Effect of Characteristics of Shared Housing in Single-Person Households on Housing Satisfaction and Shared Housing Performance

: This study quantitatively evaluated South Korean residents’ housing satisfaction and factors affecting shared housing and made recommendations for policy establishment for the supply and revitalization of shared housing in South Korea. Using an online and ofﬂine survey, the study analyzed the relationship between the satisfaction with shared housing, intention to reside again in shared housing, and intention to recommend shared housing to others among young people living in shared housing in Seoul. The results showed that shared housing characteristics signiﬁcantly affected the housing satisfaction but not the intention to reside again and intention to recommend. Housing satisfaction signiﬁcantly affected the intention to reside again and the intention to recommend. Residents of public-supply shared housing showed no signiﬁcant correlation between the shared housing characteristics and housing satisfaction; those of private-supply shared housing showed increased housing satisfaction. Housing satisfaction did not signiﬁcantly affect the intention to reside again among people in public-supply shared housing; however, it affected the intention to reside again among those in private-supply shared housing. The physical location and environment and community factors did not signiﬁcantly affect overall housing satisfaction in public-supply shared housing but signiﬁcantly affected the housing satisfaction and intention to reside again in private-supply shared housing. These results support the need for regulation and policy to guide housing adjustments and facilitate lifestyles, the need for diversiﬁcation in housing types, and the importance of uniform management and operations of public-run units.


Introduction 1.Background and Purpose of the Study
In recent years, challenges to the global housing market have been trending among international stakeholders.This scrutiny stems from adverse conditions such as urban sprawl, reduced housing stock, migratory patterns, societal aging, quality and affordability, and government policies [1][2][3][4][5].As a consequence of young people having been particularly affected as governments have not always kept pace with contemporary social changes [2], housing policies have not always addressed constraints that hinder young people from accessing equal opportunities to live independently or suitable and available accommodations [2].Currently, limited housing options have forced single young people into shared housing arrangements with mixed experiences [5,6].Bricocoli and Sabatinelli note that sharing spaces is not always a happy medium despite the economic benefits [7].One recommendation from their Mediterranean study was a need for more supportive housing policies for alternative housing.This is an acute requirement also in the U.K., where limited rentals, planning restrictions, high rents, and reduced housing benefits have forced young people into shared accommodations [4].The provision of housing benefits and subsidized rentals as in the USA [8] has been a precarious panacea to support single living because of urbanization [9].In this regard, fit-for-purpose and satisfactory living space is important in the shared housing market.In places such as Amsterdam, the housing shortage for young adults is exasperated by its reputation as a university city [3].Additionally, not all countries are well-regulated and able to provide subsidized or specialized housing options.In post-socialist countries such as Serbia, the private rental market is unregulated, and government-subsidized housing is diversified [10].Where the renewed scholarship in shared spaces aims to understand how youths engage within their economic and social environments [1,6,7], this research adds to the literature by examining the relationship between housing satisfaction and structural variables as an important driver for change among single young people within the urban city of Seoul, with the overall result of these disruptions being that young people worldwide are often left without the vital infrastructure to help them maneuver the housing market and positively affect available, satisfactory accommodation and housing policies.The renewed scholarship in shared spaces aims to understand how youths engage within their economic and social environments [1,6,7]; this research adds to the literature by examining the relationship between housing satisfaction and structural variables as an important driver for change among single young people within the urban city of Seoul.
South Korea's housing patterns are changing rapidly, with low birth rates, low growth, an increase in single-person households, and emerging housing problems for the younger generation.According to data from the National Statistical Office's "2020 South Korea Social Indicators," the South Korean population has continued to decline since 2010, while the proportion of single-person households has steadily increased from 23.9% in 2010 to 33.4% in 2021.In particular, the proportion of young single-person households in the Seoul metropolitan area of 59.2% is higher than in other regions, has more than doubled over the past 20 years, and is predicted to increase until 2047 [11].According to the Population and Housing Survey (National Statistical Office), there has been some improvement in housing poverty among domestic households.In this paper, housing poverty is defined as "the basic problems of housing shortage, poor physical conditions and overcrowding; the suitability of the dwelling stock in terms of tenure, type, size, location and other qualitative aspects of the dwelling and the neighborhood environment; and the ability of households to access suitable dwellings" [12].However, youth housing poverty is worsening because of the increase in single-person households that fail to meet the minimum housing standards.In South Korea, minimum standards are based on floor space (4 persons at a minimum of 37 sq m), facilities, structure, and environment [12].The housing poverty rate is worse for youth living in Seoul than elsewhere in the country.There is an increase in single-person youth households, exacerbating the youth housing poverty problem in two aspects: the burden of housing costs and the high-cost housing environment for young people with low economic stability [13] where single-person youth households bear housing costs about twice as much as three to four households in the family unit.
Until now, although the central government and local governments provide rent deposits to assist young people with public rental housing, South Korea's housing policy has focused on apartments for three-to four-people households, while small houses, in which most young single people live, have been highlighted as the cause of youth housing problems owing to their relatively poor environment.Nevertheless, there is a lack of housing supply because of the difficulty in securing land and relatively weak support for young people with low economic credibility.Consequently, shared housing has emerged as a new type of housing for single-person households to reduce the burden of housing costs and improve the quality of housing as social awareness of sharing has recently expanded along with the youth housing problem.
The sharing economy has expanded globally, as the paradigm of consumption has shifted from owning to sharing owing to stagnant economic growth.As such, new alternative housing types such as room sharing and shared housing are emerging in the housing market to reduce housing costs.As such, new alternative housing types such as room sharing and shared housing are emerging in the housing market to reduce housing costs.The room-sharing type of housing, which involved remodeling existing houses in the early days of its introduction in South Korea, is gradually changing into a structure that is planned, designed, developed, and operated for shared housing [14].Although the supply of shared housing in South Korea is mainly carried out by the private sector, the government has recently expanded the supply of shared housing as part of its youth housing policy by recognizing shared housing supply as a national task and promoting projects related to shared housing in Seoul [15], with the shared housing market evolving as the purpose of residents is taking a step beyond home sharing to community relations and facility sharing.
Shared housing can be defined as a type of housing that allows several non-bloodrelated tenants to live in one house and share common spaces or facilities such as living rooms and kitchens [16].Such an arrangement has the advantage of reducing the burden of housing costs, improving the residential environment through sharing physical spaces, and forming social relationships, thus relieving loneliness and increasing the preference for such housing types among young people living in single-person households [6,17].According to the "2020 Population and Generation Status Analysis" conducted by the Ministry of the Interior and Safety, while it is expected to grow to 200,000 units and KRW 1.3 trillion by 2025, as of 2020, the market size of shared housing in South Korea is approximately 20,000 units, which is about 1% of rental households for young people.In the case of developed overseas markets such as the USA and Europe, the number of young singleperson households is expected to increase because of young people's desire to live in the city, network, and experience consumption, convenience, and flexibility during their employment periods [1,6,9].Additionally, while the supply of shared housing has increased rapidly since 2015, the exact concept and guidelines for shared housing are still insufficient under South Korea's domestic law.
As such, young single-person households in South Korea are becoming increasingly interested in and aware of shared housing as an alternative.
Research on shared housing in South Korea has focused on two main aspects-the type of shared housing, mainly the characteristics of space planning of shared housing that can be introduced in South Korea through case studies [18][19][20]; and the demand preferences of single-person youth households, which are the main potential demand group for shared housing [20,21].Moreover, to identify and derive reasonable and applicable standards for the supply of shared housing, a balanced approach to the demand and supply of shared housing is necessary; while many studies exist on space planning characteristics of shared housing and demand characteristics and preferences, few studies exist on residents' satisfaction with shared housing.Therefore, this study aimed to quantitatively evaluate housing satisfaction and factors affecting residents to address problems related to shared housing in South Korea.Using a housing satisfaction survey, the study identified the differences in consumer aspects, including state-led and private-sector-led supply types, which are representative suppliers of shared housing.The difference was then confirmed by comparing and analyzing the impact of housing satisfaction on the intention to reside again and the intention to recommend to others, and the basis for the public-and private-sectorled supply and revitalization of shared housing.This can enhance the understanding of the rapidly growing domestic shared housing market and highlight implications for the rational supply of shared housing as alternative housing in the future and for improvement in the living environment.This study would prove useful for policymakers and governments in their attempts to manage urban sprawl and private real estate markets, force the design of more inclusive and sustainable cities [20], and satisfy young people's entitlement to sustainable habitation.

Scope and Method of Research
This study was limited to the active shared housing market of Seoul.Shared housing is divided into state-led and private-sector-led markets based on supply and operation; both types were adopted as subjects of this study to identify their differences.The survey was limited to residents (19-39 years old) of shared housing in Seoul.Respondents in this age group were selected because they are subject to the housing policy for young people proposed by the Ministry of Land, Infrastructure, and Transport.

Literature Review
Shared housing is a type of housing with shared spaces for cooking, and separate private spaces for resting [22], provided in a moderate residential environment at a relatively low cost.The term "shared housing" or "collective living" is a commonly understood concept internationally and originated in the 19th century with boarding houses and urban transplants.With the concept of shared housing spans, a diverse range of accommodations including transitional houses, rooming and lodging houses; life cycle changes; income levels; and housing conditions [23], the commodification of shared housing continues to represent a lucrative real estate deal for investors, especially since affordable dwelling has become a priority with the rise of cities and urban sprawl.In areas throughout the U.K. and Europe, inner-city and shared housing developed in response to workers' and students' demand for flexible and affordable accommodation [3].Nowadays, although in this study, shared housing focuses on an implicit tenant demographic, i.e., young single people within public-led or private-led shared accommodations in the dense population city of Seoul, shared housing and all its formats reflect the global shift of "urban living and urban housing markets, characteristic of late-capitalist urbanization" [24].However, in this study, shared housing focuses on an implicit tenant demographic, i.e., young single people within public-led or private-led shared accommodations in the dense population city of Seoul.
Recently, shared housing has attracted the attention of young single people, albeit the concept of shared housing in academia and housing industries is sometimes mixed with community housing, co-housing, and collective housing.Byeon [25] defines shared housing as a generic term for housing types and housing methods whereby residents share residential facilities or spaces and classifies other sub-concepts according to the degree of physical and social sharing related to housing.Shared housing is not yet included in the housing type classification system under the Domestic Building Act or the Housing Act.However, the legal definition encompasses physical and social sharing as a "new type of housing where single-person households with common characteristics or interests gather to share some space such as kitchens and living rooms".
Currently, shared housing for single youth households is divided into public-led and private-led supplier types.Under the state-led supply type, capital, such as land and buildings, is initially provided by the public sector and then entrusted to the private sector to operate and manage.Recognizing the need for an alternative housing method amid high housing costs for young people, the government is increasingly attempting to introduce shared housing types in housing support projects for single-person youth households.In the case of state-led projects, low housing costs are an advantage, but there are limitations, such as operation management after supply.Further, with private-led shared housing being supplied in the form of construction or remodeling by a private business entity without state support, there is not much public land available in the city, which makes continuous provision difficult.In terms of location, there is a wide range of choices, and businesses can be expanded quickly, but the housing stability, including housing cost savings, is relatively reduced.Domestic shared housing is at a nascent stage because shared housing methods are still uncommon, despite research on shared housing for single-person households having been actively conducted in the past five years since the supply of shared housing in South Korea [11,14].However, models of alternative housing are being attempted by various suppliers in the private sector.Related previous studies have mainly focused on the characteristics of space planning through case analysis of shared housing at home and abroad and consumers' housing needs and preferences to inform shared housing planning [24][25][26].Regarding the spatial composition and operational characteristics of shared housing, Hee-won Lee et al. [22] and Min-ho Seong [26] analyzed domestic and foreign shared housing cases to derive the architectural planning characteristics of shared housing.Oh Jung et al. [16] analyzed the demand characteristics and influencing factors of shared housing for young single-person households, and Ji Eun-young [27] suggested ways to improve young single-person households' satisfaction with shared housing.Demand characteristics and preference surveys of potential demand groups in shared housing have been conducted, often suggesting the direction of planning for shared housing.Nevertheless, few studies have been conducted on actual shared housing residents.This study aimed to differentiate itself from existing studies by selecting residents living in shared housing in Seoul as study subjects, analyzing their housing satisfaction, and deriving factors affecting the performance of shared housing by supply type.
The concept of housing satisfaction is not limited to the physical aspects of the residence but also satisfaction with the overall housing, including the surrounding environment [28].Morris and Winter's conceptual and theoretical framework of housing adjusting behavior posit that unsatisfactory housing gives rise to dissatisfaction and may result in relocation or adaptation when housing does not meet expectations [29].Some twenty years earlier, Rossi posited that mobility was a housing adjustment but of a rudimentary type [30].While authors concurred that adjustment is a multi-decision process related to adjusting housing and a decision to select an alternative, it was Morris and Winter who identified the perceived gap between actual and desired housing as a source of dissatisfaction [29].
Where overall these theories highlight the shortcomings of these models in addressing this demographic based on the difference in structural variables, i.e., young, single, childless, and in rental accommodation, Speare also interpreted housing dissatisfaction as a predictor of housing adjustments but introduced the sociodemographic variable of age, contending that young people are more likely to move than old people if they are less satisfied with their housing [31].Marans and Rodgers [32] presented a housing satisfaction model that considers the subjective factors of individual residents.Each subjective factor corresponds to an individual's demographic characteristics and becomes a criterion for evaluating the residential environment of local communities, neighborhoods, and unit houses.The conceptual model of housing satisfaction for evaluating residential environment quality has been studied by many researchers.In particular, Fried and Gleicher [33] emphasized the importance of housing satisfaction, stating that among the various concepts and forms of housing, housing satisfaction is the most appropriate criterion for evaluating housing quality.Housing satisfaction is also the perceived difference between the resident's need or desire for housing and their current housing state [34], and the extent to which residents evaluate their current state and future needs.Previous studies mainly set several variables as factors influencing housing satisfaction and verified them through structural equations and regression analysis.According to Oh Soo-cheol [35], the influencing factors of housing satisfaction were mainly economic issues, social issues, education, stability, environment, and comfort.Recently, with the subject and range of spatial aspects of research related to housing satisfaction having been expanded to neighborhood and community characteristics [36,37], Kim Bu-sung [38] set and analyzed housing satisfaction factors as economic, cultural, environmental, educational, and social and urban infrastructural factors and concluded that they all, excluding cultural factors, affect housing satisfaction.In terms of traditional housing satisfaction, demographic characteristics, housing performance, housing support services, public facilities, neighborhood living facilities, and the social environment are also important factors.
The performance evaluation of shared housing can be examined based on the intention to reside again and the intention to recommend to others.The intention to recommend to others increases the willingness to settle in the current residence and leads to conveying one's housing satisfaction positively to others.Nam Young-woo [39] states that the frequency of movement decreases with higher housing satisfaction, and higher satisfaction leads to tenants' lease extensions.In a study on housing satisfaction, long-term residence intention, and recommendation intention for residents in new cities and downtown, Moon Hyun-seung [40] analyzed that housing satisfaction had a positive (+) effect on the longterm residence intention and recommendation intention.In summary, the higher the housing satisfaction, the higher the intention to continue living in the same residence and the lower the intention to move.However, with it being necessary to analyze the influence of housing satisfaction in shared housing on residents' intention to reside again and their recommendation intention, in terms of the relationship between housing satisfaction and recommendation intention, studies have yielded varying results.Notably, the intention to reside again may be low owing to the non-housing characteristics of young single-person households, such as switching jobs.

Summary
The concept of shared housing is still in the nascent stages in South Korea.While it has not been included within domestic laws, the state has become aware of the issues related to housing poverty for young singles.Accordingly, state-led agencies have been cooperating with the private sector by constructing these units and allowing them to operate and manage them.How much young single residents are satisfied with these units forms the basis of this structural equation modeling (SEM) quantitative research.Previous international and local studies have focused on space planning, and consumer needs and preference in housing satisfaction.Theoretical and conceptual studies have hypothesized housing adjustment models in response to unsatisfactory habitation and shown robust relations; however, the models are limited in their coverage of the structural variables of this unique population.Specifically, in research that addresses these limitations and adds to the literature on policy challenges in housing for this demographic, these theories occlude single young person households.Moreover, while economic, educational, social, and social characteristics have been used as performance evaluators for traditional housing satisfaction, these evaluators have not been used within the context of returning rental residents or for recommending rental units to others.The research addresses these limitations and adds to the literature on policy challenges in housing for this demographic.

Research Model
The structural equation model is a statistical analysis method that combines a latent variable model.It represents a causal relationship between latent variables and is a measurement model in which latent variables are measured by observation variables [41].The SEM is more reliable than regression analysis in this case because it considers measurement errors, and multiple independent variables can be analyzed simultaneously for dependent variables.Therefore, a single analysis can confirm causality and direct and indirect effects.Based on the theoretical relationship between the factors affecting housing satisfaction in previous studies, the following figure suggests that physical, economic, environmental, safety, community, and operation management factors of shared housing characteristics affect residents' housing satisfaction.Further, housing satisfaction can affect residents' intention to reside again and recommend to others.
Figure 1 is the measurement model of the study.The measurement model analysis illustrates whether indicators of the variables showed acceptable suitability using a single factor model before the hypothesis test of the proposed model and evaluated concentrated validity and discriminant validity.These are the most rigorous evaluation methods.Thirdlevel variables indicate measurement errors related to latent variables in the SEM using AMOS and do not mean differences between variables.This discussion is elaborated on further in Section 4.2.
Third-level variables indicate measurement errors related to latent variables in the SEM using AMOS and do not mean differences between variables.This discussion is elaborated on further in Section 4.2.

Research Model
The researchers used a purposive and snowball sampling technique.Both techniques were included to better estimate population parameters.Snowball sampling was particularly useful to expand the sampling target by introducing other respondents that meet specific conditions.The inclusion criteria for respondents were young mixed-gender persons between the ages of 20 and 39, unmarried, and residents in shared rental housing in Seoul.
A total of 620 sample data were collected through the survey of an on/offline questionnaire, but only 595 copies were valid.A pilot study was conducted before the main survey was conducted.After making amendments to the instruments based on feedback from the preliminary pilot, the survey was conducted over four weeks after revising and supplementing the questionnaire.
The study was conducted from 10 June to 20 July 2022.The sampling culminated in young people between the ages of 18 and 40 who were residents of or had resided in shared housing in Seoul.The 595 viable questionnaires collected through the online and offline surveys were reviewed by two expert researchers to eliminate duplicates.
The case study evaluation was considered an appropriate method to develop an indepth analysis of this phenomenon.The approach required detailed collection of a variety of data-collection procedures [42].Accordingly, the study involved three stages: first, based on the literature review, a theoretical background of the concept, the encompassing characteristics of shared housing, and single-person households' satisfaction with shared housing were established.An instrument was designed that included indicators that could measure variables such as housing satisfaction and intention to reside again and were modified for use on single-person shared housing residents (see Table 1).Second, the proposed research model was pilot tested by conducting a survey of residents of

Research Model
The researchers used a purposive and snowball sampling technique.Both techniques were included to better estimate population parameters.Snowball sampling was particularly useful to expand the sampling target by introducing other respondents that meet specific conditions.The inclusion criteria for respondents were young mixed-gender persons between the ages of 20 and 39, unmarried, and residents in shared rental housing in Seoul.
A total of 620 sample data were collected through the survey of an on/offline questionnaire, but only 595 copies were valid.A pilot study was conducted before the main survey was conducted.After making amendments to the instruments based on feedback from the preliminary pilot, the survey was conducted over four weeks after revising and supplementing the questionnaire.
The study was conducted from 10 June to 20 July 2022.The sampling culminated in young people between the ages of 18 and 40 who were residents of or had resided in shared housing in Seoul.The 595 viable questionnaires collected through the online and offline surveys were reviewed by two expert researchers to eliminate duplicates.
The case study evaluation was considered an appropriate method to develop an indepth analysis of this phenomenon.The approach required detailed collection of a variety of data-collection procedures [42].Accordingly, the study involved three stages: first, based on the literature review, a theoretical background of the concept, the encompassing characteristics of shared housing, and single-person households' satisfaction with shared housing were established.An instrument was designed that included indicators that could measure variables such as housing satisfaction and intention to reside again and were modified for use on single-person shared housing residents (see Table 1).Second, the proposed research model was pilot tested by conducting a survey of residents of shared housing using the selected indicators.This involved in-person visits to distribute to this end; confirmatory factor analysis, measurement model analysis, and a measurement model validity test were performed for each variable, and the research model was tested.Third, a multi-group analysis was conducted to test the differences in each parameter according to the type of supplier of shared housing for single-person households.By analyzing the structural impact relationship between housing satisfaction, intention to reside again in single-person shared housing, and recommendation to others, the study presents recommendations for improving the living environment of shared housing.All variables were measured using the 5-point Likert scale.Specifically, the characteristics of shared housing that affect the housing satisfaction and performance of single-person shared housing comprise seven factors: physical, economic, environmental, location, safety, community, and management characteristics.These were used as detailed factors of overall housing satisfaction, which in this study was determined as the overall satisfaction level of shared housing perceived by residents living in shared housing.Five questions were used to measure participants' housing satisfaction on a 5-point Likert scale (1 point: very dissatisfied; 3 points: neutral; 5 points: very satisfied).Additionally, the performance of shared housing was defined as the result of the overall evaluation of shared housing perceived by residents living in single-person shared housing.It was measured as the intention to reside again in shared housing after moving owing to reasons such as job changes and the intention to recommend shared housing to others.To measure this, a questionnaire used by Park Jong-hee [43] and Kwon Chil-Kwan [44] was modified.The questionnaire was based on a 5-point Likert scale and comprises eight items: four items related to the intention to reside again in shared housing and four items related to recommending shared housing to others.A summary of the instrument is provided in Table 1.

Establishment of Research Hypotheses
A modified research model was used to measure the factors influencing people's satisfaction with shared housing, their intention to reside again in shared housing, and their intention to recommend shared housing to others.
Hypothesis 1.The characteristics of shared housing for single-person households will have a significant effect on their housing satisfaction.
Hypothesis 2. The characteristics of shared housing for single-person households will have a significant effect on their intention to reside again in shared housing.Hypothesis 3. The characteristics of shared housing for single households will have a significant effect on their intention to recommend shared housing to others.Hypothesis 4. Housing satisfaction will have a significant effect on the intention of single-person households to reside again in shared housing.
Hypothesis 5. Housing satisfaction will have a significant effect on the intention of single-person households to recommend shared housing to others.
Furthermore, here, an indirect effect was estimated to test the mediating effect of housing satisfaction on the relationship between shared housing characteristics and the intention to reside again in shared housing, as well as the intention to recommend shared housing to others.Additional research hypotheses were established.Hypothesis 6.The characteristics of shared housing will have a significant effect on housing satisfaction and the intention to reside again in shared housing.
Hypothesis 7. The characteristics of shared housing will have a significant effect on housing satisfaction and the intention to recommend shared housing to others.Hypothesis 8.The differences between each parameter will depend on the provision of shared housing on the structural relationship among the characteristics of shared housing, housing satisfaction, intention to reside again in shared housing, and intention to recommend shared housing to others.

Descriptive Analysis
This study analyzed the structural influence relationship between satisfaction with shared housing, intention to reside again in shared housing, and intention to recommend shared housing to others among young people in their 20s and 30s living in shared housing.Study participants' sociodemographic characteristics are shown in Table 2. * Co-living provides the option of a private bathroom and kitchen; shared housing means sharing these facilities with flatmates.+Conversion rates are: KRW 1 = USD 0.00070-KRW = EUR 0.00071.

Analysis of Descriptive Statistics for Each Variable
The results of the descriptive statistics analysis of the variables used in this study are shown in Table 3. First, the results of the mean and standard deviation for each variable were as follows: characteristics of shared housing (M = 3.50, SD = 0.622), housing satisfaction (M = 3.73, SD = 0.607), intention to reside again (M = 3.55, SD = 0.669), and recommendation intention (M = 3.68, SD = 0.773).Further, the skewness and kurtosis values of all variables were less than the absolute value of descriptive statistics of 2, and all variables formed a normal distribution.Finally, the minimum value of each variable was distributed between 1.00 and 1.89, and the maximum values were all 5.00 points.

Measurement Model Analysis
Measurement model analysis was used to investigate whether indicators of the variables showed acceptable suitability using a single factor model before the hypothesis test of the proposed model and evaluate concentrated validity and discriminant validity.These are the most rigorous evaluation methods.As a result of the measurement model analysis, the indicators were found to have a very good fit level with sufficient acceptability.The specific results were as follows: χ2 = 338.280,df = 84, p = 0.000, GFI = 0.927, AGFI = 0.896, CFI = 0.900, TLI = 0.875, RMR = 0.055, RMSEA = 0.071.Moreover, looking at the standardized regression coefficient value of the variables used in the measurement model, all variables, excluding re-reside #1 (β = 0.443) and re-reside #3 (β = 0.420), showed a value of 0.5 or above, implying a strong explanatory power.Next, the construct reliability value was calculated to investigate convergent validity.The results were as follows: shared housing characteristics, 0.890; housing satisfaction, 0.762; intention to reside again 0.801; recommended intention, 0.730.All variables showed higher than the general standard (0.7 or higher), thus securing concentrated validity [45].
As a result of the measurement model analysis, the indicators were found to have a very good fit level with sufficient acceptability.Hypothesis 4, 'Housing satisfaction will have a significant effect on the intention to reside again in shared housing', can be found in Table 4.Moreover, housing satisfaction had a statistically significant effect on the intention to reside again (t = 6.435, p = 0.000) and intention to recommend (t = 5.135, p = 0.000); thus, Hypothesis 5 was accepted.This indicates that as the housing satisfaction increases, the intention to reside again in shared housing and the intention to recommend shared housing to others increase.The research model for this can be confirmed in Figure 2.
Sustainability 2022, 14, x FOR PEER REVIEW 13 of 21 It was found that shared housing characteristics had a statistically significant effect on housing satisfaction (t = 7.375, p = 0.000); therefore, Hypothesis 1 was accepted.However, in terms of the characteristics of shared housing, it was found that there was no statistically significant relationship between intention to reside again (t = 0.477, p = 0.633) and intention to recommend (t = −0.066,p = 0.947).This indicates that the higher the characteristics of shared housing, the greater the overall housing satisfaction.Moreover, housing satisfaction had a statistically significant effect on the intention to reside again (t = 6.435, p = 0.000) and intention to recommend (t = 5.135, p = 0.000); thus, Hypothesis 5 was accepted.This indicates that as housing satisfaction increases, the intention to reside again in shared housing and the intention to recommend shared housing to others increase.

Mediation Effect Test Results
In this study, the indirect effect was estimated to test the mediating effect of housing satisfaction in the relationship between the characteristics of shared housing and the intention to reside again and intention to recommend.The results are shown in Table 7.The indirect effect of shared housing characteristics on the intention to reside again through housing satisfaction was 0.659 (0.687 × 0.959).As the 95% confidence interval did not include zero, it appears to be statistically significant (p = 0.002); thus, Hypothesis 6 was accepted.Further, it was found that there was no statistically significant relationship between the characteristics of shared housing and the intention to reside again; therefore, it was found to have a complete mediating effect.This result indicates that as the detailed  Note: In the measurement model, the parameter estimate of the measurement variable was fixed at 1.A total of five questions were asked to measure the overall satisfaction with shared housing, and among them, 1, 3, and 5 were found to be appropriate.The performance of shared housing was measured by dividing it into residential intentions and recommended intentions.For this purpose, the questionnaire consisting of four items for residential intentions and four items for residential intentions is suitable.
The average variance extracted (AVE) value was used to test the discriminant validity.The method of evaluating discriminant validity using AVE values is said to have secured discriminant validity if the square values of all correlation coefficients between variables are lower than AVE values [46].The correlation analysis in Table 5 shows that the highest correlation coefficient was 0.612, which is between the intention to reside again and recommended intention.This squared value was 0.374, and it is said that discriminant validity is secured if all AVE values are greater than this value.The results show that the AVE values of all variables used in this study were 0.374 or higher, indicating that discriminant validity was also secured.

Causal Relationship Test Results
In this study, an SEM analysis was conducted using the maximum likelihood method to verify the set research model.The goodness-of-fit test results showed acceptable suitability, with χ2 = 349.715,df = 86, p = 0.000, GFI = 0.924, AGFI = 0.894, CFI = 0.896, TLI = 0.873, RMR = 0.056, RMSEA = 0.072.Further, as shown in Table 6, three out of five hypotheses were accepted in this study, as explained below.It was found that shared housing characteristics had a statistically significant effect on housing satisfaction (t = 7.375, p = 0.000); therefore, Hypothesis 1 was accepted.However, in terms of the characteristics of shared housing, it was found that there was no statistically significant relationship between intention to reside again (t = 0.477, p = 0.633) and intention to recommend (t = −0.066,p = 0.947).This indicates that the higher the characteristics of shared housing, the greater the overall housing satisfaction.Moreover, housing satisfaction had a statistically significant effect on the intention to reside again (t = 6.435, p = 0.000) and intention to recommend (t = 5.135, p = 0.000); thus, Hypothesis 5 was accepted.This indicates that as housing satisfaction increases, the intention to reside again in shared housing and the intention to recommend shared housing to others increase.

Mediation Effect Test Results
In this study, the indirect effect was estimated to test the mediating effect of housing satisfaction in the relationship between the characteristics of shared housing and the intention to reside again and intention to recommend.The results are shown in Table 7.The indirect effect of shared housing characteristics on the intention to reside again through housing satisfaction was 0.659 (0.687 × 0.959).As the 95% confidence interval did not include zero, it appears to be statistically significant (p = 0.002); thus, Hypothesis 6 was accepted.Further, it was found that there was no statistically significant relationship between the characteristics of shared housing and the intention to reside again; therefore, it was found to have a complete mediating effect.This result indicates that as the detailed influencing factors of shared housing characteristics increase, housing satisfaction increases, and the intention to reside again also increases.The indirect effect of the shared housing characteristic factor on the intention to recommend through housing satisfaction was 0.494 (0.687 × 0.719).It was statistically significant because it did not include 0 in the 95% confidence interval (p = 0.003); therefore, Hypothesis 7 was accepted.Additionally, there was no statistically significant relationship between the characteristics of shared housing and the intention to recommend it; thus, it had a complete mediating effect.This result indicates that as the detailed influencing factors of the shared housing characteristics increase, housing satisfaction increases, and the intention to recommend to others increases.

Verification of Measurement Model Identity
In this study, a multi-group analysis using SEM was conducted to determine whether the parameter between the latent variables suggested in the research model showed a significant difference according to the supply type of single-person households-publicand private-supply.The identity test of the measurement model was performed first, as shown in Table 8.The unconstrained model refers to a value without inter-group constraints, and Model 1 is a value that equally constrains the factor load between groups.Model 2 is a value that constrains factor load and covariance equally between groups, and Model 3 is a value that constrains factor load, covariance, and error variance equally.The x 2 difference between the non-pharmaceutical model and Model 1 was 13.844 (461.632-447.788).When the degree of freedom was 11 (179-168) at the level of 0.05, the x 2 value was 19.675.In other words, as it was found to be smaller than the x 2 value corresponding to the statistical threshold, there was no problem with the equality of the factor load for the measurement tool.Additionally, in terms of the model fit of the CFI, TLI, and RMSEA indices, there was no difference in all models because the fit index of the model was almost similar [47].In other words, the cross-validity was tested, and the scale used in this study was recognized equally by those living in public-supply housing and those in private-supply housing.

Analysis Result of Multi-Group Model by Supply Type
Table 9 shows the test results for Hypothesis 8, which proposed that the relationships among the parameters will be different depending on the supply type of single-person household shared housing (whether public-or private-supply housing) regarding the structural relationship between shared housing characteristics, housing satisfaction, intention to reside again, and intention to recommend to others.After separating people living in public-supply housing from those living in privatesupply housing, a constrained model and an unconstrained model were established.Here, the unconstrained model was identified as a model superior to the constrained model, judged by the difference in χ2 value.In other words, if the difference in χ2 value based on the degree of freedom was above the threshold, it was interpreted that there is a difference between models.Here, the difference in the parameter between "public-supply" and "private-supply" housing was significant.
The χ2 value difference between the constrained model and the unconstrained model was 12.779, and the difference in degree of freedom was 5. When the difference in degrees of freedom is 5, the χ2 threshold of significance level 0.05 is 11.071, which is larger than the value in this study (p = 0.025).Thus, the difference in parameters between groups can be interpreted as being significant.Therefore, Hypothesis 8 was accepted.
Differences between parameter (DBP) values are used to interpret the differences between parameters.If this value is greater than the absolute value of 1.96, it means that the parameter differs between groups.The analysis showed that there was a difference in the parameter between shared housing characteristics and housing satisfaction (DBP = 2.075) and between housing satisfaction and the intention to reside again (DBP = 2.771).
Specifically, there was no significant relationship between shared housing characteristic factors and housing satisfaction (β = 0.251) for people living in public-supply housing.It was found that the housing satisfaction of people living in private-supply housing also increased when the detailed influencing factors of shared housing characteristics were high (β = 0.664).Further, while the housing satisfaction of people living in public-supply housing had no statistical effect on their intention to reside again (β = 0.441), the housing satisfaction of people living in private-supply housing had a significant effect on their intention to reside again (β = 0.877).
The above results indicate that when comparing private-supply single-person households with public-supply single-person households, higher-detailed influencing factors of shared housing characteristics increase the housing satisfaction and intention to reside again.

Discussion
This study was conducted to identify the relationship between young, single-person households' satisfaction with shared housing, the intention to reside again in shared housing, and the intention to recommend shared housing.The results should be useful to provide recommendations for policy establishment for the supply and revitalization of shared housing in South Korea.Specifically, the study examined the effect of the characteristics of shared housing on housing satisfaction and shared housing performance given the efforts to reduce housing costs and improve quality following the increase in single-person households.Based on the theoretical relationships among the factors affecting housing satisfaction, the study hypothesized that physical, economic, environmental, safety, community, and operation management factors, which are sub-factors of housing satisfaction, affect residents' overall housing satisfaction.It also hypothesized that housing satisfaction affects residents' intention to reside again in shared housing and their intention to recommend shared housing to others.The main results of this study are summarized as follows.

Statistical Analysis
First, this study, which was conducted using a survey analysis method, showed that the standardized regression coefficient values of the variables used in the measurement model, excluding some variables on "intention to reside again," had strong explanation power, with a value of 0.5 or higher.Further, the calculation of construct reliability yielded the following results: shared housing characteristics, 0.890; housing satisfaction, 0.762; intention to reside again, 0.801; and intention to recommend to others, 0.730.Therefore, concentrated validity was achieved given that all variables were higher than the general standard (0.7 or more).This study appeared to be similar to the work by Woo et al. [48], who investigated a similar population in South Korea to ascertain their preference for shared housing.However, their study differed in that they used a binary logistic regression model to analyze multiple determinants such as sociodemographic, personal, accommodation, and locational characteristics with a larger population (1000).More importantly, both studies were able to review these preferences across the city and diverse neighborhoods and concur on the stratification of rental throughout the city.The study advocates the need for policy to support young people in their choice of shared living arrangements.

Shared Housing Satisfaction
Second, the detailed influencing factors of shared housing characteristics were found to have a statistically significant effect on housing satisfaction.Satisfaction could be attributed to interpersonal relationships in shared accommodations, as Milić and Zhou found on the residential satisfaction of young people in Serbia [10], as did Green and McCarthy and Bricoli and Sabatinelli [4,7].However, the characteristics of shared housing were found to have no direct significant influence on the intention to reside again and the intention to recommend to others.James and colleagues found that the operations factors were important in satisfaction but mainly in the private-led market [49].Various physical, emotional, social, and convenient factors derived from the shared housing characteristics of young, single-person households can affect the housing satisfaction of residents [50].Nevertheless, it is considered that they have no statistically significant relationship between the intention to reside again and the intention to recommend to others, given that intention to reside again may decrease because of housing costs and a change in jobs.This supports the housing adjustment theories submitted by Morris and Winter, Rossi, and others that each unit had its own threshold for dissatisfaction [29][30][31].Therefore, it suggests that while respondents may have experienced some levels of discomfort, it was not enough to trigger a major housing adjustment such as relocating.Moreover, housing satisfaction had a statistically meaningful effect on the intention to reside again and the intention to recommend to others, indicating that as housing satisfaction increases, the intention to reside again and the intention to recommend to others also increase.Housing satisfaction was found to have a positive effect on both the intentions to reside again and intention to recommend to others.These characteristics indirectly suggest that the overall satisfaction level of shared housing residents of single-person households is high.

Supply and Demand in the Private-Public Sector Housing Market
Third, this study examined whether there are differences in the relationships among the parameters depending on the type of supply (public or private), which are the main suppliers of single-person housing, given the structural relationship between the characteristics of shared housing, housing satisfaction, the intention to reside again, and the intention to recommend to others.The results showed that residents of shared housing supplied by the state did not show a significant correlation between the detailed influencing factors of shared housing characteristics and housing satisfaction.However, the residents of shared housing supplied by the private sector showed increased housing satisfaction depending on the degree of positive evaluation of the detailed influencing factors of shared housing characteristics.This supports research conducted by Khozaei et al. that demonstrated students' attachment and preference for their hall when they were satisfied with the accommodations [51].In China, Li and others reported higher satisfaction among private tenants in China's urban cities still depended on a tier system [52].Further, the housing satisfaction of people living in public-supply shared housing had no statistical effect on their intention to reside again, but that of those living in private-supply shared housing had a significant effect on their intention to reside again in shared housing.According to Najib et al., this could be supported by social relationships within the co-living arrangement [53].

Recommendations Arising from the Research
Previous studies defined potential demand groups and researched young people in general, but this study is meaningful as it collected research data and analyzed experiences based on young people currently living in shared housing.Discussions on shared housing projects and policies for youth housing support are actively underway in Seoul, as evidenced by the government's recently announced "Housing Support Roadmap," which plans to supply shared housing for single young people.With supporting policies and systems in place, it is essential to increase the quantitative supply through the diversification of supply types to cater to diverse preferences.Accordingly, one recommendation is that governments consider creating and maintaining a pleasant living environment in shared housing and not focus only on economic factors.Therefore, policymakers, architects, and developers must consider designs and concepts and take into account preferences for living arrangements.These should go beyond the construction of multi-dwellings to preserve land space but include a pragmatic urban growth management strategy [5,12,52].
There is still a lack of domestic awareness of shared housing.Therefore, marketing campaigns should be developed to alert young people to the available options [13,16,18,48,54].
While the supply of shared housing is planned as a way to solve the youth housing problem, appropriate architectural design and operational guidelines have not been put in place.Therefore, the results of this study can be helpful as basic data in preparing guidelines for each supply type that reflect consumer characteristics and for informing the directions for the supply [12,18].
There is a need for state policy to regulate the provision of shared housing to ensure minimum standards are accessible to renters.While this may be difficult given the housing poverty in South Korea, state agencies can be established to assist landlords willing to conform.To this end, state policy should incorporate all dwelling types, choices, and mobility and strengthen institutional frameworks to develop and manage human settlements [12,16,55].
The preference for private-led accommodation was evident in this study; perhaps state-run agencies may consider removing barriers to encourage more private-sector entities to invest in shared housing [12,56,57].
To help with sustainable development, it might be useful to encourage the decentralization of high-paying jobs and population relocation to minor cities to assist with the efficient management of resources [58].

Limitations
This study had several limitations.The first is the regional limitation as the research was not conducted nationwide but concentrated in Seoul.However, Seoul has the largest market of shared housing and was selected as a representative area.A sample survey was conducted to derive general rather than regional characteristics.Further, the characteristics of shared housing and housing satisfaction were examined according to the operating entity, but the effect of individual preferences on housing satisfaction was not systematically analyzed.Although factors and motivations for moving into shared housing can affect residents' expectations and satisfaction in terms of space and social aspects, the effect of these factors on housing satisfaction was not analyzed.In future studies, qualitative research methods, including interviews and in-depth studies on the effect of individual characteristic factors and motivation for moving into shared housing on housing satisfaction and performance, should be conducted.In this way, a better understanding of shared housing will be possible, with detailed housing-related problems and needs identified and addressed through an enhanced qualitative study of housing and business types.Finally, the study neglected to survey young persons in shared housing in illegal or squatter housing, despite the prevalence of these sub-standard dwellings within city limits.These conditions, however, confirm an emerging market segment for this demographic.

Conclusions
Housing for young single persons remains a significant issue in South Korea.Lowincome earners coupled with the cost of high-density developments that fail to provide sufficient infrastructure for housing satisfaction are also problematic [12,14].More needs to be carried out to create partnerships with private developers and landlords as residents of private-supply shared housing appeared to have higher housing satisfaction [35,40,51].Their satisfaction depended on the degree of positive evaluation of detailed influencing factors of shared housing characteristics compared to public-supply single-person households and a higher intention to reside again.In the case of public-supply shared housing, cheap deposits and rents, which are economic characteristics, are the main drivers.These are advantages in terms of housing costs compared to private-supply housing but depend on factors such as income standards and government subsidies.This suggests that detailed influencing factors other than economic factors, such as physical, location, environment, and community factors, had no significant effect on overall housing satisfaction.Nevertheless, in the case of private-supply shared housing, the detailed influencing factors of shared housing characteristics had a significant effect on housing satisfaction and intention to reside again [19].This shows that apart from housing cost reduction, which is an important advantage of shared housing, other factors, including facility sharing and reduced social isolation, influence judgment when looking at the housing problem among young, single-person households.Clearly, shared housing can be recognized from an economic standpoint and physical and emotional quality, suggesting that living quality should be considered when supplying shared housing.Therefore, when supplying private-led shared housing, it is necessary to increase the housing satisfaction by raising the quality of living to the level of consumers' expectations, which can lead to the revitalization of shared housing and housing stability.

Figure 1 .
Figure 1.Measurement model of the study.

Figure 1 .
Figure 1.Measurement model of the study.

Table 1 .
Summary of survey questions.

Table 3 .
Technical statistics analysis results.

Table 7 .
Bootstrapping results of indirect effects.

Table 8 .
Measurement model identity test result.

Table 9 .
Multi-group analysis results by shared housing operator.