My friend John is an interaction designer for a non-profit organization. His job is to manage their website. The purpose of the website is to help users learn about various charitable organizations, enabling them to choose which ones they want to donate to. The site has grown over time and now offers information about many different types of charities.

Designing categories can be tricky business. Make sure to understand fully how the different possibilities you categorize information might affect your users usability.

John feels it’s important to categorize the charities so users can easily browse what types of charities are available, as well as to more easily choose those they’re most interested in. But he wonders about the best way to do this, since they could be grouped in various ways. For example, the charities could be grouped according to size, type, purpose, geographic location, etc. What method of categorization might be best and most meaningful to those who visit the site?

Because he is a conscientious designer, he is concerned about how different ways of categorizing might influence usability. But what might not occur to him is that the method of categorization might have other implications, too—such as which charities users actually choose.

For example, let’s say a philanthropist wants to donate to an array of children’s charities. If these are grouped by domestic versus international, she might select an equal number of charities from both groups. Alternatively, if the charities are organized into three groupings—local, national, and international—might she choose differently? To what extent might the design of how the charities are organized or grouped influence how she decides, as well as how much money she allocates to each of them?

In this article, we’ll look at how the categorization and grouping of items can influence and drive decision outcomes, sometimes in very significant ways.

Categorization of ranges

In a study designed to understand how the actual categorization of items might influence or impact decision outcomes, researchers asked participants to determine how they would allocate financial aid money according to the income levels of the families of college applicants. They anticipated that participants would lean towards giving the most money to applicants with the greatest need (e.g., with the lowest income). But they were also interested in finding out how the actual categorization of the income levels might impact decision outcomes.

To test this, they used two different ways of grouping the income ranges which were presented to participants: 

The group of participants was split so that an equal number of them selected from one or the other of these two groupings. Researchers expected that even though participants would tend to allocate more money to needier families, this would be tempered by an inclination to spread allocations over the income levels that were available to choose from.

Findings showed that this was indeed the case. More money was allocated to poorer families in the low-income grouping than in the high-income one. Specifically, the mean proportion of financial aid given to families with incomes less than or equal to $75K was 95.9 percent in the low-income grouping, as opposed to only 47.7 percent in the high-income grouping. Despite instructions given to participants that the groupings were arbitrary, they still tended to spread their allocations over the categories they had encountered.

The design had a significant influence on decision outcomes. Even something as seemingly mundane as how the ranges are categorized or grouped can have a profound effect on decision outcomes.

A Mixture: Aggregated & Non-aggregated

In a related study, researchers wanted to learn how categorization might influence how people would schedule the timing of pleasurable experiences to occur in the future. Specifically, they offered college students a chance to have three free meals during the course of the upcoming school year, and asked them to choose when they would like to consume them.

Researchers anticipated that students would be inclined to be impatient, wanting to schedule the free meals sooner rather than later in the year. But given this, researchers also wondered to what extent they might be influenced by the categorization of time periods from which to select, as well as a likely tendency to want to spread their consumption of these pleasurable experiences over time.

Since the school year is segmented into two semesters—Fall and Spring—and each of these is further divided into two terms, there are a total of four time segments that comprise the academic year. (E.g., Fall semester spans terms I and II, and Spring semester spans terms III and IV.) Researchers asked each student to select a timeframe during which they would like to partake of the complimentary meals.

To gauge the potential influence of categorization on decision outcomes, students were randomly assigned to one of two possible response sets, each of which contained three options:

Note that each of the two types of response sets includes a combination of an aggregate timeframe (e.g., semester) and a breakout of the semester (e.g., two terms). To what extent would students accommodate (or not) for this aspect of the design?

The findings showed that while all students preferred to consume the meals sooner rather than later, they were still very much influenced by the set of response options from which to select. It turned out that participants who selected from the Fall/III/IV response set were nearly four times as likely to choose most of their meals in the Spring than were participants in the I/II/Spring response set!

Notice that students were much more likely to defer consumption when the far future was segmented into smaller intervals than when the near future was segmented into smaller intervals. Needless to say, this is an interesting finding for anyone seeking to find real world ways of helping people defer consumption over time.

Multi-tiered Categorization

So far, we’ve looked at how decision making works in a single-step manner, where the level of categorization is ‘flat’ (or all on one level). But what would happen if people were deciding in a two-step manner, such that items are organized in hierarchical levels?

For example, let’s consider a situation where a philanthropist wants to make a United Way donation, allocating some money to an international fund, as well as four local funds. To what extent might her decision be influenced by whether she is first required to allocate between international versus local funds, and then take an additional step of allocating at the subordinate category level?

Two potential designs for the response set might be:

  1. A ‘flat’ (nonhierarchical) listing of funds in which all five options are presented on the same level: international funds, as well as a listing of the four local funds
  2. A hierarchical categorization requiring the respondent to first allocate between international versus local funds, and then within the ‘local’ category, in order to allocate amongst the four local funds

Note that in each of these designs, the same number of options is made available: an international fund and four local funds. The only difference is whether they have been categorized on a hierarchical basis or not.

The Effect of Design on Decision-Making

A research study that examined the effect of hierarchical versus nonhierarchical categorization showed evidence that people used the same decision strategy regardless of the design. That is, people tended to allocate evenly across the categories as they were presented, regardless of whether it was a single-step or two-step decision.

Because of this, decision outcomes differed across the two designs. In the hierarchical design, the mean donation to the international fund was 55 percent, compared to only 21 percent in the nonhierarchical design—a result that reflects the types of proportions that would be expected if people were simply allocating equally across the categories as presented.

Summary and conclusion

The results of the research studies discussed in this article show that the design of how items or options are categorized can significantly influence decision outcomes. Sometimes substantially so. People’s tendency to diversify (or seek variety) when making multiple simultaneous selections can result in allocation and choices that vary systematically with the types of groupings they encounter. This is because people tend to use the same decision strategy, regardless of how the response set is organized.

Indeed, it’s interesting to consider how something as seemingly inconsequential as how items are categorized can have such an important effect. But one thing is sure: design matters! Indeed, there is no such thing as a “neutral” design. Every design and every design decision has an effect. And for that reason, we have a responsibility to understand and appreciate how the design drives decision outcomes, and consequently, how it impacts people’s lives, sometimes in critical ways.