返回Google圖書搜尋

Collecting, Coding, and Analyzing Observational Records from Real Organizations

其他書名	“Best in Class” Analytical Techniques
出版	SAGE Publications Limited, 2023
ISBN	15296673729781529667370
URL	http://books.google.com.hk/books?id=G5HgzwEACAAJ&hl=&source=gbs_api

註釋Experimental manipulation through random assignment has important benefits for social scientists and data analysts, but this method is not always possible, feasible, or desirable. In such circumstances, researchers often analyze observational data (also termed archival data), which are typically generated and retained for purposes other than addressing research questions. Analysis of observational records generally allows researchers and practitioners to observe patterns or correlations between variables of interest, but demonstrating causality or explaining "why" a relationship exists may not be possible. Given this substantial constraint, methodologists have developed key analytical techniques that enable researchers to address certain fundamental challenges inherent in assessments of observational data. Using our previously published study of applicant endorsements in a full-time MBA program as a reference, we overview our use of several such analytical techniques, including blinded assessments, time gaps/lags, matching, selection models, and multi-method research. We close with an overview of other prominent techniques not used in our study but similarly valuable for the analysis of observational data including natural experiments, instrumental variable analyses, and longitudinal/time-varying models. For each of the analytical techniques reviewed, we direct readers to helpful resources elsewhere where they may continue learning about such techniques.