Sharing Boundary Spanning Information: The Role of Source Characteristics
and Perceived Information Quality
Cheng Chen (S3444171) University of Groningen
Author Note
Cheng Chen, Department of HRM and OB, University of Groningen.
I would like to thank prof. dr. Bernard Nijstad and dr. Yingjie Yuan for their constant support and supervision on my master thesis.
Abstract
Research has suggested that boundary spanning information transfer is not always successful, but left unclear why that is the case. This study addresses this problem by investigating
boundary spanning information sharing which refers to boundary spanners sharing a piece of
external information with the team. Integrating insights from the persuasion, communication, advice taking, and information systems literatures, I propose a mechanism that explains
boundary spanning information sharing via the individual perception of information quality
in terms of credibility, uniqueness (newness), uniqueness (rareness), and relevance. Based on the elaboration likelihood model of persuasion, I further propose that source characteristics (i.e., source expertise and source duplication) influence perceived information quality, which in turn determines boundary spanning information sharing. In addition, the study explores the influence of information valence in boundary spanning information sharing. I adopt a 2 (Source Duplication: Yes vs. No) × 2 (Source Expertise: Low vs. High) × 2 (Information Valence: Positive vs. Negative) scenario study (N = 395, Mage = 41.38 years) to test these
propositions. Results show that source expertise positively influences boundary spanning
information sharing by increasing credibility and relevance. Source duplication has indirect
effects on boundary spanning information sharing when source expertise is high and when external information is negative. Overall, this study clarifies the mechanism underlying the sharing stage of boundary spanning and highlights the role of perceived information quality and source characteristics in boundary spanning.
Keywords: boundary spanning information sharing, source characteristics, perceived
Introduction
To deal with the challenges presented by the changing environment and meet increasingly complex organizational goals (Marrone, 2010), organizational teams have to effectively manage not only internal functioning but also external functioning (Ancona & Caldwell, 1992a, 1992b; Reagans, Zuckerman, & McEvily, 2004). Especially, to obtain valuable knowledge and information, organizational teams have to interact with parties external to the team (Hansen, 1999; Joshi, Pandey, & Han, 2009). Consequently, it has become more and more common for team members to access external information across team boundaries, which is captured in the notion of boundary spanning (Ancona & Caldwell, 1992a). Previous research has demonstrated that boundary spanning improves team
performance, for both routine work and innovation (Ancona & Caldwell, 1992a; Hargadon, 1998).
Research on boundary spanning explores mostly what drives it (e.g., Joshi et al., 2009; Schepers, De Vries, Van Weele, & Langerak, 2019) while ignoring that the external information acquired by boundary spanners may not be shared with the focal team ultimately (Tushman & Scanlan, 1981). External information being shared with the team is a
prerequisite for external information to exert influence on team performance. Hence, the lack of understanding of why and how boundary spanners share external information with the team will hinder the understanding of boundary spanning effectiveness.
2007), communication (Mojzisch, Kerschreiter, Faulmüller, Vogelgesang, & Schulz-Hardt, 2014), and information systems (Nicolaou & McKnight, 2006) have shown that perceived information quality (PIQ) is a key determinant of the exchange, the sharing, and the
acceptance of information. In line with Marrone’s (2010) call for an integration of boundary spanning research and its related fields of research, and inspired by the insights from these fields of research, I propose that boundary spanners share external information with the team only if they recognize the quality of a piece of external information.
Although PIQ is likely to influence the sharing of boundary spanning information, the notion of PIQ has not been fully captured in boundary spanning research. Yet existing
research often equates the quality of boundary spanning information to its uniqueness in the sense that it is new to the team (Reagans & Zuckerman, 2001). This results in a limited understanding of PIQ in boundary spanning in terms of the role it plays and the aspects it covers. The value of unique information lies in not only its newness but also rareness in the sense that such information is not common knowledge and can bring competitive advantage (Barney, 1991). Moreover, according to relevant fields of research, PIQ is also captured by
credibility and relevance (Maltz, 2000; Mojzisch et al., 2014; Nicolaou & McKnight, 2006).
Thus, PIQ in boundary spanning should have four dimensions: credibility, uniqueness
(newness), uniqueness (rareness), and relevance.
the “negativity bias” (Rozin & Royzman, 2001) suggests that negative information is often surprising and unexpected (Fiske, 1980; Taylor, 1991) and thus receives more attention than positive information from humans in many domains including perception. This effect of negative information on humans’ perception is independent of information sources (Siegrist & Cvetkovich, 2001; Zhu, Xie, & Gan, 2011). Hence, I propose that source characteristics and information valence both influence how boundary spanners perceive the quality of a piece of information.
In the current study, I aim to investigate what drives boundary spanners to share boundary spanning information they have acquired with their teams, which I refer to as
boundary spanning information sharing. I propose that boundary spanners use information
source characteristics and information valence as heuristic cues to judge the quality of
external information which in turn determines boundary spanning information sharing. In the following session, I first review the previous studies on boundary spanning and boundary
spanning information sharing. I then analyze the obstacles to boundary spanning information sharing. Subsequently, I go into the role of PIQ and how it is influenced by source
characteristics and information valence. I investigate the effect of information valence in an exploratory way. My hypotheses are tested using an experiment.
Theory and Hypothesis Development Boundary spanning and boundary spanning information sharing
for information (Ancona & Caldwell, 1992a; Marrone, 2010). The last category is the common focus of many empirical studies (Tortoriello, Reagans, & McEvily, 2012) and the focus of the current study. Specifically, external parties refer to external members either within or outside the organizations in which the focal team locates (Marrone, 2010).
Boundary spanning is important for a number of team performance outcomes
(Tortoriello et al., 2012) including routine work (Ancona & Caldwell, 1992b) and innovation (Cohen & Levinthal, 1990; Hargadon, 1998), because external information acquired through boundary spanning enables the team to understand the business environment, and inspires the team to generate new ideas and adopt new ways of conducting work (Somech & Khalaili, 2014). However, the benefits that can potentially be created through boundary spanning are often unrealized, probably because information transfer across boundaries is not always successful (Tortoriello et al., 2012).
Previous research has explored the causes for such unsuccessful information transfer in boundary spanning. Taking a social perspective, some research suggested that boundary spanners may not always transfer external information to the team. This is because boundary spanners sometimes tend to maintain their competence and influence (Cross, Nohria, & Parker, 2002; Crozier, 1964; Gould & Fernandez 1989) by withholding external information they have acquired or because they do not want to put enough effort into transferring external information to the team (Tortoriello et al., 2012). Nevertheless, given the difficulty of
To understand this problem, we first need to understand boundary spanning
information sharing. I refer to boundary spanning information sharing as the process after
boundary spanners have acquired external information and during which they decide whether to share a piece of external information with the team. Boundary spanning information
sharing is more complex than information sharing in general and team internal information
sharing because boundary spanning information sharing is embedded in a social context that involves three parties: the external source, the boundary spanner, and the boundary spanner’s team. In particular, the complexity manifests in two ways. First, the boundary spanner is both an information recipient and a (potential) information sender. Boundary spanning information
sharing depends on the boundary spanner’s recognition of the external information after it is
received. Especially, such recognition is required only in boundary spanning information
sharing where the information sender is not the original source of the information. Second, boundary spanning information sharing is perceived as more risky than general information
sharing and team internal information sharing. This is because people have relatively little knowledge to evaluate external information which is featured by uncertainty and can result in unfavorable outcomes. For example, inaccurate information harms team performance
(Marrone, 2010) and a team member may be viewed as incompetent for sharing external information that is viewed as valueless by the team.
Consequently, the key determinant of boundary spanning information sharing is boundary spanners’ recognition of external information. Individuals are receptive to external knowledge only when they view it as potentially valuable (Dokko, Kane, & Tortoriello, 2014). It is thus reasonable to expect the sharing to be carried out only when boundary
valuable (Maltz, 2000), is an important determinant of boundary spanning information
sharing. Fields of research that are relevant to boundary spanning has shown that information
quality determines the adoption of information. For example, in the advice taking literature, judges’ evaluation of advice quality determines whether they will use this advice (Harvey, Harries, & Fischer, 2000; Jungermann, 1999). Similarly, the information systems literature has suggested that PIQ determines one’s intention of using this information (Nicolaou & McKnight, 2006). In the following section, I focus on identifying the key dimensions of PIQ that influence boundary spanning information sharing.
Perceived information quality in boundary spanning information sharing
PIQ is multidimensional (Moenaert & Souder, 1996). Research from various domains including boundary spanning (Menon & Pfeffer, 2003), information systems (e.g., Nicolaou & McKnight, 2006), advice taking (e.g., Yaniv, 2004), persuasion (Petty & Cacioppo, 1986), and communication (Mojzisch et al., 2014) has suggested that information uniqueness (newness),
uniqueness (rareness), credibility, and relevance are the key dimensions of PIQ that influence
information processing. I will explain the links between these PIQ dimensions and boundary
spanning information sharing separately in detail
Uniqueness (newness). Uniqueness (newness) refers to the extent to which a piece of
information is distinct from existing information owned by the user. Most research on boundary spanning attributes the merit of external information to uniqueness (e.g.,
Dahlander, O'Mahony, & Gann, 2016; Wong & Boh, 2014), and more specifically, newness. Acquiring information that is distinct and new to the team is the fundamental drive of people engaged in boundary spanning (Dokko et al., 2014). This is because unlike internal sources (e.g., collocated workers), external sources are more likely to provide nonredundant
“information benefits” to the team because external information embedded in the focal team’s network (Burt, 1992) is less likely to be redundant compared with team internal information. Hansen (1999) also suggested that externally-oriented boundary spanning activities increases information diversity for the focal team. Knowing the benefits of new information, people are more likely to share such information with the team.
Uniqueness (rareness). Importantly, the value of unique external information is not
only about being new to the internal team members, but also about being rare or less accessible to teams in other companies, a point that is not much explicated in the boundary spanning literature. Uniqueness (rareness) refers to the extent to which a piece of information is not commonly available. According to the resource-based view (Barney, 1991), rare
resources create competitive advantage (Brush, Greene, & Hart, 2001) by making a
difference between those who own and those who do not own it. Rare external information, as a type of rare resources, brings “information advantage” to the team (Cross & Cummings, 2004; McEvily & Zaheer, 1999) and helps the team to outcompete in the marketplace.
Rareness and preciousness of external information is also a reason why people value external information (Dahlander et al., 2016; Menon & Pfeffer, 2003). Importantly, information
rareness is independent of information newness. For example, a piece of information about
the market trend acquired from external newspapers is new to the focal team but is not rare because it is highly accessible. Boundary spanning information that is new but not rare is not perceived as valuable as information that is both new and rare. In other words, rareness also determines how people perceive the value of external information and thus influences
boundary spanning information sharing.
Credibility. Credibility refers to the extent to which a piece of information is
1982) and are aware that inaccurate external information, if being transferred to the team, can harm team outcomes (Marrone, 2010). Given that boundary spanning information sharing is a risky decision, team members will carefully evaluate external information credibility before sharing it with the team. Fields of research such as advice taking (e.g., Yaniv, 2007),
persuasion (e.g., Petty & Cacioppo, 1986), information systems (Maltz, 2000; Nicolaou & McKnight, 2006), and communication literature (Mojzisch, 2014) all suggest that information
credibility or accuracy determine the judgment of information quality and influence the
information recipient’s intention of using this information or recommending it to his or her colleagues. It is thus reasonable to expect that team members also account for information
credibility as an aspect of external information value when deciding whether to share it with
the team.
Relevance. Relevance refers to the extent to which information addresses the user’s
needs (Miller, 1996). A piece of external information is of high PIQ if it is related to the information user’s concerns and can be applied to the information user’s tasks. The
information systems literature (Huber, 1982; Maltz, 2000) has demonstrated that information
relevance determines the transfer of information across organizational units, probably
discussion above, I propose that information uniqueness (newness), uniqueness (rareness),
credibility, and relevance as four critical aspects of PIQ, which together determine a
boundary spanner’s tendency to share a piece of external information with the team.
Hypothesis 1: Perceived information quality is positively influenced by perceived information uniqueness (newness), uniqueness (rareness), credibility, and relevance.
Hypothesis 2: Perceived information uniqueness (newness), uniqueness (rareness), credibility, and relevance are positively related to the likelihood of boundary spanning information sharing.
As discussed earlier, people often rely on social cues to judge the quality of boundary spanning information because of communication boundaries (Tushman & Scanlan, 1981) and a lack of objective measures of information quality. The elaboration likelihood model of persuasion (Petty & Cacioppo, 1986) suggests that, when people are unable to process
information, they rely on simple cues such as source expertise and the number of sources that repeat this information to judge the information. Borrowing insights from this model, I attempt to explain how these source characteristics influence boundary spanning information
sharing in the following session.
Information source as social cues of perceived information quality
One characteristic of information sources is source expertise which is defined as “the extent to which a source is perceived to be capable of making correct assertions”
people know relatively little about the field the information is concerned with (Petty & Cacioppo, 1986).
Another characteristic of information sources is the number of sources that repeat a piece of information. The likelihood of elaboration model of persuasion (Petty & Cacioppo, 1986) argues that the motivation of an information recipient to attend to a message increases with the number of sources that provide this message, which is referred to as ‘multiple source effect’. Petty and Cacioppo (1986) argue that multiple source effect is common in our daily life. For example, at political rallies, the audience may see several speakers articulate the same advocacy regarding a candidate or some political issues. Yaniv and Milyavsky (2007) also suggest that the processing of multiple opinions from multiple sources frequently happens in real life. Similarly in boundary spanning, boundary spanners often have multiple external sources (Tortoriello et al., 2012) from which they receive information that is
unavailable within the team. These external sources can be, but are not limited to one’s family members, friends, ex-colleagues, and school mates (Teigland, & Wasko, 2003). Hence, it is possible that a boundary spanner hears the same piece of information from multiple sources, which I refer to as source duplication. Based on the above discussion, I propose that PIQ of a piece of boundary spanning information is a function of source
expertise and source duplication (i.e., whether sources share duplicated information with the
boundary spanner).
by a source with high expertise is likely to be uncommon. Further, people tend to show obedience to experts (Milgram, 1974) and believe relying on experts is an efficient shortcut to making good decisions (Cialdini, 2001). In other words, people tend to believe that experts’ information is relevant and important for decision making. Therefore, I propose that
source expertise positively relates to perceived information credibility, uniqueness (rareness),
and relevance, which in turn stimulates boundary spanning information sharing.
Hypothesis 3a: High (vs. low) source expertise leads to a higher likelihood of boundary spanning information sharing and this relationship is mediated by perceived information credibility.
Hypothesis 3b: High (vs. low) source expertise leads to a higher likelihood of boundary spanning information sharing and this relationship is mediated by perceived information uniqueness (rareness).
Hypothesis 3c: High (vs. low) source expertise leads to a higher likelihood of boundary spanning information sharing and this relationship is mediated by perceived information relevance.
Source duplication has a different influence on PIQ. Source duplication results in
information is highly accessible. The boundary spanner may further infer that this piece of information is not new to the team because other team members may have received it, for example, by approaching their own external information sources. Moreover, the boundary spanner may infer that this information is likely to be known by people outside his or her team (e.g., people from rival companies) and therefore can hardly bring competitive
advantage to the team. Therefore, I propose that source duplication (vs. no duplication) leads to higher perceived information credibility which in turn increases boundary spanning
information sharing. On the other hand, source duplication (vs. no duplication) leads to
lower newness and rareness, which in turn decreases boundary spanning information
sharing.
Hypothesis 4a: The positive influence of source duplication (vs. no source duplication) on the likelihood of boundary spanning information sharing is mediated by perceived information credibility.
Hypothesis 4b: The negative influence of source duplication (vs. no source duplication) on the likelihood of boundary spanning information sharing is mediated by perceived information uniqueness (newness).
Hypothesis 4c: The negative influence of source duplication (vs. no source duplication) on the likelihood of boundary spanning information sharing is mediated by perceived information uniqueness (rareness).
The influence of source expertise and source duplication, however, may not be independent of each other. First, as I mentioned earlier, source expertise and source
duplication both positively influence perceived information credibility. I expect their effects
information credibility between information shared by a single source and that shared by multiple sources. In contrast, when a source is less capable of making correct assertions (Pornpitakpan, 2004), the perceived credibility of information from such a source will be largely determined by whether it is confirmed by another source. I therefore expect source
expertise moderates the relationship between source duplication and credibility. Credibility in
turn, will mediate the interaction effect of source duplication and source expertise on the likelihood of boundary spanners sharing external information with the team. I thus expect:
Hypothesis 5a: The positive effect of source duplication (vs. no source duplication) on credibility is stronger with low (vs. high) source expertise.
Hypothesis 5b: The positive indirect effect of source duplication (vs. no source duplication) on the likelihood of boundary spanning information sharing through credibility is stronger with low (vs. high) source expertise.
Second, I have argued that boundary spanners perceive a piece of information as newer and rarer when it is shared by one source than by multiple sources. I argue that when
source expertise is high, such information would be perceived as even more so because it is
from a source who has special knowledge that is probably not available to the team and to the general public. In contrast, when the source does not have special knowledge (i.e., when
source expertise is low), people are likely to perceive the information it provides as ordinary
and probably redundant with what the team or the general public already knows, even though it is the only source that provides this information. I therefore expect source expertise
moderates the relationship between source duplication and uniqueness (including newness
and rareness). Uniqueness (newness) and uniqueness (rareness) in turn, mediate the
interaction effect of source duplication and source expertise on the likelihood of boundary
spanning information sharing. I thus hypothesize:
uniqueness (newness) is stronger with high (vs. low) source expertise.
Hypothesis 6b: The positive indirect effect of no source duplication (vs. source
duplication) on the likelihood of boundary spanning information sharing through uniqueness (newness) is stronger with high (vs. low) source expertise.
Hypothesis 7a: The positive effect of no source duplication (vs. source duplication) on uniqueness (rareness) is stronger with high (vs. low) source expertise.
Hypothesis 7b: The positive indirect effect of no source duplication (vs. source
duplication) on the likelihood of boundary spanning information sharing through uniqueness (rareness) is stronger with high (vs. low) source expertise.
The role of information valence
The “negativity bias” suggests that humans attend more to negative than positive events (Rozin & Royzman, 2001). Negative events are often surprising and unexpected (Taylor, 1991). Negative events attract more cognitive resources and trigger deeper scrutiny than positive events as people try hard to make sense of negative events (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001). The negativity bias is so salient across domains (e.g., information processing) that many researchers tend to suggest it as unconditional (e.g., Baumeister et al., 2001). When we consider the “negativity bias” in boundary spanning, it is reasonable to expect information valence to have an impact over and above that of source characteristics on boundary spanning information sharing. In the current study, I explore the effect of information valence on boundary spanning information sharing through PIQ dimensions in a similar way as I hypothesized earlier the effects of source duplication and
source expertise.
Method Design and participants
questionnaire via Amazon.com’s Mechanical Turk (MTurk) in exchange for 1.75 USD (approximately 1.55 EUR). Ten participants who either failed the attention check or the manipulation checks or both were removed from the analyses, resulting in 395 valid
observations (173 women, 222 men, Mage = 41.38 years) for the analyses. The questionnaire
had a 2 (Source Duplication: Yes vs. No) × 2 (Source Expertise: Low vs. High) × 2 (Information Valence: Positive vs. Negative) between-subjects design. Participants were randomly assigned to one of the experimental conditions. The survey took about 20 mins.
Procedures and manipulations
Participants read that the study was about dealing with external information and were asked to read a scenario in which they worked in the product development team of a
company. The team was deciding whether to launch Product A to the market. Participants imagined that to help the team make the decision, they approached people outside the team for external information inputs. The scenario in the no source duplication, high [low] source expertise, and negative information condition read as follows:
You mentioned Product A to your friend, Bill, who has an entry-level [high] position in the government sector of education [commerce]. Bill said: “I have heard that the
government plans to establish a new regulation shortly. In my opinion, this is not favorable to products like Product A, so I think it might not be a good opportunity to launch Product A to the market.”
In the source duplication condition, there was an additional paragraph manipulating source duplication, which read as follows:
In the positive information condition, the above negative impact of the foreseen government regulation was changed to positive impact (i.e., from “not favorable” to “favorable” and from “hinder” to “support”) (see Appendix).
Measures
Perceived Information Quality. I measured the four dimensions of PIQ and PIQ on
seven-point bipolar scales by asking participants how they perceive the received information in terms of provided criteria (i.e., credibility, uniqueness [newness], uniqueness [rareness], and relevance). Credibility was measured with four items (e.g., “Absolutely not credible (1) – Absolutely credible (7)” and “Absolutely not correct (1) – Absolutely correct (7)”, α = .92).
Uniqueness (newness) was measured with four items (e.g., “Absolutely not new to the
product development team (1) – Absolutely new to the product development team (7)” and “Absolutely overlaps with the existing information in the product development team (1) – Absolutely does not overlap with the existing information in the product development team (7)”, α = .86). Uniqueness (rareness) was measured with four items (e.g., “Absolutely known by rival companies (1) – Absolutely unknown by rival companies (7)” and “Absolutely common information (1) – Absolutely uncommon information (7)”, α = .90). Relevance was measured with six items (e.g., “Absolutely irrelevant for the team's decision (1) – Absolutely relevant for the team's decision (7)” and “Absolutely does not address the team’s needs (1) – Absolutely addresses the team’s needs (7)”, α = .94). PIQ was measured with six items (e.g., “Absolutely not of high quality (1) – Absolutely of high quality (7)” and “Absolutely not worthwhile (1) – Absolutely worthwhile (7)”, α = .91).
Boundary spanning information sharing. To measure boundary spanning
you received from Bill (or Bill and Chris) with the product development team?”(1 = Extremely unlikely, 7 = Extremely likely).
Manipulation check. To check the manipulation of source duplication, I asked
participants to indicate in numerals how many people outside their team shared information about Product A with them in the scenario. I also checked the manipulation of source expertise by asking participants how much they disagree or agree with the four statements about expertise of Bill (or Bill and Chris). Example items were “Bill was (or Bill and Chris
were) knowledgeable for providing information about Product A.” and “Bill (or Bill and Chris) had expertise for providing information about Product A.” (1 = Strongly disagree, 7 =
Strongly agree). To check the manipulation of information valence, I asked participants to indicate whether the received information was supporting or not supporting the idea to launch Product A.
A 2 (Source Duplication: Yes vs. No) × 2 (Source Expertise: Low vs. High) × 2 (Information Valence: Positive vs. Negative) analysis of variance (ANOVA) on the perceived expertise of information source(s) revealed that the manipulation of high source expertise led to higher perceived expertise of information source(s) (low source expertise = 0, high source expertise = 1; M = 5.15 vs. M = 5.75), F(1, 387) = 42.83, p < .001, ƞ2 = .10. For the
manipulation of source duplication, I found a significant difference (χ2 (2) = 363.66, p < .001). Participants in the source duplication condition were more likely to indicate “2” (98%) as the number of sources while those in the no source duplication condition were more likely to indicate “1” (95.4%) as the number of sources. Similarly, I found a significant difference for the manipulation of information valence (χ2 (1) = 327.94, p < .001). Participants in the positive information condition were more likely to interpret the
the idea to launch Product A. I conclude that my manipulations were successful. Six participants who failed the manipulation check of source expertise, information valence, or both, were removed from the following analyses.
Results Overview of analyses
The analyses were carried out as follows. To test the construct validity of PIQ, I first conducted a confirmatory factor analysis using the lavaan package (Rosseel, 2012) in R environment. I then conducted a structural equation modeling using the same lavaan package in R to test the indirect effects of source duplication and source expertise on the likelihood of
boundary spanning information sharing through multiple dimensions of PIQ. I further
conducted 2 × 2 × 2 ANOVAs to understand the main and interactive effects of source
duplication, source expertise, and information valence on different dimensions of PIQ in
SPSS. Finally, I tested the conditional indirect effects of source duplication on the likelihood of boundary spanning information sharing using Preacher, Rucker, and Hayes’s (2007) approach in SPSS.
Descriptive statistics
Table 1 presents the means, standard deviations, and inter-correlations for variables used in the study.
The validity of PIQ
I hypothesized earlier that PIQ has four distinctive dimensions: credibility, uniqueness
(newness), uniqueness (rareness), and relevance. To test the validity of the four-factor PIQ
model, the latent variables representing credibility and relevance were modeled in the same way as in the hypothesized model while items used to measure uniqueness (newness) and
uniqueness (rareness) were combined to form one latent variable. This model thus left out the
distinction between uniqueness (newness) and uniqueness (rareness). By comparing my hypothesized model with these two alternative models, I could test whether PIQ is multidimensional and is composed of credibility, uniqueness (newness), uniqueness
(rareness), and relevance.
The CFA for the hypothesized four-factor model had a model chi-square of 350.06 (df = 153; CFI = .96, TLI = .95, RMSEA = .07, SRMR = .06). The one-factor CFA had a model chi-square of 2913.03 (df = 153; CFI = .49, TLI = .43, RMSEA = .23, SRMR = .20). The three-factor CFA had a model chi-square of 725.79 (df = 153; CFI = .89, TLI = .88, RMSEA = .11, SRMR = .08). There were significant differences in model fit between the
hypothesized model and the one-factor model (∆χ2 = 2563.00, p < .001), and the three-factor model (∆χ2 = 375.73, p < .001). From these results, I concluded that the hypothesized model had a better model fit than the other two models and that the items were reasonable measures of their respective PIQ dimensions.
Subsequently, I regressed observed PIQ on uniqueness (newness), uniqueness
(rareness), credibility, and relevance. This produced a significant regression model, R2 = .68,
F(4, 390) = 208.82, p < .001. The effects for uniqueness (rareness) (B = .19, t = 5.11, p
< .001), credibility (B = .31, t = 9.00, p < .001), and relevance (B = .61, t = 16.26, p < .001) were significant. However, the effect for uniqueness (newness) was not significant (B = -.01,
t = -.33, p = .745). An explanation for the nonsignificant effect of uniqueness (newness) could
be that there was correlation between the uniqueness (newness) construct and uniqueness
(rareness) construct although the confirmatory factor analysis suggested treating them as
(VIF) and the VIF values for the proposed dimensions of PIQ were all around 1.5, which suggested that there was only mild multicollinearity. In other words, the multicollinearity problem was not severe enough to warrant corrective measures. Hypothesis 1 was supported.
Hypothesis testing
To test the indirect effects of source duplication and source expertise on the likelihood of boundary spanning information sharing through different dimensions of PIQ, I conducted structural equation modeling (SEM) to test multiple mediations simultaneously. This
mediation model showed a good fit with the data: χ2 (df =5) = 8.18, p = .146, CFI = .99, TLI = .97, RMSEA=.04, SRMR = .04.
Results of the effects of PIQ dimensions on the likelihood of boundary spanning
information sharing partially supported hypothesis 2: credibility (B = .14, p = .034),
uniqueness (newness) (B = .11, p = .024), and relevance (B = .54, p < .001) were positively
while uniqueness (rareness) (B = -.11, p = .022) was negatively related to the likelihood of
boundary spanning information sharing. The direct relationships among source duplication,
source expertise, PIQ dimensions, and the likelihood of boundary spanning information sharing are depicted in Figure 1.
Results of the mediated relationships among source duplication, source expertise, and the likelihood of boundary spanning information sharing are shown in Table 2. The indirect effects of source expertise on likelihood of boundary spanning information sharing were significant through credibility (B = .06, p < .05, 95% CI [0.01, 0.13]) and relevance (B = .17,
p < .01, 95% CI [0.06, 0.28]). Hypothesis 3a and 3c were thus supported: perceived
information credibility and relevance positively mediated the positive relationship between
source expertise and the likelihood of boundary spanning information sharing. In contrast,
source duplication, the indirect effects of source duplication on the likelihood of boundary spanning information sharing through credibility (p = .060, 95% CI [0.01, 0.09]), uniqueness
(newness) (p = .122, 95% CI [-0.08, -0.001]), and uniqueness (rareness) (p = .054, 95% CI
[0.01, 0.11]) were nonsignificant. Hypothesis 4a, 4b, and 4c were not supported.
I hypothesized in hypothesis 5a, 6a, 7a, and 8a that source expertise moderates the effects of source duplication on different credibility, uniqueness (newness), and uniqueness
(rareness). To test these hypotheses and explore the effects of information valence on PIQ
dimensions, I conducted 2 × 2 × 2 ANOVAs. Results are shown in Table 3.
Credibility. Inconsistent with hypothesis 5a, the interaction effect between source duplication and source expertise on credibility was nonsignificant (F(1, 387) = .09, p = .766,
ƞ2 = .00). The analysis yielded a main effect for source expertise (F(1, 387) = 24.07, p < .001, ƞ2 = .06): external information was perceived as more credible by participants in the high source expertise condition (M = 5.57, SD = .88) than those in the low source expertise condition (M = 5.13, SD = .96), which was in line with the result of testing hypothesis 3a.
Information valence showed a main effect on credibility: external information was perceived as less credible by participants in the negative information condition (M = 5.23, SD = .97) than by those in the positive information condition (M = 5.47, SD = .91). The analysis also revealed a significant main effect for source duplication (F(1, 387) = 13.75, p < .001, ƞ2 = .03) qualified by a Source duplication × Information valence interaction (F(1, 387) = 5.05,
p < .05, ƞ2 = .01) (Figure 2). Simple effects analysis showed that in the negative information
condition, participants received external information from duplicated sources perceived it as more credible (M = 5.50, SD = .88) than those received it from a single source (M = 4.96, SD = .99; F(1, 391) = 5.05, p < .001, ƞ2 = .04). This difference was, however, not present for the positive information condition (p =.267).
duplication on uniqueness (newness) (F(1, 387) = 4.51, p < .05, ƞ2 = .01) was qualified by an Source expertise × Source duplication interaction (F(1, 387) = 3.93, p < .05, ƞ2 = .01) as shown in Figure 3. Simple effects analysis revealed that in the high source expertise condition, participants receiving external information from a single source perceived it as newer (M = 5.09, SD = 1.35) than those received it from duplicated sources (M = 4.56, SD = 1.32; F(1, 391) = 8.10, p < .01, ƞ2 = .02). This difference was, however, not significant for low source expertise condition (p = .989).
In addition, the analysis yielded a significant main effect for information valence on
uniqueness (newness) (F(1, 387) = 20.59, p < .001, ƞ2 = .05): participants in the negative information condition perceived the external information as newer (M = 5.01, SD = 1.31) than those in the positive information condition (M = 4.43, SD = 1.25).
Uniqueness (rareness). The interaction effect between source duplication and source expertise on uniqueness (rareness) was nonsignificant (F(1, 387) = .47, p = .495, ƞ2 = .001), not supporting hypothesis 7a. In line with what I argued earlier, the analysis yielded a
significant main effect for source expertise (F(1, 387) = 19.26, p < .001, ƞ2 = .05) and source
duplication (F(1, 387) = 12.31, p < .001, ƞ2 = .03) on uniqueness (rareness). External information was perceived as rarer in the high source expertise condition (M = 4.96, SD = 1.24) than in the low source expertise condition (M = 4.42, SD = 1.29), and in the no source duplication condition (M = 4.91, SD = 1.27) than in the source duplication condition (M = 4.48, SD = 1.28).
In addition, the analysis showed a significant main effect for information valence on
uniqueness (rareness) (F(1, 387) = 5.94, p < .05, ƞ2 = .02). Participants in the negative information condition perceived the external information as rarer (M = 4.84, SD = 1.35) than those in the positive information condition (M = 4.54, SD = 1.22). However,
and relevance, the analysis showed a significant main effect of source expertise on relevance (F(1, 387) = 12.18, p < .001, ƞ2 = .03). Perceived information relevance was higher in the high source expertise condition (M = 6.10, SD = .94) than the low source expertise condition (M = 5.75, SD = 1.10). In addition, the analysis revealed a significant main effect information valence on relevance (F(1, 387) = 9.20, p < .01, ƞ2 = .02). Perceived information relevance was higher in the negative information condition (M = 6.08, SD = .98) than the positive information condition (M = 5.77, SD = 1.07).
In exploring the effects of information valence, I also found a main effect of
information valence on the likelihood of boundary spanning information sharing (F(1, 387) = 5.74, p < .05, ƞ2 = .015): participants were more likely to share negative external
information (M = 6.35, SD = 1.15) than positive external information (M = 6.06, SD = 1.25). To sum up results of ANOVAs, I did not find significant interaction effects between
source expertise and source duplication on credibility (hypothesis 5a), uniqueness (rareness)
(hypothesis 7a), and relevance (hypothesis 8a). Accordingly, hypothesis 5b, 7b, and 8b were not supported as the relationships proposed in hypothesis 5a, 7a, and 8a were prerequisites for those proposed in hypothesis 5b, 7b, and 8b. On the other hand, I found a significant interaction between source expertise and source duplication on uniqueness (newness) (hypothesis 6a) and a significant interaction between information valence and source
duplication on credibility. Consequently, in the following analyses, I only tested the indirect
effects of source duplication on the likelihood of boundary spanning information sharing through uniqueness (newness) moderated by source expertise (hypothesis 6b) and the indirect effects of source duplication on the likelihood of boundary spanning information sharing through credibility moderated by information valence. I conducted two moderated mediation tests following Preacher et al.’s (2007) procedure. Results are shown in Table 4 and 5.
source duplication on the likelihood of boundary spanning information sharing through uniqueness (newness) was only significant for participants in the high source expertise
condition (B = -.10, 95%, CI [-0.22, -0.03]), and was not significant for participants in the low source expertise condition (B = -.0005, 95% CI [-0.08, 0.08]) (Table 4). Hypothesis 6b was thus supported.
In exploring the role of information valence as a moderator, I found that information valence moderated the positive indirect effect of source duplication on the likelihood of
boundary spanning information sharing through credibility: the indirect effect was only
significant for participants in the negative information condition (B = .22, 95% CI [0.11, 0.38]), and was not significant for participants in the positive information condition (B = .06, 95% CI [-0.04, 0.18]) (Table 5).
Table 6 summarizes the results of hypothesis testing.
Discussion
The overarching goal of the current study is to advance the literature on boundary spanning by investigating what determines boundary spanning information sharing, which is a prerequisite for the successful absorption of external information. Borrowing insights from relevant literatures including advice taking (Yaniv, 2004; Yaniv & Milyavsky, 2007),
communication (Mojzisch et al., 2014), persuasion (Petty & Cacioppo, 1986), and information systems (Nicolaou & McKnight, 2006), I argued that perceived information quality drives boundary spanning information sharing and perceived information quality is influenced by source characteristics (i.e., source duplication and source expertise) and information valence. I tested my hypotheses and explored the effect of information valence based on a series of analyses.
credibility, uniqueness (newness), uniqueness (rareness), and relevance. Consequently, I
found that credibility, uniqueness (newness), and relevance were positively related to the likelihood of boundary spanning information sharing. However, uniqueness (rareness) was negatively related to the likelihood of boundary spanning information sharing. One possible explanation for this negative effect, as I mentioned earlier, is that there was still correlation between the uniqueness (newness) construct and uniqueness (rareness) construct and
therefore their effects might have canceled out each in predicting the likelihood of boundary
spanning information sharing. To examine this explanation, I combined all the items of
uniqueness (rareness) and uniqueness (newness) to form a variable of uniqueness and
regressed the likelihood of boundary spanning information sharing on credibility,
uniqueness, and relevance. This produced a significant regression model, R2 = .67, F(3, 391)
= 268.94, p < .001. The effects for uniqueness (B = .16, t = 4.76, p < .001), credibility (B = .32, t = 9.13, p < .001), and relevance (B = .61, t = 16.11, p < .001) were all significant. These results supported my explanation. Another explanation for the negative effect of
uniqueness (rareness) is that the external information mentioned in the scenario was about
government regulations and people may perceive sharing government-related external information as less appropriate when it was not widely known.
I also found the expected effects of source characteristics. First, I found the expected effects of source duplication and source expertise on different dimensions of perceived information quality. Source duplication positively predicted credibility while negatively predicted uniqueness (newness) and uniqueness (rareness). Source expertise positively predicted credibility, uniqueness (rareness), and relevance. Second, I demonstrated the indirect effect of source expertise on the likelihood of boundary spanning information
sharing through perceived information quality dimensions. My findings showed that people
with the team because they perceived it as credible and relevant. Third, I demonstrated an interaction effect between source duplication and source expertise on uniqueness (newness): when source expertise was high, external information shared by a single source (vs.
duplicated sources) was perceived as newer. When source expertise is low, however, whether the information is duplicated across sources will not influence the perceived information
uniqueness (newness). Further, this interaction effect predicts the likelihood of boundary spanning information sharing such that people receiving external information from a single
source (vs. duplicated sources) intended to share it with the team, because they view such information as new to the team, but only when source expertise is high.
There were also interesting findings regarding the role of information valence in
boundary spanning information sharing. First, information valence was related to all
dimensions of perceived information quality: negative (vs. positive) external information was perceived as less credible but newer, rarer and more relevant. Second, information valence interacted with source duplication to influence perceived information credibility: external information shared by duplicated sources (vs. a single source) was perceived as more credible only when the information was negative. This suggests that social confirmation helps people overcome the skepticism on external information only when such information is negative. Further, this interaction effect predicts the likelihood of boundary spanning information
sharing such that people receiving external information from duplicated sources (vs. a single
source) intended to share it with the team, because they view such information as credible, but only when the external information is negative. Third, people were more likely to share negative external information than positive external information with the team.
somehow develop certain heuristics based on source characteristics and information valence for deciding whether to share boundary spanning information with the team. These results clearly show that boundary spanners do not share external information with the team
mindlessly, but rather share it in a controlled manner based on perceived information quality. To sum up, these results have important implications for understanding the nature of sharing boundary spanning information and how it is influenced by the context it is embedded in.
Theoretical implications
By investigating the basic, micro-level boundary spanning information sharing, this study contributes to the theory of boundary spanning by highlighting the essential step for understanding boundary spanning effectiveness (losses). Previous work on boundary spanning and the transfer of boundary spanning information has overlooked the step during which boundary spanners share boundary spanning information with the team. However, since unshared boundary spanning information can hardly exert influence on team performance, this step is essential for understanding not only how boundary spanning influences team performance, but also how boundary spanners affect boundary spanning outcomes (Dahlander et al., 2016), which are not much addressed in previous research.
Another contribution lies in the notion of information quality. First, while previous research attributes the failure of boundary spanners importing external information to a lack of either external or internal networking abilities (Tushman & Scanlan, 1981), or to their tendency to withhold external information as a way to maintain power and influence (Crozier, 1964; Tortoriello et al., 2012). I demonstrate the perceived low quality of external
negative relationship between uniqueness (rareness) and the likelihood of boundary spanning
information sharing. Despite the potential overlap between uniqueness (newness) and
uniqueness (rareness), this might have two implications. First, this may suggest that uniqueness (rareness) does not have much influence over and above that of uniqueness (newness). Second, this may suggest that even valuable, rare information is not necessarily
shared with the team, which again highlights the possibility of unsuccessful sharing as a reason why boundary spanning is not always effective.
More indirectly, although my study did not address the links between these
dimensions with team performance, my conceptualization of perceived information quality enriches the understanding of the value and probably the performance implications of boundary spanning information. For example, the notion of uniqueness being composed of
rareness and newness is beneficial for the studies focused on boundary spanning’s impact on
innovation (e.g., Dahlander et al., 2016; Obstefeld, 2005). Those studies often only address the value of nonredundant (i.e., new) information in stimulating innovation while ignoring the competitive advantage associated with rare information. Similarly, the notion of
credibility and relevance may enhance our understanding of why certain boundary spanning
patterns are more effective in shaping team performance. For example, previous research suggested that a Simmelian boundary spanning tie (i.e., a boundary spanning tie involving two parties that are connected to a common third party) stimulates innovation because it facilitates cooperation and mitigates conflict between the boundary spanning parties (Tortoriello & Krackhardt, 2010). The notion of information quality, however, provides an alternative explanation for the effectiveness of Simmelian boundary spanning ties: it could be that these ties enhance perceived credibility and relevance of boundary spanning information, which is in turn more likely to be passed on to the team and then influence team performance.
the role of source characteristics in boundary spanning. Source expertise has been studied in different forms such as prestige (e.g., Dokko et al., 2014) and professional backgrounds (e.g., Ratcheva, 2009) while source duplication is less concerned in the boundary spanning
literature. In other words, the notion of source characteristics is implied but not
systematically examined in the boundary spanning literature. My study contributes to this growing body of literature by showing that source expertise and source duplication influence
boundary spanning information sharing by influencing perceived information quality.
Moreover, source characteristics belong to the properties of boundary spanning ties which have influence over and above that of network structure (Cross & Cummings, 2004). Cross and Cummings (2004) proposed that certain boundary spanning ties provide better
information in terms of relevance, uniqueness, novelty, and value than others without testing this proposition. In line with this proposition, my findings show that boundary spanning ties to external sources with high expertise are associated with higher perceived credibility,
uniqueness, and relevance. Future research can continue to identify more important tie
properties that play a role in boundary spanning. For example, future research can investigate the effect of tie strength on perceived information quality and boundary spanning
information sharing.
bias” in boundary spanning. For example, future research can investigate whether boundary spanning helps a company to monitor threats better than does it help a company to identify the opportunities. Further, the current findings point out the importance of having more content concerns when we study boundary spanning—in general boundary spanning
information and its impact is not distinguished based on its content in the boundary spanning literature. To enhance theory development, future studies should investigate the influence of information valence as well as other aspects of information content on boundary spanning.
Practical implications
This study carries also practical implications for organizations and teams that rely on boundary spanning to enhance effectiveness and maximize adaptability. My findings suggest that by encouraging team members to approach expert sources for information, managers can ensure boundary spanning efficiency and effectiveness because such information is likely to be shared with and thus leveraged by the team. This is important as inefficient or ineffective boundary spanning is associated with costs to individuals as well as the team (e.g., role overload and decreased team identity), which can decrease team productivity and harm team performance (Faraj & Yan, 2009; Marrone et al., 2007).
I also find that the sharing of positive external information is less likely than the sharing of negative external information. Given that positive external information is helpful for teams to understand the trends and opportunities embedded in their business environment (Wong & Boh, 2014), managers should remind their employees of the importance of positive external information during their boundary spanning. In this way, managers can avoid the disproportionate dissemination of positive and negative external information within teams that may bias team decision making.
Limitations and future directions
The current study has some limitations that call for future research. First, boundary
spanning information sharing was measured as intentions rather than real behavior. Although
intentions come before real behavior (Fishbein & Ajzen, 1975), future research may benefit from measuring real behavior of boundary spanning information sharing to test the validity of the current findings (Martinescu, Janssen, & Nijstad, 2019).
Second, there was still overlap between the measurement constructs of uniqueness
(newness) and uniqueness (rareness) in the current study, causing difficulties in capturing
their respective effects on boundary spanning information sharing. Future studies should develop more accurate and suitable items for measuring these two constructs in order to distinguish their effects and advance the understanding of unique information. Moreover, the negative relationship between uniqueness (rareness) and boundary spanning information
sharing calls for further investigation. It could be that people do not perceive rare information
as valuable although in reality it is. Future research may investigate this explanation or consider potential moderators (e.g., appropriateness of sharing) in this relationship.
likelihood of boundary spanning information sharing over and above the influence of perceived information quality. Further, the external sources specified in the current scenario were outside the company while external sources can also be within the same company where the focal team locates (e.g., colleagues in another department of the company). Moreover, the external sources specified in the current scenario were friends and family members who were assumed to be trustworthy. However, in real life, boundary spanners may have to acquire external information from people whom they are not familiar with and whose intentions of providing information are uncertain. Hence, future research may design different scenarios to test whether the current findings are consistent across different boundary spanning contexts.
Last, the current study focuses on the stage of sharing in boundary spanning which is followed by the stage of acceptance in which the team decides whether to accept the external information shared by boundary spanners. Accordingly, future studies could examine
whether the mechanism used to explain the sharing of boundary spanning information also explains the acceptance of boundary spanning information. In this way, we may have an even broader understanding of the role played by perceived information quality in boundary spanning.
Conclusion
This study shows that perceived information quality including credibility, uniqueness
(newness), uniqueness (rareness), and relevance, influences the likelihood of boundary
spanners sharing a piece of external information with their team. In turn, these aspects of perceived information quality are influenced by source expertise, source duplication (i.e., information is provided by duplicated sources), and information valence. Moreover, when
source expertise is high, source duplication decreases boundary spanning information
sharing by decreasing perceived information uniqueness (newness). When external
by increasing perceived information credibility. In addition, this study shows that boundary spanners are more likely to share negative external information than to share positive external information. Overall, this study clarifies the mechanism underlying boundary spanning
information sharing, which have important implications for understanding boundary
References
Ancona, D. G., & Caldwell, D. F. (1992a). Bridging the boundary: External activity and performance in organizational teams. Administrative Science Quarterly, 634-665. doi: 10.2307/2393475
Ancona, D. G., & Caldwell, D. F. (1992b). Demography and design: Predictors of new product team performance. Organization Science, 3(3), 321-341. Retrieved from http://www.jstor.org/stable/2635276
Barney, J. (1991). Firm resources and sustained competitive advantage. Journal of
Management, 17(1), 99-120. doi:10.1177/014920639101700108
Baumeister, R., Bratslavsky, E., Finkenauer, C., & Vohs, K. (2001). Bad is stronger than good. Review of General Psychology, 5(4), 323-370. doi:10.1037/1089-2680.5.4.323 Bechky, B. (2003). Sharing meaning across occupational communities: The transformation of
understanding on a production floor. Organization Science, 14(3), 312-330. doi:10.1287/orsc.14.3.312.15162
Behrend, F. D., & Erwee, R. (2009). Mapping knowledge flows in virtual teams with SNA. Journal of Knowledge Management, 13(4), 99-114.
Birnbaum, M. H., & Mellers, B. A. (1983). Bayesian inference: Combining base rates with opinions of sources who vary in credibility. Journal of Personality and Social
Psychology, 45(4), 792. doi:10.1037/0022-3514.45.4.792
Birnbaum, M. H., & Stegner, S. E. (1979). Source credibility in social judgment: Bias, expertise, and the judge's point of view. Journal of Personality and Social
Psychology, 37(1), 48. doi:10.1037//0022-3514.37.1.48
Brown, S. P., Cron, W. L., & Slocum Jr, J. W. (1998). Effects of trait competitiveness and perceived intraorganizational competition on salesperson goal setting and
Brush, C. G., Greene, P. G., & Hart, M. M. (2001). From initial idea to unique advantage: The entrepreneurial challenge of constructing a resource base. Academy of
Management Perspectives, 15(1), 64-78.
Burt, R. (1992). Structural holes: The social structure of competition. Cambridge, Mass.: Harvard University Press.
Burt, R. S. (2000). The network structure of social capital. Research in Organizational
Behavior, 22, 345-423. doi:10.1016/S0191-3085(00)22009-1
Cialdini, R. B., & Goldstein, N. J. (2004). Social influence: Compliance and conformity. Annual Review of Psychology, 55, 591-621.
doi:10.1146/annurev.psych.55.090902.142015
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128-152. doi:10.2307/2393553
Cross, R., & Cummings, J. N. (2004). Tie and network correlates of individual performance in knowledge-intensive work. Academy of Management Journal, 47(6), 928-937. Cross, R., Nohria, N., & Parker, A. (2002). Six myths about informal networks-and how to
overcome them. MIT Sloan Management Review, 43(3), 67.
Crozier, M. (1964). The bureaucratic phenomenon. Chicago: University of Chicago Press. Dahlander, L., O'Mahony, S., & Gann, D. M. (2016). One foot in, one foot out: how does
individuals' external search breadth affect innovation outcomes?. Strategic
Management Journal, 37(2), 280-302. doi:10.1002/smj.2342
Dokko, G., Kane, A. A., & Tortoriello, M. (2014). One of us or one of my friends: How social identity and tie strength shape the creative generativity of boundary-spanning ties. Organization Studies, 35(5), 703-726. doi:10.1177/0170840613508397
firms. Organization Science, 3(2), 179-202. doi:10.1287/orsc.3.2.179
Faraj, S., & Yan, A. (2009). Boundary work in knowledge teams. Journal of Applied
Psychology, 94(3), 604. doi:10.1037/a0014367
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to
theory and research (Addison-wesley series in social psychology). Reading, Mass.:
Addison-Wesley.
Fiske, S. T. (1980). Attention and weight in person perception: The impact of negative and extreme behavior. Journal of Personality and Social Psychology, 38(6), 889. doi:10.1037/0022-3514.38.6.889
Gould, R. V., & Fernandez, R. M. (1989). Structures of mediation: A formal approach to brokerage in transaction networks. Sociological Methodology, 89-126.
Greitemeyer, T., & Schulz-Hardt, S. (2003). Preference-consistent evaluation of information in the hidden profile paradigm: beyond group-level explanations for the dominance of shared information in group decisions. Journal of Personality and Social
Psychology, 84(2), 322.
Hansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharing
knowledge across organization subunits. Administrative Science Quarterly, 44(1), 82-111. doi:10.2307/2667032
Hargadon, A. B. (1998). Firms as knowledge brokers: Lessons in pursuing continuous innovation. California Management Review, 40(3), 209-227. doi:10.2307/41165951 Harvey, N., Harries, C., & Fischer, I. (2000). Using advice and assessing its
quality. Organizational Behavior and Human Decision Processes, 81(2), 252-273. doi:10.1006/obhd.1999.2874
Hollingshead, A. B. (1996). Information suppression and status persistence in group decision making: The effects of communication media. Human Communication
Research, 23(2), 193-219.
Huber, G. (1982). Organizational information systems: Determinants of their performance and behavior. Management Science, 28(2), 138-155. doi:10.1287/mnsc.28.2.138 Humphreys, P., & Berger, J. (1981). Theoretical consequences of the status characteristics
formulation. American Journal of Sociology, 86(5), 953-983.
Joshi, A., Pandey, N., & Han, G. (2009). Bracketing team boundary spanning: An examination of task‐based, team‐level, and contextual antecedents. Journal of
Organizational Behavior, 30(6), 731-759.
Jungermann, H. (1999). Advice giving and taking. Proceedings of the Hawaii International
Conference on System Sciences,32, 38-38.
Kahneman, D., & Tversky, A. (2000). Choices, values, and frames. Cambridge: Cambridge University Press.
Katz, R., & Allen, T. J. (1982). Investigating the Not Invented Here (NIH) syndrome: A look at the performance, tenure, and communication patterns of 50 R & D Project
Groups. R&D Management, 12(1), 7-20. doi:10.1111/j.1467-9310.1982.tb00478.x Lee, M. T., & Ofshe, R. (1981). The impact of behavioral style and status characteristics on
social influence: A test of two competing theories. Social Psychology Quarterly. Maltz, E. (2000). Is all communication created equal?: An investigation into the effects of
communication mode on perceived information quality. Journal of Product
Innovation Management, 17(2), 110-127.
Manev, I. M., & Stevenson, W. B. (2001). Balancing ties: Boundary spanning and influence in the organization's extended network of communication. The Journal of Business
Marrone, J. A. (2010). Team boundary spanning: A multilevel review of past research and proposals for the future. Journal of Management, 36(4), 911-940.
doi:10.1177/0149206309353945
Marrone, J. A., Tesluk, P. E., & Carson, J. B. (2007). A multilevel investigation of antecedents and consequences of team member boundary-spanning
behavior. Academy of Management Journal, 50(6), 1423-1439.
Martinescu, E., Janssen, O., & Nijstad, B. A. (2019). Self-evaluative and other-directed emotional and behavioral responses to gossip about the self. Frontiers in
Psychology, 9. doi:10.3389/fpsyg.2018.02603
McEvily, B., & Zaheer, A. (1999). Bridging ties: A source of firm heterogeneity in competitive capabilities. Strategic Management Journal, 20(12), 1133-1156. doi:10.1002/(SICI)1097-0266(199912)20:123.3.CO;2-Z
McGuire, W. J. (1968). Personality and attitude change: An information-processing theory. Psychological Foundations of Attitudes, 171, 196.
Menon, T., & Pfeffer, J. (2003). Valuing internal vs. external knowledge: Explaining the preference for outsiders. Management Science, 49(4), 497-513.
doi:10.1287/mnsc.49.4.497.14422
Milgram, S. (1975). Obedience to authority : An experimental view (Harper torchbooks, tb 1983). New York: Harper and Row.
Miller, H. (1996). The multiple dimensions of information quality. Information Systems
Management, 13(2), 79-82. doi:10.1080/10580539608906992
Mojzisch, A., Grouneva, L., & Schulz‐Hardt, S. (2010). Biased evaluation of information during discussion: Disentangling the effects of preference consistency, social validation, and ownership of information. European Journal of Social
