Some improvements for action research and information systems development

We study a real, not esoteric, nor imaginary phenomenon. We concentrate on studies with one-directional relationships between variables. They form a majority of studies, and other studies with two-directional relationships are a minority. Burrell and Morgan (1979) found four paradigms for the studies, but half assumed one-dimensional relationships, and others two-dimensional ones. We exclude the latter from further consideration. The former set of studies assumes a stable state, but in reality, there are also phenomena assuming a change, for example, information systems development (ISD) and action research (AR). Baskerville and Wood-Harper (1998) showed that 10 ISD methods are AR methods. Here, we contribute science by proposing the fifth paradigm (a shared change). We also analyze this paradigm and ISD & AR studies further. In all the studies, there are two groups: 1) traditional studies with the stable state, and 2) the ISD & AR studies trying to cause a change. This differentiation seems to describe the fundamental nature of those two study groups. This differentiation can be called backward- and forward-looking. We shall also contribute by presenting some characteristics of the backward- and forward-looking studies. The primary development of an IT artifact and standard updates are performed in software houses, requirements of special maintenance and change are carried out in an organization. In the latter and AR efforts, an organization plays a central role in many ISD/AR project activities. We analyze the beginning, guidance, and possible ending of those projects, too. Finally, we present implications for science and practice, limitations, and further research.