Can searching be hazardous to your health?
A recent article by Darren Osborne in www.news.com.au reported that, “Australian research has found that people searching the internet only take in what they want to read, potentially placing their health at risk.”
Here are some excerpts from the article:
“The research, conducted at the University of NSW (UNSW), found that people are inclined to maintain their beliefs, even if they find contrary information.”
“‘Our research shows that, even if search engines do find the ‘right’ information, people may still draw the wrong conclusions - in other words, their conclusions are biased’, UNSW Professor Enrico Coiera, said.”
“The results…also revealed that where the information appears in search results, and how much time a person spends looking at it, affects how people perceive it. The first or the last document the user sees has a much greater impact on their decisions,” Professor Coiera said.
“We know that the web is increasingly being used by people to help them make healthcare decisions…There can be negative consequences if people find the wrong information, especially as people in some countries can now self-medicate by ordering drugs online.”
So we asked two of our Alternative Search Engine experts for their responses:
“The research by the University of NSW raises some very good points about online health research. People searching for health information are often in a concerned or distressed framed of mind, so it is more critical than in any other situation that the results are accurate. Today’s search engines are not geared towards the specific needs of consumers requiring medical or health information. They rely on the user entering the correct medical term, which frequently is not the case. Results are often too broad and not filtered to ensure only relevant, reliable content from accredited sources is served.”
“At iMedix, we believe the Internet has tremendous potential to deliver the kind of patient-to-patient interaction that consumers need to improve on search results. That is why we have developed an evolutionary search technology that encompasses sophisticated search algorithms and social recommendations to dramatically increase the relevance and reliability of online search results. We have combined this search engine with a community platform that enables patients to share experiences with others.”
“Because iMedix combines search and community, it offers users the unique benefit of being able to identify others with shared interests and concerns and communicate with them in a single click. Each search result includes both relevant Web pages and relevant user profiles based on their stated interests. We believe this powerful combination will help empower consumers to make better health decisions using the Internet.”
Dr. Michael R. Alvers responds: There is an experiment that I call the “2-4-6 experiment.” It was first published by D. C. Wason in 1960 in /The Quarterly Journal of Experimental Psychology/ and was called “On the Failure to Eliminate Hypotheses in a Conceptual Task.” The experiment goes like this: an experiment conductor gives people the sequence of numbers 2-4-6 and tells participants that this is a sample sequence for which they have to find the rule. Participants may uncover the rule by offering other sequences of three numbers while the experiment conductor confirms or rejects the sample according to whether or not it fits the rule behind the sequence. The ultimate goal is to figure out the rule. Participants are asked to share the rule when they think they have found it — then and only then.
Most people start with 8-10-12. And the answer here is yes; it complies with the rule. Try 12-14-16. Correct again! Already, after only two confirmations, most people think they have deduced the rule: the distance from one number to the next is two. But they’re wrong! This is not the rule. More creative people (mostly non-academics) come up with more examples like 10-15-20 or 11-22-33 or 1‑100-1000, some even with 1-5-6 or 22-46-1005, all of which are correct. Because the rule is very simple: the numbers go in ascending order.
Observation number one: people try always to confirm their hypotheses and never try to intentionally falsify them. People do not, for example, change the order of the given examples from ascending to descending. In my experience only a few kids do this.
This is what Professor Enrico Coiera from UNSW found. Even scientific publications are created this way: by proving the hypothesis by hook or by crook. People believe in simple patterns. If a pattern is observed it is somehow stored in our brain. If it is seen a second time, the brain becomes alert: rule confirmed. If it is seen a third time, the rule is made.
So here is my answer to this dilemma: the first and most important point is the quality of the information used in making judgments. GoPubMed is based on 18 million peer reviewed scientific publications. We can thereby be sure that this information is in all conscience the best available today. To foster an objective selection process, our method of filtering or ranking done by the user (left tree) rather than by non-transparent algorithms, is a very strong asset for an improved search. Combined with the reflection of the search space (how many articles are found per category and “Hot Topics”) it helps remove bias in favor of ones own hypothesis. All our activities are aligned with our vision “towards answering questions.” Simulating dialog with an “independent” expert is the goal.
Editor’s note: An AltSearchEngines’ reader contributes:
Hi, Charles,
I was very interested to read your post about online health search. I just wrote (at http://lisaneal.com) about the Pew study looking at health search in the US. There are many concerns about health and information literacy, as you find as well, since far too many people searching for information are desperate for information and support, yet lack the skills or the awareness of the need to discern the quality and source of what they find online.
Traditional cues may be quite different online, and the need for help can overpower even common sense at times. Some of the sites I’ve seen for herbal supplements are the worst. I teach a course on Online Health Communities at Tufts University School of Medicine and would be happy to engage in further dialog about these issues.
Lisa Neal, Ph.D
January 19th, 2008 at 5:58 pm
Hi, I was very interested to read your post about search. I just wrote (at http://lisaneal.com) about the Pew study looking at health search in the US. There are many concerns about health and information literacy, as you find as well.