Role of dispositional aspects of self-identity in the process of planned behavior of outbound travel

Abstract

Identity theorists suggest that one’s self-identity affects his behavioral intent directly, while planned behavior researchers believe that one’s attitude toward a specific behavior has a direct effect on his intention to perform the behavior. Few studies have ever examined how one’s self-identity and situational attitude may interact on his behavioral intent. This study aims to fill this gap by examining the role of one’s dispositional aspects of self-identity in the process of planned behavior in the context of outbound travel. Based on a survey on Chinese outbound travel, this study verifies that self-identity does not exert direct influence on outbound travel intents; instead, the effect of self-identity is mediated by the effect of the planned behavioral constructs. This study concludes that, in the context of outbound travel, the dispositional aspects of self-identity contribute to outbound travel intentions within the structure of the theory of planned behavior. Both theoretical contributions and practical implications are discussed.

Keywords

China evaluative attitude outbound travel planned behavior self-identity

Introduction

Tourism literature has taken on the issue of self-identity in better understanding the relationships between identity and tourism. Research on tourism and visitors’ identity can date back to a special issue of Annals of Tourism Research guest edited by Keyes and van den Berghe (1984), with a focus on studying the impact of tourism on identity. Lanfant et al.’s (1995) edited volume International Tourism: Identity and Change examines the relationships between tourism and identity in more dynamic ways, for instance, looking at how identities of visitors and local people are mutually constructed.

Literature relating to tourism and identity tends to focus on identity issues, such as ‘finding yourself’ through travel. Noy’s (2004) study of Israeli backpackers’ narratives demonstrates how they used tourism to construct their identities; Thurlow and Jaworski’s (2006) study illustrates how visitors use travel to affirm identities. Research undertaken by Becken (2007) indicates that tourists see travel as part of their identity; their holidays have a great meaning to them and help them understand their positions in society.

While many tourism studies focus on how travel helps affirm visitors’ self-identities, relatively, few has ever examined the role of self-identity on one’s travel behavior. Desforges (2000) suggests that understanding identity can lead to an insight into tourism consumption, because by understanding the people and their needs and desires, it is possible to predict their future travel patterns. Hibbert et al. (2013) explored how a need for personal identity can influence one’s travel behavior and verified that identity plays a significant role in making a travel decision.

In contrast, the theories of reasoned action (TRA) (Ajzen and Fishbein, 1980) and planned behavior (TPB) (Ajzen, 1985) are about attitude–behavior relationships. The literature indicates that attitudinal evaluations may be distinguishable from identity-related symbolic outcomes (e.g. McCracken, 1986). While researchers aligned with the TRA or TPB may consider one’s self-identity to be reflected in the person’s beliefs, values, and attitudes (Sparks and Shepherd, 1992), there is a body of empirical studies indicating the importance of self-identity in the overall structure of the TRA and TPB (e.g. Hagger et al., 2007; Stets and Biga, 2003; Strack and Deutsch, 2004). For instance, Hagger et al. (2007) examined the effects of self-identity on behavioral intentions (BIs) in the context of TPB for three health behaviors: exercise, dieting, and binge drinking, but failed to identify direct effects of the identity constructs on BIs. In contrast, Biddle et al. (1987) suggest that one’s self-identity and other cognitions (such as attitudes or preferences) may not always be consistent; Charng et al. (1988) suggest that self-identity may influence one’s BI independent of evaluative attitudes.

Attitude research and identity research have both attracted a big amount of research interest in the context of tourism research. A few studies have even examined the interactive role of one’s self-identity and attitude on BIs (e.g. Hagger et al., 2007; Sparks and Shepherd, 1992). However, to the authors’ best knowledge, no empirical inquiry in this regard in the context of travel behaviors has ever been conducted, and the relationship between self-identity and attitude toward travel behaviors remains unknown. Drawing on the theories of self-identity and planned behavior, this study aims to fill this gap by testing how the aspects of self-identity are associated with people’s evaluative attitudes toward travel behaviors, in the context of outbound travel.

Literature review

A person may have a great variety of behavioral engagements, such as exercise, dieting, and binge drinking, which are health-related as investigated in a study by Hagger et al. (2007). In comparison, international travel is a type of behavior different from these activities, which, for instance, might involve an adventurous tour away from home or an experiential visit to an exotic destination. The type of tour and expected outcome from the tour may more likely reflect and affirm one’s personal identity or social identity. Given that more empirical inquiries are needed to examine the relationship between self-identity and evaluative attitude in guiding one’s behavior, and no such studies have ever been conducted in the field of tourism, this study is designed to examine how the dispositional aspects of self-identity and the TPB components interact to influence one’s outbound travel intention. To achieve the goal of this study, literature review was conducted and introduced in this section in terms of the relationships between TPB components and BIs, self-identity components and BIs, and the interactive role TPB components and self-identity components on one’s intentional outbound travel behavior.

Attitude behavior relationships

The TPB (Ajzen, 1985) has emerged as an influential social cognitive model widely applied to understand one’s BI. The main concept of TPB is that most human behavior is under volitional control; people engage in actions because they want to act in a certain behavioral way, and their conscious motives trigger them to engage in that action (Ajzen and Fishbein, 1980). The TPB suggests that behavioral achievement depends on three types of beliefs—attitude, subjective norm (SN), and perceived behavioral control (PBC). Among them, attitude toward a behavior is, at the most basic level, a function of behavioral beliefs and outcome evaluations (Ajzen, 1991). Attitudes toward a specific behavior are theorized to have a direct effect on their intentions to perform the behaviors (Ajzen, 1988).

The second TPB component is SN, which refers to one’s perceived social pressure from important others (such as relatives, friends, and colleagues) to perform or not to perform a behavior (Ajzen and Fishbein, 1980). It is believed that people are more likely to perform a behavior when they receive support from their referents than when they do not (Ajzen and Fishbein, 1980). In the context of tourism, if your close friends or relatives all think that a certain destination is more attractive or worth visiting than other destinations, it is more likely for your intention of visiting this destination to arise as compared to visiting other destinations. The third TPB component is PBC, which is the degree to which an individual feels that the performance or nonperformance of a behavior in question is under his or her volitional control (Ajzen, 1985, 1988), which, in the context of tourism, is about one’s perception of travel constraints or inhibitors, such as physical mobility, disposable income, and leisure time.

Based on the above, this study posits that the TPB’s three primary components—attitude, SN, and PBC—significantly influence international travel BIs, respectively.

Despite its effectiveness in explaining people’s behavior intentions, Ajzen (1985) suggests that the TPB model be viewed as a flexible theoretical framework to explain the psychological influence on intentional behavior, and other independent variables can be encompassed provided they increase the predictive validity of the theory. That said, the TPB model does not account for all the variance in predicted behavior, and considerable variance may remain unexplained. As such, the TPB model may not be able to provide a complete view as to why people perform a certain behavior, and researchers should look for other important variables which contribute to the predicting power of the model. In this vein, recent research has highlighted the importance of integrating the role of trait-like, dispositional aspects of self-identity in predicting people’s BIs within the structure of TPB (e.g. Hagger et al., 2007).

Self-identity behavior relationships

According to Stets and Biga (2003), a self-identity is ‘a set of meanings attached to the self that serves as a standard or reference that guides behavior in situations’ (p. 401). Self-identity refers to salient and enduring aspects of one’s self-perception (Rise et al., 2010). Self-identity is how individuals see and describe themselves with reference to others in society (Hibbert et al., 2013). ‘Self’ represents more than just a collection of individualized attributes that remain constant over time and across contexts (Niedenthal and Beike, 1997). In terms of the aspects of self-identity, Cheek and Briggs (1982) suggest that people’s self-representations tend to be either associated with their private conceptions of self or their social roles. In other words, one’s self-identity can be understood from the two aspects of personal identity and social identity. In contrast to people’s planned behavior, given that these aspects of identity are dispositional constructs, they are more likely to represent people’s more impulsive, unplanned routes to behavioral engagement (Strack and Deutsch, 2004).

Self-identity perceptions may affect individuals’ intentional behaviors by serving as a source of information when people make plans to act (Hagger et al., 2007; Wang and Xu, 2015). For example, a person who claims to be a ‘green tourist’ may be more likely to choose green-friendly destinations, accommodations, and/or travel programs. It is considered that a person with a sense of personal identity tends to develop personal beliefs of control over a given behavior. Hagger et al. (2007) suggest personal identity and social identity both affect BIs directly. In terms of travel behavior, Bond and Falk (2012) suggest that a tourist’s travel experience can be wholly motivated by his self-perceived identity-related needs.

With respect to the above discussion, it is hypothesized that personal and social identities have significant direct impacts on international travel BIs, respectively.

Interactive role of self-identity and evaluative attitude

Early studies on the role of self-identity (e.g. Biddle et al., 1987, in a study of college retention decisions; Charng et al., 1988, in studying blood donation intentions) claim that self-identity contributes to the prediction of BIs independently of evaluative attitudes. Biddle et al. (1987) suggest that self-referent identity labeling has an effect on behavior which is independent of the effect of individual preferences. These authors found that students’ ratings of the sort of person they thought they were contributed to intentions independently of the contribution of preferences. Charng et al. (1988) argue that role identity as a blood donor may influence intentions to donate blood independently of attitudes toward blood donation. Charng et al. demonstrated that regressions of intentions to donate blood on attitudes and role identity revealed an independent effect for each of the two predictor variables. However, Sparks and Shepherd (1992) expressed suspicion about the theoretical rationale for expecting an independent effect of self-identity on BIs, debating that a person’s self-identity should be reflected in that person’s beliefs, values, and attitudes. Sparks and Shepherd (1992) argued that the affirmation of a particular self-identity may be considered justifiably to be a behavioral outcome and that such outcome may not be especially salient when respondents rate behaviors on the evaluative scales used to assess attitudes toward the act (p. 397). As a result, Sparks and Shepherd (1992) posited that self-identity would influence attitudes but would not show an independent effect on BIs.

In Sparks and Shepherd’s (1992) study on green consumerism, however, the direct effects of self-identity on BIs were verified, which was contrary to their postulation and led them to a conclusion that a person’s self-identity may be distinct from his evaluative attitudes. It is observed that, in Sparks and Shepherd’s (1992) study, only direct effects were measured and the interactive effects of self-identity with evaluative attitudes were not investigated. Being aware of the contradiction, Spark and Shepherd (1992) stressed the importance of capturing both the direct and indirect effects of self-identity and recommended that a more sophisticated modeling technique (e.g. structural equation modeling) be applied in identifying other effects such as the mediating effects of the attitudinal factors.

Notably, the ‘symbolic’ role of self-identity has rarely been considered within the TPB assessment model (Stets and Biga, 2003). Stets and Biga (2003) observed that previous research did not take the role of identity into account and argued that psychology-based attitude theory only demonstrates how individuals make choices based specifically on the object or situation they are in. Hagger et al. (2007) argued that the outcome beliefs that underpin TPB’s three primary components (i.e. attitude, PBC, and SN) are likely to be influenced by a tendency to rate personal outcomes and resources more highly. For instance, a person with a higher level of personal identity may be more likely to develop personal beliefs of control over international travel behavior. Within the TPB assessment model, Hagger et al. (2007) examined the effects of people’s self-identities on their intentional exercise, dieting, and binge drinking. To be different from the previous studies (e.g. Sparks and Shepherd, 1992), their results showed no direct effects of self-identity on the health BIs; instead, self-identity’s effects on BIs were indirect, mainly through the mediating TPB’s primary components.

To measure the interactive effects of self-identity and TPB’s primary components, this study would need to measure the relationships between self-identity and TPB components, hence hypothesized that both personal and social self-identities significantly influence TPB’s attitude, SN, and PBC in terms of outbound travel, respectively.

Other than Hagger et al.’s (2007) study, not many other studies have ever examined the mediating role of TPB components when discussing the effects of self-identity on human BIs. None of such studies seems to have been conducted in the field of leisure travel, leaving the interactive effects of self-identity with evaluative attitudes on travel BIs unverified. Hence, one of the main objectives of this study is to better understand the relationships between self-identity and travel BIs within the TPB assessment model. Based on the literature review, a conceptual research framework about people’s outbound travel intentions is developed based on the posited relationships between the aspects of self-identity and the TPB components (see Figure 1).

Figure 1.

The conceptual framework of self-identity and planned behavior.

As shown in Figure 1, 11 hypothetical relations are developed as shown below, which are tested in this study.

H1 and H2: Personal identity and social identity significantly influence attitude on one’s BIs, respectively;

H3 and H4: Personal identity and social identity significantly influence SN on one’s BIs, respectively;

H5 and H6: Personal identity and social identity significantly influence PBC on one’s BIs, respectively;

H7 to H9: The TPB’s attitude, SN, and PBC significantly influence international travel BIs, respectively;

H10 and H11: Personal and social identities have significant direct impacts on international travel BIs, respectively.

Research methods

Place of study and data collection

Empirical inquiries are made in this study focusing on Chinese outbound visitors. According to China National Tourism Administration, the number of Chinese outbound tourists reached 109 million in 2014 and the overseas spending reached US$129 billion in 2013. Since 2013, the number and spending of China’s outbound tourists have both topped all the other travel markets in the world, and even with the slowdown of China’s economy, the World Tourism Organization predicted that China would still be the largest outbound travel market in 2016 (CNTA, 2016). While China has rapidly become the world’s top international tourist-generating market, there has been a growing interest in studying China’s outbound tourism. It is believed that information provided by this study will help people of interest better understand Chinese tourists’ self-identity, attitude, and their interactions in the process of outbound travel planning.

This study is designed to focus on outbound tourists instead of domestic visitors in China based on two considerations: One is that, for people not only from China but also from most of the countries today, outbound travel is still considered to be something luxurious which can help us better build self-esteem and a positive self-identity; the second consideration is that, while China is becoming a top tourist-generating market, findings of this study will be helpful for the global travel companies and professionals to better understand China’s outbound travel market.

To better understand how one’s self-identity may affect his or her outbound travel intention, a self-administered survey was conducted in China’s mega city Shenzhen, located in the southern province of Guangdong. Shenzhen is one of China’s top four outbound tourist-generating cities with the highest growing rate, about 1.21 million residents traveling overseas in 2013 (Shenzhen-HK, 2014). The research assistants who are students from a tourism college in Shenzhen helped collect the data from May 20, 2015 to July 20, 2015. The data collection sites include the metro stations along line no. 1 and line no. 3, and the streets around Bai Shi Zhou and High-tech Park, which are always gathered by lots of people. Given the limited resources and financial support for this study, it is deemed unrealistic to apply a probability sampling technique to do a survey by including every individual in a mega city like Shenzhen; moreover, based on the researchers’ past experience, it is estimated that people’s response rate will be low when conducting a survey in a busy mega city in China. As a result, a convenience sampling technique was used for data collection in the places where subjects were selected because of their convenient accessibility and proximity to the researchers. City residents who were at least 18 years old walking around the four locations were intercepted and asked for participating in the survey. As a result, a total of 547 usable surveys were collected. On average, the response rate is reported to be about 20% on the sites around the metro stations and 12.5% at Bai Shi Zhou and Hi-tech Park.

Questionnaire design

The survey instrument is a cross-sectional questionnaire, consisting of three sections. The first section contains items measuring the TPB constructs of attitude, SN, and PBC; the second section includes the measures of the two dispositional aspects of self-identity—personal identity and social identity; and the third section asks questions about respondents’ demographic characteristics.

Semantic scaling questions anchored with bipolar adjectives are used to measure the construct of ‘attitude’, as are typically used in tourism studies (Baloglu and Mangaloglu, 2001). The bipolar adjectives are unfavorable/favorable, unpleasant/ pleasant, boring/exciting, and foolish/fun. It is noted that previous studies used one single item (e.g. Song et al., 2012) to measure ‘subjective norm’. According to Armitage and Conner (2001), a single-item measure of SN may become a weak predictor of BIs; hence, three items are used to measure the construct of SN, which are ‘it is popular among my friends or family’, ‘People who are important to me would probably think that it would be good to take a holiday abroad’, and ‘Friends or family have recommended I take a holiday abroad’. The items used to measure ‘perceived behavioral control’ are adopted from Song et al. (2012), which are ‘I feel I have enough time to take a holiday abroad’, ‘I feel I have enough money to take a holiday abroad’, and ‘I feel there is nothing that prevents me from taking a holiday abroad if I want to’. Besides, two items are used to measure the construct of ‘behavioral intentions’ which are ‘I intend to take a holiday abroad in the next 1-2 years’ and ‘How likely would you take a holiday abroad in the next 1-2 years’. Notably, a holiday abroad means a leisure trip to a foreign country. Since Hong Kong and Macao are China’s special administrative zones, visiting Hong Kong or Macao is not counted as a holiday abroad in the minds of Chinese today.

The aspects of identity questionnaire conceptualize the self as a dual identity structure comprised of both personal identity and social identity, which is designed to assess differences in the importance people assigned to both internal and external aspects of identity (Cheek, 1989). Social identity was operationalized as the tendency to categorize oneself in terms of one’s aggregate group identifications and personal identity as the tendency to individuate the self as distinct from in-group memberships. To measure the two dispositional aspects of identity constructs, this survey instrument adopts six-item personal identity and six-item social identity scales adapted from Nario-Redmond et al. (2004). One example of the items measuring personal identity is ‘My feeling of being unique, distinct from others’ and one example of social identity items is ‘The similarity I share with others in my groups’.

Prior to the survey, a pretest of the questionnaire was conducted in Shenzhen. Forty completed surveys were collected and analyzed; comments from the research assistants who conducting the survey were collected. As a result, the format of the survey and the wordings of some of the questions were amended. The survey instrument was originally developed in English and then translated into Chinese by the investigator who are proficient and fluent at both English and Chinese languages. All the items relative to the measurement of the TPB and self-identity constructs were measured on seven-point Likert-type scales, with ‘1’ standing for the most negative perception and ‘7’ standing for the most positive perception. The third part of the questionnaire contains the demographic variables such as employment, income, age, education, occupation, and gender.

Data analysis techniques

The data were screened for violations of underlying assumptions based on descriptive statistics, using the Statistical Package for the Social Sciences (SPSS 23). Data were cleaned prior to data analyses. Each of the univariate distributions has skew and kurtosis within reasonable ranges (<3 and <10, respectively), the values falling within the guidelines and being regarded fairly normal for further structural equation modeling (SEM) analyses (Kline, 2005). The data were then analyzed with LISREL (8.80), which is a statistical analytic software, to run the covariance structure analysis. This approach advocates initially estimating a measurement model and then a structural equation model. The goodness of fit indicators demonstrating a good fit for the structural model was inspected, based on the indices of χ ²/df, p-value, comparative fit index (CFI), normative fit index (NFI), and root mean square error of approximation (RMSEA). This study hypothesized that the aspects of self-identity may not only exert direct effects but also exert indirect effects on BIs mediated by TPB components; thus as for the two self-identity constructs, their direct, indirect, and total effects are also examined. In addition, the Sobel test (Sobel, 1986) is conducted to assess how the effects of the aspects of self-identity on BIs are mediated by the TPB components.

Results

The results of the respondents’ demographic profiles are displayed in Table 1. Of the 547 participants, 53.6% are male and 46.4% are female. Of the respondents, 52.7% are 30 years old or younger, 33.7% are between 31 and 50 years, and 13.6% are over 50 years old. Regarding marital status, 36% are singles and under 35 years old, 29% are married and under 35 years old without kids, 26% of the respondents have kids, and the other 9% are over 35 years old who are either singles or empty nesters. About 67.3% are full-time employees, 9.1% are self-employed, 5.7% are part-time, 4.6% are job-waiting, 4.2% are retired, and 9.1% are the others. In terms of their occupations, 26.9% are clerks or white-collars, 18.1% are businessmen, 16.5% are civil servants or teachers, 11% are managers, 8.8% are students, 2.6% are blue-collar, 2% are retired, and 14.1% are the others. In terms of the respondents’ monthly household income, about 16% are below CNY5000 (approximately US$1 = CNY6.4), 50% are between CNY5000 and CNY10,000, 14% are between CNY10,001 and CNY20,000, and the other 20% are above CNY20,000.

Table 1.

Results of respondents’ demographic profiles.

		%			%
Age	18–30	52.7	Gender	Male	53.6
	31–50	33.7	Gender	Female	46.4
	Over 50	13.6	Monthly household income	Below CNY5000	16
Marriage status	Single under 35	36		CNY5000–10,000	50
	Married under 35, no kids	29		CNY10,001–20,000	14
	With kids	26		Above 20,000	20
	Over 35, singles/empty nesters	9	Occupation	Clerks/white-collars	26.9
Employment status	Full-time	67.3		Businessmen	18.1
	Self-employed	9.1		Civil servants/teachers	16.5
	Part-time	5.7		Managers	11
	Job-waiting	4.6		Students	8.8
	Retired	4.2		Blue-collar	2.6
	Others	9.1		Retired	2
	Others	9.1		Others	14.1

Based on the mean scores of the items, the summated mean scores of the constructs, from the highest to the lowest, were attitude (6.09), personal identity (5.93), social identity (5.62), outbound travel BI (5.20), SN (5.15), and PBC (5.00). The results indicate that, generally, the respondents’ perceptions of the TPB and self-identity constructs are positive, ranging from 5.0 to 6.09. Among the variables, attitude received the highest mean score, and the two identity components received higher scores than all the other variables.

The goodness of fit tests of the posited measurement model and structural model were conducted with Lisrel 8.80. Confirmatory factor analysis of the constructs—attitude toward international travel (ATT), SN, PBC, and BI—was conducted. The fit indices were as follows: χ ² = 945.33 (df = 238, p < 0.001), NFI = 0.96, CFI = 0.95, and RMSEA = 0.074. Overall, the measurement model showed a good fit for the data (MacCallum et al., 1996). The χ ² value looks to be a bit high, and this is because the χ ² test is sensitive to sample size and the χ ² statistic nearly always rejects the model when large samples are used (Jöreskog and Sörbom, 1993). Convergent validity was assessed by the significant loadings between the observed variables and each latent variable. All the observed variables were loaded above 0.40 on their delegated latent variables and were statistically significant (p < 0.01). As shown in Table 2, all the average variance extracted (AVE) values were above 0.50 ranging from 0.525 to 0.604, which supported adequate internal consistency except the latent variable of ‘social identity’ (AVE = 0.432) (Hair et al., 1998). The composite reliabilities of all constructs exceeded the cutoff value of 0.70 (Hair et al., 1998). Cronbach’s α values of the constructs range from 0.798 to 0.940; thus, the multiple item scales were acceptable for measuring each of the constructs. To compare the AVE with the squared correlations between constructs for discriminant validity testing (Fornell and Larcker, 1981), the squared correlations between each pair of constructs were all less than the AVE values. Thus, discriminant validity was satisfied. Overall, the measurement model shows a good fit to the data.

Table 2.

Measure correlations, the squared correlations, and measurement properties (N = 547).

Correlations between latent constructs (squared)
Measures	PI	SI	ATT	SN	PBC	BI
PI	1	0.64 (0.410)	0.61 (0.372)	0.63 (0.397)	0.20 (0.040)	0.40 (0.160)
SI	0.64 (0.410)	1	0.52 (0.270)	0.65 (0.423)	0.34 (0.116)	0.41 (0.168)
ATT	0.61 (0.372)	0.52 (0.270)	1	0.43 (0.185)	0.17 (0.029)	0.38 (0.144)
SN	0.63 (0.397)	0.65 (0.423)	0.43 (0.185)	1	0.21 (0.044)	0.49 (0.240)
PBC	0.20 (0.040)	0.34 (0.116)	0.17 (0.029)	0.21 (0.044)	1	0.71 (0.504)
BI	0.40 (0.160)	0.41 (0.168)	0.38 (0.144)	0.49 (0.240)	0.71 (0.504)	1
AVE	0.566	0.432	0.604	0.543	0.525	0.61
Mean	5.93	5.62	6.09	5.15	5.00	5.20
Cronbach’s alpha	0.858	0.893	0.879	0.804	0.798	0.940

PI: personal self-identity; SI: social self-identity; ATT: attitude; SN: subjective norm; PBC: perceived behavioral control; BI: outbound travel behavioral intentions; AVE: average variance extracted; RMSEA: root mean square error of approximation; CFI: comparative fit index; NFI: normative fit index.

Note: Model measurement fit: χ ² = 945.33 (df = 238, p < 0.001), RMSEA = 0.074, CFI = 0.96, and NFI = 0.95.

The structural model was estimated to examine the hypothetical relations. The results showed that the goodness of fit indices (goodness of fit statistics: χ ² = 1324.12 (df = 241, p < 0.001), RMSEA = 0.08, CFI = 0.96, NFI = 0.94) were all within their acceptable level, suggesting that the model is adequate. The constructs of personal identity and social identity explained 39%, 49%, and 12% of the variances in attitude, SN, and PBC, respectively; then, the self-identity components and TPB components altogether explained 73% of the variance in the construct of BI, indicating a good fit of the proposed structural model.

Table 3 shows the paths’ standardized coefficients and the corresponding t values. The significant paths include both the gamma paths (relationships between exogenous constructs and endogenous constructs) and the eta paths (relationships between endogenous constructs) in the model. Based on the t values, 9 out of the 11 hypothetical paths showed to be significant. It is noted that the two insignificant paths were from personal identity to PBC and BIs, respectively. In addition, unlike all the other significant paths which are all positive, the direct effect from social identity to BIs is negative.

Table 3.

Standardized maximum likelihood parameter estimates (N = 547).

Paths	Standardized coefficients	T Value	Results of hypothesis testing
PI → ATT	0.47	7.36	H1: supported
SI → ATT	0.20	3.20	H2: supported
PI → SN	0.33	5.30	H3: supported
SI → SN	0.42	6.28	H4: supported
PI → PBC	−0.06	−0.82	H5: not supported
SI → PBC	0.39	4.78	H6: supported
ATT → BI	0.18	3.40	H7: supported
SN → BI	0.35	5.39	H8: supported
PBC → BI	0.71	12.29	H9: supported
PI → BI	0.06	0.92	H10: not supported
SI → BI	−0.20	−2.65	H11: supported

Note: Goodness-of-fit statistics: χ ² = 1179.61 (df = 241, p < 0.001), RMSEA = 0.080, CFI = 0.96, and NFI = 0.94.

Both the indirect effects and total effects of personal identity and social identity were examined (see Table 4), which were compared to their direct effects as displayed in Table 3. Interestingly, it is found out that personal identity’s indirect and total effects are both positive and significant despite its insignificant direct effects on BIs; in addition, the indirect and total effects of social identity show to be positive and significant, too, in spite of the negative direct effect on BIs. Social identity’s effect on BIs is negative while its indirect effect is positive; then, the two effects cancel each out and result in a suppressed total effect which is positive and significant. This result indicates that the direct effect of social identity has been suppressed by the stronger indirect effects. As a result, both personal identity and social identity are shown to have positive total effects on BIs.

Table 4.

Decomposition of effects with standardized values.

Indirect effect					Total effect
	ATT	SN	PBC	BI	ATT	SN	PBC	BI
PI	—	—	—	0.17	0.41	0.38	−0.06*	0.24
SI	—	—	—	0.54	0.19	0.52	0.44	0.31
ATT	—	—	—	—	—			0.22
SN	—	—	—	—	—	—	—	0.34
PBC	—	—	—	—	—	—	—	0.74

PI: personal self-identity; SI: social self-identity; ATT: attitude; SN: subjective norm; PBC: perceived behavioral control; BI: outbound travel behavioral intentions.

*Effects that are not significant; all the other effects are significant.

Further, the Sobel test (Sobel, 1986) was utilized to test the significance of the mediating roles of TPB components to the relationships between the aspects of self-identity and outbound travel intentions. The results of the Sobel tests, based on the Sobel test statistics (STS) and two-tailed p-values, are listed in Table 5. As shown in the table, both attitude and SN play an important role in governing the relationships between personal identity and BI and between social identity and BI; in contrast, PBC mediates the relationship between social identity and BI, not between personal identity and BI. On the one hand, the results indicate that personal identity and social identity act as antecedents to TPB constructs’ impacts on outbound travel BI; on the other hand, although no direct effect on visit intention was observed from personal identity and social identity’s direct effect that is suppressed by its strong indirect effect, they exert positive indirect influence on BI through the mediating variable of the TPB variables. Hence, all the TPB and self-identity constructs, that is, personal identity, social identity, attitude, SN, and PBC, prove to be significant in influencing one’s outbound travel BI, either directly or indirectly.

Table 5.

Results of indirect-effect tests (Sobel test).

Independent V. → Mediator V. → Dependent V.	STS	Two-tailed p-value	Results
PI → ATT → BI	2.86	0.004	Significant
PI → SN → BI	3.92	0.001	Significant
PI → PBC → BI	−0.75	0.454	Insignificant
SI → ATT → BI	2.23	0.025	Significant
SI → SN → BI	4.27	0.001	Significant
SI → PBC → BI	4.55	0.001	Significant

PI: personal self-identity; SI: social self-identity; STS: Sobel test statistics; ATT: attitude; SN: subjective norm; PBC: perceived behavioral control; BI: outbound travel behavioral intentions.

To further specify, the Sobel tests show that the two TPB components of attitude and SN are significant in mediating the associations between ‘personal identity’ and ‘travel intention’; ‘behavioral control’ failed to mediate the relationship. Decomposition of effects indicates that personal identity has no significant direct effect on travel intention, whereas its total effect on travel intention is positive and significant. In other words, its total effect on travel intention is mainly exerted by its indirect effects mediated by attitude and SN. In contrast, social identity has both direct and indirect effects on travel intention; the indirect effects mediated by all the three TPB components. As a result, it is found that both personal identity and social identity exert positive impacts on the three TPB components; the only exception is that there is no significant impact made by personal identity on PBC. The influence of personal identity and social identity on travel intentions is mainly mediated by all the three TPB components, respectively; their direct impacts on travel intentions are either insignificant or suppressed by the indirect effects through the three TPB constructs.

Discussion and conclusion

Results of this study verify that the dispositional aspects of self-identity have significant effects on people’s outbound travel BIs. However, self-identities do not contribute to their BIs independently of their attitudinal evaluations. This result echoes the findings of Hagger et al. (2007) who focused on the effects of self-identities on people’s health behaviors but did not observe direct effects of self-identity on health BIs. This study is conducted partially in response to Sparks and Shepherd’s query: Using a more sophisticated analytical approach (e.g. SEM in this study) will make it possible to get a holistic picture of the effects of self-identity on BIs. Results of this study indicate that self-identity’s direct effects are overwhelmed by its interactive effects with the TPB components. According to Sparks and Shepherd (1992), one reason why self-identity may not predict intention after TPB components are taken into account is that self-identity concerns may not have been especially salient when people responded to the evaluative scales typically used to tap attitude toward behaviors. Thus, it is verified that, in the context of outbound travel behavior, empirical evidence of the independent effect of self-identities on people’s outbound travel intentions is nuance.

In terms of the indirect effects of self-identities on international travel intentions, findings of this study show that the personal aspect of self-identity (PI) impacts the two TPB constructs—attitude and SN—which in turn influence travel intentions; the social aspect of self-identity (SI) makes impacts on all the three TPB constructs—attitude, SN, and PBC—which in turn affects travel intentions. Thus, the effects of the aspects of self-identity on travel intentions are mediated by the TPB’s three primary components. In other words, the aspects of self-identity exert indirect effects on travel intentions within the assessment model of the TPB. Results of the Sobel tests further prove that all these indirect effects mediated by the TPB constructs are significant.

This study verifies that both the effects of PI and SI on outbound travel intentions are mediated by TPB’s attitude. In other words, one’s cognition of his personal self and social self can affect his attitude toward outbound travel intentions. For instance, a person who desires to become competent and unique (i.e. one’s PI) will be more likely to adopt the idea of traveling abroad, and a person who wants to become a better-recognizable social person (i.e. social aspect of self-identity) tends to like more the idea of outbound travel.

The effects of both personal identity and social identity on outbound travel were also found to be mediated by the variable of SN. The old Chinese proverb of ‘Read ten-thousand books, travel ten-thousand miles’, which is deeply embedded in Chinese people’s mind to become a person of wisdom, denotes that learning from books does not complement traveling for a person to become intelligent (which strongly indicates one’s PI) and respectable (which is a strong indicator of one’s SI). A person’s sense of self or self-concept is inextricably linked with the norms of the in-group which suggests that self-identity considerations are likely to influence normatively endorsed behavior (Tajfel, 1981). On the one hand, the association between the aspects of self-identity and perceived SN could be due to people’s common cognition of international travel as one’s most effective learning course to enhance knowledge; on the other hand, traveling abroad is commonly accepted in China as an effective approach to enhancing one’s social networking and status. One example to illustrate how important the travel-abroad experience is perceived by Chinese is that nowadays, it is getting popular for wealthy Chinese parents to send their teenage children to a summer camp in Australia or New Zealand.

In comparison, this study indicates that the relationship between SI and PBC is significant, while the relationship is insignificant between PI and PBC. In other words, the variable of PBC significantly mediates the effects of SI, not PI. In this regard, the Chinese term ‘Mianzi’ (meaning keeping face or being respectable) may well explain why one’s SI influences his PBC more than PI. For many Chinese people, not to lose face could be more important than not to lose a ‘fortune’. It is construed that a Chinese person with a strong sense of social identity will simply feel to lose Mianzi by admitting that he or she cannot make an outbound travel while his friends or colleagues can. When one’s friends or colleagues are planning to travel abroad, he would tend to assure himself that he is capable of making a similar trip, too. In contrast, PI (e.g. desire to become distinctive or unique) did not show a significant impact on PBC. This implies that, in the context of outbound travel, SI’s influence on Chinese people’s evaluative attitudes is more prevalent than PI’s.

To summarize the findings of this study, on the one hand, no direct effects of the aspects of self-identity on outbound travel intentions are observed; the effects of PI on people’s BIs are insignificant at the presence of the effects of the TPB constructs; the effects of SI are suppressed by the salient interactive effects of social identity with the TPB’s primary components. On the other hand, the positive effects of PI and SI on outbound travel BIs are exercised mainly through the mediating variables of the TPB’s primary components. This study empirically verifies that self-identities’ indirect effects through TPB components are more distinctive than their direct effects. Overall, results of this study reinforce the cognition that the dispositional aspects of self-identity significantly contribute to people’s outbound travel intentions within the TPB assessment model.

In terms of theoretical contributions, this study provides a different perspective in understanding the roles of self-identity in formulating one’s evaluative attitudes and BIs in the context of outbound tourism. There have been various kinds of studies reporting the roles of self-identity in different disciplines or behaviors; noticeably, some studies pinpoint self-identity’s direct effects on determining one’s BIs independently of evaluative attitudes, while the others suspect the independent role of self-identify. Findings of this study indicate that, in the context of outbound travel, the dispositional aspects of self-identity do not have significant direct effects on outbound travel intentions, but are closely associated with people’s attitudes, their perceptions of SN, and behavioral control abilities.

This study verifies that PI and social self-identity act as important antecedents to the TPB constructs in predicting outbound travel intentions. This is because the evaluative attitudes in the TPB are formulated on the basis of situational evaluations or utilitarian outcomes which are more decisive in explaining destination choice, while self-identities are based on dispositional and/or symbolic behavioral outcomes which are important in initiating outbound travel desire. By identifying the relationships between the aspects of self-identity and TPB components, this study integrates the two orientations and stresses the importance of studying one’s self-identity when assessing one’s beliefs, values, and attitudes toward outbound travel.

This study shows that the aspects of self-identity exert no direct effects on BIs in the context of outbound travel. Self-identity’s effects are either mediated or suppressed by the TPB components. Based on the results, it verifies that self-identity’s effects cannot be independent of the influence of people’s evaluative attitudes. Hence, it is recommended that self-identity should not be used as an independent predicting variable within the TPB assessment model, and studies of outbound travel behavior should better take the dispositional aspects of self-identity into consideration within the framework of people’s evaluative attitudes.

With regard to the practical implications, this study implies that Chinese tourists’ expected travel outcomes are closely associated with their needs to nurture and enhance self-identities. In other words, it means Chinese tourists will like to express their self-identities by choosing to visit the destinations with features capable of helping strengthen their perceived self-identities. Consumers often prefer the brands or the products which have a high level of congruence with self-image (Sirgy, 1986), and associating the specific personality features to the brands allows the consumer to express a certain conception of himself, so as to acquire some value-enhancing, symbolic benefits from a given consumption (Vernette, 2003). In destination marketing, the implication is that a destination management organization (DMO) should not only promote the inherent attributes and the characteristics of the destination but also identify and highlight the destination’s personality features. Articulating and promoting destination’s personality features will be a boost in attracting tourists whose self-identities are congruent with the personality features.

What’s usually covered in destination marketing campaigns, brochures, or websites are more about the physical appeal and services offered by the destination; what could be overlooked by tour companies and DMOs are visitors’ prospects of symbolic travel outcomes. Results of this study indicate that a travel destination which best meets a visitor’s desire for PI and SI enhancement will help augment his evaluative attitudes, which in turn reinforce his visit intentions. In this regard, today’s popular experiential tourism is a good example to show the importance of incorporating travelers’ desire for enhancing PI through travel; volunteer tourism is another good example illustrating the importance of meeting visitors’ SI needs in tourism, social self-identity needs. To attract Chinese tourists, one notion for travel companies and DMOs is that Mianzi is an important feature and element in Chinese culture. Destination organizations may consider incorporating some special events or activities aiming to enhance Chinese tourists’ Mianzi, for example, a welcome ceremony and/or a special memorable souvenir offered by the DMO or local government as a token of welcome and appreciation.

The effects of Chinese social self-identity seem to be more prevalent than PI as it shows to impact one’s travel intentions through all the three TPB components. It is considered that Chinese people tend to favor the destinations or trips where travel outcomes can make them feel more socially respectable or have a better Mianz. For a typical budget tourist, for example, with the travel expenses being equal, he would more likely prefer to choose a one-week tour covering five Western European countries instead of a one-week tour only covering two of the five Western European countries, even if this means having to stay in cheaper or less comfortable hotels and having much less sightseeing activities in each place. One implication for international travel companies is that, when designing outbound tours for Chinese market, a tour covering more places with the same price tends to be more hailed by a typical Chinese visitor. This is because what interests him might be that he can visit more places in one trip, and the number of visited places is believed to be closely related with his social self-identity; to put up with a low-quality board and lodging in the trip as a tradeoff will not be an issue. In contrast, nonbudget tourists increasingly favor in-depth experience and independent travel, as quality (not quantity) of a type may best reflect their social identity and Mianzi.

Tailor-made deluxe tours offering in-depth experience or independent travel would be a better choice and more attractive for a nonbudget tourist, as such tours meet with his social identity need can better reflect his social identity and make them feel more distinguished, thus create more Mianzi.

One caution of this study is that it has focused on one’s BIs, which is one step away from an actual behavior. BI is an individual’s decision or commitment to perform a given behavior which is the most proximal predictor of behavior (Ajzen and Fishbein, 1980). According to Young and Kent (1985), however, although it is generally accepted that BI could predict actual behavior, a high connection between BI and actual behavior may not always be attained. Given the difference between BI and an actual behavior, it is recommended that researchers and industry practitioners should be aware of the difference when applying the findings and consider it necessary to make more marketing and management efforts in order to convert people’s BI to an actual behavior. If conditions are favorable, future studies should apply this conceptual framework to the measurement of visitor’s actual travel behavior. Because this study was conducted with data collected from one of the most affluent populations in China, its findings may not generalize well across a wider range of demographics. In addition, further research should replicate this study in other countries, as converging evidence from these studies would not only consolidate the reliability and validity of the measures but also help create a more generalized theory on the link between people’s self-identities and planned behaviors.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Ajzen

(1985) From intentions to actions: a theory of planned behavior. In: Kuhl

Beckmmann

(eds) Action-Control: From Cognition to Behavior. Heidelberg: Springer, pp. 11–39.

Ajzen

(1988) Attitudes, Personality and Behavior. Milton Keynes, UK: Open University Press.

Ajzen

(1991) The Theory of Planned Behavior. Organizational Behavior and Human Decision Processes 50: 179–211.

Ajzen

Fishbein

(1980) Understanding Attitudes and Predicting Social Behavior. Englewood Cliffs, NJ: Prentice-Hall.

Armitage

Conner

(2001) Efficacy of the theory of planned behavior: a meta-analytic review. British Journal of Social Psychology 40: 471–499.

Baloglu

Mangaloglu

(2001) Tourism destination images of Turkey, Egypt, Greece, and Italy as perceived by U.S.-based tour operators and travel agents. Tourism Management 22: 1–9.

Becken

(2007) Tourists’ perception of international air travel’s impact on the global climate and potential climate change policies. Journal of Sustainable Tourism 15(4): 351–368.

Biddle

Bank

Slavings

(1987) Norms preferences, identities and retention decisions. Social Psychology Quarterly 50(4): 322–337.

Bond

Falk

(2012) Who am I? and why am I here (and not there)?: The role of identity in shaping tourist motivations. International Journal of Tourism Research. doi: 10.1002/jtr.1886.

10.

Charng

H-W

Piliavin

Callero

(1988) Role identity and reasoned action in the prediction of repeated behavior. Social Psychology Quarterly 51(4): 303–317.

11.

Cheek

(1989) Identity orientations and self-interpretation. In: Buss

Cantor

(eds) Personality Psychology: Recent Trends and Emerging Directions. New York: Springer-Verlag, pp. 275–285.

12.

Cheek

Briggs

(1982) Self-consciousness and aspects of identity. Journal of Research in Personality 16: 401–408.

13.

CNTA (2016) UNWTO predicts China continues to be the top outbound travel market. Available at: http://www.cnta.gov.cn/xxfb/hydt/201603/t20160314_763803.shtml (accessed 15 March 2016)

14.

Desforges

(2000) Traveling the world: identity and travel biography. Annals of Tourism Research 27(4): 926–945.

15.

Fornell

Larcker

(1981) Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research 18(1): 39–50.

16.

Hagger

Anderson

Kyriakaki

. (2007) Aspects of identity and their influence on intentional behavior: comparing effects for three health behaviors. Personality and Individual Differences 42: 355–367.

17.

Hair

Jr Anderson

Tatham

. (1998) Multivariate Data Analysis, 5th ed. New Jersey: Prentice-Hall.

18.

Hibbert

Dickinson

Gössling

. (2013) Identity and tourism mobility: an exploration of the attitude–behavior gap. Journal of Sustainable Tourism 21(7): 999–1016.

19.

Jöreskog

Sörbom

(1993) LISREL 8: Structural Equation Modeling with the SIMPLIS Command Language. Chicago: Scientific Software International Inc.

20.

Keyes

van den Berghe

(eds) (1984) Tourism and ethnicity. Special Issue of Annals of Tourism Research 11(3): 339–501.

21.

Kline

(2005) Principles and practice of structural equation modeling. (2nd ed.). New York: The Guildford Press.

22.

Lanfant

Allcock

Bruner

(eds) (1995) International Tourism: Identity and Change. London: Sage MacCannell.

23.

MacCallum

Brown

Sugawara

(1996) Power analysis and determination of sample size for covariance structure modeling. Psychological Methods 1(2): 130–149.

24.

McCracken

(1986) Culture and Consumerism: A Theoretical Account of the Structure and Movement of the Cultural Meaning of Consumer Goods. Journal of Consumer Research 13(2): 196–213.

25.

Nario-Redmond

Biernat

Eidelmann

. (2004) The social and personal identities scale: a measure of the differential importance ascribed to social and personal self-categorizations. Self and Identity 3: 143–175.

26.

Niedenthal

Beike

(1997) Interrelated and isolated self-concepts. Personality & Social Psychology Review 1: 106–128.

27.

Noy

(2004) Performing identity: touristic narratives of self-change. Text and Performance Quarterly 24(2): 115–138.

28.

Rise

Sheeran

Hukkelberg

(2010) The role of self-identity in the theory of planned behavior: a meta-analysis. Journal of Applied Social Psychology 40(5): 1085–1105.

29.

Shenzhen-HK (2014) Shenzhen, China’s #4 top outbound travel market. Available at: http://sz.szhk.com/2014/06/04/282876766382994.html (accessed 15 March 2016).

30.

Sirgy

(1986) Using self-congruity and ideal congruity to predict purchase motivation. Journal of Business Research 13(1): 195–206.

31.

Sobel

(1986) Some new results on indirect effects and their standard errors in covariance structure. Sociological Methodology 16: 159–186.

32.

Song

Lee

Norman

. (2012) The role of responsible gambling strategy in forming behavioral intention: an application of a model of goal-directed behavior. Journal of Travel Research 51(4): 512–523.

33.

Sparks

Shepherd

(1992) Self-identity and the theory of planned behavior: assessing the role of identification with “green consumerism”. Social Psychology Quarterly 55(4): 388–399.

34.

Stets

Biga

(2003) Bringing identity theory into environmental sociology. Sociological Theory 21: 398–423.

35.

Strack

Deutsch

(2004) Reflective and impulsive determinants of social behavior. Personality and Social Psychology Review 8: 220–247.

36.

Tajfel

(1981) Social stereotypes and social groups. In: Turner

Giles

(eds) Intergroup Behavior. Oxford: Blackwell, pp. 144–167.

37.

Thurlow

Jaworski

(2006) The alchemy of the upwardly mobile: symbolic capital and the stylization of elites in frequent-flyer programmes. Discourse and Society 17(1): 99–135.

38.

Wang

(2015) Influence of place-based senses of distinctiveness, continuity, self-esteem and self-efficacy on resident attitudes toward tourism. Tourism Management 47: 241–250.

39.

Young

Kent

(1985) Using the theory of reasoned action to improve the understanding of recreation behavior. Journal of Leisure Research 17(2): 90–106.