MaxDiff anchoring

In a MaxDiff exercise, respondents typically evaluate small groups of items (usually 3–6 at a time), selecting their “best” and “worst” options. These choices reflect relative preferences among the items chosen by the researcher.

However, relative judgments alone can be misleading. For example, a respondent who doesn’t eat sugar may still select a “best” and “worst” ice cream flavor, even though they would never purchase any of them. Their choices may resemble those of an enthusiastic ice-cream eater, despite very different real-world behavior.

MaxDiff anchoring addresses this limitation by introducing an absolute reference point. Anchoring helps distinguish whether items are genuinely important or unimportant, or whether respondents would actually buy or consider them, rather than only how items compare to one another.

Anchoring works by appending an additional grid question to the end of the MaxDiff exercise. This grid includes a subset of items (typically around seven), spanning from each respondent’s most to least favored options based on their earlier MaxDiff choices. Respondents then indicate which of these items they would actually buy or consider important.

This implementation of Anchored MaxDiff, known as the Direct Binary Approach Method, was proposed by Kevin Lattery and has been presented at several Sawtooth Software conferences.

Results are displayed with a utility boundary line separating important and unimportant items. Items above the line have positive utility and are considered important (e.g., buy), while items below the line have negative utility and are considered unimportant (e.g., not buy).

Responses to the anchoring question, together with the MaxDiff exercise results, are used during analysis to estimate the anchor line (utility boundary).

When MaxDiff anchoring is enabled, the exercise is analyzed using anchored MaxDiff methods rather than standard MaxDiff analysis.

1. Open the Advanced tab.
2. Toggle on the MaxDiff anchoring setting.
   

Configuring the anchoring question is straightforward.

1. Configure the Anchor question text as you would any other question text, including rich text, images, and formatting.  
2. Specify “Important” and “Not important” labels.
3. Optionally, override the number of items shown to each respondent.
   1. By default, 7 items are shown (or all items if the exercise contains only 6).
   2. A minimum of 5 items is required.
   3. The number of anchored items should depend on the length of your MaxDiff item list. As the number of items in the exercise increases, consider displaying more items in the anchoring question to help the algorithm estimate a more accurate utility boundary between important and unimportant items.

 

Anchored scores are displayed on a positive probability (ratio) scale, with the anchor indexed at 100:

• A score of 50 indicates the item is half as preferred or important as the anchor.
• A score of 100 indicates the item has an equal likelihood of falling above or below the anchor.
• A score of 200 indicates the item is twice as preferred or important as the anchor.