The masked villain: the effects of facial masking, distance, lighting, and eyewitness age on eyewitness identification accuracy

Figures & data

Table 1. The frequency of participants per age group and gender.

Age group	n	Gender	n	M	SD
Young adults	618	Female	356	31.40	7.22
		Male	239	32.30	6.72
		Other	23	32.09	8.12
Older adults	245	Female	133	54.11	7.43
		Male	98	54.82	8.08
		Other	14	60.00	10.20
Young children	266	Female	110	8.98	1.76
		Male	141	8.50	1.77
		Other	15	8.67	1.50
Older children	196	Female	96	14.46	1.89
		Male	93	13.95	1.80
		Other	7	14.14	2.19
Total	1325	Female	695	29.86	15.94
		Male	571	27.30	16.98
		Other	59	30.63	20.40

Note. A total of 1198 (90%) identified themselves as Finnish natives and 127 identified themselves as natives of various other countries.

Figure 1. Illustration of the experimental setup.

Note. Panel A: A target presented wearing a hoodie and sunglasses at 5 meters in 300 lux (i.e. optimal lighting). Panel B: A target presented wearing a hoodie and sunglasses at 12.5 meters in 300 lux. Panel C: A target presented wearing a hoodie and sunglasses at 20 meters in 300 lux.

Table 2. All participants: The frequencies of correct or incorrect line-up choices and the calculated diagnostic values per lighting level, facial masking, distance, line-up type.

Lighting	Facial masking	D	TP			TA		HR	FAR	dʹ	C
			In-correct rejection	In-correct filler ID	Correct target ID	Correct rejection	In-correct filler ID
Optimal	No mask	5	20	15	62	74	35	0.64	0.32	2.11	0.70
		12.5	38	20	45	47	55	0.44	0.54	1.34	0.83
		20	38	43	24	44	49	0.23	0.53	0.76	1.13
	Sunglasses	5	29	46	40	42	38	0.35	0.48	1.17	0.98
		12.5	13	40	18	34	65	0.25	0.66	0.73	1.03
		20	30	51	24	36	89	0.23	0.71	0.60	1.05
	Hood	5	19	9	47	41	26	0.63	0.39	1.98	0.67
		12.5	31	37	34	29	36	0.33	0.55	1.05	0.96
		20	39	51	28	54	68	0.24	0.56	0.76	1.10
	Sunglasses & Hood	5	28	39	31	39	66	0.32	0.63	0.94	0.95
		12.5	29	37	31	17	57	0.32	0.77	0.83	0.89
		20	33	46	2	33	42	0.02	0.56	−0.49	1.72
Sub-optimal	No mask	5	23	17	68	54	48	0.63	0.47	1.90	0.62
		12.5	28	41	21	51	52	0.23	0.50	0.80	1.13
		20	20	56	15	26	54	0.16	0.68	0.40	1.18
	Sunglasses	5	25	33	21	22	74	0.27	0.77	0.68	0.96
		12.5	25	53	20	42	68	0.20	0.62	0.60	1.13
		20	29	51	8	25	55	0.09	0.69	0.03	1.35
	Hood	5	40	20	62	47	58	0.51	0.55	1.50	0.73
		12.5	28	56	14	44	65	0.14	0.60	0.38	1.26
		20	23	37	8	29	42	0.12	0.59	0.26	1.32
	Sunglasses & Hood	5	33	38	26	32	68	0.27	0.68	0.75	1.00
		12.5	24	50	9	24	38	0.11	0.61	0.19	1.33
		20	24	38	9	21	41	0.13	0.66	0.25	1.26

Note. N = 1325 (4456 line-up decisions; 2260 target present responses and 2196 target absent responses). Age range: 5–90. Target = Live target observed for ∼20 s. Line-up = eight-person simultaneous line-up. Optimal = 300 lux, Suboptimal = 2 lux, Facial Masking = Disguise worn by live target, D = Distance (meters), TP = Target present line-up, TA = Target absent line-up. HR = Hit rate based on decisions from target present line-ups (Microsoft Excel formula = Target identifications / (target identifications + filler identifications + rejections)). FAR = False alarm rates based on target absent line-ups (Microsoft Excel formula = Filler identifications / (filler identifications + rejections)). dʹ = discriminability (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))). C = Decision criterion (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))/2).

Table 3. Young adults (18–44): The frequencies of correct or incorrect line-up choices and the calculated diagnostic values per lighting level, facial masking, distance, line-up type.

Lighting	Facial masking	D	TP			TA		HR	FAR	dʹ	C
			In-correct rejection	In-correct filler ID	Correct target ID	Correct rejection	In-correct filler ID
Optimal	No mask	5	6	2	27	31	17	0.77	0.35	2.45	0.48
		12.5	12	7	19	27	20	0.50	0.43	1.61	0.81
		20	19	14	11	16	18	0.25	0.53	0.83	1.09
	Sunglasses	5	10	21	19	19	18	0.38	0.49	1.24	0.93
		12.5	5	19	11	11	36	0.31	0.77	0.82	0.89
		20	8	23	12	9	40	0.28	0.82	0.68	0.93
	Hood	5	8	1	31	17	9	0.78	0.35	2.47	0.48
		12.5	16	19	17	12	15	0.33	0.56	1.03	0.96
		20	10	20	13	18	31	0.30	0.63	0.89	0.96
	Sunglasses & Hood	5	14	17	14	17	23	0.31	0.58	0.97	0.98
		12.5	13	15	14	6	23	0.33	0.79	0.86	0.86
		20	10	13	1	10	18	0.04	0.64	−0.33	1.57
Sub-optimal	No mask	5	14	6	38	37	22	0.66	0.37	2.08	0.64
		12.5	16	23	12	33	26	0.24	0.44	0.88	1.16
		20	13	23	8	8	23	0.18	0.74	0.42	1.12
	Sunglasses	5	15	16	14	14	41	0.31	0.75	0.83	0.91
		12.5	18	27	14	21	33	0.24	0.61	0.71	1.07
		20	15	21	3	11	25	0.08	0.69	−0.07	1.39
	Hood	5	17	9	40	32	31	0.61	0.49	1.81	0.64
		12.5	22	30	7	21	34	0.12	0.62	0.24	1.30
		20	13	17	2	14	14	0.06	0.50	0.00	1.53
	Sunglasses & Hood	5	22	17	17	19	39	0.30	0.67	0.86	0.95
		12.5	13	31	5	11	23	0.10	0.68	0.11	1.32
		20	14	13	5	10	24	0.16	0.71	0.34	1.18

Note. N = 618 (2108 line-up decisions; 1081 target present responses and 1027 target absent responses). Age range: 5–90. Target = Live target observed for ∼20 s. Line-up = eight-person simultaneous line-up. Optimal = 300 lux, Suboptimal = 2 lux, Facial Masking = Disguise worn by live target, D = Distance (meters), TP = Target present line-up, TA = Target absent line-up. HR = Hit rate based on decisions from target present line-ups (Microsoft Excel formula = Target identifications / (target identifications + filler identifications + rejections)). FAR = False alarm rates based on target absent line-ups (Microsoft Excel formula = Filler identifications / (filler identifications + rejections)). dʹ = discriminability (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))). C = Decision criterion (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))/2).

Table 4. Older adults (45–99): The frequencies of correct or incorrect line-up choices and the calculated diagnostic values per lighting level, facial masking, distance, line-up type.

Lighting	Facial masking	D	TP			TA		HR	FAR	dʹ	C
			In-correct rejection	In-correct filler ID	Correct target ID	Correct rejection	In-correct filler ID
Optimal	No mask	5	4	4	16	16	10	0.67	0.38	2.09	0.62
		12.5	9	2	8	5	11	0.42	0.69	1.17	0.78
		20	4	6	8	8	8	0.44	0.50	1.39	0.84
	Sunglasses	5	4	12	14	8	8	0.47	0.50	1.45	0.81
		12.5	2	6	4	5	10	0.33	0.67	0.95	0.91
		20	6	10	4	5	11	0.20	0.69	0.52	1.10
	Hood	5	4	3	9	7	7	0.56	0.50	1.69	0.69
		12.5	5	3	5	4	10	0.38	0.71	1.05	0.82
		20	8	9	5	12	9	0.23	0.43	0.86	1.18
	Sunglasses & Hood	5	6	7	12	5	13	0.48	0.72	1.29	0.69
		12.5	4	8	6	4	9	0.33	0.69	0.93	0.90
		20	9	13	0	3	5	0.00	0.63	NA	NA
Sub-optimal	No mask	5	2	3	13	3	10	0.72	0.77	1.89	0.36
		12.5	5	5	4	9	14	0.29	0.61	0.87	1.00
		20	4	12	4	7	8	0.20	0.53	0.66	1.17
	Sunglasses	5	2	4	3	3	12	0.33	0.80	0.85	0.86
		12.5	4	7	3	11	19	0.21	0.63	0.62	1.10
		20	6	11	2	4	12	0.11	0.75	0.07	1.29
	Hood	5	6	5	10	4	8	0.48	0.67	1.32	0.72
		12.5	4	9	5	8	16	0.28	0.67	0.79	0.99
		20	4	9	2	7	11	0.13	0.61	0.32	1.27
	Sunglasses & Hood	5	1	6	4	7	4	0.36	0.36	1.34	1.02
		12.5	6	9	3	9	9	0.17	0.50	0.57	1.25
		20	3	11	2	8	7	0.13	0.47	0.42	1.36

Note. N = 245 (835 line-up decisions; 432 target present responses and 403 target absent responses). Age range: 5–90. Target = Live target observed for ∼20 s. Line-up = eight-person simultaneous line-up. Optimal = 300 lux, Suboptimal = 2 lux, Facial Masking = Disguise worn by live target, D = Distance (meters), TP = Target present line-up, TA = Target absent line-up. HR = Hit rate based on decisions from target present line-ups (Microsoft Excel formula = Target identifications / (target identifications + filler identifications + rejections)). FAR = False alarm rates based on target absent line-ups (Microsoft Excel formula = Filler identifications / (filler identifications + rejections)). dʹ = discriminability (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))). C = Decision criterion (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))/2).

Table 5. Young children (5–11): The frequencies of correct or incorrect line-up choices and the calculated diagnostic values per lighting level, facial masking, distance, line-up type.

Lighting	Facial masking	D	TP			TA		HR	FAR	dʹ	C
			In-correct rejection	In-correct filler ID	Correct target ID	Correct rejection	In-correct filler ID
Optimal	No mask	5	6	7	8	19	4	0.38	0.17	1.72	1.16
		12.5	9	4	10	6	13	0.43	0.68	1.20	0.77
		20	9	17	2	11	17	0.07	0.61	−0.03	1.45
	Sunglasses	5	14	9	2	14	6	0.08	0.30	0.38	1.59
		12.5	5	13	1	13	10	0.05	0.43	−0.02	1.61
		20	13	13	2	13	19	0.07	0.59	−0.02	1.46
	Hood	5	6	5	5	12	7	0.31	0.37	1.20	1.09
		12.5	7	10	8	7	3	0.32	0.30	1.31	1.12
		20	13	13	4	13	18	0.13	0.58	0.35	1.28
	Sunglasses & Hood	5	6	7	2	12	20	0.13	0.63	0.31	1.26
		12.5	5	7	3	4	9	0.20	0.69	0.52	1.10
		20	9	7	0	11	7	0.00	0.39	NA	NA
Sub-optimal	No mask	5	4	7	12	7	11	0.52	0.61	1.48	0.69
		12.5	3	9	4	7	8	0.25	0.53	0.83	1.09
		20	2	9	1	5	8	0.08	0.62	0.04	1.40
	Sunglasses	5	4	7	3	3	14	0.21	0.82	0.47	1.03
		12.5	3	14	2	6	9	0.11	0.60	0.19	1.35
		20	2	10	1	4	8	0.08	0.67	−0.04	1.40
	Hood	5	13	4	8	3	13	0.32	0.81	0.80	0.87
		12.5	1	12	1	10	10	0.07	0.50	0.07	1.50
		20	2	4	1	1	9	0.14	0.90	0.15	1.14
	Sunglasses & Hood	5	7	6	5	3	16	0.28	0.84	0.66	0.92
		12.5	2	7	1	2	5	0.10	0.71	0.06	1.31
		20	5	5	1	1	6	0.09	0.86	−0.09	1.29

Note. N = 266 (880 line-up decisions; 443 target present responses and 437 target absent responses). Age range: 5–90. Target = Live target observed for ∼20 s. Line-up = eight-person simultaneous line-up. Optimal = 300 lux, Suboptimal = 2 lux, Facial Masking = Disguise worn by live target, D = Distance (meters), TP = Target present line-up, TA = Target absent line-up. HR = Hit rate based on decisions from target present line-ups (Microsoft Excel formula = Target identifications / (target identifications + filler identifications + rejections)). FAR = False alarm rates based on target absent line-ups (Microsoft Excel formula = Filler identifications / (filler identifications + rejections)). dʹ = discriminability (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))). C = Decision criterion (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))/2).

Table 6. Older children (12–17): The frequencies of correct or incorrect line-up choices and the calculated diagnostic values per lighting level, facial masking, distance, line-up type.

Lighting	Facial masking	D	TP			TA		HR	FAR	dʹ	C
			In-correct rejection	In-correct filler ID	Correct target ID	Correct rejection	In-correct filler ID
Optimal	No mask	5	4	2	11	8	4	0.65	0.33	2.11	0.68
		12.5	8	7	8	9	11	0.35	0.55	1.09	0.94
		20	6	6	3	9	6	0.20	0.40	0.80	1.24
	Sunglasses	5	1	4	5	1	6	0.50	0.86	1.24	0.62
		12.5	1	2	2	5	9	0.40	0.64	1.15	0.83
		20	3	5	6	9	19	0.43	0.68	1.19	0.78
	Hood	5	1	0	2	5	3	0.67	0.38	2.11	0.62
		12.5	3	5	4	6	8	0.33	0.57	1.03	0.95
		20	8	9	6	11	10	0.26	0.48	0.92	1.10
	Sunglasses & Hood	5	2	8	3	5	10	0.23	0.67	0.65	1.06
		12.5	7	7	8	3	16	0.36	0.84	0.90	0.80
		20	5	13	1	9	12	0.05	0.57	−0.15	1.54
Sub-optimal	No mask	5	3	1	5	7	5	0.56	0.42	1.76	0.74
		12.5	4	4	1	2	4	0.11	0.67	0.16	1.30
		20	1	12	2	6	15	0.13	0.71	0.23	1.23
	Sunglasses	5	4	6	1	2	7	0.09	0.78	−0.04	1.32
		12.5	0	5	1	4	7	0.17	0.64	0.44	1.19
		20	6	9	2	6	10	0.12	0.63	0.23	1.30
	Hood	5	4	2	4	8	6	0.40	0.43	1.36	0.93
		12.5	1	5	1	5	5	0.14	0.50	0.47	1.30
		20	4	7	3	7	8	0.21	0.53	0.71	1.15
	Sunglasses & Hood	5	3	9	0	3	9	0.00	0.75	NA	NA
		12.5	3	3	0	2	1	0.00	0.33	NA	NA
		20	2	9	1	2	4	0.08	0.67	0.00	1.38

Note. N = 196 (633 line-up decisions; 304 target present responses and 329 target absent responses). Age range: 5–90. Target = Live target observed for ∼20 s. Line-up = eight-person simultaneous line-up. Optimal = 300 lux, Suboptimal = 2 lux, Facial Masking = Disguise worn by live target, D = Distance (meters), TP = Target present line-up, TA = Target absent line-up. HR = Hit rate based on decisions from target present line-ups (Microsoft Excel formula = Target identifications / (target identifications + filler identifications + rejections)). FAR = False alarm rates based on target absent line-ups (Microsoft Excel formula = Filler identifications / (filler identifications + rejections)). dʹ = discriminability (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))). C = Decision criterion (Microsoft Excel formula = (−(NORMSINV(HR) + NORMSINV(FA/8))/2).

Table 7. Confidence per lighting, facial masking, distance, and age group.

Lighting	Facial masking	Distance		Young adults			Older adults			Young children			Older children
			N	n	M	SD	n	M	SD	n	M	SD	n	M	SD
Optimal	No mask	5	200	79	64.73	27.44	48	67.08	28.14	44	63.43	34.96	29	65.83	23.94
		12.5	198	82	55.79	29.70	34	56.65	30.56	40	58.73	32.03	42	56.45	24.74
		20	188	73	55.14	27.21	30	58.00	26.86	55	44.47	32.35	30	47.40	24.83
	Sunglasses	5	185	83	52.22	26.19	43	56.79	23.74	42	60.48	31.80	17	52.24	22.12
		12.5	164	80	48.63	23.82	26	48.08	22.70	40	47.95	30.15	18	48.94	22.26
		20	222	88	45.18	23.61	34	40.21	28.35	59	49.51	31.42	41	50.51	19.40
	Hood	5	136	62	72.77	21.24	28	52.71	30.96	35	59.51	35.59	11	54.45	26.57
		12.5	161	76	54.12	27.54	26	42.54	22.53	33	54.82	25.31	26	56.96	22.74
		20	228	87	45.71	26.26	38	41.58	26.63	60	41.30	31.24	43	48.23	24.20
	Sunglasses & Hood	5	195	81	44.74	25.53	41	47.10	23.50	45	55.73	31.03	28	50.11	22.68
		12.5	168	70	45.91	24.74	30	39.20	25.77	28	49.46	33.88	40	56.15	24.09
		20	152	49	35.88	25.59	29	34.31	26.66	34	45.71	32.78	40	46.73	21.88
Suboptimal	Normal	5	204	113	64.63	28.01	29	55.03	26.29	41	66.59	26.64	21	58.76	26.83
		12.5	181	108	52.30	24.64	31	52.45	25.77	27	50.85	27.77	15	52.73	26.03
		20	156	68	37.60	22.98	28	35.07	27.03	24	48.04	32.41	36	44.92	20.70
	Sunglasses	5	169	96	43.94	23.04	22	44.86	23.13	31	43.45	22.38	20	45.60	20.66
		12.5	202	111	43.38	24.29	41	41.66	28.04	33	43.33	28.80	17	42.24	32.22
		20	151	67	32.52	22.64	27	32.15	21.54	25	51.28	32.55	32	45.00	20.77
	Hood	5	221	125	61.13	24.48	31	54.35	21.64	41	55.34	23.23	24	49.83	24.59
		12.5	196	112	44.34	24.95	37	36.92	24.14	30	51.17	29.30	17	39.53	21.16
		20	126	54	32.41	23.34	28	27.21	19.68	16	51.75	31.65	28	41.14	21.74
	Sunglasses & Hood	5	192	110	43.76	24.31	21	38.38	23.78	37	48.65	31.78	24	42.75	21.55
		12.5	137	81	35.07	22.06	31	35.74	25.99	16	51.19	32.40	9	43.89	35.85
		20	117	58	27.43	21.49	24	28.46	20.92	17	46.59	31.93	18	37.06	23.85

Note. The total sample includes 4456 line-up decisions. However, there were 207 missing confidence judgements in the sample (i.e. participants that forgot or due to errors did not report their confidence). The table above contains 4249 decisions (4456–207). Age range: 5–90. Line-up type = eight-person simultaneous line-up. Optimal = 300 lux, Suboptimal = 2 lux, Facial Masking = Disguise worn by live target, D = Distance (meters), Young adults = 18–44, Older adults = 45–90, Young children 5–11, Older children 12–17.

Table 8. Response time per lighting, facial masking, distance, and age group.

Lighting	Facial masking	Distance		Young adults			Older adults			Young children			Older children
			N	n	M	SD	n	M	SD	n	M	SD	n	M	SD
Optimal	No mask	5	206	83	15.08	8.95	50	17.68	13.60	44	10.33	5.18	29	11.72	7.70
		12.5	205	85	15.58	9.07	35	17.28	9.56	42	13.97	10.15	43	12.29	9.90
		20	198	78	16.83	12.05	34	14.85	8.53	56	15.90	14.05	30	14.84	17.45
	Sunglasses	5	203	85	19.80	11.37	43	24.69	19.69	47	15.86	8.25	28	15.58	7.25
		12.5	171	71	17.41	8.92	31	19.54	9.66	28	18.98	11.74	41	12.76	7.13
		20	156	52	21.19	12.38	30	16.08	8.47	34	18.82	11.95	40	15.40	10.12
	Hood	5	195	87	20.02	10.90	46	20.37	12.49	45	17.61	11.65	17	18.64	8.10
		12.5	170	82	18.78	11.00	27	21.71	13.20	42	17.92	12.23	19	21.42	15.09
		20	230	92	17.15	8.94	36	19.86	10.63	60	16.40	10.34	42	13.33	7.66
	Sunglasses & Hood	5	142	66	13.53	8.21	30	18.25	9.54	35	14.12	10.98	11	20.01	12.61
		12.5	167	79	17.18	10.74	27	18.84	11.07	35	12.14	6.99	26	12.64	7.41
		20	240	92	16.14	9.18	43	17.52	9.57	61	14.79	9.69	44	19.31	26.39
Suboptimal	Normal	5	210	117	13.98	8.23	31	19.27	10.42	41	13.08	9.07	21	12.33	6.32
		12.5	193	110	15.17	10.47	37	17.94	7.72	31	15.22	12.04	15	16.08	7.08
		20	171	75	14.94	8.65	35	16.86	10.11	25	14.50	6.41	36	16.59	10.51
	Sunglasses	5	197	114	17.60	9.06	22	19.14	10.93	37	18.49	15.12	24	18.87	10.10
		12.5	145	83	16.28	7.28	36	19.76	11.22	17	12.45	8.11	9	15.41	11.20
		20	133	66	15.22	9.55	31	16.42	10.33	18	18.95	12.86	18	13.69	5.40
	Hood	5	227	129	14.93	8.85	33	19.89	13.01	41	14.28	7.87	24	14.87	6.87
		12.5	207	114	16.38	11.52	42	18.15	12.16	34	15.41	12.34	17	17.62	8.03
		20	139	60	14.59	8.14	33	17.75	11.85	17	15.73	9.08	29	14.73	6.79
	Sunglasses & Hood	5	175	100	18.74	9.37	24	18.93	10.98	31	15.06	9.00	20	16.67	7.52
		12.5	208	113	18.36	16.42	44	18.35	8.10	34	16.51	12.90	17	18.07	18.25
		20	168	75	14.33	6.45	35	17.72	10.34	25	11.79	7.04	33	15.76	9.09

Note. The total sample includes 4456 line-up decisions. Age range: 5–90. Line-up type = eight-person simultaneous line-up. Optimal = 300 lux, Suboptimal = 2 lux, Facial Masking = Disguise worn by live target, D = Distance (meters), Young adults = 18–44, Older adults = 45–90, Young children 5–11, Older children 12–17.

Figure 2. Identification accuracy in TP line-ups with no facial masking in optimal and suboptimal lighting (only choosers).

Note. The predicted probabilities of identification accuracy in target present line-ups (only the choosers) by distance, lighting, and age group with no facial masking. Here, the sample had been reduced to only the choosers (i.e. all decisions that were rejections had been removed). On the left, optimal lighting. On the right, suboptimal lighting. TP = Target present line-up, Optimal lighting = 300 lux, Suboptimal lighting = 2 lux. Young adults = 18–44, Older adults = 45–90, Young children 5–11, Older children 12–17. All results are based on multilevel binary logistic regressions. The line-ups consisted of eight images that were presented simultaneously; meaning that chance level of identifying the actual target was at .125.

Figure 3. Rejection accuracy in TA line-ups with no facial masking in optimal and suboptimal lighting (choosers & non-choosers).

Note. The predicted probabilities of rejection accuracy in target absent line-ups (choosers and non-choosers) by distance, lighting, and age group with no facial masking. On the left, optimal lighting. On the right, suboptimal lighting. TA = Target absent line-up, Optimal lighting = 300 lux, Suboptimal lighting = 2 lux. Young adults = 18–44, Older adults = 45–90, Young children 5–11, Older children 12–17. All results are based on multilevel binary logistic regressions. The line-ups consisted of eight images that were presented simultaneously.

Figure 4. Identification accuracy in TP line-ups with facial masking in optimal and suboptimal lighting (only choosers).

Note. Panel A: The predicted probabilities of identification accuracy in target present line-ups by distance and facial masking in optimal lighting. Panel B: The same as Panel A but in suboptimal lighting. Here, the sample had been reduced to only the choosers (i.e. all decisions that were rejections had been removed). TP = Target present line-up, Optimal lighting = 300 lux, Suboptimal lighting = 2 lux. Young adults = 18–44, Older adults = 45–90, Young children 5–11, Older children 12–17. All results are based on multilevel binary logistic regressions. The line-ups consisted of eight images that were presented simultaneously; meaning that chance level of identifying the actual target was at .125.

Figure 5. Rejection accuracy in TA line-ups with facial masking in optimal and suboptimal lighting (choosers & non-choosers).

Note. Panel A: The predicted probabilities of rejection accuracy in target absent line-ups (choosers and non-choosers) by distance and facial masking in optimal lighting. Panel B: The same as Panel A but in suboptimal lighting. TA = Target absent line-up, Optimal lighting = 300 lux, Suboptimal lighting = 2 lux. Not separated by age group because there was no main effect of age. Age groups = Young adults = 18–44, Older adults = 45–90, Young children 5–11, Older children 12–17. All results are based on multilevel binary logistic regressions. The line-ups consisted of eight images that were presented simultaneously.

Figure 6. Confidence-accuracy characteristic analyses per lighting, facial masking, distance, and confidence bin.

Note. Panels A-H: Confidence-Accuracy Characteristic Analyses per Lighting, Facial Masking, Distance, and Confidence Bin. The total sample includes 4456 line-up decisions. However, there were 207 missing confidence judgements in the sample (i.e. participants that forgot or due to errors did not report their confidence). The figure above contains information based on 4249 decisions (4456–207). Please see Table 9 for the raw values that the CAC analyses are based on. Age range: 5–90. Optimal = 300 lux, Suboptimal = 2 lux, no mask = No disguise worn by live target, sunglasses, hood, sunglasses & hood = descriptions of the disguise worn by the live target. CAC = Confidence-accuracy characteristic, Confidence = post line-up decision. Distance was measured in meters (m). The formula used to calculate the CAC value per bin (0–59%, 60–89%, 90–100%) was: target present (TP) identifications / (TP filler identifications + (target absent (TA) filler identification)/8). In other words, we assumed fair line-ups and that is why we divided by eight (i.e. the number of images in the line-ups).

Table 9. Confidence-accuracy characteristic analyses per lighting, facial masking, distance, line-up type, decisions, and confidence bin.

Lighting	Facial masking	Distance	Target present									Target absent						CAC
			Target selections			Filler selections			Rejections			Filler selections			Rejections
			0–59	60–89	90–100	0–59	60–89	90–100	0–59	60–89	90–100	0–59	60–89	90–100	0–59	60–89	90–100	0–59	60–89	90–100
Optimal	No mask	5	12	26	22	8	1	6	10	5	5	21	9	2	26	26	21	0.82	0.96	0.99
		12.5	12	19	13	12	3	4	23	8	7	37	8	6	20	19	7	0.72	0.95	0.95
		20	13	7	3	29	10	2	19	14	4	27	14	3	23	12	8	0.79	0.80	0.89
	Sunglasses	5	14	17	9	28	12	3	9	11	6	28	6	2	25	8	7	0.80	0.96	0.97
		12.5	10	8	0	25	9	4	7	5	1	49	9	4	20	10	3	0.62	0.88	0.00
		20	8	15	1	35	14	0	17	8	4	64	19	3	22	8	4	0.50	0.86	0.73
	Hood	5	10	16	19	5	2	2	11	6	1	14	6	4	18	15	7	0.85	0.96	0.97
		12.5	11	17	6	25	5	3	19	9	2	27	8	1	13	12	3	0.77	0.94	0.98
		20	15	5	7	39	9	2	26	9	4	50	15	0	28	15	4	0.71	0.73	1.00
	Sunglasses & Hood	5	18	8	3	29	6	3	20	5	1	46	14	4	24	8	6	0.76	0.82	0.86
		12.5	16	12	3	22	12	3	18	9	2	39	12	3	13	2	2	0.77	0.89	0.89
		20	1	0	0	39	5	2	23	6	3	31	8	3	22	7	2	0.21	0.00	0.00
Suboptimal	No mask	5	15	28	23	8	5	3	13	5	3	28	15	4	25	15	14	0.81	0.94	0.98
		12.5	9	9	2	25	11	1	17	8	1	37	7	4	24	17	9	0.66	0.91	0.80
		20	11	3	0	40	10	2	16	3	0	40	6	2	15	5	3	0.69	0.80	0.00
	Sunglasses	5	15	4	2	24	6	1	17	8	0	59	11	1	12	7	2	0.67	0.74	0.94
		12.5	15	4	1	38	9	3	18	4	2	47	16	4	30	9	2	0.72	0.67	0.67
		20	6	2	0	39	7	1	18	5	1	43	9	1	14	3	2	0.53	0.64	0.00
	Hood	5	22	32	6	10	7	3	23	13	2	41	13	3	21	15	10	0.81	0.95	0.94
		12.5	10	1	3	41	8	4	18	7	2	46	13	2	29	11	1	0.63	0.38	0.92
		20	4	3	0	29	2	1	14	4	1	33	7	1	21	5	1	0.49	0.77	0.00
	Sunglasses & Hood	5	15	8	2	28	7	2	19	9	4	50	16	2	24	6	0	0.71	0.80	0.89
		12.5	8	1	0	41	5	2	16	5	1	30	6	1	15	4	2	0.68	0.57	0.00
		20	6	0	1	31	5	0	17	2	1	33	2	2	16	1	0	0.59	0.00	0.80

Note: The total sample includes 4456 line-up decisions. However, there were 207 missing confidence judgements in the sample (i.e. participants that forgot or due to errors did not report their confidence). The table above contains 4249 decisions (4456–207). Age range: 5–90. Line-up type = eight-person simultaneous line-up. Optimal = 300 lux, Suboptimal = 2 lux, Facial Masking = Disguise worn by live target, D = Distance (meters). CAC = Confidence-accuracy characteristic, Confidence = post line-up decision, Distance was measured in meters (m). The formula used to calculate the CAC value per bin (0–59%, 60–89%, 90–100%) was: target present (TP) identifications / (TP filler identifications + (target absent (TA) filler identification)/8). In other words, we assumed fair line-ups and that is why we divided by eight (i.e. the number of images in the line-ups).

Figure 7. Raw data receiver operating characteristic (ROC) curves per lighting.

Note. Receiver Operating Characteristic (ROC) Curves per lighting condition. Optimal Lighting = 300 lux, Suboptimal Lighting = 2 lux. This analysis was based on 4249 observations (4456 observations minus 207 missing confidence judgements).

Figure 8. Raw data receiver operating characteristic (ROC) curves per lighting, facial masking, and distance.

Note. Panel A: Receiver Operating Characteristic (ROC) curves per distance (5 meters (m), 12.5, 20 m) in optimal lighting (300 lux). This analysis was based on 2197 observations (2283 observations minus 86 missing confidence judgements). Panel B: ROC curves per distance in suboptimal lighting (2 lux). This analysis was based on 2052 observations (2173 observations minus 121 missing confidence judgements). Panel C: ROC curves per facial masking in optimal lighting (no_mask = no facial masking worn by live target; sunglasses = facial masking (sunglasses) worn by live target, hood = facial masking (hood) worn by live target, sunglasses_&_hood = facial masking (sunglasses and hood) worn by live target). This analysis was based on 2197 observations (2283 observations minus 86 missing confidence judgments). Panel D: ROC curves per facial masking in suboptimal lighting. This analysis was based on 2052 observations (2173 observations minus 121 missing confidence judgements).

Figure 9. Raw data receiver operating characteristic (ROC) curves per lighting, facial masking, and distance.

Note. Panel A: Receiver Operating Characteristic (ROC) curves per distance (5 meters (m), 12.5, 20 m) in optimal lighting (300 lux) with no facial masking worn by target. This analysis was based on 586 observations (609 observations minus 23 missing confidence judgements). Panel B: ROC curves per distance in suboptimal lighting (2 lux) with no facial masking worn by target. This analysis was based on 541 observations (574 observations minus 33 missing confidence judgements). Panel C: ROC curves per distance in optimal lighting with sunglasses worn as facial masking by target. This analysis was based on 571 observations (595 observations minus 24 missing confidence judgements). Panel D: ROC curves per distance in suboptimal lighting with sunglasses worn as facial masking by target. This analysis was based on 522 observations (551 observations minus 29 missing confidence judgements). Panel E: ROC curves per distance in optimal lighting with a hood worn as facial masking by target. This analysis was based on 525 observations (549 observations minus 24 missing confidence judgements). Panel F: ROC curves per distance in suboptimal lighting with hood worn as facial masking by target. This analysis was based on 543 observations (573 observations minus 30 missing confidence judgements). Panel G: ROC curves per distance in suboptimal lighting with sunglasses and a hood worn as facial masking by target. This analysis was based on 515 observations (530 observations minus 15 missing confidence judgements). Panel H: ROC curves per distance in suboptimal lighting with sunglasses and a hood worn as facial masking by target. This analysis was based on 446 observations (475 observations minus 29 missing confidence judgements). Due to the lack of target identifications, we were unable to create ROC curves for 20 m for the figures in Panels G and H.

Data availability statement

The datasets analyzed for this study can be found in the Open Science Framework: https://osf.io/s5tcu/; https://osf.io/s5tcu/?view_only=0c6f9455d2434110a6c33b230ffc5377