Generating High-Quality Verifiable Relational Data About News Coverage of Social Movement Events

社会运动反映了社会对关键问题的关切，是跟踪社会变化的重要工具。然而，围绕特定社会运动的活动有起有落。重要的是要知道，当社会运动处于较低动员期或休止期时会发生什么。该项目开发了一种处理少数族裔报纸文章的电子档案的更好方法，以分类关于1994年至2016年社会运动事件的故事，并使用它收集有关这些事件的数据。这项研究填补了一个重要的知识空白，因为在这段时间内，没有关于少数群体运动事件的系统信息。创建关于这一时期事件的系统数据集不仅有助于我们更好地了解以前聚会的长期影响，而且还有助于我们更好地了解最近发生事件的背景。这些报纸数据补充了已经从主流新闻报道中收集的现有数据。这些事件本身对于了解活动人士的所作所为很重要，而有关这些事件的新闻报道对于了解新闻媒体如何描述它们以及公众对它们的了解也很重要。我们知道，少数事件得到了大量的新闻报道，对公众的认知产生了很大的影响，而大多数事件只被提到一次。比较主流新闻和小众报纸的报道，有助于我们了解不同群体对事件的不同看法，也有助于我们了解被主流新闻忽视的事件。该项目的结果将使领导人和公民都能更好地了解社会运动轨迹，从而加强民主国家的安全和安保以及政治影响力。社会运动的动态会随着时间的推移而变化，在最初的焦点重新出现之前，焦点往往会从一个话题转移到另一个话题。这些变化很难发现和核实，特别是使用标准的新闻媒体来源。该项目使用以前开发的程序来预处理报纸文章的电子副本，这些文章已被选择为包含一个或多个与社会运动相关的关键字。它创建了新的人机界面，并使用关系数据库结构，以确定事件之间的关系和有关事件的文章的方式，加快从新闻来源收集和存储事件数据的速度。之前开发的界面将进行更新，以使这项工作更有效率。然后，该项目将开发新的编码接口。人们将在文章之间匹配事件；为每个事件分配唯一的标识符；识别和标记事件之间的关系，包括活动、剧集以及主事件和子事件关系；以及检查编码员在标记事件文本方面的工作。新的界面将链接到一个关系数据库，使编码员能够访问事件、地点、活动、剧集和事件间关系的授权列表以及原始全文文章。在第二编码活动中，使用适合于每个变量(例如，大小、问题、形式、参与者类型、警察行动)的自动化例程和接口，一次一个地对变量进行编码。编程例程将使用关键字搜索和词典对变量进行粗略编码；人工编码员将使用特定于变量的接口做出最终的编码判断。这些例程和界面将用于处理从22家报纸的121,129篇报道中挑选出来的7619篇文章，这些文章是2017年2月从报纸档案中检索到的，使用的是我们的标准事件相关词汇集。在以前的资助下编码的新闻通讯社中确定的事件数据库将可供访问，以便为关于相同事件的新编码的报纸文章赋予相同的唯一识别符。对数据的分析将结合新闻通讯社和少数族裔报纸来源的事件，以更全面地了解事件，并比较这两个来源对运动的描述。该项目将使用开源工具，并将项目源代码存放在公开可用的档案中，从而允许研究人员通过报纸文章研究其他主题，以利用新开发的工具。项目发现将为有关社会运动和社会变化的社会学理论提供信息，特别是关于运动动态如何在较长时间内起伏的社会学理论。更广泛地说，这些发现将为民主参与的理论提供参考，重点是在运动特征的演变和变化中的地位差异。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Social movements reflect societal concerns with key issues and are important vehicles for tracking social change. Yet activity surrounding specific social movements ebbs and flows. It is important to know what happens when social movements are in the periods of lower mobilization, or abeyance. This project develops a better way of processing electronic archives of articles from minority newspapers to classify stories about social movement events and uses it to collect data on them between 1994 and 2016. This research fills an important knowledge gap because there is no systematic information about minority movement events during this interval. Creating a systematic dataset on events in this period helps us to better understand not only long-term impacts of prior gatherings, but also the contexts from which recent events emerged. These newspaper data supplement existing data already collected from mainstream news wire stories. The events themselves are important for understanding what activists were doing, while the news stories about the events are important for understanding how news media portrayed them and, thus, what the public learned about them. We know that a small number of events receive a great deal of news coverage and have a large effect on public perceptions, while most events are mentioned only once. Comparing mainstream news coverage with that in minority newspapers helps us understand how different groups perceive events differently as well as allowing us to learn about events that are neglected by the mainstream news. Findings from the project will allow both leaders and citizens to better understand social movement trajectories, thus enhancing safety and security as well as political influence in democracies. The dynamics of social movements change over time, with focus often shifting from one topic to another before the original focus reappears. These changes are difficult to detect and verify, particularly using standard news media sources. This project uses a previously develop program to pre-process electronic copies of newspaper articles that have been selected for containing one or more social movement-relevant keywords. It creates new human-machine interfaces and uses relational database structures to expedite the collection and storage of event data from news sources in a way that identifies relations among events and articles about events. A previously-developed interface will be updated to make this work even more efficiently. The project will then develop new coding interfaces. One will match up events between articles; assign each event a unique identifier; recognize and flag relations between events including campaigns, episodes, and master- and sub-event relations; and check the work of coders in marking text in events. The new interface will be linked to a relational database to give the coder access to an authorized list of events, locations, campaigns, episodes, and inter-event relations as well as access to the original full-text articles. In a second coding activity, variables are coded one at a time using automated routines and interfaces appropriate for each variable (e.g. size, issue, form, actor type, police actions). Programming routines will rough-code variables using keyword searches and dictionaries; human coders will make the final coding judgments using variable-specific interfaces. These routines and interfaces will be used to process 7619 articles selected from 121,129 stories in 22 newspapers that were retrieved in February of 2017 from a newspaper archive using our standard set of broad event-relevant words. A database of events identified in the news wires coded under previous funding will be accessible so that newly coded newspaper articles about the same events can be given the same unique identifier. Analysis of the data will combine events from both news wires and minority newspaper sources to give a more complete picture of events and to compare the portrayal of the movement in the two sources. The project will use open source tools and deposit project source code in publicly available archives, thus allowing researchers studying other topics via newspaper articles to make use of newly-developed tools. Project findings will inform sociological theory regarding social movements and social change, particularly regarding how movement dynamics ebb and flow over extended periods of time. More generally, the findings will inform theories of democratic participation, with a focus on status differences in the evolution and change in movement characteristics.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.