Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

26
Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies’ FIONA LEE^ Harvard University There are many ways to communicate bad news. The MUM effect (Tesser & Rosen, 1975), which is keeping mum and not transmitting the bad news at all, is only one of many possible approaches. Using P. Brown and Levinson’s (1987) politeness theory, an experimental study was conducted to address not whether bad news is being transmitted, but how bad news is being transmitted. The results show that most communicators tend to use politeness strategies when communicating bad news. Moreover, using politeness strategies to couch the message did not attenuate the informative value of the message. Focused contrasts revealed two strong interactions between gender and communication direction on strategy use. First, power differ- ences more strongly predicted strategy use for men, whereas distance differences more strongly predicted strategy use for women. Second, men used most politeness strategies when the combined effect of power and distance was the highest (as the politeness theory would predict), but the trend was reversed for women. This finding suggests that politeness theory may not be an accurate model for describing female communicators. People are often reluctant to transmit bad news. This phenomenon, known as the MUM effect, has been replicated in many studies across vari- ous settings (O’Neal, Levine, & Frank, 1979; Tesser & Rosen, 1975). Reluc- tance to transmit bad news can take several forms-for example, communi- cators are more likely to pass on good news rather than bad news (O’Reilly & Roberts, 1974; Tesser & Rosen, 1975), or they tend to distort, delay, or delegate the transmission of the bad news (Bond & Anderson, 1987; Rosen, Grandison, & Stewart, 1974). The MUM effect is especially interesting to organizational psychologists, because the omission, distortion, or delay of information can be detrimental to organizational effectiveness. Fulk and Mani (1986) state that “distortion is an important problem in organizations because modifications to messages cause misdirectives to be transmitted, nondirectives to be issued, incorrect information to be passed on, and a variety of other problems related to the quantity and quality of information” (p. 483). Lewis (1987) argues that ‘1 want to thank J. Richard Hackman and Roger Brown for their generous advice and guidance throughout the preparation of this paper. Special thanks also to two anonymous reviewers for their valuable comments on an earlier draft. 2Correspondence concerning this article should be addressed to Fiona L.ee, Department of Psychology, Harvard University, 33 Kirkland Street, Cambridge, MA 02138. 1124 Journal of Applied Social Psychology, 1993,23, 14, pp. 1124-1149 Copyright @ 1993 by V. H. Winston & Son, Inc. All rights reserved.

Transcript of Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

Page 1: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies’

FIONA LEE^ Harvard University

There are many ways to communicate bad news. The MUM effect (Tesser & Rosen, 1975), which is keeping mum and not transmitting the bad news at all, is only one of many possible approaches. Using P. Brown and Levinson’s (1987) politeness theory, an experimental study was conducted to address not whether bad news is being transmitted, but how bad news is being transmitted. The results show that most communicators tend to use politeness strategies when communicating bad news. Moreover, using politeness strategies to couch the message did not attenuate the informative value of the message. Focused contrasts revealed two strong interactions between gender and communication direction on strategy use. First, power differ- ences more strongly predicted strategy use for men, whereas distance differences more strongly predicted strategy use for women. Second, men used most politeness strategies when the combined effect of power and distance was the highest (as the politeness theory would predict), but the trend was reversed for women. This finding suggests that politeness theory may not be an accurate model for describing female communicators.

People are often reluctant to transmit bad news. This phenomenon, known as the MUM effect, has been replicated in many studies across vari- ous settings (O’Neal, Levine, & Frank, 1979; Tesser & Rosen, 1975). Reluc- tance to transmit bad news can take several forms-for example, communi- cators are more likely to pass on good news rather than bad news (O’Reilly & Roberts, 1974; Tesser & Rosen, 1975), or they tend to distort, delay, or delegate the transmission of the bad news (Bond & Anderson, 1987; Rosen, Grandison, & Stewart, 1974).

The MUM effect is especially interesting to organizational psychologists, because the omission, distortion, or delay of information can be detrimental to organizational effectiveness. Fulk and Mani (1986) state that “distortion is an important problem in organizations because modifications to messages cause misdirectives to be transmitted, nondirectives to be issued, incorrect information to be passed on, and a variety of other problems related to the quantity and quality of information” (p. 483). Lewis (1987) argues that

‘ 1 want to thank J. Richard Hackman and Roger Brown for their generous advice and guidance throughout the preparation of this paper. Special thanks also to two anonymous reviewers for their valuable comments on an earlier draft.

2Correspondence concerning this article should be addressed to Fiona L.ee, Department of Psychology, Harvard University, 33 Kirkland Street, Cambridge, MA 02138.

1124

Journal of Applied Social Psychology, 1993,23, 14, pp. 1124-1149 Copyright @ 1993 by V. H. Winston & Son, Inc. All rights reserved.

Page 2: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1125

information distortion in organizations results in an organizational climate of suspicion and mistrust, causing employees unnecessary anxiety and inse- curity, and eventually rendering employees apathetic or overly susceptible to rumors. The Stealth Bomber project is a specific example of how the MUM effect affected organizational effectiveness. The secretive nature of the pro- ject “discouraged officers from reporting bad news up the chain of com- mand” and “bad news . . . was systematically suppressed and, when it could not be held back, was ignored or finessed on a number of occasions”(Kap- lan, 1990, p. 1). As a result, senior Pentagon officials continued to fund the project despite cost overruns, schedule delays, and technical problems.

When faced with bad news, communicators must first make a choice of whether or not to transmit the information. The MUM effect suggests that communicators are more likely to make the decision not to transmit the information when it is bad news. The present study goes beyond the MUM effect by exploring the alternate choice-when communicators do decide to transmit the bad news, what choices or approaches are available? Logically, several choices are open to the communicator once the decision has been made to transmit the bad news. First, the communicator can convey all the relevant information “baldly” (P. Brown & Levinson, 1987), matter-of-factly stating the bad news in a straightforward fashion. Second, the communica- tor can lessen the “threat”of the bad news by couching the information with strategies. For example, the communicator can encourage and comfort the listener, or the communicator can reduce the importance or relevance of the bad news (P. Brown & Levinson, 1987). Third, the communicator can increase the threat of the bad news by being overly negative. For example, the communicator can be pessimistic or make negative attributions about the bad news.

Based on these choices, the present study explores the different approaches communicators use to transmit bad news by measuring the level of informa- tion, the usage of couching strategies, and the overall negativity in the transmission. P. Brown and Levinson’s (1987) politeness theory is a particu- larly useful model to explore these issues. According to this theory, polite- ness is used to redress face-threatening acts, or actions that threaten an- other’s needs to be publicly appreciated and be free of impositions. When communicating face-threatening acts, communicators can use “positive” strategies, which acknowledge the listener’s needs, and “negative” strategies, which minimize the threat. Based on this theory, politeness is appropriate when communicating bad news because bad news is a face-threatening act that directly threatens the recipient’s needs to be appreciated and acknowl- edged. Furthermore, politeness theory outlines specific individual, social, and contextual conditions that affect how face-threatening acts are commun- icated, providing a model to understand the antecedents of different com-

Page 3: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1126 FIONA LEE

munication approaches. Using politeness theory as a framework, the present study addresses two research questions: (a) what are the effects of the differ- ent communication approaches on the informative value of the message, and (b) what are the factors that affect the choice of communication approaches?

Effect of Communication Approaches on Informative Value

Information distortion is a problem for organizations when the “quality and quantity” of information are compromised (Fulk & Mani, 1986). How do different communication approaches affect the informative value of the message? Of the different approaches, perhaps couching-with-politeness strategies offer the highest potential for information distortion, because the communicator is actively trying to decrease the threat of an inherently face- threatening act. According to politeness theory, the strategies communica- tors use to redress threats to public appreciation, such as negative evalua- tions, include “claiming common ground with the listener,” “trying to fulfill the listener’s wants,” “minimizing the threat,” and “disassociating the self from the message” (P. Brown & Levinson, 1987). Whereas strategies such as “minimizing the threat” might result in information distortion, other strate- gies such as “fulfilling the listener’s wants,” “claiming common ground,” and “disassociating the self’ do not entail any compromise to the accuracy or completeness of the information. This suggests that although politeness strategies could potentially compromise the quality of the information, they do not necessarily affect the informative value of the message.

Contextual, Social, and Individual Factors Affecting Communication

Many studies on information distortion have focused on factors that increase or decrease the level of information distortion within the organiza- tional hierarchy. These factors can be roughly divided into three groups (Dansereau & Markham, 1987): (a) contextual factors, such as organiza- tional climate and structure (Athanassiades, 1973; Bacharach & Aiken, 1977; Curtis, 1989; Jablin, 1982); (b) relationship or social factors, such as trust and interdependence (Larson, 1986; Read, 1962; Roberts & O’Reilly, 1974); and (c) individual difference factors, such as the communicator’s upward mobility aspirations (Athanassiades, 1973; Cohen, 1958; Glauser, 1984; Read, 1962; Synder & Zorn, 1986) or the supervisor’s leadership and com- munication style (Level &Johnson, 1978; Luthans & Larsen, 1986).

P. Brown and Levinson’s (1987) politeness theory also considers how individual, social, and contextual factors affect communication. Past re- search on politeness theory (Baxter, 1984; R. Brown & Gilman, 1989; Field, 1991; Holtgraves & Yang, 1990, 1992; Smith-Hefner, 1988) show that usage

Page 4: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1127

of politeness strategies is a function of (a) contextual factors such as the relevance or extremity of the face-threatening act, (b) social factors such as the power of the listener over the communicator, and the social distance between the listener and the communicator, and (c) individual difference factors such as the gender of the communicator.

Contextual factor-relevance of the bad news. P. Brown and Levinson (1987) propose that as the extremity of the bad news increases, more polite- ness is needed to couch the information. However, research on information distortion typically compared the difference between transmitting good news and bad news, with little emphasis on how different types of bad news would affect the communication process. The present study focuses on how the relevance of the bad news would affect the way the information is transmit- ted. According to P. Brown and Levinson (1987), the extremity of the face threat against another’s positive face involves the amount of “pain” given to the hearer. In this sense, bad news that is relevant to both the speaker and the hearer should be less painful than bad news that is relevant only to the hearer, because the “pain” is distributed among both speaker and hearer. Therefore, bad news that is relevant to both speaker and hearer should be less extreme or threatening, and transmitted with less politeness and couch- ing than bad news that is relevant only to the hearer. However, a study found that supervisors who were more dependent on their subordinates tended to be more positive and helpful when transmitting bad news (Ilgen, Mitchell, & Fredrickson, 198 1). In other words, supervisors who perceived their subor- dinates’ poor performance as directly relevant to their own performance were actually more polite, suggesting that face threat may be higher when bad news is relevant to both the speaker and the hearer. Therefore, it is unclear how the relevance of the bad news could affect the transmission of the information.

Social factors-power and distance. In politeness theory, power refers to the power of the listener over the communicator, and distance refers to the “similarity/ difference” between the communicator and the listener (P. Brown & Levinson, 1987). As power and distance increase, more politeness is needed to communicate the face-threatening act, and more strategies will be used (R. Brown, 1990; R. Brown & Gilman, 1989).

Power and distance are espec-ially salient within organizational hier- archies. In many ways, the basic structure of the organizational hierarchy is defined by interpersonal differences in power and distance (Cohen, 1958). Power is high in upward communication (subordinate communicating to a supervisor), and low in downward communication (supervisor communicat- ing to subordinate). In terms of distance, people occupying similar levels within the organizational hierarchy are more likely to have similar organiza- tional experiences, and more likely to interact with each other informally

Page 5: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1128 FlONA LEE

(Kanter, 1977; Mintzberg, 1973). Therefore, distance is.higher in upward and downward communication situations where the communicator and the lis- tener are at different hierarchical levels, and lower in horizontal communica- tion situations where the communicator and listener occupy the same hier- archical level.

Although power and distance are typically considered as independent dimensions in politeness research, they are inseparable within the organiza- tional hierarchy. When relating to supervisors, subordinates, and peers, there are obvious implications of power differences, and also obvious impli- cations of distance differences. Thus, when considering communication within organizational hierarchies, power and distance are appropriately con- sidered in conjunction. Table 1 shows the different levels of power, distance, and the combined effects of power and distance in upward, downward, and horizontal communication. The combined power and distance effect is high- est in the upward communication, medium in downward communication, and lowest in horizontal communication. Politeness theory would suggest that politeness strategies would be used most in upward communication situations, when levels of power and distance are highest.

Individual factor-gender. In general, the politeness research provides inconclusive evidence about gender differences in the communication of face-threatening acts. Holtgraves and Yang (1992) found that female com- municators were more polite than male communicators. On the other hand, Field (1991) found that men and women did not differ in amount of polite- ness, although subjects were more polite when playing the role of female communicators than when playing the role of male communicators. There is also evidence to suggest that gender differences in communication interact with social factors of power and distance (Steckler & Rosenthal, 1985). Tannen’s (1990) theory that men and women are differentially sensitive to power and social distance was supported by Holtgraves and Yang (1992),

Table I

Levels of Power and Distance for the Three Levels of Communication Direction

Communication direction ~~ ~

Upward Downward Horizontal

Power

Distance

High Low Medium

High High Low ~

Power + Distance High Medium Low

Page 6: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1129

who found that when using politeness strategies, women placed more weight on distance than men did, whereas men placed more weight on power than women did. These studies suggest that men and women differ in how they communicate depending on the relative importance of power and distance they place on the relationship between the communicator and the listener.

Present Study: Design and Hypotheses

The present study goes beyond asking whether bad news is communicated within organizational hierarchies to describe how communicators transmit bad news. Specifically, this study explores (a) the effect of communication strategies on informative value, and (b) the effects of message relevance, communication direction, and gender on the how bad news is communi- cated. The present study has six experimental conditions-three levels of communication direction (upward, downward, and horizontal), crossed with two levels of message relevance (“individualized” bad news that is not rele- vant to the communicator, and “collective” bad news that is relevant to both the communicator and the listener). It is hypothesized that (a) using polite- ness strategies does not affect the quality or accuracy of the information transmitted, (b) when the combined effect of power and distance is highest (upward communication direction), more strategies are used to communi- cate bad news than when the combined effect is lowest (horizontal commun- ication direction), and (c) power differences have more impact on men and distance differences have more impact on women.

Method

Subjects

Sixty subjects (27 male, 33 female) participated in the study. The subjects’ ages ranged from 16 to 28. They include Harvard University undergraduate students (15 male, 19 female), graduate students (4 male, 7 female), and summer school students (8 male, 7 female). Subjects were recruited through a posted sign-up sheet, describing an experiment in which the top two per- formers would each receive $100. Overall, there was an even distribution of subjects from the humanities, social sciences, and physical sciences.

Procedure

The present study used Cohen’s (1958) procedure to experimentally create an organizational hierarchy. Subjects were randomly assigned to roles of supervisors, subordinates, and peers while performing an experimental task.

Page 7: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1130 FlONA LEE

A few modifications were made to Cohen’s experimental procedures to bet- ter simulate an organizational environment. First, subjects were asked to perform a meaningful and challenging task. Second, they were led to believe that payment would be contingent on their performance. Third, there was a time constraint for completing the task. Fourth, subjects were not only labeled as supervisors, subordinates, and peers, but were also given different levels of information power, reward power, and expert power (French & Raven, 1968).

Each subject arrived at the experimental room with a same-gender con- federate. After filling out an initial questionnaire on personal work experi- ence, the experimenter outlined the experimental instructions to both the subject and the confederate. The subjects were led to believe that the experi- ment was designed to measure specific personality traits that accurately predicted future success in the workplace. Because the “bad news” would be operationalized later in the experiment as a poor personality trait evalua- tion, this “cover” was used to create a high level of personal relevance of the bad news to the target of the evaluation.

The subject and the confederate worked as a team on a complex task that involved designing a staff reduction plan for a fictitious company. Although the subject and confederate worked as a team, they had to work in separate rooms, and could only pass written messages under the door to each other. Each team was told that 30 min would be allocated to complete the task, and the members of the team with the best performance in the staff reduction task would receive $100 each.

Each subject/ confederate team was randomly assigned to one of the fol- lowing conditions: (a) subject as supervisor, confederate as subordinate (downward communication direction); (b) subject as subordinate, confeder- ate as supervisor (upward communication direction); or (c) subject and con- federate as peers (horizontal communication direction). Before assigning roles, the experimenter spent a moment looking through the initial question- naires, in which the subject and confederate provided work experience information. This was done to create the impression that the role assignment was merit based. In the upward and downward conditions, the subject was also led to believe that the supervisor would evaluate the subordinate’s per- formance at the end of the task: Should the subordinate’s performance be rated below average, only the supervisor would receive the award, even if the team developed the best plan. In the horizontal condition, no such provi- sions were given. Information needed to perform the staff reduction task was then distributed among the team members. In the upward and downward communication conditions, the supervisor received more and better infor- mation than the subordinate. In the horizontal condition, the information was equally divided between the subject and confederate.

Page 8: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1131

After these instructions, the confederate was supposedly led away to a n adjacent room. In fact, the confederate was dismissed. The experimenter returned to the subject and gave the additional instructions depending on whether they were supervisors, subordinates, or peers. This set of instruc- tions, developed by Cohen (1958), is illustrated in Table 2.

When the task began, the experimenter went t o the adjacent room and

Table 2

Instructions to Subjects in the Upward, Downward, and Horizonal Direction Conditions (Cohen, 1958)

Upwards: “As you can see, your supervisor has the more important and critical pieces of information. He or she will give you directions on how to act. Your main job is to follow the instructions he/she gives you as best you can. Your super- visor has the more important and difficult task. His or her job requires more ingenuity, organization, and even creativ- ity, for there are many ways to reduce staff in the organiza- tion. Your job is the poorer one-more menial and routine. You are simply to follow your supervisor’s instructions as accurately as possible. Also, your supervisor will have input into your evaluation after the task is over. The evaluation of you will affect whether you will win $100.”

“Your subordinate has poorer information than you do. Your job is to integrate all the global issues and let your subordinate know what to do with the information he has. You have the difficult task of integrating all the company’s goals and strategies and forming an ideal staff reduction plan. This requires a lot of insight, organization, compre- hension, and even creativity. Your subordinate has the poorer job-more menial and routine. He or she is depen- dent on your instructions. After the task is completed, you will evaluate how well he/ she has performed according to your demands. Your evaluations will affect whether he or she will win $100.”

“Let me summarize your job again. Both you and your partner have some information about the company. Your job is to jointly integrate both of your information and come up with an ideal plan to streamline the company. Both sets of information are equally essential to completing the task. Obviously, both you and your partner have equally impor- tant roles to play. Depending on how well each of you do, you may be selected to win $100.’’

Downwards:

Horizontal:

Page 9: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1132 FIONA LEE

began passing standardized messages related to the staff reduction task to the subject. (The subject believed that the person on the other side was the confederate.) Ten minutes into the task, the subject was given an interim evaluation of personality traits. The subject was randomly given one of the following evaluations: (a) an evaluation of the confederate’s personality traits (individualized bad news, only relevant to the listener), or (b) an evalu- ation of the team’s personality traits (collective bad news, relevant to both the listener and the communicator). To create the “bad news,” the interim evaluation was very poor in both cases (see Appendix). The subject was instructed to relay the results of the evaluation to the confederate through a written message. Once the subject passed the message, the experiment was ended and the subjects debriefed.

Manipulation Checks

Six subjects participated in a pilot study, one for each experimental condi- tion. After the experimental procedure, a questionnaire containing manipu- lation checks was given to these subjects. The interim evaluation was rated equally poorly in all the experimental conditions. Pilot subjects in the collec- tive bad news conditions rated the bad news as more personally relevant than those in the individualized bad news conditions. All the pilot subjects felt that the supervisor-subordinate-peer manipulation was successful- supervisors felt that they had more power, subordinates felt that they had less power, and peers did not perceive any power differences between them- selves and the confederates. Supervisors and subordinates also felt a larger social distance from their partners than did peers. This suggests that, as expected, the communication direction manipulation affected both power and distance.

Coding of Notes

The notes that the subjects relayed to the confederates regarding the nega- tive interim evaluation were first content coded and then Q-sorted.

Content coding. The procedure for coding politeness strategies in the present study differs from past research in politeness theory, which typically coded the occurrence of general strategies developed by P. Brown and Levinson (1987). Instead of this “top-down’’ approach for coding strategies, the present study utilized a “bottom-up” approach: The strategies were selected directly from an initial reading of the messages by a coder familiar with politeness theory, but blind to the experimental conditions. This coder first read literature on politeness theory (P. Brown & Levinson, 1987; R. Brown, 1990), and then went through all the notes recording each distinct

Page 10: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1133

strategy found in the notes. Table 3 illustrates the specific strategies identi- fied from the notes, and the corresponding general strategies proposed by P. Brown and Levinson. Note that the specific strategies did not represent all the strategies proposed by P. Brown and Levinson, because they were more specific and situationally relevant than the general strategies designed to address all types of face-threatening acts. However, all the specific strategies could be subsumed under general strategies from politeness theory.

After the strategies were selected by the first coder, two other coders, also blind to the experimental conditions, rated the occurrence of these strategies for each note. Table 4 shows two notes generated from the experimental procedure. In the first note, the communicator used two strategies, decreas- ing the importance of the bad news (“Who cares?”) and directing the listen- er’s attention back to the task (“Moving right along. . .”). In the second note, the communicator offered acts for improvement (“Let’s work o n . . .”), and also offered encouragement (“I’m confident . . . and you deserve to be confident too”). Besides coding for strategies, the two coders also coded for the following variables: (a) the number and proportion of “very poor” rat-

Table 3

Politeness Strategies Coded With Corresponding P. Brown and Levinson (1987) Typology

Corresponding Politeness strategy Examples from present typology in

coded in present study study P. Brown and Levinson

1. Discredit the source directly “How would they know?” Minimize threat

2. Decrease importance “This is only interim” Minimize threat

3. Disagree explicitly “They are wrong” Minimize threat

4. Providing acts for

5. Direct attention to

improvement “Let’s work on . . .” Hedging

the task (simply continues task) Include both in activity

6 . Provide comfort “Don’t worry” Attend needs, sympathy

7. Provide encourage- ment “You can do it” Be optimistic

Page 11: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1134 FIONA LEE

Table 4

Examples of Notes Written to Communicate Bad News

Note 1: Dear Mike,

In our evaluation, we are getting nailed. Who cares? Moving right along. . . (continues with the organizational task)

Note2: J,

They say you appear to be creative and energetic and to commun- icate well. Let’s work on planning an overall strategy, handling this large amount of information, and organizing a bit more. I’m confident we can come up with a good plan-you deserve to be confident too.

ings from the interim evaluation mentioned in the note, (b) the severity of the note rated along a 0 to 4 point scale, (c) the informative value of the note rated along a 0 to 4 point scale, (d) whether the note contained a general negative summary of the evaluation (e.g., “things do not bode well for us”), and (e) whether the note contained a general positive summary of the evalua- tion (e.g., “you are doing great”).

Q-sort. Two additional coders blind to the experimental conditions Q- sorted the notes. The notes were systematically separated into seven piles (to reflect scores of 1 to 7) along each of the following dimensions: (a) amount of discounting (Q-discounting), (b) amount of encouragement (Q-encourage), (c) amount of negativity (Q-negative), (d) amount of information specificity (Q-specificity), and (e) amount of orientation toward the task (Q-task).

Preliminary Considerations

Table 5 lists the estimated reliability of all the coded variables. The Spearman Brown estimated reliability for all the content coded and Q-sort variables ranged from 0.61 to 0.99.

A step-up-by-one principal components analysis with varimax rotation was performed to understand the relationship between the various content-coded and Q-sorted dependent variables. The three-factor solution was selected. The factor loadings are listed in Table 6. The three orthogonal dimensions were negativity, informative value, and strategies used. The majority of the varia- bles loaded on one of the factors at a loading of 0.40 or higher. Three variables did not have a factor loading of more than 0.40-providing comfort, discredit- ing the source of the bad news, and decreasing the importance of the bad news. These variables were all specific politeness strategies.

Page 12: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1135

A negativity composite score and an informative value composite score were formed by averaging the standardized values of the variables under the negativity and informative value factors. To better understand the strategy variables, a second factor analysis was performed only on the strategy vari- ables, which included the variables grouped under the strategy factor in the first factor analysis, as well as the three strategy variables that did not load onto any factor. Again, a step-up-by-one principal components analysis with varimax rotation was used. The results are illustrated in Table 7. A three- factor solution was selected. The three factors are: (a) “positive” politeness

Table 5

Estimated Reliability Between Codes for All Dependent Variables

Spearman Brown estimated reliability

Average correlation Variables

Strategies Number of strategies Discrediting Decreasing Disagreeing Attention to task Acts of improve Comforting Encouraging

Other content coding items Number “very poor” Severity Informative Giving a negative comment Giving a positive comment

Q-sorts Q-negative Q-discount Q-encourage Q-specificity Q-task

0.77 0.91 0.94 0.65 0.77 0.58 0.72 0.44

0.98 0.77 0.93 0.58 0.44

0.79 0.71 0.74 0.85 0.78

0.87 0.95 0.97 0.79 0.87 0.73 0.84 0.61

0.99 0.91 0.97 0.73 0.62

0.88 0.83 0.85 0.92 0.84

Page 13: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1136 FlONA LEE

Table 6

Factor Loadings of Three Factor Solution: Factor Analysis of All Coded Variables

Factors

Negativity Informative Strategies Coded variables

Q-negative 0.86 0.08 -0.31 Severity 0.82 0.37 -0.23 Giving a negative comment 0.81 -0.08 0.08 General positive -0.68 -0.17 0.11

Informative 0.00 0.93 -0.22 Q-specific -0.04 0.92 -0.10 Number “very poor” 0.10 0.88 -0.22 Proportion “very poor” -0.14 0.43 0.21

Number of strategies Q-encourage Encouraging Disagreeing Attention to task Q-discount Acts to improve Q-task

0.10 -0.53 -0.05 0.13

0.07 -0.13

-0.23 -0.32

0.15 0.05 0.13 0.10

-0.28 0.15

-0.20 -0.37

0.91 0.68 0.66 0.58 0.55 0.53 0.49 0.48

Comforting Discrediting Decreasing

-0.02 -0.06 0.29 0.02 0.30 0.24 0.39 -0.08 0.18

strategies, such as being encouraging and comforting; (b) “negative” polite- ness strategies, such as discounting, discrediting, decreasing the importance, and disagreeing with the bad news; and (c) task-oriented strategies, such as providing specific acts for improvement and directing the attention back to the task. Again, composite variables of positive, negative, and task-oriented strategies were formed. Because communicators tend to use more strategies to indicate increased politeness (R. Brown, 1990; R. Brown & Gilman, 1989), the number of strategies was left as a stand-alone variable instead of aver- aged into a composite variable. All in all, six variables were used in the subsequent data analyses: negativity, informative value, number of strate- gies, positive strategies, negative strategies, and task-oriented strategies.

Page 14: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1137

Table 7

Factor Loadings of Three Factor Solution: Strategies Variables Only

Factors

Positive/ encourage Negative Task Coded variables -~

Q-encourage Comforting Encouraging

~~ ~~

0.73 0.43 0.3 I 0.64 -0.10 0.1 1 0.62 0.42 0.12

Q-discount Discrediting Disagreeing Decreasing

-0.36 0.91 -0.08 -0.37 0.68 -0.01 0.06 0.55 0.15 0.17 0.40 0.38

Q-task 0.08 0.06 0.87 Attention to task -0.06 0.17 0.86 Acts to improve 0.24 -0.03 0.72

Results

Before examining the specific hypotheses, it is useful to descriptively examine how subjects communicated bad news. Figure 1 shows the fre- quency distribution of the negativity composite, the informative value com- posite, and number of strategies used. Not surprisingly, when communicat- ing bad news, the messages were predominantly negative. Figure 1 shows that twice as many subjects were rated above the mid-point in negativity than below the mid-point. Another apparent pattern in Figure I is that there was a range in informative value: subjects were not limited to the options of “baldly” transmitting the bad news (high informative value) or not transmit- ting the bad news at all (low informative value). The data showed that about half of the subjects were rated as providing intermediate levels of informative value. Informative value was not an all-or-nothing dimension. In terms of number of strategies used, most subjects used at least one strategy when communicating bad news. Only 7 out of 60 subjects did not use any polite- ness strategy; 43 subjects used two or more different strategies when com- municating bad news. The average number of strategies used was 2.55.

Effect of Couching on Informative Value

How would using strategies to couch the bad news affect the informative value of the message? Intercorrelations between informative value, negativ-

Page 15: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1138 FlONA LEE

Number of Subjects

20

4

E2 Below Midpoint Above Midpoint

-2.0 - 1.5 - 1.0 -0.5 0.0 0.5 1.0 1.5 Low Negativity High Negativity Negativity Composite Score

16 i

- 1.5 - 1.0 -0.5 0.0 0.5 1.0 1.5 High Informative Value Low Informarive Value

Informative Value Composite Score

0 1 2 3 4 5 6 7 8 Number of Strategies Used

Figure 1. Frequency distribution of Negativity composite, Informative Value com- posite, and Number of Strategies Used.

Page 16: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1139

ity, and strategy usage are illustrated in Table 8. Informative value was positively related to negativity, and negatively related to strategy usage. However, none of these correlations were significant (p > .05). This analysis supports the hypothesis that strategy usage was not strongly related to informative value.

Effects of Individual, Social, and Contextual Factors on Communication

To analyze the effect of specific individual, social, and contextual vari- ables on the communication of bad news, analyses of variance were per- formed on the dependent variables (negativity, informative value, number of strategies used, positive strategies, negative strategies, and task strategies) with gender, communication direction, and relevance as independent vari- ables.

Main effects of gender, relevance and direction. Table 9 illustrates the means, Fvalues, effect sizes, and significance levels for gender and relevance main effects. Women used more positive politeness strategies than men, F( 1, 48) = 4 . 3 0 , ~ = .04, and women were marginally less negative than men, F( 1, 48) = 3 . 2 1 , ~ = .08. Although women were not more polite than men across all strategy usage variables, this finding partially supports Holtgraves and Yang’s (1992) finding that women were more polite than men. Communica- tors in the collective condition were significantly more negative than those in the individualized condition, F(l, 48) = 18.655, p < .001, although there were no differences between individualized and collective conditions in informative value and strategy usage. This finding directly contradicts Ilgen et al.’s (I98 1) finding that communicators transmitting collective bad news were less negative.

Table 8

Correlations Between Informative Value and Other Composite Variables

Informative

Negativity

Strategy usage Number of strategies Positive strategies Negative strategies Task strategies

0.24

-0.21 -0.05 -0.04 -0.22

Page 17: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1140 FIONA LEE

Table 9

Means, Fs, Effect Sizes, and Significance Levels of Gender and Message Relevance Main Effects

Gender main effects

F(1, 48) r P Mean Mean

(Males) (Females)

Composite variables

Negativity 0.21 -0.17 3.21 .250 ,080 Informative value -0.17 0. I4 1.46 ,172 .I37 Number of strategies 2.44 2.64 0.27 ,076 ,604 Positive strategies -0.19 0.16 4.30 ,287 .043 Negative strategies 0.05 -0.04 0.32 ,082 ,574 Task strategies -0.04 0.03 0.12 ,049 .733

Message relevance main effects

F(1, 48) r P Mean Mean

(Indiv) (Collect)

Negativity -0.42 0.39 18.66 .529 < .0001 Informative value -0.09 0.08 0.86 ,132 ,360 Number of strategies 2.57 2.53 0.02 ,006 .96 1 Positive strategies 0.06 -0.06 0.30 .079 ,586

Task strategies 0.06 -0.05 0.21 ,065 .649 Negative strategies 0.00 0.00 0.00 .ooo .993

No significant main effects emerged for communication direction ( p > .20 for all effects). Two focused contrasts were performed on the communica- tion direction effect based on: (a) the combined effects of power and distance (P 4- D), and (b) the differential effects of power and distance (P - D).3 Surprisingly, neither contrast reached conventional levels of significance ( p > .10 for all effects).

3The contrast weights indicating power effects in upward, downward and hor i~onta l com- munication are + I , - 1 , O (the power of the listener over the communicator is highest in upward communication, and lowest in downward communication). The contrast weights for distance effects over the three levels of communication are + l , + I , -2 (distance is high in upward and downward communication, and low in horizontal communication). Both sets of contrast weights were standardized. The contrast weights for the combined effects of power and distance (P f D) were derived from the sum of the standardized weights. The contrast weights for the differential effects of power and distance (P - D) were derived from the difference between the standardized contrast weights.

Page 18: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1141

0.6 - 0.4 -- a .-

M E o.2 --

B -0.2 -- vl g 0.0 -- .- z

- 0.4 -- -0.6 --

Direction X Gender Interaction. Although main effects for communica- tion direction did not reach conventional levels of significance, significant Direction X Gender interactions were found for the strategy variables. The means of the strategy variables are plotted in Figure 2. Focused contrasts on the Direction X Gender interaction were performed on the means on these variables, using the contrast weights for the combined effects of power and distance (P -t- D) and the differential effects of power and distance (P - D).

= Females o Males

0.6 -

.P 0.2

E 0.0

I I I I

- 0’4; 0.4 $ I

2 ln

$ -0.2

- 0.6

I I

Comm. Dir. Up Down Horizontal Comm. Dir. up Down Horizontal

Power + Disrance High Medium Low Power + Distance High Medium Low

Page 19: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1142 FlONA LEE

These contrasts revealed two findings:

1. The combined effect of power and distance on strategy usage was consistent with the predictions of politeness theory for men, but the trend was reversed for women. The contrast for the inter- action was significant for number of strategies used, F( 1, 48) = 11.64, r = .42, p = .001, positive politeness strategies, F(1, 48) = 4.06, r = .26, p = .049, negative politeness strategies, F( 1, 48) = 4.30, r = .27, p = .043, and task strategies, F( 1, 48) = 4.57, r = .28, p = .037. As Figure 2 shows, as combined power and dis- tance increased, men used more politeness strategies. However, this was not true for women: Contrary to politeness theory, as the combined effect of power and distance increased, women used fewer strategies. Thus, the hypothesis that strategy usage would be highest in upward conditions and lowest in horizontal condi- tions was confirmed only for men, but disconfirmed for women.

2. A focused contrast of the differential effects of power and distance found that power had more impact on men, whereas distance had more impact on women, supporting the hypothesis that power and distance had differential effects on men and women. This effect was found in number of strategies used, F( 1, 48) = 12.06, r = .45, p = .002, positive politeness strategies F( 1, 48) = 5.25, r = .30, p = .008, negative politeness strategies, F( 1, 48) = 4.73, r = .28,p = .0413, and task strategies, F(1,48)=6.06, r = .32,p = .004.

Discussion

The present study explores how communicators transmit bad news in organizational hierarchies. Specifically, this study explores the effect of communication strategies on informative value, and the effects of message relevance, communication direction, and gender on how bad news is com- municated. First, the results revealed that informative value was not signifi- cantly related to negativity or usage of politeness strategies, supporting the hypothesis that couching with politeness did not strongly affect informative value. Also, it was hypothesized that gender, communication direction (power and distance), and relevance of the bad news would affect how bad news is transmitted. The present study found main effects for gender (women were more positive than men) and message relevance (collective bad news were more negatively transmitted than individualized bad news).

The key findings of the present study emerged from the interactions be-

Page 20: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1143

tween gender and communication direction. First, it was hypothesized that when using politeness strategies, power would have more impact on men, and distance more impact on women (Holtgraves & Yang, 1992; Tannen, 1991). This hypothesis was supported for number of strategies used, usage of positive strategies, negative strategies, and task strategies. It was also hy- pothesized that strategy usage would increase as the combined effect of power and distance increased. Although past research (Field, 1991; Holt- graves & Yang, 1992) supported this hypothesis among both male and female subjects, the present study found support for this hypothesis only among male subjects. As power and distance increased, men used more politeness strategies and women used fewer politeness strategies.

The inconsistency of the present results with past studies could be explained by the different coding methodology (the bottom-up approach instead of the top-down approach), or the focus on a different face threat (bad news instead of requests), or artifactual effects in the experimental simulation. However, there is a logical explanation for the results as well. It seems that in addition to placing more weight on distance, women might interpret the use of politeness strategies differently than men. Whereas men might view strategies as a means of preserving the rules of interaction, women might view strategies as a means of administering to others’ needs. Thus, when power and distance are low, male communicators would not see the need for politeness strategies. However, in the same situations, women would be especially sensitive to the close distance of the listener, and would see a greater need to administer to the needs of the close “other” rather than a distant “other.” This finding on female strategy usage is consistent with Gilligan’s (1982) claim that “women do not only define themselves in a context of human relationship but also judge themselves in terms of their ability to care”(p. 17). It is therefore not surprising that women would place more weight on distance differences, and that women would find face-saving work more important in close relationships than in distant relationship^.^

Several implications emerged from the present study. First, in the field of organizational communication, there is value in a multidimensional model of information distortion. A single measurement of whether information is

4The finding that female strategy usage contradicted politeness theory might also be due to the fact that levels of power and distance were only inferred from the hierarchy manipulation, but were not directly measured. Holtgraves and Yang (1992) found evidence that communica- tor’s perceptions of power and distance might be different from objective, manipulated levels of power and distance. They argued that the perceived levels of power and distance were more important in the communicator’s use of politeness strategies than the objective levels. It is possible that men and women had different perceptions of power and distance based on the hierarchical manipulations. However, Holtgraves and Yang did not find any differences between male and female perceptions.

Page 21: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1144 FIONA LEE

transmitted does not fully capture all the different approaches communica- tors may use to transmit the message. This study suggests that three dimen- sions should be explored-informative value, negativity, and strategy usage.

Second, the study offers some new insights on the methodologies for coding politeness. Although the politeness strategies provided by P. Brown and Levinson (1987) were intended to describe general face-threatening acts, researchers interested in specific types of face-threatening acts (such as nega- tive evaluations) could generate more specific strategies that map onto the general strategies. This study used a “bottom-up” approach where strategies were directly extracted from the data, rather than from the theory. The strategies used in this study were more specific and situationally relevant, considerably improving the reliability of politeness coding.

Third, past studies on politeness (Field, 1991; Holtgraves & Yang, 1990, 1992) typically treated power and distance as independent, orthogonal dimensions. These studies were able to separate the effects of power and distance on communication strategies. However, the focus of the present study is not on separating the variables of power and distance, but on looking at communication strategies within specific social situations where power and distance are closely related. This study is different from past research because it recognizes that in some social relationships, power and distance are not as independent and orthogonal as past studies have assumed. Indeed, by creating a hierarchy, one creates differences in both power and in distance. Differences in communication strategies across the hierarchy is not attributed to individual effects of power or distance, but to their combined or relative effects.

Finally, the present study provided initial evidence that politeness theory might not be an accurate model for describing female communicators, directly contradicting past results in politeness research (Field, 1991; Holt- graves & Yang, 1992). This might be due to gender differences in the inter- pretation of the function of politeness strategies. To maintain connections with close “others,” face-saving work is especially important for women when distance is low, not when distance is high as politeness theory would predict. Overall, it makes sense that women and men are not similarly polite across different situations. More studies needs to be done to focus on gender differences in the usage of politeness strategies.

References

Athanassiades, J. (1973). The distortion of upward communication in hier- archical organizations. Academy of Management Journal, 16(2), 207-226.

Bacharach, S., & Aiken, M. (1977). Communication in administrative bureaucracies. Academy of Management Journal, 20(3), 365-377.

Page 22: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1145

Baxter, L. (1984). An investigation of compliance-gaining on politeness. Human Communication Research, 10, 427-456.

Bond, C., & Anderson, E. (1987). Reluctance to transmit bad news-Private discomfort or public display. Journal of Experimental Social Psychology,

Benedict, M., & Levine, E. (1988). Delay and distortion: Tacit influences on performance appraisal effectiveness. Journal of Applied Psychology,

Brown, P., & Levinson, S. (1987). Politeness: Some universals in language usage. Cambridge: Cambridge University Press.

Brown, R. (1990). Politeness theory: Exemplar and exemplary. In I . Rock (Ed.), The legacy of Solomon Asch: Essays in cognition and socialpsy- chology (pp. 23-38). Hillsdale, NJ: Lawrence Erlbaum Associates.

Brown, R., & Gilman, A. (1989). Politeness theory in Shakespeare’s four major tragedies. Language in Society, 18, 159-212.

Cohen, A. (1958). Upward communication in experimentally created hier- archies. Human Relations, 11, 41-53.

Curtis, R. (1989). Cutbacks, management and human relations: Meanings for organizational theory and research. Human Relations, 42(8), 67 1-689.

Dansereau, F., & Markham, S . (1987). Superior-subordinate communica- tion: Multiple levels of analysis. In F. Jablin, L. Putnam, K. Roberts, & L. Porter (Eds.), Handbook of organizational communication (pp. 343- 388). Beverly Hills, CA: Sage.

Field, S. (1991). On saying unpleasant things: An experimental investigation ofpoliteness theory. Unpublished doctoral dissertation, Harvard Univer- sity.

French, J., & Raven, B. (1968). The bases of social power. In D. Cartwright & A. Zander (Eds.), Group dynamics (pp. 259-269). New York: Harper & Row.

Fulk, J., & Mani, S. (1986). Distortion of communication in hierarchical relationships. In M. McLaughlin (Ed.), Communication yearbook 9 (pp. 483-510). Newbury Park, CA: Sage.

Gilligan, C. (1982). In a different voice. Cambridge, MA: Harvard Univer- sity Press.

Glauser, M. (1984). Upward information flows in organizations: Review and conceptual analysis. Human Relations, 37(8), 61 3-643.

Holtgraves, T., & Yang, J. (1990). Politeness as universal: Cross cultural perception of request strategies and inferences based on their use. Journal of Personality and Social Psychology, 59, 7 19-729.

Holtgraves, T., & Yang, J. (1992). Interpersonal underpinnings of request strategies: General principles and differences due to culture and gender. Journal of Personality and Social Psychology, 62(2), 246-256.

23(2), 176-187.

73( 3), 507-5 14.

Page 23: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1146 FIONA LEE

Ilgen, D., Mitchell, T., & Fredrickson, J. (1981). Poor performers: Supervi- sors’ and subordinates’ responses. Organizational Behavior and Human Performance, 27, 386-410.

Jablin, F. (1982). Formal structural characteristics of organizations and supervisor-subordinate communication. Human Communication Re- search, 8(4), 338-347.

Kanter, R. (1977). Men and women of the corporation. New York: Basic Books.

Kaplan, F. (1990, December 5). Three navy officers disciplined for hiding Stealth problems. The Boston Globe, pp. 1, 6.

Larson, J. (1986). Supervisor’ performance feedback to subordinates: The impact of subordinate performance valence and outcome dependence. Organizational Behavior and Human Decision Processes, 37, 39 1-408.

Level, D., & Johnson, L. (1978). Accuracy of information flows within supervisor/subordinate relationships. Journal of Business Communica- tion, 15(2), 12-22.

Lewis, P. (1987) Organizational communication: The essence of effective management. New York: John Wiley & Sons, Inc.

Luthans, F., & Larsen, J. (1986). How managers really communicate. Human Relations, 39(2), 161-178.

Mintzberg, H. (1973). The nature of managerial work. New York: Harper and Row.

O’Neal, E., Levine, D., & Frank, J. (1979). Reluctance to transmit bad news when the recipient is unknown: Experiments in five nations. Social Behavior and Personality, 7( I ) , 39-47.

O’Reilly, C., & Roberts, K. (1974). Information filtration in organizations: Three experiments. Academy of Management Journal, 17(2), 205-215.

Read, W. (1962). Upward communication in industrial hierarchies. Human Relations, 15, 3-15.

Roberts, K., & O’Reilly, C . (1974). Failures in upward communication: Three possible culprits. Academy of Management Journal, 17, 205-21 5.

Rosen, S., Grandison, R., & Stewart, J. (1974). Discriminatory buckpassing: Delegating transmission of bad news. Organizational Behavior and Human Performance, 12, 249-263.

Smith-Hefner, N. (1988). Women and politeness: The Javanese example. Language in Society, 17, 535-554.

Steckler, N., & Rosenthal, R. (1985). Sex differences in nonverbal and ver- bal communication with bosses, peers, and subordinates. Journal of Ap- plied Psychology, 70(1), 157-163.

Synder, B., & Zorn, T. (1986). Communication-related abilities and upward mobility: A longitudinal investigation. Human Communication Research, 12(3), 420-43 1.

Page 24: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1147

Tannen, D. (1990). You just don’t understand. New York: William Morrow. Tesser, A., & Rosen, S. (1975). The reluctance to transmit bad news. Ad-

vances in Experimental Social Psychology, 8, 192-232.

Appendix

Interim Evaluation Given to Subjects as “Bad News”

Interim Evaluation

Subject Code:

Subject Code:

1. Ability to work well with others

1 0 3 4 5 very poor average poor

2. Organizational skills

0 2 3 4 5

very poor average poor

3. Being creative

1 2 3 4 0 very poor average poor

4. Planning the task

1 2 3 @ 5 very poor average poor

5 . Ability to handle a lot of information at once

3 4 5 6 poor average

0 2 very poor

7 good

7 good

7 good

7 good

7 good

9 very good

9 very good

9 very good

9 very good

9 very good -

Continued

Page 25: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

1148 FIONA LEE

Appendix Continued

6. Being calm, not breaking down under pressure

2 3 4 5 6 7 8 9 very poor average good very poor good

7. Confident about own abilities

1 0 3 4 5 6 7 8 9 very poor average good very poor good

8. Ability to give good instructions/ Ability to follow instructions well

1 2 3 4 0 6 I very poor average good poor

9. Decision-making abilities

1 2 0 4 5 6 7 very poor average good poor

10. Being dynamic, vital, and energetic

I 2 3 4 @ 6 7 very poor average good poor

11. Ability to achieve good comprehension of materials

1 2 0 4 5 6 7 very poor average good poor

12. Problem-solving abilities

1 2 3 5 6 7 very poor average good poor

13. Providing structure, coordination

0 2 3 4 5 6 7 very poor average good poor

9 very good

9 very good

9 very good

9 very good

9 very good

9 very good

Page 26: Being Polite and Keeping MUM: How Bad News is Communicated in Organizational Hierarchies

COMMUNICATING BAD NEWS 1149

Appendix Continued

14. Communication skills

1 2 3 4 5 7 8 9 very poor average good very poor good

15. Ability to anticipate upcoming events

1 2 3 @ 5 6 I 8 9

very poor average good very poor good