Maintaining Workplace Friendships

Abstract

Workplace friendships are linked to many important processes and outcomes such as employee satisfaction, career development, creativity, and decision-making influence (Sias, 2009). Research has examined how workplace peer friendships develop and deteriorate, but largely ignored their maintenance. The studies reported here addressed this void. Study 1 identified communicative tactics individuals use to maintain friendships with peer coworkers. Study 2 assessed the perceived politeness/positive face threat of the tactics. Study 3 examined the extent to which perceived politeness, task interdependence and individual attachment style predict the likelihood an individual will use specific maintenance tactics.

Keywords

workplace relationships friendships workplace relationship maintenance politeness task interdependence attachment style

The workplace is a key site of interpersonal relationships and the informal relationships that exist in the “white spaces” of the organizational chart are fundamental contexts for organizing (Eisenberg & Goodall, 2004). Friendships among peer coworkers, in particular, are linked with important activities and outcomes including employee job satisfaction and commitment, creativity, information-sharing, decision making influence, and career advancement (Sias, 2009). Scholars have examined peer workplace friendship development and deterioration (Sias & Cahill, 1998; Sias, Heath, Perry, Silva, & Fix, 2004), yet, we know little about how organizational members maintain these important relationships at desired levels. The three studies reported here address this void by identifying a set of peer workplace friendship maintenance tactics, assessing the relative politeness/face-saving nature of those tactics, and examining predictors of maintenance tactic preferences.

Review of Relevant Literature

Workplace Friendship

As “a core aspect” of life (Fehr, 1996, p. 1), friendship penetrates all social spheres, including the workplace (Marks, 1994). In fact, because organizations bring together people with common occupational experiences and engage them in shared activities, the workplace is “. . . a somewhat natural ‘incubator’ for personal relationships that extend beyond the professional boundary” (Sias & Gallagher, 2009, p. 3). Workplace friendships are distinguished from other workplace relationships by two primary characteristics. First, they are voluntary in nature (Rawlins 1992; Sias & Cahill, 1998). Their initiation, development and dissolution are chosen and guided by the relationship partners, not imposed by organizational requirements. Second, workplace friendships are personalistic—friends acknowledge and interact with one another as “whole” persons, not just employees in prescribed organizational roles (Rawlins, 1992; Sias & Cahill, 1998). Friendships develop between supervisors and their subordinate employees but are more common among peer coworkers (Sias, 2009)—the focus of this project.

Workplace Friendship Relational Dynamics

Kram and Isabella (1985) were the first to address the friendship component of peer coworker relationships. Their study identified three primary types of peer workplace relationships. Information peer relationships are characterized by low levels of trust and self-disclosure. These relationships function largely within the work sphere with information exchange regarding work and organizational topics the focus of conversation. Collegial peer relationships incorporate moderate levels of friendship. These relationships are characterized by moderate levels of trust and increasing discussion of non-work-related topics at a moderate level of intimacy. Coworkers in special peer relationships are very close, trust each other greatly, and discuss a virtually limitless breadth of work and nonwork topics. Special peers consider themselves to be very close or best friends (Kramer, 1994; Sias & Cahill, 1998).

Like all interpersonal relationships, workplace friendships are dynamic and affected by factors that both strengthen and threaten the relationship. People may weaken old friendships and start new ones due to the novelty or potential advantages of the latter (Whitmeyer, 2002). Workplace friendships also combine often conflicting individual identities—the coworker identity and the friend identity. As these identities involve different roles and expectations, workplace friendships often experience inherent tensions (Bridge & Baxter, 1992). Workplace friendships can be, therefore, somewhat unstable or, as Whitmeyer (2002) put it, “fickle.”

While Kram and Isabella (1985) identified types of peer relationships, others have examined why and how peer coworker relationships develop into friendships. Sias and Cahill (1998) found that both personal (e.g., personality, similarity) and contextual factors (e.g., proximity, shared tasks) influence an individual’s preferences as to which peer coworkers to befriend. Like all interpersonal relationships, workplace friendships are socially constructed (Sias, 2009; Sigman, 1995). Thus, their development is accompanied by important changes in the communication between the partners, including increased communication frequency, increased breadth of topics, increased intimacy, and decreased caution (Sias & Cahill, 1998).

Several challenges make peer workplace friendships vulnerable to instability and deterioration. These relationships may deteriorate due to conflicting personalities, incidents of betrayal, life events that become so distracting for one partner that the other must carry a heavier workload, and the promotion of one partner to a position of formal authority over the other (Sias & Perry, 2004). The “blended” nature of workplace friendship also introduces tensions that challenge the relationship (e.g., excessive connection) and, if mismanaged, can result in termination of the friendships (Bridge & Baxter, 1992). Like friendship development, workplace friendship disengagement is accomplished communicatively (Sias et al., 2004).

Not all workplace friendships succumb to the challenges as discussed above, however. Instead many coworkers are able to maintain their friendships at a desired level. Below we present three studies examining peer workplace friendship maintenance. The Study 1 identified tactics employees use to maintain friendships with coworkers at a desired state. Study 2 evaluated the perceived politeness of those tactics. Finally, Study 3 assessed the likelihood employees will use the various tactics to maintain workplace friendships, and the extent to which task interdependence and individual attachment style predict tactic likelihood of use.

Study 1: Peer Workplace Friendship Maintenance Tactics

Scholars have defined relationship maintenance in many ways over the past few decades. These include simply keeping a relationship in existence (Dindia & Canary, 1993), keeping the relationship “in a specified state or condition” (Dindia, 2003, p. 3), keeping a relationship “in a satisfactory condition” (Dindia, 2003, p. 3), and keeping a relationship “in repair” (Dindia, 2003, p. 4). Keeping a relationship in existence, in satisfactory condition, or in repair all tend to refer to preventing a relationship from deteriorating or terminating. Our studies focus on maintenance in both deteriorating and escalating situations. Thus, we conceptualize relationship maintenance as maintaining a relationship at a specific level or condition.

Relationship maintenance, as conceptualized here, occurs in three main types of situations (Lee & Jablin, 1995). Escalating situations are those in which one partner feels uncomfortable with the growing closeness of the relationship. Deteriorating situations are those in which one partner is concerned that the other is distancing him- or herself from the relationship and does not want the relationship to deteriorate. Routine situations are those in which neither party feels concerned that the relationship is growing too close or too distant but nonetheless act to maintain the relationship. To limit the scope of our project, our studies focus on strategic or intentional relationship maintenance behavior. As Dainton (2003) explained, intentional relationship maintenance refers to “strategic, conscious efforts” (p. 303) while unintentional relationship maintenance “refers to routine, less conscious maintenance efforts” (p. 303). As escalating and deteriorating situations involve one partner changing the nature of the relationship, such situations are likely to draw more conscious attention, and consequently, more conscious strategic responses, than routine situations in which individuals reproduce the status quo. Thus, to examine strategic, intentional workplace friendship maintenance, we examined maintenance in escalating and deteriorating situations only.

Although research has examined relationship maintenance in certain types of nonwork relationships such as romantic relationships (e.g., Dainton, 2007; Simon & Baxter, 1993) and nonwork friendships (e.g., Bippus & Rollin, 2003), workplace relationships are unique because they are defined, in large part, by an instrumental or utilitarian function; that is, coworkers rely on each other to varying extents to do their jobs and, like it or not, spend much of their waking hours in each others’ presence at work (Sias, 2009). Thus, we turned to research on workplace relationship maintenance for guidance. Lee (1998) and Lee and Jablin (1995) provided a particularly useful starting point because they identified several tactics for employees use to maintain workplace relationships. Although their studies focused on supervisor-subordinate relationships that were not necessarily friendships, many of the tactics they identified are likely applicable to peer workplace friendships.

According to Lee and Jablin (1995), tactic use varies depending on the type of maintenance situation. In escalating situations, employees attempt to prevent the relationship from growing closer via four different tactics. Avoidance of interaction refers to intentional attempts to avoid the relationship partner (e.g., avoiding meetings the partner was expected to attend, not responding when the partner initiates a conversation). Indirect conversational refocus refers to deliberately but indirectly refocusing the topic of a conversation away from personal topics and toward work-related concerns. Direct conversational refocus refers to explicitly and directly refocusing a conversation toward work-related issues (e.g., specifically stating she or he does not want to discuss personal issues). Openness refers to direct and explicit discussion of the matter. These tactics function largely to maintain the existing nature of the boundary between the work and personal spheres. As workplace friendship is defined, in part, by a personalistic focus, such tactics are also likely to be used frequently in maintaining workplace friendships. An employee can maintain the extent of that personalistic focus by restricting the extent to which she or he discusses personal topics with the relationship partner.

Employees in deteriorating situations tend to rely on five primary tactics to prevent the partner from disengaging from the friendship (Lee, 1998; Lee & Jablin, 1995). Openness refers to direct statements of concern about issues that threaten the relationship or about the relationship’s trajectory. Creating closeness refers to tactics by which an individual intentionally tries to emphasize the personal, rather than professional relationship (e.g., asking the partner about his or her personal life, initiating informal conversations). Circumspectiveness refers to being generally cautious in how one communicates with the relationship partner (e.g., avoiding topics that the partner is uncomfortable discussing). Deception/distortion refers to intentionally misrepresenting or withholding information from the relationship partner (e.g., withholding bad or negative news). Finally, people engage in self-promotion when they emphasize their successes on the job or how hard they work. Again, given friendship’s personalistic focus, openness and creating closeness may be particularly effective in preventing peer friendship deterioration because they maintain the relationship’s focus on partners’ personal lives. Deception and circumspectiveness could also function to prevent an employee’s coworker from obtaining negative information that could damage the relationship and propel it toward deterioration.

Lee (1998) and Lee and Jablin (1995) focused solely on supervisor-subordinate relationships, however, not peer workplace friendships. Accordingly, our first study examined the following research question:

Research Question 1 (RQ1): What tactics do employees use to maintain peer workplace friendships at a desired level of closeness?

Method

Sample

A convenience sample of employees participated in two focus groups of five members each (two men, eight women). Participants were chosen because of their status as subject matter experts (SMEs), or “lay persons who have specific knowledge about the construct you are interested in studying” (Kline, 2005, p. 10). The SMEs in this case were adults with work experience who could identify and provide feedback on the relevance and realistic nature of the items and scenarios regarding peer friendship maintenance. Participants worked in several organizations in the Northwest United States and ranged from 20 to 55 years of age, M = 26.9.

Procedures

Study 1’s primary goal was to develop a set of maintenance tactic items relevant to peer workplace friendships. Using Lee and Jablin’s (1995) items as a starting point, we asked focus group members to discuss each item and, based on their experience, assess its relevance to peer workplace friendships. We then asked the focus groups to generate additional tactics not identified by Lee and Jablin (1995) but that would be relevant to peer workplace friendship maintenance. A secondary goal was to develop peer workplace friendship maintenance scenarios referencing a fictional workplace friendship between characters “Sam” and “Chris” for use in subsequent studies reported later in this manuscript.¹ Participants evaluated two maintenance scenarios—one describing a deteriorating situation and one describing an escalating situation—with respect to readability, clarity, and validity. Each group interview lasted 1 hr and was conducted at a university research facility.

Results

Focus group participants identified several Lee and Jablin (1995) relationship maintenance items that were inappropriate for examining peer friendships and also generated new items that represented realistic peer coworker behavior. Figure 1 details the maintenance tactic items removed, revised, and added based on focus group discussions. Overall, we discarded 22 of Lee and Jablin’s (1995) items; revised the wording of 11 of Lee and Jablin’s items, and added 18 new items generated by the focus groups. The final set included 40 items—20 related to escalating situations and 20 related to deteriorating situations.

Figure 1.

Original and added maintenance tactic items

The focus groups suggested clarifying the level of friendship that was supposed to exist between the scenarios’ two fictitious coworkers. Accordingly, we defined their relationship as a collegial peer relationship. We chose the collegial peer relationship because, unlike the information peer relationship, it is characterized by friendship and, unlike the special peer relationship; it has room to develop into a closer friendship. As noted above, the collegial peer relationship is characterized by moderate levels of trust and self-disclosure, sharing of information, support, and feedback, and discussion of nonwork topics. These details were added to the scenarios to clarify the nature of the friendship in question.

Study 2: Politeness and Peer Workplace Friendship Maintenance

Study 1 developed a set of peer workplace friendship maintenance tactic items, but generated no insights into why individuals may prefer some tactics over others to maintain their relationships. Politeness theory (Brown & Levinson, 1987) provides some guidance for speculation regarding why employees may choose particular tactics to maintain a workplace friendship. As coworkers must work together and interact with one another on, typically, a daily basis, some level of collegiality, civility, and cordiality in their relationships is important (Fritz, 2009). Concerns about politeness are, therefore, likely relevant to peer workplace friendship maintenance. Politeness theory centers on Goffman’s (1967) concept of face, or an individual’s public self-image. Face is a fluid, rather than stable, phenomenon that is constituted and must be continually attended to interaction (Sias, Smith, & Avdeyeva, 2003). Individuals consider both their own face and their conversation partner’s face when choosing communication strategies that they hope will enable them to achieve their message goals (Brown & Levinson, 1987).

Politeness theory considers two types of face. Positive face relates to a positive self-image based on the need to be liked by others. Negative face refers to an image of oneself as an autonomous individual based on the need to remain free from imposition. Although positive and negative faces are likely to be concerns in workplace friendship maintenance processes, to reasonably limit the scope of this study, we focused on positive face only. People who attempt to maintain a relationship are clearly concerned about damaging the relationship and those concerns will likely influence tactic choice. According to face management theory (Lim, 1994), positive face threats are likely to be more strongly associated with interpersonal relationship damage than are negative face threats because participating in a relationship requires friends to lose some autonomy and develop tolerance for criticism from one another. Consistent with this, Cupach and Carson (2002) found positive face threats in complaint messages resulted in significantly more damage to relationships than did negative face threats. In general, messages that consider the other’s feelings and desires reflect a concern for positive face; messages that show little concern for the conversation partner’s feeling or that express dislike or disapproval reflect threats to positive face (Brown & Levinson, 1987).

Scholars have not examined the extent to which workplace relationship maintenance tactics are perceived to be face-threatening. The tactics likely vary in their concern for the face of both relationship partners; however, to reasonably limit the scope of the present project, our study examined positive face threat to the target of the maintenance tactic (i.e., the extent to which a tactic presents a threat to the coworker’s positive self-image). In escalating situations, direct conversational refocus, for example, requires the speaker to express his or her desire to avoid discussion of personal topics at increasing levels, which could be interpreted by the coworker as an expression of dislike or disrespect. That tactic would, therefore, likely be more face-threatening to the hearer than the more subtle indirect conversational refocus strategy. Openness could be similarly interpreted because the coworker may interpret speaker’s statements of concern regarding increased closeness as a statement of dislike or disrespect. Avoidance of interaction may be considered as a face-threatening act because the avoidance of contact indicates dislike for the coworker. In contrast, indirect conversational refocus and circumspectiveness convey concern for the other’s feelings via subtle ways of limiting the amount of personal conversation and are, therefore, likely less face-threatening to the coworker.

Maintenance in deteriorating situations may be generally less face-threatening to the recipient because the tactics generally communicate liking and affinity for the recipient. The tactics may vary, however, with respect to politeness/positive face. Openness, for example, involves direct questioning of the coworker’s motives and behaviors, which could be interpreted as disapproval. Creating closeness, in contrast, communicates liking for the coworker, and is likely not perceived as face-threatening/impolite. Although dishonest, deception/distortion represents little face threat to the coworker because withholding negative information is not explicitly insulting or hurtful. Similarly, circumspectiveness is a subtle strategy designed to avoid interaction about topics the partner is uncomfortable discussing and designed to save face, not threaten it. To address these possibilities, we examined the following:

Research Question 2 (RQ2): What is the relative perceived politeness of peer workplace friendship maintenance tactics in deteriorating and escalating situations?

Method

Sample

Students in undergraduate public speaking classes at a midsized public university completed a questionnaire and were given extra credit for participating in the study. The majority (94%) reported that they had work experience. The questionnaire asked participants to read a peer workplace friendship scenario and rate the politeness/positive face threat of various friendship maintenance tactic items. A total of 165 people completed the surveys. Of those, eight were discarded from the study because their surveys were either incomplete or demonstrated an obvious response bias (e.g., giving the same score on every item throughout the survey). The final sample included 157 participants, 79 responded to a deterioration scenario, and 78 responded to an escalation scenario. The sample was evenly represented by males (n = 79) and females (n = 78). The average age of participants was 20 years of age (range = 18-33).

Measures

Participants read a scenario reflecting either a deteriorating or escalating workplace friendship maintenance situation involving a collegial peer relationship.² After reading the scenario, they rated a series of workplace friendship maintenance tactic items on eight positive face threat items derived from Cupach and Carson’s (2002) instrument designed specifically to measure recipient face threat—the focus of this study. Participants indicated on a 7-point Likert-type scale the extent to which they agreed that each maintenance item was: polite, rude, insensitive, disrespectful toward Chris, justified, hostile, contemptuous toward Chris, and tactful. Some items were reverse-coded to prevent response bias and for the resulting scores, a higher score represents more perceived politeness. The items were derived from the focus group study reported above. The final instruments included eight politeness items and 20 maintenance tactic items. We used two survey versions for each scenario condition, changing item order to control for potential order effects.

Analysis

Data analysis entailed a multistep process. We first conducted principal components analysis with a varimax rotation of the eight politeness items for each of the 20 maintenance tactic items (a total of 20 principal components analyses). Two politeness items failed to load consistently for each of the 20 maintenance items (hostile and tact). Removing them resulted in single-factor politeness solutions for all 20 maintenance tactics. We next created maintenance tactic items based on the politeness scales; that is, for each maintenance tactic item which originally had eight politeness ratings each, we created a single variable based on the average of the six surviving politeness items. This resulted in 20 maintenance tactic politeness items, which we submitted to a principal components analysis. We used a 50/40 criterion for creating factors; and also considered conceptual relevance when the 50/40 criterion was unclear.

Deteriorating situation

A principal components analysis followed by varimax rotation of maintenance tactic items in the deterioration scenario resulted in a five-factor solution explaining 76% of the common variance. Four items did not load onto any of the five factors. A subsequent analysis excluding these four items produced a four-factor solution accounting for 60% of the common variance (see Table 1). The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was .82, indicating the sample was adequate for factor analysis. Factor 1, Creating Closeness (variance explained = 24%), included five items reflecting attempts to maintain personal ties and closeness, α = .84. Factor 2, Openness, (variance explained = 14%), included five items reflecting an individual’s open and direct discussion of the relationship, α = .86. Factor 3, Circumspectiveness (variance explained = 12%), included three items reflecting an individual’s attempt to edit his or her behavior (e.g., avoiding negativity) and conversational topics, α = .74. Factor 4, Deception (variance explained = 10%), included three items via which an individual misrepresents or betrays the coworker, α = .64.³ In sum, 16 of the original 20 tactic items were retained. The dimensions are identical to those identified by Lee and Jablin (1995).

Table 1.

Maintenance Tactic Politeness Final Factor Solution—Deteriorating Situations

Item	Factor 1	Factor 2	Factor 3	Factor 4
Avoid saying anything that might embarrass Chris	.77*	.00	.11	.24
Offers willingness to help out	.62*	.31	.36	−.17
Has informal conversations with Chris to see how things are going	.77*	.32	.21	−.17
Talks about past experiences he and Chris have shared	.72*	.33	.28	−.16
Makes an effort to strike up conversations with Chris	.81*	.17	.12	−.08
Confronted Chris about the deterioration of their friendship	.08	.81*	.09	.22
Gave examples of how and why he was upset about the level of their relationship	.18	.77*	.27	−.07
Asked Chris if there was something wrong with their friendship	.06	.78*	.25	.17
Asked Chris if there were any issues going on between them that he wasn’t aware of	.30	.70*	.31	−.02
Directly asked Chris if a personal matter was affecting their friendship	.34	.69*	−.18	.29
Always tries to be in a good mood around Chris	.09	.21	.86*	.07
Avoids being negative when he’s around Chris and instead always tried to be in a good mood around him	.36	.18	.74*	.03
Doesn’t discuss topics that he knows Chris is sensitive	.41	.14	.60*	.07
Overexaggerates his friendships with other coworkers to try and make Chris jealous	.08	.11	.44	.68*
Pretends there is no problem with his friendship with Chris	−.05	.12	−.01	.86*
Releases “steam” about their relationship by talking about Chris behind Chris’ back	−.47	.28	−.12	.60*

Escalating situation

Principal components analysis followed by a varimax rotation of maintenance tactic items in escalating situations resulted in a five-factor solution explaining 75% of the variance. However, two factors were comprised of only one item. These items were removed from the data set and a second analysis produced a three-factor solution (see Table 2) explaining 55% of the common variance (KMO = .77). Factor 1, Openness/Direct Conversational Refocus (variance explained = 27%), included six items reflecting open and direct discussion of the relationship and/or stated desire to limit conversation to specific topics, α = .87. Factor 2, Avoidance (variance explained = 14%), included four items reflecting behaviors directed at avoiding contact and interaction with the coworker, α = .74. Factor 3, Indirect Conversational Refocus (variance explained = 14%), included five items reflecting indirect attempts to refocus conversation, α = .79. In sum, 15 of the 20 original maintenance tactic items were retained. The dimensions are identical to those identified by Lee and Jablin (1995), with the exception that their openness and direct conversational refocus are combined into one factor in our study.

Table 2.

Maintenance Tactic Politeness Final Factor Solution—Escalating Situations

Item	Factor 1	Factor 2	Factor 3
Tells Chris that he does not want to discuss either of their social activities in greater detail than they usually do	.79*	.14	.08
Tells Chris that he does not want to discuss either of their personal lives in more detail than they usually do	.82*	.25	.09
Openly told Chris that he was uncomfortable with increasing the level of closeness in their relationship	.86*	.05	.20
Expressed his feelings about the depth of their relationship in a forthright manner	.60*	−.18	.39
Provided Chris with examples of how the friendship was getting too close	.76*	.10	.15
Tell Chris not to talk so much about him in the workplace	.65*	.18	.06
Plans his work schedule so as not to encounter Chris	.16	.85*	−.11
Wanders into another location when Chris shows up in the vicinity of his work station	.15	.68*	.16
Does not communicate with Chris unless Chris initiates the conversation	−.05	.78*	.06
When Chris wants to talk about topics, they don’t typically discuss, Sam makes up excuses for avoiding the discussion	.33	.57*	.30
Tries to focus his conversation with Chris on work-related issues	.41	.40	.48*
Discourages Chris from talking to him about personal matters they’ve never discussed before	.54	.25	.60*
Tells Chris that he is not aware of the subject/topic of conversation	−.04	−.03	.74*
Changes the topic whenever Chris tries to discuss a personal matter that they don’t typically talk about	.46	.32	.46*
Says “this isn’t something I feel like discussing” when Chris brings up new topics unrelated to work	.42	.08	.62*

We examined RQ2 using a series of paired t tests comparing the tactics with one another. Tables 3 and 4 present the mean scores, range, and correlations among the maintenance tactics.

Table 3.

Descriptive and Correlations for Politeness Scores—Deteriorating Situations

Variable	Range	M	SD	1	2	3
1. Circumspectiveness	3.67-7.00	6.10	0.85	—
2. Creating closeness	3.90-7.00	6.21	0.71	0.56*	—
3. Openness	2.40-6.95	5.10	0.97	0.46*	0.50*	—
4. Deception	1.00-6.50	2.80	0.92	0.18	−0.10	0.31*

p < .01.

Table 4.

Descriptive Statistics and Correlations for Politeness Scores—Escalating Situations

Variable	Range	M	SD	1	2
1. Openness/direct conv. Refocus	2.23-6.97	4.47	1.02	—
2. Avoidance	2.00-6.85	3.44	0.95	0.31*	—
3. Indirect conv. refocus	2.76-7.00	4.71	0.98	0.72*	0.46*

p < .01.

In deteriorating situations, circumspectiveness was perceived as more polite than openness, t = 9.05, d = 1.04, p < .00, and more polite than deception, t = 25.39, d = 2.91, p < .00. Creating closeness was perceived as more polite than openness, t = 11.71, d = 1.36, p < .00, and more polite than deception, t = 24.24, d = 2.80, p < .00. Openness was perceived as more polite than deception, t = 17.97, d = 2.05, p < .00. Circumspectiveness and creating closeness did not differ significantly, t = −.47, d = .17, p > .15. In sum, participants perceived circumspectiveness and creating closeness as the most polite strategies, followed by openness and deception.

In escalating situations, indirect conversational refocus was perceived as more polite than openness/direct conversational refocus, t = −2.58, d= .31, p < .01, and more polite than avoidance, t = −10.41, d = 1.22, p < .00. Openness/direct conversational refocus was also perceived as more polite than avoidance, t = 7.36, d = .88, p < .00. In sum, participants perceived indirect conversational refocus as the most polite maintenance tactic, followed by openness/direct conversational refocus and avoidance, respectively.

Study 3: Politeness and Task Interdependence

Study 3 builds on studies one and two by assessing the factor structure of the tactic items with respect to likelihood of use, and examining the extent to which perceived politeness, task interdependence, and attachment style predict tactic likelihood of use.

Perceived Politeness and Tactic Use

As noted earlier, coworkers must work together and typically interact with one another on a daily basis. Some level of collegiality, civility, and cordiality in their relationships is, therefore, important (Fritz, 2009). In addition, friendship is characterized by liking and concern for the other as a whole person, not simply a role occupant. Employees are, therefore, likely to be concerned with politeness as they attempt to maintain workplace friendships. Based on politeness theory, and the results of Study 2, we posit the following:

Hypothesis 1a (H1a): In deteriorating situations employees will be more likely to use circumspectiveness and creating closeness than openness and deception to maintain workplace friendships.

Hypothesis 1b (H1b): In deteriorating situations employees will be more likely to use openness than deception to maintain workplace friendships.

Hypothesis 2a (H2a): In escalating situations employees will be more likely to use indirect conversational refocus than open/direct conversational refocus and avoidance to maintain workplace friendships.

Hypothesis 2b (H2b): In escalating situations employees will be more likely to use openness/direct conversational refocus than avoidance to maintain workplace friendships.

As Study 2 revealed no significant difference in the perceived politeness of circumspectiveness and openness, we posed the following research question:

RQ2: What is the relative likelihood that an employee will use circumspectiveness and openness to maintain a workplace friendship in a deteriorating situation?

Task interdependence

Workplace friendships are defined, in part, by the context in which they exist—the workplace—and this context plays an important role in workplace friendship processes (Sias & Cahill, 1998). Along these lines, Waldron (2003) noted that employees’ task characteristics are strongly associated with employee communication and employee relationships. A particularly important task characteristic with respect to friendship maintenance may be the extent to which the relationship partners work closely with one another and depend on each other to do their jobs. Although coworkers’ tasks are all somewhat interdependent in organizational systems (Weick, 1995), the level of interdependence varies across contexts. Task interdependence generally refers to the “the degree to which work flows from one job to one or more other jobs” (Kiggundu, 1983, p. 501). Applying this concept to coworkers, task interdependence refers to the extent to which the work of one individual flows to the work of another. Coworkers with highly interdependent tasks depend on one another for material and other forms of support, such as help, to complete their tasks and do their jobs well (Tagger & Haines, 2006). Given the importance of competent job performance, coworkers whose tasks are highly interdependent will likely be very concerned about maintaining effective relationships with one another so that they may minimize any potential threat to their job performance. Accordingly, individuals seeking to maintain a friendship with a coworker with whom they work closely and interdependently may show a greater preference for using less face-threatening tactics than employees with lower levels of task interdependence in both deteriorating and escalating situations. Thus, we posit task interdependence will predict workplace friendship maintenance tactic use as follows:

Hypothesis 3a (H3a): Individuals with high levels of task interdependence with their coworker/friend will be more likely to use circumspectiveness and creating closeness than will individuals with low levels of task interdependence to prevent their workplace friendship from deteriorating.

Hypothesis 3b (H3b): Individuals with low levels of task interdependence with their coworker/friend will be more likely to use openness and deception/distortion than will individuals with high levels of task interdependence to prevent their workplace friendships from deteriorating.

Hypothesis 4a (H4a): Individuals with high levels of task interdependence with the coworker/friend will be more likely to use indirect conversational refocus than will individuals with low levels of task interdependence to prevent their workplace friendship from escalating.

Hypothesis 4b (H4b): Individuals with low levels of task interdependence with the coworker/friend will be more likely to use avoidance than will individuals with high levels of task interdependence to prevent their workplace friendship from escalating.

Although Study 2 indicated that avoidance was perceived as less polite than openness/direct conversational refocus, it is unclear if the different perceptions were strong enough to be associated with task interdependence. Accordingly, we examined the following research question in lieu of a specific hypothesis:

Research Question 3 (RQ3): How is task interdependence associated with the likelihood an employee will use openness/direct conversational refocus and avoidance to prevent their workplace friendship from escalating?

Attachment Style

In addition, to contextual influences, research indicates individual characteristics, such as attachment style, are also associated with relationship maintenance. According to attachment theory, individuals vary in their approaches to interpersonal relationships (Bowlby, 1969; Hazan & Shaver, 1987). From infancy, people develop perceptions of themselves and others who provide a foundation for interpersonal interaction and relationships (Bartholomew, 1993; Bowlby, 1969). As Guerrero and Bachman (2006, p. 343) explained, “attachment theory is more than a description of personality types; the theory provides an account of how people develop cognitive schema that guide perceptions and social behavior.” Research indicates individuals tend to ascribe to one of four primary attachment styles that vary on two dimensions—anxiety and avoidance. Secures represent low levels of anxiety (thus, high self-confidence) and an approach orientation toward others and toward relationships. They are comfortable with both closeness to others and independence. Dismissives are also self-confident (thus, low in anxiety), but have an avoidance orientation toward others and relationships. They tend to avoid intimacy and highly value their independence. Preoccupieds, have low self-confidence and a positive view of others, thus they exhibit an approach orientation to relationships. These individuals are preoccupied with relationships and worry about being abandoned by others. Fearfuls tend to have a negative view of themselves and of others. They want and value interpersonal relationships, but they do not trust others and fear intimacy and commitment (Bartholomew & Horowitz, 1991).

These attachment styles are associated with a variety of interpersonal relationship processes and dynamics, including interpersonal relationship maintenance. For example, studies of romantic relationships suggest that Secures and Preoccupieds tend to rely more on prosocial maintenance behaviors (Simon & Baxter, 1993), while Dismissives and Fearfuls are more likely to rely on avoidance strategies (Guerrero & Bachman, 2006). Research has not examined attachment style with respect to workplace friendship maintenance. It is likely, however that people with high levels of self-confidence (Secures) or who are highly concerned with maintaining connection and commitment to others (Preoccupieds) may be more likely to engage in direct and open discussion of their relationships than those with low self confidence (Fearfuls) or who are less concerned with their relationship partner (Dismissives). In addition, people with an avoidance orientation toward relationships (Dismissives and Fearfuls) may be less likely to use tactics that emphasize closeness or personal connection than those with an approach orientation toward relationships (Secures and Preoccupieds).Given the lack of research in this area, we examined the following research question in lieu of specific hypotheses:

Research Question 4 (RQ4): How is attachment style associated with workplace friendship maintenance tactic likelihood of use in deteriorating and escalating situations?

Method

Overview

Data were obtained via a web-based survey that asked participants to respond to hypothetical scenarios (developed in studies one and two) of workplace friendship maintenance in deteriorating and escalating situations.⁴

Sample

We distributed the survey to 398 adults employed full-time at a variety of organizations (192—escalating situations, 206—deteriorating situations). Participants were recruited by students in a large upper-division undergraduate class at a midsized public university. After securing permission from participants, the students provided participants’ email addresses to the senior researcher who contacted participants directly, confirmed their willingness to participate, and forwarded them the survey link via email. Students were given extra class credit for participant recruitment. Of the 398 participants contacted, 265 completed the survey, for a final response rate of 67%.⁵ One hundred and eleven (42%) participants were male and 137 (51%) were female. Seventeen respondents did not report their sex. Most participants (179; 67%) were between 41 and 60 years of age, White (91%), and had worked at their current organization for an average of 10.5 years (range 2 weeks-42 years). The sample represented a wide variety of occupations and industries including lawyer, military infantry soldier, secretary, teacher, pilot, legal assistant, carpenter, stock broker, psychologist, and newspaper editor, to name a few.

Measures

Task interdependence was manipulated in the scenarios using Tagger and Haines’ (2006) conceptualization. The high interdependence scenario specified that coworkers A” and “B” “. . . have worked together for about one year. Their jobs require that they work closely together and provide material help, information, advice, and support to one another in order to do their work effectively.” The low interdependence scenario specified that “A” and “B” “. . . have worked together for about a year. They work largely independently of one another and do not depend on each other to get their work done.” Taggar and Haines’ (2006) four-item measure of task interdependence appeared at the end of the survey as a manipulation check (α = .90 for both deteriorating and escalating situations). Participants rated the extent to which they agreed “A’s” and “B’s” working relationship was characterized by the four items on 7-point Likert-type scale. T tests indicated the manipulation was successful in both the deterioration condition, t = 7.10, d = 1.24, p < .00; M high condition = 5.21; M low condition = 3.70, and the escalation condition, t = 6.16, d = 1.05, p < .00; M high condition = 5.06; M low condition = 3.64.

Attachment style

Consistent with Guerrero and Bartholomew (2006), we assessed attachment style with Bartholomew and Horowitz’s (1991) self-report attachment style prototype measure. Respondents chose which of four descriptions (representing the four attachment styles) best described them. In the deteriorating situation sample, 56 (45%) respondents identified as Secures, 47 as Dismissives (38%), seven as Preoccupieds (6%), and 15 as Fearfuls (12%). In the escalating situation sample, 61 respondents identified as Secures (49%), 52 as Dismissives (42%), two as Preoccupieds (2%), and 10 as Fearfuls (8%).

Maintenance tactic choice was assessed using the items developed in studies one and two. In this study, however, after reading the scenario, respondents rated the likelihood that a person in “A’s” place (other condition) would use the various tactics to maintain the friendship and prevent it from growing closer/more distant on a 7-point scale with higher scores indicating increased likelihood. Respondents then imagined themselves in “A’s” place (self condition) and rated the likelihood that they would use the various maintenance tactics in such a situation. We used scenarios because Study 2 demonstrated that the maintenance tactics vary with respect to politeness. Social desirability bias was, therefore, a concern such that respondents would be hesitant to report that they had actually used impolite or face-threatening tactics. Our scenario method is consistent with other research using hypothetical scenarios to study sensitive issues such as facework and politeness (e.g., Feng & Burleson, 2008; Miller & Roloff, 2008; Wilson, Kunkel, Robson, Olufowote, & Soliz, 2008). Using scenarios also enabled us to include self (direct) and other (indirect) conditions to more fully address the social desirability issue. Prior studies examining socially sensitive issues have incorporated direct and indirect/projective questioning methods and shown that they can reveal such a bias (Cruz, Shafer, & Strawser, 2000; Sias & Perry, 2004). Similar responses for the two conditions indicate lack of a response bias. Different responses indicate potentially biased responses in the direct (self) condition. In such cases, responses for the indirect (other) condition are considered as more valid because individuals are more likely to project their beliefs and evaluations in an indirect condition when social desirability bias is an issue (Fisher, 1999).

Although scenarios helped to maintain internal validity, they introduced a potential threat to external or ecological validity if the scenarios were not adequately realistic. To assess this possibility, respondents rated the realism of the scenario via an item at the end of the instrument: “On a scale from 1 to 7, how realistic is the hypothetical situation described above; that is, how likely is it that a similar situation would really occur in a workplace?” with one representing extremely unrealistic and seven representing extremely realistic. The mean realism rating was 5.44 for escalation scenario and 5.40 for the deterioration scenario, indicating the scenarios represented realistic phenomena to the respondents. These results, along with the fact that the scenarios were reviewed and revised in Study 1 by subject matter experts based on their personal work experience, constitute evidence of external validity.

Maintenance tactic factor structure

We conducted principal components analyses (using Varimax rotation) on the maintenance tactic items in both deteriorating and escalating situations to identify the underlying factor structures. Analyses were conducted for both the “other” and “self” conditions. Results for both conditions indicated virtually identical factor structures for the other and self conditions.⁶ As of the potential for a social desirability bias in the “self” condition, we chose the factor solution for the “other” condition as the basis of scale construction for both conditions. We used a 50/40 criterion to create factors, and considered conceptual relevance when the 50/40 criterion was unclear.

For deteriorating situations, initial analysis resulted in a six-factor solution explaining 63% of the common variance. Three items did not load clearly on any factor and a second analysis excluding those items resulted in a four-factor solution (see Table 5) explaining 55% of the variance (KMO = .72). Factor 1, Openness (variance explained = 20%), included four items reflecting open and direct discussion of the relationship, α = .83 for other and .93 for self. Factor 2, Creating Closeness (variance explained = 14%), included four items reflecting an individual’s attempts to maintain personal ties and closeness, α = .70 for other and .72 for self. Factor 3, Circumspectiveness (variance explained = 11%), included four items reflecting an individual’s attempt to edit his or her behavior (e.g., avoiding negativity) and conversational topics, α = .72 for other and .71 for self. These three factors are identical to those in Study 2 grouping items by their perceived politeness. The item loadings are also identical to those of Study 2, with two exceptions. Specifically, one item that loaded on creating closeness in Study 2, loaded on circumspectiveness here, and one item that loaded on openness in Study 2, did not load on any factor. Factor 4, Deception/Distortion (variance explained = 10%) included three items via which an individual misrepresents information or betrays the coworker, α = .62 for other and .46 for self. However, the factor loadings for deception/distortion based on likelihood of use were substantially different from those in Study 2; specifically, two deception items from Study 2 did not load on any factor in this study. As of this, and the scale’s low reliability coefficients, this factor was not included in hypothesis testing.

Table 5.

Friendship Maintenance Tactic Use Final Factor Solution—Deteriorating Situations

Item	Factor 1	Factor 2	Factor 3	Factor 4
Confront “B” about deterioration of friendship	.70*	.10	.01	.04
Directly ask “B” if a personal matter is affecting their friendship	.86*	.04	−.11	−.03
Ask “B” if there is something wrong with their friendship	.89*	−.03	−.09	−.08
Ask “B” if there are any issues going on between them that she or he is unaware of	.79*	.06	.00	−.15
Offer willingness to help out	.11	.74*	.10	.05
Have informal conversations with “B” to see how things are going	.09	.56*	.20	−.36
Talk about past experiences shared with “B”	.11	.66*	.10	.03
Make an effort to strike up conversations with “B”	−.05	.72*	.18	−.02
Avoid saying anything that might embarrass “B”	−.14	.10	.64*	−.15
Avoid being negative around “B” and instead always try to be in a good mood around him or her	−.10	.17	.63*	.03
Always try to be in a good mood around “BA”	.01	.38	.71*	.08
Don’t discuss topics she or he knows “B” is sensitive about	−.04	.01	.77*	.02
Release “steam” about the relationship by talking about “B” behind “B’s” back	.01	.09	−.10	.79*
No longer joke around with “B” so she or he doesn’t accidentally insult her or him	−.16	−.18	−.01	.74*
Withhold negative information from “B”	−.06	.20	.39	.63*

For escalating situations, analysis resulted in a five-factor solution accounting for 65% of the variance. However, one factor was comprised of only one item. That item was removed from the data set and a second analysis resulted in a four-factor solution (see Table 6) accounting for 62% of the common variance (KMO = .75). Factor 1, Avoidance (variance explained = 22%), included six items reflecting behaviors directed at avoiding contact and general avoidance of interaction with the coworker, α = .82 for other and .84 for self. Factor 2, Openness (variance explained = 18%), included four items reflecting open and direct discussion of the relationship, α = .85 for other and .90 for self. Factor 3, Direct Conversational Refocus (variance explained = 13%), included four items reflecting direct requests to avoid discussion of certain nonwork topics, α = .74 for other and .80 for self. Factor 4, Indirect Conversational Refocus (variance explained = 10%), included five items reflecting indirect attempts to refocus conversation on work-related topics, α = .68 for other and .67 for self. In sum, 19 of the 20 original maintenance tactic items were retained. The factors are very similar to the politeness factors identified in Study 2 with a few differences. The avoidance dimension in this study included an item that was excluded in Study 2, and one that previously loaded on the indirect conversational refocus dimension. The indirect conversational refocus dimension included an item that was excluded in Study 2; otherwise it is identical to Study 2’s version. Finally, in Study 2, the openness and direct conversational refocus items loaded together on a single item; in this study, and consistent with Lee and Jablin (1995), they loaded on separate dimensions.

Table 6.

Study 3 Friendship Maintenance Tactic Use Final Factor Solution—Escalating Situations

Item	Factor 1	Factor 2	Factor 3	Factor 4
Act like she or he doesn’t know what “B” is talking about even though she or he does	.66*	−.06	−.03	.16
Plans his or her work schedule so as not to encounter “B”	.80*	.00	−.07	−.11
When “B” wants to talk about topics, they don’t typically discuss, make up excuses for avoiding the discussion	.74*	−.16	.05	.17
Wander into another location when “B” shows up in the vicinity of his or her work station	.87*	−.09	.06	−.06
Tell “B” that she or he is not aware of the subject/topic of conversation	.74*	−.05	.07	.12
Do not communicate with “B” unless “B” initiates the conversation	.77*	−.24	−.06	.04
Openly tell “B” that she or he is uncomfortable with increasing the level of closeness in their relationship	−.13	.88*	.09	.04
Expressed his or her feelings about the depth of their relationship in a forthright manner	−.24	.74*	.16	−.26
Provide “B” with examples of how the friendship is getting too close	−.11	.71*	.18	.16
Tell “B”, “I really appreciate our friendship, but I’m uncomfortable with it getting closer”	−.15	.85*	.07	.15
Tell “B” that she or he does not want to discuss either of their social activities in greater detail than they usually do	−.18	.23	.80*	.08
Tell “B” not to talk so much about him in the workplace	−.02	−.08	.83*	−.12
Simply say “none of your business” to personal questions “B” asks about topics they don’t normally discuss	.11	.34	.51*	.09
Tells “B” that she or he does not want to discuss either of their personal lives in more detail than they usually do	.16	.54	.60*	−.13
Does not initiate or voluntarily offer personal information about topics they don’t normally discuss	.07	−.08	−.10	.74*
Try to focus his or her conversation with “B” on work-related issues	.08	−.11	−.02	.67*
Says “this isn’t something I feel like discussing” when “B” brings up new topics unrelated to work	−.04	.43	.40	.47*

Social desirability analysis

To assess the presence of a social desirability response bias in the data, paired t tests compared with each respondent’s mean “self” maintenance strategy score with his or her “other” strategy score. For deteriorating situations, results showed significant differences in “self” and “other” scores for two of the three strategies (creating closeness and circumspectiveness). For escalating situations, results showed significant differences in “self” and “other” scores for two of the four strategies (avoidance and direct conversational refocus).⁷ As these results indicated a response bias, we used only data in which others were the referent for the remaining analyses.

Perceived politeness and tactic use

H1a predicted that in deteriorating situations, employees would be more likely to use circumspectiveness and creating closeness than openness and deception to maintain workplace friendships.

Paired t tests for differences in means (excluding deception scores) supported this hypothesis (see Table 7). Respondents reported that an employee would be more likely to use circumspectiveness, t = 8.19, d = .74, p < .00, and creating closeness, t = 10.4, d = .98, p < .00, than openness. H1b predicted that, in deteriorating situations, employees would be more likely to use openness than deception to maintain workplace friendships. We were unable to test this hypothesis because of the deception scale’s low reliability.

Table 7.

Descriptive and Correlations for Tactic Use Scores—Deteriorating Situations

Variable	Range	M	SD	1	2	3
1. Circumspectiveness	2.50-7.00	5.42	0.78	—
2. Creating closeness	3.80-6.80	5.45	0.60	0.35*	—
3. Openness	1.50-7.00	4.37	1.08	−0.11	0.21*	—
4. Deception	1.67-6.67	3.76	0.98	0.11	0.07	−0.20*

p < .01.

RQ2 asked about the relative likelihood, an employee would use creating closeness and circumspectiveness to maintain a workplace friendship in a deteriorating situation. A paired t test indicated no significant difference in the likelihood of use of those tactics.

H2a predicted that in escalating situations, employees would be more likely to use indirect conversational refocus than openness/direct conversational refocus and avoidance (see Table 8). As openness and direct conversational refocus loaded on separate factors in this study, we analyzed those dimensions separately. Paired t tests indicated support for H2a. Respondents reported employees would be more likely to use indirect conversational refocus than openness, t = 12.30, d = 1.13, p < .00, direct conversational refocus, t = 18.79, d = 1.69, p < .00, and avoidance, t = 10.22, d = .92, p < .00. H2b predicted that, in escalating situations, employees would be more likely to use openness than avoidance to maintain workplace friendships. This hypothesis was not supported. Contrary to expectations, respondents reported that employees would be more likely to use avoidance than openness, t = 2.80, d =.25, p < .01, and more likely to use avoidance than direct conversational refocus, t = 6.20, d = .56, p < .00. As openness/direct conversational refocus items loaded on separate factors in this study, we compared the mean scores of openness and direct conversational refocus, and found respondents reported employees would be more likely to use openness than direct conversational refocus, t = 4.19, d = .39, p < .00.

Table 8.

Descriptive Statistics and Correlations for Tactic Use Scores—Escalating Situations

Variable	Range	M	SD	1	2	3
1. Openness	1.00-6.75	3.80	1.20	—
2. Avoidance	1.50-7.00	4.25	1.12	−31*	—
3. Direct Conv. Refocus	1.00-6.25	3.35	1.09	0.44*	−0.03	—
4. Indirect Conv. Refocus	2.80-6.60	5.28	0.69	0.09	0.30*	0.25*

p < .01.

Task interdependence, attachment style, and tactic use

H3a-b, H4a-b, and RQ3 and RQ4 addressed the extent to which task interdependence and attachment style are associated with maintenance tactic choice. These were examined in a series of ANOVAs in which task condition (low and high) and attachment style (secure, dismissive, and fearful) were predictor variables and maintenance tactics were the dependent variables. As both samples had fewer than 10 respondents who identified as “Preoccupieds,” our analysis only considered Secures, Dismissives, and Fearfuls.

Deteriorating situations

Results for the creating closeness tactic indicated a significant overall model, F = 3.52, ω² = .15, p < .01, and a significant main effect for attachment style, F = 5.68, ω² = .10, p < .01. Post hoc Scheffe’s comparisons indicated Secures, M = 5.62, and Fearfuls, M = 5.71, were more likely to use creating closeness than were Dismissives, M = 5.23. Results also indicated a significant main effect for task interdependence, F = 5.45, ω² = .05, p < .02. Respondents in the low interdependence condition were more likely to use creating closeness, M = 5.68, than were those in the high condition. The overall model for circumspectiveness was not significant, F = 1.10, ω² = .05, p < .37. However, results indicated a significant main effect for task interdependence, F = 4.65, ω² = .04, p < .04, with respondents in the low interdependence condition, M = 5.66, reporting employees were more likely to use circumspectiveness than did those in the high condition, M = 5.31. No significant results were obtained with respect to the openness tactic.

Escalating situations

Task interdependence and attachment style did not predict the avoidance, openness, and indirect conversational refocus tactics. Although the overall model for direct conversational refocus was not significant, results indicated a main effect for task interdependence, F = 4.28, ω² = .04, p < .05. Specifically, higher tactic use scores were observed in the high task interdependence condition, M = 3.64, than in the low condition, 3.09.

Discussion

Research has addressed how and why peer workplace friendships develop (Sias & Cahill, 1998) and deteriorate (Sias et al., 2004), but ignored their maintenance. Our studies, therefore, address an important void in the literature. Participants in Study 1 identified four primary tactics employees use to maintain peer friendships at a desired state in escalating situations—avoidance, indirect conversational refocus, direct conversational refocus, and openness—and four primary maintenance tactics relevant to deteriorating situations—creating closeness, circumspectiveness, deception, and openness. All these tactics involve, in part, the speaker attempting to communicatively manage the boundary between the personal and work spheres. Such a focus is particularly relevant to the maintenance of workplace friendships because of friendship’s personalistic focus; that is, friends’ recognition and treatment of one another as “whole persons” not just work role occupants. Employees concerned that their friendship with a coworker is growing too close communicate in ways that strengthen the personal/work boundary. Those concerned that their friendship with a coworker is deteriorating communicate in ways that weaken or transcend the personal/work boundary. These findings are consistent with other research that indicates communicative transformation of personal/work boundaries is a primary element in workplace friendship dynamics (Sias & Cahill, 1998; Sias et al., 2004).

Study 2 built on Study 1 by providing insights into the perceived politeness of the strategies identified in that study. Specifically, in escalating situations, indirect conversational refocus and openness/direct conversational refocus were perceived as significantly more polite than avoidance. In deteriorating situations, creating closeness and circumspectiveness were perceived as the most polite maintenance strategies, followed by openness and deception, respectively. These results indicate that, consistent with politeness theory, the more subtle or indirect the interaction, the more likely people were to perceive it as polite instead of a face threat. In addition, tactics involving interaction with the coworker (e.g., indirect conversational refocus, open/direct conversation, circumspectiveness, and creating closeness) display concern for the coworker’s positive face. In contrast, tactics that occur without interaction or contact with the coworker (e.g., avoidance, deception) are perceived as less polite and more face-threatening. This is largely consistent with politeness theory, with the exception of avoidance—a tactic typically considered polite (Brown & Levinson, 1987). The constraints of the workplace context may explain this difference. As Shea and Pearson (1986) noted, the range of permissible communication behaviors in the workplace context is narrower than the range in other nonwork relationship contexts. Along these lines, while one may be able to avoid a nonwork friend, avoiding a coworker, especially one with whom you work in close proximity and/or with whom you share interdependent tasks, is very difficult. In such cases, the act of avoidance is likely unusual and highly noticeable. Such noticeably different behavior would likely send a clear, and face-threatening message to the receiver. In sum, then, acknowledging, rather than ignoring, a coworker communicates respect for the individual and the relationship.

With the exception of deception, the factor structures of the tactics when rated for politeness were virtually identical to the factor structures of the tactics when rated for likelihood of use. Moreover, the dimensions underlying the maintenance tactic politeness items are largely similar to those developed by Lee and Jablin (1995). These results provide two important contributions to the literature. First, they indicate that politeness is a fundamental element underlying workplace relationship maintenance and should be considered in theory and research addressing various types of workplace relationship dynamics. Second, the consistency between our tactics and Lee and Jablin’s (1995) supervisor-subordinate maintenance tactics indicates that employees approach workplace relationship maintenance in generally the same way regardless of the hierarchical nature of the relationship (i.e., supervisor-subordinate or peer). Study 1 resulted in removal of some more clearly “hierarchical” tactics, but in general the tactics used for supervisor-subordinate relationships are similar to those in peer friendships and, importantly, are fundamentally organized around concerns for politeness and face.

Having identified maintenance tactics and the perceived politeness of those tactics, Study 3 provided insights into predictors of maintenance tactic choice. The strongest and most consistent predictor of the likelihood of using tactics in both deteriorating and escalating situations was perceived politeness. Respondents showed a clear preference for tactics rated higher in perceived politeness. Moreover, the effect sizes (Cohen’s d) were quite high, demonstrating that politeness is an especially powerful predictor of workplace friendship maintenance. Contrary to expectations, however, although avoidance was rated as significantly less polite than openness/direct conversational refocus in escalating situations, respondents reported that employees were more likely to use that strategy. As noted earlier, maintaining a friendship in an escalating situation poses particularly high risk to the coworker’s positive face. In fact, the relative politeness ratings of maintenance tactics in deteriorating and escalating situations demonstrate this higher level of risk—although not directly comparable, the mean tactic politeness scores were consistently higher in deteriorating situations than escalating situations (see Tables 3 and 4), indicating that preventing a friendship from growing closer is a potentially more face-threatening negotiation than preventing it from deteriorating. Accordingly, although avoidance may be perceived as less polite than open and direct interaction that increases distance, individuals may nonetheless prefer to use it because they want to avoid an uncomfortable and awkward interaction.

Although less robust than for politeness, attachment style and task interdependence were also linked to maintenance tactic likelihood of use. As noted earlier, Secures tend to have an approach orientation to interpersonal relationships, while Dismissives tend to avoid them. It is not surprising, then that Secures were more likely to expect employees to use creating closeness, a tactic that requires approaching others, albeit subtly, to prevent deterioration. Dismissives would likely be less concerned with friendship deterioration generally, and less comfortable with tactics that require approaching others. Contrary to expectations, however, Fearfuls, who have an avoidance orientation toward relationships, also reported higher likelihood scores for creating closeness. Recall, however that although Fearfuls have low levels of self-confidence and worry about being hurt by others, they still value relationships. Thus, in deteriorating situations, the less overt creating closeness tactic might be preferable.

Task interdependence also predicted tactic likelihood of use in deteriorating situations, but in directions different than expected. Specifically, those responding to low interdependence scenarios reported higher likelihood scores for creating closeness (a polite tactic) than did those in high interdependence scenarios. A closer look at the creating closeness items helps explain this finding. These items involve the employee creating opportunities for interaction and contact. High interdependence situations, by definition, require more contact and interaction between coworkers; thus, creating opportunities for contact may be less necessary than for employees in low interdependence situations. More puzzling is the finding that employees responding to low interdependence scenarios reported higher scores for circumspectiveness (another polite tactic) than did those in high interdependence scenarios. In sum, contrary to expectations, for deteriorating situations, individuals in high interdependence scenarios were less likely than those in low interdependence scenarios to prefer polite tactics. This may indicate a greater sense of security; that is, employees who work closely with a friend who is becoming more distant may be less concerned with losing that friendship because their jobs require, and thus ensure, daily contact. In contrast, if the friendship is the primary point of contact and commonality for coworkers (i.e., low interdependence), distancing or “fading away” may be easier and, an employee may, therefore, put more thought and thoughtfulness into their maintenance tactics to not risk losing all contact with their coworker.

Although perceived politeness was the best overall predictor of likelihood of tactic use in both deteriorating and escalating situations, attachment style and task interdependence played a major role in deteriorating situations, predicting all three tactics; while only one main effect (and a relatively weak one) was identified for tactic use in escalating situations. Thus, politeness was the only meaningful predictor of maintenance tactic use for escalating situations in the present study. However, as mentioned above, a desire to avoid uncomfortable interactions may also play a role in preventing friendship escalation. Our study purposely centered only on the recipient’s positive face, but results indicate employees may also consider their own face in escalating situations, choosing tactics that protect their image as caring and considerate individuals. Future studies should incorporate this concept to develop further insights into the complexities of maintaining friendships in escalating situations.

Because of friendship’s important role in organizational processes, insights into effective friendship maintenance are important for practitioners. In particular, knowledge of how maintenance behaviors are perceived by others can help employees enact behaviors that are likely to maintain, rather than damage, workplace friendships, or at least to better understand the potential consequences of various maintenance strategies. In addition, our results indicate that contextual elements such as task interdependence play a role in workplace friendship maintenance. Task interdependence has been shown to encourage the initiation of workplace friendships (Sias & Cahill, 1998) and our results indicate that it also impacts the maintenance of those relationships. Practitioners should consider the extent to which employees’ tasks are interdependent and help employees develop constructive and polite methods for maintaining their friendships in varying contexts. Managers, trainers, and other organizational communication professionals can use the present findings to help employees understand the dynamic nature of personal/work boundaries and how they communicatively transform or maintain those boundaries in ways perceived as polite and considerate of their coworkers.

Limitations and Future Research

Although the maintenance tactics items were derived from prior research and evaluated and revised by subject matter experts, Study 1 was limited to two focus groups. Thus, we cannot provide evidence of theoretical saturation regarding the maintenance tactics. The focus groups’ purpose, however, was to obtain feedback from subject matter experts regarding items established in previous research (Lee & Jablin, 1995) rather than develop entirely new themes; thus it was appropriate to use fewer participants. The maintenance items were supported by factor analyses and consistent throughout the studies, indicating the validity of the tactics established in Study 1. The research is also limited by use of a student sample in Study 2. We note, however that 94% of Study 2’s participants had work experience and, therefore, experience with workplace relationships. Moreover, Study 2’s goal was to assess perceived politeness of maintenance tactics, not use, a task for which the sample was qualified.

The use of scenarios in Study 3 limits the ecological validity of the findings. As noted earlier, however, many studies of facework and politeness rely on hypothetical scenarios such as those used here (e.g., Feng & Burleson, 2008; Miller & Roloff, 2007; Wilson et al., 2008). In addition, our analysis revealed a social desirability bias, indicating the scenario method was useful for maintaining internal validity. Thus we chose to mitigate internal validity at the risk of threatening external validity. Three factors, however, indicate the scenario method did not seriously harm Study 3’s external validity. First, participants perceived the scenarios as very realistic and very likely to occur in a real workplace. Second, our results are largely consistent with extant theory and research, indicating the data are realistic. Third, our individual difference variable, attachment style, predicted tactic choice, indicating that participants projected their own views and communication preferences, and thus likely behavior, in the “other” condition.

Despite these design limitations, the studies provide a useful starting point for future research on workplace friendship maintenance by developing maintenance tactic items, assessing the politeness of those items, and identifying predictors of tactic preferences. Research should further test the instrument by having participants recall maintenance situations they have experienced, perhaps as the target, if not the source, of the maintenance messages, with the understanding that social desirability may be an issue.

We were unable to test hypotheses regarding the use of deception in escalating situations because of reliability problems. Further work should be directed toward refining and improving the maintenance tactic instrument with respect to the use of deception perhaps by having SMEs generate additional items for future testing. The remaining tactics in both deteriorating and escalating situations were stable and showed acceptable reliability, however. Thus, the instrument developed in these studies could be used to examine a number of interesting and important issues. For example, to limit the scope of this project, we did not examine how the speaker’s face impacts maintenance tactic use. As noted above, our results suggest that in escalating situations, employees tend to avoid uncomfortable interactions that may present them as uncaring and inconsiderate and, thus, are more likely to use the less polite avoidance strategy. This suggests that, in such situations, the speaker’s positive face may be more important than the target’s. Our studies also focused only on positive face. However, negative face likely plays a role in friendship maintenance dynamics (e.g., creating closeness could be interpreted as an imposition on the target’s independence and autonomy). Future studies should examine these issues to provide additional insights into the complexities of workplace friendship maintenance.

Finally, our studies did not address the extent to which motivation may contribute to workplace friendship maintenance dynamics. It is likely that employees are more motivated to maintain some workplace friendships than others. Employees who are less motivated to maintain relationships likely approach the maintenance process in distinct ways. They may, for example, choose tactics that require relatively little effort, rather than those that are polite. This is another important area for future research.

In sum, the studies reported in this article make important and heuristic contributions to the literature by developing a set of peer workplace friendship maintenance tactics, establishing the perceived politeness and predictors of tactic preferences. These results can, in turn, guide a number of important future studies to enhance our understanding of how and why employees work to maintain important friendships with coworkers.

Footnotes

The authors declared no potential conflicts of interest with respect to the authorship and/or publication of this article.

The authors received no financial support for the research and/or authorship of this article.

Notes

References

Bartholomew

(1993). From childhood to adult relationships: Attachment theory and research. In Duck

(Ed.), Understanding relationship processes: Vol. 2. Learning about relationships (pp. 30-62). Newbury Park, CA: Sage.

Bartholomew

Horowitz

L. M.

(1991). Attachment styles among young adults: A test of a four-category model. Journal of Personality and Social Psychology, 61, 226-244.

Bippus

A. M.

Rollin

(2003). Attachment style differences in relational maintenance and conflict behaviors: Friends’ perceptions. Communication Reports, 16, 113-124.

Bowlby

(1969). Attachment and Loss: Vol. I. Attachment. New York, NY: Basic Books.

Bridge

Baxter

L. A.

(1992). Blended relationships: Friends as work associates. Western Journal of Communication, 56, 200-225.

Brown

Levinson

(1987). Politeness: Some universals in language use. Cambridge, UK: Cambridge University Press.

Cruz

C. A.

Shafer

W. E.

Strawser

J. R.

(2000). A multidimensional analysis of tax practitioner’s ethical judgments. Journal of Business Ethics, 24, 223-244.

Cupach

W. R.

Carson

C. L.

(2002). Characteristics and consequences of interpersonal complaints associated with perceived face threat. Journal of Social and Personal Relationships, 19, 443-462.

Dainton

(2007). Attachment and marital maintenance. Communication Quarterly, 55, 283-298.

10.

Dindia

(2003). Definitions and perspectives on relational maintenance communication. In Canary

D. J.

Dainton

(Eds.), Maintaining relationships through communication: Relational, contextual, and cultural variations (pp. 1-30). Mahwah, NJ: Lawrence Erlbaum.

11.

Dindia

Canary

D. J.

(1993). Definitions and theoretical perspectives on maintaining relationships. Journal of Social and Personal Relationships, 10, 163-173.

12.

Eisenberg

E. M.

Goodall

H. L.

Jr. (2004). Organizational communication: Balancing creativity and constraint (4th ed.). Boston, MA: Bedford/St. Martin’s.

13.

Fehr

(1996). Friendship processes. Thousand Oaks, CA: Sage.

14.

Feng

Burleson

B.R.

(2008). The effects of argument explicitness on responses to advice in supportive interactions. Communication Research, 35, 849-874.

15.

Fisher

R. J.

(1999). Social desirability bias and the validity of indirect questioning. Journal of Consumer Research, 2, 303-316.

16.

Fritz

J. H.

(2009). Rudeness and incivility in the workplace. In Morrison

Wright

(Eds.), Friends and enemies in organisations: A work psychology perspective (pp. 168-194). Hampshire, UK: Palgrave Macmillan.

17.

Goffman

(1967). Interaction ritual. Garden City, NY: Doubleday.

18.

Guerrero

L. K.

Bachman

G. F.

(2006). Associations among relational maintenance behaviors, attachment-style categories and attachment dimensions. Communication Studies, 57, 341-361.

19.

Hazan

Shaver

(1987). Romantic love conceptualized as an attachment process. Journal of Personality and Social Psychology, 52, 511-524.

20.

Kiggundu

M. N.

(1983). Task interdependence and job design: Test of a theory. Organizational Behavior and Human Performance, 31, 145-172.

21.

Kline

T. J. B.

(2005). Psychological testing: A practical approach to design and evaluation. Thousand Oaks, CA: Sage.

22.

Kram

K. E.

Isabella

L. A.

(1985). Mentoring alternatives: The role of peer relationships in career development. Academy of Management Journal, 28, 110-132.

23.

Kramer

M. W.

(1994). Uncertainty reduction during job transitions: An exploratory study of the communication experiences of newcomers and transferees. Management Communication Quarterly, 7, 384-412.

24.

Lee

(1998). Maintenance communication in superior-subordinate relationships: An exploratory investigation of group social context and the “Pelz Effect.” Southern Communication Journal, 63, 144-159.

25.

Lee

Jablin

F. M.

(1995). Maintenance communication in superior-subordinate work relationships. Human Communication Research, 22, 220-257.

26.

Lim

T. S.

(1994). Facework and interpersonal relationships. In Ting-Toomey

(Ed.), The challenge of facework: Cross-cultural and interpersonal issues (pp. 209-229). New York: State University of New York Press.

27.

Marks

S. R.

(1994). Intimacy in the public realm: The case of co-workers. Social Forces, 72, 843-858.

28.

Miller

C. W.

Roloff

M. E.

(2007). The effect of face loss on willingness to confront hurtful messages from romantic partners. Southern Communication Journal, 72, 247-263.

29.

Rawlins

W. K.

(1992). Friendship matters: Communication, dialectics, and the life course. New York, NY: Aldine de Gruyter.

30.

Shea

B.C.

Pearson

J. C.

(1986). The effects of relationship type, partner intent, and gender on the selection of relationship maintenance strategies. Communication Monographs, 53, 352-364.

31.

Sias

P. M.

(2009). Organizing relationships: Traditional and emerging perspectives on workplace relationships. Los Angeles, CA: Sage.

32.

Sias

P. M.

Cahill

D. J.

(1998). From coworkers to friends: The development of peer friendships in the workplace. Western Journal of Communication, 62, 273-299.

33.

Sias

P. M.

Gallagher

(2009). Developing and maintaining workplace relationships. In Morrison

Wright

(Eds.), Friends and enemies in organisations: A work psychology Perspective (pp. 78-100). Hampshire, UK: Palgrave Macmillan.

34.

Sias

P. M.

Perry

(2004). Disengaging from workplace relationships: A research note. Human Communication Research, 30, 589-602.

35.

Sias

P. M.

Heath

R. G.

Perry

Silva

Fix

(2004). Narratives of workplace friendship deterioration. Journal of Social and Personal Relationships, 21, 321-340.

36.

Sias

P. M.

Smith

Avdeyeva

(2003). Sex and sex-composition differences and similarities in peer workplace friendship development. Communication Studies, 54, 322-340.

37.

Sigman

S. J.

(1995). Order and continuity in human relationships: A social communication approach to defining “relationship.” In Leeds-Hurwitz

(Ed.), Social approaches to communication (pp. 188-200). New York, NY: Guilford.

38.

Simon

E. P.

Baxter

L. A.

(1993). Attachment-style differences in relationship maintenance strategies. Western Journal of Communication, 57, 416-430.

39.

Taggar

Haines

V. Y.

III . (2006). I need you, you need me: A model of initiated task interdependence. Journal of Managerial Psychology, 21, 211-230.

40.

Waldron

V. R.

(2003). Relationship maintenance in organizational settings. In Canary

D. J.

Dainton

(Eds.), Maintaining relationships through communication: Relational, contextual, and cultural variations (pp. 163-184). Mahwah, NJ: Lawrence Erlbaum.

41.

Weick

K. W.

(1995). Sensemaking in organizations. Thousand Oaks, CA: Sage.

42.

Whitmeyer

J. M.

(2002). A deductive approach to friendship networks. Journal of Mathematical Sociology, 26, 147-165.

43.

Wilson

S. R.

Kunkel

A. D.

Robson

S. J.

Olofuwote

J. O.

Soliz

(2008). Identity implications of relationship (re)definition goals: An analysis of face threats and facework as young adults initiate, intensify and disengage from romantic relationships. Journal of Language and Social Psychology, 28, 32-61.