THE EFFECT OF ICT ENABLED SOCIAL NETWORKS ON

PERFORMANCE

Kon Shing Kenneth Chung and Liaquat Hossain

School of Information Technologies, University of Sydney, Camperdown, NSW 2006, Australia

Keywords: ICT use, social network structure, individual performance, process of care, general practitioners.

Abstract: Research in sociology of information and technology use highlight the importance of information and

communication technology (ICT) for supporting networks among professional peers. This paper explores

the effects of different types and patterns of ICT usage for supporting professional networks. In this study,

we ask—how can different types ICT help support professional networks? What are the implications of

patterns of ICT usage on supporting professional networks? Given an association between ICT types, usage,

network structure for supporting professional networks, what are its implications on performance? In our

study, we apply the theory of structural holes and its underlying assumptions (i.e., efficiency and

effectiveness) to develop a theoretical framework and propositions for exploring interlock between types

and patterns of ICT usage, network structure, and professional support networks for measuring the

performance of effective clinical care. Empirical evidence from Australian rural general practitioners’ data

show that both network structure and Internet-enabled ICT use play a crucial role when it comes to

performance in the medical consultation.

1 INTRODUCTION

The introduction of information and communication

technologies (ICT) into organisations and

occupational communities has produced unfathomed

changes in work patterns and communication

structure (1991). Literature in Information Systems

widely recognises that ICT allows for synergistic

goal achievement when used and aligned properly

with the task at hand (Goodhue et al., 1995). In

general practice, a number of studies highlights the

extent of computerisation in general practice,

although recent studies have surveyed actual use of

ICT by general practitioners (GPs). Australian

national surveys on ICT use by GPs was conducted

in 1998 (Nielsen, 1998), 2001 (Western et al., 2001),

and most recently in 2006 (Henderson et al., 2006;

McInnes et al., 2006). Although these studies

explore the use of ICT for clinical tasks in general

practice, none questions its use for communication

with peers, especially in rural practice. Building

upon theories of social network structure, we

introduce a conceptual model for exploring the

relationship between network structure, ICT use and

attitudes to processes of medical care. We associate

density, efficiency and effective size of a GP’s

professional network with attitudes to processes of

medical care. Based on literature on impacts of

technology use, we also explore task-level effects

and communication-structural effects on attitudes to

medical care. Preliminary results show that while

professional networks are important, internet-

enabled ICT use rather than ICT use for clinical

tasks accounts higher for performance in the medical

consultation. The following section discusses the

importance of professional networks and ICT use in

the context of rural general practice.

2 CONTEXT OF THE STUDY

The first study conducted nationally in Australia

claimed that computers were present in 31% of

Australian general practices (Nielsen, 1998). In

2001, this figure increased to 86% (with a likelihood

of 95% of Australian general practices to be

computerised by 2005), indicating that Australian

general practices were highly computerised

(Western et al., 2003). The study also showed that

GPs practising in capital cities were less likely to be

computerised than in the rural practices. The

primary reason was because rural GPs were

105

Shing Kenneth Chung K. and Hossain L. (2007).

THE EFFECT OF ICT ENABLED SOCIAL NETWORKS ON PERFORMANCE.

In Proceedings of the Ninth International Conference on Enterprise Information Systems - HCI, pages 105-110

DOI: 10.5220/0002355901050110

 SciTePress

geographically distant from other health care

centres, general practices, and hospitals. McInnes et

al’s (2006) recent study found that Australian

general practice has achieved near-universal clinical

computerisation, although Henderson et al’s (2006)

study show that GPs are still reluctant to fully

embrace the technology. Both studies confirm

Western et al’s (2001; 2003) speculation that

Australian general practices are more likely to use

computers for administrative rather than clinical

tasks.

2.1 Research Questions

It is interesting that ICT is hardly used as a

communication medium amongst GPs in Australia.

Given the significant problems that characterise the

rural GP workforce, empirical studies highlighting

GPs’ professional support network structure related

with ICT use are relatively few. The motivating

questions for this research are thus —what are the

implications of patterns of ICT usage on supporting

professional networks? Is there any correlation

between patterns of ICT usage and the social

network structures of the professional networks?

Can we make any distinction between the types,

usage patterns, and network structure for supporting

professional network? Given an association between

types, usage, network structure for supporting

professional networks, what are the implications on

performance? How do these factors relate to

improved performance of the knowledge intensive

work groups?

3 THEORETICAL MODEL

In the following section, we discuss the development

of the constructs of our conceptual model based on

literature review.

3.1 Performance: Attitudes to Care

General practice is the “provision of primary

continuing comprehensive whole patient medical

care” (RACGP, 2004). Therefore, we regard the

‘process of care’ dimension provided by GPs to

patients as relevant for the study. In particular, we

are interested in the effectiveness of the GP in

delivering clinical and interpersonal care. Previous

studies show that comprehensively measuring the

GP’s attitude towards these dimensions of care

comes closest to measuring their actual behaviour

(Cockburn et al., 1987; Howie et al., 1992).

Research in social psychology suggests that a

person’s attitude towards an object may be related to

the overall pattern of a person’s response to that

object (Ajzen et al., 1980). In other words, the

attitude towards behaviours is a direct determinant

of the behavioural intention to perform the

behaviour. It can therefore be argued that

perceptions that the GPs hold are likely to influence

their behaviour in the consultation.

3.2 Network Structure

A social network consists of a group of human

actors that have relationships or ties amongst

themselves (Scott, 2000). Since 1970s, social

networks have been regarded as a promising concept

for becoming a unifying framework in clinical

practice (Erickson, 1975) and hospital settings

(Anderson et al., 1985; West et al., 1999). In

general, social networks theory and practice is

regarded useful for clinical practice in location of

resources and serves as an interpreter of help-

seeking behaviour and utilisation of services.

Numerous studies have documented GPs’ need for

immediate access to information and the importance

of the social network of peers and colleagues. As

Dee and Blazek (1993, p. 263) maintains,

“…colleagues are familiar, reliable, immediately

available, and inexpensive; they give concise,

organised answers that synthesize available

information”. The value of a social contact is hence

much more profound in rural areas, where the

existence of social networks amongst GPs is highly

valued.

Structural properties of social networks have

been known to bear significant impacts on an

individual’s performance. Granovetter (1973) argues

that individuals are more likely to benefit from novel

information if they capitalise on weak ties in their

social network. This implies actors ought to

maintain a number of weak ties in order to capitalise

on information benefits. Coleman (1988) argues that

closure, or closed networks, enables individuals to

combine forces, and provide collective sanctions and

thus create trust by ensuring that obligations and

promises are kept and norms are followed. Thus,

denser networks attain higher closure. Previous

studies have shown that denser ties in an

individual’s social network is conducive to

intellectual performance (Coleman, 1988), job

performance (Sparrowe et al., 2001) and knowledge-

sharing (Cross et al., 2004). Burt (1992) however,

takes on a structural perspective by suggesting that

dense networks are far more inefficient than sparse

networks because (1) they are costly to maintain,

and (2) they provide redundant information.

Structural holes theory is based on the idea that

ICEIS 2007 - International Conference on Enterprise Information Systems

106

actors are in a better position to benefit from their

interactions and transactions with others if actors are

connected to others who are not connected

themselves or well organised. Thus, an optimised

network yields information benefits through non-

redundancy. Optimisation of a social network of an

individual is measured by effectiveness and

efficiency of one’s personal network. Efficiency is

about the channels of access offered by a primary

contact (broker) in a brokerage position to access all

others (ie. secondary contacts) in the network.

Effectiveness is about the total number of contacts

reached along with all the primary contacts.

Effectiveness is hence the yield of the entire

network. In the context of rural GPs, we propose the

following:

 Proposition 1: Density of a rural GP’s

network is positively associated with attitudes

to process of care

 Proposition 2: Effectiveness of a rural GP’s

social network is positively related with

attitudes to process of care

 Proposition 3: Efficiency of a rural GP’s

social network is positively related with

attitudes to process of care

3.3 ICT Use: Clinical and Internet

based

ICTs are replacing traditional resources for

developing an actor’s social network (Nardi et al.,

2000) as they shape personal networks and re-draw

social boundaries. Therefore, a significant construct

in our model is ‘ICT use’. Computer supported

social networks (Wellman, 1996) that connect

people sustain ties that provide information and

social support in both specialized and broadly-based

relationships. Katz and Rice (2002) consider the

Internet as having great potential for connectivity

without much intention or social cost. This is

particularly useful for occupational groups such as

dispersed rural GPs who find maintenance of ties

with peers and communities difficult and expensive

(Pickering et al., 1995). ICT use hence fosters the

connection of ties and contributes to the growth of

social capital by supporting many-to-many

information exchanges among geographically

dispersed people.

At the task-level, research on ICT use by GPs

extend beyond informatics in hospitals and specialty

medicine to include computing in general practice

settings (Aydin et al., 1997). ICT use contributes to

process of care by providing benefits to GPs such as

better storage and retrieval of information, consistent

and accurate records, improved drug management,

and integration of clinical and administrative

functions (Nielsen, 1998). It appears that ICT is

utilised, and is a good fit with the general practice

tasks it supports (Goodhue et al., 1995). Given the

above arguments, we derive the following

propositions:

 Proposition 4: Frequency of use of ICT by

rural GP for clinical-tasks is positively

associated with attitudes to process of care

 Proposition 5: Frequency of use of ICT by

rural GP for internet-based tasks is positively

associated with attitudes to process of care

 Proposition 6: ICT use significantly

moderates the interaction between network

structure and process of care

4 RESEARCH METHODOLOGY

The following section discusses the operational

model highlighting the significance of ICT use as a

variable that moderates the inherent relationship

between network structure and process of care

attitudes.

Figure 1: Operational Model for the Research.

4.1 Egocentric Network Approach

In the egocentric approach, the actor is the “ego”

and his affiliates, advisors, or friends, are known as

“alters” (Scott, 2000). Name generators are used to

elicit alters’ names. To elicit names from a GP’s

professional network, we asked:

“Looking back over the last six months, please

identify people (up to 15 maximum) who are

important in providing you with information or

advice for providing care to patients.”

Other name interpreter items solicited were

strength of each tie, measured by “time known the

person”, “frequency of interaction”, “type of

relationship”, and “degree of closeness” (Marsden et

al., 1984). Attribute data about frequency of

THE EFFECT OF ICT ENABLED SOCIAL NETWORKS ON PERFORMANCE

107

interaction via email, telephone (including mobile),

and video conferencing were also included in the

instrument to segregate face-to-face and ICT media

interactions. Finally, we asked GPs to determine

how the members of their professional network

relate to each other based on a five point degree of

closeness scale ranging from ‘especially close’ to

‘do not know each other’.

4.2 Survey Administration

The original survey was pre-piloted amongst a

group of 5 students (in the research laboratory) and

three rural GPs. Experts in the domain of general

practice, including former president of a rural

doctor’s association in Australia, professor and head

of discipline of general practice in a renowned

university, and rural GPs vetted the survey

instrument, which was then pre-tested for

comprehension and ease-of-use. Finally, the survey

was administered in two phases: first to sixty rural

GPs practising in a southern region of rural NSW

with nineteen responding (response rate: 31%);

second to 46 GPs in a rural doctors conference with

17 responding (response rate: 37%). Administration

was personal, which allowed for capturing of survey

duration, respondent reaction and errors in the

survey, which we eventually rectified.

4.3 Measures

Density is the ratio of existing number ties to the

maximum possible ties possible. Effectiveness is a

measure of the number of alters minus the average

degree of alters within the ego network, not counting

ties to the ego. Efficiency is measured by dividing

effectiveness by the number of alters in the ego’s

network.

For performance, we adapted Cockburn et al’s

(1987) validated instrument to measure four

dimensions of care: mutuality, communication,

responsibility for decisions and appropriateness of

consultations.

Clinical ICT task measures were based on the

reliable and valid item sets used by Western et al

(2001) and includes items such as frequency of

using ICT for “Generating health summaries”,

“Writing prescriptions”, and so on, measured on a

five point scale ranging from ‘Daily use’ to ‘Never’.

Internet task items were adapted from Andrews et

al’s (2004) and included “accessing medical

journals”, “accessing databases”, “consultation with

colleagues”, and so on.

5 RESULTS

There were 36 responses in total. However, 5

specialised in areas such as gynaecology and

orthopaedics, and were excluded from the analysis.

Of the remaining 29 GPs, the mean number of years

practised in a rural setting was 17.21 years (max=40

years; min=1 year).

To associate network structure, ICT use and the

dimensions of care, we ran a multiple regression

based on the four dimensions of care –

communication, mutuality, decision responsibility,

and appropriate consultation as the dependent

variable. We used density, efficiency and effective

size of the network as the network predictor

variables (model 1) and ICT use for clinical

functions and ICT use requiring internet functions as

the second block of predictor variables (model 2).

5.1 Communication Dimension

Table 1: Model summary for Communication Dimension.

Model R

Square

Adjusted

R Square

Std. Error

of the

Estimate

.405(a) .164 .059 5.67385

.408(b) .166 -.023 5.91807

(a) Predictors: (Constant), Network Density, Effective Size,

Efficiency; (b) Predictors: (Constant), Network Density,

Effective Size, Efficiency, Internet Functions, Clinical Functions

Our results indicate a multiple R of 0.405 (model

1) and 0.408 (model 2), which shows substantial

correlation between the predictor variables and the

variable ‘communication’, where approximately

16% of the variance in ‘communication’ is explained

by the two blocks of predictor variables. In model 1,

effective size bears greatest influence on

‘communication’ (β=0.002). In model 2, ICT use

(requiring Internet) bears greatest influence on

‘communication’ (β=0.066).

5.2 Mutuality Dimension

Table 2: Model Summary for Mutuality Dimension.

Model R

Square

Adjusted

R Square

Std. Error

of the

Estimate

.417(a) .174 .071 4.11406

.431(b) .186 .001 4.26593

(a) Predictors: (Constant), Network Density, Effective Size,

Efficiency; (b) Predictors: (Constant), Network Density,

Effective Size, Efficiency, Internet Functions, Clinical Functions

ICEIS 2007 - International Conference on Enterprise Information Systems

108

There is substantial correlation between the

predictor (independent) variables and the dependent

variable ‘communication’ (R = 0.417 and 0.431

respectively), where approximately 17-18% of the

variance in ‘mutuality’ is explained by the two

blocks of predictor variables. In model 1, effective

size bears greatest influence on ‘mutuality’

(β=0.209). In model 2, effective size bears greatest

influence on ‘mutuality’ (β=0.204), followed by ICT

use (requiring Internet) (β=0.134).

5.3 Decision Responsibility Dimension

Table 3: Model Summary for Decision Responsibility.

Model R R Square

Adjusted

R Square

Std. Error

of the

Estimate

.316(a) .100 -.013 3.20666

.363(b) .132 -.066 3.28896

(a) Predictors: (Constant), Network Density, Effective Size,

Efficiency; (b) Predictors: (Constant), Network Density, Effective

Size, Efficiency, Internet Functions, Clinical Functions

The results indicate considerable correlation

between the predictor (independent) variables and

the dependent variable ‘decision responsibility’ (R =

0.316 and 0.363 respectively), where approximately

10-13% of the variance in ‘decision responsibility’ is

explained by the two blocks of predictor variables.

In model 1, effective size bears greatest influence on

‘decision responsibility’ (β=0.127). In model 2,

effective size also bears greatest influence on

‘decision responsibility’ (β=0.240), followed by ICT

use (requiring Internet) (β=0.116).

5.4 Appropriate Consultation

Dimension

Table 4: Model Summary for Appropriate Consultation.

Model R R Square

Adjusted

R Square

Std. Error

of the

Estimate

1 .070(a) .005 -.120 6.35078

2 .260(b) .068 -.144 6.42042

(a) Predictors: (Constant), Network Density, Effective Size,

Efficiency; (b) Predictors: (Constant), Network Density, Effective

Size, Efficiency, Internet Functions, Clinical Functions

There is considerable correlation between the

predictor (independent) variables and the dependent

variable ‘appropriate consultation’ (R = 0.070 and

0.260 respectively), where approximately 6% of the

variance in ‘appropriate consultation’ is explained

by predictor variables in the second model. Model 2

also indicates that efficiency (β=0.796) and network

density (β=0.779) bears greatest influence on

‘appropriate consultation’, followed by ICT use

(requiring Internet) (β=0.343). This is probably true

because GPs are known to consult with their peers

and specialists for appropriate treatment. In such

cases, GPs consult those who can provide

information benefits to them in the shortest amount

of time (efficiency). Furthermore, it also indicates

that GPs who use internet functions (eg. for

communicating with peers and for accessing

educational materials) are more likely to provide

appropriate consultations.

6 CONCLUSION

Although our results are limited and not

generalisable to the population of GPs in Australia,

it shows that while individual professional network

properties are important in explaining process of

care attitudes, ICT use that required internet access

was equally important. GPs who used the internet

for communication (with professional networks of

peers, special interest groups) and accessing online

materials were better in communicating with

patients, making decision responsibilities, and

conducting appropriate consultations. Such GPs also

shared mutuality with their patients. ICT use for

clinical tasks did not contribute to either of the

dimensions of care. This is most likely the case

because GPs in the consultation process do not

usually have time to retrieve and digest information

from most forms of ICT (Bolton, 2006). At best,

ICT for clinical tasks are useful for quality assurance

such as checking drug interactions. Our results also

confirmed that effective size (or reach in the

network) was the most important predictor of

communication, decision responsibility and

mutuality dimensions of care. Efficiency (or access

to novel and useful information) and network

density (closure), on the other hand, were most

important predictors for appropriate consultation.

REFERENCES

Ajzen, I., & Fishbein, M., 1980. Understanding Attitudes

and Predicting Social Behavior. Engelwood Cliffs

New Jersey: Prentice Hall.

Anderson, J. G., & Jay, S. J., 1985. Computers and

Clinical Judgement: The Role of Physician Networks.

Social Science & Medicine, 20(10), 969-979.

THE EFFECT OF ICT ENABLED SOCIAL NETWORKS ON PERFORMANCE

109

Andrews, J. E., Pearce, K. A., Sydney, C., Ireson, C., &

Love, M., 2004. Current State of Information

Technology Use in a US Primary Care Practice-based

Research Network. Informatics in Primary Care,

12(1), 11-18.

Aydin, C. E., & Forsythe, D. E., 1997, 25th-29th October.

Implementing Computers in Ambulatory Care:

Implications of Physician Practice Patterns for System

Design. Paper presented at the Proceedings of the

American Medical Informatics Association 1997

Annual Fall Symposium, Nashville.

Bolton, P. G. M., 2006. Information Mastery and 21st

Century General Practice. The Medical Journal of

Australia, 185(2), 94.

Burt, R. S., 1992. Structural Holes: The Social Structure

of Competition. Massachusetts: Harvard University

Press.

Cockburn, J., Killer, D., Campbell, E., & Sanson-Fisher,

R. W., 1987. Measuring General Practitioners'

Attitudes Towards Medical Care. Fam. Pract., 4(3),

192-199.

Coleman, J. S., 1988. Social Capital in the Creation of

Human Capital. The American Journal of Sociology,

94, S95-S120.

Cross, R., & Cummings, J. N., 2004. Tie and Network

Correlates of Individual Performance in Knowledge-

Intensive Work. Academy of Management Journal,

47(6), 928-937.

Dee, C., & Blazek, R., 1993. Information Needs of the

Rural Physician: A Descriptive Study. Bulletin of the

Medical Library Association, 81(3), 259-264.

Erickson, G. D., 1975. The Concept of Personal Network

in Clinical Practice. Family Process, 14(4), 487-498.

Goodhue, D. L., & Thompson, R. L., 1995. Task-

Technology Fit and Individual Performance. MIS

Quarterly, 19(2), 213-236.

Granovetter, M. S., 1973. The Strength of Weak Ties. The

American Journal of Sociology, 78(6), 1360-1380.

Henderson, J., Britt, H., & Miller, G., 2006. Extent and

Utilisation of Computerisation in Australian General

Practice. The Medical Journal of Australia, 185(2),

84-87.

Howie, J. G. R., Hopton, J. L., Heaney, D. J., & Porter, A.

M. D., 1992. Attitudes to Medical-Care, the

Organization of Work, and Stress among General-

Practitioners. British Journal of General Practice,

42(358), 181-185.

Katz, J. E., & Rice, R. E., 2002. Social Consequences of

Internet Use: Access, Involvement, and Interaction

(1st Edition ed.). London: MIT Press.

Marsden, P., & Campbell, K. E., 1984. Measuring Tie

Strength. Social Forces, 63, 482-501.

McInnes, D. K., Saltman, e. C., & Kidd, M. R., 2006.

General Practitioners' Use of Computers for

Prescribing and Electronic Health Records: Results

from a National Survey. The Medical Journal of

Australia, 185(2), 88-91.

Nardi, B. A., Whittaker, S., & Schwarz, H. (2000). It's Not

What You Know: Work in the Information Age.

Retrieved 12th January, 2005, from

http://www.firstmonday.org/issues/issue5_5/nardi/inde

x.html

Nielsen, A., 1998.

A Study into Levels of, and Attitudes

Towards Information Technology in General Practice

(No. S2587/NG2132). Sydney: Commonwealth

Department of Health and Family Services.

Pickering, J. M., & King, J. L., 1995. Hardwiring Weak

Ties: Interorganizational Computer-mediated

Communication, Occupational Communities, and

Organizational Change. Organization Science, 6(4),

479-486.

RACGP. (2004). Rural General Practice - Definition.

Retrieved 14th November, 2004, from

http://www.racgp.org.au/document.asp?id=15301

Scott, J., 2000. Social Network Analysis: A Handbook.

London: SAGE Publications.

Sparrowe, R. T., Liden, R. C., Wayne, S. J., & Kraimer,

M. L., 2001. Social Networks and the Performance of

Individuals and Groups. Academy of Management

Journal, 44, 316-326.

Sproull, L., & Kiesler, S., 1991. Connections: New Ways

of Working in the Networked Organization.

Cambridge: MIT Press.

Wellman, B., 1996. For a Social Network Analysis of

Computer Networks: A Sociological Perspective on

Collaborative Work and Virtual Community. Paper

presented at the Proceedings of the 1996 ACM

SIGCPR/SIGMIS conference on Computer Personnel

Research Denver, Colorado, United States

West, E., Barron, D. N., Dowsett, J., & Newton, J., 1999.

Hierarchies and Cliques in the Social Networks of

Health Care Professionals: Implications for the Design

of Dissemination Strategies. Social Science and

Medicine, 48, 633-646.

Western, M., Dwan, K., Makkai, T., Mar, C. d., &

Western, J., 2001. Measuring IT use in Australian

General Practice. University of Queensland: General

Practice Computing Group.

Western, M. C., Dwan, K. M., Western, J. S., Makkai, T.,

& Mar, C. D., 2003. Computerisation in Australian

General Practice. Australian Family Physician, 32(3).

ICEIS 2007 - International Conference on Enterprise Information Systems

110