Table 3.

Checklist of methodological aspects relevant for measuring the ANS.

Methodological aspect	Recommendation	Reason
Stimulus range	Use numerosities beyond subitizing range (i.e., larger 4)	Numerosities up to four are processed via subitizing and, therefore, not based on the ANS (Feigenson et al., 2004). Furthermore, also stimuli just above subitizing range might be counted easily and should, therefore, be avoided.
Presentation duration	Use short presentation durations (around 100 ms)	Inglis and Gilmore (2013) found that 16 ms were sufficient to compare two dot sets. Additionally, counting strategies are prevented.
Ratios	Use a wide range of different ratios	Too difficult/too simple tasks can result in floor/ceiling effects Note: difficulty of tasks does not solely depend on ratios, but also on age or the task employed (e.g., same-different task is more difficult than dot comparison tasks, e.g., Smets et al., 2014).
Visual confounds	Control for visual cues	Controlling for visual cues is necessary to ensure that numerosity is processed and not visual cues (Gebuis and Reynvoet, 2011).
Feedback	Avoid feedback	Feedback affects motivation, might enhance strategy use or ANS acuity (DeWind and Brannon, 2012; Lindskog et al., 2013a).
Number of trials	Employ about 400 trials Use an adaptive procedure	To reach acceptable reliability about 400 trials are necessary. An adaptive procedure is more economic (Lindskog et al., 2013a).