Table 3.

Articles and studies included in the systematic review: demographic and methodological characteristics.

ID	Article	Language	Total	Sel	Size	Participant	Age	Range	N	Match
1	van Alphen 2017	Dutch				Mix	37.9		42	No
2	Berends 2015	Dutch		High		Drisk	36	30–42	14	No
3	Bergelson 2018	NAE		algo_driven	segm	Typ	11.5	5–20	61	Yes
4	Bredin-Oja 2018+	NAE	138	Random	cont	Mix	36	28–46	6	No
5	Bulgarelli 2019+	NAE	1,932	Random	segm	Typ	6.5	6–7	44	Yes
6	Burgess 2013	NAE	465	High	cont	ASD	51	35–67	10	No
7	Busch 2018+	Dutch	240	algo_driven	cont	Typ	42	24–60	5	No
8	Canault 2016+	Fr	3,240	High		Typ	25.5	3–48	18	No
9	Caskey 2014	NAE, Sp	25			Preterm	0	32–36 wg	5	No
10	Cristia, Lavechin, et al., 2019	NAE, UKE, Tsimane	1,472	Random	cont		16.8	3–58	49	No
11	D'Apice 2019	UKE	320	High	cont	Typ	33.2	24–48	107	No
12.1	Elo 2016	Finnish	698	Random	cont	Twin	7	0	1	No
12.2	Elo 2016	Finnish	630	Random	cont	Twin	9	0	1	No
13	Ganek 2018+	Viet	100	algo_driven	cont	Mix	30.5	22–42	10	No
14	Gilkerson 2015*+	Chn	330	High	cont	Typ	12.1	3–23	22	No
15.1	Jones 2019+	NAE	120	Random	cont	ASD	78	60–96	8	No
15.2	Jones 2019+	NAE	105	Random	cont	ASD	192	168–216	7	No
15.3	Jones 2019+	NAE	2,210		segm	ASD	96	60–204	36	No
16	Ko 2016	NAE	26	algo_driven	segm	Typ	20.4	12–30	13	Yes
17	Lehet 2019	NAE	734	algo_driven	cont	Mix	20	4–34	23	No
18	McCauley 2011	NAE	150	Random		ASD			5	No
19	Merz 2019	NAE	600	algo_driven		Mix	90	61–119	10	No
20	Oetting 2009+	NAE	510	Random	cont	Low SES	42	24–60	17	No
21	Orena 2019+	Fr, NAE	945	algo_driven	segm	Bilingual	10	10–11	21	No
22	Pae 2016*	Korean	630			Typ	12.5	4–16		No
23	Schwarz 2017	Swedish	240	Random	cont	Typ	30		4	No
24	Seidl 2018	NAE		algo_driven	segm	ASDrisk			10	No
25	Soderstrom 2016	NAE	1,305	Random	segm	Typ	25	12–38	32	Yes
26	VanDam 2016+	NAE	47	algo_driven	segm	Typ	29.1		26	Yes
27	Weisleder 2013	Sp	600	algo_driven		Mix	19	0	10	No
28.1	Xu 2009*	NAE	4,200	High	segm	Mix	19	2–36	70	Yes
28.2	Xu 2009*	NAE	720	Random		Typ	10	0	1	Yes
28.3	Xu 2009*	NAE	720	Random		highL	31	0	1	Yes

Note. Regardless of how many authors a publication has, articles are identified by the first author and year of publication. Articles with an asterisk (*) denote Language Environment Analysis (LENA) affiliation; those with a plus sign (+) were published as evaluations in peer-reviewed journals. “Match” reflects whether it matches the LENA training sample. All numeric predictors have been rounded for this display. Language = native language of participants; Total = total duration of annotated samples (in minutes); Sel = type of selection; Size = size of the sample provided to the human annotator; Participant = characteristics of the participant sample; Age = mean age of participants in months; Range = age range of participants in months (except for Caskey et al. [2014], where it indicates weeks gestation [wg]); N = number of children included in the sample; mix = a mixture of any of the above; high = based on high AWC/CVC/CTC; Drisk = at risk for developmental delays; NAE = North American English; segm = segment; typ = typically developing and with none of the other characteristics; random = unrelated to LENA segmentation; cont = continuous; ASD(risk) = diagnosed with (or at risk for) autism spectrum disorder; Fr = French; Sp = Spanish; UKE = United Kingdom English; twin = having a twin; Viet = Vietnamese; Chn = Shanghai Chinese and/or Mandarin Chinese; low SES = family with low socioeconomic status; highL = high language.