The halo effect is a well documented social-psychology phenomenon that causes people to be biased in their judgments by transferring their feelings about one attribute of something to other, unrelated, attributes.

For example, a tall or good-looking person will be perceived as being intelligent and trustworthy, even though there is no logical reason to believe that height or looks correlate with smarts and honesty.

The term "halo effect" (a.k.a. halo error) was first introduced into psychological-research circles in 1920 in a paper authored by Edward Thorndike titled “A Constant Error in Psychological Ratings.” Through empirical research, Thorndike found that when people were asked to assess others based on a series of traits, a negative perception of any one trait would drag down all the other trait scores.

The halo effect works both in both positive and negative directions:

  • If you like one aspect of something, you'll have a positive predisposition toward everything about it.
  • If you dislike one aspect of something, you'll have a negative predisposition toward everything about it.

A negative halo effect is sometimes called the "devil effect" or the "pitchfork effect," but that seems to be taking the metaphor too far. We recommend using the term "halo effect" for both positive and negative biases.

In This Article:

Why Is It Called "Halo?"

The term "halo" is used in analogy with the religious concept: a glowing circle that can be seen floating above the heads of saints in countless medieval and Renaissance paintings. The saint's face seems bathed in heavenly light from his or her halo. Thus, by seeing that somebody was painted with a halo, you can tell that this must have been a good and worthy person. In other words, you're transferring your judgment from one easily observed characteristic of the person (painted with a halo) to a judgment of that person's character.

Thus, the name has nothing to do with the video game Halo :-)

Why Does the Halo Effect Exist?

The halo effect allows us to make snap judgments, because we only have to consider one aspect of a person or design in order to "know" about all other aspects.

In the age of the cave people, there might even have been some truth to these snap judgments: to grow tall a person would have had to eat lots of meat and was therefore probably a good hunter that was worth listening to. And a good-looking person would have avoided disfigurement from lost battles, animal bites, and nasty diseases, which again would make them role models for other protohumans.

Those early humans who could make fast decisions were more likely to survive to become our ancestors than anybody who had to ponder all problems for hours. Thus, we have inherited a tendency to make (overly) fast judgments based on generalizing from a small amount of data.

Websites are Impacted by the Halo Effect

The halo effect can impact organizations, locations, products and delivery/communications channels, as well as our judgments of other people. If users like one aspect of a website, they're more likely to judge it favorably in the future. Conversely, if users have a particularly bad experience with a site, they'll predict that the site will treat them poorly in the future as well and, thus, will be reluctant to return to the site. In this latter case, even if the site is later redesigned to be better, users will still carry over their negative expectations from their earlier experience.

An example we often see is that the quality of a website's internal-search results are used to judge the overall quality of the site, and, by inference, the quality of the brand behind the site and its products. Thus, a user's statements may proceed as follows if verbalized in a thinking-aloud study: "Wow, these search results make no sense and appear in seemingly random order. This site must be really poorly done. This company doesn't have its act together and doesn't care about customers. I shouldn't buy any of these products." Note that each step in this chain of inferences is at least a little bit reasonable, and yet the final conclusion doesn't follow from the initial observation. (Sometimes you'll get a good product when buying from a site with poorly implemented search.) However, users don't really progress through a logical-reasoning process. The halo effect works by shortcutting all these steps and simply allows people to make their overall judgment based on their impression of one attribute.

SImilarly, if it's horribly complicated to set up an account for a service then that bad user experience will rub off on people's expectations for the rest of the service.

A 2002 research study (Lindgaard and Dudek) asked users how they would rate the visual appeal of a group of websites. The websites that had high visual-appeal ratings were then tested for usability. On average, participants' task-failure rate on these sites was over 50%; this is an unacceptable failure rate based on our tracking of failure rates since 2000. However, despite this atrocious failure rate, participants' satisfaction ratings remained high. In this case, research indicated that the look and feel of the site had a halo effect on the entire site experience, even when these sites were poorly designed for usability.

In many instances, the trait or characteristic that a person will use to assess the whole isn’t even what best answers a particular question; this is the basis of judgmental heuristics and cognitive biases. An example: You ask someone to tell if the site is easy to use and they say, “Yes, it’s beautiful.” Just because it’s beautiful, doesn’t mean it’s easy to use. But judging beauty is often far simpler than judging ease of use. This is why task-based analysis and triangulation of data sources is so important in user-experience work.

Conclusion

It’s important to keep the halo effect in mind as you are planning sites, designing flows, defining key performance indicators (KPIs), and measuring your site because dropoffs at any one point in your users’ experience may indicate a poor first impression via design, content, site performance, and so on. Additionally, it’s important to supplement quantitative data sources with qualitative methods such as usability testing.

References: