Avoiding Informatics Overload

Peter Jones Design for Care, Information Ecology, Sensemaking

Mark Hurst posts on Good Experience the argument that information overload suppresses comprehension and creates an absence of understanding and retention: To solve info overload, make friends with The Nothing

In my experience this is true, and is moreover a testable proposition.  Mark says:

Because the only way to really make information disappear, these days, is to surround it by a sufficient amount of competing information.

Case in point is Side Effects? These Drugs Have a Few. Here the NYT references a Harvard study showing that there are, on average, 70 side effects listed on drug labels. Some labels contain over 500 side effects.

What would be the possible benefit of a drug maker listing over 500 side effects? Easy: it gives coverage in a liability lawsuit.

What’s clear is that the patient experience is harmed by these labels. Patients now know less about side effects than they did before. Sure, a drug might hypothetically bring about any of 500 side effects, but what are the few most common ones to look out for?

The drug label story teaches us that we have to change our perspective in the digital world. In a world full of information, the villain isn’t The Nothing – it’s actually The Everything.
Put another way, too many competing inputs are the same as not using the inputs at all:

In the digital world, information will find us. It’s inescapable, and if we’re not careful, The Everything will arrive and paralyze us. So the challenge is to find The Nothing, and make friends with it, to solve overload permanently. Let the bits go.

Electronic Medical Records (EMR) systems present too much information with too little context, overwhelming the attention and time of clinicians. Clinical Decision Support systems (for the most part) have poorly designed search and information navigation, requiring extensive parsing and disambiguation of similar topics and references. Online scientific research, too many papers, too little context.

My guess is that a power law function guides this relationship – similar to short term memory (7 +/2) there is a small number of objects that can be invariably retained as important, and each additional object adds load linearly, until the entire set of objects becomes useless.  Designers often limit the number of search hits to 10 per page in hopes that their relevance is sufficient to deliver a satisficing result in that priority view. I disagree with that practice, because good content and indexing is all. Since people naturally chunk information objects, we can repair context by effective clustering and visual support. But yes, as discrete information objects grow beyond (5-7), individual attention significantly degrades until the information set is truncated (cutting off everything after the first 5) or discarded (using a new resource). (These are presented as testable propositions, but are supported by my observations in UX and field research and enduring cognitive science studies).