11:00 AM - 1:00 PM
[HCG22-P02] landscape appreciation of residential areas in Munich
Keywords:landscape appreciation, residential areas in Minich, greenery and snow, winter dreary
Research method
We used the previous study of Aoki(1993).
Preparation of images for the residential area
The residential areas in Munich were presented by Prof. Valentin's laboratory at the Munich Technical University during Aoki stay in Munich and were also advised by members of Prof. Lats's laboratory. The early spring and late summer photos were taken by Aoki in 1984, and the winter photos were taken by Kroitzsch in 2003.
We prepared photographs of the 30 residential areas taken from the same spots in three seasons.
Conducting an evaluation survey on images
These 90 images were sent by e-mail to acquaintances around the world, and 18 respondents rated their preference 1-5(preferable). Respondents were interested in landscape appreciation, including 9 Japanese, 6 Europeans, and 3 Asians.
Measurement of physical quantities of images
Physical quantities related to the landscape were measured according to Aoki(1993). Table 2 shows the measurement items and measuring methods. As the snow scene was added this time, we also measured the snow coverage. A new variable was found during the preference analysis. The feeling of "winter dreary", which is the seasonal word of haiku, had an influence. However, this cannot be measured as a physical quantity, 1 is given to the image that is dreary and 0 is given to the image that is not, and it is used as a dummy variable.
Relationship between preference and physical quantities
Respondents' preferences were analyzed using image information showing these physical conditions. As a result, Table 3 showed the different influences of various quantities by the respondents. Winter dreary had a statistically significant effect on most people. Greenery also contributed preference similarly, followed by snow. The other variables affected individually differed, and no variable was affected for more than half of the respondents statistically. Some respondents were affected by the population density, by the floor height, by the coniferous trees, by the conspicuous wall surfaces, by the width to the road, by the planting shrubs, by the number of cars. Some were affected by the roof area, by the grassland, and so on. No one was affected by the size of the road, the glass walls and the height of the roof.
In this way, the responses of the respondents were diverse, and the responses of the landscape preference were different for each person. The multiple correlation coefficient for each respondent's obtained averaged 0.675, ranging from a minimum of 0.499 to a maximum of 0.802. Judging from the value of the multiple correlation coefficient, it is shown that there are still important physical variables, and further research is required.
Favorite group
When the t-value (statistical importance) of the partial regression coefficient obtained by the multiple regression analysis was grouped by the cluster analysis, it was divided into three groups (Fig.2).
The first group was a group containing many Japanese, the second group was a group of data4 and data15, and the third group was a group containing many Europeans.
From this result, there are Japanese landscape preferences and European landscape preferences, and there is another group that mainly evaluates human factors.
We investigated which variables separate into these groups. Figures 3, 4 and 5 show the relationships of these groups using the average values of the partial regression coefficients of each group. Different evaluations in the constant item and greenery were observed between the 1st and the 2nd group. And the item for winter dreary was differently observed between the 2nd and the 3rd group. The different evaluation of greenery and automobiles were observed between groups 1 and 3.
Overall landscape preference
When multiple regression analysis was performed using the average of the respondents' preferences, four variables of snow, greenery, roof area, and winter dreary were effective. Compared to the previous result (Aoki 1993), the number of valid variables has decreased, but it is a reasonable result considering that this result summarized the opinions of various types of respondents. In particular, snow, greenery, and winter dreary are variables found in more than half of the respondents. The most influential factor "winter dreary" is something that cannot be physically measured yet, so the future research with imaging analysis researchers will be needed.
We used the previous study of Aoki(1993).
Preparation of images for the residential area
The residential areas in Munich were presented by Prof. Valentin's laboratory at the Munich Technical University during Aoki stay in Munich and were also advised by members of Prof. Lats's laboratory. The early spring and late summer photos were taken by Aoki in 1984, and the winter photos were taken by Kroitzsch in 2003.
We prepared photographs of the 30 residential areas taken from the same spots in three seasons.
Conducting an evaluation survey on images
These 90 images were sent by e-mail to acquaintances around the world, and 18 respondents rated their preference 1-5(preferable). Respondents were interested in landscape appreciation, including 9 Japanese, 6 Europeans, and 3 Asians.
Measurement of physical quantities of images
Physical quantities related to the landscape were measured according to Aoki(1993). Table 2 shows the measurement items and measuring methods. As the snow scene was added this time, we also measured the snow coverage. A new variable was found during the preference analysis. The feeling of "winter dreary", which is the seasonal word of haiku, had an influence. However, this cannot be measured as a physical quantity, 1 is given to the image that is dreary and 0 is given to the image that is not, and it is used as a dummy variable.
Relationship between preference and physical quantities
Respondents' preferences were analyzed using image information showing these physical conditions. As a result, Table 3 showed the different influences of various quantities by the respondents. Winter dreary had a statistically significant effect on most people. Greenery also contributed preference similarly, followed by snow. The other variables affected individually differed, and no variable was affected for more than half of the respondents statistically. Some respondents were affected by the population density, by the floor height, by the coniferous trees, by the conspicuous wall surfaces, by the width to the road, by the planting shrubs, by the number of cars. Some were affected by the roof area, by the grassland, and so on. No one was affected by the size of the road, the glass walls and the height of the roof.
In this way, the responses of the respondents were diverse, and the responses of the landscape preference were different for each person. The multiple correlation coefficient for each respondent's obtained averaged 0.675, ranging from a minimum of 0.499 to a maximum of 0.802. Judging from the value of the multiple correlation coefficient, it is shown that there are still important physical variables, and further research is required.
Favorite group
When the t-value (statistical importance) of the partial regression coefficient obtained by the multiple regression analysis was grouped by the cluster analysis, it was divided into three groups (Fig.2).
The first group was a group containing many Japanese, the second group was a group of data4 and data15, and the third group was a group containing many Europeans.
From this result, there are Japanese landscape preferences and European landscape preferences, and there is another group that mainly evaluates human factors.
We investigated which variables separate into these groups. Figures 3, 4 and 5 show the relationships of these groups using the average values of the partial regression coefficients of each group. Different evaluations in the constant item and greenery were observed between the 1st and the 2nd group. And the item for winter dreary was differently observed between the 2nd and the 3rd group. The different evaluation of greenery and automobiles were observed between groups 1 and 3.
Overall landscape preference
When multiple regression analysis was performed using the average of the respondents' preferences, four variables of snow, greenery, roof area, and winter dreary were effective. Compared to the previous result (Aoki 1993), the number of valid variables has decreased, but it is a reasonable result considering that this result summarized the opinions of various types of respondents. In particular, snow, greenery, and winter dreary are variables found in more than half of the respondents. The most influential factor "winter dreary" is something that cannot be physically measured yet, so the future research with imaging analysis researchers will be needed.