By Mark Blomsma
Abstract:
In the course of 2006 a study book on SDI, “An Introduction to Spatial Data Infrastructures”, has been written. Thirty-five students wrote reviews on the different chapters of this book. The thesis aimed to extract the main suggestions for improving the book chapters from these reviews. It also aimed to develop and analyse methods that can be used in future book writing to analyse book chapter reviews. Four methods were eventually selected for analyses. Two of these methods were taken from existing scientific work. It was then tried to improve the second method, a scoring card, resulting in a third method. These three methods all test reviews as a whole. Using logical deduction instead of existing work, a fourth method was created from scratch that regarded not the whole review, but each single comment in each review. Analyses showed that of the four methods, three yielded better then random results. The existing scoring card method was fast and efficient at selecting reviews. The altered version of this scoring card initially performed worse, but was found to need calibration after which it out-performed the existing scoring card. It was therefore concluded that improving an existing scoring card can be efficient only if enough time for calibration exists. The fourth method was found to select good comments, and the selected comments from this method were indeed presented as main suggestions for improvement of the book chapters. It was however concluded that the selected suggestions were a very small selection and improvement of the method should result in a greater selection of comments and suggestions. It was also attempted to combine the scoring card methods with the loose comment method. It was found that these combinations yielded similar results but took 50% more time and were therefore not recommended for use.