Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Data and Information Management

4 Issues per year

Open Access
Online
ISSN
2543-9251
See all formats and pricing
More options …

A Three-way Interaction Model of Information Withholding: Investigating the Role of Information Sensitivity, Prevention Focus, and Interdependent Self-Construal

Yongqiang Sun / Dina Liu / Nan Wang
Published Online: 2017-09-29 | DOI: https://doi.org/10.1515/dim-2017-0007

Abstract

Although user information disclosure behavior in the context of social network service(SNS) has been well studied in previous literature, there is a lack of understanding about user information withholding behavior. To fill this research gap, the present study assumes that there might be a three-way interaction among information sensitivity, prevention focus, and interdependent self-construal regarding information withholding. The proposed model is empirically tested through an online survey of 479 users in the context of WeChat, one of the most popular SNSs in China. The results of hierarchical regression analysis verify the three-way interaction that prevention focus positively moderates the relationship between information sensitivity and information withholding, and interdependent self-construal strengthens the moderating effect of prevention focus. Findings in light of theoretical and practical implications as well as limitations of the study are discussed.

Keywords: social network services; information withholding; information sensitivity; regulatory focus; self-construal

1 Introduction

With the rapid development of Web 2.0 technologies, social network services (SNSs) have become increasingly popular in the past few years. SNS refers to a web-based platform that allows individuals to “construct a public or semi-public profile within a bounded system; articulate a list of other users with whom they share a connection; and view and traverse their list of connections and those made by others within the system” (Boyd & Ellison, 2007, p. 211). As a convenient way to develop and maintain interpersonal relationships without the restrictions of time and space, various SNSs such as Facebook, Twitter, and Instagram, have become an indispensable part of many individuals’ daily lives throughout the world. According to the 2015 Chinese Social Media Users’ Usage Behaviors Report issued by the CNNIC (China Internet Network Information Center) in April 2016, the coverage rate of the most popular SNS was 90.7% among Chinese netizens in 2015, and the rate is still increasing.

SNS allows users to intentionally and voluntarily disclose and share information about themselves, their opinions, and activities to others in the form of status updates, photographs, videos, and hyperlinks (Reynolds, Venkatanathan, Gonalves, & Kostakos, 2011). Information disclosure about oneself is often known as self-disclosure, and refers to the revelation of personal information to others (Derlega & Chaikin, 1977). The personal information disclosed in SNSs includes not only the static information (e.g., gender, birthday education background, hometown, e-mail, and phone numbers) required in user profile fields, but also the more dynamic information (e.g., photos, experiences, thoughts, feelings, concerns, and fears) revealed by users’ actual usage of SNSs, such as updating status, sharing, commenting, and liking (Chang & Heo, 2014; Taddei & Contena, 2013; Utz, 2015).

It is of great importance to encourage SNS users’ information disclosure because the survival and sustainability of SNSs highly depend on continuous usergenerated content (Sun, Wang, Shen, & Zhang, 2015). However, because of the increasing concerns raised by negative factors such as privacy issues, users become more reluctant to disclose their personal information in SNSs (Chang & Heo, 2014), which can be viewed as a manifestation of information withholding. To date, a burgeoning body of studies have identified factors that contribute to information disclosure in SNSs, whose different research models are based on theories of uses and gratification (Chang & Heo, 2014; Hollenbaugh & Ferris, 2014), social contract (Chang & Heo, 2014), privacy calculus (Krasnova, Spiekermann, Koroleva, & Hildebrand, 2010; Min & Kim, 2015), trust (Taddei & Contena, 2013), social capital (Trepte & Reinecke, 2013), Big Five personality (Chen, Widjaja, & Yen, 2015), or SNS flow (Kwak, Choi, & Lee, 2014). However, few studies have investigated the determinants of information withholding by SNS users.

Information withholding is found to be a common phenomenon in the research on health information (Agaku, Adisa, Ayo-Yusuf, & Connolly, 2014), scientific information (Martine R Haas & Park, 2010), and financial information (Dedman & Lennox, 2009), and is interpreted as an incomplete disclosure, information hiding, or information privacy protection behavior (Debatin, 2011; John, Barasz, & Norton, 2016; Squicciarini & Griffin, 2012). Although people usually regard information withholding as the opposite of information disclosure, we suggest that information withholding and information disclosure are two related but conceptually distinct constructs. Although the antecedents may be shared by both information withholding and information disclosure, the underlying mechanisms and effects of the antecedents are likely to be extremely different. Specifically, from an effortwithholding perspective (Kidwell & Bennett, 1993), this study considers information withholding in SNSs as the willingness of giving less than full effort on disclosing information. In other words, information disclosure may capture the extent to which users would like to disclose their personal information, while information withholding reflects the difference between the amount of information that users can provide and the amount of information that they actually provide. Because most studies have focused on why SNS users are motivated to engage in information disclosure, an understanding of why people are willing to withhold the information they are able to provide will offer some additional insights.

Information withholding can be taken as a strategy of privacy management and protection for online users (Jin, 2012; Metzger & Miriam, 2007). As a result, factors affecting information withholding online are closely associated with users’ privacy concerns, such as information sensitivity (Yang & Wang, 2009). Hence, this study identifies information sensitivity as a predictor of SNS user information withholding in terms of privacy concerns. Several studies have found conflicting results because SNS users’ privacy concerns do not always lead to their information withholding behavior. Many users continue to disclose their personal information in SNS regardless of their stated privacy concerns (Taddei & Contena, 2013; Tufekci, 2008). A possible explanation for this phenomenon is that some users view the risks of privacy issues as less important than others in SNS. On the contrary, those who view privacy risks as more important, are more willing to withhold information. The varying results of users’ view of risks can be illustrated by regulatory focus theory, which illustrates two distinct self-regulatory orientations (e.g., promotion focus and prevention focus) (E. T. Higgins, 1997). Promotion-focused individuals focus on attaining positive outcomes, while prevention-focused individuals focus on preventing negative outcomes (E. T. Higgins, 1998). In this regard, we presume that SNS users with a prevention focus tend to care more about their privacy risks because of their concerns of negative outcomes. Because they perceive an increasing level of information sensitivity, they are more likely to withhold information online. Further, we suppose that prevention-focused users may not always engage in information withholding behavior caused by information sensitivity, unless they are concerned about social interactions with others. For those who only focus on themselves, even with a dominant prevention focus, they may be more willing to disclose information to distinguish themselves from others in SNS. Therefore, we assume that the moderating effects of prevention focus vary in individuals with distinct self-identity orientations (e.g., viewing the self as independent from others, or viewing the self as associated with social context), which is conceptualized as self-construal (e.g., independent self-construal and interdependent self-construal) (Markus & Kitayama, 1991).

The aim of this study is to investigate SNS user information withholding behavior by examining the moderating effects of individual differences on the relationship between information sensitivity and information withholding. Prevention focus and interdependent self-construal are regarded as two individual factors, drawing upon the theories of regulatory focus and self-construal. It is expected that there is threeway interaction among information sensitivity, prevention focus, and interdependent self-construal when examining the user relationship with information withholding.

This study contributes to the existing literature in several ways. First, our study provides a new perspective for understanding SNS user information withholding by considering it as an intentional choice of withholding or concealing information; the underlying reasons for this behavior are different from those for information disclosure, which has been widely investigated in prior studies. Second, this study examines the moderating effect of prevention focus on the relationship between information sensitivity and information withholding. Third, this study further investigates the complex three-way interaction among information sensitivity, prevention focus, and interdependent self-construal in SNS user information withholding behavior by proposing interdependent self-construal as a possible candidate moderator that strengthens the moderating effect of prevention focus.

Section 2 presents the theoretical background and rationales for our research hypothesis regarding the interaction among information sensitivity, prevention focus, and interdependent self-construal. Sections 3 and 4 describe the data collection process and the data analysis results. Finally, Section 4 discusses the findings in light of theoretical and practical implications, as well as limitations.

2 Theoretical background and hypotheses

2.1 Information withholding

Information withholding in organizational literature is defined as the intentional process of failing to share potentially useful information with others (Martine R. Haas, 2010). A unified definition of the concept of information withholding in the online context does not currently exist. Generally, information withholding is considered as a means of managing and protecting privacy online (Jin, 2012; Metzger & Miriam, 2007). Individuals are more likely to withhold information that is perceived to be risky to disclose (Petronio, 2002). Information withholding in the context of e-commerce and e-health is usually measured by counting the frequencies of stated information categories required by service providers, including personal identification information, online experiences, personal hobbies, financial data or health records, and so on (Jin, 2012; Metzger & Miriam, 2007).

However, compared to e-commerce and e-health, disclosing or withholding information is to some degree voluntary in SNS (Chang & Heo, 2014). E-commerce and e-health users will not voluntarily reveal personal information unless they are requested to provide specific information to complete a transaction or to get medical advice, whereas SNS users are encouraged to proactively disclose and share their daily activities, inner thoughts, and emotional states in the form of status updates, pictures, and other web content (Hollenbaugh & Ferris, 2014). Therefore, existing documents always examine information withholding in terms of static information (e.g., name, phone number, e-mail address) required by service providers, whereas this study focuses on the dynamic information that users may disclose or withhold in the process of their voluntary use of SNS, like updating status, sharing, commenting, and liking.

Although information withholding often appears to be the opposite of information disclosure in that individuals either disclose or withhold information, we contend that they are not the opposites of each other, but rather two distinct constructs, the underlying mechanisms and motivations of which are quite different. For example, when a user fails to disclose certain information simply because he or she does not have the information, it is not a demonstration of information withholding, but rather a lack of information to disclose. In contrast, information withholding occurs when a user is in possession of the information to disclose, but intentionally withholds or conceals it. That is, instead of a lack of information to disclose, information withholding is the user’s intentional attempt to withhold or conceal information that he or she is able to provide. This is in line with a perspective of effort withholding, which refers to the willingness with which a person exerts less than full effort to perform a task (Kidwell & Bennett, 1993). In particular, information withholding in SNSs can be viewed as a specific form of effort withholding, which refers to the willingness of a person to exert less than full effort to disclose information to others in an SNS.

Individuals may conduct information disclosure and information withholding simultaneously. They may disclose some information that is less sensitive, and withhold other information that is more sensitive (Yang & Wang, 2009). This is an individual’s intentional choice of concealing certain information while revealing other information. Because users have the freedom to choose what information they disclose and what information they withhold in the SNS, their choice to some extent depends on the level of information sensitivity (Squicciarini & Griffin, 2012).

2.2 Information sensitivity

Information sensitivity is usually defined as “the level of privacy concern an individual feels for a type of data in a specific situation” (Weible, 1993, p. 30). Many researchers have claimed that the level of privacy concern depends on the type of information requested. Generally speaking, when users are requested to provide various types of information, they are likely to be more concerned about their financial data, medical records, and personal identification information (e.g., names, addresses, social security numbers) than demographic and lifestyle information (Phelps, Nowak, & Ferrell, 2000; Sheehan & Hoy, 2000; Ward, Bridges, & Chitty, 2005). In other words, the former types of information are considered as more sensitive information than the latter types. In addition, compared with less sensitive information, more sensitive information will have a more negative impact on users’ willingness to reveal their personal information (Malhotra, Kim, & Agarwal, 2004).

Because the levels of information sensitivity vary in different types of information, existing studies usually take various information types to indicate different levels of information sensitivity. For example, Yang and Wang (2009) created two scenarios in the online shopping context by considering demographic information as having a low level of sensitivity, and demographic combined with financial information as having a high-level of sensitivity.

These prior studies, which used information types as indicators of information sensitivity, ignored the fact that the perception of sensitivity varies with different individuals and situations (Sheehan & Hoy, 2000). For instance, demographic information with relatively lower levels of sensitivity, such as age, may be perceived as more sensitive for some users, especially for single female users. Referring to a stimulus-organism-response (S-O-R) framework, individuals’ internal cognitive processes play an important role in their response to external stimulus (Mehrabian & Russell, 1974). Similarly, users may react differently to the same information based on their distinct personal perception of it. Therefore, instead of indiscriminately dividing the levels of information sensitivity based on general information types, it is more appropriate to measure information sensitivity by taking users’ perception of sensitivity into account. Some scholars define information sensitivity as the perceived intimacy level of information (Lwin, Wirtz, & Williams, 2007). More intimate information is perceived as riskier to disclose because it may lead to potential losses, including psychological (e.g., loss of self-esteem), physical (e.g., loss of health), and material (e.g., loss of property and assets) aspects (Moon, 2000). Therefore, this study defines information sensitivity as the extent to which the information is perceived as sensitive due to the potential losses incurred by its disclosure.

Many previous studies have demonstrated that users’ willingness to disclose or withhold information depends on the sensitivity of the information (Malhotra et al., 2004; Phelps et al., 2000; Yang & Wang, 2009). For example, Metzger and Miriam (2007) proposed that e-commerce consumers would be more likely to withhold more sensitive information as a way to protect their privacy. Bansal, Zahedi, and Gefen (2010) regarded information sensitivity as an important predictor of users’ privacy concerns about health information disclosure in web-based health service. In the context of SNS, users’ choices on information disclosure or withholding also rely on their perceived sensitivity of the information (Squicciarini & Griffin, 2012). Thus, the more sensitive the information is, the greater the extent of SNS users’ intention to withhold the information. Therefore, we hypothesize that:

H1: Information sensitivity is positively associated with information withholding.

The impact of information sensitivity on information withholding may vary across different individuals. For example, individuals who view the self as interdependent of others and focus on preventing negative outcomes like privacy risks may be more likely to be motivated by information sensitivity to withhold information in SNSs. Drawing upon the theories of regulatory focus and self-construal, our study further identifies prevention focus and interdependent self-construal as two individual moderating factors, and explores the three-way interaction among information sensitivity, prevention focus, and interdependent self-construal.

2.3 Regulatory focus theory

Regulatory focus theory elaborates the operation of the basic hedonic principle that people are motivated to approach pleasure and avoid pain (E. T. Higgins, 1997). Specifically, regulatory focus refers to the different ways in which the hedonic principle operates to regulate pleasure and pain, including promotion focus and prevention focus (E. T. Higgins, 1998). Promotion focus is driven by the need for nurturance and development, whereas prevention focus is driven by the need for insurance and security (Johnson, Chang, & Yang, 2010). Individuals with promotion focus strive to achieve the “ideal self” by fulfilling expectations, ambitions, and aspirations. They are concerned with the presence of positive outcomes such as gains, successes, and advancements. Individuals with prevention focus strive to become the “ought self” by realizing obligations, duties, and security. They are concerned with the presence of negative outcomes such as failures, losses, and mistakes (E. T. Higgins, 1998).

Regulatory focus theory posits that individuals demonstrate disparate strategic tendencies for achieving their desired goals (E. T. Higgins, 1997, 1998). Promotion-focused individuals tend to focus on strategies that match desired goals of promoting positive outcomes, whereas prevention-focused individuals tend to focus on strategies that match desired goals of preventing negative outcomes (Higgins et al., 1994). E. T. Higgins (2000) proposed regulatory fit to describe the fit between individuals’ strategies for pursuing desirable goals and their dominant regulatory focus (e.g., promotion or prevention). When individuals experience a regulatory fit, they will show enhanced value perceptions of desired goals, perceiving an attractive goal as more attractive, and an unpleasant goal as more unpleasant (E. T. Higgins, 2006).

In the SNS context, information withholding can be viewed as a kind of prevention strategy, because its purpose is to avoid negative outcomes such as privacy concerns, and is therefore more likely to be adopted by users with a prevention focus. Because there is a fit between users’ regulatory focus and the strategies by which the desired goal is achieved, users with a prevention focus are inclined to perceive the negative outcomes of information disclosure as more unpleasant, and are more motivated by the costs of withholding information. Hence, we propose that prevention focus will moderate the relationships between information sensitivity and users’ information withholding behavior. Information sensitivity is defined in terms of the potential losses caused by information disclosure in this study, which is a kind of negative outcome from the perspective of prevention focus. Due to the regulatory fit, prevention-focused users tend to perceive information sensitivity as more unpleasant, and feel a stronger urge to avoid the potential losses by means of information withholding. As a result, as prevention focus increases, information sensitivity becomes more motivating and has a stronger influence on SNS users’ information withholding. Therefore, we hypothesize that: H2: Prevention focus strengthens the relationship between information sensitivity and information withholding.

2.4 Self-construal theory

Self-construal is conceptualized as individuals’ divergent views about the self, which depicts the extent to which they think of themselves as distinct from others, or as connected with others (Markus & Kitayama, 1991; Singelis, 1994). Self-construal consists of an independent self and an interdependent self, which are exemplified in Western (individualist) and Asian (collective) cultures, respectively (Kim et al., 1996). Independent self-construal describes the individual as a bounded, stable, autonomous entity, and views the self as separate from social context. Interdependent self-construal, on the other hand, focuses on individuals in the context of social relationships, associating the self within the external social environment (Markus & Kitayama, 1991). Although independent and interdependent self-construal originate from cross-culture research on individualism and collectivism, these two ways of viewing the self co-exist within every individual in any culture, and vary in their relative strengths in determining the overall self-construal of an individual (Singelis, 1994). Individuals with a dominant independent self-construal seek to distinguish themselves positively from others, with a focus on the positive aspects of the self and potential gains in situations. In contrast, individuals with a dominant interdependent self-construal strive to get along harmoniously with others in the social context, and focus on the negative aspects of the self and potential losses in situations. It is suggested that individuals with high independent self-construal are consistent with a promotion focus, viewing a gain-framed situation as more important, whereas those with high interdependent self-construal are consistent with a prevention focus, viewing a loss-framed situation as more important (Aaker & Lee, 2001; Lee, Aaker, & Gardner, 2000).

In this regard, we propose that the interaction of prevention focus and information sensitivity depends on the level of interdependent self-construal. A high level of interdependent self-construal will strengthen the interaction of prevention focus and information sensitivity, (which is hypothesized in H2), and prevention focus positively moderates the relationship between information sensitivity and information withholding. Individuals with high interdependent self-construal tend to care about social interactions with others, and focus on the negative aspects of the situation, which is compatible with a prevention focus. They view the loss-framed situation (for example, information sensitivity in this study) as more important. As a result, higher interdependent self-construal results in a stronger interacting effect between prevention focus and information sensitivity, subsequently leading to a stronger willingness to withhold information in the SNS environment. Considering the interplay of information sensitivity, prevention focus, and interdependent self-construal, we assume that under conditions of higher interdependent self-construal, the interaction between prevention and information sensitivity will be more significant. In other words, high interdependent self-construal will make SNS users with higher prevention focus more sensitive to negative outcomes like information sensitivity, and as a result, information sensitivity becomes more significantly motivates users to withhold information. Therefore, we hypothesize that:

H3: Interdependent self-construal strengthens the moderating effect of prevention focus such that when interdependent self-construal is high, the moderating effect of prevention focus is stronger.

Ultimately, the proposed three-way interaction model of information withholding is shown in Fig. 1.

A three-way interaction model of information withholding.
Figure 1

A three-way interaction model of information withholding.

3 Methods

3.1 Data collection

WeChat, a mobile SNS developed by Tencent in China, first released in January 2011, was used in this study to examine users’ information withholding. It is well-known that Facebook and Twitter are popular in America and Europe, whereas WeChat has become one of the most universal SNS applications in China in recent years. According to Tencent’s 2016 first quarter operating results, WeChat has 762 million monthly active accounts around the world, a 39% increase over same period in the previous year. Apart from conventional services (e.g., text messaging, hold-to-talk voice messaging, broadcast [one-to-many] messaging), WeChat provides a special function called “Moments” that allows users to post image and text, share music, videos, articles, and links. Only the friends in the user’s contacts can view the information that WeChat users disclose in Moments.

We conducted an online survey to collect data using sojump.com, which is a professional questionnaire survey agency that has the largest number of online panels in China. The questionnaire was divided into two parts: the demographics and SNS usage information of respondents, and the scales for the four variables in the research model. An invitation to the online questionnaire, including the URL at sojump.com, was posted on a number of WeChat online social groups. The participants were assessed with screening questions to ensure that they were current active users of WeChat. In addition, IP addresses were recorded and checked to exclude duplicate respondents in this survey. As a result, 479 valid responses were obtained and used in the data analysis. The demographics of the respondents are summarized in Table 1.

Table 1

Demographics.

Table 2

Measurements for the study variables in the research model.

Non-response bias was tested by comparing the demographics and means of the factors in the first 1/3 and last 1/3 of the sample using the Chi-square test and t-test, and the results suggested that no significant differences were found, indicating that non-response bias was not a critical issue for this study.

The male participants slightly outnumbered the female participants. The age of most participants ranged from 18 to 40, and accounted for about 90% of the total participants. More than 70% of the participants were of graduate education level or higher. Nearly 80% of participants have used WeChat for more than one year and 90% use WeChat less than 4 hours per day.

3.2 Measurements

Four variables were measured using multi-item perceptual scales in this study. Most items were adapted from previous studies according to the context of this research. Specifically, the scales for information sensitivity were derived from Lwin, Wirtz, and Stanaland (2016). Four items for prevention focus were tailored from Lockwood, Jordan, and Kunda (2002). The scales for interdependent self-construal were adapted from Wang, Ma, and Li (2015). Because there is no applicable existing scale for information withholding in an SNS to assess users’ intentional attempts to withhold the information they are able to provide, two items for information withholding were modified from the scale of knowledge withholding in Tsay, Lin, Yoon, and Huang (2014), which was also developed from the perspective of effort withholding. All the items were measured on a seven-point Likert scale ranging from 1=strongly disagree to 7=strongly agree.

3.3 Statistical analysis

The measurement model was examined by SmartPLS to evaluate the reliability, discriminant validity, and convergent validity for the measures underlying each variable. The structural model was tested using SPSS Statistics for Windows v20.0. Specifically, the interactions among information sensitivity, prevention focus, and interdependent self-construal were analyzed by hierarchical regression analysis with the stepwise method in SPSS.

4 Results

4.1 Measurement model

Table 3 displays the means, standard deviations, and zero-order correlations of all the study variables. Before conducting the hierarchical regression analysis, we evaluated the measurement model for the variables using structural equation modeling (SEM). The reliability of the measurements was examined using average variance extracted (AVE) and composite reliability (CR). The critical values for AVE and CR are 0.5 and 0.7, respectively (Fornell & Larcker, 1981). Table 4 shows that the minimum values of AVE and CR were 0.563 and 0.789 respectively. Each value was higher than the recommended value, suggesting that all variables were reliable.

Table 3

Descriptive statistics and correlations among the study variables.

Table 4

AVE, CR and factor loadings among the study variables.

Discriminant validity can be tested using a correlation matrix and square roots of AVE for the variables. The square root of the AVE for each variable should be higher than the correlation of the specific variable with all the other variables in the model (Fornell & Larcker, 1981). Table 3 shows that the square roots of all AVEs were greater than the correlations with other variables, suggesting sufficient discriminant validity.

Convergent validity was assessed by seeing whether the item loadings on the respective constructs were high enough (Gefen & Straub, 2005). As shown in Table 4, factor loadings for each observation variable exceed the cutoff value of 0.50, suggesting that there is some evidence of convergent validity.

Because all the constructs were measured with the selfreported items at the same time from the same respondents, common method bias may be an issue. According to Podsakoff et al. (2003), this issue was evaluated through Harman’s single factor test. The results showed that four primary components were derived with eigenvalues greater than 1 and the first primary component explained 29.64% of the total variance, suggesting that common method bias was not a critical concern for this study.

4.2 Structural model

We then conducted hierarchical regression analysis to test the hypotheses, centering interdependent variables around their grand mean to facilitate the interpretation of the main effects in models containing interaction terms (Aiken, West, & Reno., 1991). The predictors were entered into regression in four models: (a) gender, age, education, usage experience, and usage frequency as control variables; (b) information sensitivity, prevention focus, and interdependent self-construal; (c) the two-way interaction; (d) the three-way interaction. Information withholding was regarded as the dependent variable.

The results of the regression analyses are indicated in Table 5. In Model 1, not all of the control variables significantly predicted information withholding. In Model 2, information sensitivity (β=0.116, p<0.05) and prevention focus (β=.306, p<.01) were significant, accounting for an additional 13.3% of the variance (ΔF=24.699, p<0.01) in information withholding. We can see that information sensitivity was significantly related to information withholding (β=0.116, p<0.05), which provided support for H1. In Model 3, the two-way interaction of information sensitivity and prevention focus was found to be significant, accounting for an additional 2.6% of the variance (ΔF=4.997, p<0.01). This provided support for H2. In Model 4, the three-way interaction of information sensitivity, prevention focus, and interdependent self-construal was significant, and explained an additional 1% of the variance (ΔF=5.617, p<0.05).

Table 5

Hierarchical Linear Regression.

To understand the three-way interaction more intuitively, Figure 2 delineates the interaction effect among information sensitivity, prevention focus, and interdependent self-construal. For individuals with low interdependent self-construal, information sensitivity displayed a non-significant relationship with information withholding under conditions of both high and low prevention focus. On the other hand, for individuals with high interdependent self-construal, information sensitivity was shown to have a significant effect on information withholding under conditions of high prevention focus, whereas the effect under conditions of low prevention focus was not significant. Therefore, when interdependent self-construal is high, the effect of information sensitivity on information withholding is stronger with higher prevention focus, suggesting that interdependent self-construal strengthened the moderating effect of prevention focus on the relationship between information sensitivity and information withholding. This provided support for H3.

Three-way interaction effect. Note: INFS = Information Sensitivity, PRVF = Prevention Focus.
Figure 2

Three-way interaction effect.

Note: INFS = Information Sensitivity, PRVF = Prevention Focus.

5 Discussion

5.1 Discussion of findings

The present study provided an integrative understanding of information withholding behavior of SNS users. We began with the assumption that information sensitivity is an important predictor of information withholding, which is in line with previous findings that as the level of information sensitivity increases, users are less likely to reveal information in terms of their privacy concerns (Yang & Wang, 2009). However, it has been found that there is a discrepancy between users’ stated privacy concerns and their actual information disclosure behavior (Taddicken, 2013). Considering that the effect of factors related to privacy concerns (e.g., information sensitivity) on information withholding may vary in different individuals, we took additional aspects of the self into account (e.g., regulatory focus, self-construal). Drawing upon theories of regulatory focus and self-construal, this study considered prevention focus and interdependent self-construal as two possible moderators of the relationship between information sensitivity and information withholding. Specifically, prevention focus was chosen because prevention-focused individuals tend to focus on negative outcomes, such as information sensitivity, in this study. Interdependent self-construal was considered due to the social attribute of SNS, which is consistent with the goals of individuals with interdependent self-construal, emphasizing interpersonal relationships in social context. As a result, the present study supposed that there is a three-way interaction among information sensitivity, prevention focus, and interdependent self-construal regarding their relationship with information withholding.

The results of hierarchical regression analysis offered pronounced support for the proposed threeway interaction that interdependent self-construal strengthens the moderating effect of prevention focus on the relationship between information sensitivity and information withholding. That is, the engagement of high interdependent self-construal SNS users in information withholding behavior will be positively motivated by information sensitivity under conditions of higher prevention focus. This agrees with our expectations. Due to the goal compatibility between self-construal and regulatory focus, users with high interdependent self-construal are inclined to focus on harmonious social relationships and avoid negative outcomes, which is more compatible with prevention-focused goals (Lee et al., 2000). Because these users perceive prevention-focused goals as more important, interdependent self-construal will enhance the effect of prevention focus. Subsequently, prevention-focused users tend to perceive negative outcomes (e.g., information sensitivity in this study) as more unpleasant because of a regulatory fit (E. T. Higgins, 2006). As a result, when prevention focus is high, information sensitivity will more significantly stimulate SNS users to withhold information.

5.2 Theoretical implications

This study has several theoretical implications. First, this study extends our understanding of information withholding in the SNS context. Although the underlying mechanisms and antecedents of SNS users’ information disclosure behavior has been extensively probed in the existing literature, little attention has been devoted to the research on SNS users’ withholding behavior. Noting that information withholding and a lack of information disclosure may appear quite similar but have very different underlying mechanisms and motivations, we regard information withholding as an intentional choice of giving less than full effort on disclosing information in the SNS. Therefore, aiming at exploring the intentional attempts of SNS users to withhold information, this study offers a disparate perspective to understand SNS users’ information withholding. In addition, unlike most prior studies, which examined the information in SNS in terms of the static information in profile fields requested by SNS providers, we focus on the information intentionally and voluntarily revealed or concealed by users’ actual usage of SNS, such as updating status, sharing, commenting, and liking.

Second, this study demonstrates that the effect of information sensitivity on information withholding is not invariant across individuals, but is affected by users’ prevention focus. Because the level of information sensitivity has a positive influence on SNS users’ information withholding, we find that prevention focus moderates the relationship between information sensitivity and information withholding. Prevention-focused users who are more concerned about negative outcomes, tend to perceive information sensitivity as more unpleasant due to a regulatory fit, so information sensitivity will have a stronger influence on information withholding of SNS users.

Third, this study proposes a three-way interaction among information sensitivity, prevention focus, and interdependent self-construal to investigate SNS users’ information withholding behavior. Because prevention focus moderates the relationship between information sensitivity and information withholding, we also found that the moderating effect of prevention focus depends on the level of users’ interdependent self-construal. For SNS users with high interdependent self-construal, prevention focus has a stronger moderating effect on the relationship between information sensitivity and information withholding, leading to a more pronounced influence of information sensitivity on information withholding. Results of the proposed model can be taken as a foundation for future research.

5.3 Practical implications

This study may also provide guidance to practitioners who are interested in promoting user information disclosure and preventing withholding behavior in the SNS environment. Because SNS users encounter trade-offs between information disclosure and withholding, SNS practitioners need to take effective measures to encourage users to disclose more and withhold less. It is shown in this study that a high level of information sensitivity leads SNS users to withhold information in terms of their privacy concerns, especially those with dominant interdependent self-construal and prevention focus. Therefore, on the one hand, SNS practitioners should delay or avoid asking users for highly sensitive information, unless such information is necessary to run a specific task, for example, providing location information to search for people nearby. On the other hand, to ensure security of the information that users have disclosed in SNS, SNS practitioners need to show users that they are committed to the confidentiality of users’ personal information and provide users with strong written privacy policies.

5.4 Limitations and directions for further research

Despite the theoretical and practical implications, there are some limitations in this paper that can be further improved in future studies. First, to investigate a threeway interaction model for information withholding, we focused on the role of information sensitivity, prevention focus, and interdependent self-construal in shaping SNS users’ information withholding behavior. However, to gain a more comprehensive understanding of this issue, there may be other important factors that need to be considered in future research.

Second, the three-way interaction is only empirically examined in the context of WeChat, which is a communication-oriented service in China. There are various SNSs that are relationship-oriented services (e.g., Facebook) or information-sharing services (e.g., Twitter), and the proposed model should be validated in different types of SNSs. Research results of other kinds of SNSs will be useful in improving and verifying the generalizability of the findings in this study.

6 Conclusion

Although the factors influencing users’ information disclosure behavior have been well examined in prior studies, the reasons that users withhold their personal information is not well understood. This study proposes and empirically tests a research model of information withholding by considering the three-way interaction effects of information sensitivity, prevention focus, and interdependent self-construal. The findings show that prevention focus moderates the relationship between information sensitivity and information withholding, and interdependent self-construal further moderates the moderating effect of prevention focus. This study enriches the social media literature and lays the foundation for research on information withholding.

Acknowledgment

The work described in this paper was partially supported by grants from the Humanities and Social Sciences Foundation of the Ministry of Education, China (Project No. 17YJC630157, 16YJC870011) and the Research Fund for Academic Team of Young Scholars at Wuhan University (Project No. Whu2016013).

References

  • [1]

    Aaker, J. L., & Lee, A. Y. (2001). “I” seek pleasures and “we” avoid pains: The role of self-regulatory goals in information processing and persuasion. Journal of Consumer Research, 28(1), 33-49. CrossrefGoogle Scholar

  • [2]

    Agaku, I. T., Adisa, A. O., Ayo-Yusuf, O. A., & Connolly, G. N. (2014). Concern about security and privacy, and perceived control over collection and use of health information are related to withholding of health information from healthcare providers. Journal of the American Medical Informatics Association, 21, 374-378. CrossrefGoogle Scholar

  • [3]

    Aiken, L. S., West, S. G., & Reno., R. R. (1991). Multiple regression: Testing and interpreting interactions: Sage. Google Scholar

  • [4]

    Bansal, G., Zahedi, F. M., & Gefen, D. (2010). The impact of personal dispositions on information sensitivity, privacy concern and trust in disclosing health information online. Decision Support Systems, 49(2), 138-150. CrossrefGoogle Scholar

  • [5]

    Boyd, D. M., & Ellison, N. B. (2007). Social network sites: Definition, history, and scholarship. Journal of Computer-Mediated Communication, 13(1), 210-230. CrossrefGoogle Scholar

  • [6]

    Chang, C. W., & Heo, J. (2014). Visiting theories that predict college students’ self-disclosure on Facebook. Computers in Human Behavior, 30, 79-86. CrossrefGoogle Scholar

  • [7]

    Chen, J. V., Widjaja, A. E., & Yen, D. C. (2015). Need for affiliation, need for popularity, self-esteem, and the moderating effect of Big Five personality traits affecting individuals’ self-disclosure on Facebook. International Journal of Human-Computer Interaction, 31(11), 815-831. CrossrefGoogle Scholar

  • [8]

    Debatin, B. (2011). Ethics, privacy, and self-restraint in social networking Privacy online (pp. 47-60): Springer. Google Scholar

  • [9]

    Dedman, E., & Lennox, C. (2009). Perceived competition, profitability and the withholding of information about sales and the cost of sales. Journal of Accounting and Economics, 48(2), 210-230. CrossrefGoogle Scholar

  • [10]

    Derlega, V. J., & Chaikin, A. L. (1977). Privacy and selfdisclosure in social relationships. Journal of Social Issues, 33(33), 102-115. CrossrefGoogle Scholar

  • [11]

    Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39-50. CrossrefGoogle Scholar

  • [12]

    Gefen, D., & Straub, D. (2005). A practical guide to factorial validity using PLS-graph: Tutorial and annotated example. Communications of the Association for Information Systems, 16(1), 91-109. Google Scholar

  • [13]

    Haas, M. R. (2010). To share or not to share? Professional norms, reference groups, and information withholding among life scientists. Organization Science, 21(4), 873-891. CrossrefGoogle Scholar

  • [14]

    Haas, M. R., & Park, S. (2010). To share or not to share? Professional norms, reference groups, and information withholding among life scientists. Organization Science, 21(4), 873-891. CrossrefGoogle Scholar

  • [15]

    Higgins, Tory, E., Roney, Christopher, J. R., Crowe, Ellen, … Charles. (1994). Ideal versus ought predilections for approach and avoidance: Distinct self-regulatory systems. Journal of Personality & Social Psychology, 66(2), 276-286. CrossrefGoogle Scholar

  • [16]

    Higgins, E. T. (1997). Beyond pleasure and pain. American Psychologist, 52(12), 1280-1300. CrossrefGoogle Scholar

  • [17]

    Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational principle. Advances in Experimental Social Psychology, 30(2), 1-46. CrossrefGoogle Scholar

  • [18]

    Higgins, E. T. (2000). Making a good decision: Value from fit. American Psychologist, 55(11), 1217-1230. CrossrefGoogle Scholar

  • [19]

    Higgins, E. T. (2006). Value from hedonic experience and engagement. Psychological Review, 113(3), 439-460. CrossrefPubMedGoogle Scholar

  • [20]

    Hollenbaugh, E. E., & Ferris, A. L. (2014). Facebook selfdisclosure: Examining the role of traits, social cohesion, and motives. Computers in Human Behavior, 30(1), 50-58. CrossrefGoogle Scholar

  • [21]

    Jin, S. A. A. (2012). “To disclose or not to disclose, that is the question”: A structural equation modeling approach to communication privacy management in E-health. Computers in Human Behavior, 28(1), 69-77. CrossrefGoogle Scholar

  • [22]

    John, L. K., Barasz, K., & Norton, M. I. (2016). Hiding personal information reveals the worst. Proceedings of the National Academy of Sciences, 113(4), 954-959. CrossrefGoogle Scholar

  • [23]

    Johnson, R. E., Chang, C. H., & Yang, L. Q. (2010). Commitment and motivation at work: The relevance of employee identity and regulatory focus. Academy of Management Review, 35(2), 226-245. CrossrefGoogle Scholar

  • [24]

    Kidwell, R. E., & Bennett, N. (1993). Employee propensity to withhold effort: A conceptual model to intersect three avenues of research. Academy of Management Review, 18(3), 429-456. Google Scholar

  • [25]

    Kim, M. S., Hunter, J. E., Miyahara, A., Horvath, A. M., Bresnahan, M., & Yoon, H. J. (1996). Individual-vs. culture-level dimensions of individualism and collectivism: Effects on preferred conversational styles. Communication Monographs, 63(1), 29-49. CrossrefGoogle Scholar

  • [26]

    Krasnova, H., Spiekermann, S., Koroleva, K., & Hildebrand, T. (2010). Online social networks: Why we disclose. Journal of Information Technology, 25(2), 109-125. CrossrefGoogle Scholar

  • [27]

    Kwak, K. T., Choi, S. K., & Lee, B. G. (2014). SNS flow, SNS self-disclosure and post hoc interpersonal relations change: Focused on Korean Facebook user. Computers in Human Behavior, 31(1), 294-304. CrossrefGoogle Scholar

  • [28]

    Lee, A. Y., Aaker, J. L., & Gardner, W. L. (2000). The pleasures and pains of distinct self-construals: The role of interdependence in regulatory focus. Journal of Personality & Social Psychology, 78(6), 1122-1134. CrossrefGoogle Scholar

  • [29]

    Lockwood, P., Jordan, C. H., & Kunda, Z. (2002). Motivation by positive or negative role models: Regulatory focus determines who will best inspire us. Journal of Personality & Social Psychology, 83(4), 854-864. CrossrefGoogle Scholar

  • [30]

    Lwin, M., Wirtz, J., & Stanaland, A. J. S. (2016). The privacy dyad: Antecedents of promotion- and prevention-focused online privacy behaviors and the mediating role of trust and privacy concern. Internet Research, 26(4), 919-941. CrossrefGoogle Scholar

  • [31]

    Lwin, M., Wirtz, J., & Williams, J. D. (2007). Consumer online privacy concerns and responses: A power–responsibility equilibrium perspective. Journal of the Academy of Marketing Science, 35(4), 572-585. CrossrefGoogle Scholar

  • [32]

    Malhotra, N. K., Kim, S. S., & Agarwal, J. (2004). Internet users’ information privacy concerns (IUIPC): The construct, the scale, and a causal model. Information Systems Research, 15(4), 336-355. CrossrefGoogle Scholar

  • [33]

    Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224-253. CrossrefGoogle Scholar

  • [34]

    Mehrabian, A., & Russell, J. A. (1974). An approach to environmental psychology: MIT. Google Scholar

  • [35]

    Metzger, & Miriam, J. (2007). Communication privacy management in electronic commerce. Journal of Computer-Mediated Communication, 12(2), 335-361. CrossrefGoogle Scholar

  • [36]

    Min, J., & Kim, B. (2015). How are people enticed to disclose personal information despite privacy concerns in social network sites? The calculus between benefit and cost. Journal of the Association for Information Science and Technology, 66(4), 839–857. CrossrefGoogle Scholar

  • [37]

    Moon, Y. (2000). Intimate exchanges: Using computers to elicit self-disclosure from consumers. Journal of Consumer Research, 26(4), 323-339. CrossrefGoogle Scholar

  • [38]

    Petronio, S. (2002). Boundaries of privacy: Dialectics of disclosure. New England Journal of Medicine(15), 1505-1506. Google Scholar

  • [39]

    Phelps, J., Nowak, G., & Ferrell, E. (2000). Privacy concerns and consumer willingness to provide personal information. Journal of Public Policy & Marketing, 19(1), 27-41. CrossrefGoogle Scholar

  • [40]

    Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N.P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879-903. CrossrefGoogle Scholar

  • [41]

    Reynolds, B., Venkatanathan, J., Gonalves, J., & Kostakos, V. (2011). Sharing ephemeral information in online social networks: Privacy perceptions and behaviours. Paper presented at the IFIP International Federation for Information Processing 2011, Berlin Heidelberg. Google Scholar

  • [42]

    Sheehan, K. B., & Hoy, M. G. (2000). Dimensions of privacy concern among online consumers. Journal of Public Policy & Marketing, 19(1), 62-73. CrossrefGoogle Scholar

  • [43]

    Singelis, T. M. (1994). The measurement of independent and interdependent self-construals. Personality & Social Psychology Bulletin, 20(5), 580-591. CrossrefGoogle Scholar

  • [44]

    Squicciarini, A. C., & Griffin, C. (2012). An informed model of personal information release in social networking sites. Paper presented at the Privacy, Security, Risk and Trust. Google Scholar

  • [45]

    Sun, Y., Wang, N., Shen, X. L., & Zhang, J. X. (2015). Location information disclosure in location-based social network services: Privacy calculus, benefit structure, and gender differences. Computers in Human Behavior, 52, 278-292. CrossrefGoogle Scholar

  • [46]

    Taddei, S., & Contena, B. (2013). Privacy, trust and control: Which relationships with online self-disclosure? Computers in Human Behavior, 29(3), 821-826. CrossrefGoogle Scholar

  • [47]

    Taddicken, M. (2013). The ‘privacy paradox’ in the social web: The impact of privacy concerns, individual characteristics, and the perceived social relevance on different forms of selfdisclosure. Journal of Computer-Mediated Communication, 19(2), 248–273. Google Scholar

  • [48]

    Trepte, S., & Reinecke, L. (2013). The reciprocal effects of social network site use and the disposition for self-disclosure: A longitudinal study. Computers in Human Behavior, 29(3), 1102-1112. CrossrefGoogle Scholar

  • [49]

    Tsay, H. H., Lin, T. C., Yoon, J., & Huang, C. C. (2014). Knowledge withholding intentions in teams: The roles of normative conformity, affective bonding, rational choice and social cognition. Decision Support Systems, 67, 53-65. CrossrefGoogle Scholar

  • [50]

    Tufekci, Z. (2008). Can you see me now? Audience and disclosure regulation in online social network sites. Bulletin of Science, Technology & Society, 28(1), 20-36. CrossrefGoogle Scholar

  • [51]

    Utz, S. (2015). The function of self-disclosure on social network sites: Not only intimate, but also positive and entertaining self-disclosures increase the feeling of connection. Computers in Human Behavior, 45, 1-10. CrossrefGoogle Scholar

  • [52]

    Wang, Y., Ma, S., & Li, D. (2015). Customer participation in virtual brand communities: The self-construal perspective. Information & Management, 52(5), 577-587. CrossrefGoogle Scholar

  • [53]

    Ward, S., Bridges, K., & Chitty, B. (2005). Do incentives matter? An examination of on-line privacy concerns and willingness to provide personal and financial information. Journal of Marketing Communications, 11(1), 21-41. CrossrefGoogle Scholar

  • [54]

    Weible, R. J. (1993). Privacy and data: An empirical study of the influence of types of data and situational context upon privacy perceptions. Mississippi State University. Google Scholar

  • [55]

    Yang, S., & Wang, K. (2009). The influence of information sensitivity, compensation on privacy concern and behaviour intention. The DATA BASE for Advances in Information Systems, 40(1), 38-51. CrossrefGoogle Scholar

Footnotes

    About the article

    Received: 2017-05-02

    Accepted: 2017-06-03

    Published Online: 2017-09-29


    Citation Information: Data and Information Management, Volume 1, Issue 1, Pages 61–73, ISSN (Online) 2543-9251, DOI: https://doi.org/10.1515/dim-2017-0007.

    Export Citation

    © 2017 Yongqiang Sun, Dina Liu, Nan Wang. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. BY-NC-ND 3.0

    Comments (0)

    Please log in or register to comment.
    Log in