Table 2.

Mean response scores of debriefing questions, by study arm and combined study arms[Link]

Study arm	Overlap	Color coding	The questions were clear	I could easily identify the differences between health states	I could easily chose between the health states	There were too many choice tasks in the survey	The survey's topic was interesting
				Separately for each study arm
1	No	No color	5.5 (base)	5.0 (base)	3.9 (base)	3.6 (base)	5.4 (base)
2	No	Shades of purple	5.6	5.1	4.1	3.6	5.5
3	Yes	No color	5.7**	5.4***	4.4***	3.3***	5.4
4	Yes	Highlighting	5.7*	5.4***	4.2**	3.3***	5.6**
5	Yes	Shades of purple	5.7*	5.4***	4.4***	3.3***	5.5*
				Separately by level overlap
1, 2	No	—	5.5 (base)	5.0 (base)	4.0 (base)	3.6 (base)	5.4 (base)
3–5	Yes	—	5.7***	5.4***	4.3***	3.3***	5.5
				Separately by color coding
1, 3	—	No	5.6 (base)	5.2 (base)	4.2 (base)	3.5 (base)	5.4 (base)
2, 4, 5	—	Yes	5.6	5.3**	4.2	3.4	5.6**

All answers were provided on 7‐point Likert response scales ranging from 1 (I totally disagree) to 7 (I totally agree).

^* Null hypothesis of equal means is rejected with p < 0.1 (one‐sided).

^** Null hypothesis of equal means is rejected with p < 0.05 (one‐sided).

^*** Null hypothesis of equal means is rejected with p < 0.01 (one‐sided).