返回Google圖書搜尋

Optimal Matching Analysis

出版	SAGE Publications Limited, 2020
ISBN	15264210389781526421036
URL	http://books.google.com.hk/books?id=WdRdzwEACAAJ&hl=&source=gbs_api

註釋Optimal matching analysis (OMA) is used to analyse longitudinal data. It is of particular use when the order of the data is important, for example, the timing of different events. OMA is the most widely used method of analysing sequential data. A typical example is an individual (the case) and their employment status (the variable) measured at discrete points in time (say every year) over a certain period. Not just individuals but organisations, groups, nations, and so on, can constitute the case. The variable for each case is measured over time and thus represents a connected ordered sequence, a historical representation of how that case has evolved. The connectedness of the values that the variable takes over time is important as each sequence is able to articulate complex changes in status - some cases may have a volatile character with many status changes, while others are relatively stable, with low levels of variability from one time point to the next. This procedure can handle longitudinal complexity in distinctive ways: (1) it does not require a focus on a specific event or duration, (2) it uses all available data when processing sequence comparisons (i.e., all time points in the sequence are used and treated as a single dependency), and (3) data processing facilitates the identification of patterns or structures among the sequences. The value of OMA is in being able to process data so as to preserve its longitudinal complexity and to provide a framework for the exploration of structural patterns within sequential data.