Understanding the Determinants and Motivations for Collaborative Consumption in Laundromats

Abstract

A laundromat is a collaborative consumption alternative that is representative of a larger shift toward a sharing economy. The present study aimed to investigate determinants of laundromat use and develop a theoretical model based on the Theory of Planned Behavior to provide insights into consumer intentions regarding laundromats. This study also investigated differences among consumer motivations concerning laundromats in terms of their washing practices and sociodemographics using correspondence analyses. Data were collected from laundromat users. The model results indicated that consumer intentions were driven primarily by attitude, subjective norms, perceived behavioral control, and perceived usefulness. Further group analysis between the respondents only using laundromats and those using laundromats together with other washing choices showed different significant intention determinants. Convenience and speed were the most practical reasons for using laundromats. The correspondence analyses indicated divergent motivations of various customer segments. Our findings could be used to support laundromats and their marketing campaigns by highlighting the uniqueness of their services to gain customers at different segments and also to maintain their existing consumer base.

1. Introduction

A shift from ownership to access-based consumption has been suggested as a viable alternative in the movement toward sustainable consumption due to its environmental benefits, including material-demand reduction, resource-efficiency, and product longevity [1,2]. Bardhi and Eckhardt [3] defined access-based consumption as “transactions that may be market mediated in which no transfer of ownership occurs”. Various access-based consumption styles have been proposed; for example, consumers can pay for rental products, memberships in sharing services, and rental available spaces as alternatives to ownership [3,4]. Another example of access-based consumption is suppliers holding on to property and allowing exchanges or access for its short-term use without transferring property rights [5]. Access-based consumption is considered to be a similar concept to collaborative consumption and sharing, which has been shown to contribute to the circular economy [6]. These consumption modes are based on shared use and collaboration as opposed to personal ownership [4], which is related to a product-service system (PSS) [7]. As defined by Goedkoop et al. [8], a PSS is “a marketable set of products and services capable of jointly fulfilling a user’s needs”, and includes product leasing, renting, and sharing, e.g., car renting and laundromat [2,9]. These collaborative consumption modes are gaining popularity and surpassing traditional consumption methods, particularly since the development of information and communication technologies that enable internet-based services, e.g., Uber and Airbnb [3,10].

The number of studies on collaborative consumption has increased dramatically over the last decade, coinciding with the emergence of the sharing economy concept [7,11]. Scholars have attempted to conceptualize, theorize, and classify emerging collaborative consumption styles based on their characteristics [3,4,12]. The studies have focused on people’s motivations [6,13], personal sharing via an online platform [14,15], access-based platforms of renting and sharing [16], car sharing [17], bike sharing [18], and more. For business applications, particularly in the industrial sectors, the most focused areas are machinery and equipment, mobility (e.g., car and bike sharing), and real estate (e.g., living space) [7,11]. These studies have been mainly conducted in industrialized countries in Europe and North America [11]. Retamal [19] showed that collaborative consumption has already been established in Hanoi, Metro Manila, and Bangkok. However, collaborative consumption in emerging economies in which consumption is rapidly growing, such as in developing countries, is still underexplored.

A shared laundry, which can include a communal laundry room, self-service coin-operated laundry, or a laundromat (LM), has been discussed as an alternative consumption style in a transformation toward a circular economy [20,21]. A shared laundry typically sells washing and drying services; it also has the potential to reduce household environmental loads [22,23] and building space [23] compared with private ownership of washing machines. The general concept of LMs, as facilities that provide washers and dryers for use to customers, is not new; LMs have been used for decades in Europe and the United States [24]. With dramatic technological improvement in the laundry industry, recent industrialized washers and dryers have become much more efficient in energy and water use; they have also become faster. The LM industry also has been able to adapt to modern consumer lifestyles to increase its popularity. Modernized LM franchises have been shown to grow dramatically and sustainably in many southeast Asian countries, including Thailand [25,26]. Under the Ministry of Commerce, Thailand’s Department of Business Development promoted LM as one of the most attractive emerging businesses in 2021 due to positive consumer support by younger generations’ lifestyles and urbanization [27]. Some brands have even posted their services to an online platform via a mobile application, allowing consumers to check the availability of machines and control their laundry cycles in advance. In Thailand, several hundred new franchise LMs are planned to open across the country by 2023–2024 [25,26].

Unlike other sectors of collaborative consumption (e.g., car sharing, bike sharing, space sharing), few studies have examined this new type of laundry service [19,28,29], and those studies mainly focused on washing methods and use patterns. A systematic model to provide insights into the determinants of LM use has not yet been done, to the authors’ best knowledge. With this motivation, this study aims to investigate the factors that influence consumer intentions to use LMs in Chiang Mai, Thailand, and discuss the major psychological factors and consumer motivations for accelerating LM services.

The rest of the paper is organized as follows. Section 2 reviews the theoretical background and research hypotheses proposal. Section 3 explains the research methodology, including the questionnaire details, study area, data collection, and analysis. The results of data analysis, discussion, limitations, and future research are presented in Section 4. Finally, the conclusion is provided in Section 5.

2. Theoretical Background and Research Hypotheses

One of the most commonly applied theoretical foundations for explaining the origins of behavior is the Theory of Planned Behavior (TPB), proposed by Icek Ajzen [30]. TPB suggests that the target behavior is determined by the behavioral “intention”, which refers to a person’s willingness to perform the behavior. The “intention (INT)” is determined by three variables: “attitudes (ATT)”, “subjective norms (SN)”, and “perceived behavioral control (PBC)”. Attitudes are favorable or unfavorable perceptions about the behavior. Subjective norms refer to the recognition of social pressure from particularly important individuals or groups to perform the behavior. Lastly, PBC refers to one’s perception of the level of difficulty in performing the behavior or confidence in one’s ability to perform [30].

TPB has been used to understand other collaborative consumption behaviors, such as purchasing remanufactured laptops [31], car sharing [17], using collaborative apparel [32], and goods sharing [14]. In those previous studies, it was shown that attitudes and subjective norms had a significant positive influence on intention. More recently, Zhang and Luo [31] showed that PBC had a significant influence on the intention to buy remanufactured products. Based on these findings, we propose the following hypotheses, H₁–H₃:

Hypothesis 1 (H1).

Attitudes toward collaborative consumption have a positive influence on the intention to use LMs.

Hypothesis 2 (H2).

Subjective norms have a positive influence on the intention to use LMs.

Hypothesis 3 (H3).

PBC has a significant influence on the intention to use LMs. Consumers who feel it is easy to use LMs have a high intention to use LMs.

In addition to the influence of subjective norms, “descriptive norms (DN)”, which are defined as the perceptions about what others do or what are commonly done by most people, have been considered to be a crucial influential factor on behavior. Descriptive norms can provide a shortcut to decision making based on the presumption that it is wise to follow what other people are doing [33]. The perceived use of LM by others or by society as a whole can influence one’s intention to use LMs. Therefore, we propose the following hypothesis, H₄:

Hypothesis 4 (H4).

Descriptive norms have a positive influence on the intention to use LMs.

It has also been reported that “motivation” plays a vital role for the intention of collaborative consumption. Motivation can be classified as either intrinsic or extrinsic [34]; intrinsic motivation is when someone is driven to do something by inner pleasure and satisfaction, whereas extrinsic motivation is when someone is driven to doing something in order to obtain a particular outcome [34]. Previous studies on collaborative consumption measured the intrinsic motivation by “enjoyment”, and showed that it was significantly correlated with intention [15,17,18]. Extrinsic motivation includes economic benefits, environmental benefits, convenience, reputation, and more. Economic benefits, including money-saving, are perceived as critical motivations for collaborative consumption, particularly for short-term access such as renting [6,16]. Barnes and Mattsson [17] showed a positive influence of perceived environmental benefits on the intentions to participate in car sharing; they were also shown by Edbring et al. [6] to be key factors for renting children’s products and furniture. Such environmental and economic benefits had positive influences on intention through the recognition of usefulness (UF) [17,18]. The greater the perceived usefulness of LMs based on various benefits, the greater the intentions to use them. In addition to the monetary (MON) and environmental (ENV) benefits, another influence on the intention of using LMs is the cleanliness and hygiene (HG) benefits provided by industrial machinery. Therefore, we propose the following hypotheses, H₅–H₉, as an extension of the TPB model:

Hypothesis 5 (H5).

Enjoyment has a positive influence on the intention to use LMs.

Hypothesis 6 (H6).

Perceived environmental benefits have a positive influence on the perceived usefulness of LMs.

Hypothesis 7 (H7).

Perceived monetary benefits have a positive influence on the perceived usefulness of LMs.

Hypothesis 8 (H8).

Perceived hygiene benefits have a positive influence on the perceived usefulness of LMs.

Hypothesis 9 (H9).

Usefulness has a positive effect on the intention to use LMs.

Based on these hypotheses, we propose the hypothetical model shown in Figure 1.

In addition to the socio-psychological factors discussed above, previous studies showed the influences of personal traits, including materialism, volunteering, and sociability, to be essential determinants for collaborative consumption [14,16,35]. Belk [36] defined materialism as the importance a person attaches to possessions—the tendency to control or own tangible assets, other persons, and experiences. Materialists are defined as those who believe possessions are their sources of satisfaction. Lawson et al. [16] found that the groups with higher scores in aspects of materialism showed relatively higher scores in attitudes and intention for collaborative consumption. This is contrary to several previous studies which showed that a person with lower scores in aspects of materialism would be more inclined to share resources and more interested in collaborative consumption [36,37]. Collaborative consumption covers a wide range of products and services; in fields such as fashion and technology, customers can use services to try various products to stay up to date with the latest trends. Materialism can serve as a facilitator in collaborative consumption as opposed to an impediment [16]. Further research is required to better understand the relationship between materialism and collaborative consumption. While materialism is an important personality trait for commercial sharing, sociability—interacting with others—is a crucial trait that has a positive influence on attitudes and intentions, particularly in non-commercial sharing [14]. Based on these previous findings, we propose the following hypothesis, H₁₀:

Hypothesis 10 (H10).

Materialism and sociability have positive effects on the intention to use LMs.

3. Materials and Methods

3.1. Questionnaire Design

For this study, a questionnaire was used that consisted of three parts: (1) questions on current practice of laundry activities, (2) questions about psychological variables and personal traits, and (3) questions on socio-demographics.

In the first part, respondents were asked about frequency of use of each option in daily laundry practices and frequency and duration of LM use. Respondents were also asked to provide up to three reasons for using LMs in an open-ended answer format. In the second part, questions about the psychological variables and personal traits discussed in Chapter 2 were asked using 21 items, as shown in

Table 1. List of variables and items to measure assumed influential factors.

Variable	Item	Questionnaire Statement	Ref.
Intention (INT)	Int1	I want to use LMs in the future.	[15]; [17]
	Int2	I expect to continue using LMs in the future.
Attitudes (ATT)	Att1	Sharing goods and services are a positive thing.	[15]
	Att2	Sharing or CC is a better mode of consumption than selling and buying individually.
	Att3	I find participating in CC of goods and services such as using LM is a good move.
PBC	Pbc1	LMs are flexible, I can do laundry anytime.	[31]
	Pbc2	Using LMs is easy.
Subjective norm (SN)	Sn	My family or people close to me encourage me to use LMs.	[17]
Descriptive norm (DN)	Dn	Other people use LMs. *
Enjoyment (EJ)	Ej1	Using LMs is pleasant.	[17]
	Ej2	I enjoy using LMs.
Environmental benefits (ENV)	Env	Using LMs is a sustainable consumption choice.	[15]
Monetary benefits (MON)	Mon	Using LMs is worth the money.	[17]
Hygiene (HG)	Hg1	Using LMs is clean. *
	Hg2	Using LMs is hygienically safe. *
Usefulness (UF)	Uf	The advantages of LMs outweigh the disadvantages.	[17]
Materialism (MAT)	Mat1	I would be happier if I could afford to buy more things.	[14], [36]
	Mat2	I want to be rich enough to buy anything I want.
	Mat3	I would rather buy something I need than borrow or rent it.
Sociability (SOC)	Soc1	I prefer working with others rather than being alone.	[14]
	Soc2	I love to be with people.

* Created by this study.

. Respondents were asked to score each item using a 6-point scale, ranging from “strongly agree (6)” to “strongly disagree (1)”. In the third part, questions were asked about the respondents’ socio-demographics, such as age, gender, education level, occupation, and type of living accommodation. The questionnaire was created using Google Form, and the respondents accessed the questionnaire via a quick response code from their smartphones.

3.2. Study Area and Data Collection

The questionnaire survey was conducted for LM users in Mueang Chiang Mai district, Chiang Mai, Thailand. Chiang Mai is the largest city in northern Thailand, with an area of 152 km² and 1.63 million people; 14% of its population lives in the Mueang Chiang Mai district [38]. As of September 2021, 58 LM stores were in service in Chiang Mai city, as shown in Figure 2. The stores were located throughout the city, and three modernized franchised brands owned 69% of these stores. In this study, we targeted the western part of Chiang Mai city and surveyed users of 26 franchised stores, green squares in Figure 2, covering the low- to high-population density areas. The survey started from July to September 2021. A total of 469 valid responses were obtained and used for further analysis.

Details of the respondents’ characteristics are provided in Appendix A (

Table A1. Characteristics of the respondents.

Items	Category	All	LMonly	LMplus a	Difference between LMonly and LMplus
Number of samples		469	239	230
Gender	Male	35.2%	34.7%	35.7%	Chi-square = 0.006, df = 1, p = 0.94
	Female	64.0%	63.6%	64.3%
	Not specified	0.9%	1.7%	0.0%
Age	18–24 yr	49.5%	54.8%	43.9%	Mann–Whitney U = 23,055.5, p = 0.001
	25–29 yr	27.1%	28.9%	25.2%
	30–34 yr	12.4%	9.2%	15.7%
	35–39 yr	4.7%	3.3%	6.1%
	40–44 yr	3.0%	2.1%	3.9%
	45–49 yr	1.1%	0.8%	1.3%
	50–54 yr	1.7%	0.4%	3.0%
	55–59 yr	0.6%	0.4%	0.9%
Education level	Primary	0.2%	0.0%	0.4%	Chi-square = 0.066, df = 2, p = 0.968
	Secondary	6.8%	6.7%	7.0%
	Diploma	1.1%	1.3%	0.9%
	Bachelor	78.5%	78.2%	78.7%
	Higher	13.4%	13.8%	13.0%
Family size	1 person	32.0%	38.1%	25.7%	Mann–Whitney U = 20,872.5, p < 0.001
	2 people	40.5%	45.2%	35.7%
	3 people	10.7%	6.7%	14.8%
	>4 people	16.8%	10.0%	23.9%
Occupation	Company employee	21.5%	24.7%	18.3%	Chi-square = 23.016, df = 5, p < 0.001
	Government employee	8.7%	7.1%	10.4%
	Self-employed	14.9%	9.6%	20.4%
	Freelance	9.2%	8.8%	9.6%
	Student	42.4%	48.5%	36.1%
	Others	3.2%	1.3%	5.2%
House type	Condo/Apartment	16.8%	18.8%	14.8%	Chi-square = 49.721, df = 2, p < 0.001
	Detached house	24.5%	10.9%	38.7%
	Room rental	57.4%	69.0%	45.2%
	Tenement	1.1%	0.8%	1.3%
	Others	0.2%	0.4%	0.0%
Experience of LM use	<1 month	7.9%	2.1%	13.9%	Mann–Whitney U = 21,433.5, p < 0.001
	1–3 months	9.0%	7.5%	10.4%
	4–6 months	8.1%	7.1%	9.1%
	6 months–1 year	12.2%	10.9%	13.5%
	1–2 years	24.1%	28.0%	20.0%
	>2 years	38.8%	44.4%	33.0%
Frequency of LM use	≥4 times per week	1.3%	0.8%	1.7%	Mann–Whitney U = 20,613, p < 0.001
	2–3 times per week	23.5%	30.5%	16.1%
	1 time per week	53.9%	56.1%	51.7%
	every 2 weeks	15.8%	12.1%	19.6%
	every 3 weeks	1.1%	0.4%	1.7%
	Occasional	4.5%	0.0%	9.1%
Service use at LM	Wash and dry	74.0%	77.0%	70.9%	Chi-square = 23.987, df = 2, p < 0.001
	Wash only	20.7%	21.8%	19.6%
	Dry only	5.3%	1.3%	9.6%
Expenses	THB per wash	44.6	41.3	48.5	Mann–Whitney U = 21,213.5, p = 0.014
	THB per dry	42.8	41.2	44.5	Mann–Whitney U = 15,025.5, p = 0.001

^a Use LMs with private washing machine (42.2%); with laundry service (20.9%); with hand wash (11.3%); with more than two choices (25.7%).

). Most respondents were female (64%), aged 18–24 years old (49.5%) and 25–29 years old (27.1%), had acquired a bachelor’s degree (78.5%), lived in a rented room (57.4%), and had a family size of two people (40.5%). Half of the respondents (51.0%) used only LMs for their daily laundry choice, while the others (49.0%) used LMs together with other choices such as private washing machines, laundry services, and hand washing. Half of the respondents (54.1%) visited LMs once a week. More than half (62.9%) of the respondents had used LMs for more than one year, and 75.0% of the respondents came to LM stores to use washing and drying services.

3.3. Data Analysis

The analysis steps of this study are described as follows. First, descriptive statistical analysis was conducted for the respondent profiles and the measurement variables of psychological and personal traits using SPSS Statistics version 28.0 (IBM Corp., Armonk, NY, USA). Second, to determine the factor influencing consumers’ intention to use LM, we employed a factor analysis (FA) and a structural equation modeling (SEM), using SPSS Statistics version 28.0 and SPSS AMOS version 28.0 (IBM Corp., Armonk, NY, USA), respectively. These analyses used input data from the questions on the current practice of laundry activities and psychological variables, and personal traits. Third, to examine consumers’ reasons for using LMs, we used the correspondence analysis (CA) method via a CORRESPONDENCE syntax of SPSS Statistics version 28.0. Data from the questions on the current practice of laundry activities and socio-demographics were used as inputs for CA. Details of each step are explained as follows.

For data preparation, the questionnaire responses based on the 6-point scales of the psychological variables and personal traits questions were converted into scores ranging from 1 (“strongly disagree”) to 6 (“strongly agree”) before using for the analyses. The respondents were divided into two groups regarding their daily laundry choice: (1) only LM use for daily laundry (LM_only) and (2) LM use in addition to other options (LM_plus). A Mann–Whitney U non-parametric test was applied for the comparison of the two groups because the data did not fall under a normal distribution.

The FA was conducted for five personality items targeting materialism and sociability to verify the appropriateness of the measured items. The FA was performed using the maximum likelihood method with the Promax rotation. Two latent factors were extracted as expected for materialism and sociability items with 72.6% of the total variance explained. The result of Bartlett’s test of sphericity was significant (p < 0.001), and the Kaiser–Meyer–Olkin (KMO) statistic was 0.66, which was greater than the minimum requirement of 0.50 for the appropriateness of FA [39].

The other 16 psychological factors were tested by SEM with the maximum likelihood estimates and bootstrapping (BS). Since our data violated the assumption of the multivariate normal distribution for the maximum likelihood estimation, we ran BS, which can be used under non-normal distributions, with 2000 iterations to achieve robust statistics in SEM [40,41]. To assess the measurement model validity, the following criteria were examined: the standardized factor loadings should be greater than 0.5, the average variance extracted (AVE) should be greater than 0.5, the construct reliability (CR) should be greater than 0.7, and Cronbach’s alpha should be greater than 0.7 [39]. The discriminant validity was also examined for the construct uniqueness by comparing the AVE with the squared correlation estimate between constructs to determine if the AVE was greater [39]. Four latent factors met the above criteria as expected; however, ‘enjoyment’ did not, as shown in Appendix A (

Table A2. Factor loadings and convergent validity test.

Factor	Item	Std Loading a	Cronbach’s α	CR	AVE
Intention (INT)	Int1 Int2	0.96 0.94	0.94	0.93	0.90
Attitude (ATT)	Att1	0.84	0.91	0.91	0.74
	Att2	0.88
	Att3	0.85
PBC	Pbc1	0.88	0.79	0.84	0.65
	Pbc2	0.72
Hygiene (HG)	Hg1	0.82	0.87	0.85	0.74
	Hg2	0.90
Criteria		≥0.5	≥0.7	≥0.7	≥0.5

^a A 2000-bootstrapping with 95% bias-corrected confidence intervals, p < 0.001; CR: construct reliability; AVE: average variance extracted.

and

Table A3. Discriminant validity test.

Construct	INT	ATT	PBC	HG
INT	0.95
ATT	0.53	0.86
PBC	0.43	0.53	0.80
HG	0.49	0.54	0.47	0.86

Diagonal items represent the squared root of average variance extracted for each construct. The items below the items in bold font represent the correlations between constructs with a 2000-bootstrapping and all significant levels < 0.01. INT: intention to use LMs, ATT: attitudes toward collaborative consumption, PBC: perceived behavioral control, DN: descriptive norm, ENV: environmental benefits, and HG: hygiene.

). Two items for ‘enjoyment’, Ej1 and Ej2, showed cross-loadings, and the AVE of ‘enjoyment’ was smaller than the squared correlation with PBC; therefore, they were removed from the model analysis. We then examined the hypothetical model using the entire dataset (n = 469) and performed a group analysis to compare the two user groups, LM_only (n = 239) and LM_plus (n = 230).

The CA was conducted to understand the relationship between the reasons for using LM and the relationship between the reasons and personal attributes. CA is a compositional technique used on nominal data to examine the associations between two groups of variables. It uses a table of frequencies to calculate the associations between row and column categories using Chi-square statistics, and then, it involves plotting the standardized measured scores in a correspondence map. On the map, proximity indicates the level of association; the closer the score, the higher the level of association between categories [39]. For this analysis, we first performed a qualitative content analysis in which we grouped and categorized the input texts received for reasons for using based on the meaning, and then, we summarized the number of keywords about reasons for using LMs into a cross-tabulation table that aggregates by personal attributes. The cross-tabulation table was then utilized in the CA.

4. Results and Discussion

4.1. Analysis of the Measurement Variables

The distribution of the respondents is shown in Appendix A (Table A1). As presented above, 51.0% of the entire respondents (n = 469) used only LM for their daily laundry (LM_only) while 49.0% used LM together with other laundry methods (LM_plus) which included a private washing machine (42.2%), laundry service (20.9%), hand washing (11.3%), and more than two choices (25.7%). No significant difference was observed in the proportions of gender and education level between LM_only and LM_plus (p > 0.05). LM_only were significantly younger, had a smaller family size, showed a higher frequency of LM use (56.0% used it weekly), and had longer experience using LMs (44.0% had used them for more than two years). No respondents in LM_only visited LMs solely for the purpose of drying, while 9.5% of LM_plus visited LMs strictly for drying. The average price for one service use of washing and drying was THB 44.6 and THB 42.8, respectively (about USD 1.3 for each washing and drying, THB 1 ≅ USD 0.03). LM_plus showed significantly higher payments for one service use (p < 0.05), possibly due to the fact that LM_plus used LMs less frequently and the amount of washing for a one-time use can be larger.

The scores of 21 items are shown in Figure 3. The items for environmental benefits (Env) showed relatively lower scores than other items. High scores of over five points were observed in the PBC items (Pbc1 and Pbc2).

Comparing between the two groups, the LM_only scores were higher than the LM_plus scores in all items. Significantly higher scores were observed in intention (Int1 and Int2), PBC (Pbc1 and Pbc2), subjective norms, monetary benefits, hygiene (Hg1 and Hg2), usefulness, and enjoyment (Ej1) (Mann–Whitney U, p < 0.05).

To investigate the influence of personality traits on the intention to use LMs, respondents were divided into two groups (Low and High) based on their scores of each factor (Mat_Low and Mat_High; Soc_Low, and Soc_High). Figure A1 in Appendix A shows the ratio of Low and High groups in each socio-demographic group. As seen in the figure, the share of Mat_High was significantly higher in younger (χ² = 19.32, df = 7, p < 0.01) and female groups (χ² = 14.14, df = 2, p < 0.01), while these differences were not significant for sociability. The average scores for intentions to use LMs of Mat_High, Soc_High, Mat_Low, and Soc_Low were 4.81 ± 1.22 (n = 239), 4.84 ± 1.05 (n = 227), 4.44 ± 1.00 (n = 230), and 4.44 ± 1.18 (n = 242), respectively. These scores indicate that the respondents with higher materialism or higher sociability showed higher intentions to use LMs. The differences between the Low and High groups were statistically significant for both the personality traits (materialism: Mann–Whitney U = 20440, p = 0.000; sociability: Mann–Whitney U = 21476, p = 0.000). These results support our hypothesis (H₁₀) and are consistent with the findings of previous studies, such as [14,16]. Among the two LM user groups, LM_only showed higher materialism (5.23 ± 0.74) and higher sociability (4.34 ± 1.07) than LM_plus (5.10 ± 0.87 and 4.23 ± 1.24), but the differences were not statistically significant (materialism: Mann–Whitney U = 25,395, p = 0.145; sociability: Mann–Whitney U = 26,685, p = 0.579).

4.2. Model Evaluation

The overall model, which assumes the same path coefficients in all samples, showed acceptable model indices, as shown in

Table A4. Model fit indices.

Indicator	Criteria	Overall	Group Analysis
Chi-squared/df	≤5	3.806	2.744
RMSEA	<0.08	0.077	0.061
GFI	>0.80	0.940	0.919
CFI	>0.90	0.958	0.948
NFI	>0.90	0.945	0.922
TLI	>0.90	0.932	0.915
PNFI	>0.50	0.581	0.567
PCFI	>0.50	0.590	0.583

df: degree of freedom, RMSEA: Root Mean Square Error of Approximation, GFI: Goodness of Fit Index, CFI: Comparative Fit Index, NFI: Normed Fit Index, TLI: Tucker–Lewis Index, PNFI: Parsimony Normed Fit Index, PCFI: Parsimony Comparative Fit Index.

. The comparative fit index (CFI) was 0.958 (>0.90), and the root mean square error of approximation (RMSEA) was 0.077 (<0.08). The estimation results of the overall model are provided in Appendix A (Figure A2) and the total effect from each variable on the intention is listed in the first three columns in

Table 2. Total effect of each variable on the intention to use LMs.

Items	Standardized Effects on INT
	Overall (n = 469)			Group Analysis (n = 469)
				LMonly (n = 239)			LMplus (n = 230)
	Direct	Indirect	Total	Direct	Indirect	Total	Direct	Indirect	Total
ATT	0.41 ***	-	0.41 ***	0.35 ***	-	0.35 ***	0.46 ***	-	0.46 ***
PBC	0.19 ***	-	0.19 ***	0.16 *	-	0.16 *	0.24 ***	-	0.24 ***
SN	0.16 ***	-	0.16 ***	0.05	-	0.05	0.24 ***	-	0.24 ***
DN	-	0.09 ***	0.09 ***	-	0.06	0.06	-	0.11 ***	0.11 ***
ENV	-	0.16 ***	0.16 ***	-	0.13 ***	0.13 ***	-	0.18 ***	0.18 ***
MON	-	0.03 ***	0.03 ***	-	0.06 ***	0.06 ***	-	0.01	0.01
HG	-	0.06 ***	0.06 ***	-	0.11 ***	0.11 ***	-	0.02	0.02
UF	0.14 ***	-	0.14 ***	0.23 ***	-	0.23 ***	0.05	-	0.05

Bootstrap confidence intervals (bias-corrected percentile method), *** p < 0.01, * p < 0.10. INT: intention to use LMs, ATT: attitudes toward collaborative consumption, PBC: perceived behavioral control, SN: subjective norm, DN: descriptive norm, ENV: environmental benefits, MON: monetary benefits, HG: hygiene, and UF: usefulness.

. The ATT, SN, and PBC showed significant positive direct effects on the INT to use LMs as hypothesized (H₁, H₂, and H₃); the standardized path coefficients were 0.41 (p = 0.001), 0.16 (p = 0.003), and 0.19 (p = 0.004), respectively. The direct effects of DN on the INT were found to be insignificant (p > 0.10); however, significant indirect effects were observed on the INT (p = 0.001). The direct effect of UF on the INT and the influences of MON and HG benefits on UF were significant, as hypothesized (H₇, H₈, and H₉). However, contrary to the hypothesis (H₆), ENV showed a positive indirect effect (0.16, p = 0.001) on the INT via ATT as opposed to influences related to UF. This result was similar to those of Hamari et al. [15] in which perceived environmental benefits were shown to have a significant positive effect on the attitudes toward collaborative consumption, but the effect was not significant on the intention. The tests of a mediating effect of ATT, SN, and UF using a bias-corrected BS method are presented in

Table 3. Mediation effects.

Path	Overall (n = 469)			LMonly (n = 239)			LMplus (n = 230)
	Effect	SE 1	p	Effect	SE 1	p	Effect	SE 1	p
via attitude
ENV → ATT → INT	0.148	0.036	0.001	0.113	0.045	0.001	0.184	0.057	0.001
via subjective norms
DN → SN → INT	0.100	0.038	0.008	0.026	0.038	0.510	0.131	0.046	0.002
via usefulness
DN → UF → INT	0.031	0.015	0.005	0.039	0.021	0.008	0.013	0.022	0.358
MON → UF → INT	0.033	0.015	0.007	0.066	0.030	0.004	0.009	0.015	0.306
HG → UF → INT	0.067	0.027	0.007	0.137	0.053	0.004	0.019	0.029	0.434

¹ Bootstrap standard errors, p-value < 0.05 are in bold.

. For the entire sample analysis, as shown in the first three columns of the table, all mediating effects were supported (p < 0.01): ENV → ATT → INT; DN → SN → INT; DN, MON, HG → UF → INT.

The group analysis, which assumes different coefficient values between groups, was performed with the two user groups, LM_only and LM_plus. The model fit indices for the group analysis were acceptable: CFI = 0.948 > 0.9 and RMSEA = 0.061 < 0.08; the results are shown in Figure 4 and Table 2. As with the overall model, ATT showed the largest significant positive effect on INT in both groups: the standardized total effects were 0.35 (p = 0.002) for LM_only and 0.46 (p = 0.001) for LM_plus, respectively. The PBC and SN showed the second largest effects on INT after the ATT for LM_plus (0.24, p < 0.01), whereas the effect of SN was small and insignificant for LM_only (0.05, p = 0.546). For LM_only, the UF showed the second largest effect on INT (0.23, p = 0.006), while its effect for LM_plus was very small and insignificant (0.05, p = 0.453). The mediating effect of ATT between ENV and INT was similarly significant for both groups (Table 3). The influence of DN on the INT via the SN was significant only for LM_plus, while the mediating effects of UF were supported only for LM_only. Significant differences in the path coefficient between the two groups were observed in the paths of SN → INT (p = 0.040), UF → INT (p = 0.045), and HG → UF (p = 0.055). The result of the group analysis indicated that attitude had the strongest influence on the intention to use LMs for both consumer groups. For the respondents who only use LMs, perceived usefulness was an important predictor of their intention as well as attitude, while the perceived pressure from close people and PBC were more significant for those who were not restricted to LM use only.

4.3. Correspondence Map of Reasons for LM Use

To understand the reasons for the use of LMs, a CA was conducted. Prior to the analysis, the keywords observed in the open-ended answers were categorized into five groups based on the 21 items, as shown in

Table 4. Reasons for using LMs.

Category	No.	Label	Reason	Count
Necessity	(1)	NoSpace	No space for washing machine/drying	6
	(2)	NoDryer	No dryer	8
	(3)	NoWM	No washing machine	50
Quality of LM service	(4)	Safe	Feeling LM as safe place to use	6
	(5)	Machine	Good quality of machines	3
	(6)	BigSize	Availability of big loading size	15
	(7)	Clean	Cleanliness	85
	(8)	D_smell	Dryer use to avoid unpleasant smell	10
	(9)	Service	Extra laundry services (i.e., folding and service staff)	12
	(10)	FreeServ	Other free services (i.e., Wi-Fi and waiting area)	17
Money-saving	(11)	Price	Affordable price	55
	(12)	Saving	Saving on products and utilities used at home (i.e., detergent, water, and electricity)	14
Time-saving	(13)	Time	Time-saving/speed	158
	(14)	TimeDry	Time-saving to use dryer	93
Convenience	(15)	Access	Good access from living places	57
	(16)	Maintenance	Unnecessity of maintenance	8
	(17)	Convenience	Convenience	295
	(18)	Anytime	Usage at anytime	13
	(19)	LessEffort	Less efforts for laundry	32
	(20)	Ease	Easy to use	66
	(21)	CloseCon	Good access to convenience stores	6
Total count				1009

. The five categories of the reasons for LM use included the necessity to use machines, service quality of LM, money-saving, time-saving, and convenience. Based on each item counting, convenience (No. 17), time-saving (No. 13 and No. 14), and cleanliness (No. 7) were the main reasons to use LMs. The count values were then divided by LM usage conditions and socio-demographics, and the cross-tabulation table was used for the correspondence analysis. Details of LM usage conditions and socio-demographics and their labels used in the CA maps are listed in

Table 5. Categories used in the correspondence analysis.

Category	Sub-Category		Label	Details
LM usage condition	(a) Service use at LM	(a-1)	W	Wash only
		(a-2)	D	Dry only
		(a-3)	WnD	Wash and dry
	(b) Experience of LM use	(b-1)	<1 month	<1 month
		(b-2)	1–3 month	1–3 months
		(b-3)	4–6 month	4–6 months
		(b-4)	6–12 month	6 months–1 year
		(b-5)	1–2 years	1–2 years
		(b-6)	>2 years	>2 years
	(c) Frequency of LM use	(c-1)	<1/week	<1 time per week
		(c-2)	1/week	1 time per week
		(c-3)	>1/week	>1 time per week
Socio-demographics	(d) Age	(d-1)	20 s	18–29 years old
		(d-2)	30 s	30–39 years old
		(d-3)	40 s	40–49 years old
		(d-4)	50 s	50–59 years old
	(e) Gender	(e-1)	Female	female
		(e-2)	Male	male
	(f) Family size	(f-1)	F1	1
		(f-2)	F2	2
		(f-3)	F3	3
		(f-4)	F4	more than 4

. For LM usage conditions, three points were involved: (a) service use at LMs, (b) experience of LM use, and (c) frequency of LM use. For socio-demographics, additional points of (d) age, (e) gender, and (f) family size were included.

Figure 5 displays the associations between LM usage conditions and socio-demographics and the reasons to use LMs plotted by their correspondence scores. The total variation explained by the two dimensions was 53.8%, which was a sufficiently high explanation power to be discussed. The closer the plots, the higher the association between them. Since symmetrical normalization was applied, the respondents’ characteristics and associated reasons could be compared [42]. “NoDryer” was shown to be the most distinguished reason on the first dimension and was more likely to relate with the respondents who visit LMs strictly to use the dryers (D). In contrast, the most likely reason for the respondents who visit LMs only for washing (W) was having no washing machine “NoWM”. Other reasons such as “Convenience”, “Clean”, “Ease”, “Time”, and “LessEffort” showed similar scores as they accumulated together and were the most probable motivations for the respondents in their 20s, female, and who had used LMs less than one month, 1–2 years, or more than 2 years. “BigSize” was more likely the reason for males and those with three family members to use LMs. Respondents living alone were more associated with the reasons of “FreeServ”, “BigSize”, and “Saving”, which included saving money from purchasing a washing machine as well as saving electricity consumption, water use, and detergent use at their own home. These cost-saving benefits were of greater concern than the respondents who lived alone. “Service” was more related to “Anytime” and “TimeDry”; these were more likely the concerns of the respondents with a bigger family size and the respondents in their 40s. Respondents in their 30s, however, used LMs for reasons such as “Maintenance”, “CloseCon”, and “Safe”.

The associations between the reasons for using LMs and the respondents’ characteristics were further examined between the LM_only and LM_plus groups. As can be seen in Figure 6, the plot points for of LM_only (black diamonds) and LM_plus (green triangles) are clustered around different locations on the axes. The most likely reasons for the LM_only included “Clean”, “Price”, “NoWM”, and “Convenience”, while the most likely reasons for the LM_plus were “BigSize”, “Service”, “Anytime”, “TimeDry”, and “D_smell”. These differences can be explained by the fact that most LM_plus consumers would have their own washing machines and visit LMs for more specific options in addition to ordinary washing, such as dryer use and extensive-load washing. A dryer is an uncommon household appliance among Thai households; typically, clothes are dried naturally. In contrast, LM_only consumers use LMs to meet the basic need of clean clothes as most do not own a washing machine. Similar motivations of “Ease” and “Machine” were observed for using LMs between LM_only and LM_plus by the respondents who were relatively new to LMs.

4.4. Discussion

The results of the study showed that most LM users were in their 20s, and the respondents with higher intentions to use LMs exhibited more qualities related to materialism and sociability. These characteristics of collaborative consumption users are in line with those discussed in the existing literature. Respondents who were experienced in using sharing services, such as ride and food sharing, were relatively young, with higher tendencies toward materialism and normative personalities [13]. Regarding the higher percentage of younger LM users, one reason may be that they are more flexible and more likely to use new services [13]. As discussed earlier, the modernized LMs offering both washers and dryers are an emerging business sector in Thailand. LMs may be more attractive to the younger generation due to their advantages that include convenience, speed, cleanliness, and ease of use (Figure 5). Another explanation may be that the younger consumers are more likely to have constraints on owning their own washing machine, such as budget or space availability. As seen in Figure 6, respondents in their 20s who only use LMs are primarily motivated by the fact that they do not own a washing machine, whereas they are more likely to be motivated by the affordable pay-per-use of LM. This young demographic of LM users may no longer define themselves by their ownership of possessions; in addition, they want to avoid the ownership-related hassle of paying maintenance costs [12,43].

Regarding personal traits, our results indicated that stronger characteristics of materialism do not inhibit the intention to use sharing services such as LMs. Akbar et al. [44] demonstrated that the negative impact of materialism on the intention to use sharing services is negligible for highly desired goods or unaffordable products. In the present study, the temporary on-demand accessibility of LMs could meet the users’ need to clean their clothes, particularly for those who cannot afford their own washing machines. In addition, using LMs may be seen as a more attractive option by those with stronger materialistic qualities as they are more able to access various and up-to-date laundry technologies and services. Consequently, they may gain favorable opinions of short-term accessed services as opposed to ownership. Bucher et al. [14] found sociability to be a primary motivation for both commercial and non-commercial sharing purposes. Our results similarly demonstrated that respondents with higher sociability showed higher intention to use LMs.

Our results also showed that attitudes, subjective norms, and PBC were significant determinants of intention to use LMs, which supported the TPB framework. In addition to that original framework, our model results indicated that the intention was also positively influenced by the perceived usefulness, which was in turn affected by the perceived monetary benefits, hygiene benefits, and descriptive norms. Attitudes were the most prominent determinant of the intentions to use LMs for all respondents; the higher the attitude, the higher the intention to use LMs. Comparing between the two user groups, the LM_only was more likely to use LMs for their usefulness whereas the LM_plus was more influenced by the subjective norms and PBC to use LMs. These different determinants for the intentions between the two groups were also observed in the correspondence analysis results (Figure 6), which revealed that affordable price and cleanliness were strong motivators for LM_only to use LMs, while the LM_plus users were more motivated by the time flexibility and ease of use relative to PBC.

Convenience and speed were the most probable reasons for using LMs (Table 4). In terms of convenience benefits, it could be obtained for several reasons, including ease of accessibility to LMs, attractive locations of LMs nearby other shops, anytime accessibility, ease of machine use, and maintenance-free responsibility. Promising these benefits provided would attract consumer engagement with LM stores. In addition, specific motivations corresponding to different consumers’ characteristics were observed, e.g., age and family size (Figure 5), particularly between LM_only and LM_plus (Figure 6). The results can provide insights into approaches for promoting greater LM use. For instance, the LM_only were more motivated to use LMs for their money-saving benefits, i.e., affordable price and saving other things expenses, than were LM_plus. This is likely due to the fact that LM_plus use LMs for extra or additional services, such as dryers and extensive size loading, which are costly per use. Promoting their affordable prices would help LMs be seen as an attractive, price-friendly alternative to ownership. In contrast, providing special services, such as dryers and extensive size loading, would attract those engaged in various washing choices to use LMs. To attract new consumers, LMs should more effectively promote their superior quality of industrialized washers and dryers and ease of use, as the results show that these are the two primary motivations of the less experienced LM users (Figure 5 and Figure 6).

Our findings provide insights on several possible motivations and drivers for practical implications for policy and marketing strategies encouraging LM use. Nonetheless, the present study has some limitations. Firstly, our measures were based on self-reports. The self-report method might have a ubiquitous bias of social desirability leading to over- or under-reporting [45]. Secondly, the present study was cross-sectional, which is a helpful approach to examining and understanding the current situation of people’s behaviors, their determinants, and ways for improvement. However, the result may be constrained to future changes, as people’s behaviors and preferences may change over time and affect key behavior predictors. Lastly, the present study added to the literature on collaborative consumption in the context of developing society, but was limited to Chiang Mai, Thailand. Future studies in other developing countries can validate and expand upon the insights gained from this study, leading to improved modes of collaborative consumption in society.

5. Conclusions

The present study investigated the determinants of consumers’ use of LMs as an alternative collaborative consumption model in a shift toward a sustainable sharing economy. Key predictors of people’s intention to use LMs were determined by modifying the TPB. The associations between people’s characteristics and motivations to use LMs were also analyzed by a correspondence analysis. A total of 469 respondents were obtained among 26 LM store customers in Chiang Mai province, Thailand. The overall model estimation indicated that attitude was the strongest predictor of the intention to use LMs, followed by PBC and subjective norms. Perceived usefulness also showed a significant impact on the intention to use LMs in addition to the original TPB determinants. Considering the two different user groups, LM_only and LM_plus, attitude was still the most significant influence on the intention for both groups. The perceived usefulness was more impactful for LM_only, while subjective norms and PBC were vital in predicting the intention of LM_plus. Most people use LMs due to their convenience and speed. The correspondence analysis results indicated divergent motivations between customer segments. Convenience, cleanliness, and money-saving benefits were the strongest motivations for people who had no washing machines and only used LM for daily laundry. Using additional services, such as dryers and extensive size loadings, was the primary benefit to motivate those engaged in various washing choices to use LMs. Less experienced customers use LMs due to their superior machine quality and ease of use.