December 30th, 2012 by Ελευθέριος Αγγελής
The important role of Software Cost Estimation and the impact of accurate predictions of cost for both managers and customers have been portrayed through the evolutionary introduction of a plethora of methodologies, so far in the literature. Despite the variety of proposed models and their promising results, the problem of inconsistent findings and conclusion instability have swiftly attracted the researchers’ interest. Among to a variety of significant limitations of the past studies are certainly the lack of appropriate and well-established statistical frameworks and the absence of automated tools that would reinforce and lead the whole comparison process.
Here, we present StatREC, a Graphical User Interface statistical toolkit, which offers a great flexibility for analyzing results from experimental studies through both traditional and modern statistical hypothesis testing procedures. In addition, the toolkit provides an easy-to-use way for the visualization of prediction performance that can be used from both project managers and non-experts in order to facilitate the strategy for a wise decision-making concerning several issues of a new software project. The applications through StatREC and the formal framework illustrate the benefits and the significant information that are revealed from the utilization of the proposed toolkit.
The first prerequisite of StatREC is R.
StatREC is not at liberty to redistribute R, but it is downloadable for free from:
StatREC also needs .Net Framework 4.5 (Free Microsoft Software) which is essentially an update to .Net Framework 4.0 downloadable from
— Tip — StatREC’s Installer will automatically download .Net Framework 4.0 if it is not installed on the computer.
Warning #1: To download any and all of the above, internet connection is needed
Warning #2: StatREC sends a report to developers in case of crash.