计算机科学
概念化
推论
原始数据
定性性质
数据科学
口译(哲学)
分辨率(逻辑)
多元方法论
定性研究
数据挖掘
大数据
管理科学
人工智能
机器学习
数学
程序设计语言
社会科学
数学教育
社会学
经济
作者
Alex Gillespie,Vlad Petre Glăveanu,Constance de Saint Laurent,Tania Zittoun,Marcos José Bernal Marcos
标识
DOI:10.1177/15586898241284696
摘要
A recent challenge is how to mix qualitative interpretation with computational techniques to analyze big qualitative data. To this end, we propose “multi-resolution design” for mixed method analysis of the same data: qualitative analysis zooms-in to provide in-depth contextual insight and quantitative analysis zooms-out to provide measures, associations, and statistical models. The raw qualitative data is transformed between excerpts, counts, and measures; with each having unique gains and losses. Multi-resolution designs entail transforming the data back-and-forth between these data types, recursively quantitizing and qualitizing the data. Two empirical studies illustrate how multi-resolution design can support abductive inference and increase validity. This contributes to mixed methods literature a conceptualization of how mixed analysis of the same big qualitative dataset can create tightly integrated synergies.
科研通智能强力驱动
Strongly Powered by AbleSci AI