Integrated Data for Events Analysis

***This site requires Microsoft Internet Explorer version 5 or above to view properly***

Background

The IDEA event form typology is a conceptual framework for use in coding social, economic and political events data. The IDEA framework is an extension and a refinement of, and is congruent with the World Event / Interaction Survey or WEIS. Like WEIS, IDEA is nominally scaled, but unlike WEIS the event forms in IDEA are not bound to state actors, though some event forms are intrinsically bound to specific actors like armed forces when they engage in a battle. To be clear, the WEIS event form, reduction in relations, represents a diplomatic behavior and therefore represents diplomatic action, but the IDEA equivalent, reduce routine activity, refers to such reductions by individuals, groups or organizations, both state and non-state.

The congruence with WEIS is embodied in what are referred to as WEIS "cue" categories, the 2-digit numeric codes numbered 1-22. Aside from some minor label changes, these 22 event form categories remain essentially the same (but free of the actor bound definitions) to WEIS; and like WEIS, we generally consider these cue categories to be residuals, to be coded only if the lower level distinctions cannot be made.

The extension of WEIS is represented by 2-digit residual categories beginning with number 23, and extending (with breaks) to 99. The first additions made were the PANDA events that embodied the contentious and coercive but not yet violent aspects of conflict. PANDA was superseded by IDEA in its first version. Next, the event forms from Charles Taylor's World Handbook of Social and Political Indicators were added, administrative adjustments, for example. In addition, we added categories deemed important in that they represented distinctive sets of behavior (for example, voting and judicial behaviors), and these are true residuals in that we have created them primarily as placeholders until we can engage scholars in these areas to help us flesh out the sub-forms. We also added in the previous version of IDEA a handful of economic behaviors (with very crude lower-form distinctions, again as placeholders). Likewise, we added a number of detailed event forms for mass protest behaviors, guided largely by the work of Craig Jenkins. Finally, we added a series of non-human phenomena (like natural disasters and accidents) and (animal) behaviors, used primarily to track biomedical phenomena such as the benign or malicious spread of toxins or disease.

The current (July 2002) version of IDEA adjusts a number of operational definitions, merging a few events and adding numerous others, based on our experience over the past year. This July 2002 version also explicitly integrates virtually all of the CAMEO event forms as published the January 2002 release available from the Kansas Events Data System Project website and updated slightly in the March 2002 ISA paper by Deborah Gerner and others. Thus cooperative actions, particularly those involving mediation, are much better covered in this version. With this version too, we have also made available the IDEA equivalents for each of the following event data frameworks: CAMEO, MID, WEIS and World Handbook.

IDEA event forms represent the lowest level of specificity that current (circa 2002) machine coding can accomplish with precision at least equal to large-scale human coding exercises. Our goal in this collaborative effort is to offer a coherent framework with the lowest level or "common denominator" event forms used in the major nominal frameworks cited above. Equivalent events for the major ordinally-scaled alternative framework, COPDAB, can be generated by using a weight applied to each nominal event form similar to what Joshua Goldstein did in his 1992 JPR article.   framework. Thus by generating events using the IDEA framework, one can map the coding back to any of the major event data frameworks still in use.

At the very least, we hope this approach facilitates peer review and independent assessment of alternative events data sets. Such an approach stands in contrast to the present situation in which review and testing is not possible across the numerous, competing efforts being conducted today, including proprietary schemes (when done in the commercial sector), secret (when done in some agencies of governments) and even across the open frameworks used by many in the academy, in part because they are not cross-mapped, and in part because human coding does not lend itself well to replication by others. 

The July 2002 version is the third major revision since IDEA was first outlined in 1998. The framework below is posted as an ongoing draft, with comments, suggestions and extensions welcome. Thanks in advance. Please e-mail questions or comments to Doug Bond.

Coding Instructions

IDEA Framework Typologies

Relationship Between IDEA and Other Frameworks

Copyright 1998-2010 IDEA Project. All rights reserved.
Last Revised: April 1, 2010