Joint vs. Separate Evaluation

When shown two dictionaries together, most people choose the one with more entries, even if its cover is torn. This reveals a quirk in how we judge value: our preferences change based on whether we evaluate something in isolation (separate evaluation) or alongside alternatives (joint evaluation). When comparing options side-by-side, we tend to over-focus on attributes that are easy to compare (like the number of entries), even if we wouldn't have cared about them otherwise, leading to inconsistent choices.

Ask this: What am I comparing this to, and what am I ignoring?

The core mechanism behind this effect is evaluability. Some attributes are easy to judge on their own, while others are only meaningful in comparison.

In separate evaluation, when you see one option alone, you rely on features that are easy to assess in a vacuum. A torn cover is an obvious flaw, so you might devalue that dictionary. The absolute number of entries—10,000 vs. 20,000—is hard to evaluate. Is 10,000 good? Is 20,000 necessary? Without context, the number is abstract.

In joint evaluation, seeing both options together makes the number of entries suddenly very easy to compare: 20,000 is clearly double 10,000. This highly evaluable, quantifiable attribute becomes a dominant factor, often overriding the concern about the torn cover.

This applies to many real-world decisions. When hiring, a candidate's GPA is highly comparable in joint evaluation, so a manager might overweigh a 3.9 from a state school versus a 3.7 from an Ivy League, ignoring the harder-to-evaluate context of school prestige. Similarly, when choosing a TV, a technical spec like 'contrast ratio' is meaningless to most people in isolation but becomes a key decision point when comparing models side-by-side in a store, even if the real-world viewing difference is negligible.

Research centers on the "evaluability hypothesis": an attribute's influence depends on how easily it can be judged. Attributes that are hard to evaluate in isolation often become decisive when a direct comparison makes them salient.

• Initial studies found that salary, hard to evaluate alone, became the deciding factor when job seekers compared two offers side-by-side. [1] (1996)
• The classic dictionary problem showed the number of entries (hard to evaluate) dominated decisions in joint evaluation, while a torn cover (easy to evaluate) mattered more in separate evaluation. [2] (1999)
• Recent work shows cognitive traits can moderate the effect; individuals higher in cognitive reflection are less susceptible to the bias. [3] (2017)

Experts with well-calibrated internal standards are less susceptible, as the effect is weaker for attributes that are easy to evaluate in isolation (e.g., severe pain). Joint evaluation can also backfire, causing a focus on trivial but comparable metrics over more important qualitative factors.