How Does Self-Efficacy in Communication Affect the Relationship Between Intercultural Experience, Language Skills, and Cultural Intelligence

Cultural intelligence (CQ) is a multidimensional socio-mental construct that allows a person to effectively explore and understand different cultures for the benefit of the entire organization. Therefore, many studies investigate which factors support or hinder CQ development. Our study anticipates that intercultural social contracts, travel experience and knowledge of foreign languages affect CQ. We further explore how these direct relationships are mediated by the variables of self-efficacy in communication. To confirm or reject our hypotheses we used a sample (n = 190) of university students studying at a private Czech university and a public Russian university that is analyzed using the statistical technique PLS-SEM. Our results confirm that intercultural social contacts, the experience of traveling abroad, and knowledge of foreign languages are positively related to CQ, and that self-efficacy partially positively mediates these relations. The study can serve as the base for other research concentrating on the mediators of the factors supporting CQ.


Introduction
Cultural intelligence (CQ) investigates how a person is able to cope with a culturally diverse environment (Ang et al., 2007).Due to the current globalization, CQ is considered an essential workplace competence for the 21st century.Therefore, CQ began to be developed and trained in academic courses at many universities (Barnes et al., 2017;Fischer, 2011;Jura´sek & Wawrosz, 2021); in addition, organizations hold various workshops for their employees dedicated to the training of intercultural skills (Van Dyne et al., 2016) and to the development of cultural intelligence (Livermore, 2011;Lovvorn & Chen, 2011).Aware of the importance of CQ for the competitiveness of organizations in international markets, many studies are trying to find factors that contribute to or hinder CQ.Some results are the following: studying or working abroad is a more effective factor for increasing CQ than just (leisure) tourism abroad (Crowne, 2008); previous stays and traveling abroad increases CQ (R. L. Engle & Crowne, 2014); and nonworking international experience has a greater influence on the development of CQ than working experience (Moon et al., 2012).J. Lee et al. (2018) summarize the main factors affecting CQ: (1) number of completed academic courses focused on the development of intercultural skills; (2) knowledge of foreign languages; (3) how frequent are social contacts with people from foreign countries; and (4) scope of experience with traveling abroad.The study concluded that knowledge of foreign languages and academic development of intercultural skills are positively related to CQ.It was also stated that it depends more on the intensity (depth) than the frequency of intercultural interactions.When it comes to the experience of traveling abroad, the duration of stay in a foreign country and the number of countries visited contribute, compared to the frequency with which a person travels abroad, more to the development of CQ.However, the results of this study largely contradict several other conclusions from similar research.For example, Tarique and Takeuchi (2008) stated that the number of countries visited predicts an increase in CQ in all its four dimensions (metacognitive, cognitive, motivational, and behavioral), while the duration of stay predicts only the cognitive and metacognitive component of CQ.Consistent results were not obtained even on the impact of the frequency of foreign trips on the development of CQ: while J. Lee et al. (2018) confirmed a positive (significant) correlation, Tarique and Takeuchi (2008) did not.
These inconsistent and contradictory claims are usually caused by other variables mediating relationships between the CQ and its antecedents.Our text builds on previous research using a new variable: self-efficacy in communication.We examine whether the direct relationship between CQ and its antecedents (intercultural experience and knowledge of foreign languages, when all antecedents positively affect CQ), is further positively mediated by self-efficacy in communication.The use of this mediator distinguishes our study from others, or in other words, self-efficacy in communication is considered to be an explaining variable of the relationship between biographical variables and CQ for the first time.Our study thus helps to understand the specific role of self-efficacy as a driver that directs attention and effort toward learning about and functioning in a situation characterized by cultural differences.Many self-efficacy scales suitable for different situations and contexts have been constructed (Bandura, 2006).CQ researchers have applied various types of self-efficacy (SE) construct in their studies: creative SE (AlMazrouei & Zacca, 2021), sales SE (Charoensukmongkol & Pandey, 2020), expatriate's SE (Hu et al., 2018), general SE (MacNab & Worthley, 2012;Nguyen et al., 2018) and task-specific SE (Rehg et al., 2012).However, the list of self-efficacy types that regulate cultural intelligence is not complete.In our view, more self-efficacy facets fit well into the character of the CQ construct.Therefore, we focus on self-efficacy in communication that has not been used in CQ research before.
We develop a new conceptual model explaining relationships among intercultural experience (represented by travel, frequency of intercultural contacts, development of intercultural skills), language skills, self-efficacy, and CQ.The originality of the paper, from our point of view, consists in the proposed theoretical framework that explores a new facet, that is, self-efficacy construct as an explaining variable (mediator) of the relationship between the biographical variables and CQ.Moreover, assuming the positive effects of self-efficacy in communication on CQ the study raises the question of which biographical variable fuels self-efficacy in communication most and thus can be used for practical purposes.Our hypotheses are further tested on the sample of students studying in the Czech Republic and Russia who are usually not the subject of the CQ research.As a result, our study extends the CQ literature in various ways: it helps understand the direct relationship between some biographical variables and CQ, and it introduces a new explanatory variable to this relationship with the focus on a very important role of cross-cultural issues, that is, communication.
The article is organized as follows.Review of the literature briefly discusses our main variables-cultural intelligence, intercultural experience, and self-efficacy.The next chapter describes how research usually investigates the relationship between all variables and formulates hypotheses on this basis.The material and method describe our sample, research methods, and measures.The results are introduced in the following chapter and are further explored in discussion.Conclusion summarizes main findings.

Cultural Intelligence
CQ investigates how a person is able in a culturally diverse environment (Ang & Van Dyne, 2015;Earley & Ang, 2003).It is a multidimensional socio-mental construct that allows a person to effectively explore and understand different cultures for the benefit of the entire organization (Darvishmotevali et al., 2018).CQ as a certain ability creates conditions for cooperation in situations characterized by cultural diversity regardless of race, ethnicity, and nationality (Mahembe & Engelbrecht, 2014).The research of cultural intelligence is based on the work of Sternberg and Detterman (1986), who created an integrative model of multiple intelligences, that is, abilities of mental, motivational, and behavioral nature with special emphasis on solving intercultural problems.Cultural intelligence belongs to the variables measuring a person's ability to accommodate a culturally new and unknown environment (Earley et al., 2006).CQ is not related to the culture of the individual's origin, but it comes from a person's own knowledge and experience (S xahin et al., 2014).
Cultural intelligence shows how flexible a human is and what competence a human has in the following three areas: how (s)he knows a foreign culture, how (s)he is able to notice cultural nuances, similarities, and differences, and to act adequately and naturally in the new cultural environment based on the correct interpretation of these observations (Yitmen, 2013).The topic of CQ began to be developed at the beginning of the 21st century.It is considered (Ang & Van Dyne, 2015) a more comprehensive approach than others-for example, the concept of intercultural competence, which began to be used in the 1980s primarily in communication literature.The concept of CQ is mainly investigated in economic management studies (Triandis, 2006) or in the environment of a particular organization (Earley & Ang, 2003).

Intercultural Experience
The intercultural experience is researched by various approaches, such as the cross-border cross-cultural competence model (see Figure 1).This model emphasizes that the lives of temporary residents-foreigners are especially affected by the following intertwining factors: (1) differences of individuals in various aspects (e.g., personality traits); (2) characteristics of foreign experience (social contacts, perceived language discrimination, duration of stay, level of language skills); and (3) internal processes of a person (e.g., reflections on cultural experiences) in context sociocultural environment.The model is based on three pillars: (1) cultural awareness (how a person is aware of his/her cultural worldview and attitudes toward cultural differences or similarities of people from specific countries and cultures); (2) cultural knowledge (whether a person understands different worldviews and sociohistorical context and practices connecting with specific cultural groups and countries); and (3) cultural skills (if a person is effectively able to act, communicate, cooperate and establish contacts with people from other cultures).
The development of cross-border cultural competence is based on three factors: (1) individual personality traits; (2) experiences of intercultural character (characteristics of contact with or a certain immersion into another culture/merging with another culture); and (3) active reflection of intercultural experiences.L. Wang et al. (2017) selected curiosity, exploration, and perseverance (i.e., enthusiasm for long-term goals) in their study for the first category as certain personality traits that are important for well-being, motivation, and the learning process in an intercultural environment.In the second category (immersion in a foreign culture and associated experiences), K. T. Wang et al. (2015) chose the following: duration of stay, perceived language discrimination (objective and subjective language skills), and relationship with the host culture (specifically social connection with mainstream society).Frequent contact with locals will increase CQ; on the contrary, perceived language discrimination and anxiety can significantly reduce CQ, although CQ will rise sharply again after some time (up to 2 months after arrival).

Self-efficacy
The author of the self-efficacy concept is Canadian-American psychologist Albert Bandura (Bandura, 1977).The concept concentrates on how a person is selfconfident in his/her ability and skills to plan and act in a way necessary to achieve a certain goal, to manage a situation or task in the broadest sense.If a person has a low self-efficacy score (s)he probably avoids social interactions in an unknown (cultural) environment.They are worse at adapting to the new context.Self-efficacy can be seen as a predictor of better intercultural adaptation in a culturally different environment.Higher self-efficacy allows people to perform their newly learned behavior in foreign situations while subsequently receiving feedback on this behavior.As a result, they reduce their uncertainty in expectations and manage to behave in a culturally appropriate way, leading to better intercultural adaptation.During stays abroad, people with higher self-efficacy usually show better self-regulation, wellbeing, and state of health.Students with study experience abroad receive more job offers, perhaps because they are more empathetic, patient, and confident.
Literature (Ng & Earley, 2006) found that self-efficacy is an important attribute of motivational CQ, specifically: (1) the ability to handle (MacNab & Worthley, 2012) the challenges of the international environment; (2) the issues of motivation; and (3) the setting of goals (Barakat et al., 2015).High self-efficacy helps a person to cope well with problems and obstacles.Such a person does not afraid of an unknown international (intercultural) environment, (S)he actively seeks opportunities and social interactions and considers them as a valuable source of experience.Higher self-efficacy is also correlated with a stronger intention to go abroad to work and to develop CQ (Camargo et al., 2020).Some studies (e.g., Alexandra, 2018) indicate that internal motivation (self-efficacy) is even more important for CQ development than managerial and practical experience abroad.

Intercultural Experience (Travel, Frequency of Intercultural Contacts, Development of Intercultural Skills) and CQ
When a person is exposed to the influence of a foreign culture, the person merges better with the values, norms, and religiosity of a certain culture.This influence can be either superficial (e.g., through travel, reading, watching TV shows, studying the relevant literature, talking directly to someone from this culture) or deeper (e.g., through a long work stay in the country, migration, frequent business trips to countries of a certain culture, studies, long-term stays in another country, humanitarian missions or foreign military service) (Sousa & Gonc xalves, 2017).It can be argued that individuals (e.g., international students) will feel more culturally competent abroad if their sense of communication and connection with mainstream society (i.e., with insiders/locals) is stronger (K.T. Wang et al., 2015).Livermore and Soon (2015) described four steps to how a person can become culturally intelligent.Studying foreign languages and mastering them at an advanced level, regular trips abroad and gaining experience from contact with foreigners (J. Lee et al., 2019), or more frequent communication with people from other cultures contribute to the development of CQ (Urnaut, 2014).International travel, work stays, nonwork (recreational) trips abroad or study stays abroad, as well as perceptions of own efficiency, were confirmed as CQ antecedents (Crowne, 2008;MacNab & Worthley, 2012;Tarique & Takeuchi, 2008).It can be said that the nature of international experience does not matter that much.All forms of stay abroad (nonwork, tourist, short-term) develop CQ (Crowne, 2008;Douglas & Jones-Rikkers, 2001; R. L. Engle & Crowne, 2014;Takeuchi et al., 2005).On the other hand, it depends on the destination of the trip.R. L. Engle and Nash (2016) showed that it is better for the development of cultural intelligence (and all its components) to spend time in countries that are relatively culturally foreign to one's own culture (i.e., when the cultural values of this destination and the domestic culture are not too similar).Thanks to the experience gained from the intercultural environment, people can better adapt during their new stay abroad; in other words, foreign stays and trips moderate the relationship between CQ and intercultural effectiveness (adaptation) (R. L. Engle & Crowne, 2014).In addition, it depends on the individual's personal opinions and characteristics.For example, it was found that the relationship between CQ and (a) knowledge of foreign languages, (b) teaching skills of intercultural cultures at the university, (c) daily contact with people of foreign cultures, and (d) experience with international travel is weaker in individuals who show a lower degree of ethnocentrism (J. Lee et al., 2018).Table 1 summarizes the antecedents (including relevant studies) that have an impact on the development of intercultural skills.
Based on previous knowledge and research, we assume the following: H1: CQ and intercultural social contacts with people from foreign cultures are positively correlated.H2: Travel experience and CQ are positively correlated.

Language Proficiency and CQ
The cognitive component of CQ enables one to identify and get to know the cultures (in the sense of cultural norms, practices, and conventions), their similarities, and difference.The cognitive factor of CQ can be seen as the critical factor of overall CQ because high cognitive CQ allows a person to better interact with people from culturally diverse backgrounds.Language knowledge is rather important in this respect; language skills (especially good knowledge of English) are positively connected with overall satisfaction and atmosphere of multicultural working groups (Cramer, 2018).
The relationship between CQ and foreign language skills has been tested in several studies.Knowledge of the first foreign language was a variable to validate CQ measurements of Colombian university students (Robledo-Ardila et al., 2016).In another study that validated CQ measurements in Ukraine, the authors used the following variables: international experience and knowledge of foreign languages (Johnson, 2014).L. Wang et al. (2017) found in a sample of Taiwanese students studying in the United States that CQ grows when students master the language of the host country, when they are more curious and happier to discover, reflect on their cultural experiences, and seek (or maintain) social contacts with the majority population of the country.The study emphasized that perseverance in discovering new cultural dimensions of the country is related to metacognitive and motivational CQ.Thanks to all the CQ components (except behavioral), the students were able to notice and realize that they are to some extent overlooked by the locals due to poorer knowledge of the local language (in this case, English).It is also true that the metacognitive and motivational components of CQ increase with the length of stay in a foreign country.Previous research that has examined CQ shaping factors has been the basis for an extensive empirical study by Tharapos et al. (2019).In a sample of accounting teachers working at universities in Australia, they confirmed the relationship between CQ (that acts as a dependent variable) and many independent variables: employment at a university in the state capital or territory, gender, age, length of work experience, number of foreign languages known (high proficiency), work experience abroad with teaching at foreign universities (in Asia and the Anglo-Saxon world) and long-term stay abroad.Another study (Miele & Nguyen, 2020) confirmed that the ability to speak another foreign language is a strong predictor of CQ, but this does not apply to extracurricular (university) international and multicultural activities.
Language skills (as well as living in diverse cultural environments and work experience from another cultural environment) predict a relationship with CQ (R. Engle & Nehrt, 2012;Triandis, 2006).Working in a multicultural team (Robledo-Ardila et al., 2016) contributes to the increase of metacognitive and motivational CQ components.Alon et al. (2018) used another measurement of CW, the Business Cultural Intelligence Quotient for the investigation of CQ antecedents.According to the results of their study, the most important factors for the development of cultural intelligence of business professionals (practitioners) are the following (in this order): the number of countries in which businessmen lived for more than 6 months, their level of education, and the number of languages they speak.
Finally, a developed CQ helps with learning foreign languages; students with a higher CQ are more motivated to learn about foreign cultures, which includes acquiring language skills.Therefore, they achieve better results in language tests (Alahdadi & Ghanizadeh, 2017).In this sense, it has previously been found that CQ correlates positively with the proficiency achieved when learning foreign languages (English) (Khodadady & Ghahari, 2012).Based on the above-mentioned research, we assume: H3: CQ and knowledge of foreign languages are positively correlated.

Self-Efficacy as a Mediator of CQ and Biographical Quantities (Knowledge of Foreign Languages and Frequency of Social Contacts)
General self-efficacy is a key variable in predicting the successful development of culturally intelligent behavior (MacNab & Worthley, 2012).Self-efficacy and CQ are related to the success of university studies (Iskhakova, 2018), for example, learning foreign languages or confidence in communication in various social situations.Selfefficacy is positively correlated with intercultural adaptation (Peterson et al., 2011).The positive relationship of CQ to adaptation to culturally foreign conditions has been confirmed by many studies (Ang et al., 2007;Che Rose et al., 2010;Huff, 2013;Huff et al., 2014;Imai & Gelfand, 2010;Jyoti & Kour, 2015;Kumar et al., 2008;L. Lee, 2010; L. Y. Lee & Kartika, 2014; L. Y. Lee & Sukoco, 2010;Malek & Budhwar, 2013;Ramalu et al., 2010;Subramaniam et al., 2011;Zhang & Oczkowski, 2016).H4: Self-efficacy in communication mediates the direct relationship between CQ and knowledge of foreign languages, the frequency of intercultural social contact with people from different cultures, and the experience of traveling abroad.
Figure 2 explains our theoretical research model.In our study, we focus on communication self-efficacy, which may be related to self-efficacy in general.It is true that not only self-efficacious intercultural communicators but also individuals with a high CQ act confidently and effectively in several unknown situations.Conversely, people with low CQ or those who do not consider themselves good intercultural communicators perceive an unfamiliar environment as a certain threat and try to avoid stress, for example, by minimizing or eliminating intercultural social interactions.Persons with good language skills and high proficiency in the host country's language can be reasonably expected to be more and more easily involved in intercultural social interactions; they will be better and more effective (in terms of internal motivation) intercultural communicators.Therefore, we assume the following (see Figure 2): 1. (Better) language skills, (higher) contact frequency of relationships with foreigners and (often) traveling abroad directly affect CQ (increase CQ) and contributes to (higher) self-efficacy too.2. (Higher) self-efficacy supports (higher) CQ.

Materials and Methods
We examine our hypotheses using an online questionnaire in English.It was filled in by 220 respondents who study at two universities: University of Finance and Administration in Prague (a private university in the Czech Republic) and State University of Management in Moscow (a Russian state university).Students with good English proficiency were asked to complete the questionnaire voluntarily.We assume that this requirement was met for most of the respondents because they studied in an English study program.It can be generally summarized that most of them come from Russia and other post-Soviet countries, from the Czech Republic and from China.As part of the data cleaning process, the structure of the data was examined using standard deviations, and those cases where the respondents were not very involved in filling in the questionnaire (i.e., unengaged respondents) were omitted from the questionnaire.In cases where, according to general recommendations (Hair et al., 2016), blank items per respondent accounted for more than 5% of the total, these responses (30 persons) were excluded from the analysis.In other cases, incomplete data were replaced by the average method.The responses from 190 respondents were analyzed in total.It is necessary to emphasize that the participation in the survey was completely anonymous and voluntary; it was not associated with any positive motivation (e.g., the possibility of getting credits).Respondents agreed to participate in the survey and the questionnaire followed standard ethical rules (see Israel, 2014 for details).
The gender structure of the respondents is the following: female 123 (64.74%), male 67 (35.26%).Most of the students (156, i.e., 82.10%) were aged 18 to 25, specifically 64 (33.68%) aged 18 to 20, 77 (40.53%) aged 21 to 23, and 15 (7.89%) aged 24 to 25.The respondents were from 25 countries, mainly from the Czech Republic (36.32%),Russia (24.74%),Ukraine (8.95%) and Kazakhstan (7.89%).Some of them had relatively extensive international experience-50 (26.32%) students spent more than 1 year abroad (either for work or study), 21 (11.05%)students more than 2 years, 44 (23.16%) students more than 3 years, and 19 (10%) students more than 4 years.Twentytwo students had experience with the Erasmus study program, 84 (44.21%) students worked abroad.CQ was measured using the 9-item Mini-Cultural Intelligence Scale (Mini-CQS) developed by Ang and Van Dyne (2008).The scale is a holistic measure of CQ in the four dimensions: metacognitive, cognitive, motivational, and behavioral CQ.For example, an item states: ''I am sure I can deal with the stresses of adjusting to a culture that is new to me.''A seven-point Likert-type scale ranging from 1 (strongly disagree) to 7 (strongly agree) was used for expressing the respondent's position.A study of Van Dyne et al. (2008) demonstrated that Mini-CQS has good reliability and generalizability across multiple student samples.Eisenberg et al. (2013), J. Lee et al. (2018) show that the scale reliability alpha coefficient of the Mini-CQS exceeded the standard cutoff value of 0.70 in previous research.Cronbach's alpha for our scale was .77.
Intercultural travel experiences were measured using items that included questions like: ''How often do you interact with people from different countries or cultures?''(1: never -7: always), ''How many times have you traveled internationally?''(1: zero -7: more than 6), and ''How long did you stay in foreign countries for education, work, vacation or another?''(1: never -7: 3 years or more).However, since the last two elements were below 0.7, they were removed for high reliability of indicator loadings and the travel experience variable was measured by only one element (J. Lee et al., 2018).
Intercultural social contact (J. Lee et al., 2018) was assessed using three elements: ''In your daily life (e.g., work, school, neighborhood, etc.), how often did or do you interact face to face with people from different cultures or countries?''(1: never -7: always), ''In your daily life (e.g., work, school, neighborhood, etc.), ''How intensive would you describe your interactions with people from different countries or cultures?''(1: not interactive -7: very interactive), and ''To what extent do you interact with people from different countries or cultures on social media?''(1: never 2 7: always).Cronbach's alpha for this scale was .78.
In the case of foreign language skills participants were asked how many foreign languages they speak (1: none -7: more than three) (J. Lee et al., 2018).Language proficiency was measured using nine elements that were used in the study by Luna et al. (2008).Using a scale of 1 = very low to 5 = like a native speaker, the respondent was asked to self-evaluate how fluent they are in speaking and listening in English.In this respect, their language skills are above average.Then they were asked to indicate how well they think they can do things such as ''understand newspaper headlines'' or ''read popular novels without using a dictionary'' (1 = very bad, 5 = very well).Cronbach's alpha for this scale was .93.
An adapted questionnaire (Peterson et al., 2011) was used to measure self-efficacy in communication.Fifteen items with a loading factor greater than 0.6 were selected.Examples of items: ''How well can you inspire others to gain new insight when you communicate with them?''or ''How well can you think of possible outcomes before you speak?''(1 = not very well, 5 = very well).Cronbach's alpha for this scale was .73.
Our research used partial least-squares structural equation modeling (PLS-SEM) to confirm or reject the hypothesis.PLS-SEM consists of a broad class of methods that model relationships between sets of observed variables using latent variables.PLS-SEM tries to maximize explained variance (R 2 ) in the dependent variable, and to minimize the residual variance of endogenous variables in any regression run by the model, and to evaluate the utility of the data by evaluating the measurement model (Hair et al., 2016).PLS-SEM seems to be, in comparison with covariance-based SEM (CB-SEM) models, an appropriate statistical modeling technique.It is a more flexible model due to the minimal demands on measurement scales, sample size, and residual distributions.Researchers can, by using PLS-SEM analyze multiple hypotheses that involve single and multiple-item measurements at the same time.Data investigated by the model does not have to be normally distributed and can also contain smaller sample sizes (Puyod & Charoensukmongkol, 2019).Furthermore, PLS-SEM does not estimate all model parameters simultaneously; instead, it only estimates partial model structures, one equation at a time (Sharma, 2019;Vlajcic et al., 2019).Testing a mediating effect of the communication selfefficacy variable on cultural intelligence, the sampling distribution was bootstrapped.Bootstrapping ''makes no assumptions about the shape of the variable's distribution [.], can be applied to small sample sizes and yields higher levels of statistical power compared with the Sobel test'' (Hair et al., 2016).The PLS-SEM technique is suitable to provide the researchers with three types of validity: conclusion validity (to demonstrate whether there is a relationship between two variables), internal validity (assuming that there is a relationship between the explored variables, the causality between variables is indicated) and construct validity (evaluating the measurement models of all variables).
Accuracy of data entry, missing values, outliers, and multicollinearity was examined for all variables.Incomplete data were addressed in two ways: we either removed all cases from the analysis that include missing values in any of the indicators used in the model (casewise deletion) and data were replaced by mean value because there were less than 5% values per indicator (Hair et al., 2016).The sharpness coefficient of all variables was also determined using SPSS software.If this coefficient does not lie in the interval \22, + 2., it would mean that the respondents answered the question very similarly.In our case, this coefficient did not exceed the required limit; this indicates that the respondents answered the given questions with a certain variance, which is desirable for further statistical analysis.Skewness does not need to be measured for latent variables measured on the Likert scale.Sometimes it can happen that a respondent does not show much interest in the research and answers all items (questions) with the same number (i.e., does not read the individual items, does not think about them, answers automatically); these are unengaged respondents.These types of responses were found in Excel by calculating the standard deviation of latent variables.If the standard deviations found were less than 0.5, the responses were removed from the analysis based on a thorough examination and consideration of these borderline cases.Extreme values (remote observation of outliers in continuous variables) were found in SPSS.Because the questions were closed, the respondents did not have the opportunity to ''make a mistake'' (e.g., wrong hand placement on the keyboard) and there were no extreme values in demographic information (such as age).Evaluating our conceptual (mediation) model, all quality criteria of the measurement models suggested by Hair et al. (2016) have been explored and met.

Results
In addition to the PLS-SEM algorithm for iterative estimation of latent variable scores (Hair et al., 2011), we further use the bootstrapping technique to evaluate the statistical significance of the path coefficients (Gabel-Shemueli et al., 2019).The measurement model presented in Table 2 satisfies the required reliability and validity criteria.The data correspond to various benchmarks (Hair et al., 2016) for reflective measurement models such as composite reliability 0.70 (internal consistency), factor loadings 0.70 (indicator reliability), average variance extracted (AVE) .0.50 (convergent validity), square root of variance extracted from each construct .any vertical or horizontal correlation (Fornell-Larcker criterion -discriminant validity), HTMT ratios \.90, and HTMT ratios confidence intervals do not include 1 (discriminant validity).
Convergence reliability was confirmed using factor loadings.Table 2 shows that all Cronbach's alpha values (..7) and the loadings of all items indicate the reliability of the indicators.All items that did not meet the threshold (or minimum requirement) for the reliability of the indicators (loadings .0.5) were removed.All AVEs (Average Variance Extracted) .0.5 indicate convergence reliability because the AVE values range between 0.554 and 0.699 (not considering the single-item construct of Traveling).All composite reliability values (CR .0.7) indicate internal consistency.
Discriminant validity was confirmed-the average variance extracted (AVE) was compared with the squared correlation coefficient.The Fornell-Larcker criteria (Fornell & Larcker, 1981) ensure that the constructs are truly distinct from each other.The square root of the AVE (the bold diagonal in Table 3) must be greater than the correlations between the constructs.We further checked discriminant validity with hetrotrait-monotrait (HTMT) ratios.All ratios are \0.9 and their confidence of intervals do not include 1 which satisfies the discriminant validity criteria (but for the sake of clarity the values are not reported in Table 4).
Multicollinearity was investigated by using full variance inflation factor (VIF) statistics.The highest value of full VIF was found at 3.512 and it is lower than the maximum threshold (see Table 3 for details).The common method bias test (CMB) was finally made by Harman's one-factor test, as suggested (Jarvis et al., 2003).The results showed that a single factor in extracting is 26.66% of the variance.The value is far less than 50% and we can conclude that there is no threat of CMB (Podsakoff & Organ, 1986).
PLS regression analysis is presented in Table 4. H1 predicted a positive relationship between intercultural contact frequency and CQ.The results show a positive relationship, which is also statistically significant (b = .158;p \ .1).Thus, H1 was supported.H2 predicted the positive relationship between travel and CQ.The H2 was supported (b = .145;p \ .1).H3 predicted a positive relationship between language skills and CQ.
The result supported a positive relationship that was also statistically significant (b = .248;p \ .001).Thus, H3 was supported.
As can be seen in Table 4, all the biographic variables explored in this study are related to the variable of selfefficacy in communication.Specifically, there is a positive relationship between contact frequency and intercultural self-efficacy in communication (b = .139;p = .082),between travel and self-efficacy in communication (b = .175;p \ .05),and between language skills and selfefficacy (b = .347;p \ .01).H4 predicted that self-efficacy mediates the direct relationships between CQ and language skills, contact frequency, and travel experience.The result shows that self-efficacy partially mediates the relationship between language skills and CQ (b = .075;p \ .05).The mediation is complementary and not extraordinarily strong.This is the only mediating effect of the explored relationships that has been found statistically significant.Thus, H4 was only partially supported.

Discussion
Cultural intelligence antecedents can be classified as individual (e.g., international experience, personality traits, certain states such as anxiety) and situational (nature of the task, team or organization environment, job roles, etc.).Our article deals with the first category of CQ antecedents: intercultural experience in the form of knowledge of foreign languages, frequency of social contacts in an intercultural environment and experience resulting from traveling or staying abroad.Our results are largely in line with previous research (Li, 2017;Moon et al., 2012;Pawlicka et al., 2019): our research confirmed that the variables examined are predictors of cultural intelligence and contributed to the CQ literature with new knowledge on what factors matter in terms of the development of intercultural skills.This can be used, in the interest of the development of intercultural skills, by university teachers or guarantors of study programs and courses to create new educational modules and change the syllabi of existing courses.Based on the acquired knowledge, training workshops and courses at universities should focus on the development of language skills, support student internships abroad (work or study) and stimulate intercultural social contacts, for example, by hosting regular faculty ''International Days,'' where students from different countries will present their country and culture.
Although individuals do not benefit in the same way from acquired international experience and previous contact with foreign culture (e.g., travel or other intercultural interactions), these factors undoubtedly affect the development of CQ (Li, 2017).Intercultural experience has a beneficial effect on the development of CQ (Pawlicka et al., 2019).CQ is a predictor of intercultural effectiveness, which concerns mental well-being in intercultural situations, intercultural adaptation, the potential of multicultural teams, the negotiation process, team cohesion, and integration.CQ also predicts the frequency of intercultural interactions (Pawlicka et al., 2019); our study also points to the opposite predictive relationship: intercultural social interactions (their frequency) contribute to the development of CQ.
Previous travel experience abroad and the intensity of social contacts with foreigners (i.e., culturally different people) are directly related to CQ (J. Lee et al., 2019).It was stated that non-work experience (e.g., personal and language-focused stays) has a stronger influence on CQ than work experience from abroad (Moon et al., 2012).This can be explained by the fact that certain limitations caused by work responsibilities and tasks to some extent limit a person in freely expanding his or her cultural horizons.A certain role is also played by the individual's predispositions and characteristics and how well he or she is able to participate in the course of events and cognitively process the signals coming from the new cultural environment (S xahin et al., 2014).In this sense, language skills and the inner desire to participate proactively and interactively in the life of the local community play an important role.Based on the results of this study, it can be stated that people who can speak well (i.e., who have a high level of language communication skills) will be more willing to establish communication with foreigners, which will subsequently have a positive effect on the development of their intercultural skills (or CQ).This study adds a new dimension to the self-efficacy construct explored by CQ researchers earlier, focusing on it from the point of view of communication.It has been found that this specific type of self-efficacy construct has an unquestionable impact on the models with the CQ variable.Our findings relativize the need to specify the selfefficacy variable.Obviously, the construct works well under different circumstances.Self-efficacy conceptualized by many as a state-like construct (Rehg et al., 2012) is likely to be spread out to different areas.This explains why a variety of types of self-efficacy (including our contribution to the list of different types of self-efficacy construct) can be used when explaining the relationship between CQ and some other variable.As a result, this finding has a practical implication for decision makers and managers in an organization when they think about sending their employees for a foreign assignment: by demonstrating communication skills (or competency in communication) with compatriots, people are likely to be active (and successful) in communicating or dealing with foreigners in an unfamiliar cultural setting when some preconditions (such as good foreign language skills) are met.It is not surprising that self-efficacy as a kind of ''inner urge'' or motivation will persistently nurture an individual's desire to consciously expand their knowledge and subsequently transform it into skills that can be used practically in a new cultural environment (Kang et al., 2019).

Conclusions
Our research confirmed the findings of previous studies (e.g., Li, 2017;Moon et al., 2012;Pawlicka et al., 2019) that intercultural social contacts, the experience of traveling abroad, and knowledge of foreign languages are positively related to CQ.We further investigated whether these relationships are mediated by self-efficacy in communication.This variable was placed in a new position within a theoretical model to explain the relationship explored earlier between some biographical variables and CQ.The benefits of this theoretical approach were twofold: not only the indirect effects of the biographical variables on CQ (via the mediator of self-efficacy in communication), but also the direct effects between the biographical variables and self-efficacy in communication respectively CQ were tested.It is comprehensible (and our results confirm it) that self-efficacy in communication from all monitored biographical variables is closest to the means of communication, that is, knowledge of foreign languages (we found that self-efficacy partially mediates the relationship between language skills and CQ).Both variables complement each other, create a synergic effect and contribute to the development of CQ.
It must be emphasized that other variables (specifically travel experience and frequency of intercultural social contacts) also help to the development of self-efficacy.Although our study did not find their direct impact on CQ development, further research is needed to confirm or reject our findings.The research should also consider factors such as ethnocentrism (J. Lee et al., 2018) or culture shock (Jura´sek & Wawrosz, 2021).The dependence between CQ and self-efficacy can be also the opposite than it is investigated in our study, that is, CQ acts as the predictor of self-efficacy in communication.
Our results are based on the answers to the questionnaire.The method raises the possibility of responses being affected by common method variance (CMV) (Jarvis et al., 2003).To avoid such potential distortion other studies should include multiple sources of observations or control for social desirability, which is commonly assumed to cause CMV (Alexandra, 2018).Last but not least, one threat to the validity of our findings might be associated with the fact that the accessible population was used.Only further research based on different samples (more representative in terms of nationality, age or gender) that hold for the broader population can confirm our findings and result in a higher degree of external validity and generalizability of our conclusions.Generally, only longitudinally designed studies could reveal more definitive conclusions on the issue which antecedents can increase CQ.
Source.Own research.

Table 3 .
Discriminant Validity and Correlation Matrix.