Get the latest articles and downloads sent to your inbox in a monthly newsletter.

Get the latest articles and downloads sent to your inbox in a monthly newsletter.

Understanding Information Trustworthiness in the Networked Information Ecosystem

Clifford A. Lynch, Coalition for Networked Information | Jan 01, 2010

As readers are confronted with an evergrowing and ever more overwhelming set of content offerings, accessible through an ever-multiplying set of channels and services, one hears a great deal of concern about the ability to identify “trustworthy” information. It’s clear that a key critical skill for the 21st century is the ability to assess the “trustworthiness” of information. There is a great deal of discussion about the shifting roles of various players — notably, libraries and publishers — in establishing trustworthiness in the information ecosystem, and of new mechanisms such as social networking and filtering systems that can help in identifying and assessing information.

These concerns are not new. But the volume of information has grown enormously, and the number of contexts in which it can be provided has expanded more rapidly and more extensively than the volume itself. Further, the properties of the Internet and of discovery tools in common use such as search engines mean that information can be quickly and easily stripped of one context and placed into another.

Defining “trustworthiness”

The term “trustworthiness” itself is complex, and conflates a number of related properties. Achieving a consensus on these properties, and trying to examine themmore specifically, would be a major step forward; this brief piece is an attempt to do this, though I cannot say that I am entirely satisfied with my characterizations. Progress here would point towards an agenda of metadata standardization, and of discussions about how different participants within the overall ecosystem could better help readers to evaluate information resources available to them.

The role of infrastructure

One aspect of trustworthiness is really about stability and transparency; it’s primarily a property of the infrastructure that publishers and libraries have worked together to establish over the centuries, and one in which roles are now changing and where new players (such as subject repositories) are defining new roles. Here, I think that the key characteristics are the ability to cite or reference an information object persistently and reliably across time; the assurance that responsible provision has been made for the preservation of the information object and its access across time; and the assurance that the integrity of the information object will be managed in some transparent and formal way.

Note that integrity does not necessarily mean that an information object will never change, but rather that changes are controlled and documented (for example, through versioning, where versions are clearly identified, preserved and linked one to the other, or through change logs or errata). Other than through empirical experience, it’s hard for a reader to tell if a specific delivery or access context is providing the desired levels of stability and transparency; these are properties that need to be evaluated over long periods of time, and often involve behind the scenes cooperation between multiple parties.

The role of metadata

There’s one additional aspect to infrastructural stability and transparency. Certain metadata should be attached to information objects, and readers want this metadata to be available and correct. Notably, this would include the date an object first entered the infrastructure (such as a submission date), and the dates of subsequent events (e.g., public accessibility, revisions). Such metadata would include the identity of the author or authors of the object; this is actually a highly complex matter because of the possible use of strong authentication, name authority control and related matters. (Who serves as guarantor of author identity, and why should we trust this guarantor?)

And perhaps metadata should include some indication of the nature of the vetting or review processes, if any, that the information object has been subjected to; this might include one of various forms of peer review, editorial assessment, fact checking by a publisher or endorsement by one or more parties. Of course the reader may also want to know who takes responsibility for the rigor and neutrality of a peer-review process, or who takes responsibility for the editorial evaluation of a work. It’s interesting to note that virtually all of this metadata is established as a by-product of article publication in a traditional journal, though often much of the metadata is implicit in the sense that it’s part of the journal’s policies.

Correctness and importance

Whether a given information object is credible, that is, whether the statements and conclusions it makes are likely to be true, is a complex judgment that must rest with the reader. But having the metadata just discussed, and being able to make a separate judgment about confidence that this metadata is accurate, is a significant part of making the credibility judgment. Sometimes it may be enough: A reader may be prepared to put a very high level of faith in a specific author, or in the rigor of a specific review process. Other times, the metadata will be just one factor in judging credibility, along with the identification of corroborating or contradictory materials in other information objects, or other assessments such as reviews or ratings, of the information object in question.

Finally, I want to strongly distinguish between ideas like trustworthiness and credibility on one hand, and attention-worthiness on the other. In many situations, as we attempt to deal with the overwhelming amounts of information available, we are really more concerned with attention-worthiness or importance.

Trustworthiness and testimonies of credibility may or may not be prerequisites for attention-worthiness, depending on the context. We may be interested in rumors, in unsubstantiated reports; we may want to know about controversial results while their correctness is still under debate, and perhaps even to take part in these debates. Readers may be prepared to make their own judgments about attention-worthiness based on authorship or topic of a work, or they may rely on the judgment of others; these judgments may be expressed in the direct personal recommendations of colleagues, the collective recommendations of a social filtering system or the public recommendations of editors or reviewers.

Attention-worthiness, like credibility, is ultimately a personal and subjective judgment, though certainly one around which a community may ultimately achieve a consensus.

Comments