1# RECORD 757bd683e867096bd3de6214d8e05c02d711dbcc2# TITLE: Effects of Educational and Cultural Facilities on Housing Prices in Seoul from an Accessibility Perspective3# YEAR: 2021.0 VENUE: Journal of people, plants, and environment DOI: 10.11628/ksppe.2021.24.5.5294# AUTHORS: Minki Sung; Junghoon Ki5# PAGES: 16 CANDIDATE TABLE PAGES (1-based): [6, 9, 10, 11]6# ABSTRACT: Background and objective: A great deal of previous research has highlighted the value of educational and cultural facilities embedded in housing prices, by taking a large spatial area as the focus, such as the city or district level. However, few studies have investigated the extent to which educational and cultural facilities influence the formation of housing prices from an accessibility perspective. This study aims to identify the value of educational and cultural facilities embedded in the housing prices in Seoul Metropolitan City with a focus on the concept of the residents’ neighbourhood and accessibility. Methods: To this end, this research used a spatial regression model with educational and cultural facilities as the independent variables and housing prices as the dependent variable. The model assessed the accessibility of cultural and educational facilities by considering geographic effects. Results: The findings are as follows. First, the spatial error model was found to be the best fit for multi-unit housing, while the spatial lag model was more appropriate for single-unit housing and apartments. Second, private educational facilities and art museums had positive effects on single- and multi-unit housing prices, while historical sites had a negative effect. Finally, private educational facilities positively influenced apartment prices, whereas public libraries and urban park areas had a negative effect. Conclusion: These findings indicate that the accessibility of educational and cultural facilities reflects residents’ preferences and needs, which will ultimately influence housing prices.789===== PAGE 1 (score 12, numeric-paren 0, table-words 1, var-words 10) =====1011Minki Sung and Junghoon Ki12Journal of People, Plants, and Environment Vol. 24, No. 5, 2021∙52913J. People Plants Environ. Vol. 24 No. 5: 529-544, October 202114https://doi.org/10.11628/ksppe.2021.24.5.52915JPPE16Effects of Educational and Cultural Facilities on Housing Prices in 17Seoul from an Accessibility Perspective18Minki Sung1 and Junghoon Ki2*191Postgraduate student, Department of Archaeology, Durham University, Durham DH1 3LE, United Kingdom202Professor, Department of Public Administration, School of Social Science, Myongji University, Seoul 03674, South Korea21ABSTRACT22Background and objective: A great deal of previous research has highlighted the value of educational and cultural facilities23embedded in housing prices, by taking a large spatial area as the focus, such as the city or district level. However, few 24studies have investigated the extent to which educational and cultural facilities influence the formation of housing prices 25from an accessibility perspective. This study aims to identify the value of educational and cultural facilities embedded in26the housing prices in Seoul Metropolitan City with a focus on the concept of the residents’ neighbourhood and accessibility.27Methods: To this end, this research used a spatial regression model with educational and cultural facilities as the 28independent variables and housing prices as the dependent variable. The model assessed the accessibility of cultural and 29educational facilities by considering geographic effects. 30Results: The findings are as follows. First, the spatial error model was found to be the best fit for multi-unit housing, while31the spatial lag model was more appropriate for single-unit housing and apartments. Second, private educational facilities32and art museums had positive effects on single- and multi-unit housing prices, while historical sites had a negative effect.33Finally, private educational facilities positively influenced apartment prices, whereas public libraries and urban park areas34had a negative effect. 35Conclusion: These findings indicate that the accessibility of educational and cultural facilities reflects residents’ 36preferences and needs, which will ultimately influence housing prices.37Keywords: neighbourhood; public facilities; residents’ preference; spatial regression model38This work was supported by the National Research Foundation of Korea [grant number: NRF-2017R1D1A1B03032415]39Received: July 6, 2021, Revised: August 10, 2021, Accepted: September 13, 202140First author: Minki Sung, mingi5678@nate.com, 41 https://orcid.org/0000-0001-6665-938342*Corresponding author: Junghoon Ki, johnki70@gmail.com, 43 https://orcid.org/0000-0001-7554-957344Ⓒ 2021 by the Society for People, Plants, and Environment. This is an Open Access article distributed under the terms of the Creative Commons 45Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, 46and reproduction in any medium, provided the original work is properly cited.47Introduction48How is housing appraised? Are the physical character-49istics of housing the only indicators of its value? While these 50have been ongoing questions in housing research, many re-51searchers have also discussed housing prices per se or in 52relation to locational characteristics (Dong et al., 2019; Li 53et al., 2019; Wittowsky et al., 2020). In reality, housing prices 54reflect not only the physical characteristics of a house itself 55but also the various elements of its location and surrounding 56environment (Adair et al., 2000; Dong et al., 2019; Lan et al., 572018; Li et al., 2019; Zahirovic-Herbert and Gibler, 2020). 58In particular, the neighbourhood plays an important role 59(Bramley et al. 2008; Lan et al., 2018; Wittowsky et al., 2020), 60in that housing prices vary significantly among different sub-61markets segmented according to accessibility (Adair et al., 2000; 62Feng et al., 2020; Ki and Sung, 2020; Sung and Ki, 2021).63Previous Seoul-based case studies, which mainly focused 64on how urban public facilities affect housing prices over a 65large area, such as a residential district (Jin and Sohn, 2005; 66Kim, 2012; Kim and Lee 2007; Lee, 2010), have primarily 67addressed the extent to which access to educational facilities 68influences housing prices. Further, recent studies on cultural 69facilities and green spaces have explicitly demonstrated that 707172===== PAGE 6 (score 14, numeric-paren 3, table-words 3, var-words 5) =====7374Effects of Educational and Cultural Facilities on Housing Prices in Seoul from an Accessibility Perspective75534∙Journal of People, Plants, and Environment Vol. 24, No. 5, 202176where Y denotes the housing price, x is a matrix of observations 77on the explanatory variables, and ρwy is the spatial lag (or 78error) variable. ρ assesses the degree of spatial correlation and 79w is the spatial weights matrix. The value of ρ ranges be-80tween -1 and 1, with ρ = 0 signifying no spatial correlation. 81Under a positive ρ, housing prices are expected to be higher if 82spatially adjacent housing prices are higher, while ε is a vector 83of the independent and equivalently distributed error terms 84(Ward and Gleditsch, 2008).85Spatial regression models include SLM and SEM, which 86use spatial regression dependencies and spatial regression 87errors, respectively. We determined which model was more 88suitable by comparing them with the results of the OLS model. 89The OLS model was tested for suitability under the assumptions 90of normality, equal variability, and spatial independence of 91errors (Park and Ham, 2018). Similarly, the spatial depend-92encies of dependent variables and regression model error 93terms were examined using the Lagrange multiplier (LM) lag 94and error. If the LM lag and error are meaningful at the 5% sig-95nificance level, the null hypothesis of the spatial independence 96of dependent variables and error terms is rejected. If only one 97of either the LM lag or error is statistically significant, selecting 98a more adequate model should be considered. If both values 99are significant, the robust LM lag and robust LM error statistics 100test functions are used for reselection, as a last resort (Park 101and Ham 2018). Park and Ham (2018) relied on the log-like-102lihood, Akaike information criterion (AIC), and Schwarz 103criterion (SC) results to determine the suitability of spatial 104regression models. Compared to the results of the OLS mod-105el, the model with a higher log-likelihood and lower AIC 106and SC was considered appropriate.107To determine the effects of educational and cultural factors, 108we used the number of public libraries and private educational 109facilities as a proxy for the educational factor (independent 110variable). Regarding the cultural factor, art museums, museums, 111historical sites, and urban parks are used as independent 112variables. Table 3 and Table 4 present detailed statistical 113information for the independent and dependent variables.114Results and Discussion115As shown in Fig. 3, we illustrated housing prices and 116their independent variables using ArcGIS 10.8. On average, 117the number of private educational facilities was large (31.65) 118and showed a high standard deviation (17.74) compared 119to other facilities. Fig. 3 also shows that private educational 120facilities tend to be clustered in the south-eastern part of 121Seoul, where housing prices are generally higher regardless of 122the housing type. However, public libraries had a relatively 123low standard deviation (1.26), implying that a similar number 124of public libraries are distributed across districts, as can 125also be seen in Fig. 3.126As for cultural facilities, art museums, museums, and 127Fig. 2. Global and local Moran’s I analysis results; MULTI: multi-unit housing, SINGLE: single-unit housing, APT: apartment.128129130===== PAGE 7 (score 13, numeric-paren 0, table-words 2, var-words 8) =====131132Minki Sung and Junghoon Ki133Journal of People, Plants, and Environment Vol. 24, No. 5, 2021∙535134Variable135Dong (sub-district)136Average137Standard deviation138Max139Min140Dependent 141variable142Price by housing type143Multi-unit housing144249688516.3145155834809.21461100000000147104363359.5148Single-unit housing149831456068.7150617399014.61514288716364152190000000153Apartment154517997051155290439973.61561956893793157183663265.3158Independent 159variables160Educational factor161Public library1621.431631.26164201650166Private educational 167facility16831.6516917.741701951710172Cultural factor173Art museum1740.111750.5717671770178Museum1790.261801.14181111820183Historical site1840.351853.16186491870188Number of urban parks1896.051902.53191271920193Urban park area1940.06 km2 (4%)1950.9719614.55 km2 (80%)1970.00003 km2 (0.01%)198Note: All independent variables are expressed per 10,000 population. Urban park areas are expressed as a proportion of the total sub-district (dong) area. The 199unit of housing price is Korean won.200Table 4. Descriptive statistics of independent and dependent variables201Variables202Definition203Dependent 204variable205Price by 206housing type207Multi-unit housing208Real estate transaction 209price210Housing types are classified as follows: single-unit 211housing, multi-unit housing, or apartment. Single- and 212multi-unit housing tends to be occupied by less affluent 213families, whereas relatively affluent families reside in 214apartments (Kang et al., 2019). More than half of the 215population (50.7%) in Seoul resides in apartments (Kang 216et al., 2019). Furthermore, Jeon (2015) suggested that 217most residents in Korea prefer apartments because an 218apartment complex that has access to educational and 219cultural facilities is more attractive than single- or 220multi-unit dwellings when raising children.221Single-unit housing222Apartment223Independent 224variables225Educational 226factors227Public library228Number of public libraries 229in Seoul230Public library means a library established and operated by 231local government for information, reading activities, and 232cultural activities by the public. The libraries we 233employed as an independent variable are local public 234libraries, not university or special libraries, specially built 235for residents.236Private educational facility237Number of registered 238private academies and 239study centres in Seoul240Private educational facilities in this study are categorized 241into two types: those with curricula that facilitate entrance 242into prestigious universities and those providing lifelong 243or vocational education.244Cultural factors245Art museum246Number of art museums in 247Seoul248The reason for distinguishing museums from art museums 249is that the Museum and Art Gallery Support Act of South 250Korea (Government of the Republic of Korea, 2018b) 251considers them to be different.252Museum253Number of museums in 254Seoul255Historical site256Number of historical sites 257in Seoul258In this study, historical sites cover every historical 259property registered by central and local governments. In 260South Korea, historical sites can be categorized into three 261types: nationally assigned cultural properties, city 262assigned properties, and nationally registered properties.263Number of urban parks264Number of urban parks in 265Seoul266The park concept we employed in this study includes 267parks established and managed within the urban residents’ 268immediate neighbourhood structure; it includes 269small-sized, children’s, and neighbourhood parks.270Urban park area271Urban park areas in Seoul272Table 3. Definitions of independent and dependent variables273274275===== PAGE 9 (score 34, numeric-paren 4, table-words 10, var-words 6) =====276277Minki Sung and Junghoon Ki278Journal of People, Plants, and Environment Vol. 24, No. 5, 2021∙537279Table 7 presents the analysis results of the OLS model, 280SLM, and SEM using single-unit housing price as the de-281pendent variable. The Jarque-Bera test statistic of 1286.0811 282(p < .05) signifies an abnormality in the spatial autocorrelation 283test of error terms. Therefore, this test leads to spatial re-284gression models instead of the OLS model. Using the results of 285the spatial regression model analysis, we selected the more suit-286able model between SLM and SEM. The R2 value (0.303476) 287of the SLM is higher than that of the SEM, and is also more 288robust. Additionally, the log-likelihood value of the SLM 289(-9,084.95) is higher and their AIC and SC are lower than 290those of the SEM. Furthermore, the likelihood ratio test of 291the SLM generates statistically significant results, indicating 292it is more suitable than the SEM.293For this reason, the SLM was considered the most suitable 294model based on the analysis results. The SLM results reveal 295that educational (private educational facilities) and cultural 296(art museums) factors exert positive influences on single- 297unit housing prices. However, historical sites adversely affect 298single-unit housing prices.299Table 8 shows the analysis results of the OLS model, SLM, 300and SEM using apartment price as the dependent variable. 301The Jarque-Bera test statistic of 1,008.7673 (p < .05) signifies 302an abnormality in the spatial autocorrelation test of the error 303term. Moreover, the Breusch-Pagan test statistic (26.1987, 304p < .05) indicates heteroscedasticity in the error term. Therefore, 305spatial regression models are preferrable to the OLS model.306The spatial regression model analysis allowed the most 307suitable model between the SLM and SEM to be selected. 308The R2 value (0.171330) of the SLM was more robust and 309higher than that of the SEM. Moreover, the log-likelihood 310value of the SLM (-8,803.23) was higher and the SC lower 311Dependent variable (multi-unit housing)312OLS313SLM314SEM315Constant3162.53016e + 008***3172.30529e + 008***3182.52693e + 008***319Educational factor320Public library321-1.14233e + 007322-1.18353e + 007323-1.21706e + 007324Private educational facility325548281**326546127**327544701**328Cultural factor329Art museum3308.81226e + 007***3318.72462e + 007***3328.69073e + 007***333Museum3344.83881e + 0063356.14577e + 0063366.9372e + 006337Historical site338-1.37727e + 007***339-1.39005e + 007***340-1.39798e + 007***341Number of urban parks342-1.28692e + 006343-1.21817e + 006344-1.16495e + 006345Urban park area346-0.996065347-0.957873348-0.914984349Rho3500.0890597351Lambda3520.092414353R23540.0811623550.0855083560.085678357Adj. R23580.065664359Log-likelihood360-8561.41361-8560.71362-8560.70363Akaike information criterion36417138.836517139.436617137.4367Schwarz criterion36817171.236917175.937017169.8371Koenker-Bassett3723.5660373Breusch-Pagan37419.1758**37518.838237618.8002377White test37860.6184**379Moran’s I3801.3146381Lagrange multiplier (lag)3821.4785383Robust LM (lag)3840.0085385Lagrange multiplier (error)3861.4964387Robust LM (error)3880.0265389Likelihood ratio test3900.237623911.4261392Note:* p < .05, ** p < .01, *** p < .001. OLS: ordinary least squares; SLM: spatial lag model; SEM: spatial error model; LM: Lagrange multiplier.393Table 6. Analysis of multi-unit housing394395396===== PAGE 10 (score 34, numeric-paren 7, table-words 7, var-words 7) =====397398Effects of Educational and Cultural Facilities on Housing Prices in Seoul from an Accessibility Perspective399538∙Journal of People, Plants, and Environment Vol. 24, No. 5, 2021400than that of the SEM. For this reason, the SLM was re-401garded to be the most suitable based on the analysis results. 402The results reveal that the educational factor exerted both 403positive (private educational facilities) and negative (public 404libraries) influences on apartment prices. This finding is 405important, because while public libraries were statistically 406significant in the spatial regression model, they were insig-407nificant in the OLS one. This indicates that spatial re-408gression analysis considers spatial effects and must be ap-409plied to measure the spillover effects of sites with a specific 410space.411Dependent variable (single-unit housing)412OLS413SLM414SEM415Constant4166.52325e + 008***4175.36521e + 008***4186.5393e + 008419Educational factors420Public library421-2.79845e + 007422-2.76873e + 007423-2.86462e + 007424Private educational facility4252.79395e + 008***4267.4858e + 006***4277.5038e + 006***428Cultural factors429Art museum430258232550.280***4312.68802e + 008***4322.68357e + 008***433Museum4346.32358e + 0074356.68108e + 0074366.65248e + 007437Historical site438-3.71002e + 007***439-3.78481e + 007**440-3.84802e + 007**441Number of urban parks442-6.80577e + 006443-6.92999e + 006444-6.8878e + 006445Urban park area446-6.46759447-5.90286448-5.8235449Rho4500.0890597451Lambda4520.10986453R24540.2938184550.3034764560.298736457Adj. R24580.281906459Log-likelihood460-9087.08461-9084.95462-9086.07463Akaike information criterion46418190.246518187.946618188.1467Schwarz criterion46818222.546918224.347018220.5471Koenker-Bassett4722.5375473Breusch-Pagan47412.465147511.86374762.0222477White test47841.6741479Moran’s I4801.5495481Lagrange multiplier (lag)4824.6786*483Robust LM (lag)4843.9209*485Lagrange multiplier (error)4862.1215487Robust LM (error)4881.3638489Likelihood ratio test4904.2727*4912.0222492Note: * p < .05, ** p < .01, *** p < .001. OLS: ordinary least squares; SLM: spatial lag model; SEM: spatial error model; LM: Lagrange multiplier.493Table 7. Analysis of single-unit housing494Dependent variable (apartment)495OLS496SLM497SEM498Constant4994.76082e + 008***5004.84887e + 008***5014.76125e + 008***502Educational factors503Public library504-2.39781e + 007505-2.39659e + 007*506-2.39323e + 007*507Private educational facility5082.5649e + 006***5092.56447e + 006***5102.5631e + 006***511Cultural factors512Art museum5134.78958e + 0075144.83697e + 0075154.81688e + 007516Museum5173.89872e + 0075183.87268e + 0075193.88224e + 007520Historical site521-5.82057e + 006522-5.82256e + 006523-5.81122e + 006524Number of urban parks525-678318526-668196527-678375528Urban park area529-6.0852**530-6.10283**531-6.09184**532Table 8. Analysis of apartments533534535===== PAGE 11 (score 23, numeric-paren 3, table-words 6, var-words 6) =====536537Minki Sung and Junghoon Ki538Journal of People, Plants, and Environment Vol. 24, No. 5, 2021∙539539Conclusion540We analysed Seoul Metropolitan City’s dong-unit data-541base and obtained the following results. First, while the 542SEM is suitable for analysing multi-unit housing prices, 543the SLM is suitable for single-unit housing and apartment 544prices. Second, private educational facilities and art museums 545have positive influences on most housing prices, while historical 546sites have a negative influence. Finally, public libraries and urban 547park areas have negative effects on apartment prices. Thus, this 548study can imply that historical sites, public libraries, and urban 549park areas can be interpreted as less influential factors in 550housing prices compared to private educational facilities 551and art museums.552Table 9 presents a comparison of our research results with 553those of Ki and Sung (2020). Both studies adopted the same 554independent variables but at different spatial scales: the res-555idents’ immediate neighbourhood (dong) and the residential 556district (gu), respectively. The test of the SLM is more robust 557in terms of residents’ immediate neighbourhood (dong) analysis, 558especially between single-unit housing and apartments, sug-559gesting that the most proximate public facilities can be con-560sidered a relevant factor in explaining residential housing 561prices, which is also consistent with the results of Wittowsky 562et al. (2020).563Based on this comparison, we suggest the following policy 564implications. First, private educational facilities positively 565influence all types of housing at the gu and dong levels, 566which is a finding consistent with most of the previous studies 567(Cho et al., 2020; Ha and Lee, 2013; Kim and Kim, 2019). 568The residents of Seoul closely associate private educational 569facilities with students’ academic performance and consider 570them as essential educational facilities (Ha and Lee, 2013). 571Consequently, parents of students tend to relocate close to 572private educational facility clusters (Park and Lee, 2011) to 573allow easy access to education for their children, thereby 574positively influencing housing prices in Seoul at both the 575gu and dong levels.576Second, art and general museums have positive effects on 577the prices of all housing types in a residential district (gu). 578Art museums have a positive effect on single- and multi-unit 579housing prices in residents’ immediate neighbourhood (dong). 580In particular, art museums positively affect social and edu-581cational environmental satisfaction, and in turn, residential 582environmental satisfaction (Ki and Sung, 2020). For example, 583a recent study (Han, 2020) suggested that art museums have 584Dependent variable (apartment)585OLS586SLM587SEM588Rho589-0.0169392590Lambda591-0.006901592R25930.1711845940.1713305950.171206596Adj. R25970.157204598Log-likelihood599-8803.26600-8803.23601-8803.253223602Akaike information criterion60317622.560417624.560517622.5606Schwarz criterion60717654.960817660.960917654.9610Koenker-Bassett6116.2613612Breusch-Pagan61326.1987***61426.0083***61526.1088***616White test61736.1292618Moran’s I619-0.0090620Lagrange multiplier (lag)6210.0560622Robust LM (lag)6230.2264624Lagrange multiplier (error)6250.0083626Robust LM (error)6270.1787628Likelihood ratio test6290.05286300.0079631Note: * p < .05, ** p < .01, *** p < .001. OLS: ordinary least squares; SLM: spatial lag model; SEM: spatial error model; LM: Lagrange multiplier.632Table 8. Continued633634635===== PAGE 12 (score 15, numeric-paren 0, table-words 1, var-words 13) =====636637Effects of Educational and Cultural Facilities on Housing Prices in Seoul from an Accessibility Perspective638540∙Journal of People, Plants, and Environment Vol. 24, No. 5, 2021639gained in popularity as art education program venues for 640nearby elementary schools. Therefore, the art museums in Seoul 641serve not only as exhibition spaces but also as educational 642facilities. Additionally, another study (Seresinhe et al., 2016) 643showed that the image of art represented by art museums 644can affect housing prices. These findings indicate that the 645accessibility of art and general museums reflects residents’ 646needs and preferences, and in turn, influences housing prices.647Third, public libraries negatively influence apartment prices 648at both the gu and dong levels. This implies that the percep-649tion of public libraries needs to change from ‘book lending 650library’ to ‘community cultural complex’ (Noh and Kim, 2019). 651Notably, residents regard public libraries as community spaces 652for cultural activities (Seoul Metropolitan City, 2019). If public 653libraries do not meet their expectations, they lose their appeal 654and fail to motivate residents to work within the community 655for cultural reinforcement (Adkins and Hussey, 2006), which 656would have an adverse effect on apartment prices. Although 657apartment residents accept public libraries as a type of edu-658cational facility, they consider them less important than pri-659vate educational facilities. Therefore, public libraries are re-660quired to serve as both educational and cultural facilities for 661residents. Enduring efforts and prompt support from both local 662and central governments are necessary to achieve this con-663version for public libraries. Beyond this, policy-makers could 664adopt measures to transform public libraries from basic fa-665cilities to complex systems with various educational and 666cultural programs (Gu, 2019; Seoul Metropolitan City, 2019).667Fourth, historical sites negatively influence the prices of 668most housing types, at both the gu and dong levels. Historical 669sites are often perceived as a hindrance to local physical 670development (Been et al., 2014) because the disadvantages 671of construction restrictions appear to outweigh their cultural 672benefits, thus adversely affecting social environment sat-673isfaction (Ki and Sung, 2020). In other words, historical 674sites fail to attract public interest as cultural heritage sites 675(Hyundai Research Institute, 2018), due to the lack of under-676standing of the inherent value and information they provide 677(Korea Cultural Heritage Administration, 2018). To address 678this, public efforts must be made to improve the quality of the 679interpretive content of signs, interpretation of cultural heritage, 680and relevant cultural programs to help residents appreciate 681the value of historical sites.682Finally, urban park areas negatively affect the prices of all 683housing types at the gu level, and of apartment prices at 684the dong level. Although urban parks were initially created to 685meet residents’ needs (Kim et al., 2006), negative external-686ities, such as noise or disturbances from frequently used com-687munity sports facilities in the park, can cause inconvenience 688to residents, adversely affecting cultural environmental sat-689isfaction (Ki and Sung, 2020) and in turn, the housing pri-690ces in the area (Lin, 2016). For instance, apartment resi-691dents often register civil complaints against park users due 692to noise disturbances (Sung and Ki, 2021). In particular, 693considering that most apartments in Seoul Metropolitan 694City have their own green spaces, urban park areas - which 695negatively affect apartment prices at the dong level - seem 696unnecessary. Therefore, urban park planners must find al-697ternative means to minimize negative externalities, such as 698employing landscape treatments to screen out disturbances 699(More et al., 1988).700Despite the significance of this study, we acknowledge its 701limitations. First, we assumed that the ‘other location factors’ 702in the spatial dataset are similar, because Seoul’s dong-level 703Dependent Variables704Spillover Effects of Independent Variables705Best Model706Positive Effects707Negative Effects708Residents’ immediate 709neighbourhood 710(dong)711Multi-unit housing712SEM713Private educational facility, art museum714Historical site715Single-unit housing716SLM717Private educational facility, art museum718Historical site719Apartment720SLM721Private educational facility722Public library, urban park area723Residential district 724(gu)725Multi-unit housing726SEM727Private educational facility, art museum, museum728Historical site, urban park area729Single-unit housing730SEM731Private educational facility, art museum, museum732Historical site, urban park area733Apartment734SEM735Private educational facility, art museum, museum Public library, historical site, urban park area736Note: SLM: spatial lag model, SEM: spatial error model.737Table 9. Comparison of research results738