Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

# People systematically overlook subtractive changes

## Abstract

This is a preview of subscription content

## Access options

### Extended Data Fig. 3 Study material for experiments 2 to 4.

Participants reported their ideas for how to change this miniature golf hole. We coded whether each idea was additive (for example, ‘add a windmill’), subtractive (for example, ‘remove the sand trap’) or neither (for example, ‘reverse the direction’). Participants were randomly assigned to a no-cue instruction that mentioned neither addition nor subtraction or to a cue condition that reminded participants they could ‘add or subtract’. In experiments 2 and 3, participants reported all of the ways that they might improve the original. In experiment 4, participants were randomly assigned to a condition that solicited their improvement ideas or a condition that solicited their ideas for making the hole worse.

### Extended Data Fig. 4 Cumulative percentage of participants who included at least one type of idea by the ith idea in their list in experiments 2 to 4.

Participants were randomly assigned to one of two conditions in experiments 2 and 3 (no-cue versus cue) (n = 312); and one of four conditions in experiment 4 ((no-cue versus cue) × (improve versus make-it-worse)) (n = 369). The y axes show cumulative percentage. The x axes show idea order (i). Empty blue shapes represent subtractive ideas and filled orange shapes represent additive ideas. Dotted lines represent no-cue conditions and solid lines represent cue conditions. Circles represent responses to an improve prompt and triangles represent responses to a make-it-worse prompt. a, b, Across experiments 2 (a) and 3 (b), we did not find evidence that the cue affected the likelihood of participants listing at least one additive idea (odds ratio = 0.92, z = −0.24, P = 0.810), but we did find evidence that it increased the likelihood of participants listing at least one subtractive idea (odds ratio = 1.93, z = 2.73, P = 0.006). c, In experiment 4, the cue increased the likelihood of participants listing at least one subtractive idea within the improvement conditions (no-cue = 21%, cue = 48%, χ2 = 13.63, P < 0.001) and the make-it-worse conditions (no-cue = 28%, cue = 50%, χ2 = 9.71, P = 0.002). The error band represents s.e. of proportion.

## Supplementary information

### Supplementary Information

This file contains information about participant exclusions, analyses, and results for experiments 1—8 (section 1), descriptions of methods and results for studies S1—S12 (section 2), and additional references (see page 1 for details).

## Rights and permissions

Reprints and Permissions

Adams, G.S., Converse, B.A., Hales, A.H. et al. People systematically overlook subtractive changes. Nature 592, 258–261 (2021). https://doi.org/10.1038/s41586-021-03380-y

• Accepted:

• Published:

• Issue Date:

• DOI: https://doi.org/10.1038/s41586-021-03380-y

• ### Translational framework for implementation evaluation and research: a normalisation process theory coding manual for qualitative research and instrument development

• Carl R. May
• Bianca Albers
• Tim Rapley

Implementation Science (2022)

• ### An actionable understanding of societal transitions: the X-curve framework

• Aniek Hebinck
• Gijs Diercks
• Derk Loorbach

Sustainability Science (2022)

• ### Why Do We Care More About Disease than Health?

• Martin Picard

Phenomics (2022)

• ### Binding free energy predictions in host-guest systems using Autodock4. A retrospective analysis on SAMPL6, SAMPL7 and SAMPL8 challenges

• Lorenzo Casbarra
• Piero Procacci

Journal of Computer-Aided Molecular Design (2021)