Supporting Selection of Statistical Techniques
In this paper we describe the necessity for a semi-structured approach towards the selection of techniques in quantitative research. Deciding for a set of suitable techniques to work with a given dataset is a non-trivial and time-consuming task. Thus, str
- PDF / 197,030 Bytes
- 9 Pages / 439.36 x 666.15 pts Page_size
- 35 Downloads / 177 Views
Abstract In this paper we describe the necessity for a semi-structured approach towards the selection of techniques in quantitative research. Deciding for a set of suitable techniques to work with a given dataset is a non-trivial and timeconsuming task. Thus, structured support for choosing adequate data analysis techniques is required. We present a structural framework for organizing techniques and a description template to uniformly characterize techniques. We show that the former will provide an overview on all available techniques on different levels of abstraction, while the latter offers a way to assess a single method as well as compare it to others.
1 Introduction Researchers and students engaging in quantitative analysis of their data are always faced with a set of decisions as to which techniques to use for the data set at hand. The decisions made in quantitative analysis are motivated by financial or temporal efficiency, trends in previous choices of other researchers or plain convenience. They are certainly—at least to some extent—motivated by technical and functional aspects. Up to now, there are no standards to support decisions of that kind. It is important to mention here that our approach does not aim to make a decision for a researcher or student, but to support the decision making process. Despite the fact that some consider elegant data analysis an art, the aforementioned constraints often stand in the way of treating it accordingly. Especially in early stages of their education, students may not have acquired the expertise to handle data analysis problems adequately yet. K.F. Hildebrand () European Research Center for Information System (ERCIS), Münster, Germany e-mail: [email protected] M. Spiliopoulou et al. (eds.), Data Analysis, Machine Learning and Knowledge Discovery, Studies in Classification, Data Analysis, and Knowledge Organization, DOI 10.1007/978-3-319-01595-8__39, © Springer International Publishing Switzerland 2014
359
360
K.F. Hildebrand
Furthermore, it can be difficult to identify suitable techniques, because there is little effort from experts developing new quantitative methods to make them available to non-expert communities in an understandable way. Being understandable in this context does not mean to strip the techniques from their core contents and present them in a superficial fashion. On the contrary, we are going to propose an approach that will preserve all necessary complexity but only reveal it in a stepwise manner. Thus, more complexity is only introduced when required for decision in favor of or against a data analysis technique. The remainder of the paper will be structured as follows: We will first analyze, what previous work has been done with regard to a framework for data analysis and a description structure for data analysis techniques (Sect. 2). We will then present our approaches to these two items in Sects. 3 and 4. A discussion as well as an outlook on upcoming research will conclude the paper.
2 Related Work With regard to providi
Data Loading...
