A model of factors affecting an information system's change in state

Abstract

SUMMARYThis paper utilizes a theoretically-grounded model of information systems change together with data from 1891 maintenance projects to test the effects that four factors have on the volatility index of application software. The volatility index is a measure of the relative cost of doing maintenance on the deep structure of a system. Two factors were found to be associated with a higher volatility index, the age of the system and the size of the system. One factor was found to be associated with a lower volatility index, the use of higherlevel languages of implementation. One factor was found to be unrelated to the volatility index, the time period when the change was made. Some factors may be under management control, e.g. the language of implementation, while others are usually system characteristics, e.g. system size. Knowledge of the effects of these factors could influence how researchers work with information systems and how information systems management does its work.

Determinants of software volatility: a field study

“…This answer may vary from company to company. Heales [13,34] remarks that there is a threshold x for each system such that when the system volatility index exceeds x the system changes states. The literature suggests that enhancement changes contribute to the high cost of software maintenance.…”

Theoretical perspectives in IS research: from variance and process to conceptual latitude and conceptual fit

“…Putting these two pieces together, one could critique the fact that the model currently 'stops' at organizational impact, and propose instead that a successful IS is not only reflected in its impact at a point in time, but also in an organization's commitment to rectifying negative impacts, for example, by engaging in adaptive or corrective maintenance (Heales, 2002;Clark et al, 2007). This would correspond with the view that performance outcomes 'are just way-stations in ongoing processes … [and] interpretations of performance can have important effects on subsequent actions' (Langley & Abdallah, 2011, p. 211).…”

“…• Second, note that one way to extend the three-stage model of creation→ use → consequences would be to recognize that systems are not just created once and for all, but maintained over time (Heales, 2002).…”

Fat tails, long memory, maturity and ageing in open-source software projects

“…A popular measure, called software volatility, measures the number of enhancements per unit of time over a specified time frame [8]. High volatility is often associated to high maintenance costs; according to this point of view, when the volatility exceeds some threshold it may be more economical to rewrite the entire system from scratch instead of maintaining an aged (and unstable) software system [9,10]. We show here however that the probability distribution of scaled waiting times of the software projects converges towards a universal function, providing a contrario a measure of software maturity.…”