In the traditional quantitative research, the researchers will firstly propose several hypotheses of the subject according to their experiences, and then determine whether there exists a statistical significance in some hypothesis in a try-and-error manner. The quality of the result using the manner, of course, is in accordance with the quality of the hypotheses made by the researchers, and the process of finding the statistically significant differences is highly dependent on researchers’ intuition and experience. For example, in a questionnaire survey of elementary school students’
Internet usage behavior, a researcher might make a hypothesis, “There is a difference between different genders about the hours they access the Internet every week,” and then use appropriate inferential statistics method according to his/her knowledge of statistics to test this hypothesis. Without making the hypothesis, the statistically significant difference can not be found even if it really exists. Therefore, how to acquire the knowledge and experience of senior researchers might be helpful for the junior researchers.
Besides, granularity of original questionnaire data may not be good enough to find the statistically significant differences. For example, in a questionnaire survey of elementary school students’ Internet usage behavior, if there is no significant difference between different resident regions about the hours students access the Internet every week, the researcher might conclude there is no significant difference between different resident regions. However, each region may contain several counties in geography. By drilling down the student’s resident dimension to lower levels of granularity, it is still possible to find a significant difference between different counties.
When researchers analyze a questionnaire data, they have to make some hypotheses of the possible statistically significant differences from the data and make some selections of appropriate inferential statistic methods to test their hypothesis according to their intuition, experience, and knowledge.
However, the statistic-related assistant tools nowadays are not intelligent. These tools do not support users assistants when they analyze questionnaire, for example give some advices for the selections of statistical methods. They only provide assistants on computations of statistical methods.
In our previous research, we have proposed some techniques on assistant questionnaire analysis. But these research are theoretical, we have not use these research on a real system for users to use. We want to use these research to design and implement an intelligent questionnaire analysis assistant system.
In order to assist junior researchers in analyzing questionnaire data, we acquire the knowledge and experience of senior researchers to construct a forward-chaining rule-base expert system. This expert system assists researches in making hypotheses of the possible statistically significant differences from the data and making selections of inferential statistic methods to test researchers’ hypotheses. According to the experts’ experiment and knowledge, three indicators, Increase, StepDown, and Dice, are designed to help researchers finding the possible statistically significant differences. The experts also define metarules to determine degree of indicators.
Hence, a significant difference viewer is constructed based on the indicators and metarules in the expert system to assist junior researchers in exploring data. For the need of selecting appropriate inferential statistic methods to test researchers’
hypothesis, the expert system is also designed to give suggestions for appropriate statistic methods to test hypotheses. The expert system gives explanations about why these methods are appropriate to analyze the data. Since the methods are suggested from the expert system, researchers may not necessarily understand what these methods are and how to use them. But researchers may want to know the detail information of the methods, for example, meaning of the method, how to use the method. The expert system is designed to provide a learning platform for junior researchers to learn these methods. Therefore, junior researchers can learn which methods are appropriate to use.
One of the issues in designing a real online system is how to design a system having good maintainability and extensibility. This is the most difficult part for
system design and implementation. We need to design functions provided for users to assist them, design architecture of the system supports those functions, and formulate clearly and completely those diagrams about system design, for example, use case diagram. Besides, when implementing this system, we need to integrate and adjust those techniques used in the system to adapt for requirements of the system. This is also a difficult part in system implementation.
The rest of this thesis is organized as follows. In Chapter 2, we introduce some preliminaries about the techniques used in the system, data warehouse and OLAP, significant difference, indicator, DRAMA/NORM, Ontology-based Learning Sequences Construction Algorithm [8], Ontology-based Adaptive Learning Sequences Construction Algorithm [5], and the e-learning architecture proposed by Chang [5].
Chapter 3 presents the methods we used to design and implement the system and functions, some schemas we designed for the system. Chapter 4 shows the overall system architecture and describes the system in detail. Chapter 5 gives the results of the experiments we designed to evaluate the system. Finally, concluding remarks are given in Chapter 6.