This is not a typical research output article but a commentary article written by Howard Wainer, a distinguished research scientist at the National Board of Medical Examiners. Being one of the authors of the popular book recently published on testlet response theory, he not only wrote commentary on testlet theory and its applications but also wrote some based his observations and experience. With the title “14 conservations about three things”, in this article the author shared at the beginning of 2010 a total of 14 pieces of his personal views on the history and future development in the field of educational and psychological measurement.
He has provided his comments in a number of sections. One of the most interesting sections was written on his views on some areas that do not worth much more attention of researchers in the field in the future. The reason behind is that he thinks there has too much, and enough, research being done in these areas such as differential item functioning (DIF) and the Rasch model, i.e. areas such as these have received many great development in the past but . Whether an area is worth more research and development or less is an area open for discussion and debate. In particular, for the area of DIF, he mentioned correctly there are two approaches for studying DIF and they are capable of dealing with any situations that require DIF analysis. However, the development in DIF is probably not limited in approaches of detecting DIF.
Besides flagging the possible of different item response behaviours between manifest groups in a sample, DIF, in a broad sense, does not necessarily exist among manifest groups only. The heterogeneity within sample may not exist as a function of the manifest groups, and in such cases, common procedures in DIF detection, i.e. by testing item parameter invariance between sub-groups, may be inadequate to exclude the possibility of DIF that may exist without any association with manifest groups. Despite that the author’s relatively pessimism about the future of DIF development, there are, therefore, other areas of DIF that still worth research findings to fill the current knowledge gap.