This research compares formal and informal organizational communication structures, specifically focusir^ on salience, channel factors, and channel usage. The major hypotheses of this research were partially supported with data collected from a large, technically oriented governmental agency (n = 380).
Differences Between Fbrmal and
Informal Conuniinication Channels
J. David Johnson
William A. Donohue
Charles K. Atkin
Mtchtgan State Untverslty
Lansing Community College
lis research focuses on a central problem in organizational comJL munication structures, the relationship between formal and informal structures (Hartman & Johnson, 1990). An organization’s communication structure consists of formal and informal elements, as well as other ingredients, and is not reducible to either (March & Simon, 1958). However, to most organizational researchers this fundamental distinction captures two different worlds within the organization, worlds that have different premises and outlooks and most importantly, different fundamental assumptions about the nature of interaction (AUen, 1977; Dow, 1988).
Very few research studies have attempted tocompare these approaches directly, to assess how they may differ along critical dimensions. A recent attempt to systematically compare formal and informal groupings and their impact on the levels of role ambiguity found more similarities than differences and suggested a complex set of contingencies in which one or the other would have the most impact on organizational variables (Hartman & Johnson, 1990).
A formal structure identifies individuals who are the official sources of information and the information that is their special concern. This has been the traditional view of managers and professional business communicators. Since relationships are determined by one’s role, structure is viewed by managers as a static entity which conforms to a top down configuration (Monge & Eisenberg, 1987). This perspective, which has been termed the configurational view, emphasizes the authoritative coordination of work in the service of stated organizational objectives Ul
TTie Journal ofBusiness Communication
(Dow, 1988). Recent reviews suggest that formal approaches focus on the configurations resultingfiromformal authority relationships represented in the organizational hierarchy (Dow, 1988; Jablin, 1987), from differentiation of labor into specialized tasks (Dow, 1988; Jablin, 1987), andfiromformal mechanisms for coordination of work (Dow, 1988). These characteristics, along with the notion of goal or purpose, have been seen by Schein (1965) as representing the very essence of an organization. On the other hand, informal approaches recognize that a variety of needs, including social ones, underlie communication in organizations and that, as a result, the actual communication relationships in an organization may be less rational than formal systems (Johnson, 1993).
Informal structures function to facilitate communication, maintain cohesiveness in the organization as a whole, and maintain a sense of personal integrity or autonomy (Smelser, 1963). The coactivational perspective recognizes that communication rdationships are not solely based on the positions individuals occupy within formal organizations. Informal groups often arise out of a combination of human needs and formal factors (Schein, 1965). Fbr example, increasingly business communicators have focused on the role of informal communication in generating innovations within organizations (Johnson, 1990).
Scholars in the two camps have completed many research projects, hut have rarely attempted to examine the relationships between the two perspectives, especially in terms of specific organizational factors such as salience, channel factors, and channel usage.
Three means of characterizing salience, personal, effect, and cultural, will be used in this research to compare the relative importance of formal and informal channels.
“Information is valued to the degree it is salient. Salience to an individual means the perceived applicability of information to a problem that he or she faces” (Evans & Clarke, 1983, p. 239). Individuals use particular channels for a variety of motives, including some intensely personal ones, such as securing information related to advancement and merit increases. Using a channel to seek feedback about performance can be seen as an instrumental response on the part of employees who desire to eam positive evaluations (Ashford & Cummings, 1985). Individuals who are skilled at seeking feedback are more likely to be successful (Ashford & Cvunmings, 1985). H i ^ performers within research and develc^ment laboratories have been found to be much more likely to communicate widely within their organizations, using a variety of informal sources (AUen, 1977). Accordingly,
Fbnnal and Informal Channels / Johnson, Donohue, Atkin, Johnson
Hla: Employees will perceive informal channels as more effective than formal channels for advancing their personal agendas.
An individual’s beliefs about the outcomes of channel usage are also important. Individuals ask themselves: Can I do something? A belief that there is no procedure available for con&ondng their problems would impede channel usage. Somewhat relatedly, if an individual believes there are costs (for example, loss of self-esteem) associated with a particular channel, this will reduce bis or her level of channel usage (Bdiga & Jaeger, 1984; Miller & Jablin, 1991; Reinsch & Beswick, 1990). Because of the capacity of information channels to be shaped to fit individual needs and circumstances we expect to find that, Hlb: Employees will evaluate the effect salience of informal channels more highly than formal channels. Cultural Salience
In some cases cultural norms may identify which information is salient and should be gathered in decision m a ^ g situations (Feldman & March, 1981). Strong cultures can also severely restrict the content and the communication partners available to in^viduals. But, because of the sophistication of shared understandings, strong cultuires can increase the effectiveness of communication by clearly delineating roles, relationships, and contexts within which individuals use channels. Thus, a strong, monolithic organizational cultuire would likely enhance perceptions ofthe effectiveness of formal channels of communication. Hlc: Formal channels will be perceived as more culturally salient in achieving organizational goals than will informal ones.
The channel factors are drawn from a model of Media Exposure and Appraisal (MEA) that has been tested on a variety of information carriers and in a variety of international settings (Johnson & Meischke, 1993). These fectors focus on the characteristics and utUity of channels. Characteristics
Characteristics, such as editorial tone £ind communication potential, relate primarily to message content attributes. Editorial tone reflects an individual’s perception of the credibility and intentions of a source. If individuals perceive that a source has motives other than the mere provision of information, then this will weigh heavily in their exposure decisions. Scholars have argued that source credibility and trustworthiness affect anticipatory socialization (Jablin, 1987) and the upward fiow of information in organizations (Glauser, 1984).
The Journal (^Business Communication 31:2 1994
Communication potentitd, the other dimension examined in prior research, refers to an individual’s perception of the manner in which information is presented. This dimension relates to issues of style and comprehension. For example, is an article in a company newsletter visually stimulating and well written? Comprehension has been found to be a critical factor in determining the selection of technical reading material of engineers (Allen, 1977). Because of the mixed-motives often attributed to formal communication (Eisenberg, 1984) and the greater understandability of informal channels.
Hid: Employees will evaluate informal channels more highly than formal ones for channel characteristics of comprehension and credibility. Utiiity
Is the information contained in the channel relevant, topical, and important for the individual’s purposes? Atkin (1973) has argued that mass media exposure will resultfiroma combination of receiver needs and message attributes. Utility is a central concern in determining formal structures that filter the upward fiow of information in organizations, since reducing information overload requires choices about which information should be provided to decision makers (Glauser, 1984). Because of the greater flexibility inherent in the choice of informal channels, which can be tailored more to fit individual needs. Hie: Employees will evaluate informal channels as more useful than formal channels.
Employees use channels in several different ways within formal and informal structures. For example, Lenz (1984) has identified three different dimensions of information seeking: method, scope, and depth. Method relates to the channel selected—in this current research, formal and informal channels. The two major variables examined here related to technical corrununication are scope, in terms of the number of different peoplefiromwhom information is sought, and depth, in terms of the number of different messages involved. Diversity in the number of people contacted and the depth of each contact have both been found to be very important to successful performance within research and development laboratories (Allen, 1977). Here we will argue that the level of individual initiative differs for formal and informal channels, with much more fireedom of action for the individual within informal channels. Accordingly,
H2; In comparison to formal channels, employees will report
that they receive fewer informal-channel messages firom
Formal and Informal Channels / Johnson, Donohue, Atkin, Johnson fewer people and send more informal-channel messages to a larger number of people.
The primary research question this study addresses concerns the different perceptions of various organizational factors and how those relate to formal and informal communication channels. As we have seen formal and informal channels have inherently different capabilities, they also have different historical patterns of development and usage, which should result in individuals evaluating them differently. METHODS
This research was conducted in a large midwestern state governmental agency that is charged with providing engineering and technical services. This study was part of a much larger project designed to comprehensively assess current communication practices, to make recommendations, and to implement change strategies designed to improve both internal and external communication within this agency. Before developing the questionnaires thirty employees, selected ficom various levels and functional divisions within the organization, were interviewed at length (30 minutes to three hours duration) by the coauthors. After these interviews and a review of appropriate agency documents (for example, mission statement, standard operating pjrocedures) questionnaires were pretested on agency employees.
C!opies of the final version of the questionnaire were distributed to bureau executives who then distributed them to their administrators and supervisors. In tum, administrators and supervisors distributed questionnaires to employees along with plain, brown envelopes in which the questionnaires were to be returned. The surveys included a cover letter from the Interim Director of the agency stressing the importance of the survey and the confidentiality of the responses. The employees returned the sealed envelopes to their supervisors who then returned them to the researchers.
The respondents (n=380) were characterized as follows: the median report of education was college graduate, with 18 percent reporting at least some post graduate education. Thirty-three percent reported they were engineers; 32 percent, other technical specialties; and 15 percent, administrative. Eighty-seven percent of the respondents were Caucasians, 6 percent were Afiican-Americans, and 7 percent reported other or did not report. Eighty percent of the respondents were male. There was a 26 percent refusal rate among the 513 individuals drawn for the total sample.
All psychometric scales were based on eleven point bipolar type items embedded in a larger questionnaire. Questions for each of lAie scales
The Journal ofBusiness Communication 31:2 1994
were developed based on items derived fixtm the academic literature; firom the initial exploratory, qualitative interviews; and firom formal agency documents. The actual question wordings for the final individual items are
contained in Table 1.
Scale Items, Factor Loadings, and Reliabilities
for Formal and Informal Channels
Personal Salience (alpha = .78, .80)
When I communicate with this channel I get recognition
which can help me in getting promoted,
When I communicate with this channel I get recognition
which can help me in getting a merit increase,
Effect Salience (alpha = .91, .88)
It is pointless to communicate using tUs
channel since no one listens.
Nothing ever happens when I communicate by
means of this channel,
I dont get information helpful to me
in performing my job if I use this channel.
Cultural Salience (alpha = .91, .82)
Using this channel I can help _
in performing its missions,
By using this channel I can help
better serve the public,
By using this channel I can help
inform the public,
By using this channel I can help people in
omer units do their jobs,
By iising this channel I can develop teams of people
to tackle spediic problems within
Characteristics (alpha = .86, .85)
Accurate (10)-Inaccurate (0)
Well-intentioned (10), questionable intentions (0)
Very understandable (10), very difficult to understand (0)
Material clearly presented (10), not deiu-ly presented (0)
Utility (alpha = .85, .84)
Personally significant (10), not personally significant (0)
Very useful (10), not at all useful (0)
Very easy to get information from (10),
very difficult to get information (0)
Ebrtremely valuable source of information (10),
not valuable at all (0)
Note. Effect salience items were reverse coded. Items that show no value in
the “informal” coliimn were omitted because of parallelism problems.
Formal and Informal Channels / Johnson, Donohue, Atkin, Johnson
The measurement model was analyzed by means of the confirmatory &ctor analysis subroutine of the PACKAGE computer p n ^ a m (Hunter & Iim, 1987). Confirmatory £actor anal}rsis is a superior technique when the a priori specification of items expected to duster t ( ^ t h e r is possible (Pink & Monge, 1985; Hunter & Grerbing, 1982). We assessed unidimensionality with the three criteria proposed by Hunter (Hunter, 1980; Hunter & Gerbing, 1982), homc^neity of item content (fece validity), internal consistency (for example. Spearman product rule), and parall ^ s m (for example, “flat” corrdation matrix). In addition, ihe scree test was also used to determine the unidimensionality of factor structures (Van de Geer, 1971). Tbsts of unidimensionality are essential to scale development since alpha provides an unbiased estimate of reliability only if scale items are unidimensional (Hunter, 1980; Hunter & Gerbing, 1982). Only psychometric scales that met these criteria were included in thefilialtests. Naturally, during the process of scale development, items that did not meet the psychometric criteria were dropped from the respective scales. Respondents were given the following instructions to help them in distinguishing between formal and informal commimication for each scale item:
Informal communication usually does notfollowtheorganizational chart and tends to be more personal, fbr example, work-related discussions with co-workers, calling Mends in another work unit on how to handle a work problem, etc.
Formal communication is considered to be “official” such as oral communication up and down theorganizational chart and written communication contained in formal memoranda and departmental directives. Table 1 presents actual question wordings and the results for the psychometric scales for salience. The personal scale (2 items, alpha = .78 for formal and .80 for
informal), the effect scale (3 items, alpha = .91 for formal and .88 for informal), and cultural scale (5 items, alpha = .91 for formal and 3 items, alpha = .82 for informal) were used as indicators of salience. The cultural scale items were drawnfix)mthe agency’s Mission Statement which had been recently developed. Both the process of development and the dissemination of the Mission Statement have made values very central to the lives of organizational members, a conclusion that is buttressed by the findings of the exploratory interviews. Scores for this scale were adjusted to refiect the differing number of items available for each channel.
Channel factors (see Table 1) were represented by two scales: characteristics (4 items, alpha = .86, .85) and utility (4 items, alpha = .85, .84) (Johnson & Meischke 1993).
The Journal ofBusiness Communication 31:2 1994
Channel usage was operationalized by items refiecting the number of peoplefromwhom technical information was initiated and received and tbe number of messages sought and received.
T-tests were used to determine if there were significant differences for each of the separate organizational factors between channels (See Table 2). Hla was not supported (p > .05), with very low means and an exceptionally high Pbarson correlation (r = .84) between personal salience for the two channels. Hlc, cultural salience, and Hid, channel characteristics, had significant differences, although not in the direction predicted by the hypotheses. H2, which dealt with messages initiated and received and the numbers of people involved was only supported for the number of messages initiated. Hlb, effect salience, and Hie, utility, were supported as originally predicted.
T-Test8 for the Hypotheses
for Formal and Informal Channels
iVote. Results supported hypotheses concerning Effect Salience (Hlb), Utility (Hie), and Messages Initiated (H2). Resxilts were statistically significant but contrary to prediction for hypotheses concerning Cultural Salience (Hlc), Characteristics (Hid), and Messages Received (H2). * p