20 Aesthetics and Safety of Road Landscape: are they Related? Irina Matijošaitienė1*, Kristina Navickaitė2 1 Kaunas University of Technology, Faculty of Civil Engineering and Architecture, Studentu st. 48, LT-51367 Kaunas, Lithuania 2 Kaunas University of Technology, Faculty of Civil Engineering and Architecture, Studentu st. 48, LT-51367 Kaunas, Lithuania * corresponding author: irina.matijosaitiene@ktu.lt The problem of unsafe roads is very actual in Lithuania. The solution of the problem is usually found in construction of road equipment. The paper covers another solution: the authors analyze road safety through aesthetic features of landscape. The identification of relations between safety and landscape aesthetic features will enable us to enhance safety by modeling road landscape. This untypical approach would lead to both achievements: decrease of road traffic accidents and increase of visual quality of road landscape. The correlation analysis enabled us to identify weak relations between the quantity of car accidents and some aesthetic properties of road landscape. Regression analysis revealed the factors described by aesthetic properties which influence the quantity of car accidents on Lithuanian roads. Keywords: Road landscape, aesthetics, safety, correlation analysis, linear regression analysis. DARNIOJI ARCHITEKTŪRA IR STATYBA 2012. No. 1(1) JOURNAL OF SUSTAINABLE ARCHITECTURE AND CIVIL ENGINEERING ISSN 2029–9990 1. Introduction According to Eurostat statistical findings (Transport … 2012) about road accidents in EU member countries during 2008, the highest road fatality rates were recorded in Lithuania (148), Poland (143), Romania (142), Bulgaria, Greece and Latvia (all 139) (here the values of road fatality rates are expressed as the number of deaths per million inhabitants). Though during the last decade the number of car accidents on Lithuanian roads decreased from 5972 in 2001 (1715 accidents per million inhabitants) to 3312 in 2011 (1035 accidents per million inhabitants) (Eismo … 2012) and the fatality rates also decreased from 706 in 2001 (202 accidents per million inhabitants) to 297 in 2011 (93 accidents per million inhabitants) the overall road safety in Lithuania is not high, having in mind Lithuania’s position in the context of other EU member countries. For instance, road fatality rates in Sweden and the United Kingdom are both 43, in the Netherlands 41 and in Malta 37 per million inhabitants. Lithuania try to solve the problem of unsafe roads through speed restriction on dangerous segments of roads, installation of safety islands, roundabouts etc. Though many countries (USA, Germany, Great Britain, Australia etc.) involve road landscape aesthetics and design into enhancing of road safety. In these countries creation of an aesthetic road landscape is an essential part of creation a safe driving on the road. Literature review revealed some analysis of road safety through the prism of landscape aesthetics. According to the National Association of Australian State Road Authorities (NAASRA) the main goal of road landscaping is „to produce roadways to high safety standards which will also aesthetically integrate with the environment“ (Road … 1997). Though the Road landscape manual (1997) presents the assessment of the road landscape at visual-aesthetic, ecological (environmental) and cultural heritage consideration stages, road safety aspect has to be considered at all stages, including construction, operation and maintenance of roads (Road … 1997). According to American experience billboards impact the visual quality of the highway because they obstruct the views of scenic features and the natural landscape (I-15 … 2005). Advertising can also distract drivers through messages and products which are not relevant to travelling (Road … 1997). J. Edquist (2008) analyzed the effect of visual disorder on road safety. Chaotically located road signs, advertising, buildings, electrical transmission lines etc. are called here as visual disorder. The scholar carried out the research of simulation of drivers’ behavior at day time. The result of her research revealed that visual disorder in road landscape decrease drivers’ attention while driving and negatively affect safety on the roads. H. Antonson with a group of researchers (2009) analyzed the reliance of drivers’ behavior and safety on the road landscape type – open, woodlands or mixed. For the research they used the simulator of driving in these types of landscape. Then eighteen research participants were asked to answer the questions about their feelings while driving. The research results indicated that in the open landscape the speed of driving is faster, and that road safety depends on the landscape through which http://dx.doi.org/10.5755/j01.sace.1.1.2613 21 road passes. H. Drottenborg (2002) analyzed traffic safety and driving behavior in 10 beautiful and 10 ugly traffic environments. The results confirmed that “aesthetically rewarding traffic environments seem to be beneficial for traffic safety”, and that driving speed is lower in beautiful rather than ugly traffic environments due to more stops in beautiful landscape. Research in Germany demonstrated that the most part (68%) of all the car accidents happen due to the wrong design of road and its landscape, and because of insufficient informativity of road and its landscape. According to Küllers’ model of the basic emotional process, “driving behavior is related to the physical environment, other road users, the driving task, the individual factors and own abilities, and to the interaction among them” (Drottenborg, 2002). Therefore, this research is concentrated on the physical environment, id est the landscape through which the road passes. Accordingly, the aim of the research is to identify if there is a relation between landscape aesthetics and safety on Lithuanian roads. 2. Methods The research objects are the main Lithuanian highways which are marked as the European motorway network corridors or the European highways: A1 road Vilnius-Kaunas-Klaipėda (311.40 km), A2 road Vilnius- Panevėžys (135.92 km), A3 road Vilnius-Minsk (Belarus) (33.99 km), A5 road Kaunas-Marijampolė-Suwalki (Poland) (97.06 km), A6 road Kaunas-Zarasai-Daugavpils (Latvia) (185.40 km), A8 road Panevėžys-Aristava-Sitkūnai (87.86 km), A9 road Panevėžys-Šiauliai (78.94 km), A10 road Panevėžys-Pasvalys-Bauska (Latvia) (66.10 km), A11 road Šiauliai-Palanga (146.85 km), A12 road Riga (Latvia)- Šiauliai-Tauragė-Kaliningrad (Russia) (186.09 km), A13 road Klaipėda-Liepaja (Latvia) (45.15 km) and A16 road Vilnius-Prienai-Marijampolė (137.51 km) (Fig. 1). The total length of the researched roads is 1512.27 km. Only the road segments within the Lithuanian borders were considered. The research of aesthetic properties of road landscape was conducted in spring 2010 by employing the photo- fixation of road landscape and the qualitative survey (Matijošaitienė 2011). Selected photos of road landscape were used for the qualitative survey. Also aesthetic properties of road landscape were used for the survey and for the analysis of aesthetics and road safety: interesting, natural, visually safe, skittish, beautiful, outstanding, harmonious, sophisticated, relaxing, majestic, pleasant, elements match for surrounding environment, left an intense positive impression, willing to drive on this road. These properties were measured by the 5-point semantic differential scale, where 1 meant the least acceptance and 5 meant the most acceptance (Matijošaitienė and Stankevičė 2011). Thus, the aesthetic properties of road landscape are based on respondents’ emotions and their opinion about a certain road landscape. The number of respondents was N=486. PASW Statistics 17.0 software was applied for the correlation and regression analysis of the data. Correlation analysis was applied for the identification of relations between all or separate variables, and if there was a relation this analysis will enable us to identify the strength of the relation. The variables describing demographic, financial and marital status of respondents are nominal and interval (for the measurement of respondents’ income per month), the variables describing the aesthetic properties of roadscape are ordinal (rank), and the variables describing the quantity of car accidents on the researched roads are scale. Therefore, the Kendall’s tau_b and Spearman’s correlation coefficients were counted. Kendall’s tau-b correlation coefficient is used to measure the association between two measured quantities. Kendall’s tau-b, unlike tau-a, makes adjustments for ties and is suitable for square tables. In our case we have 15x15 table (according to the number of variables), thereby the table is square. Values of Kendall’s Aesthetics and Safety of Road Landscape: are they Related? 24 electrical transmission lines etc. are called here as visual disorder. The scholar carried out the research of simulation of drivers’ behavior at day time. The result of her research revealed that visual disorder in road landscape decrease drivers’ attention while driving and negatively affect safety on the roads. H. Antonson with a group of researchers (2009) analyzed the reliance of drivers’ behavior and safety on the road landscape type – open, woodlands or mixed. For the research they used the simulator of driving in these types of landscape. Then eighteen research participants were asked to answer the questions about their feelings while driving. The research results indicated that in the open landscape the speed of driving is faster, and that road safety depends on the landscape through which road passes. H. Drottenborg (2002) analyzed traffic safety and driving behavior in 10 beautiful and 10 ugly traffic environments. The results confirmed that “aesthetically rewarding traffic environments seem to be beneficial for traffic safety”, and that driving speed is lower in beautiful rather than ugly traffic environments due to more stops in beautiful landscape. Research in Germany demonstrated that the most part (68%) of all the car accidents happen due to the wrong design of road and its landscape, and because of insufficient informativity of road and its landscape. According to Küllers’ model of the basic emotional process, “driving behavior is related to the physical environment, other road users, the driving task, the individual factors and own abilities, and to the interaction among them” (Drottenborg, 2002). Therefore, this research is concentrated on the physical environment, id est the landscape through which the road passes. Accordingly, the aim of the research is to identify if there is a relation between landscape aesthetics and safety on Lithuanian roads. 2. Methods The research objects are the main Lithuanian highways which are marked as the European motorway network corridors or the European highways: A1 road Vilnius-Kaunas-Klaipėda (311.40 km), A2 road Vilnius- Panevėžys (135.92 km), A3 road Vilnius-Minsk (Belarus) (33.99 km), A5 road Kaunas-Marijampolė-Suwalki (Poland) (97.06 km), A6 road Kaunas-Zarasai-Daugavpils (Latvia) (185.40 km), A8 road Panevėžys-Aristava- Sitkūnai (87.86 km), A9 road Panevėžys-Šiauliai (78.94 km), A10 road Panevėžys-Pasvalys-Bauska (Latvia) (66.10 km), A11 road Šiauliai-Palanga (146.85 km), A12 road Riga (Latvia)-Šiauliai-Tauragė-Kaliningrad (Russia) (186.09 km), A13 road Klaipėda-Liepaja (Latvia) (45.15 km) and A16 road Vilnius-Prienai- Marijampolė (137.51 km) (Fig. 1). The total length of the researched roads is 1512.27 km. Only the road segments within the Lithuanian borders were considered. Fig. 1. Research objects The research of aesthetic properties of road landscape was conducted in spring 2010 by employing the photo-fixation of road landscape and the qualitative survey (Matijošaitienė 2011). Selected photos of road landscape were used for the qualitative survey. Also aesthetic properties of road landscape were used for the survey and for the analysis of aesthetics and road safety: interesting, natural, visually safe, skittish, beautiful, outstanding, harmonious, sophisticated, relaxing, majestic, pleasant, elements match for surrounding environment, left an intense positive impression, willing to drive on this road. These properties were measured Fig. 1. Research objects 22 tau_b range from -1 to +1. Spearman’s correlation coefficient (Spearman’s rho) is a non-parametric measure of statistical dependence between two variables. It assesses how well the relationship between two variables can be described using a monotonic function. If there are no repeated data values, a perfect Spearman correlation of +1 or −1 occurs when each of the variables is a perfect monotone function of the other. For instance, when X was increasing Y monotonously increases (not necessarily linearly) or decreases. The Spearman correlation coefficient is defined between the ranked variables. For the both correlation coefficients the correlation can be: a) very strong when the value is -1 or +1, b) strong when the value is from -1 to -0.7 or from +1 to +0.7, c) moderate when the value is from -0.7 to -0.5 or from +0.7 to +0.5, d) weak when the value is from -0.5 to -0.2 or from +0.5 to 0.2 and e) very weak when the value is from -0.2 to 0 or from +0.2 to 0. A value of 0 indicates the absence of relation. Multiple linear regression analysis was applied for the identification of visual characters of road landscape which influence the quantity of car accidents. Because all the variables which represents the visual character of road landscape are ordinal (rank) we make the assumption that the intervals between the ranks are equal. The biggest advantage of regression analysis is that regression model (function which connects variables) is composed. Regression model is a statistical model which let forecast the values of one variable through the values of other variables. English geneticist F. Galton used the term of regression for the first time during his research on the relation between height of children and their parents (Čekanavičius 2008). The literature review revealed many cases of application correlation and regression analyses for the research of landscape. For instance, T. Daniel (1976) applied correlation analysis for the research of Arizona (USA) woodlands. Regression analysis is often used for the practical research: for the forecast of election winners for the political purposes, for the identification of consumer opinion about a product or a service, for the research of landscape. P. Cook (1995) applied multiple regression and correlation analyses for the analysis of landscape scenery of Great Plains in the USA, R. Clay (2004, 2000) applied correlation, regression and factor analyses for the research of factors which describe Californian road scenery. 3. Results The characteristics of demographic, financial and marital status of respondents correlate with variables describing road landscape very weakly: the highest correlation coefficients are rtau_b=0.178 (p=0.000<α=0.05) and rs=0.190 (p=0.000<α=0.05) at the significance level of 0.01. These correlations are between the variable outstanding road landscape and the marital status of respondents. Therefore, the conclusion is that demographic and financial characteristics of respondents as well as their marital status do not affect respondents’ opinion and assessment of road landscape views. The analysis of the Kendall’s tau_b correlations between road landscape describing variables demonstrate that almost all the describing variables correlate to each other. The exception is the variables beautiful and sophisticated – these variables do not correlate at all rtau_b=0.000 (p=0.000<α=0.05). The most of the correlations are moderate, weak and very weak. The strongest correlation (though it is just a strong correlation) is between the variables interesting and skittish rtau_b=0.848 (p=0.000<α=0.05), majestic and left an intense positive impression rtau_b=0.818 (p=0.000<α=0.05), relaxing and willing to drive rtau_b=0.758 (p=0.000<α=0.05), interesting and left an intense positive impression rtau_b=0.727 (p=0.000<α=0.05) all at the significance level of 0.01. There are some opposite very weak correlations between the variables visually safe and elements match for surrounding environment, visually safe and beautiful, outstanding and elements match for surrounding environment, harmonious and sophisticated, sophisticated and elements match for surrounding environment: the better assessment is for one variable the worse assessment is for another variable in the pair (table 1). The higher quantity of car accidents on the road correlates weakly with relaxation rtau_b=-0.303 (p=0.000<α=0.05), visual safety rtau_b=-0.212 (p=0.000<α=0.05), beauty rtau_b=0.364 (p=0.000<α=0.05), sophistication rtau_b=-0.333 (p=0.000<α=0.05) and elements match for surrounding environment rtau_b=0.273 (p=0.000<α=0.05) (table 1). It is interesting that the more road landscape is relaxing, visually safe and sophisticated the less car accidents happen. The analysis of the Spearman’s rho correlations between road landscape describing variables demonstrate that almost all the describing variables correlate to each other. The most of the correlations are strong, moderate, weak and very weak. The strong correlation is between the variables skittish and interesting rs=0.951 (p=0.000<α=0.05), majestic and left an intense positive impression rtau_b=0.930 (p=0.000<α=0.05), relaxing and willing to drive rs=0.888 (p=0.000<α=0.05), interesting and left an intense positive impression rtau_b=0.881 (p=0.000<α=0.05), skittish and left an intense positive impression rtau_b=0.853 (p=0.000<α=0.05), skittish and outstanding rtau_b=0.846 (p=0.000<α=0.05), outstanding and left an intense positive impression rtau_b=0.839 (p=0.000<α=0.05), pleasant and beautiful rtau_b=0.813 (p=0.000<α=0.05), interesting and outstanding rtau_b=0.811 (p=0.000<α=0.05), harmonious and elements match for surrounding environment rtau_b=0.804 (p=0.000<α=0.05), beautiful and elements match for surrounding environment rtau_b=0.776 (p=0.000<α=0.05), majestic and willing to drive rtau_b=0.762 (p=0.000<α=0.05), pleasant and skittish rtau_b=0.760 (p=0.000<α=0.05), pleasant and interesting rtau_b=0.760 (p=0.000<α=0.05), majestic and interesting rtau_b=0.741 (p=0.000<α=0.05), majestic and skittish rtau_b=0.734 (p=0.000<α=0.05), relaxing and visually safe rtau_b=0.720 (p=0.000<α=0.05), pleasant and willing to drive rtau_b=0.718 (p=0.000<α=0.05) all at the significance level of 0.01, left an intense positive impression and willing to drive rtau_b=0.706 (p=0.000=α=0.05) at the significance level of 0.05. There are some opposite weak and very weak correlations between the variables visually safe and beautiful, visually safe and elements match for surrounding environment, outstanding and elements match for 23 surrounding environment, harmonious and sophisticated, beautiful and sophisticated, sophisticated and elements match for surrounding environment: the better assessment is for one variable the worse assessment is for another variable in the pair (table 2). The hihger quantity of car accidents on the road correlates weakly with pleasure rtau_b=0.214 (p=0.00<α=0.05), relaxation rtau_b=-0.406 (p=0.00<α=0.05), visual safety rtau_b=-0.315 (p=0.00<α=0.05), harmony rtau_b=0.217 (p=0.00<α=0.05), naturalness Kendall’s tau_b correlation coefficient values pleasant relax- ing safe skittish interes- tin outstandi harmo- niou majestic natural beauti- ful sophis- ticat left posit impr elemen mat willin to drive car ac- cidents pleasant 1.00 .351 .076 .595** .626** .382 .473* .473* .504* .687* .137 .473* .534* .595** .168 relaxing .351 1.00 .545* .152 .303 .242 .394 .515* .030 .091 .364 .394 .242 .758** -.303 safe .076 .545* 1.00 .121 .212 .030 .182 .364 .121 -.182 .394 .364 -.030 .485* -.212 skittish .595** .152 .121 1.00 .848** .667** .091 .576** .455* .394 .424 .697** .121 .394 .061 interestin .626** .303 .212 .848** 1.00 .636** .182 .606** .545* .364 .394 .727** .152 .424 -.030 outstandi .382 .242 .030 .667** .636** 1.00 .061 .545* .182 .303 .394 .667** -.030 .364 .091 harmonio .473* .394 .182 .091 .182 .061 1.00 .091 .273 .394 -.061 .152 .667** .394 .121 majestic .473* .515* .364 .576** .606** .545* .091 1.000 .273 .273 .364 .818** .061 .576** .000 natural .504* .030 .121 .455* .545* .182 .273 .273 1.00 .576** .182 .394 .303 .212 .121 beautiful .687** .091 -.182 .394 .364 .303 .394 .273 .576** 1.00 .000 .273 .606** .333 .364 sophisticat .137 .364 .394 .424 .394 .394 -.061 .364 .182 .000 1.00 .424 -.152 .364 -.333 left posit impr .473* .394 .364 .697** .727** .667** .152 .818** .394 .273 .424 1.00 .061 .576** .000 elements matc .534* .242 -.030 .121 .152 -.030 .667** .061 .303 .606** -.152 .061 1.00 .364 .273 willing to drive .595** .758** .485* .394 .424 .364 .394 .576** .212 .333 .364 .576** .364 1.00 -.061 car accidents .168 -.303 -.212 .061 -.030 .091 .121 .000 .121 .364 -.333 .000 .273 -.061 1.00 **. Correlation is significant at the 0.01 level *. Correlation is significant at the 0.05 level Strong relation Moderate relation Weak relation Very weak relation Table 1. Kendall’s tau_b correlation coefficient values Spearman’s rho correlation coefficient values pleas- ant relax- ing safe skittish interes- tin out- standi harmo- niou majes- tic natu- ral beauti- ful sophisti- cat left posit impr elemen mat willin to drive car ac- cidents pleasant 1.00 .410 .175 .760** .760** .504 .609* .616* .623* .813** .203 .669* .687* .718** .214 relaxing .410 1.00 .720** .259 .399 .308 .455 .615* .126 .077 .497 .531 .203 .888** -.406 safe .175 .720** 1.00 .140 .308 .098 .301 .483 .161 -.203 .573 .510 -.126 .595* -.315 skittish .760** .259 .140 1.00 .951** .,846** .140 .734** .545 .559 .552 .853** .126 .510 .098 interestin .760** .399 .308 .951** 1.00 .811** .294 .741** .615* .510 .594* .881** .168 .573 -.056 outstandi .504 .308 .098 .846** .811** 1.00 .077 .727** .266 .385 .476 .839** -.021 .483 .147 harmonio .609* .455 .301 .140 .294 .077 1.00 .091 .399 .455 -.189 .224 .804** .531 .217 majestic .616* .615* .483 .734** .741** .727** .091 1.000 .385 .399 .559 .930** .056 .762** -.063 natural .623* .126 .161 .545 .615* .266 .399 .385 1.00 .685* .245 .476 .448 .266 .238 beautiful .813** .077 -.203 .559 .510 .385 .455 .399 .685* 1.00 -.014 .406 .776** .399 .476 sophisticat .203 .497 .573 .552 .594* .476 -.189 .559 .245 -.014 1.00 .587* -.378 .448 -.413 left posit impr .669* .531 .510 .853** .881** .839** .224 .930** .476 .406 .587* 1.00 .056 .706* .056 elements matc .687* .203 -.126 .126 .168 -.021 .804** .056 .448 .776** -.378 .056 1.00 .399 .357 willing to drive .718** .888** .594* .510 .573 .483 .531 .762** .266 .399 .448 .706* .399 1.00 -.098 car accidents .214 -.406 -.315 .098 -.056 .147 .217 -.063 .238 .476 -.413 .056 .357 -.098 1.00 **. Correlation is significant at the 0.01 level *. Correlation is significant at the 0.05 level Strong relation Moderate relation Weak relation Very weak relation Table 2. Spearman’s rho correlation coefficient values 24 rtau_b=0.238 (p=0.00<α=0.05), beauty rtau_b=0.476 (p=0.00<α=0.05), sophistication rtau_b=-0.413 (p=0.00<α=0.05), and elements match for surrounding environment rtau_b=0.357 (p=0.00<α=0.05). It is also interesting that according to the Spearman’s rho correlation coefficient values we got the same result as according to the Kendall’s tau_b correlation coefficient values: the more road landscape is relaxing, visually safe and sophisticated the less car accidents happen. The application of the multiple linear regression analysis leads to one regression model. The quantity of car accidents is the dependent variable. According to the ANOVA and Coefficients tables prepared by the PASW Statistics software we find the point estimates for the regression equation. The statistical acceptance of the coefficients of the model (p-value shall not have to exceed α=0.05) was estimated. Then the unstandardized coefficient B as well as the variables, which influence the quantity of car accidents were identified. Quantity of car accidents = -33.93 + + 40.98*Beautiful –33.05*Interesting + + 32.78 * Willing to drive – – 29.97 * Elements match for surrounding environment – – 19.11*Relaxing + 14.8*Natural + 8.27*Harmonious – – 8.2*Sophisticated + 7.11*Majestic – 1*Visually safe (1) The linearity of the regression equation is approved (according to ANOVA p=0.000<0.05). The hypothesis that the coefficients are equal to zero was rejected (p=0.000<0.05), it means that the regression lines are suitable for making predictions. 4. Discussion Values of the both Kendall’s tau_b and Spearman’s correlation coefficients demonstrate weak or very weak relations between the variable Quantity of car accidents and other variables describing aesthetics of road landscape. This could happen due to not very detailed information about aesthetic properties of each road landscape. The collected data on aesthetic properties describe the landscape of the whole road instead of description of separate sections of each road. More over the regression equation explains only 50.3% of the dispersion of all the variables. Whereas the other very important factors which influence the quantity of car accidents remain unknown. On other hand, the regression equation contains too many independent variables (10 of 14), and the constant is very high (-33.93). In consideration of these facts, more detailed analysis should be carried out. The landscape of each road has to be divided into separate segments according to the landscape type. Then for each segment more detailed aesthetic properties has to be evaluated. Finally the data on landscape aesthetics has to be compared with the number and types of car accidents for each segment of the road. Also more detailed data on road landscape aesthetic properties will let us create the guidelines for the design of an aesthetic and safe road landscape, and to forecast potentially dangerous (unsafe) places of road landscape. 5. Conclusions The results of the correlation analysis (both Kendall’s tau_b and Spearman’s correlation coefficients evaluated) revealed that the road safety described through the quantity of car accidents is weakly related with some aesthetic properties of road landscape. Actually, the more landscape is pleasant, beautiful, harmonious, natural and elements match for surrounding environment the more car accidents happen, and the more road landscape is relaxing, visually safe and sophisticated the less car accidents happen. These aesthetic properties have to be considered by the planners while designing and creation of safe road landscape. According to the multiple linear regression analysis the more landscape is beautiful, willing to drive, natural, harmonious and majestic the more car accidents happen on the road, and the more road landscape is interesting, elements match for surrounding environment, relaxing, sophisticated and visually safe the less car accidents happen on the road. Still, the coefficients of the independent variables and the constant are too high, therefore the equation need to be revised in further research. (I. Gurauskiene, 2006, Eco-design methodology for electrical and electronic equipment industry) Acknowledgements The authors acknowledge Student Research Fellowship Award from the Lithuanian Science Council, as well as Lithuanian road administration under the Ministry of transport and communication for the provided information about car accidents on Lithuanian roads. References Antonson H., Mardch S., Wiklund M., Blomqvist G. 2009. Effect of surrounding landscape on driving behavior: A driving simulator study. Journal of Environmental Psychology, 29 (4), 493-502. Clay G. R., Smidt R. K. 2004. Assessing the validity and reliability of descriptor variables used in scenic highway analysis. Landscape and Urban Planning, 66 (4), 239-255. Clay G. R., Daniel T. C. 2000. Scenic landscape assessment: the effects of land management jurisdiction on public perception of scenic beauty. Landscape and urban planning, 49 (1-2), 1-13. Cook P. S., Cable T. T. 1995. The scenic beauty of shelterbelts on the Great Plains. Landscape and Urban Planning, 32 (1), 63-69. Čekanavičius V., Murauskas G. 2008. Statistika ir jos taikymai 2 [Statistics and its application 2]. Vilnius, TEV. Daniel T. C., Boster R. S. 1976. Measuring Landscape Esthetics: The Scenic Beauty Estimation Method. USDA Forest. Service, Rocky Mountain Forest and Range Experiment Station, Research Paper RM-167, Fort Collins. Drottenborg H. 2002. Are beautiful traffic environments safer than ugly traffic environments? Doctoral thesis. Lund University: Studentlitteratur. Available at: http://www.getcited.org/ pub/103430785 (accessed 16 May 2012). Edquist, J. 2008. The effects of visual clutter on driving performance. Doctoral thesis. Monash University. Available 25 at: http://www.tml.org/legal_pdf/Billboard-study-article.pdf (accessed 16 May 2012). Eismo įvykių statistika [Transport accident statistics], 2012. Lietuvos automobilių kelių direkcija prie Susisiekimo ministerijos. Available at: http://www.lra.lt/lt.php/eismo_ saugumas/eismo_ivykiu_statistika/27 (accessed 16 May 2012). I-15 landscape and aesthetics corridor plan. 2005. Nevada DOT. Available at: http://www.ndotprojectneon.com/FEIS/I-15_ Landscape_Corridor_Plan.pdf. (accessed 16 May 2012). Matijošaitienė, I. 2011. Automobilių kelių hedonomiško kraštovaizdžio formavimo principai. Daktaro disertacija [The principles of formation of hedonomic road landscape. Doctoral thesis]. Kaunas: Technologija. Matijošaitienė, I., Stankevičė, I. 2011. Hedonomic Roadscapes in the Context of Urban Sprawl. The Scientific Journal of Riga Technical University, 5, 70-76. Road Landscape Manual. Version 1. 1997. Queensland Government, Department of Main Roads. Available at: http:// www.tmr.qld.gov.au/Business-and-industry/Technical- standards-and-publications/Road-landscape-manual.aspx. (accessed 16 May 2012). Transport accident statistics, 2012. European Commision Eurostat. Available at: http://epp.eurostat.ec.europa.eu/statistics_ explained/index.php/Transport_accident_statistics. (accessed 16 May 2012). Received 2012 05 14 Accepted after revision 2012 09 03 Irina MATIJOŠAITINĖ – lecturer at Kaunas University of Technology, Faculty of Civil Engineering and Architecture, Department of Architecture and Land Management. Main research areas: road and urban landscape, space syntax, Kansei engineering, hedonomics, ergonomics, statistics. Address: Kaunas University of Technology, Faculty of Civil Engineering and Architecture, Studentu st. 48, LT-51367 Kaunas, Lithuania. Tel.: +370 606 88819 E-mail: irina.matijosaitiene@ktu.lt Kristina NAVICKAITĖ – bachelor student at Kaunas University of Technology, Faculty of Civil Engineering and Architecture. Main research areas: urban crime, road landscape. Address: Kaunas University of Technology, Faculty of Civil Engineering and Architecture, Studentu st. 48, LT-51367 Kaunas, Lithuania. Tel.: +370 624 39155 E-mail: kristina.navickaite.fsa.statius@gmail.com