Intercultural Attitudes Embedded in Microblogging: Sentiment and Content Analyses of Data from Sina Weibo

Abstract

This study analyzed 2421 microblogs posted between the year 2012 to March 2022 reflecting the microbloggers’ attitudes toward different cultures. Results indicated that (1) the number of microblog posts expressing the users’ intercultural attitudes increased distinctly from 2019 to March 2022, with females users in general posting more microblogs than males; (2) females posted more microblogs encompassing positive emotions to show their interest and motivation to learn about foreign cultures, and the tendency to value and appreciate cultural differences, whereas males created more sentimentally neutral posts that revealed their recognition of the existence of cultural differences, and females and males posted a similar number of microblogs containing negative emotions; and (3) more posts involved “small c” culture were posted than those containing themes belonging to the “Big C” culture. Gender gap was further observed regarding the cultural themes concerned by the microbloggers. Implications were discussed in the context of intercultural education.

1. Introduction

Why do we need to be attitudinally prepared to embrace multiculturalism? When people and things around are greatly different from those in one’s own native cultural milieu, insecure feelings may arise because one’s internal belief system is hard to be fully affirmed (Furnham 2019). It is thus natural for a person to react with defense mechanism to protect the self from emotional conflicts or stress when faced with cultural differences (Lebedko 2014). Typical defensive reactions include, but are not limited to, denying the existence of cultural differences (Bennett and Hammer 2017), a tendency to judge using one’s own culture as the benchmark (Tori and Bilmes 2002), and trying to make sense based on overgeneralized and secondhand beliefs (i.e., stereotypes) (Amoretti 2018). Unfortunately, these internal attitudes can be radiated via a person’s speech, body language, and even tone of voice, which can further lead to a cycle of negative reaction and result in communication breakdown or conflict (Zhu and Bresnahan 2018).

Over the past decade, research on cultivating positive attitudes toward other cultures has proliferated in various fields. In healthcare, studies like Karasu et al. (2022) demonstrated the importance of cultural sensitivity in improving patient outcomes and reducing health disparities. In business, Moradi and Ghabanchi (2019) highlighted how cross-cultural understanding boosts international collaboration and market success. In tourism, Kirillova et al. (2015) showed that cultural appreciation enhances travelers’ experiences and promotes sustainable tourism. In education, Tarchi and Surian (2022) emphasized the role of intercultural competence in fostering inclusive learning environments and improving student engagement. However, in the exploration of intercultural attitudes of individuals, quantitative approach relying on self-reported measures occupied a large portion of prior studies (e.g., Idris 2020; Liu et al. 2017). Beyond any doubt, self-reported measures are versatile and indispensable for capturing differentiated patterns of human attitudes (Hendrick et al. 2013), whereas they also bear drawbacks. Given the fact that self-report on survey items describing psychological processes is under respondents’ control, a common criticism against self-report measures has been their tendency to be susceptible to socially desirable responses (Holtgraves 2004). That is, the respondent might engage in faking behaviors intentionally and unintentionally to forge the ideal image of the self. This can be understood through Goffman’s (1967) theories of impression management and facework. Individuals regularly engage in impression management, consciously or unconsciously adjusting their behaviors to project a favorable image. This includes unintentional actions driven by ingrained social mechanisms that help maintain a positive identity. Thus, both intentional and unintentional faking are part of broader social practices aimed at managing self-presentation, including in intercultural contexts. As such, scholars have issued calls on researchers to use more diverse approaches when investigating intercultural dispositions, such as behavioral observations, text mining, and implicit association tests, to obtain a more nuanced and accurate understanding of intercultural attitudes (Deardorff 2017). Such methodologies can provide deeper insights by minimizing the influence of impression management and capturing more genuine intercultural behaviors and dispositions (Leung et al. 2014).

Microblogging has been a valuable data source for researchers to understand individual attitudes and beliefs about a particular topic. For example, based on tweets or posts collected from social media sites where users share information, feelings, and perspectives (e.g., X [formerly Twitter], Facebook), studies were conducted to understand individual’s attitudes toward MOOCs (massive open online courses) (Zhou 2022), vaccination (Tavoschi et al. 2020), global warming (Qiao and Jiang 2021), LGBTQ-related issues (Adamczyk and Liao 2019), among other topics. As Kumar (2021) pointed out, a strength of unstructured data such as microblogging is that it does not contain any predefined rules. Translated to the current context, unlike structured survey items assessing intercultural attitudes that preset intercultural situations (e.g., “I respect the ways people from different cultures behave.”, Intercultural Sensitivity Scale, Chen and Starosta 2000), the revealed attitudes via microblogging derived from what a person saw and heard that formed his/her unique intercultural experiences (e.g., “I asked the teachers if they wanted to try my cookies. The Chinese teachers were all like, ‘No, no thanks, I’m good’, but before I even finished my question, the foreign teacher’s hand was already reaching for a cookie. It was so cute!”, an analysis item of this study).

Hence, the microblogging data are deemed to be more authentic regarding intercultural situations due to their real-time, personal nature, which captures spontaneous expressions of cultural behaviors, customs, values, and beliefs (Tufekci 2014). For example, posts about local festivals offer insights into how individuals celebrate their traditions and share their experiences with a global audience. Microbloggers might also express admiration or curiosity about other cultures, as seen in posts from individuals in the U.S. sharing their excitement about attending a traditional Japanese tea ceremony, or in users from India posting their reflections on attending a Thanksgiving dinner in the U.S. Additionally, discussions on cultural values provide perspectives on how people perceive and interact with different cultural norms. To the author’s best knowledge, however, no study to date has employed microblogging data to investigate the attitudinal aspect of individuals’ intercultural dispositions. As Armitage and Christian (2003) maintained, attitudes, when shaped in positive way, can influence behaviors to achieve a favorable outcome. Indeed, positive attitudes toward foreign cultures are the preconditions of voluntary cognitive skills enhancement, and effective behavioral and emotional regulation that would ultimately lead to successful intercultural interactions (Griffith et al. 2016). The purpose of this study was thus to explore Chinese microblogging users’ attitudes toward foreign cultures/cultural differences and their utmost attention to certain themes of culture. The findings of this study will inform educators who aim to adopt purposeful and effective strategies to develop positive intercultural attitudes of learners.

2. Literature Review

2.1. Intercultural Attitudes

Individuals who wish to thrive in the multicultural but interconnected globe are expected to possess a decent level of intercultural competence (IC). Intercultural attitudes, as the attitudinal dimension of IC and the focus of this study, was defined as a collection of traits that drive individuals to evaluate intercultural situations as beneficial and remain engaged in or initiate intercultural interactions despite the innate uncertainty and ambiguity (Byram 2021; Deardorff 2006; Griffith et al. 2016). These IC experts further proposed a variety of elements of intercultural attitudes to operationalize the definition. Specifically, in possession of favorable intercultural attitudes, a person would respect and value cultural diversity; show open-mindedness to cultures of and people from other countries while withholding judgment; and show curiosity and discovery for intercultural learning and interactions. Opposite to these elements were ethnocentric beliefs, stereotypes and prejudices, low tolerance for ambiguity, and even dislike or rejection of foreign people and cultures, which constitute typical causes of intercultural tensions and breakdown (Spitzberg and Changnon 2009).

2.2. Types of Culture: ‘Big C’ and ‘Small c’ Culture

Culture is defined as “the learned and shared values, beliefs, and behaviors of a community of interacting people” (Bennett and Bennett 2004, p. 147). To further sort out different themes of culture, Paige et al. (2003) suggested two broad types of culture: “Big C” and “small c”. “Big C” culture refers to a set of facts that can be easily observed and remembered, such as a country’s history, economy, social system, education, festivals, and art. By contrast, “small c” consists of aspects that cannot be easily and immediately observed or understood, such as people’s behavioral patterns, thinking patterns, values and beliefs, customs, language, and everyday living. These two culture types have been used by researchers to extract cultural content in various text materials.

For example, Kang-Young (2009) analyzed the texts and pictures in eleven high-school EFL conversation textbooks and found that most of the textbooks showed a strong preference for the teaching of “Big C” culture (e.g., education, holidays, food), yet paid little attention to “small c” culture (e.g., individualist beliefs, interpersonal relationships). Similarly, Labtic and Teo (2019) investigated the cultural information contained in six English textbooks and reported that “Big C” culture (e.g., literature, geography, and races) is more frequently presented than “small c” culture (e.g., everyday living, ritual behavior). A review of literature showed that the framework relying on “Big C” and “small c” cultures were more frequently adopted in the analyses of educational materials (e.g., Yue et al. 2020; Zakiyah and Rukmini 2022). Nonetheless, this framework is also believed to be valid for analyzing microblogging texts because when users posted specifically on intercultural topics, the notion of culture would be more tangible for them as such different cultural themes would be more explicitly presented in their postings.

2.3. Sentiment and Content Analysis of Microblogging Data

Sentiment analysis identifies subjectivity (neutral vs. emotionally loaded) and the polarity (positive vs. negative semantic orientation) of a text (Pawar et al. 2016). Appraisal theory posited that emotions can be elicited by evaluations of concrete situations that produce specific reactions on various individuals (Scherer et al. 2001). For example, happiness felt when a romantic relationship starts could be elicited by the appraisals that something desired has been obtained (Berscheid and Walster 1978). The appraisal theory lays the foundation for structured sentiment extraction that is based on appraisal expression (Bloom 2011). As platforms for users to express themselves, the prevalence of microblogs has aroused the research in the field of sentiment analysis. According to Korenek and Šimko (2014), the appraisal theory also has its strength in supporting the analysis of microblogers’ attitudes toward any object/person/event (e.g., foreign cultures or people in this study), because attitudes express a person’s current mindset when writing a post that contains his/her affect, appreciation or judgment.

Qualitative content analysis is an analytical tool used to “systematically analyze texts by processing the material step by step with theory-based category systems developed on the material.” (Mayring 2002, p. 114). The deductive content analysis is often used to test existing categories and concepts. In other words, the structure of analysis is operationalized on the basis of previous knowledge (Elo and Kyngäs 2008). For example, some studies analyzed the microblog posts based on existing categories of health-related messages (Pei et al. 2016), learning and teaching content (Fischer et al. 2019), and attitudes toward sports for women (Esmundo 2021). Given that the analytic targets of this study (i.e., dimensions of intercultural attitudes and culture) have already been maturely conceptualized and validated by both theoretical and empirical studies, the deductive approach was chosen to perform the content analysis of microblogging data.

2.4. Gender Differences in Intercultural Attitudes and Microblogging

The American Psychological Association (2011) defined gender as “the attitudes, feelings, and behaviors that a given culture associates with a person’s biological sex” (p. 1), implying that men and women differ psychologically in the way they perceive and act. Some past studies reported that women were more likely to (a) respect cultural diversity and be open-minded about cultural differences (Vreckova et al. 2020), (b) be interculturally empathetic (Solhaug and Kristensen 2020), (c) possess stronger foreign language proficiency (Wightman 2020), and (d) hold ethnorelative perspectives (Goldstein and Kim 2006) than their male counterparts. In contrast, some studies documented that men experienced a lower level of anxiety and had better control of their interactions with culturally different others (Moradi and Ghabanchi 2019). Concerning the posting behaviors on social media, compared to men, women were found to express more emotions on social media (Tifferet and Vilnai-Yavetz 2014) that were either distinctly positive or negative (Jalonen 2014). Moreover, researchers observed that women tend to post more topics related to everyday living such as family fun, holidays, and festivals, whereas men post more about topics such as deep thought, politics, and religion (Liu et al. 2018; Wang et al. 2013). All the evidence indicated the need to attend to gender differences in the sentiment and content of microbloggers’ posts.

2.5. The Present Study

This study made the first attempt to generate a comprehensive picture of how Chinese people feel about foreign cultures and people, and more specifically, what themes of cultural content were concerned, and whether gender gap existed in these issues, by closely examining microblog posts on relevant topics from the launch year of the selected microblogging platform through 2022. The general objective of this study is to uncover how microbloggers’ attitudes and thematic focus on foreign cultures have evolved over time and to determine the role that gender differences play in shaping these perspectives. Five research questions were proposed to guide this study:

(1)

What was the overall microblog activity about users’ attitudes toward foreign cultures and people during 2009 and 2022?

(2)

What were the sentiment values of the eligible posts that reflected the microbloggers’ attitudes toward foreign cultures and people?

(3)

Were there any gender differences in microbloggers’ attitudes toward foreign cultures and people?

(4)

What themes of culture under the “Big C” and “small c” culture types were included in the eligible microblog posts?

(5)

Were there any gender differences in the culture themes concerned by the microbloggers?

3. Methods

3.1. Data Source and Data Collection

Sina Weibo, the Chinese equivalent of X (formerly Twitter), was launched by Sina Corporation in August 2009. As reported by Thomala (2022, June 21), up to the first quarter of 2022, monthly active users of Sina Weibo have surpassed 582 million (252 million daily active users). It has been the most popular microblogging platform in China for users to share and receive information. Microblogging data were extracted using Houyi Crawler Version 4.0.1.0 (https://www.houyicaiji.com/, accessed on 7 May 2022), a free software developed based on an artificial intelligence algorithm to identify and collect information from websites. The search terms used for data collection were “外国文化” (foreign culture), “外国人”/”老外” (foreigner), “其他国家”/”别的国家” (other countries), “西方国家” (Western countries), “西方人” (Westerners), “东西文化差异” (cultural differences between Eastern and Western cultures), “中外文化差异” (cultural differences between China and foreign cultures), “和我们的文化相比” (compared to our culture), “和中国文化相比” (compared to Chinese culture), “和中国人相比” (compared to our Chinese), and “过洋节” (celebrate foreign festivals).

The timeframe of searching was set between August 2009 and March 2022. A total of 9464 microblogging posts were initially identified. The data were exported to and processed in Microsoft Excel. In the phase of data pre-processing, the dataset was cleaned by removing repeated posts by the same author (n = 45), political comments on international wars (n = 2839) or the pandemic (n = 2055), news (n = 63), advertisements (n = 284), and irrelevant posts (n = 1757). The final dataset for subsequent analyses included 2421 microblog posts by 2295 Sina Weibo users. The gender information was obtained from user profiles, with 1410 microblogs posted by females and 1011 microblogs posted by males.

3.2. Sentiment Values Analysis and Content Analysis

After the data pre-processing phase, each microblog was coded regarding its sentiment value (i.e., positive, neutral, or negative). Previous studies suggested that sentiment analysis of microblogging data could have been automated through using natural language processing techniques such as Linguistic Inquiry and Word Count tool (e.g., Onan 2018) and machine learning algorithms (e.g., Gupta et al. 2017). Nevertheless, this study chose to use manual coding as the effort required was acceptable. More importantly, according to Van-Atteveldt et al. (2021) and Boukes et al. (2020), human coding tends to produce higher accuracy as compared to automated coding. The appraisal dictionary suggested by Bloom (2011) was used to assist the coding of emotional quotients associated with the microblog posts. Specifically, posts that expressed obvious positive emotions or contained words such as “glad” and “interesting” were coded as positive; posts that revealed ambiguous or mixed emotion or included words such as “different” and “differences” were coded as neutral, and posts that expressed obvious negative emotions or used words such as “frustrated” or “afraid” were coded as negative. Example microblogs that were coded as positive, neutral, or negative, respectively, are presented in

Table 1. Example microblog posts with different sentiment values.

Original Posts	English Translation	Sentiment Values
“今年的我在冬奥遇到了很多朋友，不论是一起工作的大家，还是许许多多来自不同文化背景的外国朋友。我时常感到足够温暖与幸运。”	“This year, I made many friends at the Winter Olympics, including colleagues and people from various cultural backgrounds. I often feel warm and fortunate because of these experiences.”	Positive
“看了一晚上视频，外国人在中国生活，中国人在外国生活，文化差异很有趣。”	“Spent the whole evening watching videos about foreigners living in China and Chinese people living abroad. The cultural differences are really fascinating.”
“环境的差别竟最终导致了如此巨大的东西方文化差异，以至于到了现在人们都还生活在不同的文化环境和政治制度中，这便使人有了差别。”	“The environmental differences have created significant cultural differences between the East and West. Even now, people live in different cultural contexts and political systems, which shapes their behaviors and perspectives.”	Neutral
“美国人和我们国家的人相比，他们吃的食物多数是高热量的，也就是肉食比较多，在宴会上，也会使用一次性手套来吃食物。”	“Compared to people in my country, Americans eat a lot more high-calorie foods, like meat, and they use disposable gloves at banquets.”
“开始半个月的实习了被分到了一个全是外国人的组本社恐+语障已经快死了。”	“Started my internship for half a month and assigned to a group full of foreigners. I am dying as a sociophobic nerd with language barrier.”	Negative
“看一圈美食纪录片就知道，国外很多食材和做法根本没法和中国相比。”	“Watch some food documentaries then you will know that many foreign food ingredients and recipes are simply not comparable to those in China.”

. All the translation was conducted by following Brislin’s (1970) back-translation technique, which involves translating the text from the original language to the target language and then translating it back to the original language. This approach helps to identify and correct discrepancies in meaning, tone, or context, ensuring that the translated content retains its original intent and nuances. By using this method, the study aimed to minimize shifts in meaning that could affect the accuracy and fidelity of the emotional coding, thereby enhancing the reliability of the analysis of intercultural sentiments in microblogging data.

To further discover the dimensions of intercultural attitudes embedded in the microblog post and the associated themes of culture, deductive content analysis (Mayring 2015) was adopted to scrutinize the posts content. The codebook was composed of the theoretically and empirically differentiated categories of intercultural attitudes and culture (i.e., “Big C” and “small c”). To improve the reliability of coding, two PhD candidates in education were invited to perform the manual coding. The coders were paid 1.2 USD for each microblog post. Before commencing the formal coding, the two coders spent two days familiarizing themselves with the codebook and worked independently to code 200 randomly selected microblog posts to examine the relevance of each code, and to establish inter-coder reliability. The sequence of assigning codes for each post was the (a) sentiment value, (b) dimension of intercultural attitudes, and (c) theme of culture. Figure 1 delineates the data analysis procedures.

Inconsistencies were discussed after the coding, and the codebook was further refined. Then, the two coders applied the refined codebook (see Appendix A) to independently coded all microblog posts under analysis. The Krippendorff’s α for the coding of sentiment values (α = 0.90), dimensions of intercultural attitudes (α = 0.87), and themes of culture (α = 0.93) were calculated in SPSS 26.0 (De Swert 2012).

4. Results and Discussions

4.1. The Overall Microblog Activity about Users’ Intercultural Attitudes

As shown in Figure 2, during the period from 2012 to 2018, only a handful of microblogs were posted to express the users’ attitudes toward foreign people and cultures. From 2019 to 2021, the number of microblog posts increased steadily, but from the end of 2021 to March 2022, the number of microblog posts expressing the users’ intercultural attitudes dramatically increased. The turning points of growth in the number of relevant microblog posts were very likely related to the worldwide outbreak of the pandemic in 2019. Probably, during the lockdown period, many users spent increased time on active social media use to reduce boredom (Sun et al. 2022) such that through extensive viewing activities (e.g., news) on microblog platform, users were exposed to culturally different reactions to the epidemic (e.g., face masks issues), which stimulated their willingness to express their thoughts or opinions about people and cultures of other countries.

The gender-based comparison indicated that from 2019 to March 2022, females (64.4%, 55.8%, 57.4%, 59.5%, respectively) posted more microblogs than males (35.6%, 44.2%, 42.6%, 40.5%, respectively) that revealed their intercultural attitudes. This corroborated the findings of previous studies that females had a greater tendency to engage in active social media use (i.e., posting) and self-disclosure (i.e., the process of revealing intimate information about oneself to others such attitudes, feelings, desires, experiences, and thoughts) than their male counterparts (Towner et al. 2022).

4.2. Sentiment Values and Dimensions of Chinese Microbloggers’ Intercultural Attitudes: Gender-Based Comparison

The results of the sentiment analysis demonstrated that, in general, 53.9% (n = 1304) microblog posts contained obvious positive attitudes toward foreign people and cultures; 25.5% (n = 618) posts revealed ambiguous attitudes that were labeled as neutral; and 20.6% (n = 499) posts involved obvious negative attitudes toward foreign people and cultures. As shown in Figure 3, when the sentiment values of posts were compared across gender groups, females (73.8%, n = 962) were found to post high rate of sentimentally positive microblogs, nearly three times higher than that of males (26.2%, n = 342). In contrast, the rate of sentimentally neutral microblogs posted by males (69.9%, n = 432) was more than twice as that posted by females (30.1%, n = 186). Moreover, the rates of sentimentally negative microblogs posted by females (52.5%, n = 262) and males (47.5%, n = 237) were of minor differences, with females having slightly higher rate than males.

The finding that females in general exhibited more obvious emotion than males was congruent with Park et al.’s (2016) analytical results of 15.4 million posts created by 68,000 Facebook users. Their findings suggested that women use words more emotionally than men as reflected in women’s predominant use of words describing emotions (e.g., excited, happy) and intensive adverbs (e.g., sooooo, ridiculously); and their tendency to use warmer and more positive words. Similarly, Rao et al.’s (2010) examination of X (formerly Twitter) users’ language use reported that expressions such as “OMG” and “lol” were more frequently used by females, and the word “yeah” is predominantly used by men to affirm a fact.

A closer look was then taken to investigate the dimensions of intercultural attitudes as embedded in the sentimentally positive, neutral and negative microblog posts. As presented in Figure 4, sentimentally positive posts reflected the microbloggers’ curiosity and discovery about and respect for foreign people and cultures. The results indicated that among these posts, more microblogs were posted by females to show their interest and motivation to learn about foreign cultures (females: 69.9%, n = 589; males: 30.1%, n = 254), and also the tendency to value and appreciate cultural differences (females: 80.9%, n = 373; males: 19.1%, n = 88), as compared to their male counterparts. This finding was consistent with previous empirical studies that compared to males, females are more likely to initiate intercultural contact and are more motivated to understand, appreciate, and interact with different cultures (Tompkins et al. 2017; Vreckova et al. 2020). Scholars have attempted to understand this gender gap based on the assumptions of social role theory. That is, men are expected to be highly agentic because of their assertive and competitive behaviors; women, on the other hand, are seen as communal for being more friendly and caring (Eagly and Wood 2012). Thus, when approaching foreign people/cultures, women are more likely to exhibit friendly and empathetic attitudes and evaluate the intercultural situations as favorable.

Sentimentally neutral posts displayed the microbloggers’ openness toward cultural differences. That is, among these microblogs, compared to females, more males posted to express their recognition of the existence of cultural differences, without revealing obvious emotion (males: 69.9%, n = 432; females: 30.1%, n = 186). According to Epstein (2003), women and men relied on different systems (experiential versus rational) to interpret the world. The experiential system adopted by women is innate and adaptive; hence, women prefer to make sense of the world by learning from outcomes of affective experiences that stimulate positive emotions, for example, to recognize, understand, and appreciate cultural differences through engaging in authentic intercultural contacts. In comparison, men applied a rational system that operates based on evidence and is considered to be analytical. Therefore, men are apt to translate the reality into symbols or words that are transmitted culturally such as through education (Sladek et al. 2010), for instance, to indicate the acceptance of cultural differences by rationally summarizing the available evidence (e.g., history of cultural diversity).

However, mixed results were found regarding sentimentally negative microblogs. Comparison across gender groups suggested that more microblogs were posted by males that involved negative cultural orientation (males: 66.2%, n = 45; females: 33.8%, n = 23) whereas more posts were created by females that implied their low tolerance to differences (females: 55.5%, n = 239; males: 44.5%, n = 192). Specifically, more microblogs posted by males displayed the beliefs that their own culture is better than that of other countries (i.e., ethnocentric beliefs) and unwillingness to approach people/culture of other countries (i.e., inherent resistance). However, more posts created by females revealed their negative attitudes toward intercultural situations accompanied by feelings of confusion or anxiety (i.e., culture shock). Similar findings were documented in prior studies. For example, perhaps for generally being more assertive and less trusting, men were found to have higher level of ethnocentrism than women (Salisbury et al. 2010) and were less willing to communicate interculturally by viewing differences as a threat to one’s identity and view of the world (Göncz 2018). On the other hand, women are usually more socially anxious and less psychologically secure than men (Vervoort et al. 2010). Thus, intercultural situations that are inherently stressful were more likely to cause anxiety for women, which may further stimulate their intention to avoid uncertainty (Broeder 2022).

4.3. Themes of Culture Concerned by Chinese Microbloggers: Gender-Based Comparison

Of the relevant microblog posts, 27.5% (n = 666) posts were about “Big C” culture; 50.9% (n = 1233) posts were about “small c” culture; 9.1% (n = 221) posts did not specify the types of culture, culture was only mentioned literally as “culture’” in these posts; 12.4% (n = 301) posts did not include any cultural content. Instead, they expressed the microbloggers’ willingness to interact with culturally different people. Results of gender-based comparison of culture themes concerned by the microbloggers are shown in Figure 5 and Figure 6.

Regarding “Big C” culture, Figure 5 showed that more females are interested in different countries’ festivals (females: 66.4%, n = 241; males: 33.6%, n = 122), art such as movies, literature, and music (females: 57.5%, n = 111; males: 42.5%, n = 82), and food (females: 61.2%, n = 41; males: 38.8%, n = 26); whereas males posted more about history (males: 82.5%, n = 52; females: 17.5%, n = 11), the physical appearance of people (males: 60.9%, n = 14; females: 39.1%, n = 9), and the social or education systems (males: 52.0%, n = 13; females: 48.0%, n = 12) of different countries.

As for “small c” culture, females posted more about foreign language learning (females: 84.6%, n = 170; males: 15.4%, n = 31), and culturally different people’s behavioral patterns (females: 51.9%, n = 161; males: 48.1%, n = 149), everyday living (females: 75.4%, n = 52; males: 24.6%, n = 17), interpersonal or intimate relations (females: 62.5%, n = 35; males: 37.5%, n = 21), personal traits (females: 56.6%, n = 30; males: 43.4%, n = 23), and customs (females: 52.0%, n = 17; males: 48.0%, n = 12). On the other hand, males posted more about values and beliefs (males: 66.3%, n = 183; females: 33.7%, n = 93), and thinking patterns (males: 73.7%, n = 129; females: 26.3%, n = 46) of people from different cultures.

These results, to a varying extent, corroborated findings of previous studies that uncovered gender differences in posting behaviors on social media. For example, in Liu et al. (2018) and Wang et al. (2013)’s studies that investigated topic preferences of Facebook users, females were found to post more on topics related to detailed aspects of their everyday life such as family fun, intimate and interpersonal relationships, festivals, and daily routines, whereas males post more about abstract concepts such as deep thoughts, ideology- and system-related topics. Notably, as reported above, more women displayed negative attitudes toward intercultural experiences due to the culture shock they have suffered. A scrutinization of relevant posts suggested that women’ anxious feelings were mostly caused by their perceived difficulty in comprehending culturally different thinking patterns (e.g., logic of thinking) or values and beliefs (e.g., individualism, egoism, values of sex). A plausible explanation for these findings is that men and women possess different cognitive styles. Prior research (Smith et al. 2008) found that men are more disposed to abstract thinking to recognize patterns (e.g., thinking styles of Easterners and Westerners), analyze the essence of things (e.g., historical/religious roots of cultural differences), and synthesize the information to ultimately solve problems (e.g., make sense how certain values and beliefs shape the shared behavioral patterns in a cultural group). On the other hand, women tend to engage more in context-specific thinking, in aspects of concrete situations (e.g., festivals, food, and everyday living) and relationships (e.g., family relationships, friendships, and romantic relationships) in different cultures. Probably, women who were less skilled at higher-order thinking were more likely to experience culture shock related to abstract intercultural information processing.

5. Conclusions and Implications

This interdisciplinary study contributed to both fields of intercultural competence and microblogging. On the one hand, the current study made the first attempt to understand individuals’ attitudes toward foreign people and cultures through the analyses of a large number of microblogging data. In particular, the findings unsealed the percentages of microblogs with different sentiment values and the corresponding theoretically defined aspects of intercultural attitudes. The gender-based comparison further improved our understanding of the gender gap in the attitudinal dispositions needed to embrace multiculturalism. On the other hand, our findings regarding gender differences in emotional disclosure and posting topics on social media added to relevant research realm, especially from the angle of microbloggers’ demonstration of their intercultural attitudes and cultural themes of interests.

Practically, findings of this study were informative about the strengths and weaknesses, respectively, of females and males when responding to cultural differences. In general, the observed gender gaps in intercultural attitudes urge researchers and educational practitioners to take this issue into account when designing interventions to enhance individuals’ intercultural competence, so as to maximize the intervention effectiveness by attending to potential gender differences. Specifically, the inspection of microblog posts with negative sentiment values alerted that males were more likely to hold the beliefs that their own cultural group is superior to others and were less willing to approach culturally dissimilar people or situations. According to prior intervention studies, web-based exploratory activities (e.g., group discussion on academic essays, news reports, personal blogs, and videos) that involved learners’ own culture and a different culture (Stockwell 2016, 2018) and international online collaborations (i.e., students from two different cultures worked in teams to discuss and deliver PowerPoint presentation on similarities and differences on various aspects of their cultures) (Boehm et al. 2010) were effective in reducing individuals’ ethnocentric beliefs and fostering more favorable and flexible attitudes toward different cultures.

Moreover, attention also needs to be paid to the negative attitudes of a crowd of women toward intercultural experiences, as can be seen in their microblog posts (i.e., anxiety and confusion about cultural differences in thinking patterns, and values and beliefs). As suggested by Carroll and Harris (2020), repetitive instructional intervention can be applied to improve learners’ ability to establish connections between materials, thereby increasing learners’ accuracy and confidence in interpreting complex issues. Translated into the current context, instead of requiring learners to memorize materials describing typical characteristics of a certain culture, educators can guide learners to repeatedly build connections between the abstract aspects (e.g., values and beliefs) and the concrete aspects (e.g., behavioral patterns, customs) of people in a culture—that is, to decipher the impact of a group of people’s cultural values and beliefs on their shared cognitive, emotional, and behavioral responses when functioning in their community. This would help improve learners’ higher-order thinking skills to accurately dissect and understand different worldviews and minimize the risk of experiencing culture shock.

It is necessary to note that the findings of this study need to be interpreted with cautions. First, although microblogging content was less strictly structured than self-reported survey items, individual’s self-disclosure on social media can also be biased. The Self-Affirmation Theory (Steele 1988) held that individuals had fundamental needs for positive self-regard. As such, individuals can be more inclined to display their positive images more often than negative ones on social networking platforms (Li 2019), and consequently information published on the platforms can often be overly positive rather than negative (Verduyn et al. 2021) to maintain their self-worth. Therefore, it is possible that some microblog users have fabricated their posts to downplay their confusion about cultural differences. Second, one notable limitation of this study is the potential for inaccuracies in gender information provided by users, as they may deliberately misrepresent their gender or select categories that do not align with their true identity. This discrepancy can significantly impact the accuracy of the gender-based analysis, leading to potential biases in the results. Similarly, the absence of age information for both creators and users further limits the study’s comprehensiveness. Age can influence both the content created and the reception of that content, potentially interacting with gender in ways not captured by the current analysis. Future research should incorporate age demographics alongside gender to provide a more nuanced understanding of how these factors interact and affect content creation and user engagement. Third, the microblogging data used in this study only contained data from Sina Weibo (Chinese X [formerly Twitter]). The findings may be mitigated by involving data from other microblogging platforms. Last, the sample of this study only included Chinese microblog users, and the findings thus might not be able to be generalized to population of other cultures. It would be interesting for future studies to investigate this topic using a larger dataset that includes more culturally diverse samples.