Gordon Walker.

Network position and cognition online

. (page 3 of 4)
Online LibraryGordon WalkerNetwork position and cognition → online text (page 3 of 4)
Font size
QR-code for this ebook

indicated a high level of product market uncertainty; therefore,
product-related problems could be typed as ill-structured. The growth rate
and rate of technological change also led to specifying the term dimension as
a practical parameter of product goals. The Intensive technology of the firm.


made it more likely that members in many nominal roles had some Interaction
with the firm's customers. Finally, the types of product worked on were
specific only to software firms.

The influence of nominal roles on cognition may increase and become less
sensitive to the network as 1) the problem-solving situation studied becomes
more structured or routine and 2) the labor intensity of the firm's
technology decreases. Routine problems may evoke cognitive processes which
depend less on input from interpersonal relations and are more strongly
associated with task rules (see Van de Ven and Delbecq and Koenig (1976) for a
presentation of this phenomenon in cask groups). Furthermore, members of
organizations with less labor intensity than the software firm studied here
may not have sufficient access to others in the firm for the network relation-
ships to overcome the effect of nominal role pressures. Sparse and fragmented
organization-wide networks may therefore have no influence on cognition.
Task-oriented ties between individuals in local work groups may be influential,

The results are those for the influence of firm and industry tenure on
the salience of the term dimension are less limited. These effects point to
the need for relatively high executive turnover in fast growth industries
experiencing technological change. Once within a firm, veterans lose their
perspective and fail to distinguish between the different ways of achieving
product success in the short and long term.

The effect of firm tenure on causal inference models, which was contrary
to that hypothesized, may have been a result of the degree of structure in the
judgment situation studed here. Organizational members may learn solutions to
well-structured problems early in their tenure and expand on this knowledge in
innovative but incremental ways as the influence of socialization diminishes.
On the other hand new members may feel forced to structure highly uncertain
judgment situations in which older members are willing to see more ambiguity.


Thus the results of the present study regarding socialization in the firm may
represent the tendency of older members to appreciate the ill-structured
nature of the judgment task. However, because the task was not specific to
the firm but applied to the software industry in general, experience in the
industry provided Information which increased perception of differences
between short and long term success. Industry and firm tenure therefore
influence judgment for badly structured industry-specific problems in opposite

It is interesting, moveover, that differences in the perception of the
term dimension were much better predicted by the network than those of the
boundary dimension. This result suggests that individual judgments of time as
an implicit parameter of product goals depend more on input from other firm
members than judgments concerning differences between user and organizational
perspectives on successful product outcomes. Organizing activity in the firm
may be related to member perspectives towards present and future goals in a
way which is qualitatively different from orientations towards the boundary
dimension, and subgroups whose members perceive the term dimension strongly
should be located similarly in the structure of the network. This general
location should correspond to the upper levels of the management hierarchy, as
proposed by Parsons (1960) and Williamson (1981).

To the extent that netv/ork position determines causal inferences, it
affects the inputs to decision making and therefore to a degree decision
outcomes. On the other hand, a member's position in the network itself
reflects the global pattern of individual choices about product related issues
as represented by the types of task-oriented relations. These choices
furthermore are a consequence of previous decisions about the projects the
organization would undertake. Organizational members use their contacts to
accomplish project aims (Kotter, 1932). The present study shows that this
process itself deteraiines individual orientations towards product success and


therefore constrains cognitive input to new lieclsion-maklng situations. Thus
the network transforms the implementation of one set of decisions into the
premises by which another set is made.

To what degree can this process be controlled? This question has
provoked a number of qualitative models of organizational (as opposed to
individual or small group) decision processes. Among these are Gouldner's
(1959) distinction between the rational and natural system perspectives (see
also Thompson, 1967), Cohen, March and Olsen's (1972) garbage can model and
Weick's (1976) model of loosely coupled systems. Both Cohen, March and
Olsen's and Weick's models approach the problem of control by specifying where
control is absent.

Weick points out that interaction among organizational members and
therefore the network of interactions may be a tightly coupled aspect of an
organization. The association of means and ends may be either tightly or
loosely coupled. Since the results of the present research show that the
network is related to the way organizational members perceive the uniqueness
of means and ends linkages, tightly and loosely coupled systems may be joined
as the content of interpersonal interaction becomes the content of causal
inferences. The latter may involve either clear or ambiguous means/ends
associations (March, 1977), which In turn should be related to the degree of
choice flexibility inherent in the system.

Cohen, March and Olsen's garbage can model specifies four loosely
coupled decision situation parameters. These parameters are: choice
situations, participants, problems, and solutions. The results of the present
study show that the position of participants in an interaction network
determines to an extent their orentiation towards problems and solutions
proposed in a choice situation. In a sense, the network acts as a selection


and retention mechanism in the evolution of decision-making within the firm.
(Notice that problems and solutions with ambiguous means and ends may be
selected for.)

The network model in the present study also can be seen as a bridge
between Gouldner's rational and natural system models of organization. The
link to the natural system follows from the discussion above of hov; the
network relates to the garbage can and loosely coupled systems models of
organizational decision making. Furthermore, studies of network structure
over time and over populations of actors (Boorman and White, 1975) have shown
regularities of development which can not be Interpreted as controllable by
one or a group of actors.

The network is related to the rational system model through the
covarlance of the various types of task-oriented tie. Task-oriented
relationships can be partially manipulated by those with pov/er to prescribe
formal ties between individuals. Informal ties should be correlated with
formal relationships, and so the structure of the network is to some extent
determined by the powerful group. Assessing the covariance among formal and
informal types of tie and the relative contribution of each type to individual
differences in cognition are important analytical tasks for evaluating how
strongly the network has been rationalized.

Furthermore, task-oriented relationships measured in the present study
involved both information and direction relevant to the accomplishment of
product goals. The transmission of direction indicates an Influence process
which some organizational members may manage by maintaining strategically
located positions in the network for one or more types of relation (Marsden,
1981). Thus control may be exerted both by specifying formal relationships
throughout the firm which affect in various degrees the development of other
types of ties and by managing the flow of task-related direction through


structurally determined influence. Further research should show whether these
aspects of control over organizational decision-making affect Inferences about
product goals independently.

Finally, the present study has not examined the ways in which affect
might constrain or mediate the relationship between organization and cognition.
Both affective input to decision-making and the network of interpersonal
relationships based on affect may be empirically separable from the type and
determinants of cognition studied in the present research (see, for example,
Zajone, 1981). Separability may not imply independence, however. Further
research might identify how inferences and feelings about events are related
to position in a network composed of both instrumental and expressive ties.
Helse (1979, p. 32-34) suggests, for example, that an individual develops his
or her network of interpersonal relationships by selecting others with or
through whom affect can be controlled and that cognition is a function of this
process. Boorman and IThite (1976) found in their study of network structures,
however, that members of formal organizations tended to suppress negative
sentiment in comparison with actors in informal settings. These results
suggest that the demands of task accomplishment in formal organizations
constrain network development the kind proposed by Heise. Task demands may
themselves be a function of affect management strategies, as organizational
members establish task-oriented relationships with others who share
sufficiently their orientations towards accomplishing product goals. An
analysis of the relationships among positions in a task-oriented network
should show whether such a correspondence between relational and cognitive
continuities exists, and if so, whether the strucutre of positive effect ties
can be matched to it.


The present study has developed and tested a theory relating aspects of
an individual's work experience to perceptions of how different types of
software product goal are achieved. The most important predictor of
perceptions was the position, defined in terms of the principle of structural
equivalence, an individual occupied in the network of task-oriented relations
within the organization. This result indicated that cognition in ill-
structured judgment situations is strongly influenced by interpersonal
relationships whose content is relevant to the judgment made. The influence
of the network superceded the effects of nominal role, job tenure and
experience with various types of product but not those of industry and firm
tenure. The results may be generalized to organizations with similar labor
intensity, market growth rate, and rate of technological change. In addition,
the results can be related to other organizational decision-making models.
The network can be seen as a link between making and implementing product-
related decisions and at the same time as a bridge between the natural and
rational systems approaches to organizing activity.



1. Anderson, T. W. , An Introduction to Multivariate Statistical Analysis ,

New York: Wiley, 1958.

2. Arable, P., S. A. Boorman and P. R. Levitt, "Constructing Blockmodels:

How and Why," Journal of Mathematical Psychology , 1978, 17:21-63.

3. Ajzen, I., "Intuitive Theories of Events and the Effects of Base Rate

Information on Predictions," Journal of Personality and Social
Psychology , 1977, 35:303-314.

4. Bartlett, M. S. , "Tests of Significance in Factor Analysis," British

Journal of Psychology, Statistical Section , 1950, 3.

5. Bavelas, A., "Communication Patterns in Task-Oriented Small Groups,"

Journal of Acoustical Society of America , 1950, 22:725-30.

6. Berger, C. and L. L. Cummlngs, "Organization Structure, Attitudes, and

Behavior," in Rese arch in Organizational Behavior , Staw, B. M.
(ed.). Vol. 1, Grennwich, Conn.: JAI Press, 1979.

7. Breiger, R. L., "Career Attributes and Network Structure: A Biockmcdel

Study of a Biomedical Research Network," American Sociological
Review, 1976, 41:117-35.

8. Breiger, R. L., S. A. Boorman, and P. Arable, "An Algorithm for

Clustering Relational Data, With Applications to Social Network
Analysis and Comparison With Multidimensional Scaling," Journal of
Mathematical Psychology , 1975, 12:328-383.

9. Breiger, R. L. and J. G. Ennis, "Personal and Social Roles: The Network

Structure of Personality Types in Small Groups," Social Psychology
Quarterly , 1979, 42:262-70.

10. Brooks, F. P., The Mythical Man Month , Reading, Mass: Addison-Wesley, 1975.

11. Boorman, S. A., personal com:aunication.


12. Boorman, S. A. and H, C. White, "Social Structure fron Multiple

Networks. II. Role Structures," American Journal of Sociology ,
1976, 81:384:1446.

13. Burt, R. S., "Cohesion Versus Structural Equivalence as a Basis for

Network Subgroups," Sociological Methods and Research , 1978,

14. Burt. R. S., "Models of Network Structure," Annual Review of Sociology ,

1980, p. 6.

15. Burt. R. S., "Positions in Networks," Social Forces , 1976, 55:93-122.

16. Canetti, E., Crowds and Power , New York: Continuum, 1962.

17. Carrington, P. J. and G. H. Hell, "COBLOC: A Hierarchical Method for

Blocking Networks," Journal of Mathematical Sociology , 1981,

18. Carroll, J. D. and P. Arable, 'Multidimensional Scaling," Annual Review

of Psychology , 1980, 31:607-650.,

19. Carroll, J. D. and J. J. Chang, "An Analysis of Individual Differences

In Multidimensional Scaling via an N-way Generalization of
Eckart-Young Decomposition," Psychometrika , 1970, 35:283-319,

20. Chappie, E. and L. R. Sayles, The Measure of Management , New York:

MacMlllan, 1961.

21. Cohen, M. D., J. G. March and J. P. Olsen, "A Garbage Can Model of

Organizational Choice," Administrative Science Quarterly , 1972,

22. Comstock, D. E. and W. R. Scott, "Technology and the Structure of

Subunlts: Distinguishing Individual and Work Group Efforts,"
Administrative Science Quarterly , 1977, 22:(7)-202.

23. Connolly, T., "Information Processing and Decision Making in

Organizations , " in Inev Directions in Or ga nizational Behavior ,
Staw, B. M. and G- R. Salanclk (eds.), 1977, pp. 205-234,


24. Cronbach, L. J., G. Gleser, H. Nanda, and N. Rajaratnam, The

Dependability of Behavioral Measurements; Theory of
Generallzabllity for Scores and Profiles , New York; John Wiley and
Sons, 1972.

25. Cyert, R. M. and T. G. March, A Behavioral Theory of the Firm , Englewood

Cliffs, N. J.: Prentice-Hall, 1963.

26. Dearborn, D. C. and H. A. Simon, "Selective Perception; A Note on the

Identification of Executives," Sociometry , 1958, 21:140-144.

27. Delbecq, A. L., A. H. Van de Ven and D. H. Gustafson, Group Techniques

for Program Planning , Scott-Foresman, 1975.

28. Evans, P. G., "Multiple Hierarchies and Organizational Control,"

Administrative Science Quarterly , 1975, 20; 250-259.

29. Farrar, D. E. and R. R. Glauber, "Multicollinearity in Regression

Analysis: The Problem Revisited," Review of Economics and
Statistics , 1967, 49:202-217.

30. Fombrun, C, Networks and Organization: Perspectives and Pragmatics for

Assessing the Political Structure of Organizations , Ph.D.
Dissertation, Columbia University, 1980.

31. Gouldner, A. W., "Organizational Analysis," in Sociology Today ,

Merton, R. K. , L. Broom and L. S. Cottrell, Jr. (eds.). New York:
Basic Books, 1959.

32. Goetz, M. A., "Software Products: The Key to Cost-Effective Use of

Computers," Computerworld , October 16, 1978.

33. Hayes, R. H. and W. J. Abernathy, "Managing Our Way to Economic

Decline," Harvard Business Review , 1980, July-August, 67-68.

34. Heise, D. R., Understanding Events , Cambridge, UK: Cambridge University

Press, 1979, pp. 32-34.


35. Herman J. B. , R. B. Dunham and C. L. Hulin, "Organizational Structure,

Demographic Characteristics and Employee Responses," Organizational
Behavior and Human Performance , 1975, 13:206-232.

36. Hofer, C. W. and D. Schendel, Strategy Formulation: Analytical

Concepts , West, 1978.

37. Holland, P. W. and S. Leinhardt, "The Structural Implications of

Measurement Error in Sociometry," Journal of Mathematical
Sociology , 1973, 3:85-111.

38. Holland, P. W, and S. Leinhardt, "An Omnibus Test for Social Structure

Using Triads," Sociological Methods and Researc h, 1978, 7:227-36.

39. Kim, J. 0. and J. Curry, "The Treatment of Missing Data in Multivariate

Analysis," Sociological Methods and Research , 1977, 6:215-240.

40. Kim, J. S. and W. C. Hamner, "Effect of Performance Feedback and Goal

Setting on Productivity and Satisfaction in an Organizational
Setting," Journal of Applied Psychology , 1976, 61:48-57.

41. Kotter, J. P. Tl ie General Managers , New York: The Free Press, 1982.

42. Lecht, C. P., The Waves of Change: A Techno-Economic Analysis of the

Data Processing Industry , McGraw-Hill, New York, 1977.

43. Lincoln, J. R. and J. Miller, "Work and Friendship Ties in

Organizations: A Comparative Analysis of Relational Neworks,"
Administrative Science Quarterly , 1979, 24:181-199.

44. Lorrain, F. P. and H. C. White, "Structural Equivalence of Individuals

in Social Networks," Journal of Mathematical Sociology , 1971,

45. March, J. and H. Simon, Organizations , John Wiley and Sons, 1958.

46. Marsden, P. V., "Introducing Influence Processes Into a System of

Collective Decisions," American Journal o f Sociology , 1981,


47 » Mervis, C. B. and E. Rosea, "Categorzlation of Natural Objects," Annual
Review of Psychology , 1981, 32:89-116.

48. Miles, R. N., "Role-Set Configuration as a Predictor of Role Conflict

and Ambigurity in Complex Organizations," Sociometry , 1977,

49. Neisser, U., Cognition and Reality , W. H. Freeman, 1978.

50. Newell, A. and K. A. Simon, Human Problem Solving , Englewood

Cliffs, N.J.: Prentice-Hall, 1972.

51. Ouchi, W. C, "The Transmission of Control Through Organization

Hierarchy," Academy of Management Journal , 1978, 21:173-192.

52. O'Reilly, C. A. and K. H. Roberts, "Task Group Structure, Communication

and Effectiveness in Three Organizations," Journal of Applied
Psychology , 1977, 62:674-681.

53. Parsons, T., Structure and Process in Modern Societies , New York: The

Free Press of Glencoe, 1960.

54. Porter, M. E., Competitive Strategy, The Free Press, 1980.

55. Rosch, E. and B. B. Lloyd, Cognition and Categorization , Hillsdale,

N. J.: Laurence Erlbaum Assoc, 1978.

56. Schein, E, H., Career Dynamics: Matching Individual and Organizational

Needs , Addison-Wesley, 1978.

57. Schiffman, S. S., M. L. Reynolds and F. W. Young, Introduction to

Multidimensional Scaling , New York: Academic Press, 1981.

58. Schwartz, D. and E. Jacobsen, "Organizational Communication Network

Analysis," Organization Behavior and Human Performance , 1977,

59. Shaw, M. E., "Some Effects of Problem Complexity Upon Problem Solving

Efficiency in Different Communication Nets," Journal of
Experioontal Psycholog,y , 1954, 48:211-17.


60. Simon, H., "On the Concept of the Organizational Goal," Administrative

Science Quarterly , 1964, 9:1-22.

61. Simon H., Administrative Behavior , 2nd Edition, New York:

MacMillan, 1976.

62. Simon H. A., "Discussion: Cognition and Social Behavior," in Cognition

and Social Behavio r, Carroll, J. S. and J. W. Payne (eds.),
Hillsdale, N.J.: Lawrence Erlbaum Assoc. , 1976, pp. 253-268.

63. Sonnenfeld, J., "Executive Apologies for Price Fixing: Role Biased

Perceptions of Causality," Academy of Management Journal , 1981,

64. Spier, M. J., "A Pragmatic Proposal for the Improvement of Program

Modularity and Reliability," International Journal of Computer and
Information Science , 1975, 4:133-149.

65. Thompson, J. D., Organizations in Action , McGraw-Hill, 1967.

66. Tichy, N. M., "An Analysis of Clique Formation and Structure in

Organizations," Administrative Science Quarterly , 1973, 18:194-208.

67. Tichy, N., Tushman, M. L. and C. Fombrun, "Social Network Analysis for

Organizations," Academy of Management Review , 1979, 4:507-519.

68. Tushman, M. L. and T. J. Scanlan, "Boundary Spanning Individuals: Their

Pvole in Information Transfer and Their Antecedents," Academy of
Management Journal , 1981, 24:289-305.

69. Tversky, A. and D. Kahnemann, "Causal Schemas in Judgments Under

Uncertainty," in Progress in Social Psychology , M. Fishbein, (ed.),
Lawrence Erlbaum Assoc. , 1980.

70. Ungson, C. R., D. N. Braunstein and P. D. Hall, "Managerial Information

Processing," Administrative Science Quarterly , 1981, 26:116-132.

71. Van de Ven, A, H., A. Delbecq and R. Koening, "Determinants of

Coordination Modes Within Organizations," American Sociological
Review , 1975, 41:322-338.


72. Van Tassel, D. , Program Style, Design. Efficiency, Debugging, and

Testing , Prentice-Hall, 1978.

73. Weick, K. E. , "Educational Organizations as Loosely Coupled Systems,"

Administrative Science Quart e rly , 1976, 21:1-19.

74. White, H. C. , S. A. Boorman R. L. Breiger, "Social Structure from

Multiple Networks. 1. Blockmodels of Roles and Positions," American
Journal of Sociology , 1976, 81:730-780.

75. Williamson, 0. E., "The Modern Corporation: Origins, Evolution and
Attributes," Journal of Economic Literature , 1981, 19:1537-1568.

76. Zajonc, R. B., "Feeling and lainking: Preferences Need No

Inferences," American Psychologist, 1980, 35:151-175.


Appendix A
Measurement of Causal Models

Table Al presents the results of the two-dimensional CANDECOMP scaling
solution • The locations of the events and the types of success are shown
together in the tw^o dimensional space. An interesting characteristic of this
scaling solution is the high correlation between the dimensions. CANDECOMP
does not necessarily produce orthogonal axes, and a commonly cited advantage
of the technique is the rotation invariance of the solution (see Carroll and
Arable, 1980, p. 631). It was not possible, therefore, to rotate the axes to
increase their interpretability.

The horizontal axis can be labeled the boundary dimension since along it
product development types of success are clearly differentiated from the user
types. Likewise, the vertical axis can be labeled the term dimension, since
the long term types of success lie dbove the short term types. Note that the
development types of success are much closer together on the boundary
dimension than the user types, and the user types of success are closer
together on the term dimension than the development types.

Although the term dimension is interpretable, the quadrilateral
predicted by the product sucess typology is not perfectly produced: currency
lies between coherence and performance rather than occupying the fourth corner
of what should be a rhombus. The location of currency makes the inter-
pretation of the axes more difficult, especially the interpretation of the
term dimension.

'Nine CANDECOMP runs with different randomly chosen initial conf igruations
were made. For all runs, the three dimensional solution explained approxi-
mately five percent more of the variance in subject scores than the two
dimensional solution. The two aimensional solution was therefore selected
for interpretation. The run which offered the best interpretation in terms
of the type of success weights was selected to test the hypotheses. The
patterns of events and types of success in all solutions were quite similar,



CANDECOMP Scaling of Thirty-one Events and Four Types of Product

Success - R - .30





" 19

• •


1 •






1 J




2o! .^











^ 1










Numerals refer \.o events in Table 1.


The distribution of the events reinforces the interpretation of the
dimensions based on the locations of the types of success. Users become mora
prevalent in events moving from right to left on the horizontal axis,

1 3

Online LibraryGordon WalkerNetwork position and cognition → online text (page 3 of 4)