A lot of people have flagged Benedict Carey’s piece in yesterday’s Times, “In Battle, Hunches Prove to Be Valuable,” and with good reason: it’s a fascinating report on research into the way the brain combines visual data and emotional responses to shape the sort of instant-gut-reaction decisions that soldiers make as they evaluate threats.
The examples the piece draws on are from U.S. soldiers’ experiences in Iraq, where every stray boulder or trashheap by the roadside could be hiding a deadly bomb.
Reading Carey’s story, I thought of parallels in the distinctly less lethal — but still occasionally perilous — informational environment of the Web. What are the little signals that tell us, “You can trust this page”? And what are the red flags that tell us, “Watch out, something’s off here”?
These are important. Of course, they can help us protect ourselves from outright scammers (phishers who build lookalike bank websites to try to steal your passwords, and so on). But they can also help us sift and sort through the news and information that flows through our browsers, focusing on the good and discarding the bad.
Some of these signals are glaringly obvious (no “About” page? come on!). Others are subtler (are the writer’s arguments logical? Are statements of fact documented by links?).
What are some of the tools you use? I’ll be teaching a workshop this coming weekend as part of the Stanford Professional Publishing Course, and would love to hear your suggestions.
Post Revisions:
This post has not been revised since publication.
That’s a great article.
I have a similar conversation ever time I’m trying to teach some students how to evaluate the popularity of blogs and websites. There are the imperfect empirical measures (PageRank, Technorati, incoming links, Compete, Alexa and so forth), but I really emphasize the anecdotal stuff. These are marketers, mostly, but ‘popular’ and ‘authoritative and trustworthy’ usually go hand in hand:
* Number of comments
* Design aesthetic–is it a default template or have they got a customer header et al?
* Do they have their own domain?
* How much social media stuff is in their sidebar?
* Is it blanketed with cheesy ads?
I mistrust sites that have no search–until I spend enough time there to discover that there’s not really enough to need a search function. I mistrust blogs that have no comments and Twitter feeds that have no followers. I’m being followed by someone who has never tweeted, who follows over 200 people but no one follows him/her. I feel a little unsafe, like I’m looking over my shoulder when I tweet. I mistrust sites that have old data–of their own or someone else’s. “A 2005 study shows…” that sort of thing. That may be laziness rather than a desire to fool me but it still undercuts credibility. Ugly sites? I just feel sorry for them but there’s so many of them out there.
For websites and blogs, the first indicator is spelling and grammar. Although these problems are not limited to sites that aren’t trustworthy, they do raise an immediate flag.
For Twitter, the number (and list) of followers is a primary indicator. As Rebecca said, if no one is following the user, there is probably a reason why.
For the rest, well, it’s a matter of judgment. Anyone can purchase a domain, and blog/website templates are easily accessed and developed. Assessment requires intellectual curiosity: Why is this site in existence? What message is it sending? What does the site owner want?
I look at where the links point, and in particular, any link that I’m asked directly to click on. If the links point to an unexpected domain, or if they are obfuscated via javascript, then I get cautious.
Spelling and grammar are helpful, but only if you are able to filter them to separate garbage from the ordinary usages of non-native speakers.
I flag broken links: one is okay, more is not. I tend to report them too; sites that respond to that are real — those that don’t are junk.
Search is more a measure of how sophisticated the webmaster is, to my mind. I like Contact Us — if all they offer is email, then I am suspicious. A phone number doesn’t prove anything (especially if I don’t dial it); but lack of phone or physical delivery address is serious.
Finally – if there is any good chance this page was a mistake (typo, domain confusion, etc) it should have a clear disambiguation spot prominently displayed. Lacking this is a clear sign that I don’t want to deal with these folks. If XYZ.COM doesn’t let you know that XYZ.GOV is a different group, there is something wrong.
Another thing I look for is what the links are. E.g., today I ran across a site that was supposedly about professional game development, but there were -many- links to gambling sites. And the information was scanty, to say the least. Google-bait, I assume.
Many scam sites still have the garish web design of the very early web: animated logos, flashing text, glaring white background, cheesy script or cartoon fonts, etc. A really creepy example (actually there are a least 3 sites with the same business model) is the “discount on your next purchase” button which auto-subscribes you to a monthly paid subscription discount service when you think you are getting a one-time rebate after an individual purchase on a legitimate site.
I was saved from this scam because when you click the button, you get to the scam site (without warning; you might think you are still on the site where you made the purchase) which has garish web design and is obviously not trustworthy. See: http://sporkworld.tumblr.com/post/169517800/beware-of-discounts-offered-after-an-online-purchase As I said there, this won’t go away unless people boycott the “legit” sites that refer to the scam site.