A number of studies have researched the effects of tourism on transportation system, hotel industry, economic efficiency and environment. This paper examines the influence of the information and communication technologies development on the inbound tourism intensity. The correlation and regression analysis has been used to identify the relationship between the Travel and Tourism Competitiveness Index, the Information and Communication Technology Development Index and International tourism arrivals. The results demonstrate that there is a close link between the countries' tourist attractiveness and the level of their information and communication development. However, it is not equal for different countries, which are grouped by the level of intensity of tourism arrivals, the level of the country's attractiveness and its information and communication technologies development. Besides, the country's information and communication technologies development has little effect on the inbound tourism intensity
The contemporary unification of the world society into a single information and communication network and the transformation of information technologies into the generative force of socio-economic development contributes to the close interconnection between countries, regions and societies of different nations. Most countries are actively using Internet space to shape the country’s image (including tourist image). Mainly because it is one of the most important tools for creating an appropriate image of the country as a tourist destination that can greatly increase the intensity of the inbound tourism flow. In this regard, the study of the impact of the country’s information and communication technologies development on the tourism and travel development is relevant and requires further theoretical and practical research.
Tourism development affects the development of a range of other areas of economic activity. Inbound and outbound tourism has bidirectional causality with air transportation (Syed Abdul Rehman Khan and other, 2017). Air transport and tourism are highly connected. Researches show that tourist-oriented airports may achieve higher efficiency levels than non-touristic ones (Xosé Luis Fernández, Pablo Coto-Millán, Benito Díaz-Medina, 2018).
The level of tourism development is estimated by the global index. The Travel & Tourism Competitiveness Index has been the subject of some methodological criticism, such as the arbitrary weighting of variables. There is an alternative methodology for calculating this index based on two points of reference to propose a new standardization. А synthetic index that measures the state of the pillar in the worst position, as well as other alternative indices, is calculated (Juan Ignacio Pulido-Fernández, Beatriz Rodríguez-Díaz, 2019).
Depending on how the variables are included in the underlying technology specification, the same tourism index can be oriented towards the assessment of either the private or the public sector’s effectiveness (Walter Briec and other, 2018).
Trade openness, climate change and intensity of market competition increase tourism efficiency in China. Tourism efficiency improvement in China was mainly driven by technological improvement (Sami Chaabouni, 2019).
UNESCO’s World Heritage inscription is considered to positively influence tourism demand. However, relevant econometric research has yielded inconsistent results. A sub-group analysis identifies different factors in developing vs. developed countries and cultural vs. natural WHS types. (Yang LanXuе, Thomas E.Jones, 2019).
The expansion of tourism translates into an environmental deterioration of the destination (risk dimension) and, furthermore, it substantiates that there are specific variables connected to environmental sustainability (regulatory dimension) that contribute to greater tourism growth, so that the relationship between tourism and environmental sustainability is bidirectional (Juan Ignacio Pulido-Fernández, Pablo Juan CárdenasGarcía, Juan Antonio Espinosa-Pulido 2019).
The studies suggest that the effect of growth rate of total foreign tourist arrivals on hotel equity return is asymmetric and state-dependent, conditional on the distributions of hotel equity return. The study further identified that GTA has a significant influence only on equity returns of hotels with a small size (Ming-Hsiang Chen, 2016).
The shift of our view on information technology in tourism research from a primarily a marketing-driven tool to a knowledge creation tool due to new technological conditions such as the smartphone, drone, wearables, new connectivity and big data is recognized. Some possible future research problems and challenges regarding our existing views of the relationship between information technology and tourism are studied (Zheng Xiang, 2018).
Not only ICTs empower consumers to identify, customise and purchase tourism products but they also support the globalisation of the industry by providing effective tools for suppliers to develop, manage, and distribute their offerings worldwide (Buhalis, 1998).
Buhalis (1998) stated that potential tourists have become moreindependent and sophisticated on using a wide range of tools to arrange for their trips (such as Expedia, Googleand Kayak, visitbritain.com), web 2.0portals, wayn and tripadvisor, kelkoo).
Information Search is a significant part of the purchase decision process and was revolutionised as a result of the Internet. ICTs not only reduce uncertainty and perceived risks but also enhance the quality of trips (Fodness &Murray, 1997).
The quality of the website, Digital Marketing, Social Networking, Multimedia, Mobile Technologies and Intelligent Environments are the main keys factors of ICT in Tourism (Elisabete Paulo Morais & other, 2016).
A Virtual Travel Community (VTC) makes it easier for people to obtain information, maintain connections, develop relationships, and eventually make travel-related decisions (Stepchenkova, Mills & Jiang, 2007).
Increasingly the impacts of ICTs are becoming clearer, as networking, dynamic interfaces with consumers and partners and the ability to re-develop the tourism product proactively and reactively are critical for the competitiveness of tourism organizations (Buhalis, D., & Law R., 2008).
The analysis of the mentioned resources has allowed hypothesizing the following: Hypothesis 1. Information development of the society contributes to the improvement of the country’s tourist attractiveness.
Hypothesis 2. The development of information and communication technologies in the countries across the globe positively influences the inbound tourism intensity. 3
The methods of multivariate statistical analysis, such as Descriptive Statistics, the multiple regression, the cluster analysis were used to study the influence of information and communication technologies on tourism. These statistical methods were implemented with the StatSoft's software package Statistica. This package is well balanced with the “power / convenience ratio”, has a wide range of functional data analysis algorithms and has wide graphical capabilities for data visualization.
To carry out the research, the global indices and variables of tourism development were selected:
The Travel and Tourism Competitiveness Index (TTCI), which reflects the level of the country’s attractiveness for both tourists and also investors and representatives of the tourism business. This index includes the characteristics of the following framework: Enabling Environment, Travel and Tourism Policy and Enabling Conditions, Tourism and Transport Infrastructure, Natural and Cultural Resources [12];
The Information and Communication Technologies Development Index (ICT) reflects the level of networked infrastructure and access to ICTs, the level of use of ICTs in the society and more efficient and effective ICT use [11].
International tourism arrivals (ITA) is one of the main indicators that reflects the effectiveness of all the measures adopted for the development of tourism in the country [3].
The objects of research are 80 countries of the world. The variables are the data for 2016. The countries without sufficient data were excluded from the database. 4
In order to study the influence of the country’s information and communication development on tourism development, the following algorithm of the research has been proposed:
Stage 1. Selection of the initial variables.
Stage 2. Research of the basic statistical characteristics of the selected variables.
Stage 3. Verification of the first hypothesis on the basis of the correlationregression analysis methods.
Stage 4. Verification of the second hypothesis on the basis of the correlationregression and cluster analysis methods for the whole array of initial data and within the scope of separate groups of countries, which are similar according to the level of tourism activity.
For implementation of the first stage of the algorithm, the following variables were selected: Travel and Tourism Competitiveness Index (TTCI), Information and Communication Technologies Development Index (ICT) and International Tourism Arrivals (ITA).
The descriptive statistics was used to process, systematize and provide quantitative description of the empirical data by means of the main statistical indicators. The implementation of the second stage of the study presupposed the calculation of the following characteristics: Mean, Median, Mode, Frequency of Mode, Minimum, Maximum, Variance, Standard Deviation, Coefficient of Variation, Skewness, Kurtosis, as well as histograming. The results of calculation are presented in Table.1. 3,0 3,5 4,0 4,5 5,0
5,5
X <= Category Boundary -10000 0
According to the results of the analysis of the obtained statistical characteristics (see Table 1) and distribution histograms (see Figures 1-3), the following conclusions can be made: - the TTCI variable has a distribution close to normal. This is evidenced by the proximity of the mean, mode and median, as well as small values of the skewness and kurtosis. This variable has the least value of the coefficient of variation (14,52); - the ICT variable also has a distribution rather close to normal. Its average value is close to the median. But unlike the TTCI variable, it has a slightly larger range (from 3.03 to 8.98) and a larger coefficient of variation (24.95). It should be noted that almost half of the world’s countries (38 out of 80) fall into the last two intervals with values of 7 to 8 and from 8 to 9. This means that a significant part of the countries in the considered group has a high level of development of information and communication technologies (ICT); - the ITA variable is significantly different from the previous two. First of all, it has a completely different unit of measurement and dimension, therefore, during further research with the simultaneous use of the TTCI and ICT variables, the calculations will be made on the basis of the standardized data. Secondly, the distribution of this variable is quite distant from normal. This is evidenced by the large difference between the mean and the median (12 684.2 and 5 460.0 thousand persons respectively), as well as the statistical criteria of the Kolmohorov-Smirnov (K-S test), Shapiro-Wilk test and Lillifors test. For the distribution of this variable the right-side bias is characteristic (the skewness equals 2.37) and significant elevation (the kurtosis is equal to 5.69). In 2016, this variable was significant (from 121 to 82,600 people), more than 50 countries had the value of this variable up to 10,000 thousand people, in France this value was more than 80,000, and in Spain and United States – it ranged from 70,000 to 80,000 thousand people.
The verification of the first hypothesis that the information development of the society contributes to the improvement of the country’s tourist attractiveness was carried out during the implementation of the third stage of the study. A pair correlation coefficient between the TTCI and ICT variables was calculated according to the data from all 80 countries. In 2016 it was equal to 0.711 that indicates a fairly close direct linear relationship between these variables. The graphic representation of this connection is given in Fig. 4.
The dispersion field proves a linear relationship between TTCI and ICT, therefore, we can accept the hypothesis 1 that the greater the information development of the country is, the better the tourism and travel sector is developed in this country. This allows putting forward the second hypothesis that the development of information and communication technologies in the countries of the world positively influences the intensity of inbound tourism.
Thus, the implementation of the fourth stage involves verification of the second hypothesis that involves the following sequence of steps:
Step 1. Determination of the pair correlation coefficients between the resulting ITA variable and factor variables of TTCI and ICT.
Step 2. Construction of the multiple regression based on the standardized data of the type: where a1 та a2 are unknown parameters that are evaluated by the least squares method.
Step 3. Distribution of countries into homogeneous groups according to the ITA,
TTCI and ICT variables on the bassis of the cluster analysis methods.
Step 4. Construction of the multiple regression (
In the result of implementation of the first step the following values of the pair correlation coefficients have been received: ; . These values mean that there is a direct linear link of the moderate level between ITA and TTCI, but between ITA and ICT there is a direct but weak link.
The obtained values are not sufficient for accepting or rejecting the hypothesis 2. Therefore, during the second step, the following equation of multiple regression was constructed:
This regression equation is statistically significant in terms of Fisher’s criterion (F = 40.32), and separate parameters according to Student’s criterion ( , ). The coefficients of the multiple correlation (R = 0.713), the determination ( )) and the corrected determination coefficient ( ) indicate a sufficient quality of the model. There is no autocorrelation of the errors in this model (the statistics of Darbine-Watson are approximately equal to 2, and the cyclic coefficient of autocorrelation is close to 0). Thus, this model can be used for analysis and forecasting.
We have analysed the problem under study according to this model. As can be seen from the obtained regression equation, compared with the pair correlation coefficients, there is a significant increase of the influence of TTCI on ITA (from 0.6574 to 0.9293), the simultaneous change of direction and the increase of the influence of ICT on ITA (from + 0.2734 to -0.3915). To answer the question whether these changes are only due to the multicollinearity that are present in the model, or in fact there is an inverse relationship between ICT and ITA, partial correlation coefficients have been calculated and their statistical significance checked. The results of calculations are given in Table 2.
As can be seen from Table 3, the value of the partial coefficients is: and they are statistically significant according to the
During the next, third step, using the cluster analysis methods, we obtain homogeneous groups of countries. The grouping of countries is based on the hierarchical method of full communication, which allows clearly divide the countries into two, three, or four clusters. The division into 2 clusters is not informative. If we divide countries into 4 clusters, then the last cluster will consist of only three countries (28, 69 and 77). Therefore, it is rational to divide countries into three clusters, which corresponds to the logical distribution of countries with high, medium and low intensity of foreign tourists’ arrivals.
Based on the iterative method of clustering k-means, the following cluster results have been obtained. The first cluster includes 30 countries with the low inbound tourism activity (Cluster contains 30 cases). These countries are listed in Table 3. C_25 SadElovlra- 0,4453 C_53 Mneognrtoe- 0,585 C_80 Vietnam 0,348
Besides, these countries have a very low level of tourism potential (TTCI), information, and communication technologies development (ICTs). Under the current conditions of certain instability, Ukraine is referred to this group.
The second cluster includes 38 countries with an average level of inbound tourism activity (Cluster contains 38 cases) and is presented in Table 4.
Countries in the cluster 2 are characterized by the highest level of tourism potential (TTCI) and the average level of ICT development.
The third cluster includes 12 countries with the highest level of inbound tourism activity (Cluster contains 12 cases). This claster is presented in Table 5. C_28 C_30 C_40 C_42
1,230744 0,654982 0,171680 0,932930 C_73 C_74 C_76 C_77
0,926716 0,916710 0,678250 0,968470
Countries of the latter cluster are characterized by the highest level of tourism potential development (TTCI) and rather high level of ICT.
It should be noted that the received country grouping by the level of tourism activity is sustainable, since the hierarchical method of complete dependence and the kmedium method yielded identical results, except for the country number 78 (Uruguay) which, according to the first method, was reffered to the low-income countries development, and, according to the second method, it was reffered to countries with an average level of development. As the final result we accept the one that gives the kmedium method, since this method minimizes intragroup variance and maximizes the intergroup, thus providing higher-quality clusterization.
The average means of the variables, according to which the clusterization was carried out, are presented in Fig. 6
Analysis of the means given in Fig. 6, allows to draw the following conclusions. First, the cluster number 1 is formed by the countries with the lowest values of the TTCI, ICT and ITA variables. The second cluster consists of the countries with an average level of TTCI and ITA, but with the highest level of ICT. The third cluster includes the countries with the highest levels of TTCI and ITA with high (but not the highest) level of ICT. In addition, the countries of the second and third clusters are characterised by the inverse dependence between the ITA and ICT variables. This is also confirmed by the calculation of the pair correlation coefficients between the variables for each cluster separately. The results of calculations are presented in the Table 6. Let us analyze the means of the obtained coefficients in more detail.
Thus, cluster 1, in comparison with other clusters, is characterised by the strongest direct dependence between TTCI and ITA (0.4156) and the inverse dependence between ICT and ITA (-0.1822). Besides, the countries within this cluster have almost no dependence between TTCI and ICT (-0,0763), that is, for countries of this group we reject the hypothesis 1.
The absence of multicollinearity between the TTCI and IST factors enables to
construct a two-factor regression model (
This regression equation is statistically significant in general according to Fisher’s criterion (F = 3.395, Significance F = 0.048). According to Student’s criterion, only the influence of the TTCI factor ( ) is statistically significant, whereas the influence of ICT is not statistically significant ( ). The coefficients of the multiple correlation (R = 0.442), the determination ( ) and the corrected determination coefficient ( ) indicate insufficient quality of the model. Thus, on the basis of the above stated, for the countries with the low level of tourism activity the hypothesis 2 is rejected.
Cluster 2 is characterised by a weak dependence between TTCI and ICT (0.4883), as well as between TTCI and ITA (0.3444). There is a weak inverse dependence between the ICT and ITA values (-0.1534), that is, for countries of this group we accept the hypothesis 1.
The lack of multicollinearity between the TTCI and ICT factors within the cluster
1 allows to construct a two-factor regression model (
This regression equation is statistically significant in terms of Fisher’s criterion (F = 6.143, Significance F = 0.005). According to Student’s criterion, the influence of both factors is statistically significant: і ТТСІ ( ), і ІСТ ( ). The coefficients of the multiple correlation (R = 0.504), the determination ( ) and the corrected determination coefficient ( ) indicate insufficient quality of the model. Thus, on the basis of the above stated, the hypothesis 2 for the countries with the average level of tourism activity is rejected.
Cluster 3 is characterised by a weak dependence between TTCI and ITA (0.33931), as well as by a very weak dependence between ICT and ITA (0.1444). There is a strong direct dependence between TTCI and ICT (0.8328), that is, for the countries of this group we accept the hypothesis 1.
The presence of strong multicollinearity between the factors of TTCI and ICT does
not allow to construct a two-factor regression model (
Thus, on the basis of the aforementioned, the hypothesis 2 for the countries with the high level of tourism activity is rejected.
Let us analyze the means of the obtained coefficients in more detail.
Thus, cluster 1, in comparison with other clusters, is characterised by the strongest direct dependence between TTCI and ITA (0.4156) and the inverse dependence between ICT and ITA (-0.1822). Besides, the countries within this cluster have almost no dependence between TTCI and ICT (-0,0763), that is, for countries of this group we reject the hypothesis 1.
The absence of multicollinearity between the TTCI and IST factors enables to
construct a two-factor regression model (
This regression equation is statistically significant in general according to Fisher’s criterion (F = 3.395, Significance F = 0.048). According to Student’s criterion, only the influence of the TTCI factor ( ) is statistically significant, whereas the influence of ICT is not statistically significant ( ). The coefficients of the multiple correlation (R = 0.442), the determination ( ) and the corrected determination coefficient ( ) indicate insufficient quality of the model. Thus, on the basis of the above stated, for the countries with the low level of tourism activity the hypothesis 2 is rejected.
Cluster 2 is characterised by a weak dependence between TTCI and ICT (0.4883), as well as between TTCI and ITA (0.3444). There is a weak inverse dependence between the ICT and ITA values (-0.1534), that is, for countries of this group we accept the hypothesis 1.
The lack of multicollinearity between the TTCI and ICT factors within the cluster
1 allows to construct a two-factor regression model (
This regression equation is statistically significant in terms of Fisher’s criterion (F = 6.143, Significance F = 0.005). According to Student’s criterion, the influence of both factors is statistically significant: і ТТСІ ( ), і ІСТ ( ). The coefficients of the multiple correlation (R = 0.504), the determination ( ) and the corrected determination coefficient ( ) indicate insufficient quality of the model. Thus, on the basis of the above stated, the hypothesis 2 for the countries with the average level of tourism activity is rejected.
Cluster 3 is characterised by a weak dependence between TTCI and ITA (0.33931), as well as by a very weak dependence between ICT and ITA (0.1444). There is a strong direct dependence between TTCI and ICT (0.8328), that is, for the countries of this group we accept the hypothesis 1.
The presence of strong multicollinearity between the factors of TTCI and ICT does
not allow to construct a two-factor regression model (
Thus, on the basis of the aforementioned, the hypothesis 2 for the countries with the high level of tourism activity is rejected. 5
Thus, 1) the tourist attractiveness of the country increases if there are developed information and communication technologies, because the quality of tourist information resources (including Internet resources), formation of the comfortable information environment, the mass use of platforms for travel services searching and comparing of their prices, the development of e-commerce in tourism in general contributes to the improvement of the tourism infrastructure;
2) there is no dependence of this type between tourism attractiveness and the development of information and communication technologies in the first group of countries, which is characterized by low TTCI, ICT and low intensity of tourist arrivals. There is a significant dependence in the 2nd group of countries. There is a strong direct dependence in the third group, characterised by the highest level of tourist arrivals and the highest level of tourist attractiveness. That is, the more the country is attractive for tourism, the stronger is the interdependence between the indicators of TTCI and ICT;
3) there is a weak inverse dependence between the countries’ information and communication technologies development and the intensity of tourist arrivals. Moreover, this connection is not observed in any of the 3 distinct groups of countries. That is, it can be argued that the development of information and communication technologies almost does not affect the intensity of inbound tourism. 3. Fodness, D., & Murray, B.: Tourist information search. Annals of Tourism Research 24 (3), 503-523 (1997). 4. Juan Ignacio Pulido-Fernandez, Beatriz Rodriguez-Diaz: Reinterpreting the World Economic Forum's global tourism competitiveness index. Tourism Management Perspectives, vol. 20, 131-140 (2016). 5. Juan Ignacio Pulido-Fernández, Pablo Juan Cárdenas-García, Juan Antonio EspinosaPulido: Does environmental sustainability contribute to tourism growth? An analysis at the country level. Journal of Cleaner Production. Vol. 213, 309-319 (2019). 6. Elisabete Paulo Morais, Carlos Rompante Cunh, João Paulo Sous: Information and communication technologies in tourism: Challenges and trends. Conference: Proceedings of the 27th International Business Information Management Association Conference - Innovation Management and Education Excellence Vision 2020: From Regional Development Sustainability to Global Economic Growth, IBIMA. Italy. (2016) Homepage, https://www.researchgate.net/publication/316830871_ Information_ and_communication_technologies_ in_tourism_Challenges_and_trends. 7. Ming-Hsiang Chen: A quantile regression analysis of tourism market growth effect on the hotel industry. International Journal of Hospitality Management. Volume 52, 117-120 (2016). 8. Syed Abdul Rehman Khan Dong Qianli, Wei SongBo, Khalid Zaman, Yu Zhang: Travel and tourism competitiveness index: The impact of air transportation, railways transportation, travel and transport services on international inbound and outbound tourism. Journal of Air Transport Management, vol. 58, 125-134 (2017). 9. Sami Chaabouni: China's regional tourism efficiency: A two-stage double bootstrap data envelopment analysis. Journal of Destination Marketing & Management. Vol. 11, 183-191 (2019). 10. Stepchenkova, S., Mills, J. E., & Jiang, H.: Virtual travelcommunities: Self-reported experiences and satisfaction. In M. Sigala, L. Mich, & J. Murphy (Eds.). Information and communication technologies in tourism 2007. New York: Springer Wien. 163-174 (2007). 11. The ICT Development Index 2017 Homepage, https://www.itu.int/net4/ITU
D/idi/2017/index.html#idi2017rank-tab. 12. The Travel & Tourism Competitiveness Report 2017 Homepage, https://www.weforum.org/reports/the-travel-tourism-competitiveness-report-2017. 13. Walter Briec, Kristiaan Kerstens, Diego Prior, Ignace Van de Woestyne: Testing General and Special Fдre-Primont Indices: A Proposal for Public and Private Sector Synthetic Indices of European Regional Expenditures and Tourism. European Journal of Operational Research. European Journal of Operational Research, Volume 271, Issue 2, 756-768 (2018). 14. Xosé Luis Fernández, Pablo Coto-Millán, Benito Díaz-Medina: The impact of tourism on airport efficiency: The Spanish case. Utilities Policy, Volume 55, 52-58 (2018). 15. Yang LanXuе, Thomas E.Jones: Tourism-enhancing effect of World Heritage Sites: Panacea or placebo? A meta-analysis. Annals of Tourism Research. Vol. 75, 29-41 (2019). 16. Zheng Xiang: From digitization to the age of acceleration: On information technology and tourism. Tourism Management Perspectives. Volume 25, 147-150 (2018).