The 90% Doxa

It has recently and (too?) frequently come to my attention that “90% of information is conveyed by titles” of peer-reviewed articles such as those found in bibliographic databases, say Scopus-like databases. Now I always am wary of rash judgments like these for several reasons (please note that I provide a bibliography as a separate attachment):

  1. The assertion fails to substantiate explicitly its underpinnings (i.e. where does it derive its own information from?) I have found nothing in the literature that could back this up (Mizzaro)
  2. Neither does it articulate the view it is, I believe, supposed to reflect; namely that 90% of what is stated in the full text of an article may be expressed through a title

Concerning 1., I’d be glad to have any of you point me to the right, reliable source that conclusively establishes that the information content, as it is called in the literature (Bary,149), can be worked out at 90%.

The assertion (2.) also begs the question of why it would matter to have as much information crammed into titles. Information content of document representations such as titles, abstracts and indexing terms is geared to letting readers know whether represented full-text articles are relevant enough that they justify perusal. To borrow Bary’s phrase (1293), document representations may be assessed in terms of how well they perform as clues to relevancy of full texts.

That issue begs another question: what is meant by relevancy? Relevance? Pertinence?

If, at the end of the day, the relevancy of articles is not meant or glanced at, then I fail to see what could be here. Surely, the assertion doesn’t refer to the amount of substantives used in titles as opposed to stop words (Tocatlian, 346) as that would not prove that 90% of the information contained in the full text represented by a title represents 90% of it.

Now there’s also the dimension of the effectiveness of document representations, when it comes to carrying information (see Byrne). But measuring effectiveness doesn’t equal to making approximately all of the content of a full text accessible through its title.

A review of the literature I could find on the topic indicates that abstracts rather than titles fare highly in readers’ estimate, when it comes to the relative importance (Barry, 1295)

Or could it be that by reading the title of an article one knows 90% of there’s to learn from a full-text article? That seems an unreasonable claim to make on the back of my own experience making a living looking for the right info.

Just a quick parting shot: Would you say that “The Devil in the Dark Chocolate,” by an unknown author and retrieved from PubMed [PMID: 18156011], is a good key to its aboutness, considering no abstract is available for it?


~ by iinformationvoyager on February 28, 2010.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: