Copyright
George Franklin Farris.

Supervisory roles, colleague roles, and innovation in scientific groups online

. (page 1 of 2)
Online LibraryGeorge Franklin FarrisSupervisory roles, colleague roles, and innovation in scientific groups → online text (page 1 of 2)
Font size
QR-code for this ebook


uevvp:\/





SUPERVISORY ROLES, COLLEAGUE ROLES, AND INNOVATION
IN SCIENTIFIC GROUPSl

November, 1971 No. 575-71

Revision of Working Paper No. 552-71



MASSACHUSETTS

INSTITUTE OF TECHNOLOGY

50 MEMORIAL DRIVE

.MBRIDGE, MASSACHUSETTS 021



SUPERVISORY ROLES, COLLEAGUE ROLES, AND INNOVATION
IN SCIENTIFIC GROUPSl

November, 1971 No. 575-71

Revision of Working Paper No. 552-71



MASS. INST. TECH
DEC 10 1971

DEWEY LIBRARY



no, SlS'li



RECE'vrn
DEC 8 1971
M. I. T. LiLJhMrtltS



Abstract

Innovation of groups of scientists was related to performance of
"fcol league roles," utilizing a model of organizational decision making
proposed by Farris (1971). In more innovative groups, supervisors were
named less often by group members as useful for original ideas, but more
often for providing critical evaluation. Members of more innovative groups
named one another more often as useful for providing technical information
and help in thinking about technical problems. Organizational information
was available from fewer sources inside or outside the more innovative
groups. Supervisors of more innovative groups named more outside sources
as useful to them for original ideas. These trends are very consistent with
Maier's (1967) emphasis on the leader's "integrative function" in group
proljlem solving.



G33653



In most research and development laboratories, work is carried out
by groups of scientists and engineers. Within a given organization some
groups are often cited consistently as being more innovative than others
in their R&D work. What factors distinguish these more innovative from
less innovative R&D groups?

Previous research has considered factors such as diversity of group
members (Pelz, ]967; Pelz & Andrews, 1966), group age (Shepard, 1956; Wells I
Pelz, 1966) and characteristics of the supervisor (Andrews & Farris, 1967).
Although research on small group problem solving (for reviews, see Cart-
wright & Zander, 1968; Hoffman, 1965; and Collins & Guetzkow, 196A) suggests
that characteristics of the interaction among group members are important
determinants of group performance, little research has been devoted to the
problem-solving process of scientific groups.

Recently, Farris (1971) proposed that the interaction among members of
an organization in making decisions can be conceptualized in terms of the
roles they perform for one another in this process. His model considers
three stages in the problem-solving process: suggestion, proposal, and
solution. (See Figure 1.) Different "colleague roles" — activities



Insert Figure 1 About Here



performed by one scientist which facilitate the problem solving of another
— are hypothesized to be particularly important during each stage. (All
roles may be somewhat helpful at any stage.) Providing original ideas,
technical information, and administrative information are said to be impor-



tant colleague roles which help a scientist to come up with a suggestion.
Help in thinking through a problem and critical evaluation are important
in shaping the suggestion into a proposal. And, assuring a fair hearing
and providing administrative help are colleague roles which can help in
turning a proposal into a solution which is implemented in the organiza-
tion. Research to date (Farris, 1971; Swain, 1971) has focused on indi-
viduals who perform these roles, examining their personal characteristics,
working environments, and career development. The present research extends
this conceptual approach to the group level.

A central concern in much of the literature on group problem solving
has been the relative importance of the formal leader and group members
in performing various roles important for innovation. One school of thought
(e. g., Maier, 1967) has emphasized the importance of roles performed by
the leader of a problem-solving group. Another, often considering "leader-
less" groups, has emphasized roles which can be performed by any group
member (e . g.. Bales, 1950; Benne & Sheats, 1948). Bowers and Seashore
(1966) discuss both "supervisory leadership" and "peer leadership." If,
following French (1956), leadership is considered to be the ability of
one person to influence the behavior of another, then three parties may
exert leadership in the problem solving of scientific groups: the super-
visor, the group members, and people from outside the group. The relative
importance of each is subject to empirical investigation.

In the present study the innovation of scientific groups will be
related to the performance of colleague roles for group members by three
parties: the supervisor, other group members, and scientists from outside
the group. Then the roles performed by the supervisor will be examined in
more detail, relating the innovation of his group to his orientation inside
and outside his group. Finally, the problem-solving processes of more and



less innovative groups will be compared by examining the performance of
each colleague role at each stage. As shown in Figure 2, group innovation



Insert Figure 2 About Here

will be related to seven colleague roles performed by:

1. The supervisor for his group.

2. Group members for other members.

3. Outsiders for group members.

4. The supervisor for outsiders.

5. Outsiders for the supervisor.

6. Group members for the supervisor.

METHOD
The study was conducted in a division of a NASA research center engaged
in a wide variety of R & D activities related to aerospace. Their tasks
ranged from basic research on physical and chemical processes to the conduct
of atmospheric and deep space experiments using rockets and satellites. One
hundred and seventeen professionals participated in this study, including 87
bench scientists in 14 groups, headed by first-line supervisors. The mean
group size was 6.2 members, excluding the supervisor, and half the groups
contained fewer than 5 members. The groups ranged in size from 2 to 17
members .

As part of an extensive questionnaire describing aspects of their
working environment and motivation, the participants in the study were asked
to name individuals they saw as being useful to them for seven colleague roles;
Considering the technical activities you have been
involved in over the past few years, which people
have been most useful to you for the following:
(The same person may be named as many times as



seems appropriate.)

A. Locating relevant technical information you did not know
about previously. (Spaces for up to eight names were provided
in each part) .

B. Helping your thinking about technical problems — e. g., picking
out fruitful problems, clarifying the nature of a problem,
changing the direction of your thinking about a problem.

C. Critical evaluation of your ideas.

D. His own original ideas.

E. Making sure your ideas get a fair hearing or preventing com-
peting ideas from winning out prematurely.

F. Providing administrative help in getting you needed resources
and facilities.

G. People from whom you learn about technical and administrative
developments happening in (name of division.)

For each role, six scores were determined for each group, corresponding
to the possibilities shown in Figure 2.

1. The per cent of group members who mentioned their supervisor.

2. The percent of possible choices of group members by other
group members. The number of possible choices was N(N-l),
where N = the number of bench scientists in a group.

3. The average number of scientists outside the group mentioned
by a group member.

4. The number of times the supervisor was mentioned by "outsiders,"
(people from outside the group who participated in the study).

5. The number of outsiders mentioned by the supervisor.

6. The per cent of group members who were mentioned by their
supervisor.



7

The innovation of each group member was rated by judges who claimed
to be familiar with the scientist's work. Innovation was defined for the
judges as the extent the scientist's work had "increased knowledge in his
field through lines of research or development which were useful and new".
Judges were supervisors or senior-level non-supervisors. An average of
7.6 judges, working independently, used a modified rank-ordering procedure
to rate the innovation of each scientist's work. Since the judges showed
reasonably good agreement (Spearman-Brown estimate for reliability of a
multiple item scale = .87), their evaluations were combined into a single
percentile score for each scientist. These percentile scores were then
adjusted to remove effects attributable to two background factors: time at
R&D center, and degree (B. S., M. S., or Ph. D.). Group innovation
scores were then calculated by determining the mean adjusted innovation
score of the group members (excluding the supervisor) . Details on these
types of procedures for collecting, combining, and adjusting measures of
scientific performance are more fully described in Pelz and Andrews (1966).

RESULTS

The groups were divided at the median innovation score into high
and low- innovation categories. The scores on the seven colleague roles were
then examined to determine 1) whom the group members find helpful for
performing colleague roles, 2) the supervisor's orientation toward his
group and outsiders in the performance of colleague roles, and 3) for each

colleague role, at each stage in the problem-solving process, the differences

2
which occur between the more and less innovative groups.

Roles performed for group members

Figures 3-5 show the extent to which members of high and low innovation

groups have found three parties to be helpful in their technical work:



8

their supervisor, other group members, and people outside the group.



Insert Figure 3 About Here



Figure 3 shows the per cent of group members who mention their super-
visor for colleague roles. On the average, slightly more than fifty per
cent of them mention their supervisor. This figure ranges from a high of over
60% for help in thinking and critical evaluation to a low of less than A0%
for original ideas. Supervisors are mentioned quite frequently for both
technical and administrative roles.

In general the high innovation groups mention their supervisor more
than the low innovation groups. Differences are most pronounced for
critical evaluation and slightly smaller for help in thinking and adminis-
trative help. There appears to be a tendency for the low innovation groups
to mention their supervisor more often as being useful for his original ideas.

Figure 4 shows the choices of group members by other group members for



Insert Figure 4 About Here



the seven colleague roles. Overall, group members choose one another quite
often, but because there is also a high number of possible choices, the
percentages shown in Figure 4 are quite low. They range from 1% to 14%,
with an average a little over 6%. Group members tend to find one another
useful chiefly for technical roles and least for administrative help and
making sure their ideas receive a fair hearing.

Members of the high innovation groups tend to choose one another more
often for two technical roles: locating technical information and help in
thinking about technical problems; members of low innovation groups tend to



find one another as more useful for administrative roles, especially pro-
viding news of developments in the R&D division.

Figure 5 shows the number of scientists outside the group who are men-



Insert Figure 5 About Here



tioned by group members for the seven colleague roles. Overall, they mention
about one outsider per man. Outsiders are seen as most useful for providing
technical information and help in thinking and least useful for assuring
a fair hearing for group members' ideas.

Although the differences are small, there is a trend for the low
innovation groups, more than high innovation groups, to mention more out-
siders as useful to them. Outsiders are especially more useful to the low
innovation groups for help in thinking and assuring a fair hearing for their
ideas.

To summarize these trends, it appears that all three parties — the
supervisor, other group members, and outsiders — perform colleague roles for
these scientific groups. Group members are named most often, but the super-
visor receives a higher percentage of possible choices. The supervisor and
outsiders provide both technical and administrative help, while group
members are helpful chiefly in technical areas. The high innovation groups
appear to solve problems more as teams. Members find one another more
useful for technical roles but less for administrative roles; outsiders are
mentioned less often as useful; and their supervisor is more useful to them.
This finding indicates that the supervisor may be very much a member of
the high innovative groups as they engage in technical problem solving.
Let us examine the supervisor's roles in more detail.
Supervisor's inside-outside orientation

Figure 6 shows the average number of times the supervisors were mentioned



10

for the colleague roles by scientists outside their group. Across all
roles, supervisors are mentioned by an average of a little over one outside
person. Outsiders mentioned the supervisors most often for locating tech-
nical information and least often for providing a fair hearing or administra-
tive help.



Insert Figure 6 About Here



Overall, supervisors of low innovation groups tend to be mentioned
more often by outsiders than supervisors of high innovation groups. The
strongest differences in this direction occur for the technical roles,
especially help in thinking and providing original ideas. For two admin-
istrative roles — providing a fair hearing for ideas and administrative
help — the trend is reversed.

Figure 7 shows the average number of outsiders mentioned by the



Insert Figure 7 About Here



supervisors for the various colleague roles. Across all roles, super-
visors mention slightly under three outsiders per role. Outsiders are
mentioned most often by supervisors as helpful to them for providing tech-
nical information and least often as useful for original ideas.

Compared to supervisors of high innovation groups, supervisors of low
innovation groups mention more outsiders as useful to them for five of the
seven roles. Outsiders are especially more useful to supervisors of low
innovation groups for help in thinking about technical problems and providing
news of developments in the division. Supervisors of high innovation groups
tend to mention more outsiders as useful to them for their original ideas.

Finally, Figure 8 shows the per cent of group members who are mentioned



11

by their supervisor as helpful to them in their technical problem solving.



Insert Figure 8 About Here



Virtually no group members are mentioned by their supervisor as helpful for
administrative roles. For the four technical roles, an average of about one
in five group members is mentioned by his supervisor. Group members are
especially useful to their supervisor for help in thinking and original
ideas; they are least helpful for providing critical evaluation.

The differences between the high and low innovation groups are striking.
Supervisors of high innovation groups mention more group members as
helpful to them for all technical roles: locating technical information,
help in thinking about technical problems, critical evaluation, and original
ideas .

In summary, these trends indicate that the supervisors of the high
innovation groups are a more integral part of their groups' technical
problem solving and less oriented toward the outside for technical roles.
They find their groups more useful to them for their own technical problem
solving, and, as shown in Figure 3 above, their groups mention them more
often as helpful. Moreover, they are mentioned less often by outsiders for
technical roles, and they mention fewer outsiders for technical roles
(except original ideas). For the more organizationally oriented roles, the
trends are mixed. Let us explore the trends for each role in greater de-
tail.
Roles in the decision-making process

Recall that Farris' (1971) model of organizational decision making consider
the process in three stages: suggestion, proposal, and solution (see Figure 1).



12

Different colleague roles were said to be more important for each stage:
original ideas, technical information, and administrative information
for coming up with a suggestion; help in thinking and critical evaluation in
shaping the suggestion into a proposal; and a fair hearing and administrative
help in the executive decision to make the proposal an actual solution. How
does this process differ in the high and low innovation groups? Although
the present study was not longitudinal so that colleague roles and problem-
solving stages could be investigated over time, an examination of colleague
roles according to the stages with which they theoretically should be most
strongly associated produced some interesting^. trends .

Table 1 recasts the data on colleague roles according to the stage of



Insert Table 1 About Here



the decision making process. Let us examine each stage separately.

Suggestion stage. Supervisors of the high innovation groups name
more colleagues — both group members and outsiders — as helpful to them
for providing original ideas. On the other hand, they tend to be named
less often by others — both outsiders and group members — as helpful
for their own original ideas. No differences were found in the original
ideas colleague role as performed by other group members or outsiders.

Supervisors of the high innovation groups tend to name more group
members and fewer outsiders as useful to them for locating technical infor-
mation. They are named slightly less often by outsiders for this role.
There is a tendency for members of the high innovation groups to name
one another more often as useful for providing technical information,
but no differences occur in the extent to which they name their supervisor
or outsiders.



13

Both the supervisor and the members of the high innovation groups name
fewer colleagues — group members, outsiders, or the supervisor — as
useful to them for providing news of technical and administrative events
in the organization.

To summarize, in the suggestion stage it appears that the super-
visors of the high innovation groups bring to their groups original ideas from
more sources (other scientists' ideas, not their own) and technical and
organizational information from fewer sources. Group members furnish one
another with more technical information and less organizational information.
Thus, in the suggestion stage, the high innovation groups appear to have
available original ideas from more sources inside and outside the group
but not from the supervisor himself, more technical information generated
within the group, and organizational information from fewer sources of any
kind.

Proposal stage . Supervisors of the high innovation groups name fewer
outsiders and more group members as useful to them for help in thinking
about technical problems. Similarly, they are named less often by outsiders
and more often by their group for this role. Also there are tendencies for
group members to receive more help from one another and less from outsiders.

A similar pattern occurs for critical evaluation. Compared to super-
visors of the low innovation groups, supervisors of the high innovation groups
name their groups more and outsiders less as helpful in this role. These
supervisors in turn are named much more often by their groups and less often
by outsiders as helpful for critical evaluation. No difference occurs in the
performance of the critical evaluation role for the group by either other
group members or outsiders.

In summary, all members of the high innovation groups, including the
supervisor, are apt to help one another in thinking about technical pro-



14

blems. For giving and receiving critical evaluation, the supervisor is
similarly more oriented toward his own group than outsiders. Moreover,
the extent to which the supervisor provides critical evaluation for his
group is a key factor in distinguishing the high and low innovation teams.

Solution stage. The differences between the high innovation and the
low innovation groups are smaller at this stage of the problem-solving process.
Both the group and outsiders tend to name the supervisors of the high innova-
tion groups more often for providing administrative help. Very small differ-
ences in the same direction occur for the role, "providing a fair hearing
for your ideas." Supervisors of the high innovation groups tend to name
outsiders less often for a fair hearing and more often for administrative help.

DISCUSSION

The findings of this exploratory study indicate that the problem
solving of scientific groups is facilitated by their supervisors, fellow
group members, and scientists from outside the groups. Group members are
especially helpful in performing technical roles; supervisors and outsiders
are helpful in both technical and administrative areas.

A comparison of the relatively high and low innovation groups in
this laboratory indicated that the high innovation groups tend to work more
as a technical team. Members name each other more often as helpful for
performing technical roles. Detailed examination of the supervisor's role
nets indicated that he is very much a part of that team. The supervisors of
the high innovation groups were named more often by their groups for per-
forming technical and administrative roles; they were named less often by
outsiders as helpful for most roles; and, in turn, they received more
technical help from their groups.



15

As the technical decision-making process evolves, it appears more apt
to result in innovative work if certain colleague roles are performed by
certain parties, especially in the stages of problem solving which theoret-
ically occur earlier (see Farris, 1971). During the idea suggestion stage,
the roles most associated with group innovation are the supervisor's
receiving original ideas from more outside sources but having fewer original
ideas himself, group members providing each other with technical information,
and the availability of organizational information from fewer sources inside
or outside the group. During the proposal development stage, the high innova-
tion groups are characterized by greater exchange of help among themselves
in thinking through technical problems and greater usefulness of their
supervisors in critically evaluating their ideas.

Like most field research in organizations, this exploratory study
suffers from the common problems of small sample size, failing to sample
from a finite population, and inability to determine causality. Thus, these
findings, although they are based on consistent trends, should be regarded
as tentative. To the extent that they accurately describe colleague inter-
action in the problem-solving process of these scientific groups, however,
they have some intriguing implications for theories of problem solving and
leadership as well as some practical applications. Let us turn to these now.
Theoretical implications

Three types of theories of group problem solving were mentioned earlier
in this paper: those which emphasize the role of the leader (e. g., Maier, 1967
and Bowers & Seashore, 1966), those stressing peer leadership in roles per-
formed by group members for one another ( e. g., Benne & Sheats, 1948; Bales,
1950; and Bowers & Seashore, 1966), and those which emphasize the group
in its organizational context. The tentative findings of this study have
implications for each type of theory.



16

Maier (1967) suggests that a group is most apt to succeed in its
problem-solving efforts when its leader performs an integrative function
analagous to that of the nerve ring of the starfish. He does not dominate
the discussion and produce the solution, but instead serves as an integrator
by receiving information, facilitating communication among group members,
relaying messages, and integrating ideas so that a single unified solution
can occur. Moreover, "the idea-getting process should be separated from
the idea-evaluation process because the latter inhibits the former". (Maier,
1963, p. 247.)

Supervisors of the high innovation groups in this study were seen as
behaving very much in the way Maier says they should. They were named
more often by their groups as useful for facilitating thinking and pro-
viding critical evaluation, two roles which can be considered integrative


1

Online LibraryGeorge Franklin FarrisSupervisory roles, colleague roles, and innovation in scientific groups → online text (page 1 of 2)