed the topic [3,4,5]. In a survey of 626 participants, Yeğin and Ikram [22] identified factors influencing the intention to purchase an EV, including environmental concerns, attitudes toward EV ownership, purchasing capacity, social approval, and the environmental benefits of EVs. Another study by Kocagöz and İğde [19] examined factors such as driving pleasure, noise reduction, reputation, range, charging, and safety, finding that the purchase value and characteristics like gender and marital status influenced the intention to buy an EV.

This study contributes to the literature by investigating the direct effects of Image and Compatibility on Perceived Ease of Use, Perceived Usefulness, and Intention to Use. The objective of this study is to explore the factors influencing the intention to adopt EVs in Türkiye, including the following points:

• Hypotheses to understand the factors of Turkish users’ intention for EVs were prepared based on the extended Technology Acceptance Model, including the effect of Social Norms, Image, and Compatibility.
• A questionnaire was prepared using the relevant literature, and data were collected accordingly.
• A Partial Least Square Structural Equation Model was established, and several tests were conducted in SMART PLS4 [29] to verify the model’s quality.
• Hypotheses were tested to identify the factors influencing the intention to use electric cars.

The second section of this study outlines the research model and details the collected data. The third section presents model results, followed by a discussion in the fourth section. The fifth section concludes the main findings briefly.

2. Materials and Methods

In the methods chosen to examine the relationships between the factors affecting the intention to use electric cars, it can be seen that structural equation models, which use user characteristics and behavior data obtained through survey studies based on technology acceptance or behavioral theories, come to the fore (Table 1). The adoption of new technologies has long been studied within behavioral sciences. Ajzen and Fishbein [30] developed the Theory of Reasoned Action, which examines the relationship between attitudes and behavior [30]. In 1989, Davis [31] proposed the Technology Acceptance Model (TAM), identifying two key factors influencing users’ acceptance of an innovation: Perceived Usefulness and Ease of Use. Due to criticisms that TAM did not account for external factors, the model was later expanded, leading to the development of the Unified Theory of Acceptance and Use of Technology (UTAUT) and UTAUT 2, which incorporated external variables such as Subjective Norms, Image, Voluntariness, and Result Demonstrability [32,33]. Moore and Benbasat [34] further examined innovation acceptance through the lens of Rogers’ [35] Diffusion of Innovation Theory, identifying factors such as Relative Advantage, Compatibility, Trialability, and Image Visibility as influential in the adoption process.

2.1. Method

This study will analyze users’ intention to adopt electric vehicles in Türkiye using the Technology Acceptance Model (TAM) developed by Davis [31]. Initially created to examine and explain computer usage behavior in the workplace, TAM is rooted in Fishbein and Ajzen’s [16] “Theory of Planned Behavior”. Over time, as technological innovations emerged and new fields of inquiry were explored, the model’s scope expanded. Many researchers have since used TAM to clarify the acceptance of various technological innovations [36].

The primary goal of TAM is to explore how external variables influence internal factors such as beliefs, attitudes, and intentions. Davis [31] identified two critical determinants in explaining a user’s intention to adopt a new system: “Perceived Ease of Use” (PEU) and “Perceived Usefulness” (PU). The relationship between these factors and their role in predicting user behavior is central to the TAM structure, as illustrated in Figure 2.

Perceived Usefulness refers to an individual’s belief that using a new system will enhance work efficiency and provide benefits. The other construct, Perceived Ease of Use, reflects the effort a person believes they will need to exert to use the system effectively [31].

These concepts could be applied similarly in electric vehicles, which can be seen as innovative due to their unique charging methods and advanced features. Based on this information, the first three hypotheses of the study are formulated as follows:

H1: 

Perceived Ease of Use has a positive effect on Perceived Usefulness.

H2: 

Perceived Usefulness has a positive effect on Usage Intention.

H3: 

Perceived Ease of Use has a positive effect on Usage Intention.

External Factors

The Technology Acceptance Model (TAM) has been extensively utilized in various research domains, including online shopping, electric vehicle acquisitions, and virtual card utilization [19,37]. Despite its benefits, TAM has encountered criticism for its failure to consider external factors. Bagozzi [38] contended that the model’s simplicity fails to account for complex human behaviors and environmental influences, resulting in an inadequate understanding of these elements. Likewise, Chuttur [39] observed that TAM research has reached a saturation point, with numerous studies emphasizing the model’s strengths while overlooking its limitations. In response to these critiques, researchers have expanded TAM to incorporate the impact of external factors [31,32,40]. Venkatesh and Davis [40] introduced TAM2 by including the concepts of Subjective Norm and Image into TAM, which Fishbein and Ajzen [30] used while developing the Theory of Reasoned Action. Subjective Norm is defined as a person’s perception of the opinions of people important to him/her about whether the person should or should not use an innovation [30]. Moore and Benbasat [34] defined Image as the extent to which a person’s social status improves by using an innovation. To explore if electric vehicle usage is perceived as a status symbol in the social environment, it is assumed that the Image of EV ownership and Subjective Norms could influence Perceived Ease of Use, Usefulness, and the Intention to Use. Accordingly, the following hypotheses have been developed:

H4: 

Image has a positive effect on Perceived Usefulness.

H5: 

Image has a positive effect on Perceived Ease of Use.

H6: 

Image has a positive effect on the Intention to Use.

H7: 

Subjective Norm has a positive effect on Perceived Usefulness.

H8: 

Subjective Norm has a positive effect on Perceived Ease of Use.

H9: 

Subjective Norm has a positive effect on the Intention to Use.

Compatibility is the degree to which an innovation is perceived to be compatible with existing values, experiences, and potential needs. In their comprehensive study, Karahanna et al. [41] defined Compatibility in four dimensions: (1) compatibility with the user’s current work process, (2) compatibility with preferred work style, (3) compatibility with prior experience, and (4) compatibility with the user’s dominant value system. One of the most significant innovations for electric vehicle users is the charging process and the range these vehicles offer [41]. The experience dimension of Compatibility is especially relevant when considering users’ previous experiences with traditional fuel supply and vehicle range. The following hypotheses were added to explore if Compatibility influences the intention to use EVs:

H10: 

Compatibility has a positive effect on Perceived Usefulness.

H11: 

Compatibility has a positive effect on Perceived Ease of Use.

H12: 

Compatibility has a positive effect on the Intention to Use.

The constructs used in the study and the hypotheses used to test the relationships between constructs are shown in Figure 3.

When addressing structural models that involve a predictive approach, do contain multiple measurement constructs, or do not follow a normal distribution, Hair et al. (2019) recommend using the Partial Least Squares Structural Equation Modeling (PLS-SEM) method. This method is particularly suited for models focusing on theory development rather than theory testing [42]. This study will employ PLS-SEM to determine whether the hypothesized relationships exist. The model will be developed and analyzed using the Smart PLS-4 software [29].

2.2. Data

A questionnaire was conducted to test the research model and hypotheses presented earlier (

Table 2. Measurement Items.

Part 1—Multiple Choice Questions
Do you know about electric cars?
Have you used electric cars before?
Part 2—Answer Questions between 1–5 (1 Do not agree strongly; 2 Do not agree; 3 Neither disagree nor agree; 4 Agree; 5 Agree strongly)
PU 1	Perceived Usefulness	Electric cars are useful because they eliminate dependency on fossil fuels.
PU 2	Perceived Usefulness	Electric cars are useful because they are environmentally friendly
PU 3	Perceived Usefulness	The advantages of using an electric car outweigh the disadvantages.
PU 4	Perceived Usefulness	Electric cars are useful because they support technological development.
PEU 1	Perceived Ease of Use	I think an electric car is easy to use.
PEU 2	Perceived Ease of Use	I think it is easy to learn how to use an electric car.
PEU 3	Perceived Ease of Use	I think it is easy to learn how to charge an electric car.
PEU 4	Perceived Ease of Use	I think I can easily charge an electric car anywhere.
UI 1	Intention to Use	I have a desire to buy an electric car in the future
UI 2	Intention to Use	I plan to buy an electric car in the future.
UI 3	Intention to Use	I will buy an electric car in the future.
UI 4	Intention to Use	I prefer an electric car if there were enough charging stations in Türkiye
SN 1	Subjective Norm	My social circle (family, friends, neighbors, colleagues, etc.) thinks I should drive an electric car.
SN 2	Subjective Norm	My social circle would be happy if I drove an electric car.
IM 1	Image	Driving an electric car affects my reputation in my circle.
IM 2	Image	People who drive electric cars are prestigious.
IM 3	Image	Owning an electric car is a symbol of high status.
CMP 1	Compatibility	I do not see it as a problem that it takes longer to charge an electric car than to refuel.
CMP 2	Compatibility	Charging an electric car will not cause a problem in my daily life.
CMP 3	Compatibility	I find the range of electric cars satisfactory.
Part 3. Demographic Questions
Your Gender
Your Age
Your Marital Status
Your Graduation Status
Your Monthly Income

). In addition to 5 demographic questions, 22 items were designed to assess participants’ electric vehicle usage habits. These items were categorized under Perceived Usefulness, Perceived Ease of Use, Purchase Intention, Image, Social Norm, and Compatibility. Participants rated each statement on a 5-point Likert scale (1—strongly disagree, 2—disagree, 3—neither agree nor disagree, 4—agree, and 5—strongly agree). The questions were adapted from various sources [19,22,40,43].

According to Hair et al. [44], the minimum sample size required for PLS-SEM analysis is ten times the maximum number of indicators used to measure a construct or the maximum number of paths directed toward a construct. Based on this approach, the minimum sample size required for the analysis has been set at 50. In another approach, considering the sample sizes suggested for different R² values (0.10, 0.25, 0.50, and 0.75) and significance levels (1%, 5%, and 10%), as recommended by Cohen [45] and tabulated by Hair et al. [44], the maximum suggested minimum value is found to be 212. The minimum number of sample requirement is satisfied once access to more than this number of individuals is obtained. A total of 414 valid responses were collected between 16 March 2024 and 18 April 2024.

3. Results

3.1. Descriptive Analysis of Questionnaire

The demographic characteristics of the participants are summarized in

Table 3. Demographic Characteristics of Participants.

Demographics	N	%
Gender
Female	153	36.96
Male	261	63.04
Age
18–25	71	17.15
26–35	172	41.55
36–45	87	21.01
46–55	61	14.73
>55	23	5.56
Marital Status
Single	205	49.52
Married—w/o kids	35	8.45
Married—w kids	174	42.03
Education Status
Primary	4	0.97
High School	40	9.66
Pre-graduate	31	7.49
Graduate	276	66.66
Postgraduate	63	15.22
Monthly Income * ($)
<531	51	12.32
531–781	24	5.80
781–1094	60	14.49
1094–1563	130	31.40
1563–2344	83	20.05
>2344	66	15.94

* According to the Central Bank of the Republic of Türkiye, 1 US dollar was worth approximately 32 Turkish Liras when the study was conducted.

. The study surveyed 414 individuals, comprising 36.96% females (153 participants) and 63.04% males (261 participants). Most participants (41.55%) were aged 26–35, followed by 17.15% aged 18–25. Those in the 36–45 age group comprised 21.01%, 14.73% were aged 46–55, and 5.56% were over 55. Regarding marital status, nearly half (49.52%) of the respondents were single. A significant proportion (42.03%) were married with children, while 8.45% were married without children. Most participants (66.66%) had completed graduate education, while 15.22% had postgraduate degrees. Smaller percentages had completed pre-graduate (7.49%), high school (9.66%), or primary education (0.97%). Monthly income varied, with the largest group (31.40%) earning between USD 1094 and USD 1563. About 20.05% earned between USD 1563 and USD 2344, while 15.94% had incomes exceeding USD 2344. Additionally, 12.32% earned less than USD 531, 5.80% earned between USD 531 and USD 781, and 14.49% earned between USD 781 and USD 1094.

Of the participants, 299 people, representing 64.4%, stated that they knew electric vehicles, while 155 participants (24.8%) indicated they had detailed knowledge. The 50 participants who declared they did not know were excluded from the study. Those who knew were asked whether they used electric cars; the results are presented in

Table 4. Participants’ Electric Car Usage Characteristics.

Do You Have Any Information about Electric Cars?	N	%
Yes, I have some information	299	64.44
Yes, I have detailed information.	115	24.78
No, I do not have any information	50	10.78
Have you driven an electric car?
Yes	93	22.46
No	321	77.54

.

The distribution of the questions related to measuring their intentions to use an electric car is summarized in

Table 5. Distribution of Participants’ Responses to Questions Regarding Their Electric Car Usage Intentions.

	1—Disagree strongly		2—Disagree		3—Neither Agree nor Disagree		4—Agree		5—Agree Strongly		Mean	Standard Dev.
Items	N	%	N	%	N	%	N	%	N	%
PU 1	31	7	26	6	72	17	129	31	156	38	3.85	1.21
PU 2	30	7	25	6	58	14	122	29	179	43	3.95	1.21
PU 3	32	8	46	11	109	26	115	28	112	27	3.55	1.21
PU 4	15	4	21	5	62	15	162	39	154	37	4.01	1.02
PEU 1	17	4	30	7	104	25	131	32	132	32	3.80	1.09
PEU 2	9	2	15	4	66	16	135	33	189	46	4.16	0.96
PEU 3	15	4	19	5	67	16	141	34	172	42	4.05	1.04
PEU 4	88	21	123	30	130	31	42	10	31	7	2.53	1.15
UI 1	47	11	45	11	89	21	115	28	118	29	3.51	1.31
UI 2	55	13	51	12	113	27	88	21	107	26	3.34	1.34
UI 3	55	13	61	15	120	29	76	18	102	25	3.26	1.33
UI 4	33	8	34	8	74	18	128	31	145	35	3.77	1.23
SN 1	77	19	99	24	138	33	58	14	42	10	2.73	1.21
SN 2	38	9	63	15	140	34	105	25	68	16	3.25	1.17
IM 1	79	19	66	16	119	29	81	20	69	17	2.99	1.34
IM 2	126	30	67	16	111	27	67	16	43	10	2.60	1.34
IM 3	140	34	86	21	102	25	50	12	36	9	2.41	1.30
CMP 1	133	32	97	23	73	18	54	13	57	14	2.53	1.41
CMP 2	111	27	105	25	96	23	60	14	42	10	2.56	1.30
CMP 3	116	28	105	25	124	30	46	11	23	6	2.41	1.17

and Figure 4. Table 5 presents participant responses across various items, with percentages for each response category (from “Disagree strongly” to “Agree strongly”), along with the mean and standard deviation for each item. Generally, the highest levels of agreement were found in the Perceived Usefulness and Perceived Ease of Use, while the Compatibility and Image sections displayed the highest disagreement.

Figure 4, which visualizes Table 5, presents the distribution of responses more clearly. Accordingly, when examining the items related to PU (Perceived Usefulness), it can be seen that the distribution is similar for all questions. Participants generally reported a high level of positive opinions regarding the PU questions. In the case of PEU (Perceived Ease of Use), the distributions are similar for all questions except for PEU4. The PEU4 question (“I think I can easily charge an electric vehicle anywhere”) is related to the ease of charging and has a lower average (2.53) compared with the other PEU questions (with averages of 3.80, 4.16, and 4.05, respectively). This indicates that individuals who generally do not have problems using the vehicle or learning the charging process expressed negative opinions regarding the ease of charging. The same situation applies to UI (Usage Intentions). The IU4 question, which states that the intention to purchase a vehicle will increase with the increase in number of charging stations, shows a slightly different distribution compared with the other three questions in the construct (IU1, IU2, IU3). This suggests that the expansion of charging infrastructure may positively influence participants’ intentions to purchase vehicles.

3.2. Results Regarding the PLS-SEM Model

The model analyzed using PLS-SEM was examined with Smart PLS-4 software and is assessed in two components: the measurement model and the structural model. The evaluation of the model involves assessing each component separately to verify if the model meets the necessary criteria. Quality controls are conducted based on the recommendations of Hair et al. [42].

3.2.1. Evaluation of the Reflective Measurement Model

As suggested by Hair et al. [42], it is recommended that the indicator loadings of the variables should be greater than 0.708. In this way, it is determined that the structure can explain more than 50% of the variable variance. Most of the indicator loadings are found to be greater than 0.708 (

Table A1. Indicator loadings of the first proposed model.

	COMP	IM	IU	PEU	PU	SN
CMP-1	0.846
CMP-2	0.895
CMP-3	0.749
IM-1		0.884
IM-2		0.945
IM-3		0.910
IU-1			0.918
IU-2			0.950
IU-3			0.936
IU-4			0.745
PEU-1				0.799
PEU-2				0.710
PEU-3				0.695
PEU-4				0.701
PU-1					0.873
PU-2					0.862
PU-3					0.845
PU-4					0.785
SN-1						0.915
SN-2						0.915

).

For internal consistent reliability and Convergent Validity, it has been stated that the composite reliability (CR) values suggested by Jöreskog [46] are acceptable between the values of (0.6 and 0.7) in exploratory studies [26]. The values between 0.7 and 0.9 indicate a satisfactory–good reliability. According to Hair et al. [42], Diamantopoulos et al. [47], and Drolet and Morrison [48], values of 0.95 and above reduce the validity of the structure and are therefore considered problematic. Another reliability measure is Cronbach’s alpha value. This value includes less sensitivity than CR. Another measurement that falls between these two values and is more balanced is the ρA value.

If the Average Variance Extracted (AVE) value used for the Convergent Validity control exceeds 0.5, the structure explains 50% of the variable variances. Factors presented in

Table A2. Construct reliability and validity.

	Cronbach’s Alpha	Composite Reliability (rho_a)	Composite Reliability (rho_c)	Average Variance Extracted (AVE)
COMP	0.776	0.796	0.871	0.693
IM	0.901	0.909	0.938	0.834
IU	0.910	0.916	0.939	0.794
PEU	0.712	0.719	0.818	0.529
PU	0.863	0.866	0.907	0.709
SN	0.806	0.806	0.911	0.837

demonstrate strong reliability and Convergent Validity for the reflective measures of this study.

The HTMT (heterotrait–monotrait) measurement proposed by Henseler et al. [49] is used to determine how discrete the structures in the structural model are (discriminant validity). An HTMT value greater than 0.9 indicates that the model is problematic regarding discriminant validity. The HTMT values of the research model are given in

Table A3. Discriminant validity heterotrait–monotrait ratio (HTMT).

	COMP	IM	IU	PEU	PU	SN
COMP
IM	0.598
IU	0.502	0.605
PEU	0.524	0.459	0.658
PU	0.430	0.528	0.750	0.633
SN	0.575	0.717	0.770	0.600	0.602

. The values are below the recommended threshold of 0.85, providing good discriminant validity among the constructs.

Estimating a series of regression equations determines the structural model coefficients for the relationships between the structures. Before evaluating the structural model, possible collinearity in the relationships should be checked to avoid bias in the regression models. The Variance Inflation Factor (VIF) values are recommended to be checked for assessing the degree of collinearity among the variables. VIF values greater than 5 are considered to indicate collinearity issues. In this study VIF for IM-2 (5.403), IU-2 (7.726), IU-3 (5.856) were found to be greater than 5, so IM-2 and IU 2 were removed from the model. After that, VIF for IU-3 was found to be 3454 less than 5, like all the other items of the model (

Table A4. Collinearity statistics (VIF).

	VIF
CMP-1	1.925
CMP-2	2.064
CMP-3	1.350
IM-1	2.124
IM-2	5.403
IM-3	4.392
IU-1	4.388
IU-2	7.726
IU-3	5.856
IU-4	1.547
PEU-1	1.630
PEU-2	1.773
PEU-3	1.476
PEU-4	1.208
PU-1	3.122
PU-2	3.044
PU-3	1.917
PU-4	1.600
SN-1	1.834
SN-2	1.834

).

3.2.2. Evaluation of the Structural Model

The controls recommended to consider by Hair et al. [42] in the evaluation of the structural part of the model are the coefficient of determination (R²), Q² (blindfolding-based cross-validated redundancy measure), statistical significance, and relevance of path measurements. The PLS Predict procedure is also recommended to determine the out-of-sample predictive power of the model.

The coefficient of determination (R²) values of the structures are checked. According to Shmueli and Koppius [50], the R² value indicates the model’s explanatory power [42]. It is also expressed as the in-sample predictive power. The R² takes values between 0 and 1. The appropriate value of R² may vary according to different study areas. In their study addressing the importance of effect size in quantitative studies, Sullivan and Feinn [51] defined the effect size of the coefficient of determination R² as 0.04 small, 0.25 medium, and 0.64 large [42]. From this perspective, the R² values of the proposed model shown on the construct in Figure 3 are satisfactory. The Predictive Accuracy (Q²) value is also one of the values that should be checked. This value is obtained by deleting a single datum in the data matrix, replacing it with another datum formed according to the data average, and estimating the parameters. Minor differences between the estimated and original values indicate a significant Q² value, i.e., higher estimation accuracy. Amounts of 0, 0.25, and 0.50 indicate small, medium, and large estimation accuracies, respectively. For the model, these values are 0.287, 0.294, and 0.470 for PEU, PU, and IU, respectively. The R² and Q² values of the constructs ensure that the model represents the empirical data and a satisfactory level of predictive capability.

The mean absolute error (MAE) values of indicators calculated by the software for the PLS-SEM and Linear Models are compared to assess the PLS-SEM model’s predictor power. Hair et al. [26] suggested that if all of the errors of the PLS-SEM are greater than the LM errors, the model has no predictive power. Most of them being greater shows that the predictive power is low. If minor or equal errors are greater than, the model has moderate power; while none are greater, the model has strong predictive power.

Table 6. PLSPredict summary.

	Q2predict	PLS-SEM_MAE	LM_MAE
IU-1	0.406	0.813	0.817
IU-3	0.409	0.808	0.795
IU-4	0.265	0.816	0.813
PEU-1	0.148	0.816	0.820
PEU-2	0.009	0.755	0.754
PEU-3	0.041	0.793	0.781
PEU-4	0.312	0.776	0.709
PU-1	0.192	0.827	0.832
PU-2	0.158	0.844	0.845
PU-3	0.291	0.823	0.808
PU-4	0.171	0.713	0.716

shows that out of eleven indicators, six have greater PLS SEM errors, indicating that the model has a moderate prediction power. According to Table 6, all of the Q²_predict values of indicators are greater than zero, providing the most naïve benchmark.

3.3. Hypotheses Testing Analysis

This study proposed twelve hypotheses, as shown in Figure 3. A bootstrapping process of Smart PLS was used by resampling 5000 samples to determine the significance of the relationships between constructs. Two of the twelve hypotheses were not supported due to p values greater than 0.05 (

Table 7. Structural Model Results.

Hypotheses	Relationships	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics (|O/STDEV|)	p Values	Decision
H1	PEU → PU	0.313	0.315	0.058	5.395	0.000	Supported
H2	PU → IU	0.384	0.386	0.044	8.817	0.000	Supported
H3	PEU → IU	0.137	0.138	0.040	3.400	0.001	Supported
H4	IM → PU	0.194	0.195	0.054	3.568	0.000	Supported
H5	IM → PEU	0.120	0.119	0.060	2.007	0.045	Supported
H6	IM → IU	0.099	0.097	0.047	2.120	0.034	Supported
H7	SN → PU	0.216	0.216	0.055	3.929	0.000	Supported
H8	SN → PEU	0.298	0.300	0.052	5.686	0.000	Supported
H9	SN → IU	0.331	0.329	0.049	6.693	0.000	Supported
H10	CMP → PU	0.030	0.030	0.051	0.585	0.558	Not supported
H11	CMP → PEU	0.240	0.240	0.050	4.824	0.000	Supported
H12	CMP → IU	0.022	0.022	0.040	0.534	0.594	Not supported

).

It has been confirmed that all hypotheses, except for H10 and H12, are valid. Hypotheses H10 and H12 are those suggesting that Compatibility affects Perceived Usefulness and Perceived Ease of Use. According to the results, Compatibility has a direct effect on Intention to Use, but its impact on Perceived Usefulness is indirect and occurs through its effect on Perceived Ease of Use, as demonstrated by the findings in Table 7 and

Table 8. Total Indirect Effects.

	Original Sample (O)	Sample Mean (M)	Standard Deviation (STDEV)	T Statistics (|O/STDEV|)	p Values
CMP → IU	0.073	0.074	0.025	2.895	0.004
CMP → PU	0.075	0.075	0.020	3.849	0.000
IM → IU	0.106	0.106	0.026	4.065	0.000
IM → PU	0.038	0.037	0.020	1.878	0.060
PEU → IU	0.120	0.121	0.025	4.758	0.000
SN → IU	0.160	0.161	0.030	5.356	0.000
SN → PU	0.093	0.095	0.025	3.664	0.000

.

Perceived Ease of Use, one of the core elements of Davis’ [31] Technology Acceptance Model (TAM), plays a significant role in both Perceived Usefulness and Intention to Use. The results show that Ease of Use has a direct effect on Usefulness, supporting the validity of hypotheses H1, H2, and H3. Additionally, the constructs of Image and Social Norm have been identified as influential on the three key components of the Technology Acceptance Model: Ease of Use, Perceived Usefulness, and Intention to Use. This is reflected in hypotheses H4, H5, H6, H7, H8, and H9. The final model, illustrating the relationships between these variables, is presented in Figure 5.

Based on the final model, Perceived Usefulness was identified as the most influential factor on Intention to Use, with a weight of 0.384. This was followed by Social Norm (0.331) and Perceived Ease of Use (0.137). Social Norm and Image were found to be the most impactful factors on Perceived Usefulness (PU). As shown in Table 6, Compatibility had no direct effect on either Intention to Use or PU. However, Image was the strongest factor affecting Perceived Ease of Use, with Compatibility also playing a significant role, holding a weight of 0.24, just behind Image.

4. Discussion

As electric vehicles (EVs) become increasingly popular, it is important to identify the factors influencing users’ choices and what they consider when deciding to use or purchase them. EVs, whose use is growing in China, the USA, and Europe, stand out due to their cost advantages and innovations in terms of usage. Beyond the environmental and economic benefits, EVs are notable for the innovations they bring. While it is clear that EVs are less harmful to the environment compared with traditional fuel-powered vehicles, their disadvantages are also debated alongside their advantages [18]. However, significant investments spurred by almost all car brands producing electric models and the emergence of new brands specifically for this purpose overshadow the controversial aspects of EVs.

In recent years, innovations such as autonomous driving and shared vehicles in the transportation sector have attracted as much attention from users as EVs [52,53,54,55]. In Türkiye, autonomous driving or shared applications are not yet widely used. A society that is open to trying innovations and is curious about new experiences is one that is eager to experience such innovations happening around it. In this sense, examining the factors considered and providing accurate and rational guidance is critical for society and the environment. As mentioned, introducing electric vehicles as an innovation is enough to draw interest. However, understanding the motivations behind society’s approach to new technology will ensure that steps related to this issue (such as investments in charging stations or factory establishments) are made within a well-planned framework.

Although common factors influence how societies respond to innovations, the importance and weight of these factors can vary from one society to another [3,11,14,20,22]. How a technology or innovation is adopted can differ based on a country’s social, economic, and cultural characteristics. Therefore, it is essential to determine the approaches of societies with different dynamics. According to the results, the effect of Perceived Ease of Use on Perceived Usefulness and their effect on Intention to Use are proved again. Subjective Norms influenced by a person’s social environment were found to affect Perceived Usefulness, Perceived Ease of Use, and Usage Intention. Another factor influencing Perceived Usefulness was identified as Image. Although Compatibility, another factor examined, does not directly affect Perceived Usefulness or Usage Intention, it was found to have an indirect effect on intention. After Image, Compatibility was the second most influential factor in Perceived Ease of Use.

Cultural differences significantly impact how EVs and innovations are perceived. When we look at studies on Subjective Norms (SN), we see different results across different countries. Notably, studies from China [11,14] show that SN either had no effect or had an indirect effect, while European studies show that SN was influential [4,12,16,22,24]. In the study by Adu-Gyamfi et al. [55], it was found that SN had an effect on PU, PEU, and UI, and emerged as the most influential factor on UI, which is consistent with the results of our study. Venkatesh and Davis [40] stated in their study that Subjective Norms have little effect on intention when usage is voluntary. However, despite EV usage being voluntary in our study, Subjective Norms emerged as an influential factor on intention, unsurprising for Türkiye, a country where social interaction and pressure are intense. It is suggested that SN be studied based on the characteristics of societies for future research. While it is a limitation of our study that not enough interpretation can be made with the current results, as Venkatesh and Davis [40] noted, the influence of SN decreases as experience with innovations increases. This finding of Venkatesh and Davis [40] aligns with the fact that SN has no effect or only indirect effects in a country like China, where EVs are widely used, while it is effective in Türkiye, where EVs are newly spreading.

Another influential factor in EV adoption is the perception of EVs as symbols of prestige and status in society, aligning with findings by Khurana et al. [56]. Like Venkatesh and Davis [40], Khurana et al. [56] involved Image as an influential factor in their studies. According to their results, Image influenced usefulness, ease of use, and intention. One of the findings of our study is that electric vehicles are perceived as a symbol of prestige and status for individuals in society, which is similar to Khurana et al.‘s [56] study regarding the effects on the key factors of TAM.

Although Compatibility does not directly influence PU or UI, it emerged as the second most influential factor on PEU after Subjective Norms. The most challenging features of EVs are the limited range of the cars and the charging procedure. When we examine EVs regarding Compatibility with the user’s previous habits, we see that low scores were given for the relevant variables. The average scores for CMP1, CMP2, and CMP3 were 2.53, 2.56, and 2.41 out of 5, respectively, which are the lowest values compared with the averages of other variables (Table 5). Although Compatibility does not directly affect intention, another result of our study is that this factor indirectly impacts Ease of Use and Intention. Developments such as shorter charging times and extended range, which make daily use more convenient, will positively affect PEU and indirectly affect Intention to Use.

When examining the studies in the literature (Table 1), TAM and its subsequent versions, together with SEM, emerge as frequently used methods in the context of electric vehicle adoption. According to the results of these studies, key concepts such as permitted priorities, environmental concerns, trust, and social influence have been highlighted.

This study offers a different perspective by analyzing the effects of Compatibility, Image, and Social Norms. In Tu and Yang’s study, the impact of the concept of Compatibility on behavior-oriented attitudes was examined. In this context, exploring the effects of Compatibility on Perceived Ease of Use, Perceived Usefulness, and Intention to Use is one of the distinctive aspects of this study. Similarly, another concept that is rarely encountered in TAM and electric vehicle adoption studies is Image. The effects of the concept of Image included in the research by Khurana [56] on attitude and intention have been examined. The presented study is also significant in terms of investigating the effects of the Image concept on PEU (Perceived Ease of Use), PU (Perceived Usefulness), and UI (Intention to Use).

One of the limitations of our study is that, since EV usage is not yet widespread, characteristics such as participants’ gender, age, and education level could not be thoroughly considered during the sampling process, as participants are required to have knowledge about EVs. So, the results obtained cannot be generalized to Türkiye due to the sample size and characteristics. Another limitation is the inability to establish a connection with participants’ actual purchasing behavior. It is recommended that future studies include a more detailed analysis of demographic factors and actual purchasing actions.

5. Conclusions

This study aims to identify the factors shaping the increasing use of electric vehicles in Türkiye. After determining that the data collected to test the hypotheses in the study are reliable and valid, a PLS-SEM (Partial Least Squares Structural Equation Modeling) was developed to identify the factors influencing the usage intention of EV users in Türkiye. The results of the reliability and validity tests for the model indicate that a reliable and valid model has been established. This study examines the role of society, the importance of individuals’ positions within society, the differences brought about by new technologies, and their impact on the intention to use EVs. According to the results, the most critical point that decision-makers should pay attention to is that the most significant source of information for users is other users. At this point, it is essential to determine people’s expectations and perceptions about electric vehicles. As Heffner et al. [6] stated, the meaning each household attributes to electric vehicles may be different. People should be informed about whether they are aware of having shared expectations when influenced by their social environment.

Perceived Ease of Use is the most effective factor on Perceived Usefulness. Looking at the measurement values, the variable with the lowest satisfaction in terms of perceived ease of use is the question “I can easily charge my car anywhere“ which is the only question regarding the charging location among the ease-of-use questions. Since satisfaction with the charging process was also high, it is thought that the issue people are having trouble with is not the charging process but the charging station network. In addition, low satisfaction ratings regarding Compatibility and the impact of Compatibility on perceived ease of use also reflect user concerns regarding charging and range. As a step towards solving this problem, expanding the charging station network and focusing on technologies that will improve charging processes are among the recommendations obtained as a result of this study.

EVs offer significant environmental benefits, producing fewer emissions than traditional vehicles. However, challenges remain, such as the environmental impact of battery production and technological limitations like limited range and long charging times. Ongoing innovations, such as faster charging and longer battery life, are expected to improve user experience and adoption rates.

A significant criticism and concern surrounding electric vehicles is that the resources allocated for battery production may contradict the environmental benefits these vehicles aim to provide. Moreover, the adoption of less expensive compounds is anticipated to lead to reduced battery and vehicle costs. Furthermore, considering the observed decline in the impact of social norms as EV usage grows, it is recommended that similar studies be conducted to identify the factors influencing the intention to use electric vehicles, analyzing both familiar and novel external effects.