Comparing Tourism Activity Patterns Influenced by a Tourism Information Source: A Case of the Gyeonggi Province, South Korea

Abstract

With the development of the internet and mobile devices, tourists have been able to obtain information and then travel to various places or participate in diverse tourism activities in one travel excursion. This study investigates the patterns of tourist participation in tourism activities according to a tourism information source used through a relational approach using a social network analysis (SNA). This study utilizes raw data from 2021 Gyeonggi tourist survey data distributed by the Gyeonggi Tourism Organization. Our results indicate different patterns of participation in tourism activities between tourists using online sources and those using offline sources. In addition, the tourism activity distribution patterns of the centralities were hierarchically structured and differentiated depending on the tourism information source. The present study then offers theoretical implications and future study avenues, as well as practical implications, focused on developing tourism marketing strategies and content for the Gyeonggi province, South Korea.

1. Introduction

Most tourists decide on their destination first when planning a tour and prioritize searching for information about their goal destination. After deciding where to go, the second step is usually to search for information about the potential destinations, which can be found using many information sources [1]. This is because the search for destination information influences decisions on overall tourism, such as participation in tourism activities, selection of tourist destinations, transportation, tourism consumption, and accommodation, based on the tourism information searched and collected from various information sources [2]. Tourism information plays an important role not only in the tourism planning process but also in tourists’ activities at their tourism destination [3]. Therefore, based on their expectations and images of tourist destinations, tourists choose tourist destinations with a positive and differentiated image that can best satisfy their travel motives and needs through an active and intensive tourism information search [4]. In addition, many scholars have proposed that tourism information directly and indirectly affects tourists’ tourism satisfaction and future behavioral intentions [4,5,6,7].

Prior to the development of information technology, most tourists acquired tourism information mainly through offline sources such as relatives/family/colleagues, mass media and print media (e.g., TV, radio, newspaper, travel-related books), travel agencies, and personal experiences [8]. However, the rapid development of information and communication has a lot of influence on the travel field. In particular, due to the recent changes and development of the Internet environment, not only has it become very easy to collect travel information or establish travel plans unlike in the past, but new travel methods such as virtual travel are emerging [9]. In other words, online sources of information profoundly impact tourists tourism information search behaviors by providing them with a platform to rapidly share an abundance of tourism information and experiences [10,11]. For example, social media is an important information source for the tourism sector and different types of travelers. In particular, YouTube, the main video-sharing solution in the online environment, provides travelers with a new custom of viewing video materials and selecting destinations, and thus, YouTube is important for the development of the tourism industry [11,12].

However, the increased dependence on online information sources has caused a flood of tourist information to be provided to tourists, and problems such as exposure to privacy, the uncertainty of information, and increased efforts to search for information are occurring [13]. For this reason, certain potential tourists are reluctant to search for tourism information using online information sources and instead use offline information sources more than online information sources [14]. In addition, according to several researchers who have researched tourism information sources and tourism behavior, participation in tourism activities, the selection of tourist destinations, and the setting of tour routes vary according to the tourism information sources used by tourists [9,11,12,13,14]. Through these preceding studies, it can be seen that tourists rely on tourism information collected in advance when selecting the type of tourism activity to participate in, choosing a destination, and setting a travel route. Therefore, it can be regarded that the type of tourist participation in tourism activities is determined according to the information obtained from various sources.

Meanwhile, with the development of the Internet and mobile devices, tourists have been able to obtain information and travel to various places or participate in various tourism activities in one travel excursion [15,16]. Tourists’ participation in tourism activities shows various patterns due to constraints such as the type of tourism information, tourists’ preferences, purpose, and time and distance. These patterns can be further subdivided due to the advancement of information and communications technology (ICT) and changes in the social environment in the future [17]. Understanding the patterns of tourists’ participation in tourism activities is important because it can improve the competitiveness of the tourism market, such as through the development of tourism products and related tourism content and the establishment of tourism marketing strategies. In addition, tourists’ participation patterns are important because they can affect overall management, such as the operation of tourism destinations, improvement of the quality of tourism services, and future action intentions [4,17]. To understand the importance of tourists’ participation patterns in tourism activities, a deeper understanding of the tourism behavior of tourists who participate in tourism activities is required, and there is a need for vigorous research on this subject. However, there is a lack of research on how the participation patterns of tour guides change depending on whether the tourism information source is online or offline. Most initial studies have focused on verifying the influence of the relationship between the tourism information source and participation in tourism activities (e.g., [11,17,18]).

This study investigates tourists’ patterns of participation in tourism activities according to the tourism information source used through a relational approach using social network analysis (SNA). Specifically, first, a density analysis was performed, and the connection patterns between the tourism activity types included in the online and offline information source networks and the basic characteristics of the networks were identified. Second, a bootstrap paired sample t-test was performed to statistically verify the similarity of the connection patterns between the two networks. Third, through degree and eigenvector centrality analysis, the current study identifies which tourism activity is located at the center of the network and what kind of connection it demonstrates. Lastly, the present study investigates the simultaneous participation rate among tourism activity types. The significance of this study is that it will contribute to the development of a research model that can be applied to follow-up studies related to participation in tourism activities in the future in a situation where ICT is constantly changing. In addition, this study provides basic data for the development of tourism content based on the tourism information source by clarifying the types of tourist participation in tourism activities according to the information source used.

2. Literature Review

2.1. Tourism Information Source

Tourism information is used as basic data for potential tourists to establish tourism-related plans (transportation, accommodation, tourism products, images of tourist destinations, selection of tourist destinations, etc.). In the past, simple tourism information was recognized as information provided to help tourists make decisions [8], but the function and scope of tourist information are expanding because of the complexity of the tourism environment and the diversity of tourism needs [4]. In particular, with the rapid development of ICT and the advent of the Internet and the world wide web, the physical barriers of space and time have disappeared, allowing the tourism industry to connect with consumers online and provide tourist information online [9]. The number of Internet users worldwide is over 5 billion, accounting for 63.1 percent of the worlds’ population. Of these users, 4.7 billion people, or 59% of the worlds’ population, are reported to be using social media, and the Internet is one of the most influential technologies changing tourist behaviors [19]. The Internet is a common means of obtaining a wide range of information, and tourists can directly access online tourism information to learn about destinations, tourist attractions, restaurants, accommodations, shopping, etc. [20]. With this huge amount of information available, Internet searches have become an increasingly dominant tool for travelers to find travel destination information [21,22].

The accelerated evolution of ICT has led to the development of social media networks, blogs, and YouTube, providing diverse content directly to tourists visiting tourist destinations [4,9]. In addition, this social tourism information platform contains various types of content produced by consumers, as well as by companies, and is widely used by tourists who want to obtain tourism information on the Internet [23]. In particular, location-based social networks (LBSNs) allow consumers to connect and share information and their locations through content, including individual photos, text messages, and videos. LBSN is a feature that tracks check-ins in real time based on your location [9]. In other words, online social media (e.g., Twitter, Facebook, blogs, Tumblr, etc.), containing various types of user-produced content based on a social network that understands the interdependent relationships of individuals or groups as a network, has established itself as a major means of transporting and acquiring tourism information [23]. As such, LBSNs, which recently emerged through the development of ICT, have transformed the process of searching for tourism information and establishing and determining tourism plans.

2.2. Tourism Activity

Tourism is an on-site activity that a tourist participates in that occurs at a tourist destination [24]. Therefore, tourism activities may differ depending on individual preferences or on tourism purposes, preferences, needs, and motives, and the composition or characteristics of tourism resources at tourist destinations may also impact tourism activities [25]. In addition, when tourists participate in tourism activities, various travel patterns appear because of constraints such as their travel purpose, preferred travel style, time, and distance, and these travel patterns may be further subdivided due to future IT technology advancements and social environment changes [17,26].

Understanding tourists’ participation patterns in tourism activities is important because it can improve the tourism market’s competitiveness, such as through developing tourism products and linked tourism content and establishing tourism marketing strategies. Additionally, tourism patterns should be considered important because they can affect overall management, such as the operation of tourist attractions, the quality of tourism services, and predicting future tourists’ behavioral intentions [4,17,26]. Therefore, tourism activity patterns will diversify and develop over time. Previous studies have shown that tourists’ patterns of tourism activities are volatile and can be easily affected by various external factors, relationships between tourist destinations, and a series of intertwined factors [17,27]. In addition, recently, tourists have been increasingly inclined to make multi-purpose trips in which they engage in many tourist activities to experience a variety of experiences in one trip. By visiting various tourist destinations and participating in diverse tourism activities, the utility value of tourism activities can be maximized and satisfaction with travel can be increased [28]. In other words, tourists are expanding their tourism destinations because of various and abundantly collected tourism information, and at the same time, they visit several destinations in one trip and participate in multiple tourism activities [4].

As such, various tourism activities have changed from being supplier-centered to consumer-centered due to the development of ICT and transportation. Therefore, it is increasingly important to develop tourism content, revitalize local tourism, and improve the quality of tourism at a time when individual tourists are increasing, as tourists directly plan tourism selection, tourism activity selection, and tourism movement design.

2.3. Tourism Information Sources and Tourism Activity Research Trends

Previous studies have determined that the tourism information source relates to tourism activity. For example, using the SNA from [4], the tourism information category was hierarchically organized from the central tourist information category to the purpose-of-trip networks. Further, it showed how travel activities change according to purpose-of-trip networks, composition-of-travel-party networks, frequency-of-visit networks, length-of-stay networks, and number-of-attractions-visited networks. In addition, Xiang et al. [22] noted that most travelers use online information resources for travel planning, but Generation Y is more frequently using the Internet (social media). Further, the development of ICT has also changed travel behavior because tourists have learned to delay important decisions or use various channels in a specialized way until they arrive at their destination because of increased access to information. Talpur and Zhang [17] collected data on tourism activities (e.g., photographs, text messages, and videos) involving tourists visiting Singapore using the Foursquare application, and as a result of analyzing these data, the tourists’ tourism activity patterns in the morning and afternoon were shown to be different. Furthermore, Fernández et al. [3] presented 27 tourism information sources for travelers from four major countries (France, Germany, Italy, and the United Kingdom) who mainly visit Spain and found that family and friend recommendations were the most influential in destination selection, and tourism activities at the destination were found through Internet searches.

Previous studies have researched travel information searches and tourism behavior, travel information source and customer behavior, and travel information source and travel activity due to the development of ICT, and information source is reported to be closely related to travel behavior. On the other hand, there are few studies on the differences in tourism behaviors that divide the minority of tourism information into offline and online. Therefore, on the basis of the literature review and previous studies, this study developed the following research hypotheses from the macro (RQ1) and micro (RQ2) level perspectives.

RQ1: There is a significant difference in the patterns of participation in tourism activities depending on the tourism information source.

RQ2: There is a significant difference in normalized centrality distributions depending on whether a network is an online or offline source.

RQ2-1: There is a difference in normalized degree centrality distributions depending on whether a network is an online or offline source.

RQ2-2: There is a difference in normalized eigenvector centrality distributions depending on whether a network is an online or offline source.

3. Study Design

3.1. Study Data

The Gyeonggi province is the most preferred tourist city by Koreans and foreigners because of its abundant tourist resources such as mountains, temples, waterfalls, beaches, rivers, culture, historical sites, museums, art galleries, resorts, parks, golf courses, hot springs, and the DMZ [29]. Therefore, in this study, the Gyeonggi province was selected as the research target in order to contribute to further improving the tourism competitiveness of the Gyeonggi province.

This study utilized original data from the 2021 Gyeonggi Province tourist survey distributed by the Gyeonggi Tourism Organization. Specifically, the population of raw data is Koreans aged 15 or older who visited Gyeonggi-do in 2021. The data consisted of items with the purpose of travel, tourism information sources, tourism activities participated in during travel, destination visited during travel, travel expenses, travel satisfaction, future intentions, and other diverse items. Data collection was conducted through face-to-face interviews using structured questionnaires (Gyeonggi Tourism Organization, 2021). Also, the survey was conducted in the first quarter (27 February 2021–26 March 2021), the second quarter (5 April 2021–20 June 2021), the third quarter (4 July 2021–28 September 2021), and the fourth quarter (1 October 2021–29 December 2021). A multi-stage stratified sampling method was used to refine the data, and the total number of the sample was 6,088 domestic tourists [29].

The purpose of this study is to understand domestic tourists’ patterns of participation in tourism activities based on the tourism information source used. Therefore, the present study conducted a data pre-processing procedure to extract the final research data. First, based on the criteria regarding travel type, this study extracted 6028 respondents. Second, the present study created and analyzed a tourism activity network based on the relationship between activities that is formed by tourists participating in diverse activities to investigate tourists’ patterns of participation in tourism activities. Thus, this study extracted only responses from tourists who participated in three or more tourism activities (N = 2704). Lastly, based on the criteria regarding the information source used, the current study classified respondents as follows: (1) respondents who used offline tourism information sources (N = 574); and (2) respondents who used online tourism information sources (N = 1460).

3.2. Measurement

Among the various question items in the 2021 Gyeonggi Province tourist survey, this study used the items “information source” and “tourism activities participated while traveling to Gyeonggi.” Specifically, respondents responded to the question, “Which information source are you utilizing?” and the questionnaire about the tourism activities participated in during travel to Gyeonggi included “Which tourism activities did you participate in? Please choose the tourism activities you participated in.” In their responses to this item, respondents selected or listed tourism activities that they participated in, as shown in the Appendix A.

This type of questionnaire is generally used to obtain social network data [30]. However, social network data obtained through a survey is attribute data and must be transformed into relational data for SNA [15,31]. Therefore, the current study conducted the following procedure to transform attribute data into relational data. First, this study used 31 types of tourism activities presented by the original data such as river/sea/mountain/lake, valley/waterfall, park/arboretum/recreation forest, culture/historical experience, rural/ecological experience, acid/castle/royal tomb, and so on. Secondly, this study represented tourism activity–tourist relationships through a sociometric choice matrix that expressed the existence or nonexistence of the linkage between nodes [32]. The matrix indicated with a number 1 that a tourist participated in a particular tourism activity, while 0 means that the tourist did not participate in the tourism activity. Thirdly, for the offline source network, this study created a matrix of 574 tourists by 31 tourism activities. The matrix was converted to a matrix of 31 tourism activities by 31 tourism activities, which represented an offline source network. For the online source network, the present study made a matrix of 1460 tourists by 31 tourism activities. The matrix was converted to a matrix of 31 tourism activities by 31 tourism activities, which represented an online source network. Lastly, this study constructed tourism activity networks (offline source and online source networks) through this procedure and also utilized them in the SNA.

3.3. Data Analysis

The purpose of this study is to analyze the connection relationship between tourism activities participated in by tourists and to investigate the characteristics of the connection relationship. With regard to data analysis methods, social network analysis (SNA) methods can be appropriate because SNA focuses on investigating the nature of the relationship between two actors rather than individual actors [33]. Therefore, employing the SNA techniques allow us to understand the relationship between tourism activities, providing new insights to better understand tourists’ participation behavior in tourism activities [4].

3.3.1. Density and Centrality Analysis

This study constructed information source networks by utilizing tourists’ tourism activity participation data after classifying the data according to the information source, which affects the tourism activity participated in. Then, this study carried out SNA to investigate the structural characteristics of the information source networks.

A density analysis is used to explore the degree of linkages between nodes within a network. The density indices can be used to interpret the structure of the entire network [33]. If networks have high-density indices, it can be interpreted that the relationships between the nodes are densely formed [15].

Because of their capability of discovering the characteristics or roles of nodes, centrality analyses have been widely used in tourism research [34]. The centrality analysis is a method of quantifying how central a specific node is within a network [33]. It includes, generally, four centrality analyses: (1) degree centrality analysis, which is calculated according to the direct connectivity between nodes; (2) closeness centrality analysis, which is calculated using the amount of direct and indirect connectivity between nodes; (3) betweenness centrality, which is calculated according to the brokerage between nodes; and (4) eigenvector centrality analysis, which is calculated using not only the amount of direct connectivity but also the degree centrality score of the connected nodes [15,33]. Specifically, this study’s data do not contain information on directionality information or the order of the tourism activities that tourists participate in. Thus, this study used the degree centrality and eigenvector centrality indices among the four centrality indices to confirm which nodes have a core and stronger influence than other nodes within networks.

3.3.2. Bootstrap Paired Sample t-Test

This study used density indices to employ and verify the permutation analysis. In other words, a bootstrap paired sample t-test was conducted on the differences between the population averages of the two networks composed of identical nodes. Specifically, the bootstrap paired sample t-test used in this study can be distinguished from traditional statistical methods that utilize statistical software such as SPSS. Traditional statistical methods presume independence and random sampling in the distribution of variables. However, traditional statistical methods such as a traditional t-test cannot be applied because the network data violates independence assumptions and random sampling [35]. Therefore, a bootstrap method was used to extract countless samples, and a sampling distribution was constructed using statistics from the extracted samples. Then, a permutation verification was conducted to statistically verify the differences between the two networks.

3.3.3. SNA Results and Their Applications in Tourism Activity Networks

In this study, the SNA data were applied as follows. First, “nodes” that made up two networks refer to the “tourism activity” that tourists participated in, and the “links” between nodes refer to the “connection between tourism activities.” Secondly, density indices were used to discover the connection between tourism activities that make up networks (online source network, offline source network), and higher density indices indicate that the connection between tourism activities does not lean toward a specific tourism activity, but the relationship was formed densely [4,15]. In this study, higher density indices indicate that tourists were not focused on participating in a specific tourism activity but rather diverse activities, and they imply wider, multiple participating activity patterns. Lastly, based on Kang et al. (2018) [15], this study classifies tourism activities with a high degree of centrality and eigenvector centrality indices as core activities that have a strong influence within a network.

3.3.4. Analytical Tools

A frequency analysis was conducted using SPSS 18.0 to investigate the frequency of participation in tourism activities. In addition, a bootstrap paired sample t-test was performed using UCINET 6.682 to statistically verify the two networks constructed according to the information source used, which affects participation in tourism activities. UCINET 6.682 was used to examine the structural characteristics of the tourism activity network via a density analysis and a centrality analysis. In addition, this study employed Netdraw software to visualize the online and offline source networks.

4. Results

4.1. Frequency Distribution of Top 10 Tourism Activities in Gyeonggi Province by Information Source

Table 1. Frequency distribution of top 5 tourism activities in Gyeonggi Province by information source.

Offline Information Source (N = 574)			Online Information Source (N = 1460)
Rank	Tourism Activities	Frequency (%)	Rank	Tourism Activities	Frequency (%)
1	TA1	405 (19.481)	1	TA3	869 (15.340)
2	TA29	317 (15.248)	2	TA1	859 (15.163)
3	TA3	299 (14.382)	3	TA29	644 (11.368)
4	TA28	175 (8.418)	4	TA28	364 (6.425)
5	TA7; TA18	106 (5.099)	5	TA7	305 (5.384)

Note: Multiple responses allowed.

displays the distribution of the five most frequently participated-in tourism activities in the Gyeonggi Province among the 31 tourism activities found by using offline information sources and online information sources. First, tourists using offline information sources most frequently participated in activity TA1 (river/sea/mountain/lake), but tourists using online information sources most frequently participated in TA3 (park/arboretum/recreation forest). Secondly, TA18 (fortress/castle/royal tomb) was found to be a tourism activity in which only tourists using offline information sources frequently participated. Lastly, the five tourism activities with a high frequency of participation were common to tourists using offline information sources and those using online information sources, but the rankings were different.

4.2. Statistical Verification of Tourism Information Source Networks

To verify RQ1, “There is a significant difference in the patterns of participation in tourism activities depending on the tourism information source”, the present study used density indices obtained from the density analysis to confirm the difference in group averages between the online and offline source networks. In other words, a bootstrap paired sample t-test was conducted to investigate whether there is a difference in patterns of participation in tourism activities between tourists using online sources and tourists using offline sources.

Table 2. Results of bootstrap paired sample t-test.

	Network Density Indicates	Bootstrap Standard Error	t Value	p Value
Online source network	0.697	0.051	5.190	0.000 ***
Offline source network	0.520	0.058

Note. Number of bootstrap samples: 5000; t value: test statistics from the permutation; p value: proportion of absolute differences as large as observed, *** p < 0.001.

displays the results of the bootstrap paired sample t-test disclose a different pattern of participation in tourism activities between tourists using online sources and tourists using offline sources (t = 5.190, sig. = 0.000, p < 0.001). The density index of the online network was 0.697, and the density index of the offline network was 0.520. This difference expresses that the offline source network was more sparsely connected than the online source network. Thus, tourists using online sources cohesively participated in multiple tourism activities, and tourists using offline sources tended to show hierarchical patterns of participation in tourism activities [15].

4.3. Comparison of Online and Offline Source Network Centralities

To verify RQ2, “There is a significant difference in normalized centrality distributions depending on whether a network is an online or offline source”, the current study employed a centrality analysis among SNA methods to measure the degree centrality, eigenvector centrality, and betweenness centrality indices, and then determined the characteristics of tourism activities utilizing those indices. In addition, this study applied normalized centrality indices, which were developed to compare nodes in two and more different networks and the differences in the centrality indices of tourism activities distributed within the online and offline source networks [15,33].

To verify RQ2-1, “There is a difference in normalized degree centrality distributions depending on if a network is an online or offline source”, the present study employed a normalized degree centrality analysis, as displayed in

Table 3. Comparison of the top 5 normalized degree centralities of tourism activities in the two networks.

Rank	Offline Source Network		Online Source Network
	Tourism Activities	Normalized Degree Centrality	Tourism Activities	Normalized Degree Centrality
1	TA1	0.933	TA1, TA3, TA10, TA28, TA29	0.933
2	TA29	0.869	TA4, TA23	0.900
3	TA28	0.833	TA7, TA17	0.867
4	TA3	0.899	TA9, TA11, TA18	0.833
5	TA4	0.767	TA2, TA5, TA24, TA26	0.800

. First, Table 3 shows that the offline source network includes five tourism activities, while the online source network includes sixteen. Second, TA1 (river/sea/mountain/lake) ranks first in both networks. Therefore, river/sea/mountain/lake activities are key tourism activities in the network located in the Gyeonggi Province. Third, the tourism activities that ranked second, third, fourth, and fifth in the offline source network differ from those ranked in the online source network. Fourth, in contrast to offline source networks, online source networks include several tourism activities from the first rank to the fifth rank. Lastly, TA10 (filming location) was not included in the list of highly influential tourism activities in the offline source network, but its influence ranking was the first in the online source network. These results show that there are differences between tourism activities that strongly depend on the tourism information source used by the tourists.

To verify RQ2-2, “There is a difference in normalized eigenvector centrality distributions depending on whether a network is an online or offline source”, the present study conducted a normalized eigenvector centrality analysis, as represented in

Table 4. Comparison of top 5 normalized eigenvector centralities of tourism activities in the two networks.

Rank	Offline Source Network		Online Source Network
	Tourism Activities	Normalized Eigenvector Centrality	Tourism Activities	Normalized Eigenvector Centrality
1	TA1	35.771	TA1, TA3, TA10, TA28, TA29	30.599
2	TA29	35.060	TA4	30.141
3	TA28	34.251	TA17	29.671
4	TA3	33.900	TA23	29.664
5	TA4	33.185	TA7	29.140

. Normalized eigenvector centrality considers the local network of tourism activities directly adjacent to the focal tourism activities. Therefore, eigenvector centrality is a more extended version of normalized degree centrality [15,33]. The results of the normalized eigenvector centrality analysis are as follows. First, TA1 (river/sea/mountain/lake) ranked first in both information source networks (i.e., online and offline). Therefore, river/sea/mountain/lake activities can be regarded as prestigious tourism activities in the network. In other words, tourism activities that can be experienced by visiting the Gyeonggi Province depend on the rivers, seas, mountains, and lakes. Second, TA3 (park/arboretum/recreation forest), TA10 (filming location), TA28 (local food experience), and TA29 (gastroventure: gastronomy + adventure) also ranked first alongside TA1 in the online source network. In addition, one difference between online and offline networks is that TA4 ranked second in the online source network but fifth in the offline network. Lastly, like the normalized degree centrality results, the normalized eigenvector analysis results show that two or more tourism activities were included in each ranking from first to fifth in the online source network. These results imply that the centrality index of the tourism activity increases by participating in more diverse tourism activities when tourists use online information sources rather than offline information sources.

4.4. Visualization of Online and Offline Source Networks

This study visualized the two networks using Netdraw to examine the differences in connection patterns between the online and offline source networks and the differences in the influence of tourism activities. Figure 1 and Figure 2 express the size of the node, which means the tourism activity in the tourism information source network, based on the normalized degree centrality index. In other words, the larger the size of the node, the higher the normalized degree centrality index.

5. Discussion and Implications

5.1. General Discussion

Tourism information plays an important role not only in the decision-making process related to tourism planning but also in tourism activities involving tourists in tourist attractions [3,4]. Recently, the behavior of tourism is changing due to the increased access to various channels’ specialized methods and information [22]. Additionally, several researchers who have studied tourism behavior and tourism information sources report that participation in tourism activities and selection of tourist destinations vary depending on the tourism information sources used by tourists [11,12,13,14].

Therefore, most of the existing studies [1,3,5,6,8,21] focus on verifying the influence of the relationship between tourism information sources and participation in tourism activities. However, research on how tourism guides participation patterns varies depending on whether they are online or offline is insufficient. Therefore, after classifying tourists as tourists using either online tourism information sources or offline tourism information sources, a network of information sources was established and analyzed to investigate the difference in tourists’ participation patterns. Additionally, the characteristics and/or roles of the tourism activities within the offline and online information source networks were compared. In other words, in order to understand the importance of tourist participation patterns in tourism activities, a deeper understanding of whether tourists participating in tourism activities are online or offline, and a study to analyze the difference between the two were conducted. This study’s results are as follows. First, the results of the bootstrap paired sample t-test conducted to examine the differences in the patterns of tourism activity participation between tourists using offline information sources and those using online information sources revealed a difference in their tourism activity participation patterns (t = 5.190, p = 0.000). Secondly, our findings that the tourism information source influences tourists’ participation in tourism activities support existing studies [11,17,18]. Thirdly, this study provides the results of normalized centrality (degree centrality and eigenvector centrality) analyses to compare the characteristics and/or roles of the tourism activities located in the information source networks, which were constructed on the basis of tourists’ participation in tourism activities found using offline and online information sources. The results reveal differences in the tourism activities’ rankings within the two information source networks. In other words, the characteristics and/or roles of tourism activities differ depending on the tourism information source.

5.2. Theoretical Implications

This study’s theoretical contributions can be classified into three perspectives. First, many tourists participate in two or more tourism activities in one travel excursion [16,17], but most studies related to tourism activities focus on identifying the influence of the relationship between variables or verifying the differences by applying traditional statistical analysis techniques regarding participation in specific or single tourism activities [3,22]. Therefore, previous studies have a limitation in that they have not found any hidden values (e.g., characteristics, rules, roles, etc.) in the relationships between multiple tourism activities. To overcome these limitations, this study expands the understanding of tourists’ participation in multiple tourism activities by applying SNA techniques to reveal tourists’ participation patterns and identify the characteristics or roles of tourism activities. This constitutes a significant contribution to the existing literature related to tourism activities.

Secondly, with the development of ICT technology, the importance of tourism behavior networks based on tourist information sources used by tourists is increasing. However, there is a lack of understanding of these networks and the tourism activities they are actually participating in [4]. This study aims to identify two network reflecting tourism information sources and patterns of tourism activities in which tourists actually participate. These studies contribute to the research related to tourism content development, planning, and marketing on the demand side.

Lastly, the results of the frequency analysis of the multiple tourism activities that tourists participated in, the normalized degree centrality analysis, and the normalized eigenvector centrality analysis confirm that there is a difference in the rankings of tourism activities. These findings suggest that a relational approach is needed to understand tourists’ participation in tourism activities because, through advanced ICT technology and diversified tourism information sources, many tourists are continuously increasing their participation in various tourism activities in one trip [9,20,21,22,23]. The previous survey that covered most tourists’ tourism activities took an individualistic approach to investigating how many people participated in tourism activity A and how many people participated in tourism activity B. However, regarding the behavior of tourists participating in tourism activities, tourists who participate in tourism activity A can simultaneously participate in tourism activity B and then participate in tourism activity C in succession. Therefore, if the tourism activities in which tourists participate in are investigated with an individualistic approach, an error of underestimating or overestimating the number of tourism activities participated in may occur. Therefore, if the relationship between tourism activities is analyzed through a relational approach, the importance and characteristics of tourism activities that cannot be derived through traditional statistical analysis can be conceptualized [4,34]. In this respect, this study is meaningful in that it serves as a guideline for the application and approach of new research methodologies in research related to tourism activities.

5.3. Practical Implications

This study’s practical implications can be classified into four perspectives. First, the response analysis shows that the preferred tourist destinations for both online and offline information sources are similar. However, looking at offline information sources, TA18 (fortress/castle/royal tomb) is included in the top five preferred tourist destinations. The fact that offline information sources include fortresses/castles/royal tombs indicates that these locations are historical attractions that do not change with the trends. Tourists rely on online information sources when searching for information on new travel experiences that fit the trends, but they prefer offline information sources when they want to visit historical tourist attractions. Therefore, relevant working-level officials such as tourism content developers and marketers in the Gyeonggi Province should actively prepare offline promotional strategies such as promotional brochures and TV advertisements rather than social media so that tourists can easily recognize historical tourist attractions.

Secondly, the normalized degree centrality analysis results reveal that the pattern of tourism activities clearly differed depending on whether the source of the tourism information was online or offline. According to the top five, where tourists visiting the Gyeonggi Province engage in tourism activities found on online information sources, it was cohesive with rank 1 (TA1, TA3, TA10, TA28, TA29), rank 2 (TA4, TA23), rank 3 (TA7, TA17), rank 4 (TA9, TA11, TA18), and rank 5 (TA2, TA5, TA24, and TA26). However, offline information sources appeared hierarchically connected in the order of TA1, TA29, TA28, TA3, and TA4. Looking at these results, they can be compared with the amount of information. It is easy to obtain vast amounts of diverse information online, but limited information can be obtained offline. In other words, the more information tourists have, the more tourism activities they participate in. In response, working-level officials, including those in tourism content development and marketing and tourism officials in the Gyeonggi Province, should try to provide online information sources to tourists through various channels as a way to induce participation in more tourism activities. In addition, efforts should be made to provide accurate and reliable information so that tourists do not suffer damage from unrefined information pouring out of social media.

Thirdly, we found a significant difference in the normalized eigenvector centrality distribution depending on whether the network was an online or offline information source. The most preferred tourism activity pattern in both networks was TA1 (river/sea/mountain/lake), and TA1, TA3 (park/arboretum/recreation forest), TA10 (filming location), TA28 (local food experience), and TA29 (gastroventure) ranked first in the online information source network. In particular, TA10 was not included in the list of influential tourism activities in the offline information source network but ranked first in the online information source network. These results indicate something special about visiting the Gyeonggi Province. With the development of Korean films and dramas, interest in filming locations has been high on social media, and many foreign tourists who like Korean dramas and movies are visiting. In particular, online information sources account for a large proportion of foreign tourists who want information on Korean film locations. Therefore, policymakers related to tourism in the Gyeonggi Province should consider establishing strategies for online promotion that can induce the expansion of participation in tourism activities through the development and promotion of new tourism content related to filming locations.

In addition, TA1 (river/sea/mountain/lake) topped both online and offline information sources. These results show that activities involving rivers, seas, mountains, and lakes are the most popular and prominent tourism activities that can be experienced by visiting the Gyeonggi Province. However, one difference between the online and offline networks is that TA4 (cultural/historical experience) ranks second in online information source networks but fifth in offline information source networks. In other words, TA4 is the most important to tourism in the Gyeonggi Province after TA1 in the online source network.

One representative cultural and historical experience in the Gyeonggi Province is Dulle-gil. The Gyeonggi Province’s Dulle-gil is 860 km long, with 60 courses passing through 15 cities and counties, and visitors can experience ecology, culture, and history on foot [36]. Dulle-gil, Gyeonggi-do, began reorganizing the existing road in November 2018 and completed it on 15 November 2021, and it was introduced on various social medias centering on its Internet website, making it known to tourists and citizens of the Gyeonggi Province. Although it was newly created, the ripple effect of social media appeared quickly. It was introduced to many people in a short period of time and has now established itself as a key tourism activity in the Gyeonggi Province. Therefore, policymakers related to tourism in the Gyeonggi Province should take a policy approach to encourage the expansion of participation in tourism activities by developing new tourism products and promoting them using social media.

5.4. Limitations and Further Research

Although this study’s results contribute to the research on tourism information sources and tourism activity networks, it has a few limitations. First, tourists’ tourism behaviors are affected by human factors, physical factors (e.g., the placement of tourist attractions, and transportation to tourist destinations), and time factors [37]. However, the current study only considered factors related to tourism information sources among the diverse factors that may affect participation in tourism activities when it differentiated tourists to analyze their tourism activity patterns. Thus, future studies on tourism activity behavior should inclusively consider tourists’ length of stay, gender, and age; the time of year when tourists visit; transportation; and other factors.

Secondly, this study was conducted from the perspective of tourism in the Gyeonggi Province among various regions in South Korea. Therefore, future research should consider the region adjacent to the Gyeonggi Province as a single tourism destination and develop a tourism activity network linked to the region adjacent to the Gyeonggi Province to provide more abundant and meaningful implications.

Lastly, this study utilized basic and advanced SNA techniques to study the tourists’ tourism activity patterns. However, it did not analyze tourists’ satisfaction with participation in multiple tourism activities or their future behavioral intentions. Therefore, future studies should apply indices resulting from SNA to a theoretical model to practically study cause-and-effect relationships, which will provide an in-depth understanding of the tourist behaviors of those who participate in multiple tourism activities.