A simulation approach for the analysis and forecast of software productivity

Measuring and predicting software productivity: A systematic map and review

“…The models yielded expected results that concur with facts that have been found in the literature [58,59,34]. A limitation is that the studies [57,59,34] did not use the simulation to predict based on an actual industrial case. Only one study [58] compared the simulated results with actual data, showing a very promising outcome as there was very little difference between actuals and the simulation.…”

“…The solution presented in [57] was very limited with regard to variables as only size and effort were simulated. The models yielded expected results that concur with facts that have been found in the literature [58,59,34].…”

“…Continuous/system dynamic models Table 12 provides an overview of studies evaluating continuous models to predict software productivity. Romeu [57] built a simulator that pools confounding factors into a single random variable that adjusts values of size and effort by the random variable. Running the simulation several times will produce pairs of predicted size and predicted effort.…”

“…First, a reference set of articles was created including the articles we knew should be found by the search, and are relevant to be included in the review. The reference set consisted of the following references: [17,20,23,24,26,30,32,34,37,54,55,57]. Secondly, the reference lists of the included articles were reviewed for papers referring to measurement/prediction of software development productivity/performance based on titles.…”