学术活动 > 研讨 Academic Discussions

“政经下午茶”第六场

2017-3-8 22:01:50 点击:



“政经下午茶”是2016年5月开始兴办的一系列研讨活动。这一活动的主要形式是“主题汇报+开放式讨论

 

主题汇报部分主要由约定的青年学者、在读博士硕士研究生或从事学术写作的其他同学来系统性介绍其近期从事的一项研究,一般时长20-30分钟。

开放式讨论主要由参加下午茶的成员来推动。我们不设门槛,只设主题,希望参加人员能够围绕主题完全开放地展开学术讨论。

 

参加要求:(1)这个活动不设旁听席,我们希望每一位参与者都能够踊跃发言。(2)自带茶杯。

 

如果你想成为“政经下午茶”的汇报人,请联系封凯栋 k.feng@pku.edu.cn

如果你想赞助茶点或其他经费,或者提前了解下一轮活动的安排,请联系姜子莹13120192069@163.com

 


时间地点:2017年03月11日14:00,廖凯原楼436

主题汇报人:李诗涵

汇报主题:文本分析及定量转化:QDA Miner入门操作及应用

 

背景介绍


As one of the core methodologies in social science, textual analysis provides researchers a way to gather information about how members of various cultures make sense of who they are and how they fit into the world in which they live. Traditional textual analysis focuses on interpreting texts (films, TV programs, press releases, speeches, interviews, and so on) based on its qualitative features, such as its tone (positive, negative, neutral), political context, subjects and objects, and transferring linguistic symbols into quantitative elements. Yet in most cases text mining, quantitative content or statistical analytic techniques cannot be conducted concurrently, due to boundaries in mixed methods, lack of joint quantitative/qualitative displays and unstandardized dictionaries.


To step over these obstacles, QDA Miner is developed to offer unparalleled options for mixed methods analyses, opening up doors for handling heterogeneous data in native formats. It’s a powerful tool for conducting robust code-based analysis, allowing researchers to visualize coding patterns and trends, explore relationships in coding applications, and test statistical hypothesis. Its text-based search tools providing coding assistance include keyword in context searching, section retrieval, query by example searching, and a cluster extraction and coding tool. And reciprocal learning, assessing and suggesting process between user and the tool contributes to ensuring reliability and consistency of coding work. As to the statistical analysis part, QDA Miner allows researchers to identify coding occurrences (cross-table), assessing relationships between coding and numerical or categorical properties (frequency), and displaying clustering, proximity, similarity, multidimensional scaling, heatmaps, correspondence analysis and sequence analysis. Thus, researchers undertaking mixed methods seeking to integrate qualitative and quantitative analytic strategies in complex ways can pick and choose from an array of tools in QDA Miner, to suit the particular needs of the research design.


作为目前社会科学领域的一个重要的研究方法,文本分析手段拓展了研究者们获取信息并开展分析的途径。传统的文本分析方法侧重于定性分析,强调对影片、电视节目、出版物、演说、访谈等材料中类似于语气、政治情景、主题、关注事项等信息的量化。但是在以往的体系中,文本挖掘、内容分析以及统计分析之间并不能很好地联立开展,因为各种方法之间的边界明显,缺乏通用的方法和术语。

QDA Miner作为一个大型的文本分析软件工具,正是为了克服障碍,融合不同分析方法而被发展起来的。QDA Miner为多种研究方法的贯联提供了卓越的空间,从而为处理分析异质性的信息打开了大门。它支持研究者开展文本编码、相关的可视化呈现及趋势分析、关系分析以及相关的统计检验等工作;在文本编码、关键词检索、机器检索、范例检索以及族簇输出等工作上为研究者提供了辅助工具。软件的交互式学习、评估以及智能建议等功能也提高了研究者应用软件的可依赖性和一致性。在统计分析方面,QDA Miner允许研究者们界定编码出现的频次,对编码关键词在族簇、相近性、相似性、排列关系、热图、一致性和序列等维度上提供分析功能。这样,研究者可以很好地把定性研究的量化手段和定量分析工具结合起来,满足研究者相对复杂的研究设计需求。