Data Hazard: Irrelevant Information Contaminates Judgment

Imagine I asked you whether the number of steps to the top of the Eiffel Tower is more or less than 250 and then asked you to give me your estimate of the actual number. Your estimate would probably be higher if I first asked you whether that number is more or less than 5,000.

The Eiffel Tower example (which has 1,710 steps, by the way) illustrates the anchoring effect, one of the most important phenomena studied by researchers in the area of judgment and decision making. It occurs when exposure to a (often arbitrary) number serves as an unconscious reference point that influences subsequent judgments.

Anchoring is frequently studied in relation to value estimates. One experiment asked participants to write down the last three digits of their phone number multiplied by one thousand (e.g. 678 = 678,000). Results showed that people’s subsequent estimate of local house prices were significantly influenced by the arbitrary anchor, even though they were given a 10-minute presentation on facts and figures from the housing market at the beginning of the study.

Experiments in the area of judgment and decision making usually have one limitation: It is unclear whether the anchoring effect persists beyond the duration of the experiment.

New research by Sangsuk Yoon and Nathan Fong investigates what happens to the anchoring effect over time.

In their experiments, the researchers asked participants to draw a single card from a deck with numbers ranging from 0 to 99 (the anchor). Participants later had to indicate how much they were willing to pay for different goods, such as a bluetooth keyboard or an art book. Willingness to pay was measured in the initial session, then again after different time intervals.

Results of one study showed that the anchoring effect dropped after a one-week delay, but the effect was still statistically significant and remained stable for a duration of four to eight weeks. (A separate experiment found that neither participants’ motivation to give consistent answers over time nor their memory of the initial anchor could account for the persistence of the effect.) Even after eight weeks, a one-unit increase in the anchor (the 0 to 99 number on the card) still caused a 14 cent increase in willingness to pay, on average.

This research suggests that an arbitrary anchor does not just distort the scale that people apply to their value judgments; they don't just temporarily "think bigger" or "think smaller" than they would without the anchor. The information also leaves a long-lasting imprint that isn’t easily corrected. Anchors essentially contaminate your judgment. That’s why anchoring may well be one of the stickiest traps decision makers can fall into—and a powerful tool that others can use to manipulate your decisions.