zation, the National Bureau of Standards, which would result in a profile of ..... presented most content areas covered in employee publications.'⢠(A list ·of.
Public Relations A journal of IC.c!lictarch and Conuucnt Winterl977
Vol.3, No. 4
Measurement in Public Relations-An (h-crvicw
james E. Grunig
5
Corpomte "Cse of Re:..:arch In Public Relations
Ouo Lerbinger
11
Measuring The Bell Sy~;tcm's Public Relations
James F. Tirone
21
~lcasuring the Effects Of Corporate Ad,·crti~:~inl!
Robert C. Grass
39
ll lcastu·ement. S ystem
Ro bert K. Marker
iii
Evaluating E mployee Communicallon In A Research Operalion
James E. (i runif'!
61
An XDS Internal Communications Study: A Comment
RichardS. Franzen
83
~laxwell ~lcCombs
89
Phillipj. Tichenor, George A. Donohue, and Clarice N. Olien
~)6
The Armsl rong'PR Data
Agenda Selling Function
onlass ~ledia
Community Research and E,·aluating Comnlllnil y Relations
Coorientm.ionul .M cas1 1remcnt.
nf Public Issues
G len tv!. Broom
llO
Stmtcgics for F. valuating Public Relations
Keith R. Stamm
120
Public Relations Evaluath·c Research: Summary Statcrncnl
~lark
129
P. McElreath
1
Evaluating Employee Communications in a Research Operation TJre mOSl common way of conducting r"L.lttluutive researciJ is to take a program as gwen and measure and euuluatt its con.sequences. Gfwrig says m t/115 article. Tf1r problem with this approach. he add:,. i.!> tlwt tit~ program beir~g c-valuat£>d may not have be:en tlu?most appropriate i~1 the /lrst place. Instead. he presc>nts a nl('tllOtlology for dtnJeioping a profile of tlr4'tommutJkation problems and informatiou nrvd.s with wlriclr thi! f>R practitiat1£'r must dl. a,td lt might St4ggcst new proJ{rams which tht prarri tioner had not c:tmsidl'r~d. Grunig outlines this approach and pres(mts originul results from a study of the e-mployee communication program Cll th~ Nil tiona} Bureau o{Sumdards. titinR it as rm example of how tit£- sam£- approach can b4' used in other- PR ~ttings. In th£- followhJg article, Richard Franuu of the National Bureau of Stauda.rds discus~s how tllis approach worked at the Burt"nu GrwJig holds a PI!. D. in Mil$$ CommunicAtion from thr Univt!'t'sityof WiSCOIISm and is Associa" Pro/.ssor of /ournali1 th• University of M.>ryland.
T
he objective of employee communications in a large research organization can be characterized as the need to integrate two divergent pattems of communication behavior. On the one hand, the organization must attempt to facilitate communication between subsystems of the organization if it is to maintain itself as an integrated system. On the other hand, the organiz,uion generally must take on a decentralized structure and grant its employees job autonomy if the organization is to produce research innovations.' Autonomy. however, usually means that employees have the f.rmom to choose their own communication pattems. As a result, employees generally will communicate more with employees with similar work interests than they will participate in system-wide networks that make coordination possible.'
61
Public Relation~ Rcvi.,,.. To add to the problem, research organizations do not consist entirely of scientists and professional employees. They also employ large numbers of technicians, administrative and clerical assistants, and other support personnel. Whereas professional employees are best managed through a decentralized structure, the support personnel are generally most productive if they are managed through a centralized structure which grants them less autonomy. As a result, the most frequent communication among professionals is probably horizontal communications with peers or two-way communication with superiors, while the most frequent communication among support personnel is probably downward communication of orders and instructions from superiors to subordinates. To the manager of employee communications thus falls the problem of reconciling the communication needs of the organization and the conflicting communication patterns of employees. The results of these efforts, however, are often haphazard. Sometimes employee media proliferate as the result of efforts to meet all of the divergent information needs of employees . At other times media follow traditional patterns (such as the babies and bowling scores format of an employee newspaper) which reach some employees but not others.• This inquiry was designed as an applied study of one large research organization, the National Bureau of Standards, which would result in a profile of the organization's employee publics and their information needs, and which would provide an evaluation of the organization's employee media. • Its results go beyond those of most applied studies, however, and provide insight into the coordination problem o f a research organization, and into the different kinds of employee information needs in such an organization.
The Study Design Applied research may be approached in at least two ways. The most common approach is to take a set of practices, or programs, as given and then observe the effects of those programs. The research results can then be used to revise the programs in order to promote desirable and reduce undesirable consequences. One shortcoming of this approach is that the programs studied may not have been the most appropriate in the first place. When existing programs are evaluated in isolation, alternative programs are seldom considered, and rarely will the results of the research suggest new programs. A second shortcoming is that the results usually have application only in the situation from which they came and cannot be applied to other situations. In this study we took a second approach to applied research. We believe that an applied researcher should begin with a theory which, when measured in a research setting, would result in a profile of communication problems and needs which the practitioner must deal with. This profile can then be used as a yardstick to explain why some existing programs are effective and others are not. It might also suggest new programs which the practitioner may not previously have considered .
62
- ·~r
Evaluating Employee Communication
liSt entirely ~numbers
of of · ~pport per· 't>agh a decen· c.:ctive if they
-
less auton-
ror-essionals is
:;::-.:nunication ! support per-
:"Xtions from
w problem of
conflicting however, • "~ result of J'ftS. At other ""'1ing scores r"ft:S but not :ts
llral'dt o rgani-
m ~ profile of
i> a.Ttd which 1.:2 • Its results lo InSight into • .!ht different
hr :nost com-
Mid then obm be used to
• .r.'ldesirable v=s studied '!,en existing seldom conprograms. A yin the situ-
INIS.
L
\\'e believe measured
pd others are otr may not
The theory used in this study was designed to explain when people will communicate and the nature of their communication behavior when they do communicate. The theory has been used to explain the communication behavior of several different kinds of systems.' The most important application for this study, however, has been its use in identifying and targeting publics on the basis of similarity in people's communication behaviors.• Two studies of employee communication have already been reported using the theory.' The assumption· behind the theory is that a professional communicator should concentrate on communicating with publics which are most likely to be communicating with him, and that he should provide information to publics which are most likely to need and to seek that information. Information cannot affect the behavior of people unless they first process that information . Therefore, any communication program designed for people who are unlikely to process the information represents a waste of resources. In this study, the theory should make two conclusions possible. It should first indicate whether organization-wide communication is possible in a research organization or whether the communication of employees in such an organization is limited to information directly related to their work. Secondly, it should allow identification of specialized employee publics with in the organization, each of which will have different communication behaviors, information needs, and patterns of using employee media. The theory also assumes that communication is a situational behavior- i.e.,
that people communicate as a consequence of the way they perceive a specific situation and not because of cross-situational personality traits, attitudes, etc. Thus, the theory allows a researcher to do two things. First, by looking at how different people perceive several situations, he can predict which people will seek information about each of the situations. Second, by looking at the nature of those situations, he can predict which kind of information wiU be needed by people who are actively communicating-i.e., they generally need information relevant to the situations about which they are actively communicating. In this study, the theory was used to isolate different types of NBS employees with different kinds of information needs. These e~ployee types were then compared by their demographic characteristics in order to make it possible to identify them . They were then compared by their media use, their content preferences for NBS media, and their knowledge of news items appearing in NBS publications. In order to understand the results of the study, it is useful at this point to describe the Grunig theory in more detail. The theory uses the combination of several variables to indicate the extent to which a person will actively communicate about a particular situation. The first of these concepts, problem recognition. represents the extent to which a person recognizes that something is missing or indeterminant in a situation so that he stops to think about the situation. The second concept, constraint recognition, represents that extent to which a person perceives constraints in a situation which limit his freedom to construct his own behavior. Generally, the theory holds, people will com-
63
a•abUe Rdationl'\1 Review municate about situations in which they perceive something to be missing and in which they do not perceive constraints which limit their ability to personally affect the situation. The combination of these two concepts can be viewed as lour types of situ· ations, situations in which people's behaviors can be expected to be similar. The behaviors resulting in these four situations have been all problem-facing behavior (high problem recognition, low constraints), contrained behavior (high problem recognition, high constraints). routine habit behavior (low problem recognition, low constraints), and fatalistic behavior (low problem recognition, high constraints). Within each of these situations, a person may or may not make use of a referent criterion, the third variable in the theory. A referent criterion, in essence, exists when a person knows what to do in a situation. He might have knowledge or ~perience from similar situations, or he might have a goal, a solution, or an attitude which he carries from situation to situation . The effect of the referent criterion is to reduce a person's need to communicate about a situation because it indicates to that person what his behavior should be in the situation.
Finally, a person can either be involved or not involved in each of these four basic situations. This fourth variable in the theory, lroel of involvement, is important because it cha nges a person's communication behavior from active information s~eking to passive information processing. A person who per· ccives himself to be inyolved in a situation actively-and selectively-looks fo r information to deal with the situation-if he also perceives the situation in a way that leads to problem-facing or constrained behavior. When involvement is low, a person is unlikely to seek information a t his own initiative. T he low-involvement problem-facing situation, however, often bri ngs forth "curious" people who seek information about situations that do not involve them. On the other hand, a person will process information about situations which do not involve him if he is exposed to it without any effort on his part. For example, a person waiting in a waiting room may process information in an available magaz.ine although he would not otherwise seek out that informa· lion . T hese four variables can be grouped into 16 different combinations. com· binations which can be viewed as 16 different ways of perceiving a situation. Previous research has shown that the probability that a person will seek or process information differs across these 16 combinations. Table I presents the most accurate set of probabilities that have been calculated to date. The com· munication professional can use this theory to identify, in this case. how each type of employee views each of several types of situations confronted on the job. Once he knows how these employee types view the job, he can consult the table of probabilities to determine his chances of communicating with these employee types about the situations of concern. This theory, then, suggests several hypotheses about the nature of employee
communication in a research organization : 1. Since professional employees in a decentralized management structure can be expected to seek information about those situations which involve
64
~missing and :-to personally
- types of si I u.., to be similar. rroblem -facing ;u,ned behavior ~avior (low low problem =Wq
Evaluatin~
so be likely to , nstrained b~ M>CI thus can be to have little hypothesized . h~ ve dcvei.C~ which """Y still peru:;ually will urganization-A
will neither
Employee Co•mlUlieat.lou
TABLEl MEAN SCORES ON f-QUR SITUATIONAL VARIABLES FOR U SITUATIONS' Problem Re:cognition
LstantiaJ board The
.J r..r~
Pre-
·~. As
its
program
News of appointm~nts orawardo; NtwS about pay :md
Ad.ministratin Content
Rftevch
Employ«
Conten t
Content
Factor
Factor
F.ctor
.40
.04
~5
bcndit pbns Stories about social
.33
- .03
.•8
activities at NBS
.00
.OS
.80
-.01
.04
.71
.17
.16
.62
.66
.21
.04
.20 .64
.03
.70
.21
.33
.76
.20
.23
.30
.63
.25 .42
.31 .03
.20
.77
.04
.1(
.76
.02
,18
.70
,04
.14
.63
.rn
.43
,47
.25
Ntw> about SEBA sports activit its Ltuers to editor and rtsponses
Nnoo-s on COtlgress;onaJ legislation affecting NBS
HUJ]"'an lnlerest Stories about NBS em.ploy«S Mes.saget from the Director News from the Oepoutment of Commerce affecting
NBS Historical artido
about the Bureau the NBS budget Nf¥o"$ ol ttchnical progr-ams Nt~-s about
in the Institute for
Basic Standards No"Jnvolvf.-d Vari.ate
Z7
.69
- .53
.31
- .13
.,()e Bureau i!>.dy to be , the lnsti'J>e consists • Adminis·~ split be-~found .;,.j,.,~
nonin· that the t.lw fringes ""thus ilk concerned
Employee Co,..munleation
relation analysis. When asked which Bureau publication they would read if they could read only one, the research type chose the Technical Calendar and Dimensions magazine. The other three types chose the Technical Calendar and the Standard in that order, although both the employee type and the noninvolved type had Dimensions in third place (both types also expressed an interest in research content). The administrative type also expressed interest in the Administrative Calendar. Thus, all types tend to choose the two principal NBS news publications first, but there is an across-the-board interest in a research-oriented publication-i.e., Dimensions. Responses of each of the types to the series of questions asking them to compare and evaluate various NBS media indicated that the research type said "'other employees" are its first source of information whereas other types chose the NBS newspaper, the Standard. Likewise, the research type said its supervisor is the most dependable source of information and the others chose the Standard or the Technical Calendar. Also, the research type said the supervisor provides the most complete information while the others chose the Standard or the Technical Calendar. Otherwise, all types say the Technical Calendar is most useful, aU would choose the Standard to communicate information about themselves, and all think the Standard presents the most accurate picture of Bureau events and people. Finally, the four types were compared on the media from which they had heard the items of information included in the knowledge questions. Most of the differences here were in whether the types had heard the items, as already discussed in presenting the canonical correlation results. When employees had heard the information, all types tended to hear it from the same place. Likewise, these knowledge sources showed no dominant medium. Rather, employees with knowledge of Bureau news tended to pick it up from a number of sources. Conclusions and Recommendations The results of this study seem to provide support for the three sets of hypotheses of the study and, as a result, seem to provide important insights for employee communication managers in research organizations. First, the results suggest that NBS employee media contain information about situations which employees do not perceive to involve them. Second, the study has isolated four types of employee publics with different communication behaviors, different information needs, and for which different communication strategies are necessary, types which parallel those hypothesized. The first conclusion is an important one for a scientific organization like the Bureau which has a high proportion of professional employees. What this study seems to have shown is that involvement in one's own work is accompanied by a low level of involvement in the overall administration of the organization and in the activities of other employees. In a decentralized organization, internal communication is especially important if the activities of employees are to be coordinated. However, it is difficult to force employees to communicate and if their level of involvement is low they will seldom actively
79
P .. bUe Relatloa" Review communicate-seek information-about the situation.s which do not involve them. The study also indicates, however, that employees will process information about low-involvement problems. ln other words, employees can be informed about administrative and research activities of the Bureau but such information is not really as critical to them as is information directly relevant in their day-to-day jobs. While internal communication in such a setting cannot secure active cooperation among professionals and between administrators and other employees, it can secure awareness of and perhaps appreciation of one another's problems and activities. The nature of the information that employees will process. however, differs by types of employees. As we had hypothesized, the research-oriented typewhich consists mostly of scientists- will process research information and to some extent administrative information, particularly such administrative information as budgets, Congressional legislation. and decisions in the Department of Commerce and the Director's office_ As hypothesized, the study also isolated an administration-oriented type which consists both of top administrators and of lower-level administrative and clerical aides. Although this type thinks most about administrative problems. it also pays attention to employee and research problems. They are highly involved in employee situations and moderately involved in administrative situations. Thus they will process-and perhaps actively seek-administrative information and employee information and are most likely to use administrative media. They are also predisposed to process research information, but in practice do not seem to have time to do so. The last two types isolated support the third hypothesis but also add information to our expectations. The employee oriented type seems to consist of employees from many levels and positions of the Bureau who are probably less involved in their own work than the first two types and as a result are mostly concerned with their status as employees. Although they indicate an interest in employee and research information, in practice they demonstrate only a moderate awareness of this type of information. Th• fourth, noninvolved type, thinks about most of the situations which were measured but bt'cause of a low lt>Vel of perceived involvement or high perceived constraints seeks or processes littl• information. This type consists mostly of nonprofessional employees from several different organizational levels. and thus closely resembles what our third hypothesis had predicted. This last type does use the principal NBS media, however, and thus processes the information available in these media. They are most likely. however, to process such employeerelated information as items about social and recreational activities and employee bt>nefits rather than information about the research or administrative activities of the Bureau. The use employees make of NBS media begins to make sense when it is compared with each employee type's perception of Bureau situations. The two most fr•quently used publications. the Technical Calendar and the NBS Standard, are publications used by all types, but particularly by the employeeoriented and noninvolved types. These are broad-scale media which, in es-
80
'C1
IUftdao Ec:.blor.>io 'K.ool
._;.....,~
whdt~
Evaluatin~ 1Y'
anvolve
intormation
wormed onformamtheir
Employee Comm11uleation
sence, indicate what is on the Bureau's agenda. Reading enough to know what is on the agenda is typical behavior for noninvolved people-e.g., watching television news or scanning newspaper headlines. Thus, the roles of these two media are simply to let people know what is happening. Likewise, the noninvolved types also are most likely to scan bulletin boards. The researchoriented and administrative-oriented employees, however, use more specialized media-addresses, seminars, and Dimensions magazine for research and the Administrative Calendar, Administrative Issuances, and addresses for administrative information. Thus, there seems to be litlle overlap in NBS media. Some simply tell people what is on the agenda, some provide research information, and others provide administrative information. However, there do seem to be some gaps which either modified or new media should fill. Most of the employee types-but especially the research and administrative types-are interested in information about budgets, Congressional legislation, decisions in the Department of Commerce, decisions in the Directo(s Office, and administrative information which directly affects their status in the organization. If such information were provided more and in more depth in the Standard, and in newspapers like it in other research organizations, it could become a publication that is sought out rather than simply processed. Finally, a large group of people need information about the research activities in the Bureau, and no single publication now seems to do this. If possible, it would seem that a research-oriented publication written for the informed layman- such as the NBS Dimensions-should be distributed throughout such an organization as this and in particular to the scientists that make up the research type. Second, more research items could be included in the employee newsletter. Finally, open houses, in which employees can observe research activities of others, would also seem desirable. Some recommendations for the employee newspaper in a research organization seem in order. Most commercial newspapers follow a format in which the news most relevant to all readers is put on the front page and in which specialized information is placed in specialized sections. This format also seems ideal for the employee newspaper. The front pages should contain the most relevant information for everyone-budgets, legislation, major decisions, etc. Inside there should be a specialized research S«tion for research-oriented employees, an employee news section for employee-oriented employees, and an administrative news S«tion for administrative-oriented employees, although the administrative section might be less necessary because most organizationsincluding NBS-have specialized media for administrative personnel. Footnotts
Cf. Jerald Ha.gc Olnd Mich:lel Aiken, Sociul Cl1ange in Cornplu Organi%ahonl {New York: Random House. 1910). Al:w William Kotnhauser. Sci~tltists in Industry (Berkeley: University of 1
Caluomia Press. 1963) 'Koltz. and Kahn, (or example. tonduded: *'By ud ~ ~ N:turt and t"Xtt:nl of exc:h.a.n&ts amon, proplt at the wmt ltvtl should be related to the obj«tivts of the various subsystems in which they are involved. with the primary focus on their own major task." {Daniel Katz and
81
Pabllc Relatio ns R e vie w RolK:rt l,., Kahn, The Soda/ Psychology of Organi:..ution.t (New York : John Will-y&. Sol\$, 1966), p. 59. Simon, likewist. itAted: ln hiuaf'C'hic systems, w~ can distinguish between the int&actions among tM subsystems, on th~ one h.tncl and the inttnctions within subsystems-tNt is. arnOn& the parts of those w.btystems-on tht other. Tht int.tractions at tJw difkm..t levels may~. and often wiJI be, of c:lifftm\t orders ol ma.gnitu
