Antônio Tarcísio Reis, Faculty of Architecture - PROPUR, Federal University of Rio Grande do Sul

Ana Flávia Panzenhagen, Faculty of Architecture, Federal University of Rio Grande do Sul

Isadora Martinez Diniz, Faculty of Architecture, Federal University of Rio Grande do Sul

INTRODUCTION

Streets in traditional urban settings tend to be bounded by buildings facades with doors and windows (Bentley, 2005; Reis, 2014a, 2014b). However, this direct relationship between buildings and public open spaces has been altered by modernist urban planning, where buildings tend to be located within blocks and often with blind walls facing the streets (Reis, 2014a, 2014b; Trancik, 1986). These characteristics have been replicated in modernist housing estates, in accommodation for athletes participating in Pan American and Olympic Games, and university campuses in Brazil and elsewhere (Reis, 2014a). In addition, ground floor spaces, recognized as the most important spaces in the relationship between the building and the street (Gehl, 2010), have been used as garages in distinct cities, instead of uses that can contribute to the urban vitality.

Moreover, gated condominiums enclosed by walls have spread out in areas with different levels of urbanization around the world, provoking a rupture in the traditional streets in consolidated urban areas characterized by buildings facades with doors and windows facing the street, with negative consequences for the use and the aesthetics of public open spaces (e.g., Becker & Reis, 2004). Although other studies have also shown the negative impact of such physical and visual barriers on urban vitality and aesthetic quality (e.g., Bentley, 2005; Gehl, 2010; Jacobs, 1984), they continue to be built in many cities. It is also not clear how far these barriers would affect preferences for a place to live.

In addition, while some findings show differences in the aesthetic evaluations by architects and laypeople (e.g., Fawcett, Ellingham, & Platt, 2008; Nasar, 1998), others show that there are no significant differences among aesthetic evaluations of architects, non-architects college graduates and non- college graduates (e.g., Reis, Biavatti, & Pereira, 2011, 2014). Therefore, there is a need to deepen the knowledge regarding the aesthetic preferences of streets with different levels of permeability by people with distinct educational background, as well as the preferences of such streets as places to live.

Hence, the objective of this paper is to analyze the preferences for streets with different levels of permeability regarding aesthetics and as places to live by three groups of people with different levels and types of educational background, namely: architects; non-architects college graduates and non-college graduates. These different levels of permeability mean: buildings inside the blocks with no direct relationship with the street; buildings on the block perimeter with doors and windows facing the street; buildings with blind walls facing the street; gated condominiums’ walls facing the street; and garages’ doors facing the street.

METHODOLOGY

Data were collected through questionnaires made available in the LimeSurvey software via the internet. Respondents were invited to participate by e-mail sent to unions of employees of federal educational institutions (Federal University of Rio Grande do Sul, Federal University of Health Sciences of Porto Alegre, and Federal Institute of Rio Grande do Sul), to the Deans and Departments of many Faculties at the Federal University of Rio Grande do Sul, including the Faculty of Architecture, and to educational institutions in charge of preparing for the university entrance examination. In total, 250 people with different levels and types of educational background filled in the questionnaire, as follows: 62 architects, 169 non-architects college graduates and 19 non-college graduates.

Considering urban scenes with distinct levels of permeability, questions about preferences of scenes´ appearance and preferences of scenes as places to live, as well as multiple-choice questions about the justifications for such preferences were used, such as: "Rank the scenes from the most (1) to the least (3) preferred in appearance: ( ) Scene A, ( ) Scene B, ( ) Scene C"; "Please state the main reasons that justify the most preferred scene´s appearance: ( ) Presence of buildings with openings facing the street, ( ) Direct relationship of the buildings with the street, ( ) Others: ……”; "Please, state the main reasons that justify the least preferred scene: ( ) Presence of walls, ( ) Lack of direct relationship of the buildings with the street, ( ) Others: ……”; “Rank the scenes with the places, from the most (1) to the least (3) preferred to live: ( ) Scene A, ( ) Scene B, ( ) Scene C”; “Indicate the main reasons that justify the most preferred scene to live in: ( ) Presence of buildings with openings facing the street, ( ) Direct relationship of the buildings with the street, ( ) Others: ……”; “Indicate the main reasons that justify the least preferred scene to live in: ( ) Presence of walls, ( ) Lack of direct relationship of the buildings with the street, ( ) Others: ….”.

Nine street scenes in three sets constituted by three scenes each, representing distinct levels of street permeability, were edited and presented in the questionnaires, namely: street scenes A (buildings inside the blocks with no direct relationship with the street), B (buildings on the block perimeter with doors and windows facing the street) and C (buildings with blind walls facing the street) (Figure 1); street scenes D (Gated condominiums´ walls facing the street), E (gated condominiums´ walls facing the street - buildings on the block perimeter with doors and windows facing the street) and F (buildings on the block perimeter with doors and windows facing the street) (Figure 2); and street scenes G (garages´ doors facing the street), H (garages´ doors facing the street and buildings on the block perimeter with doors and windows facing the street) and I (buildings on the block perimeter with doors and windows facing the street) (Figure 3).

Each scene is formed by two edited and combined color photographs of streets of Porto Alegre (Brazil). Some items, such as distinct sky appearance, rubbish bins, poles and light wires, that might affect the aesthetic evaluation of the urban interfaces, were removed from the scenes through the Adobe Photoshop CS6 program. In order to avoid great contrasts and shadows in the buildings, the scenes also have similar levels of natural lighting, with little or no solar incidence. With the purpose of obtaining the right combination of two photographs in each of the nine scenes representing the urban interfaces, and the adequate questions, three pilot tests were carried out.

The use of color photographs to represent the different urban interfaces is based on their recurrent utilization in studies involving aesthetic evaluations and the adequacy of these photographs to simulate a real environment (Sanoff, 1991). The photos that make up the scenes have different perspectives (including distinct widths of sidewalks and traffic lanes of vehicles) due to the fact that these photos were taken from points that allowed an adequate visualization of the interfaces. The selected points varied from street to street due to the different types of obstacles and existing visual barriers (e.g., trees, rubbish bins, poles and light wires). However, the different viewing angles do not have a significant effect on the aesthetic evaluations, since the existing architectural elements are the determinants for such evaluations (Stamps, 2000). Corroborating this argument, the variations between the perspectives of the scenes did not compromise preferences for scenes in each set (with three scenes), as revealed by the pilot tests carried out to verify the suitability of the scenes to represent the different categories of interfaces considered in the research. Moreover, in the justifications for the most and least preferred scenes, the alternative 'Other: ....' was included in the questionnaire, allowing the respondent to mention, for example, a justification related to a particular difference among the scenes.

The data in the LimeSurvey program were transferred to and analyzed in the PASW Statistics 18 software through non parametric statistical tests, such as Kruskal-Wallis and Kendall W. The first identifies, for example, the existence of a statistically significant difference (sig. ≤ 0.05) among the preferences for each of the nine scenes by the three groups of respondents. The second reveals, for example, the existence of a statistically significant difference among the preferences for the three scenes in each of the three sets by each of the three groups of respondents. Moreover, correlations (Spearman's rank correlation coefficient) between preference for street scene´ appearance and preference for a street as a place to live were carried out.

RESULTS

Regarding the appearance of scenes A (buildings inside the blocks with no direct relationship with the street), B (buildings on the block perimeter with doors and windows facing the street) and C (buildings with blind walls facing the street) (Figure 1, Table 1), architects prefer the scene with buildings on the block perimeter with doors and windows facing the street (B), closely followed by the scene with buildings inside the blocks with no direct relationship with the street (A) which is also the preferred scene by non-architects and non-college graduates.

Although scene A received the lowest number of points according to its preference by the architects (Table 1), scene B was preferred by a larger number of architects (23) than scene A (22). The preference of architects for scene B is due to the direct relationship of the buildings with the street (87% of 23) and the existence of openings facing the street (60.9% of 23), while the aesthetic preferences for scene A are related mainly to the perception of the presence of buildings with openings to the street (54.5% of 22 architects; 41.7% of 108 non-architects college graduates; 50% of 10 non-college graduates) and the perception by non-architects of the existence of a direct relationship with the street (42.6% of 108 non-architects college graduates; 40% of 10 non-college graduates).

Figure 1. Urban scenes A, B and C

Table 1. Order of preference for scenes´ appearance

Note: Scene A - buildings inside the blocks with no direct relationship with the street; Scene B - buildings on the block perimeter with doors and windows facing the street; Scene C - buildings with blind walls facing the street; mrv K = mean rank values obtained through Kendall W Test; these values should be compared in the column, considering the lowest value as an indicator of the highest preference; mrv K-W = mean rank values obtained through Kruskal-Wallis Test; these values should be compared in the line of each scene, considering the lowest value as an indicator of the highest preference; the values in parentheses represent the sum of the points received by each scene in each of the three groups of respondents, each of them assigning values from 1 (for the preferred scene) to 3 (for the least preferred scene); thus, the smaller the value in parenthesis, the greater the preference for the scene.

The scene with buildings with blind walls facing the street (C) is the least preferred by architects and non-architects college graduates, while scene B, closely followed by scene C, is the least preferred by non-college graduates. Scene C tends to be less preferred due to the presence of walls (58.3% of 36 architects; 33.3% of 75 non-architects college graduates; 75% of 4 non-college graduates) and lack of direct relationship with street (75% of 36 architects; 66.7% of 75 non-architects college graduates; 50% of 4 non-college graduates). The lowest preference for scene B by non-college graduates is due to the perception of lack of space between the buildings (50% de 6) and lack of direct relation of the buildings with the street (33.3% de 6).

These differences in preferences for the appearance of the three scenes (A, B and C) by each group (architects - Kendall W, test statistic = 34.125, sig. = .000; non-architects college graduates - Kendall W, test statistic = 107.324, sig. = .000; non-college graduates - Kendall W, test statistic = 11.091, sig. = .004) and the differences among the preferences for each scene (A - Kruskal-Wallis test, test statistic = 16.675, sig.= .000; B - Kruskal-Wallis test, test statistic = 21.031, sig.= .000; C - Kruskal-Wallis test, test statistic = 7.048, sig.= .029) by the three groups are statistically supported.

Scene A is the most preferred as a place to live for any of the three groups, although it is closely followed by scene B as the second most preferred by architects due to direct relationship of buildings with the street (86.4% de 22) and to presence of buildings with openings facing the street (68.2% de 22) (Table 2). The reasons for the preference for scene A are, mainly, due to the presence of buildings with openings facing the street (63.2% of 19 architects; 45.4% of 108 non-architects college graduates; 54.5% of 11 non-architects college graduates) and to the perception of direct relationship of buildings with the street (10.5% of 19 architects; 39.8% of 108 non-architects college graduates; 27.3% of 11 non-architects college graduates).

On the other hand, scene C (buildings with blind walls facing the street) is by far the one with the least preferred place to live by architects, and to a lesser extent also by non-architects college graduates, due: to lack of direct relationships of buildings with the street (75% of 32 architects; 58.1% of 74 non-architects college graduates) and to the presence of walls (59.4% of 32 architects; 33.8% of 74 non-architects college graduates).

Scene B, closely followed by scene C, is the least preferred by non-college graduates due to perception of lack of direct relationships of buildings with the street (50% of 6), while the justification for scene C being the second least preferred is related to presence of walls (80% of 5) and to lack of direct relationships of buildings with the street (40% of 5).

These differences in preferences for the three scenes as places to live by each group (architects - Kendall W, test statistic = 22.435, sig. = .000; non-architects college graduates - Kendall W, test statistic = 104.993, sig. = .000; non-college graduates - Kendall W, test statistic = 16.545, sig. = .000) and the differences among the preferences for two scenes as a place to live (A - Kruskal-Wallis test, statistical test = 22.746, sig.=0.000; B - Kruskal-Wallis test, statistical test = 15.876, sig.=0.000) by the three groups are statistically supported.

Table 2. Order of preference for scenes with the places most and least preferred to live

Note: Scene A - buildings inside the blocks with no direct relationship with the street; Scene B - buildings on the block perimeter with doors and windows facing the street; Scene C - buildings with blind walls facing the street; mrv Kendall = mean rank values obtained through Kendall W Test; these values should be compared in the column, considering the lowest value as an indicator of the highest preference; mrv K-W = mean rank values obtained through Kruskal-Wallis Test; these values should be compared in the line of each scene, considering the lowest value as an indicator of the highest preference; the values in parentheses represent the sum of the points received by each scene in each of the three groups of respondents, each of them assigning values from 1 (for the preferred scene) to 3 (for the least preferred scene); thus, the smaller the value in parenthesis, the greater the preference for the scene.

Correlations between preference for scene´s appearance and preference for scene as a place to live were found regarding each of the groups of respondents for each of the three scenes, with exception of the lack of correlation concerning such preferences for scene A by non-college graduates (Table 3).

Table 3. Correlation between preference for scene´s appearance and preference for scene as a place to live

Regarding the appearance of urban scenes D (gated condominiums´ walls facing the street), E (gated condominiums´ walls facing the street - buildings on the block perimeter with doors and windows facing the street) and F (buildings on the block perimeter with doors and windows facing the street) (Figure 2, Table 4), scene F is the most preferred by the two groups with college education due the presence of buildings with openings facing the street (90.2% of 41 architects; 89% of 73 non-architects college graduates), while scene D appearance is the most preferred by non-college graduates due to existence of grassy areas (100% of 4) and trees (50% of 4) together with scene F due to presence of buildings with openings facing the street (75% of 4).

Scene D appearance is the least preferred by architects and by non-architects college graduates, due to presence of walls (97.6% of 41 architects; 93.8% of 64 non-architects college graduates), while scene E appearance is the least preferred by non-college graduates due to presence of walls (75% of 4). These differences in the preferences of the three scenes´ appearance by architects (Kendall W, test statistic = 52.898, sig. = .000) and by non-architects college graduates (Kendall W, test statistic = 23.747, sig. = .000) are statistically supported. The differences among the preferences of scene D (Kruskal-Wallis, test statistic = 19.159, sig. = 0.000) and scene F (Kruskal-Wallis, test statistic = 12.150, sig. = 0.002) by the three groups are also statistically supported.

Figure 2. Urban scenes D, E e F.

Table 4. Order of preference for scenes´ appearance

Note: Scene D - gated condominiums´ walls facing the street; Scene E - gated condominiums´ walls facing the street - buildings on the block perimeter with doors and windows facing the street; Scene F - buildings on the block perimeter with doors and windows facing the street; mrv K = mean rank values obtained through Kendall W Test; these values should be compared in the column, considering the lowest value as an indicator of the highest preference; mrv K-W = mean rank values obtained through Kruskal-Wallis Test; these values should be compared in the line of each scene, considering the lowest value as an indicator of the highest preference; the values in parentheses represent the sum of the points received by each scene in each of the three groups of respondents, each of them assigning values from 1 (for the preferred scene) to 3 (for the least preferred scene); thus, the smaller the value in parenthesis, the greater the preference for the scene.

Regarding the preferences for scenes D (gated condominiums´ walls facing the street), E (gated condominiums´ walls facing the street - buildings on the block perimeter with doors and windows facing the street) and F (buildings on the block perimeter with doors and windows facing the street) as places to live (Table 5), architects prefer scene F due to presence of buildings with openings facing the street (87.2% of 39). Scene D is the most preferred by non-architects college graduates and by non-college graduates due to existence of grassy areas (66.7% of 54 non-architects college graduates; 66.7% of 6 non-college graduates) and trees (70.4% of 54 non-architects college graduates; 33.3% of 6 non-college graduates).

Scene D is the least preferred by architects due to presence of walls (92.1% of 38 architects). Scene F is the least preferred by non-architects college graduates and by non-college graduates due to lack of grassy areas (61.8% of 55 non-architects college graduates; 33.3% of 6 non-college graduates) lack of trees (63.6% of 55 non-architects college graduates ) and presence of walls (33.3% of 6 non-college graduates). The differences in preferences for the three scenes (D, E and F) as places to live by architects (Kendall W, test statistic = 38.164, sig. = .000) are statistically supported. In addition, differences among the preferences for scenes F (Kruskal-Wallis Test, statistical test = 22.030, sig.=0.000) and D (Kruskal-Wallis test, statistical test = 20.661, sig.=0.000) as places to live by the three groups are also statistically supported. In turn, preferences for scenes D, E and F as places to live by non-architects college graduates are very close to each other.

Table 5. Order of preference for scenes with the places most and least preferred to live

Note: Scene D - gated condominiums´ walls facing the street; Scene E - gated condominiums´ walls facing the street - buildings on the block perimeter with doors and windows facing the street; Scene F - buildings on the block perimeter with doors and windows facing the street; mrv Kendall = mean rank values obtained through Kendall W Test; these values should be compared in the column, considering the lowest value as an indicator of the highest preference; mrv K-W = mean rank values obtained through Kruskal-Wallis Test; these values should be compared in the line of each scene, considering the lowest value as an indicator of the highest preference; the values in parentheses represent the sum of the points received by each scene in each of the three groups of respondents, each of them assigning values from 1 (for the preferred scene) to 3 (for the least preferred scene); thus, the smaller the value in parenthesis, the greater the preference for the scene.

Correlations between preference for scene´s appearance and preference for scene as a place to live were found regarding each of the groups of respondents for each of the three scenes, revealing again that the appearance of the scene have a clear effect on its preference as place to live (Table 6).

Table 6. Correlation between preference for scene´s appearance and preference for scene as a place to live

Considering the preference for the appearance of scenes G (garages´ doors facing the street), H (garages´ doors facing the street and buildings on the block perimeter with doors and windows facing the street) and I (buildings on the block perimeter with doors and windows facing the street) (Figure 3, Table 7), scene I clearly is the most preferred by each of the three groups due to direct relationship between the building and the street (73.8% of 42 architects; 58.7% of 90 non-architects college graduates; 57.1% of 7 non-college graduates) and to the presence of buildings with openings facing the street (64.3% of 42 architects; 52.2% of 90 non-architects college graduates; 57.1% of 7 non-college graduates).

In turn, scene G is the least preferred by the three groups due to presence of walls (61.9% of 42 architects; 50.6% of 21 non-architects college graduates; 71.4% of 10 non-college graduates) and lack of direct relationship between the building and the street (61.9% of 42 architects; 51.7% of 21 non-architects college graduates; 57.1% of 10 non-college graduates). These differences in preferences for the three scenes´ appearance (G, H and I) by architects (Kendall W, test statistic = 64.478, sig. = .000) and by non-architects college graduates (Kendall W, test statistic = 99.770, sig. = .000), are statistically supported. In turn, no statistically supported differences among the preferences for each scene by the three groups (Kruskal-Wallis test) were found.

Figure 3. Urban scenes G, H and I

Table 7. Order of preference for scenes´ appearance

Note: scene G - garages´ doors facing the street; scene H - garages´ doors facing the street and buildings on the block perimeter with doors and windows facing the street; scene I - buildings on the block perimeter with doors and windows facing the street; mrv K = mean rank values obtained through Kendall W Test; these values should be compared in the column, considering the lowest value as an indicator of the highest preference; mrv K-W = mean rank values obtained through Kruskal-Wallis Test; these values should be compared in the line of each scene, considering the lowest value as an indicator of the highest preference; the values in parentheses represent the sum of the points received by each scene in each of the three groups of respondents, each of them assigning values from 1 (for the preferred scene) to 3 (for the least preferred scene); thus, the smaller the value in parenthesis, the greater the preference for the scene.

Scene I visibly is the most preferred as a place to live by the college graduates, and it is also the most preferred by the non-college graduates (Table 8) due to direct relationship of buildings with the street (78.9% of 38 architects; 60% of 85 non-architects college graduates; 60% of 5 non-college graduates) and to the presence of buildings with openings facing the street (63.2% of 38 architects; 57.6% of 85 non-architects college graduates; 60% of 5 non-college graduates).

On the other hand, scene G clearly is the least preferred place to live by the three groups due to presence of walls (59% of 39 architects; 51.9% of 81 non-architects college graduates; 66.7% of 6 non-college graduates) and lack of direct relationship of buildings with the street (61.5% of 39 architects; 49.4% of 81 non-architects college graduates; 83.3% of 6 non-college graduates).

These differences in preferences for the three scenes (G, H and I) as places to live by architects (Kendall W, test statistic = 45.957, sig. = .000) and by non-architects college graduates (Kendall W, test statistic = 75.446, sig. = .000) are statistically supported. Preferences for each scene (G, H and I) as a place to live were not found to be statistically significant different among the three groups (Kruskal-Wallis test).

Table 8. Order of preference for scenes with the places most and least preferred to live

Note: scene G - garages´ doors facing the street; scene H - garages´ doors facing the street and buildings on the block perimeter with doors and windows facing the street; scene I - buildings on the block perimeter with doors and windows facing the street; mrv Kendall = mean rank values obtained through Kendall W Test; these values should be compared in the column, considering the lowest value as an indicator of the highest preference; mrv K-W = mean rank values obtained through Kruskal-Wallis Test; these values should be compared in the line of each scene, considering the lowest value as an indicator of the highest preference; the values in parentheses represent the sum of the points received by each scene in each of the three groups of respondents, each of them assigning values from 1 (for the preferred scene) to 3 (for the least preferred scene); thus, the smaller the value in parenthesis, the greater the preference for the scene.

Correlations between preference for scene´s appearance and preference for scene as a place to live were found regarding each of the groups of respondents for each of the three scenes (G, H and I), emphasizing the effect of scene´s appearance on its preference as a place to live (Table 9).

Table 9. Correlation between preference for scene´s appearance and preference for scene as a place to live

Therefore, regarding the appearance of nine scenes, statistically supported differences among the most preferred scenes by the three groups of respondents were found regarding scenes A, B, C, D, and F (Table 10). Results show that in all the three sets of scenes, architects prefer the scene´s appearance with buildings on the block perimeter with doors and windows facing the street (scenes B, F and I; Table 10) . The main reasons are consistently related to: the direct relationship between the buildings and the street (scenes B and I); and the existence of buildings with openings facing the street (scenes B, F and I). The preference for scene B is closely followed by the preference for scene A (buildings inside the blocks with no direct relationship with the street scene) due to the perception of the presence of buildings with openings to the street.

Table 10. The most and least preferred scenes regarding appearance and as a place to live

Note: street scenes A (buildings inside the blocks with no direct relationship with the street), B (buildings on the block perimeter with doors and windows facing the street) and C (buildings with blind walls facing the street) (Figure 1); scenes D (Gated condominiums´ walls facing the street), E (gated condominiums´ walls facing the street - buildings on the block perimeter with doors and windows facing the street) and F (buildings on the block perimeter with doors and windows facing the street) (Figure 2); and street scenes G (garages´ doors facing the street), H (garages´ doors facing the street and buildings on the block perimeter with doors and windows facing the street) and I (buildings on the block perimeter with doors and windows facing the street) (Figure 3); 1a = most preferred scene; 1b = second most preferred scene, closely following scene 1a; 3 = least preferred scene; K test = Kendall W test; sig = significant; K-W test = Kruskal-Wallis test.

The preferred scenes´ appearances by non-architects college graduates are those with buildings inside the blocks with no direct relationship with the street (A), and with buildings on the block perimeter with doors and windows facing the street (F; I) (Table 10). The reasons for such preferences are: perception of the existence of a direct relationship between the building and the street (scenes A and I); and perception of the presence of buildings with openings facing the street (scenes A, F and I).

In its turn, non-college graduates´ preferences for scenes´ appearance are characterized by: buildings inside the blocks with no direct relationship with the street (A); gated condominiums´ walls facing the street (D; closely followed by F - buildings on the block perimeter with doors and windows facing the street); and buildings on the block perimeter with doors and windows facing the street (I) (Table 10). The reasons for such preferences are: perception of the presence of buildings with openings facing the street (scenes A, F, and I); perception of the existence of a direct relationship between the building and the street (scenes A, and I); existence of grassy areas (scene D); and existence of trees (scene D).

The least preferred scenes´ appearance by architects are those formed by: buildings with blind walls facing the street (C); gated condominiums´ walls facing the street (D); and the scene with garages´ doors facing the street (J) (Table 10). The reasons for these scenes being the least preferred are: presence of walls (scenes C, D , and G); and lack of direct relationship with street (scenes C and G).

The least preferred scenes´ appearance by non-architects college graduates are those with: buildings with blind walls facing the street (C); gated condominiums´ walls facing the street (D); and the scene with garages´ doors facing the street (G) (Table 10). The reasons for these scenes being the least preferred are: the presence of walls (scenes C, D and G); and lack of direct relationship with street (scenes C and G).

In its turn, non-college graduates´ least preferred scenes regarding their appearances are those: with buildings on the block perimeter with doors and windows facing the street (scene B; closely followed by C - buildings with blind walls facing the street); with gated condominiums´ walls facing the street and buildings on the block perimeter with doors and windows facing the street (E); and the scene with garages´ doors facing the street (G) (Table 10). The reasons for these scenes being the least preferred are: perception of lack of space between the buildings (B); perception of lack of direct relation of the buildings with the street (B); presence of walls (scenes C, E and G); and lack of direct relationship with street (scenes C and G). These differences in preferences for the three scenes´ appearance in each set (A, B and C; D, E and F; and G, H and I) by each group of respondents, with exception of the preference for the appearance of three scenes in set ‘D, E and F’ and in set ‘G, H and I’ by non-college graduates, are statistically supported (Table 10).

Concerning the nine scenes as a place to live, statistically supported differences among the most preferred scenes by the three groups of respondents, were found regarding scenes A, B, F and D (Table 10). The most preferred scenes as places to live by architects are those with buildings on the block perimeter with doors and windows facing the street (scenes F and L), with exception of scene A (buildings inside the blocks with no direct relationship with the street). However, this is closely followed by scene B which is characterized by buildings on the block perimeter with doors and windows facing the street (Table 10). The main reasons are related to: direct relationship of buildings with the street (scenes A, B and I); and presence of buildings with openings facing the street (scenes A, B, F and I).

The preferred scenes as places to live by non-architects college graduates are those with: buildings inside the blocks with no direct relationship with the street (A); gated condominiums´ walls facing the street (D); and buildings on the block perimeter with doors and windows facing the street (I) (Table 10). The main reasons are related to: presence of buildings with openings facing the street (scenes A and I); perception of direct relationship of buildings with the street (scenes A and I); existence of grassy areas (D); and existence of trees (D).

The preferred scenes as places to live by non-college graduates are those with: buildings inside the blocks with no direct relationship with the street (A); gated condominiums´ walls facing the street (D); and buildings on the block perimeter with doors and windows facing the street (I) (Table 10). The main reasons are related to: presence of buildings with openings facing the street (scenes A and I); perception of direct relationship of buildings with the street (A); existence of grassy areas (D); existence of trees (D); and direct relationship of buildings with the street (I).

The least preferred scenes as places to live by architects are those with: buildings with blind walls facing the street (C); gated condominiums´ walls facing the street (D); and garages´ doors facing the street (G) (Table 10). The main reasons are related to: lack of direct relationships of buildings with the street (C); presence of walls (scenes C, D, and G); and lack of direct relationship of buildings with the street (G).

The least preferred scenes as places to live by non-architects college graduates are characterized by: buildings with blind walls facing the street (C); buildings on the block perimeter with doors and windows facing the street (F); and garages´ doors facing the street (G) (Table 10). The main reasons are related to: lack of direct relationships of buildings with the street (C); presence of walls (scenes C and G); lack of grassy areas (F); lack of trees (F); and lack of direct relationship of buildings with the street (G).

The least preferred scenes as places to live by non-college graduates are constituted by: buildings on the block perimeter with doors and windows facing the street (B), closely followed by scene C (buildings with blind walls facing the street); buildings on the block perimeter with doors and windows facing the street (F); and garages´ doors facing the street (G) (Table 10). The main reasons are related to: perception of lack of direct relationships of buildings with the street (B); presence of walls (scenes C, F and G); lack of direct relationships of buildings with the street (scenes C and G); and lack of grassy areas (F). These differences among the preferences of the three scenes as a place to live in set ‘A, B and C’ by each group of respondents, in set ‘D, E and F’ by architects, and in set ‘G, H and I’ by those with a college degree, are statistically supported.

In its turn, correlations between preference for scene´s appearance and preference for scene as a place to live were found regarding each group of respondents for each of the nine scenes, with exception of the lack of correlation concerning such preferences for scene A by non-college graduates, revealing that the appearance of the scene tend to have an effect on its preference as a place to live.

CONCLUSION

Results show that the preferences as well as the justifications for such preferences were not affected by the variations in the perspectives of the scenes (including differences in the widths of the streets), corroborating the argument that the angles of vision used for the photographs do not have a significant effect on aesthetic evaluations (Stamps, 2000).

Therefore, when significant differences among the preferences of the three groups were found, these are related to a greater preference by architects for buildings on the block perimeter with doors and windows facing the street and a greater preference, mainly by non-college graduates, for the presence of some green in the scene. Nonetheless, when the green disappeared and the visual and physical barriers were maintained, as in the case of scene with garages´ doors facing the street, all the three groups rejected this scene in favor of that with buildings on the block perimeter with doors and windows facing the street.

Therefore, there is a tendency for these results to show no support to those that reveal the existence of differences between the aesthetic evaluations by architects and by laypeople (e.g., Fawcett, Ellingham, & Platt, 2008; Nasar, 1998) and to corroborate those results that reveal no significant differences among aesthetic evaluations of architects, non-architects college graduates and non-college graduates (e.g., Reis, Biavatti, & Pereira, 2011, 2014).

In addition, the findings of this study tend to emphasize the fact that physical and visual barriers represented by buildings with blind walls facing the street, gated condominiums´ walls and garages´ doors facing the street negatively affect people´s perception of a place to live and its appearance. This is due to the lack of visual stimulation of these elements as already evidenced in other studies (e.g., Becker & Reis, 2004; Reis, Seadi, & Biavatti, 2016). Moreover, physical and visual barriers do not attract pedestrians and, so, do not favor movement and permanence of people in public open spaces, clearly reducing the potential for urban vitality. Yet, as already pointed out in other studies (e.g., Gehl, 2010; Jacobs, 1984), the very presence of people in public open spaces makes these spaces more attractive to other people. Consequently these results can contribute, mainly in the case of Brazilian cities, to reduce urban interventions characterized by these physical and visual barriers.

REFERENCES

Becker, D.; Reis, A. (2004). O Impacto das barreiras físicas de condomínios horizontais fechados no espaço urbano. In: Anais da 1ª. Conferência Latino Americana de Construção Sustentável, 10º Encontro Nacional de Tecnologia do Ambiente Construído (pp. 1-15). São Paulo: ANTAC.

Bentley, I.; Mcglynn, S.; Smith, G.; Alcock, A.; Murrain, P. (2005). Responsive environments: A manual for designers. Oxford: Elsevier Ltd.

Fawcett, W.; Ellingham, I.; Platt, S. (2008). Reconciling the Architectural Preferences of Architects and the Public: the ordered preference model. Environment and Behavior, 40(5), 599-618.

Gehl, J. (2010). Cities for People. Washington: Island Press.

Jacobs, J. (1984). The Death and Life of Great American Cities - The failure of Town Planning. Harmondsworth, Middlesex, England: Penguin Books.

Nasar, J. (1998). The Evaluative Image of the City. Thousand Oaks, CA: SAGE.

Reis, A. (2014a). Forma urbana tradicional e modernista: Uma reflexão sobre o uso e estética dos espaços urbanos. ARQUISUR Revista, (6), 70-87.

Reis, A. (2014b). Urban form and its implication for the use of urban spaces. In V. Oliveira, P. Pinho, L. Batista, T. Patatas & C. Monteiro (Eds.), Our common future in Urban Morphology (pp. 1433-1443). Porto: FEUP.

Reis, A. T., Seadi, M. L., & Biavatti, C. (2016). Views From Apartment Buildings: An Analysis by Architects and Non-Architects College Graduates. In C. Palasar & A. A. Fox (Eds.), Conference Proceedings of the 47th Annual Conference of the Environmental Design Research Association (pp. 105–114). Raleigh: EDRA.

Reis, A.; Biavatti, C.; Pereira, M. L. (2014). Composição arquitetônica e qualidade estética. Ambiente Construído, 14(1), 191-213.

Reis, A.; Biavatti, C.; Pereira, M. L. (2011). Estética Urbana: uma análise através das ideias de ordem, estímulo visual, valor histórico e familiaridade. Ambiente Construído, 11(4), 185-204.

Sanoff, H. (1991). Visual Research Methods in Design. New York: Van Nostrand Reinhold.

Stamps, A. (2000). Psychology and the Aesthetics of the Built Environment. Massachusetts: KAP.

Trancik, R. (1986). Finding Lost Space: Theories of Urban Design. New York: Van Nostrand Reinhold.

ACKNOWLEDGEMENTS

This research was conducted with scientific initiation scholarships by CNPq (National Council for Scientific and Technological Development), a Brazilian governmental institution sponsor of scientific and technological development. Additionally, thanks are due to the undergraduate student in architecture Victoria Luvizetto Gerson for taking part in this research.