Two categories of generalizing result from case study to theory pointed out by Yin(1984, p.38) are statistical analytic generalization and statistical generalization. The method of generalization is analytic generalization in which a previously developed theory is used as template with which to compare the empirical results of case study. If two or more cases one shown to support the same theory, replication maybe claimed. On the other hand, in statistical generalization an inference is made about a population(or universe) on the basis of empirical data collected about a sample. From those two methods, developmental research is more close to analytic generalization with some case studies. It will be clearly illustrated by an article bellow.

Definition The definition of development research is disciplined inquiry conducted in the context of the development of a product or program for the purpose of improving either the thing being developed or developer: capabilities to develop better things of this kind, or both.

Methods The examples of 'Development Research' are used in group problem-solving(Treisman, 1983); develop computer program(White,B.,1984); sesame stree(Flag, 1990); Linda(about action research) that have purpose to develop more effective programs use the method that has some steps as follow: by selecting the situation; do observation and ask some questions; interpret and process the description; build the prototype and use it with the target; observing and questioning; and revised the prototype. Last example (Bresler, et.al) is action research; study the success of an arts program. Here evaluation done well after the program had been developed.

But, those researches have weakness such as sample size to small, intuitive situation, mainly by observations only, unformal procedures, no tests, no explicit theoretical framework. It means they lack of quality if judged by qualitative research indicators or naturalistic study such as eight issues in Smith&Glass(1987).

But there are some distinctive problems( 6 examples that question using HOW and WHAT) for methodological challenges of development research that make different with qualitative study and naturalistic study, in other word only development research methods will resolve these problem better than other method can. Let take an example of the problem: To WHAT range of students and situations can the findings of the study be generalized? Strategy for solve this problem using replication, uses cases studies of small samples, etc. Contrary to traditional scientific educational research, developmental research can be valuable by the standard of improving the plans and materials understanding and insight into ways to design effective materials, programs, or products.