“While You Still Think,I Already Type”: Experienced Social Power Reduces Deliberation During E-Mail Communication

Abstract

E-mail allows individuals to deliberate on their communication before sending it off. For instance, communication partners can easily take their time to ponder how best to frame a request before they actually send a message. Individuals at times strategically exploit this opportunity to deliberate in order to tailor messages to their communication partner, such as when communicating with a relatively more powerful person. As social power reduces concerns about impression management, we predicted that individuals deliberate more while composing e-mail messages under low (vs. high) power. This assumption was tested with well-established power priming. As such, we expected that experienced power in one context would diminish deliberation times during a subsequent e-mail communication. An experiment manipulating the experience of (low vs. high) power and measuring deliberation times during e-mail composition supported this hypothesis. The findings thus indicate how social power alters deliberation times. More importantly, the results show that individuals not only strategically deliberate during e-mail communication in line with their current situation, but also in line with their social standing in a previous situation (here, their experience of power).

Introduction

Imagine a student and a professor, each composing an e-mail to the other in order to obtain information on a shared project. While composing this e-mail, the student and the professor likely differ in their behavior: the student having low power and depending on the professor might carefully construct the message and think about potential side effects (e.g., how to make a good impression). In contrast, the professor experiencing high power will probably not devote as much time to deliberation, but promptly finish the request to get the information.

As this example illustrates, e-mail allows communication partners to deliberate on their behavior and selectively construct messages.^1,2 Individuals at times strategically do so to pursue their immediate communication goals (e.g., obtaining relevant information), but also to take care of their strategic concerns (e.g., making a good impression on the other person).² These strategic concerns, such as impression management, are more relevant when depending on one's communication partner—that is, in the context of social power, when one individual has asymmetric control over the outcomes of another.³ Being powerless (vs. powerful) increases concerns about how one is being evaluated by others.^4–7 Individuals might thus take more time to deliberate on their e-mails under low (vs. high) power.

Though this argument^1,2 relies on power differences, the impact of power on deliberation during e-mail communication has never directly been studied. Therefore, we examine the effect of power with a well-established procedure from psychological research: inducing an experience of power, based on priming. Research indicates that subjectively experienced power in one situation can carry over to an unrelated context (e.g., subsequent e-mail communication). Experienced power guides subsequent behavior identically to actually possessed power.⁸ This suggests that individuals might not only strategically deliberate more when communicating with a relatively less (vs. more) powerful partner, but should also spontaneously think more when they experience low (vs. high) power. The present research thus investigated how the experience of power affects deliberation times during an unrelated e-mail communication.

Strategic deliberation during e-mail communication

Like other asynchronous media, e-mail enables individuals to simulate actions (e.g., posing a request), to weigh different strategies (e.g., ways to phrase it), and to anticipate likely outcomes (e.g., whether it will be granted) before and while composing a message—more specifically, before getting started and while typing.¹ Thereby, individuals can selectively construct messages and exchange information with more forethought.^1,2,9

Individuals use this opportunity for deliberation when striving for impression management with their communication partner.² Similarly, an e-mail may be phrased more politely when including high (vs. low) imposition requests to a powerful counterpart (compared to voicemail).⁹ Taken together, individuals strategically devote time to deliberate on their e-mails in line with their current goals: if the writer's goal is to make a good impression on the recipient, the writer will think more while composing the e-mail. Such concerns can also be activated by the experience of power.

How social power facilitates spontaneous deliberation

Compared to elevated power, being powerless enhances concerns about impression management^4,7 and may thereby alter thinking during e-mail composition. Based on psychological research on social power, however, we argue in the following that feeling powerless can also spontaneously promote deliberation across different situations.

Powerless individuals depend on the resources that power holders control (e.g., food, money, praise, information),³ and are more restricted by social constraints, whereas power holders are independent and have greater freedom to act. Being powerless thus activates inhibition tendencies and a focus on potential threats or punishments, whereas having power evokes approach tendencies and a focus on rewards.⁷ As a consequence, power promotes action and directs attention to one's focal goals.^5–7

Indeed, individuals low in power expect goal attainment to take longer,¹⁰ perceive more obstacles on the way,¹¹ react more sensitively to potential negative results,¹² and follow other's expectations more strongly than power holders.¹³ These results indicate that power promotes disinhibited action, presumably without too much deliberation over how one might be evaluated by others.^7,14 Applied to e-mail communication, a powerless person likely ponders potential consequences of his/her behavior, while a power holder might quickly compose a message to receive the information needed. We thus expect that individuals deliberate more on their e-mails under low (vs. high) power.

Do individuals actually need to possess low (vs. high) power within the e-mail context for this effect to occur? The previously described effects do not only apply to situations in which actual power is provided. Power represents a feature of a social relation between individuals,^3,7,15 but it is also mentally represented—that is, the concept of power is strongly associated with a set of affective, cognitive, and behavioral reactions.¹⁴ Following this logic, whenever the experience (or concept) of power is activated, the behavioral tendencies associated with power also become activated: when individuals experience power in one situation,^5,14,16 they subsequently think, feel, and act in the same way as if they actually possessed power.⁸ These effects do not depend on prior experience with power: manipulated power (e.g., among power-inexperienced students) similarly guides behavior as power in real-life settings (e.g., among supervisors vs. subordinates).^10,17–20

Taken together, individuals experiencing low (vs. high) power in one situation generally become more inhibited and concerned about outcomes. Consequently, they should engage in more deliberation in a subsequent, unrelated e-mail communication. We thus predict that the experience of power in one situation impacts deliberation times in another (e-mail) context. As such, the effect of power on deliberation times should emerge spontaneously (i.e., though individuals are not currently communicating with a more or less powerful e-mail partner).

Findings in support of this hypothesis would provide, for the first time with a behavioral indicator, evidence for the idea that power diminishes thinking before initiating action.¹⁴ More importantly, research on e-mail communication would be extended: previous research has largely focused on how strategic concerns related to the e-mail communication (e.g., the imposition of a request, the partners' status or desirability)^2,9 guide behavior. Similarly, prior research indicated, for instance, how power alters attitudes toward using e-mail in the first place²¹ or affects negotiation outcomes,²² as well as how (face-to-face and virtual) communication behavior vice versa signals power²³ or promotes support toward a power holder.^24–27 Our research goes one step further, investigating how writers' use of e-mail properties might depend on their social standing beyond the e-mail context—here, their experienced social power, activated by the previous situation. The present research could thereby contribute to an understanding of why—in a factually identical e-mail setting—one communication partner might still deliberate more than the other.

The Present Study

We predicted that low (vs. high) power promotes deliberation times while composing e-mail messages in an unrelated context. This prediction was tested in an experiment in which we first activated the experience of low versus high power. Following standard procedures, participants received a powerless or a powerful role on a dyadic creativity task.^5,14,28

We then measured deliberation times in an unrelated business setting, in which participants composed e-mail requests to fictitious receivers. These receivers were identical for all participants and thus did not depend on nor relate to participants' power in the previous (creativity task) context. This design enabled us to investigate whether the experience of power affects subsequent deliberation times, independently of individuals' power relation to their current communication partners. As such, this experiment studied the impact of spontaneous, rather than strategic, effects of power on deliberation.

Design and participants

This study implemented two experimental conditions (low vs. high power). Forty-nine undergraduates (38 female; M_age=23.65 years; range 18–31 years) participated as part of a 1 hour study package in return for €8 (approximately US$11).

Procedure

This study was presented as two unrelated studies to separate the power manipulation from the e-mail context. The first study activated the experience of power, the second study assessed deliberation times during e-mail composition. Power was manipulated via role assignment with an established procedure.^5,28 Participants randomly received a powerful (manager) or powerless (employee) role. They learned that they would solve a creativity task in dyads of one manager and one employee with a partner from another lab room. They filled in an ostensible leadership questionnaire and received bogus feedback. Participants learned that some people (i.e., employees) are especially capable of following instructions and solving tasks, whereas others (i.e., managers) are good at giving instructions and telling others what to do. Accordingly, the employee's task was to generate solutions, which were to be judged by the manager. Managers would receive a fixed amount of €3 for this study; employees would receive €2. The manager could supposedly let the employee earn up to €3 as well, depending on the manager's evaluations. As such, power implied asymmetric control over the partner's outcomes. Manipulations such as these activate the experience of power (e.g. similar to recalling a powerful/powerless experience,¹⁴ or sitting on a powerful/powerless person's office chair¹⁶) and have been extensively validated, resulting in similar effects as when indicators for actual power are examined.^17,18,20,29 We then assessed a power manipulation check.

Afterwards, participants saw on screen how they were supposedly connected to another lab. They learned that their partner was currently still involved in another study but would be available later. Participants were asked to continue with the second study while waiting, and to finish the creativity task afterwards.^5,28

For this “unrelated” second study, they composed four e-mails in a business setting. Participants imagined working in a large company. Their task was to compose e-mails to (fictitious) communication partners within the company, whose relative position was not specified. Importantly, the receivers of these e-mails were thus identical for all participants. Their e-mails requested, in counterbalanced order: (a) a room reservation for an already booked room, (b) a flipchart for a meeting from the owner, (c) a library book from the person who had lent it, and (d) assistance on a software error. While composing their requests, the computer recorded the total composition time, the contents of sent e-mails, and deliberation times.

Measures

Power manipulation check

Participants indicated how much in charge they would be and how much influence they would have in the creativity task (1=“not at all” to 7=“very much”), r(49)=0.72.

E-mail communication

For all e-mails, total composition time (i.e., from opening a new e-mail by clicking a “new e-mail” button to sending it by clicking the “send” button) and content length (i.e., number of characters in the sent e-mail) were recorded. As indicators for deliberation during e-mail composition, a deliberation index was assessed by dividing the time spent on thinking before starting to type (i.e., pressing the first key) by the total composition time. Similarly, based on the time spent on actually typing (i.e., action), we calculated a no-deliberation index. For this purpose, we divided the time participants actually typed e-mail text (i.e., keys pressed at least every 2 seconds) by the total composition time. Higher numbers on the no-deliberation index thus indicate less time spent on thinking and more time spent on typing. Note that these two indices do not add up to the value 1, as total composition time also included time for not typing before sending the message.

Exploratory measures

In addition, we assessed self-reported deliberation. After sending each message, participants completed two items (e.g., “While composing this e-mail request, I thought about what I could frame differently”: 1=“completely disagree” to 7=“completely agree”). These items were embedded within filler items and were highly correlated for all e-mails, r(49)=0.85–0.91.

Regarding potential implications of deliberation times, we explored the quality of participants' composed e-mails. Two independent raters blind to hypothesis and power condition rated message persuasiveness (as performance indicator) and message politeness (1=“not at all persuasive/polite” to 7=“very persuasive/polite”). For message persuasiveness, raters focused on how likely they would be to agree to the request. Message politeness referred to how politely the request was framed. The codings were correlated—r(49)=0.31 and 0.43; p<0.001—and thus averaged.

Results

Preliminary analyses

Participants reported less power in the low power (M=3.35, SD=0.92) than in the high power condition (M=5.35, SD=0.99), t(47)=7.29, p<0.001, d=2.00, indicating that the manipulation was successful.

Power affected neither total composition time, t(47)=0.46, p=0.652, d=0.13, nor content length, t(47)=1.04, p=0.304, d=0.30. Thus, participants in both conditions spent a similar total time on composing e-mails and sent e-mails with a comparable length.

Main analyses

We hypothesized that power lowers deliberation times. A 2×2 (Power: low vs. high×Index: deliberation vs. no deliberation) mixed model analysis of variance with repeated measurement on the last factor yielded a significant main effect of index, F(1, 47)=5533.09, p<0.001, η_p²=0.99, qualified by the expected power–index interaction, F(1, 47)=5.07, p=0.029, η_p²=0.10. As predicted, low power participants spent relatively more time on deliberation (M=0.090, SD=0.05) than high power participants (M=0.065, SD=0.03), F(1, 47)=5.31, p=0.026, η²=0.10. Conversely, high power participants used their time more for not deliberating (i.e., more for typing; M=0.899, SD=0.03) than low power participants (M=0.876, SD=0.05), F(1, 47)=4.06, p=0.050, η²=0.08. Taken together, experiencing low (vs. high) power promoted the time invested in deliberating before getting started and while composing the e-mails.^a

Additional analyses

Results for self-reported deliberation showed a trend in line with predictions: low power participants reported deliberating more (M=5.73, SD=0.64) than high power participants (M=5.18, SD=1.43), t(35, 679)=1.76, p=0.088, d=0.50, Levene's test, F(1, 47)=17.18, p<0.001.

Power did not affect message politeness, t(46)=0.32, p=0.753, d=0.09, or message persuasiveness, t(46)=1.51, p=0.138, d=0.43. If anything, there was a slight trend that higher deliberation times among the powerless (vs. powerful) rather harmed than helped persuasiveness. When controlling for message politeness, power even enhanced message persuasiveness, F(1, 45)=4.75, p=0.035, η²=0.06. Low power participants communicated their requests less convincingly (M=5.44, SD=0.90) than high power participants (M=5.76, SD=0.56).

The correlation between the no-deliberation index, the deliberation index, and message persuasiveness indicates that not deliberating too much but instead using the time to type contributed to message persuasiveness (as an indicator for performance; see Guinote²⁸ and Sanna³⁰ for similar arguments on how deliberation may harm performance), r(48)=0.292, p=0.044, whereas deliberation did not necessarily do so, r(48) = −0.185, p=0.209.

Discussion

E-mail provides the opportunity to deliberate on one's communication behavior,¹ which individuals strategically use when required (e.g., to communicate an imposing request to a power holder).⁹ To extend these findings, the current research examined if experienced power alters deliberation times during an unrelated e-mail communication. More specifically, we sought to demonstrate that individuals at times also spontaneously deliberate, independently of the context in which they are currently communicating. Based on psychological research on power, we proposed that experiencing low (vs. high) power promotes deliberation not only for strategic purposes (e.g., to make a good impression on the power holder with whom one is communicating). As power evokes a general propensity toward less thoughtful, inhibited behavior beyond the given power context,^7,8 we predicted that those experiencing low (vs. high) power deliberate more during e-mail communication. The findings supported this prediction.

As a major strength, thinking times during e-mail composition^1,2 as a behavioral indicator were assessed without participants being aware of it, thus ruling out potential demand effects. This pattern was mirrored in self-reported deliberation: individuals also tended to report more thought when experiencing low rather than high power. Interestingly, as high and low power participants spent a similar total time on composing requests, the results in sum indicate that individuals experiencing high (vs. low) power used their time more efficiently to compose convincing requests.

Extending power research, the current findings indicate that the general tendency to act promptly, which power comes with,^14,28 is paralleled by a reduced tendency to deliberate on the potential outcomes. Most importantly, the findings have implications for research on e-mail communication. Prior research on deliberation largely focused on the functional use of opportunities provided by e-mail, that is, how individuals strategically adapt within the current communication context.^2,9 Extending these findings, we showed that deliberation times can also depend on individuals' social standing beyond the current e-mail communication—here, their experienced social power. In other words, communication partners at times also deliberate longer without being strategically motivated to do so.

This effect is likely not limited to (experienced) power, but should generalize to any relational experience or social role that evokes a mental state applicable to the e-mail context. For instance, an experienced competition (vs. cooperation) may activate a competitive thinking style in one situation³¹ and thereby guide impression management concerns in an unrelated, subsequent setting. Similarly, individuals' motivation to avoid mistakes (rather than approach gains)^32,33 or create a positive self-image relative to others^34,35 in one context (e.g., when performing in class) may spill over to a subsequent situation, potentially making individuals more thoughtful in e-mail exchanges. As such, our research certainly only made a first step toward an integrative approach on how experiences from previous contexts predict the subsequent use of media properties.

The exploratory results on message persuasiveness point toward the question of when deliberation is (not) useful: forethought can help to anticipate drawbacks and optimize messages.^30,36,37 As the powerless use this deliberation opportunity more extensively, they might at times communicate more effectively. This, however, may be limited to situations that actually require individuals to construct their messages carefully, such as when formulating an especially imposing request⁹ or composing a message with potential large-scale implications. This was not the case in our study, which might explain why deliberation did not help but rather harmed persuasiveness. In contrast, a tendency to deliberate when experiencing low power throughout situations might as well have dysfunctional effects (e.g., missed opportunities to act^14,28). In this regard, our findings on powerless individuals' lower message persuasiveness are in line with results that the powerless are often more distracted and perform worse than power holders.^6,28 Future research should investigate these potential virtues and vices of the impact of power on deliberation. Power holders' more persuasive communication might, for instance, in the long run imply that they use communication to stabilize hierarchical relations (e.g., receive more support from followers).

To conclude, e-mail communication partners at times spontaneously adapt deliberation to their experienced social standing, such as power relations within another situation. Therefore, to understand when individuals use the properties provided by e-mail, researchers may need to take both individuals' strategic concerns and their experiences activated by previous contexts into account.

Notes

a. The results indicate that high power participants spent relatively more time on typing, whereas low power participants spent relatively more time on deliberation during typing. Despite this effect of power condition, participants' e-mails in both conditions had a similar content length, which may have resulted from the fact that typing also included deleting and overwriting text passages.

Footnotes

Acknowledgments

We thank Sandra Herrmann and Konstantin Schmid for their help in data coding, and Manfred Knobloch for his help in programming the e-mail client.

Author Disclosure Statement

No competing financial interests exist.

References

Walther

. Computer-mediated communication: impersonal, interpersonal, and hyperpersonal interaction. Communication Research, 1996; 23:3–43.

Walther

. Selective self-presentation in computer-mediated communication: hyperpersonal dimensions of technology, language, and cognition. Computers in Human Behavior, 2007; 23:2538–2557.

Fiske

, Berdahl

. (2007) Social power. In Kruglanski

, Higgins

, eds. Social psychology: handbook of basic principles. 2nd ed. New York: Guilford Press, pp. 678–692.

Fiske

. Controlling other people: the impact of power on stereotyping. American Psychologist, 1993; 48:621–628.

Guinote

. Power affects basic cognition: Increased attentional inhibition and flexibility. Journal of Experimental Social Psychology, 2007; 43:685–697.

Guinote

. (2010) The situated focus theory of power. In Guinote

, Vescio

, eds. The social psychology of power. New York: Guilford Press, pp. 141–173.

Keltner

, Gruenfeld

, Anderson

. Power, approach, and inhibition. Psychological Review, 2003; 110:265–284.

Smith

, Galinsky

. The nonconscious nature of power: cues and consequences. Social and Personality Psychology Compass, 2010; 4:918–938.

Duthler

. The politeness of requests made via email and voicemail: support for the hyperpersonal model. Journal of Computer-Mediated Communication, 2006; 11:500–521.

10.

Weick

, Guinote

. How long will it take? Power biases time predictions. Journal of Experimental Social Psychology, 2010; 46:595–604.

11.

Whitson

, Liljenquist

, Galinsky

, et al. The blind leading: power reduces awareness of constraints. Journal of Experimental Social Psychology, 2013; 49:579–582.

12.

Inesi

. Power and loss aversion. Organizational Behavior and Human Decision Processes, 2010; 112:58–69.

13.

Galinsky

, Magee

, Gruenfeld

, et al. Power reduces the press of the situation: implications for creativity, conformity, and dissonance. Journal of Personality and Social Psychology, 2008; 95:1450–1466.

14.

Galinsky

, Gruenfeld

, Magee

. From power to action. Journal of Personality and Social Psychology, 2003; 85:453–466.

15.

Emerson

. Power-dependence relations. American Sociological Review, 1962; 27:31–41.

16.

Chen

, Lee-Chai

, Bargh

. Relationship orientation as a moderator of the effects of social power. Journal of Personality and Social Psychology, 2001; 80:173–187.

17.

Fast

, Chen

. When the boss feels inadequate: power, incompetence, and aggression. Psychological Science, 2009; 20:1406–1413.

18.

Gruenfeld

, Inesi

, Magee

, et al. Power and the objectification of social targets. Journal of Personality and Social Psychology, 2008, 95:111–127.

19.

See

, Morrison

, Rothman

, et al. The detrimental effects of power on confidence, advice taking, and accuracy. Organizational Behavior and Human Decision Processes, 2011; 116:272–285.

20.

Scholl

, Sassenberg

. Where could we stand if I had…? How social power impacts counterfactual thinking after failure. Journal of Experimental Social Psychology, 2014; 53:51–61.

21.

Linjun

, Ming-Te

, Wong

. The impact of power distance in email acceptance: evidence from the PRC. Journal of Computer Information Systems, 2003; 44:93–101.

22.

Johnson

, Cooper

. Power and concession in computer-mediated negotiations: an examination of first offers. MIS Quarterly, 2009; 33:147–170.

23.

Wakslak

, Smith

, Han

. Using abstract language signals power. Journal of Personality and Social Psychology, 2014; 107:41–55.

24.

Fan

, Chen

, Wang

, et al. E-leadership effectiveness in virtual teams: motivating language perspective. Industrial Management & Data Systems, 2014; 114:421–437.

25.

Mayfield

, Mayfield

. The role of leader motivating language in employee absenteeism. Journal of Business Communication, 2009; 46:455–479.

26.

Menegatti

, Rubini

. Convincing similar and dissimilar others: the power of language abstraction in political communication. Personality and Social Psychology Bulletin, 2013; 39:596–607.

27.

Seyranian

. Social identity framing communication strategies for mobilizing social change. Leadership Quarterly, 2014; 25:468–486.

28.

Guinote

. Power and goal pursuit. Personality and Social Psychology Bulletin, 2007; 33:1076–1087.

29.

Ferguson

, Ormiston

, Moon

. From approach to inhibition: the influence of power on responses to poor performers. Journal of Applied Psychology, 2010; 95:305–320.

30.

Sanna

. Defensive pessimism, optimism, and stimulating alternatives: some ups and downs of prefactual and counterfactual thinking. Journal of Personality and Social Psychology, 1996; 71:1020–1036.

31.

Sassenberg

, Moskowitz

, Jacoby

, et al. The carry-over effect of competition: the impact of competition on prejudice towards uninvolved outgroups. Journal of Experimental Social Psychology, 2007; 43:529–538.

32.

Carver

, Scheier

. Origins and functions of positive and negative affect: a control-process view. Psychological Review, 1990; 97:19–35.

33.

Gray

. Brain systems that mediate both emotion and cognition. Cognition & Emotion, 1990; 4:269–288.

34.

Butler

. What young people want to know when: effects of mastery and ability goals on interest in different kinds of social comparisons. Journal of Personality and Social Psychology, 1992, 62:934–943.

35.

Wood

. Theory and research concerning social comparisons of personal attributes. Psychological Bulletin, 1989; 106:231–248.

36.

Armor

, Taylor

. The effects of mindset on behavior: self-regulation in deliberative and implemental frames of mind. Personality and Social Psychology Bulletin, 2003; 29:86–95.

37.

Pham

, Taylor

. From thought to action: effects of process- versus outcome-based mental simulations on performance. Personality and Social Psychology Bulletin, 1999; 25:250–260.