Attitudinal Factors Associated with the Use of Bicycles and Electric Scooters

Abstract

The use of personal mobility vehicles (PMVs) has increased, given the consumption characteristics of the population as well as the impact on the environment that other means of transport generate. In this work, the attitudinal factors associated with the use of PMVs are identified, making use of the theory of planned behaviour and a modified technological acceptance model. For the collection of information, a sample of 457 individuals was used. For the analysis of this information, a structural equation model was generated through SmartPLS 4. The results reveal that of the attitudinal factors associated with the use of these vehicles, green attitudes, perceived green value and loyalty stand out, with the latter being one of the latent predictive variables of the model, which is why feelings of enjoyment, usefulness generated and a perception of caring for the environment by PMV users have a greater influence on their loyalty to this type of green technology.

1. Introduction

The increase in traffic in cities has generated environmental problems, such as air pollution [1] and noise pollution [2] in urban environments, causing detrimental effects on people’s quality of life. Additionally, technologies have emerged that use alternative fuels and can counteract such effects [3,4,5]. These technologies seek to reduce the use of cars in urban areas; cars are usually the cause of traffic problems and are a technology that part of the population (not all) can substitute with public transport or nonconventional means of transport [6].

Jansson et al. [3], in their study of sustainable consumption, report that combustion “represents 80% of the anthropogenic emissions of greenhouse gases, such as carbon dioxide (CO₂)” (p. 176). For this reason, in recent years, alternative means of energy have been considered for replacing fossil fuels, and technologies such as electric vehicles (EVs) have been adopted [7,8]. To cover short distances, new modes of mobility, such as EVs, have emerged, among which are electric scooters and electric bicycles, i.e., personal mobility vehicles (PMVs) [9,10].

As described by Hasegawa et al. [9] in their study on PMVs, these vehicles “provide economic, social and environmental benefits, such as reducing congestion in urban centres, reducing harmful emissions and providing a means of transportation for people with reduced mobility” (p. 256). In addition, Ando and Li [11] noted that PMVs had been the subject of different studies that consider them important in the “revitalization of mobility in city centres and as solutions for global warming and energy shortages” (p. 164). Therefore, the acceptance of this technology is significantly influenced by the perception of being a useful tool to reduce urban traffic as a means of environmental protection [12,13].

Regarding technological change, authors such as Moons and De Pelsmacker [14] have noted that a consumer’s thinking is fundamental because it affects the expected results and promotion of the implementation of environmentally friendly technologies. Other authors have analysed bicycle sharing [15], innovative electricity alternatives [4] and energy policy measures [16] related to the adoption of EVs as PMVs.

The gap on which this study is based is related to the scarcity of empirical research on the different aspects related to the use of LMPs, specifically in developing countries. As stated by Dibaj et al. [17], in-depth studies are required to address the attitudinal elements related to using LDCs. In addition, particularly for the city of Medellín and the Metropolitan Area of Aburrá Valley (Área Metropolitana del Valle de Aburrá—AMVA), Colombia, there are no studies that relate to these factors. In this sense, the study aims to identify the factors associated with using PMVs in Medellín and the AMVA.

The contribution of this study is to identify the main factors that encourage the use of VMPs in the AMVA (green attitudes, perceived green value and loyalty). In this sense, on the one hand, this study is a crucial source of information for local authorities in the elaboration of public policies specifically designed when taking into account the three factors identified in such a way as to encourage the massification of the use of VMPs in both devices and users. On the other hand, this study contributes to the literature by using a model of integration between attitudinal factors and technological acceptance that, through the incorporation of different constructs, makes it possible to contemplate different scenarios of behaviour.

2. Reference Framework

The expansion of the vehicle fleet and its direct effects on air quality have led to the creation of new means of transport specifically designed for short distances, for example, PMVs [18]. Hasegawa et al. [9] and Ulrich [19] describe PMVs as vehicles that utilise an electric motor to assist human beings in their personal travel on short-distance trips. PMVs currently function as auxiliary vehicles for daily use, offering users lower fuelling and maintenance costs and shorter travel times, allowing, in turn, a reduction in the use of cars in congested urban environments [6]. To learn about the use of the new micromobility technologies, some authors have studied the associated factors that consider the attitudes of users after a user experience [20,21] as well as public opinion, presenting the real needs of users [12].

Although the benefits of these vehicles with regard to reducing greenhouse gases have already been reported, several studies relate consumer behaviour in favour of the environment with attitudinal factors that influence the adoption of EVs [3,5] because these, in turn, are associated with the behaviour of environmental consumers [7]. Sánchez Castañeda [22], in his study of pro-environmental or green behaviours, notes that there are three blocks of variables related to environmental behaviours: I. sociodemographic and cognitive (knowledge about the environment); II. environmental intervention (people’s knowledge about environmental issues and the skills to carry out environmental actions); and III. psychosocial (attitudes, beliefs and values that predict behaviour).

Sociodemographic variables have not played a leading role in the studies related to adoption; however, they have been related ambiguously to consumer innovation [5]. Knowledge about environmental problems and interventions has been studied and related to positive attitudes towards the environment [23], which can be studied with psychosocial variables to identify the factors related to environmental awareness and green behaviour.

2.1. Attitudinal Factors

The psychosocial model includes a set of values, beliefs and norms that compose attitudinal factors, together with green attitudes [5]. Beliefs, values and social norms have been shown to be relevant in understanding the adoption of alternative fuel vehicles [3,7] because although consumer behaviour determines the use or disposal of a product, behaviour can also be influenced by environmental awareness [5] and green behaviour that induces the purchase and consumption of goods that generate a positive environmental impact [22].

Usually, studies of attitudinal factors for environmental consumer behaviour are based on the importance of green attitudes, values [24], subjective norms and personal norms [25], proving to be important in the adoption of EVs [3,4,7,14,16,24].

2.1.1. Green Attitudes

Green attitudes are considered a psychological tendency expressed when evaluating intentions or pro-environmental behaviours predictive of consumer behaviour [3]. They evaluate the personal benefits and associated costs as an expected result of the use of renewable energy and shared use, where attitudes towards behaviour are associated with performance [24]. Similarly, green attitudes are influenced by knowledge and experience [26].

Attitude is a theoretical construct proposed in the theory of planned behaviour (TPB), which the authors defined as a positive or negative attitude of the individual to perform a behaviour [27]. The evaluation that an individual makes regarding the adoption of a technology is known as acting; hence, it comes to be considered as the users’ perception towards the adoption of VMP [28]. Attitude–value–behaviour relationships have been established in the literature, where attitude acts as the most important predictor of behaviour by incorporating a judgement, whether a behaviour is considered bad, good or indifferent [29].

Some authors, such as Jansson et al. [3] and Petschnig et al. [4], relate green attitudes to behavioural intentions and, consequently, with a positive attitude of consumers towards the adoption of EVs, with the shared use model being a similar case [24]. In TPB, attitude relates to intention, subjective norms, and perceived behavioural control in the context of green purchasing [30]. In the technological acceptance model (TAM), an attitude has a relationship with perceived usefulness [31] and, in more recent theories, with perceived enjoyment [32].

2.1.2. Personal Beliefs

Personal beliefs are shared beliefs of the perception of individuals about how to act. Internally, personal beliefs define whether a behaviour deserves a reward or punishment [3,24]. Altruistic behaviour, in turn, is conceptualised as a personal norm for which an individual is aware of the consequences of an act, feelings of self-sufficiency [33] and the possibility of acting without generating the perceived consequences that warrant a punishment [7].

Other authors argue that personal beliefs are associated with the attitude of individuals in a specific situation; that is, if an individual believes that an object or a situation is “good”, the evaluation carried out will influence a favourable perception [22]. In conjunction with subjective norms, personal beliefs act as predictors of the intention of pro-environmental behaviour [23] and the adoption of environmentally friendly technologies [5].

2.1.3. Perceived Green Value

In the extended decomposed theory of planned behaviour by Moons and De Pelsmacker [14], the authors define values as beliefs that lead to desirable behaviours or end states, where a specific situation guides the evaluations of behaviour. The predictors are actions that display concern for the environment and the will to participate in environmental actions [26]. Studies, such as that by Sánchez Castañeda [22], explore the values closely linked to pro-environmental behaviour; values can explain, motivate, and justify pro-environmental behaviour through beliefs [24].

These values are influential in driving individuals to implement or even change their attitudes in favour of the environment [14], developing conscious attitudes through values to express and act on them [3] from feelings of moral obligation [34]. Consumers, when they value environmental protection, are likely to buy products that are respectful of the environment [22].

2.1.4. Norms

Norms define the normative evaluation of a particular behaviour, and a relationship exists between norms and behavioural intention [23]. Studies agree that norms are important in influencing the adoption of environmentally friendly technologies, such as EVs [3,4,25]. Norms are divided into two types: social norms and personal norms.

Personal beliefs are affected by social norms—also called subjective norms—which are essential in motivating an individual to adopt a particular behaviour and influencing changes in others [25]. Thus, consumers want to comply with the perceived preferences of individual referents, as outlined by Petschnig et al. [4] in their alternative fuel adoption study. Subjective norms are less internalised; they are based on the perceived expectations of rewards or punishments by society and the pressure the perceptions exert on an individual [3]. In general, the pressure stems from the people who the individual considers important and influential in his or her life, such as family and friends, influencing the individual to adopt a new means of transport [7]. Other authors mention that social norms are predictors of environmental behavioural intentions [23] and, in turn, are fundamental in behavioural intention [26] and changes in behaviour in favour of the environment [25].

Personal norms allow us to understand and deal with issues related to climate change [7]. They are internalised in each individual, differentiate what is correct from what is incorrect and are closely related to values [3]. Individuals accept or reject environmental care behaviours [23]. These norms are derived from their own expectations, concerns and moral obligations and influence decision-making [25]. The environmental actions taken by pro-environmental consumers are driven by personal norms and morals and are closely related to personal beliefs about the environmental conditions and actions that each individual can employ, perform and correct in favour of the environment, such as the adoption of technologies, with a high level of sustainability [4]. The aforementioned attitudinal factors are derived from various theories about human behaviour, as explained below.

2.1.5. Green Loyalty

Loyalty to environmentally friendly technologies is achieved when a repurchase process exists [35]. According to Chen’s definition [36], it refers to “the level of repurchase intentions prompted by a strong environmental attitude and sustainable commitment towards an object, such as a product, a service, a company, a brand, a group, or so on” (p. 298). Previous studies have analysed the influence of loyalty on EV adoption in light of the advantages it offers for sustainable mobility [37].

Recent studies have also associated loyalty with the continued use of shared electric scooters [38]. Similar studies were conducted as a function of electric bicycles. Liu et al. [39] found that a person’s loyalty has been influenced by factors such as perceived green value and perceived pleasure. Studies of this type of help to better understand the factors influencing the use of shared bicycles in addition to helping policymakers formulate effective demand-side targeting strategies [40].

2.2. Technological Acceptance Model—TAM

The TAM, in the words of Chen and Chao [41], is a theory that shows a high predictive level regarding the use of technology; therefore, it is relevant for evaluating the factors associated with the use of PMVs, taking into account constructions, such as perceived green values and consumer loyalty to clean transportation technologies. This is one of the most commonly used theories to examine the acceptance of technology and was proposed by Davis [31].

While various methods have been used in the literature to examine technology acceptance, such as the diffusion of innovations model (DOI), the use and gratification model (UG), or the expectation–confirmation model (ECM), the TAM remains relevant and has been adapted and extended to address new technologies and diverse contexts [42]. The characteristics of this model, which make it suitable for use in this study, are related to the focus on the user and their perceived usefulness and ease of use as relevant predictors of the intention to use technology. In addition, it is based on psychological and behavioural theories, including the theory of reasoned action (TRA), which provides it with theoretical soundness [43].

Additionally, this model has been widely validated and used in different contexts and cultures, which makes it reliable and valid for use in various studies [44]. The configuration and composition of this model make it an easy, clear and accessible alternative to analyse the acceptance of technologies [45].

To evaluate these factors, the TAM considers another construction that affects the intention of adopting innovations, i.e., the consumers’ perceived usefulness, which affects the attitude toward the use of technologies such as PMVs [6]. Subjective norms and perceived control also influence the intention of adopting new modes of transport models for short distances, and perceived control is affected by values [15].

2.3. Theory of Planned Behaviour—TPB

Previous studies have evaluated the use of PMVs and have demonstrated the significance of the sociopsychological factors in their acceptance [11]. The model proposed by Ajzen [27] has been used extensively to predict and explain predicted behaviour in a variety of disciplines. The shared use of PMVs is also associated with the value-belief-norm theory [24]. The characteristics of consumers who engage in the early adoption of innovations and disseminate their opinion on their own initiative have also been studied [3].

Different models and theories have been used to examine people’s behaviour. Among them, the TRA describes how attitudes and subjective norms influence behavioural intention [46], and the expectancy-value theory (EVT) describes how expectations and values influence behavioural intention [47]. However, the TPB represents an alternative with wide acceptance in the literature because it is supported by solid theoretical elements that make it a suitable option for examining people’s behaviour in different contexts.

In particular, this theory focuses on behavioural intention as the main predictor of actual future behaviour [48]. It corresponds to an integrative and flexible theory, which can be adapted to different contexts and situations, demonstrating a high predictive capacity for people’s behaviour [49], and finally, this theory can serve as a guide for intervention since it identifies the factors that can be modified to influence people’s intention and behaviour [50].

The TPB has been employed by various authors [3,4,7,11,14,16,23,24,25]. The original theory was formulated by Chen [15] and is one of the theories most used to evaluate the factors associated with the use of technologies; however, in this study, it is necessary to employ a modified TAM model to perform more thorough research, taking into account the eight concepts derived from studies carried out by Chen [15], and Shao and Liang [51], and adapted to the objectives of this study.

For a long time, adoption models and theories have been adapted according to context. From the TPB, this study takes the variables attitude, subjective norm and perceived behavioural control, to which the variable innovation is added. From the TAM, the variables taken are as follows: perceived usefulness, and the external variables, perceived pleasure, loyalty, and perceived green value (see Figure 1).

2.4. Hypothesis Development

A perceived green value, as mentioned by Chen [15], is usually accompanied by functional, economic, emotional and social values and, in turn, has a significant impact on consumer loyalty and perceived control. Perceived usefulness, in the case of PMVs, addresses the benefits provided to the environment through their adoption. Perceived enjoyment, in the context of this study, refers to the user’s experience when using PMVs and, thus, is related to the perceived value, leading to the following hypothesis:

H1. 

Perceived green value is positively associated with perceived enjoyment of use, perceived usefulness, subjective norms, perceived behavioural control and loyalty to PMV use.

Shao and Liang [51] note that norms drive effective PMV sharing through behaviours influenced by the social ethics that are related to intentions to adopt. Loyalty, in the scope of this study, refers to the intention of individuals to continue the shared use of PMVs, depending largely on their enjoyment, usefulness, degree of innovation, control and perception offered to the user [15]. As such, the following hypothesis is proposed:

H2. 

Perceived enjoyment of use, perceived usefulness, subjective norms and perceived behavioural control are positively associated with loyalty to PMV use.

Green attitudes are influenced by norms and values, as established in the TPB, which can be used to interpret human behaviour and predict the future actions of individuals in relation to environmental awareness [3]. In a study of perceived value using the TPB and a modified TAM, Chen [15] relates green attitudes to subjective norms, perceived control and loyalty in behavioural intentions. As such, the following hypothesis regarding green attitudes is proposed:

H3. 

The general attitude towards caring for the environment can moderate loyalty relationships through the perceived green value, perceived usefulness, perceived enjoyment of use, subjective norms and perceived behavioural control for the use of PMVs.

The results of some studies indicate that the adoption or shared use of PMVs not only focuses on green attitudes but also on the degree of innovation of these technologies [3] because some individuals have a greater disposition to purchase and accept new products in the market with innovative characteristics. As such, the following hypothesis is proposed:

H4. 

Consumer innovation is positively associated with the general attitude toward environmental concerns.

3. Materials and Methods

This study was carried out with a quantitative approach to assess the relationships among variables. For this, referencing the studies by Jansson et al. [3], Chen [15] and Shao and Liang [51], a questionnaire was developed to assess the use of PMVs in the city of Medellín and the metropolitan area. The questionnaire was reviewed and approved by 10 experts in the research field, who identified and resolved possible problems with the questionnaire. Subsequently, 20 university students were contacted to participate in a clarity test, the results of which indicated full comprehension; thus, it was concluded that the questionnaire had high content validity.

Once the preliminary tests were carried out, the formal study was conducted at the study site, obtaining a sample of 457 people to test the hypotheses established in the theoretical model. In addition to completing the questionnaire, the respondents were asked to provide sociodemographic information, including age, gender, education level, socioeconomic stratum, occupation, municipality and commune of residence (

Table 1. Sociodemographic information of the sample.

Demographic Category	Factor	Percentage
Gender	Male	44.4%
	Female	55.6%
Occupation	Study	15%
	Works	64%
	Study and work	21%
Age	Under 20 years	25.40%
	Between 21 and 30 years	55.10%
	Between 31 and 40 years	12.30%
	Between 41 and 50 years	5%
	Between 51 and 63 years	2.20%
Socioeconomic stratum	Low-low	8.80%
	Low	31.60%
	Medium-low	44.40%
	Medium	10.90%
	Medium-high	4.3%
Municipality	Barbosa	1.1%
	Bello	8.1%
	Caldas	2%
	Copacabana	2%
	Envigado	6.1%
	Girardota	1.5%
	Itagüí	3.7%
	La estrella	2.2%
	Medellín	68.7%
	Sabaneta	4.6%
Most used modes of transportation	Bus	28%
	Metro	23%
	Metroplús (articulated bus)	4.6%
	Private car	11.6%
	Motorcycle	23.4%
	Bicycle	0.7%
	Taxi/Uber	1.3%
	Walk	6.6%
	Other	0.9%
Approximate time most frequent destination	Between 5 and 30 min	49.7%
	Between 31 and 45 min	18.2%
	Between 46 and 60 min	22.3%
	Between 61 and 90 min	8.5%
	Between 91 and 180 min	1.3%
Do you use multiple modes of transportation?	Yes	39.8%
	No	60.2%
What are these modes of transportation?	Bus and bicycle	1.7%
	Bus and bus	12.2%
	Bus and walking	3.3%
	Bus and taxi/Uber	3.3%
	Metro and bicycle	2.2%
	Metro and bus	51.1%
	Integrated metro system	14.4%
	Metroplús (articulated bus) and bus	7.2%
	Other	4.5%
Would you be willing to trade in your traditional short-distance vehicle for a VMP?	Yes	69.4%
	No	30.6%
How much would you be willing to pay for a VMP-sharing service?	Less than the established value	43.8%
	The established value	22.8%
	More than the established value	5.7%
	Not willing to pay	27.8%

).

For the factor analysis, a structural equation model (SEM) was used. A five-point Likert scale was used: 1, completely disagree; 2, disagree; 3, neither agree nor disagree; 4, agree; and 5, completely agree. For the analysis and calculation of the correlation statistics, the SmartPLS 3 program was used.

4. Results

Regarding the variables obtained from the TPB and the modified TAM, the empirical measurement analysis of the data was carried out using a partial least squares SEM (PLS-SEM) because it has been shown to be effective in previous studies of purchase intention [52], behaviours towards the sustainable use of shared electric bicycles [53] and loyalty to the shared use of public bicycles [15]. Two analyses were carried out, i.e., an analysis of the measurement model to ensure the reliability and validity of the construct and an analysis of the structural model to evaluate the hypotheses, using SmartPLS 4 software. This software is used because it is one of the best statistical calculation tools, offering advantages in terms of interface, ease of use and analytical and graphical capabilities in the techniques based on components or partial least squares [54].

4.1. SEM

Ávila and Moreno [55] and Ruiz et al. [56] define SEMs as multivariate statistical models that allow for estimating the effect and relationships between multiple variables, giving a higher level of confidence to research results and revolutionising empirical research. As stated in previous studies, the reliability should be evaluated based on Cronbach’s alpha (CA) and composite reliability (CR) to obtain a reflective measurement model; likewise, convergent validity should be calculated to evaluate the reliability of the indicator and the average variance extracted (AVE). Regarding the discriminant validity, the calculation is given by evaluating the square root of the AVE (looking for each construct to be higher than the squared correlation) and the factor loads (FLs) [54,55].

4.1.1. Measurement Model Analysis

Henseler et al. [57] and Hulland [58] suggest that CA and CR should be equal to or greater than 0.7. Hair et al. [59] and Ávila and Moreno [55] suggest FL values greater than 0.7 and an AVE greater than 0.5 are required for a measurement model to be accepted. To guarantee this, the moderating effects of green attitudes and their relationship with loyalty were excluded. The results are reported in

Table 2. Indicator reliability and composite reliability.

	Mean	SD	FL	CA	CR	AVE
Perceived Green Attitudes				0.891	0.925	0.755
GA1	0.817	0.034	0.818
GA2	0.895	0.018	0.895
GA3	0.879	0.023	0.879
GA4	0.881	0.021	0.882
Perceived Control				0.784	0.872	0.696
PC1	0.756	0.030	0.752
PC2	0.881	0.010	0.881
PC3	0.863	0.018	0.863
Consumer Innovation				0.711	0.835	0.630
CI1	0.825	0.024	0.826
CI2	0.663	0.039	0.664
CI3	0.877	0.012	0.876
Loyalty				0.819	0.893	0.736
LH1	0.875	0.016	0.874
LH2	0.795	0.026	0.796
LH3	0.900	0.012	0.900
Subjective Norm				0.741	0.885	0.794
SN1	0.897	0.012	0.897
SN2	0.885	0.014	0.885
Perceived Enjoyment				0.702	0.830	0.620
PE1	0.833	0.015	0.832
PE2	0.807	0.020	0.807
PE3	0.720	0.035	0.719
Perceived Usefulness				0.700	0.833	0.626
PU1	0.781	0.030	0.780
PU2	0.730	0.036	0.730
PU3	0.859	0.016	0.857
Perceived Green Value				0.850	0.899	0.690
PGV1	0.816	0.024	0.817
PGV2	0.848	0.017	0.847
PGV3	0.797	0.032	0.795
PGV4	0.864	0.014	0.863

Note: CA > 0.7, CR > 0.7 and AVE > 0.5. The data in the table were obtained from Ringle et al. [60], http://www.smartpls.com (accessed on 10 March 2023).

.

The Fornell–Larcker criterion, according to Ávila and Moreno [55], considers that the square root of the AVE of each latent variable should be greater than the correlations it has with the rest of the variables [59]. The results are provided in

Table 3. Fornell–Larcker criterion for discriminant validity.

	Green Attitudes	Perceived Control	Consumer Innovation	Loyalty	Subjective Norm	Perceived Enjoyment	Perceived Usefulness	Perceived Green Value
GA	(0.869)
PC	0.392	(0.834)
CI	0.729	0.575	(0.794)
LH	0.685	0.664	0.724	(0.858)
SN	0.412	0.680	0.587	0.633	(0.891)
PE	0.535	0.710	0.712	0.749	0.713	(0.788)
PU	0.587	0.559	0.625	0.708	0.550	0.660	(0.791)
PGV	0.703	0.641	0.770	0.873	0.663	0.738	0.733	(0.831)

Note: The square root of the AVE value is displayed in parentheses. The data in the table were obtained from Ringle et al. [60]. http://www.smartpls.com (accessed on 10 March 2023).

.

Conesa et al. [61] also mention that the factor loadings should have a higher value with their own variables than with the others that are also evaluated in the model. This is evidenced in

Table 4. Cross-factor loads.

	Green Attitudes	Perceived Control	Consumer Innovation	Loyalty	Subjective Norm	Perceived Enjoyment	Perceived Usefulness	Perceived Green Value
GA1	0.818	0.283	0.560	0.541	0.267	0.373	0.452	0.546
GA2	0.895	0.392	0.656	0.611	0.393	0.478	0.505	0.626
GA3	0.879	0.321	0.658	0.623	0.374	0.503	0.567	0.633
GA4	0.882	0.362	0.652	0.602	0.389	0.496	0.509	0.632
PC1	0.213	0.752	0.355	0.401	0.489	0.479	0.366	0.408
PC2	0.386	0.881	0.546	0.662	0.619	0.668	0.559	0.627
PC3	0.353	0.863	0.508	0.554	0.580	0.603	0.445	0.536
CI1	0.551	0.415	0.826	0.551	0.454	0.569	0.543	0.635
CI2	0.394	0.566	0.664	0.616	0.512	0.635	0.492	0.580
CI3	0.726	0.449	0.876	0.596	0.471	0.551	0.486	0.638
LH1	0.608	0.656	0.670	0.874	0.608	0.674	0.630	0.774
LH2	0.567	0.459	0.543	0.796	0.449	0.575	0.612	0.689
LH3	0.589	0.582	0.643	0.900	0.561	0.673	0.583	0.779
SN1	0.428	0.552	0.559	0.576	0.897	0.641	0.512	0.609
SN2	0.303	0.663	0.485	0.551	0.885	0.629	0.466	0.572
PE1	0.621	0.529	0.689	0.717	0.552	0.832	0.571	0.710
PE2	0.289	0.655	0.524	0.555	0.639	0.807	0.479	0.551
PE3	0.279	0.507	0.419	0.447	0.498	0.719	0.509	0.431
PU1	0.436	0.425	0.445	0.513	0.376	0.472	0.780	0.565
PU2	0.373	0.465	0.470	0.530	0.464	0.550	0.730	0.511
PU3	0.566	0.444	0.560	0.630	0.465	0.548	0.857	0.654
PGV1	0.585	0.486	0.539	0.728	0.485	0.571	0.591	0.817
PGV2	0.541	0.591	0.655	0.727	0.609	0.657	0.656	0.847
PGV3	0.603	0.455	0.665	0.634	0.519	0.555	0.600	0.795
PGV4	0.613	0.586	0.698	0.803	0.582	0.660	0.589	0.863

Note: Factor loads must have a higher value with their own variable than with the others that are evaluated in the model. The data in the table were obtained from Ringle et al. [60], http://www.smartpls.com (accessed on 10 March 2023).

.

4.1.2. Structure Model Analysis

To evaluate the relationship between the constructs, the first step was to evaluate the model by verifying the collinearity between indicators [53]. For this evaluation, the variance inflation factor (VIF) test is used; ideal values are greater than 0.5, with a tolerance level of less than 0.2. However, when the VIF is greater than 1, multicollinearity is present [56]. In this study, the data demonstrate multicollinearity because the VIF values vary between 1.254 and 2.796.

The second step involves the evaluation of the relationships between the constructs, determining the value of the level of significance (R²), with values greater than 0.19 indicating a weak explanatory relationship, greater than 0.33 indicating a moderate relationship and greater than 0.67 indicating a substantial relationship. The cross-valid redundancy (Q²) must exceed the value 0 for the model to have predictive relevance [61].

Values of 0.02, 0.15 and 0.35 indicate little relevance, moderate relevance and substantial relevance [58]. In the model of factors associated with the use of PMVs, the R² values for perceived control and the subjective norms are 0.412 and 0.440, respectively, and the values for green attitudes, perceived usefulness and perceived enjoyment are 0.531, 0.537 and 0.544, respectively, indicating a moderate explanatory capacity; the R² value for loyalty is 0.804, indicating substantial relevance, meaning that the proposed model has relevant predictive precision [53]. The Q² values for perceived control, perceived enjoyment and perceived usefulness are 0.316, 0.324 and 0.330, respectively, demonstrating moderate predictive relevance; the values for the subjective norms and green attitudes are 0.387 and 0.390, and that for loyalty is 0.482, indicating substantial predictive relevance [55].

For testing the hypotheses, the path coefficient must exceed 0.005, the T statistic value must be greater than 1.96, and the p-value must be less than 0.05 for the significant relationship to be considered [53,55]. The results can be seen in

Table 5. Path analysis and significance testing.

	Path Coefficients	T Statistics	p Value	Acceptation
Perceived Green Value → Perceived Enjoyment	0.713	15.696	0.000	Yes
Perceived Green Value → Perceived Usefulness	0.633	12.947	0.000	Yes
Perceived Green Value → Subjective Norm	0.737	16.821	0.000	Yes
Perceived Green Value → Perceived Control	0.722	15.954	0.000	Yes
Perceived Green Value → Loyalty	0.634	12.845	0.000	Yes
Perceived Enjoyment → Loyalty	0.165	3.153	0.002	Yes
Perceived Usefulness → Loyalty	0.089	2.296	0.022	Yes
Subjective Norm → Loyalty	−0.029	0.706	0.480	No
Perceived Control → Loyalty	0.110	2.339	0.019	Yes
Perceived Green Attitudes → Perceived Enjoyment	0.035	0.685	0.493	No
Perceived Green Attitudes → Perceived Control	−0.115	2.329	0.020	Yes
Perceived Green Attitudes → Perceived Green Value	0.704	18.835	0.000	Yes
Perceived Green Attitudes → Perceived Usefulness	0.142	3.000	0.003	Yes
Perceived Green Attitudes → Subjective Norm	−0.106	2.269	0.023	Yes
Consumer Innovation → Perceived Green Attitudes	0.729	23.222	0.000	Yes

Note: Path > 0.005; T statistics > 1.96 and p valor < 0.05. The data in the table were obtained from Ringle et al. [60], http://www.smartpls.com (accessed on 10 March 2023).

.

The analysis indicates model reliability because most of the constructs have Cronbach’s alpha values greater than 0.7, except for perceived usefulness, which has a value very close to 0.7 and is thus considered questionable [51]. This can be explained, in part, by the geographic conditions of the city of Medellín and the metropolitan area because the mountainous environment possibly influences the perceived usefulness of PMVs.

The analysis of the structural model indicates multicollinearity; in addition, the R² value indicates a good fit for the model. Regarding the factors associated with using PMVs, all R² values are greater than 0.19, demonstrating moderate explanatory capacity for perceived control, subjective norms, green attitudes, perceived usefulness and perceived enjoyment, and a substantial explanatory capacity for loyalty. Therefore, the structural model has explanatory validity. The Q² values for valid cross-redundancy exceed 0.15 for perceived control, perceived enjoyment and usefulness, presenting medium relevance, and subjective norms, green attitudes and loyalty, presenting great relevance. This indicates that the structural model has predictive relevance.

5. Discussion

The results obtained show that perceived green value positively influences perceived enjoyment, perceived usefulness, subjective norms, perceived control and loyalty, fully supporting H1. Loyalty is positively influenced by perceived enjoyment and perceived control; however, perceived usefulness and subjective norms do not have the same influence, which partially supports H2. Finally, consumer innovation has a positive influence on green attitudes; in addition, the attitude was shown to be relevant when concerning perceived green value (mainly), usefulness, and perceived pleasure; however, a negative relationship with the factors of subjective norm and perceived control is evident. Thus, H4 is partially supported. Overall, of the fifteen relationships, twelve were confirmed, and three were not. The relationships can be seen in Figure 2.

The findings indicate that, for users of PMVs, the perceived green value, perceived enjoyment, perceived control and green attitudes are positively associated with perceived loyalty; additionally, the perceived green value could directly explain all the constructs. For this reason, based on the empirical results of the TPB and modified TAM, the feelings of enjoyment, usefulness generated and perception of care for the environment by the users of PMVs have a substantial influence on their loyalty to this type of green technology. Additionally, green attitudes also proved to be highly relevant to the factors of perceived usefulness, perceived enjoyment and perceived green value.

The results of this study support Chen’s [15] model, evidencing an important relationship of perceived green value with respect to loyalty to the use of PMVs, as well as with perceived usefulness, perceived control, subjective norms and the enjoyment of using PMVs, potentially demonstrating the environmental sense of consumers. Furthermore, innovation—the variable that was added to the model—has a strong relationship with green attitudes, which explains the adoption of technologies with innovative characteristics, such as PMVs [10].

Similarly, this study provides a perspective of attitudinal factors pertaining to the use of PMVs in the context of an emerging economy, a topic that has been little explored, and thus, the results herein serve as a basis for future studies related to the subject. Taking into account the models of the shared use of new technologies such as PMVs, authors such as Campisi et al. [12] explore the use of PMVs from the perspective of public opinion, and Hashimoto et al. [6] relate the perception of use with a reduction in traffic in urban environments from an environmental perspective. The findings from this study offer evidence, through the TPB and a modified TAM, of causal relationships in the perception of environmentally friendly technologies that could increase the environmental awareness of PMV users.

It is essential to mention that this study did not measure the perception of users of different types of VMP separately. In this regard, there are few studies in the literature that have done so. Some have slight approaches, although they are related to the operation or technical aspects of VMPs, which are the attitudinal factors that affect their use. For example, Nocerino et al. [62] exposed some differences between bicycles and scooters regarding their battery life and the possibility, due to the size of both devices, of preferring the use of scooters over bicycles. On the other hand, Boglietti et al. [10] analysed the sustainability of e-kick scooters in comparison with other types of PMVs in aspects associated with their lifetime and environmental impacts.

This study has limitations. First, the context in which the study was carried out is a limitation. Therefore, future studies should analyse the use of PMVs in other emerging economies, given that this type of analysis has only been carried out in the context of developed economies. Similarly, other factors, such as behavioural use intention and behavioural use, should be considered. An integrated model can also be proposed from the perspectives of other theories of technology adoption, such as the diffusion of innovations theory (DOI) or the unified theory of acceptance and use of technology (UTAUT and UTAUT2). Finally, along the lines of future research, it is proposed that more in-depth studies will be carried out to make direct comparisons between the different types of PMVs to identify the possible differences in the factors that encourage their use. Similarly, addressing the issue in other countries in the region would allow a comparative analysis to be made with larger samples, as well as longitudinal studies to examine the changes in attitudes towards the use of PMVs over time, including new factors, such as policy or contributory factors, thus proposing improvement actions to strengthen the use and loyalty of environmentally friendly vehicles for the mitigation of the individual carbon footprint.

6. Conclusions

The aim of this research was to determine the attitudinal factors associated with the use of PMVs in the AMVA. The green attitudes, perceived green value and loyalty have a substantial influence on the use of PMVs. Perceived control, subjective norms, perceived enjoyment and consumer innovation have a moderate influence, and perceived usefulness does not have an influence, which could be due to the topography of the area.

In the literature, the attitudinal factors associated with the use of EVs generally comprise beliefs, values and norms, as well as green attitudes. Regarding the attitudinal factors associated with the use of PMVs, the constructs of the TPB and those of the TAM, such as perceived usefulness, perceived control, perceived enjoyment, loyalty and consumer innovation, are also used. Due to their importance and robustness in previous studies, the two models were used in this study, modifying the TAM to add TPB constructs, thus becoming a modified TAM. The consumer innovation construct was added to the model and was determined to be influential in using PMVs. The moderating effects of green attitudes as a mediating variable in the relationships with loyalty were excluded to ensure the reliability and validity of the construction.

The SEM-PLS model, as a measurement tool for the analysis of the variables, is reliable and valid. Likewise, the explanatory capacity and predictive relevance highlight loyalty as one of the constructs that obtained the best results; therefore, significant future use of the model for the shared use of PMVs in the AMVA market is expected.

Research that includes other cities in the country is recommended because the application of this model reveals associations among the constructs. In addition to studying the motivations, the barriers must also be known; that is, the factors associated with the non-use of this type of vehicle, which is information that could be important because these technologies are in their infancy. Finally, there are no conflicts of interest; therefore, there are no moral situations in which the authors’ direct or indirect personal or professional interests interfere or prevent them from acting objectively.