identification of the precise relationship between chosen variables in a controlled environment using qualitative analytical techniques. He noted that the idea here is that
field experiments try to construct an experiment in a more realistic environment than in laboratory setting.
According to Galliers 1992, the major advantage of the approach is the ability to control and isolate a small number of variables, which may then be studied
intensively. The major weakness of the approach is the difficulty to identify relationships which are exist in the real world due to over-simplification of the
experimental situation and the isolation of such situations from most of the variables that exist in the real world. In addition, the problem of replication as it is extremely
difficult to achieve sufficient control to enable replication of the experiment with only the study variables being altered. The difficulty of finding organisations prepared to
be experimented on also becomes the weakness of this approach. In any experimental design, randomisation and experimental control become
essential elements Zmud et al., 1989 in Braa et al., 1999. Randomisation is to allocate the people or units being studied to the experimental group, or to a control
group, on an entirely random basis, without taking account of their characteristics or preferences. Experimental control is to take appropriate steps to get rid of nuisance
variables, which are factors other than the independent variables that might be responsible for observed changes in the dependent variable. There are two types of
field experiments Cook et al., 1979; Zmud et al., 1989 in Braa et al., 1999. The first is true experimental design, which has the criteria of multiple treatments or one
treatment and a control group, randomisation, and experimental control. The second is quasi experimental design, which does not meet the three criteria rather attempt to
maintain as many of the properties of true experimentation as possible, given the constraints of the research setting.
1.20.4. Action Research
Unlike other experimental methods, action research bases on an interpretative philosophical framework Susman et al., 1978 in Baskerville et al., 1998. The
response of state-change must be interpreted in the social context as the same change in different social may produce different response because of the multivariate nature
of human social interaction. Baskerville and Wood-Harper 1998 then summarises the characteristic of action research as follow: multivariate social setting, highly
interpretive assumptions about observation, intervention by researcher, participatory observation, and the study of change in the social setting.
According to Checkland 1991 in Baskerville et al., 1996, action research is a cycle of continuous inquiry concerning interaction between theory and practice.
Checkland proposes a model comprising of three essential element i.e. an intellectual framework of linked ideas F - a theory, a methodology for using this framework M
and area of application A - research question. In Figure 3.3 Checkland depicts how the model cycles the research element of F and M through A to generate reflection,
action and scientific finding.
Figure 3.3 The Cycle of Action Research Checkland, 1991
In order to achieve scientific rigour on action research, Baskerville and Wood- Harper 1996 suggested to use Susman and Evereds 1978 cyclical five phase i.e.
1 diagnosing, 2 action planning, 3 action taking, 4 evaluating, and 5 specifying learning, which require the establishment of a client-system infrastructure
or research environment. The client-system infrastructure is the specification and agreement that constitutes the research environment. It provides the authority, or
sanctions, under which the researchers and host practitioners may specify actions and provides the legitimation of those actions as beneficial to the client or host
organization. Considerations include the boundaries of the research domain, and the entry and exit of the scientists. It must also patently recognize the latitude of the
researchers to disseminate the learning gained in the research Baskerville et al.,
1996. Baskerville and Wood-Harper 1996 then note that Hult and Lennung 1980
define action research concisely: Action research simultaneously assists in practical problem-solving and expands scientific knowledge, as well as enhancing the
competencies of the respective actors, being performed collaboratively in an immediate situation using data feed back in a cyclical process aiming at an increased
under-standing of a given social situation, primarily applicable for the understanding of change processes in social systems and undertaken within a mutually acceptable
ethical framework.
1.20.5. Action Case