importancia para minimizar el error tipo 1 porque todas las pruebas se

correlacionaron entre sí. Se realizaron pruebas Post Hoc (HSD de

Tukey) cuando el efecto principal de Grado o Grupo fue significativo. Se

realizaron pruebas de efectos simples para examinar el efecto del

Tiempo para cada Grado o Grupo cuando hubo interacción Tiempo x

Grado o Tiempo x Grupo. También calculamos los tamaños del efecto (

gramo; Hedges, 2007) como la diferencia media estimada del modelo

ajustada por el error residual en otoño. Similar a la d de Cohen, la g de

Hedges es el número de desviaciones estándar basado en el modelo

que difieren dos medidas (controlando todos los demás efectos en el

modelo).

Sintaxis expresiva en inglés (CELF-WS)

los niños tenían una puntuación estándar promedio de 2,7, más de 2 DE por

debajo de la media normativa. Solo el 11% de los niños puntuaron dentro de

1 DE de la media normativa. Los niños del grupo de solo habla mostraron

mejoras desde el otoño (media = 2,7) hasta la primavera (media = 3,5,gramo

=0.28), pero no hubo efecto principal de Grupo. La Figura S1 muestra que los

niños aumentaron el número total de elementos correctos desde el otoño

hasta la primavera, pero ambos grupos obtuvieron puntajes cercanos al

piso.

Solo hablado y bimodal

Sintaxis receptiva de ASL

del 61,9 al 69,9% correcto,gramo=0,53. Los alumnos de kindergarten obtuvieron una

puntuación más baja (media = 56,0 %) que los de primer grado (media = 65,0 %,gramo=

0,60) que puntuaron más bajo que los alumnos de segundo grado (media = 79,3%,gramo=

0.90.) Como es evidente enTabla 2y la Figura S1, las puntuaciones de los niños en los

grupos bimodales y de solo signo fueron muy similares y mostraron una mejora constante

desde el otoño hasta la primavera y entre los grados. Con un promedio de 82.0% correcto

en la primavera, muchos alumnos de segundo grado estuvieron cerca del techo.

A lo largo del año escolar, las puntuaciones aumentaron

Resultados

Los grupos puntuaron en el rango promedio para el CI no verbal sin

diferencias de grupo. Como era de esperar, los grupos diferían en sus

habilidades del habla (vertabla 1). Las tablas S3–S14 (disponibles en

línea) presentan los resultados estadísticos de los ANOVA y las pruebas

de seguimiento. Debido al volumen y la complejidad de los datos, y

para mejorar la legibilidad, resumimos estos resultados en Tabla 2.

Describimos sólo aquellos efectos que fueron estadísticamente

significativos.

Resumen de resultados de idioma

sintaxis expresiva en inglés, los niños DHH se retrasaron en comparación con las normas

de la prueba. El vocabulario fue un punto fuerte en comparación con el conocimiento de la

sintaxis de los niños. La sintaxis expresiva en inglés se retrasó gravemente y casi todos los

niños obtuvieron calificaciones muy por debajo del rango promedio. En general, los niños

mostraron mejoras en las puntuaciones estándar desde el otoño hasta la primavera,

aunque los tamaños del efecto fueron pequeños, con la excepción del aprendizaje de

vocabulario por parte de los niños de jardín de infantes (gramo=0,46). Las puntuaciones

estándar fueron similares en todos los grados. Los niños que solo hablan y los bimodales

obtuvieron puntajes similares en las pruebas de vocabulario y sintaxis. Los niños que solo

usan señas obtuvieron puntajes de vocabulario similares a los de los otros dos grupos,

pero puntajes más bajos en la sintaxis del inglés receptivo (gramo=0,45–0,50). En la

prueba de ASL, los niños bimodales y de solo señas mostraron una mejora notablemente

similar y constante a lo largo del tiempo y el grado.

Sobre vocabulario y receptivo y

Idioma

Tabla 3muestra las medias de la puntuación del idioma, las SD y el porcentaje de

puntuación de la muestra dentro del rango promedio. Las tablas S2–S4 muestran

los resultados del modelo.

Vocabulario (EOWPVT)

La muestra total obtuvo una puntuación estándar de 83,2, con un 44 % de niños

con una puntuación dentro o por encima de 1 SD de la media normativa (es decir,

por encima de 85). A lo largo del año escolar, las puntuaciones estándar de

vocabulario para la muestra total aumentaron de 81,2 a 84,9,gramo=0.25. Los

alumnos de kínder aumentaron sus puntajes de 80.8 a 86.2,gramo=0.46. Los

alumnos de primer grado aumentaron sus puntajes de 81.9 a 85,gramo=0.19. En

segundo grado, las puntuaciones estándar fueron planas. Por lo tanto, las

ganancias de vocabulario a lo largo del año escolar fueron mayores que las

ganancias esperadas en la muestra normativa en jardín de infantes y primer grado

y similares a las ganancias esperadas en la muestra normativa en segundo grado.

A pesar de las mejoras ocurridas durante el año escolar, no hubo efectos

significativos para el grado. Los niños de los tres grados comenzaron el año

escolar con puntajes estándar de alrededor de 80. El gráfico de puntajes

brutos que se muestra en la Figura S1 (disponible en línea) muestra un

aumento continuo en la cantidad de palabras correctas con el tiempo (es

decir, de otoño a primavera y de grado al grado) y también muestra que el

crecimiento fue similar para los tres grupos. La única excepción fueron los

niños de jardín de infantes que solo usaban señas y que comenzaron con

puntajes bajos de vocabulario, pero se pusieron al día con los otros dos

grupos al final de su año de jardín de infantes.

Lectura

Tabla 4muestra las medias de la puntuación estándar de lectura, las SD y el

porcentaje de puntuación de la muestra dentro del rango medio. Las tablas S5–S7

muestran los resultados del modelo.

Lectura de una sola palabra (WJ Word ID)

La muestra total obtuvo una puntuación media de 91,2 con el 67 % de los niños

con una puntuación dentro de 1 SD de la norma (en o por encima de 85). Los

alumnos de jardín de infancia (media = 95,7) obtuvieron puntajes más altos que los

de segundo grado (media = 84,9,gramo=0,65). Solo los niños de kindergarten

aumentaron significativamente sus puntajes desde el otoño (media = 94.5) hasta la

primavera (media = 97.1),gramo=0.14. En general, sólo hablado (media = 96,1,

gramo=0,65) y niños bimodales (media = 92,9,gramo=0,45) puntuaron más alto

que los niños de solo signos (media = 85,9), sin diferencias entre los dos primeros

grupos. Los niños que solo hablan aumentaron sus puntuaciones estándar desde

el otoño (media = 94,7) hasta la primavera (media = 97,6,gramo=0,15). Las

puntuaciones W aumentaron para los tres grupos (ver Figura S2). Sintaxis

Sintaxis receptiva en inglés (TACL)La muestra total tuvo una puntuación

estándar promedio de 5,8 con el 37% de los niños dentro de 1 SD de la

media normativa. Los puntajes aumentaron de 5.6 a 6.0 a lo largo del año

escolar,gramo=0.17. Niños en el solo hablado (media = 6.2, gramo=0,45) y

bimodal (media = 6,4,gramo=0,50) los grupos puntuaron de manera similar

entre sí pero más alta que los niños en el grupo de solo signos (media = 4,9).

La figura S1 (en línea) muestra que cada grupo aumentó el total de

elementos correctos con el tiempo.

Lectura sin palabras (WJ Word Attack)

La muestra total obtuvo una puntuación media de 82,2 con un 46 % de niños con

una puntuación dentro de 1 DE (85 o más). Las puntuaciones estándar

disminuyeron desde el otoño (media = 83,5) hasta la primavera (media = 81,0,

gramo=0,19). Los niños de kindergarten (media = 93.2) tenían un estándar más alto

Tabla 3Medias y desviaciones estándar por tiempo, grado y grupo en medidas de lenguaje

Prueba Puntaje Grupo Jardín de infancia

Caer Primavera

METRO(DAKOTA DEL SUR) METRO(DAKOTA DEL SUR)

Primer grado

Caer Primavera

METRO(DAKOTA DEL SUR) METRO(DAKOTA DEL SUR)

Segundo grado

Caer Primavera

METRO(DAKOTA DEL SUR) METRO(DAKOTA DEL SUR)

Total %≥85/%≥7

METRO(DAKOTA DEL SUR)

Vocabulario

(EOWPVT)

Estándar

puntaje

Hablado

bimodal

Firmar

Total

%≥85

Hablado

bimodal

Firmar

Total

%≥7

Hablado

bimodal

Total

%≥7

bimodal

Firmar

Total

82,8 (16,1)

82,1 (13,4)

76,6 (16,3)

80,8 (15,4)

37%

6,5 (2,9)

6.6 (3.1)

4.7 (2.9)

5.9 (3.1)

43%

3.1 (2.6)

2,5 (2,9)

2.9 (2.7)

11%

51,7 (15,3)

50,6 (18,5)

51,1 (17,0)

85,5 (16,6)

86,2 (14,8)

86,9 (16,6)

86,2 (15,9)

80,4 (12,5)

83,8 (11,2)

81,2 (18,2)

81,9 (14,6)

82,8 (13,1)

86,5 (15,2)

86,0 (17,7)

85,3 (15,6)

79,5 (10,2)

78,2 (14,6)

84,0 (11,3)

81,0 (12,2)

78,9 (11,2)

81,7 (14,9)

85,4 (11,3)

82,6 (12,7)

82,3 (14,2)

83,5 (14,1)

83,6 (15,8)

83,2 (14,8)

43%

44%

45%

55%

6.8 (3.2)

7,0 (3,2)

4.8 (2.8)

6.2 (3.2)

44%

3.9 (3.4)

2.3 (2.2)

3.2 (3.0)

dieciséis%

36%

5,5 (2,6)

6.1 (2.6)

4.4 (2.6)

5.2 (2.7)

34%

2.2 (2.1)

2.1 (2.5)

2.1 (2.3)

6%

49%

5,9 (2,6)

5.8 (2.8)

5.2 (2.7)

5.6 (2.7)

34%

3.2 (2.9)

2.2 (2.6)

2.6 (2.7)

11%

41%

5,7 (2,5)

6,0 (2,9)

5.1 (2.8)

5,5 (2,7)

29%

2.6 (2.2)

2.1 (3.1)

2.3 (2.7)

7%

41%

6.7 (2.7)

6.8 (2.4)

5.7 (2.6)

6.3 (2.6)

37%

3.2 (2.4)

2.3 (3.0)

2.7 (2.8)

14%

44%

6.2 (2.9)

6.4 (2.9)

4,9 (2,7)

5,8 (2,9)

37%

3.1 (2.8)

2.3 (2.7)

Ing. rec.

sintaxis (TACL)

Estándar

puntaje

40%

47%

28%

Ing. Exp.

sintaxis

(CELF-WS)

Estándar

puntaje

13%

9%

2.7 (2.7)

11%

64,5 (16,5)

66,9 (19,3)

65,8 (18,1)

Rec. de ASL

sintaxis (Schick,

2013)

Por ciento

correcto

60,4 (14,8)

61,4 (17,9)

60,9 (16,4)

60,3 (14,5)

61,0 (18,0)

60,7 (16,4)

69,9 (13,5)

69,3 (17,8)

69,6 (15,8)

73,6 (13,2)

78,7 (9,6)

76,7 (11,3)

79,5 (9,8)

83,9 (9,4)

82,0 (9,7)

Nota: norma de prueba EOWPVT:METRO=100; SD = 15. Normas de prueba TACL y CELF-WS:METRO=10; SD = 3. Los números en negrita son medias y SD de toda la muestra en cada medida de lenguaje, así

como el porcentaje de toda la muestra con una puntuación igual o superior a 1 SD de la media normativa de la prueba. El porcentaje se puede comparar con el 85 % de la muestra normalizadora de

audición que puntuó en este nivel. A todos los niños se les administró la prueba de vocabulario y sintaxis receptiva en inglés. Solo a los niños bimodales y de solo señas se les administró la prueba de

sintaxis receptiva de ASL.

Tabla 4Medias y desviaciones estándar por tiempo, grado y grupo en puntajes estándar de medidas de lectura

Prueba Puntaje Grupo Jardín de infancia

Caer Primavera

METRO(DAKOTA DEL SUR) METRO(DAKOTA DEL SUR)

Primer grado

Caer Primavera

METRO(DAKOTA DEL SUR) METRO(DAKOTA DEL SUR)

Segundo grado

Caer Primavera

METRO(DAKOTA DEL SUR) METRO(DAKOTA DEL SUR)

Total %≥85

METRO(DAKOTA DEL SUR)

Una sola palabra

(ID de palabra WJ)

Estándar

puntaje

Hablado 95,6 (16,5) 101.2

(14.1)

96,2 (15,1) 98,3 (14,3) 91,6 (14,3) 90,3 (13,1) 90,3 (15,9) 88,1 (16,3) 92,9 (15,0)

91,7 (14,4) 91,6 (13,8) 86,0 (14,1) 86,4 (14,4) 80,2 (15,8) 78,6 (16,7) 85,9 (15,6)

94,5 (15,4) 97,1 (14,6) 90,8 (14,9) 90,7 (14,5) 85,5 (15,4) 84,3 (15,7)91,2 (15,7)

76%81%66%69%49%53%67%

103.7

(15.2)

97,8 (20,5) 95,0 (17,6) 80,3 (16,1) 81,3 (15,5) 79,6 (20,7) 77,4 (21,1) 85,6 (19,9)

81,4 (14,1) 73,6 (12,7) 66,1 (12,9) 62,7 (14,8) 62,2 (18,6) 55,6 (21,2) 66,0 (17,6)

96,6 (18,8) 90,4 (19,3) 78,8 (19,0) 78,0 (20,0) 74,7 (21,5) 71,2 (24,3)82,2 (22,2)

67%60%34%36%36%37%46%

97,0 (12,8) 96,6 (15,7) 86,4 (15,5) 85,4 (12,0) 81,6 (8,3)

96,6 (13,0) 94,0 (16,6) 84,5 (14,2) 80,9 (14,3) 79,0 (15,9)

95,1 (12,3) 89,3 (14,2) 79,6 (17,0) 75,9 (14,9) 71,9 (13,0) 68,6 (16,4) 80,4 (17,3)

96,3 (12,6) 93,3 (15,7) 83,0 (15,9) 80,1 (14,4) 76,4 (13,5) 74,4 (15,1)85,2 (16,7)

86%75%47%36%29%25%53%

97,4 (14,5) 98,1 (13,9) 89,6 (11,0) 89,7 (9,0) 96,1 (14,4) 81%

bimodal

Firmar

Total

%≥85

Hablado

72%

53%

Sin palabras (WJ

Ataque de Palabra)

Estándar

puntaje

102.7

(14.0)

97,4 (14,3) 96,6 (13,4) 91,3 (11,7) 91,7 (11,2) 98,7 (14,4) 83%

bimodal

Firmar

Total

%≥85

Hablado

bimodal

Firmar

Total

%≥85

49%

13%

Lectura comp.

(Paso WJ.

Comp.)

Estándar

puntaje

81,1 (8,3)

77,3 (14,7)

90,1 (14,5)

86,4 (16,4)

sesenta y cinco%

56%

43%

Nota. Normas de las subpruebas WJ:METRO=100; SD = 15. Los números en negrita son medias y SD de toda la muestra en cada medida de lectura, así como el porcentaje de toda la muestra con una

puntuación igual o superior a 1 SD de la media de la prueba normativa. El porcentaje se puede comparar con el 85 % de la muestra normalizadora de audición que puntuó en este nivel.

puntuaciones

alumnos de segundo grado (media = 73.0,gramo=1.01). Los niños que solo hablan

(media = 98,7) obtuvieron puntajes estándar más altos que los bimodales (media =

85.6,gramo=0.75) y niños de solo signos (media = 66.0,gramo=1.85). Los niños

bimodales obtuvieron puntajes estándar más altos que los niños de solo signos,

gramo=1.10. Las puntuaciones de los niños que solo usan señas disminuyeron de

68,0 a 64,2 a lo largo del año escolar,gramo=0.44. La figura S2 muestra que los

niños que solo hablan hicieron aumentos constantes en las palabras sin palabras.

que primero calificadores (media = 78.4,gramo=0.73) y lectura a través del tiempo y el grado. Los niños bimodales lograron ganancias

menores. Los niños que solo usan señas progresaron más lentamente y sus

puntajes W disminuyeron en segundo grado.

Comprensión de lectura (comprensión de pasaje WJ)

La muestra total tuvo una puntuación media de 85,2 con el 53 % de los niños

puntuando dentro de 1 SD. Las puntuaciones disminuyeron de 86,5 a 83,8

Tabla 5Medias y desviaciones estándar por tiempo, grado y grupo en medidas de conciencia fonológica hablada

Prueba Puntaje Grupo Jardín de infancia

Caer Primavera

METRO(DAKOTA DEL SUR) METRO(DAKOTA DEL SUR)

Primer grado

Primavera

METRO(DAKOTA DEL SUR)

Segundo grado

Caer Primavera

METRO(DAKOTA DEL SUR) METRO(DAKOTA DEL SUR)

Total %≥7

Caer

METRO(DAKOTA DEL SUR) METRO(DAKOTA DEL SUR)

Sonido

fósforo

(CTOPP)

Estándar

puntaje

Hablado

bimodal

Firmar

Total

%≥7

Hablado

bimodal

Total

%≥7

7.8 (2.1)

7.2 (3.3)

4.3 (3.7)

6.4 (3.4)

67%

4.8 (4.9)

8.3 (2.1)

7.2 (3.3)

6.1 (3.1)

7.2 (3.0)

66%

9.4 (3.4)

7.4 (2.3)

6.3 (2.5)

4.5 (2.3)

5,8 (2,6)

32%

8.0 (4.0)

7.7 (2.9)

6.5 (2.2)

4.9 (2.1)

6.1 (2.6)

34%

9.4 (2.0)

6.3 (2.2)

6.4 (2.5)

5,0 (2,5)

5,7 (2,5)

38%

7,0 (3,3)

6,7 (0,6)

6.3 (2.4)

4,8 (2,6)

5.7 (2.4)

40%

7,9 (2,7)

7.7 (2.3)

6,7 (2,8)

4,9 (2,9)

6.3 (3.0)

50%

7.6 (4.1)

74%

54%

29%

Mezcla

(CTOPP)

Estándar 70%

puntaje 4.4 (4.5) 4.8 (5.2) 3.2 (4.0) 3.7 (4.2) 4.3 (4.0) 4.4 (3.6) 4.1 (4.3) 34%

4.6 (4.7)

42%

7.3 (4.9)

66%

5.3 (4.6)

44%

6.1 (4.4)

60%

5,5 (3,9)

46%

6,0 (3,6)

48%

5,8 (4,5)

52%

Nota. Normas de las subpruebas CTOPP:METRO=10; SD = 3. A todos los niños se les administró la prueba de coincidencia de sonido. Solo se les administró la prueba de blending a los niños bimodales y de

lenguaje hablado. Los números en negrita son las medias y las DE de la muestra completa en cada medida de conciencia fonológica hablada, así como el porcentaje de la muestra completa con una

puntuación igual o superior a 1 DE de la media de las medias de las pruebas normativas.

a lo largo del año escolar,gramo=0.19. Los niños de kindergarten (media =

94,8) obtuvieron puntajes más altos que los de primer grado (media = 81,6,

gramo=0,90) que puntuaron más alto que los de segundo grado (media =

75,4,gramo=0,42). Hijos que solo hablan (media = 90,1,gramo=0,59) y niños

bimodales (media = 86,4,gramo=0,39) puntuaron más alto que los niños de

solo signos (media = 80,4), sin diferencias entre los dos primeros grupos. Los

niños que solo usan señas disminuyeron sus puntuaciones desde el otoño

(media = 82,4) hasta la primavera (media = 78,3,gramo=0,33) al igual que los

niños bimodales (media de otoño = 87,8; media de primavera = 85,0,gramo=

0,21). Por el contrario, las puntuaciones de los niños que solo hablan se

mantuvieron estables durante el año escolar (media de otoño = 90,4, media

de primavera = 89,8;gramo=0,02). La Figura S2 muestra que, si bien los tres

grupos aumentaron sus puntajes W de comprensión de lectura con el

tiempo, la brecha entre sus puntajes y los de la muestra normativa aumentó

con el grado.

Coincidencia de sonido (coincidencia de sonido CTOPP)

un puntaje promedio de 6.3 con el 50% de los niños obteniendo puntajes dentro o

por encima del rango promedio. Los niños que solo hablan por señas (media = 4.9)

obtuvieron puntajes más bajos que los que solo hablan (media = 7.7,gramo=0,88) y

niños bimodales (media = 6,7,gramo=0,60), sin diferencia entre estos dos últimos

grupos.

La muestra total tuvo

Mezcla de sonido (mezcla CTOPP)

Los niños DHH tuvieron un puntaje promedio de 5.8 con el 52% de los niños

obteniendo puntajes dentro del rango promedio. Las puntuaciones aumentaron

desde el otoño (media = 5,1) hasta la primavera (media = 6,6),gramo=0.31. Hijos

que solo hablan (media = 7.6,gramo=0,86) obtuvieron puntuaciones más altas que

los niños bimodales (media = 4,1) y aumentaron sus puntuaciones desde el otoño

(media = 6,2) hasta la primavera (media = 9,0),gramo=0,53. La Figura S3 muestra

que ambos grupos aumentaron sus puntajes combinados desde el otoño hasta la

primavera, pero la brecha entre sus puntajes y los del grupo de normalización

auditiva aumentó con el grado.

Bimodal y solo hablado

Resumen de resultados de lectura

Si bien la mayoría de los niños DHH de kínder obtuvieron puntajes dentro del rango

promedio en las tres habilidades de lectura evaluadas, los puntajes estándar disminuyeron

con el grado, con tamaños de efecto grandes. Las habilidades de lectura de los niños de

kínder fueron similares a las de la muestra normalizada, con un 75 % con puntajes de

comprensión de lectura de primavera dentro del rango promedio. Por el contrario, solo el

25% de los alumnos de segundo grado obtuvieron calificaciones dentro del rango

promedio en la primavera. Los tres grupos diferían ampliamente en las habilidades de

lectura. Los niños en el grupo de solo señas obtuvieron puntajes más bajos que los niños

en los grupos bimodales y solo hablados en las tres pruebas de lectura, con tamaños de

efecto medianos a grandes. Los niños bimodales y de lenguaje hablado diferían en la

lectura de no palabras, pero no en la lectura de una sola palabra o la comprensión de

lectura. Para la primavera de segundo grado, los niños de los tres grupos tenían mejores

habilidades de lectura de palabras que de comprensión de lectura. Para el segundo grado,

aunque los puntajes disminuyeron con el grado, los niños bimodales y solo hablados

todavía tenían habilidades de lectura de palabras cercanas a las apropiadas para su edad,

mientras que los niños que solo usaban señas obtuvieron 1 SD por debajo de los

promedios normativos. Los tres grupos obtuvieron calificaciones por debajo de las normas

auditivas en comprensión de lectura.

Spoken PA results summary

dren had spoken phonological awareness skills within the average

range and performed equally well on sound matching and blending.

Bimodal children performed much better on sound matching than

sound blending. Sign-only children scored more than 1.5 SD below the

test norms on sound matching.

The majority of spoken-only chil-

Fingerspelling PA

Table 6 displays percentage of items correct for the sign-only and

bimodal children. Tables S11–S13 display model results.

FS imitation

over the school year, g = 0.58. Kindergarteners had lower scores (mean

= 22.3%) than first graders (mean = 39.0%), g = 0.79, who had lower

scores than second graders, (mean = 61.2%), g = 1.03. Figure S4 shows

the steady increase in percentage correct items for sign-only and

bimodal children with time.

Children increased their scores from 34.0 to 45.3%

Spoken Phonological Awareness (PA) FS blending

over the school year, g = 0.50. Kindergarteners (mean = 9.0%) scored

lower than first graders (mean = 20.0%, g = 0.60) who scored lower

than second graders (mean = 37.4%, g = 0.97).

Children increased their scores from 17.0 to 26.7%,

Table 5 displays spoken PA standard score means, SDs, and the

percentage of the sample scoring within the average range or higher.

Tables S8–S10 display model results.

Downloaded from https://academic.oup.com/jdsde/advance-article-abstract/doi/10.1093/deafed/enz050/5734969 by University of Exeter user on 04 May 2020

Table 6 Means and standard deviations by time, grade, and group on fingerspelling phonological measures

Test Score Group Kindergarten

Fall Spring

M (SD) M (SD)

First grade

Spring

M (SD)

Second grade

Fall Spring

M (SD) M (SD)

Total

Fall

M (SD) M (SD)

F.

Imitation

Percent

correct

Bimodal

Sign

Total

Bimodal

15.8 (17.6) 28.8 (22.9) 28.9 (22.0) 41.6 (21.0) 54.4 (23.1) 73.4 (16.5) 37.7 (27.3)

40.9 (26.8)

39.5 (27.0)

19.5 (21.4)

22.6 (25.1)

21.3 (23.6)

14.9 (24.7)

16.2 (25.1)

15.6 (24.9)

17.1 (16.4)

16.5 (16.9)

6.3 (13.0)

3.5 (7.7)

4.7 (10.5)

2.4 (11.4)

4.4 (13.3)

3.5 (12.4)

27.4 (21.2)

28.0 (21.9)

36.3 (24.9)

32.9 (23.8)

49.7 (25.1)

45.8 (23.4)

56.8 (20.8)

55.9 (21.6)

62.6 (20.8)

67.0 (19.7)

F.

Blending

Percent

correct

14.0 (19.5)

11.9 (19.2)

16.8 (18.7)

16.2 (19.5)

18.0 (17.9)

28.1 (27.6)

27.0 (21.6)

35.5 (24.7)

43.2 (21.7)

Sign 43.6 (23.2)

Total

Bimodal

Sign

Total

12.8 (19.2) 16.5 (19.0) 23.4 (23.9) 32.1 (23.7) 43.4 (22.4)

  1. Elision Percent

correct

7.0 (14.2)

4.9 (11.8)

5.8 (12.9)

4.3 (8.2)

8.5 (15.6)

6.5 (12.8)

18.0 (22.9)

20.1 (26.3)

25.0 (28.5)

25.4 (27.5)

41.7 (35.1)

37.1 (32.9)

19.1 (24.6) 25.2 (27.7) 39.0 (33.6)

Note. All means are percentage of correct items. The bold numbers are means and SDs of the whole sample on each fingerspelling phonological measure.

Figure S4 shows the steady increase of percentage correct items for

both groups with time.

to spoken language made gains in expressive English syntax, though

they scored well below age norms. Those students who had access to

sign made steady gains in ASL syntax.

Our sample achieved an average expressive vocabulary standard

score of 83, lower than the standard score of 91 reported by Geers et

  1. (2009), but similar to the average score of 81 achieved by children 8

years of age as reported by Hayes et al. (2009). Less than 50% of the

children scored within the average range (standard scores 85–115)

somewhat worse than the 58% reported by Geers et al. Children in all

three grades made progress during the course of the school year, but

kindergartners and first graders made greater than expected progress

(based on test norms), while second graders made expected progress.

Our sample showed better outcomes than that of Connor et al. (2000)

who reported that, although vocabulary growth was positive, growth

was less than expected based on test norms. In first and second

grades, our total sample showed annual gains of 3–5 standard score

points, which was substantial, but less than the 8 standard scores in

annual gain reported by Hayes et al. (2009). It is important to note that

children did not differ on vocabulary knowledge based on their group.

Despite greater than expected progress during the school year, the

gap between the DHH students and their typical hearing peers

remained wide across grades. This suggests that elementary-aged

DHH children may be losing the gains they made during the school

year over the summer as has been shown for preschoolers (Scott,

Goldberg, Connor & Lederberg, 2019).

English syntax was the weakest area for all children as reported

also by other researchers (Geers et al., 2003; Hay-McCutcheon et al.,

2008). The children in our sample were 1.5 SD below age norms in

receptive English syntax and 2.5 SD below in expressive English syntax.

Access to spoken language (in this case spoken English) was an

advantage in learning receptive English syntax, ref lected in the similar

and higher scores of spoken-only and bimodal children compared to

the sign-only children. However, all three groups showed some

improvement (shown by raw scores) in receptive English syntax over

the school year. While progress might be expected of the children who

had auditory access to English, the sign-only children without such

access also mastered additional items. We speculate that these

children may be developing bilingual skills and learning to understand

English word order from adults around them who use contact sign.

Consistent with findings by other researchers (e.g., Geers et al.,

2003, 2009), access to spoken language was not sufficient

FS elision

19.9% over the school year, gramo=0.56. Kindergarteners did not

increase their scores. First graders increased their scores from 6.5 to

19.1%, gramo=0.72. Second graders increased their scores from 25.2 to

39.0%, gramo=0.80. Kindergarteners scored lower (mean = 4.7%) than

second graders (mean = 31.9%), gramo=1.49, but not first graders. First

graders (mean = 12.8%) scored lower than second graders, gramo=

1.05. As is evident in Figure S4, while most kindergarteners could not

do this task, there was a steady increase in percentage correct

responses from first grade fall through second grade spring.

Children increased their scores from 11.3 to

FS summary Both sign-only and bimodal children made similar and

steady gains on these three tasks. Imitation was the easiest task, while

elision was the hardest. By second grade, children did not reach ceiling

for any of the fingerspelling tests.

Discussion

The purpose of this research was to describe the language and literacy

of a large sample of young DHH children and examine their progress

longitudinally over the course of a school year, cross-sectionally from

grade to grade, as well as across groups. We purposefully included

DHH students who had different access to language; thus we included

children who had access to spoken or sign language only and also

those who had access to both. The latter group particularly tends to be

little studied. We discuss key findings arranged by the major

constructs: language, reading, and spoken and fingerspelled

phonological awareness.

Idioma

While children started in kindergarten delayed compared to the

hearing test norms, they made greater than expected gains during the

school year, as shown by increases in standard scores in all areas of

language from fall to spring. Substantially more children achieved

within the average range in vocabulary than in English syntax. In

vocabulary and English syntax, children made expected progress

between kindergarten and second grade, compared to age-based

hearing norms. Children who had access

Downloaded from https://academic.oup.com/jdsde/advance-article-abstract/doi/10.1093/deafed/enz050/5734969 by University of Exeter user on 04 May 2020

for acquiring age-appropriate expressive English syntax. Only 11% of

children in the spoken-only and bimodal groups scored within the

average range on the CELF. However, the spoken-only children made

more than expected progress from fall to spring, while the bimodal

children made expected progress. The raw scores show that the

spoken-only children mastered about five additional items during the

school year; the bimodal children mastered less than one additional

item. These differences may ref lect a different instructional focus

across classrooms; teachers may spend more time in direct or indirect

instruction on English syntax in spoken-language-only classrooms.

Both groups of children who had access to sign (bimodal and signonly)

showed substantial improvement in ASL receptive syntax from fall

to spring and from grade to grade. In fact, by the end of second grade,

children were close to ceiling on this assessment. There was no effect

of group, suggesting that access to spoken language did not appear to

inhibit the bimodal children’s learning of ASL syntactical structures.

that their W scores increased over time suggests they may use visual

and kinesthetic means to acquire knowledge of spoken phonemes (

Haptonstall-Nykaza & Schick, 2007;Kyle, Campbell & MacSweeney, 2016

;Kyle & Harris, 2010). Sign-only children also showed lower

achievement in word reading and reading comprehension compared

to both bimodal and spoken-only children. Given that the three groups

did not differ in overall language ability (Lederberg et al., 2019), and

that bimodal and sign-only children were exposed to the same sign

input in the classroom, it appears that access to spoken language

facilitated learning to read. Sign-only children need to learn to

translate English words into ASL which is a complex process (

Hoffmeister y Caldwell-Harris, 2014).

Phonological Awareness

As expected, auditory perception was an advantage for spoken

phonological awareness. Spoken-only children scored within the low

average range on both sound matching and blending. These results

are similar to those reported by James et al. (2008) for early implanted

CI children and what some researchers have found for DHH

kindergarteners (Kyle y Harris, 2011; Nittrouer et al., 2012). The

bimodal children were similar to the spoken-only children on sound

matching but scored considerably lower in sound blending. Their

poorer performance on blending might have been due to speech

production issues. Although the sign-only children received standard

scores about 3 SD below the test norms on sound matching, the raw

scores for this group at each grade level were greater than 0 and

increased from fall to spring. As with nonword reading, these results

might indicate that sign-only children are using visual information such

as speechreading and fingerspelling to identify some spoken

phonemes.

This is the first study to examine developmental change in

fingerspelling PA. Fingerspelling increased across the school year and

grade to grade. Imitation was a relatively easy task while elision was

difficult. The ability to manipulate the sublexical features of words, as

shown in the blending and elision tasks, increased across time, but

even second graders were correct on less than half of the items. This is

worse than the spoken-only children did on the parallel task of sound

blending. It may be that blending fingerspelling is a more difficult task,

but it also may be a result of lack of instruction on this skill in signing

classrooms. Both the sign-only and bimodal children performed in a

remarkably similar manner on the fingerspelling tasks achieving

similar percentage scores and showing significant growth from fall to

spring and developmental differences across grades. This may suggest

the bimodal children could access the sublexical features of English

words through both auditory and visual means, and access to one did

not seem to inhibit learning in the other.

Lectura

Reading, to a much larger extent than language, was affected by both

grade and group. Kindergarteners showed age-appropriate skills in all

three areas assessed: single-word reading, nonword reading, and

reading comprehension. At kindergarten, these tests assess children’s

emerging literacy skills, namely, alphabetic knowledge (letter-name,

letter-sound) and the ability to understand concepts depicted by

rebuses (not printed words). Other researchers (e.g., Cupples et al.,

2013) have also reported that hearing loss and language delay does

not prevent acquisition of these early skills. The children in our sample

increased their reading skills from fall to spring and from grade to

grade as indicated by their W scores but made less than expected

progress from kindergarten to second grade, consistent with findings

of other researchers (Harris et al., 2017b;Kyle y Harris, 2011).

Group had a large effect on reading. Spoken-only children had ageappropriate

nonword and single-word reading skills with 80% of

children (across grades) scoring in the average range, very close to the

85% for the normative sample. Thus, spoken-only children were able to

develop the knowledge of how to sound out and blend fairly complex

words by second grade. Lederberg et al. (2013) suggest that phonics

instruction may be particularly effective with DHH children who have

auditory access to spoken language because letters can serve as visual

support for their learning of spoken phonology. Despite strength in

word reading, spoken-only children showed a decline of 16 standard

points in reading comprehension from kindergarten to second grade.

Bimodal children had weaker nonword reading skills than spokenonly

children, with 50% scoring within the normative average range.

Despite differences in phonological and spoken language skills,

bimodal and spoken-only children had similar word reading and

reading comprehension achievement. These results are in contrast to

researchers that report differences favoring children using spoken

language (Dillon & Pisoni, 2006) but similar to Geers (2003)) who found

no effect of communication modality on reading skills for children with

CIs.

Sign-only children were delayed in all three reading skills and

showed greater delays than spoken and bimodal children. These

children’s difficulty with nonword reading would be expected given

their lack of access to spoken language. The fact

Assessment Issues

Every researcher is faced with the issue of how to effectively and

ethically assess DHH children on tests that are not specifically designed

or normed for them. We chose to administer mostly standardized tests

of language and reading that were well validated and widely used with

hearing children because these provided us with interpretable data to

examine growth over a school year and across grades. We purposefully

did not administer every test to each child but assessed children on the

tests that had items that were accessible to them and that they had the

opportunity to acquire based on their language

Downloaded from https://academic.oup.com/jdsde/advance-article-abstract/doi/10.1093/deafed/enz050/5734969 by University of Exeter user on 04 May 2020

and communication environment. We judged it unethical to take time

away from classroom instruction for assessments that were clearly

invalid for specific groups of children. Thus, we did not assess the signonly

children on the spoken phonological awareness blending subtest

because it required spoken responses. We assessed bimodal children

on phonological awareness, because the ESP data indicated that they

had auditory access to spoken English even if they were in a signing

educational environment. We did not assess spoken-only children on

ASL receptive syntax or fingerspelled phonological awareness, because

they were not exposed to and therefore could not acquire sign or

fingerspelling.

Modifying standardized tests raises questions about whether

standardized scores and test norms are appropriate to use. On the

other hand, administering standardized tests in one invariant manner

to all DHH children raises questions about assessment validity. We

believe that making some reasonable modifications to allow the range

of DHH children access to the test items and ensuring that these

modifications were consistently used ensured, as much as possible,

that the test scores we obtained were valid (Sechrest, 2005). In order to

ensure that assessments were valid for each group of children, we

made standard adaptions to each test to accommodate the different

language and communication access and environments, but we

ensured that these adaptations were not idiosyncratic or dependent on

the skill of individual examiners. We ensured that all examiners were

trained to administer each test and took extreme care with scoring.

Despite the variability in children’s language and communication

access, we obtained high interrater reliability for all tests for all groups.

Despite these precautions, it is possible that some assessment results,

such as spoken phonological blending, were inf luenced by children’s

expressive speech capabilities. However, we tested many of these

validity questions for the fall time point (Lederberg et al., 2019) and

found that the tests had good psychometric properties for their

intended constructs and had no strong evidence of bias across groups.

standing both the specific language skills and the level of proficiency of

these skills necessary to improve outcomes.

Implications for Instruction

The documentation of language and reading status and progress

indicates some areas in which we can focus instructional practice. In

the area of language, although vocabulary is a relative strength, the

majority of children have lexicons that are considerably smaller than

their typical hearing peers. English syntax is a notable weakness even

for children with access to spoken language. Systematic and intensive

instruction in each of these areas is probably beneficial for all children.

On the other hand, children with access to sign in the classroom

appear to be making progress in ASL syntax, indicating that, when the

structure of language is visual and therefore completely accessible,

current instructional practices may be sufficient to ensure acquisition.

Classroom instruction and teacher talk can make a difference in

children’s language learning. Duncan and Lederberg (2018) examined

teacher talk in classrooms and reported that explicit vocabulary

instruction and teachers’ reformulation of children’s utterances were

predictors of gains in children’s vocabulary and expressive English

syntax. These authors also reported that teachers were highly variable

in the amount of explicit vocabulary instruction in which they engaged

and that explicit morphosyntax instruction occurred rarely. Given DHH

children’s delays in vocabulary and English morpho-syntax, it might be

necessary for teachers to engage in explicit direct instruction and for

researchers and practitioners to collaborate on developing and

evaluating instructional practices in these areas.

In the area of reading, all children acquired the simple skills, such

as letter naming and matching rebuses to pictures that are required at

the kindergarten level, but had difficulty with word reading and

comprehension that are required by the second grade. According to

the SVR, reading comprehension requires both f luent decoding and

linguistic comprehension (Gough & Tunmer, 1986;Tunmer & Chapman,

2012). The challenge, as always, is identifying the kind of instruction

that will help DHH students make sufficient progress in both these

areas. Ongoing research by the authors documenting reading

instruction in classrooms should assist in identifying the interaction

between instruction and children’s progress in reading. Such

knowledge should help determine effective teaching strategies and the

amount of time children need to be engaged in different types of

reading instruction.

Implications for further research

study was that the bimodal children in our sample, a group seldom

studied, are learning both English and ASL and making progress in

both languages. However, this finding needs to be replicated with

other samples of bimodal children. We need to also address how

development, and perhaps instruction, in both languages inf luences

reading development. For example, it would be informative to study

exactly how bimodal children use spoken and sign language while

reading. It would also be informative to study whether exposure to

English-like sign facilitates reading. Another area that needs further

study is the inf luence of fingerspelling on reading decoding for DHH

children. Our previous research showed that reading was strongly

related to fingerspelling (Lederberg et al., 2019) and there is some

evidence that signing DHH children benefit from connecting

fingerspelled words to print (Haptonstall-Nykaza & Schick, 2007).

However, evidence is needed on the role of systematic instruction in

this area.

We also need to explore further why children made comparatively

little progress in reading comprehension. Even spokenonly children

who showed close to age-appropriate language on word reading and

nonword reading showed delays in reading comprehension by the end

of second grade. One explanation might be that our sample displayed

weak English syntax skills; the SVR suggests that both decoding and

language are equally important to comprehension. Detailed analyses

of reading comprehension profiles (Kyle & Cain, 2015) hold promise for

under-

An important finding of this

Limitations

A limitation of this study is that we did not include a comparison group

of typically hearing children and thus were only able to compare the

DHH children’s scores to published test norms. Because our sample is

also not matched to the test sample, a comparison group matched on

variables such as socioeconomic status or ethnicity would have

provided more precise information on overlap, progress, and specific

areas of delay in language and reading. Technically, standard scores

are also not on a longitudinally consistent metric. However, the effects

of standardization are likely to be small for the current shortterm

longitudinal design. Our sample also does not represent the larger

population of DHH children; rather, we focused on children in selfcontained

programs. These children are likely to be those most in need

of intensive instruction, but our results should not be generalized to all

DHH children. Another limitation is that we only received self-reports

from teachers about

Downloaded from https://academic.oup.com/jdsde/advance-article-abstract/doi/10.1093/deafed/enz050/5734969 by University of Exeter user on 04 May 2020

their use of ASL or signed English. We were not able to obtain

observational or other data to confirm the kind of signing occurring in

classrooms.

Although we examined differences due to time, grade, and group

on several outcomes, there are many other malleable variables that inf

luence individual children’s language and reading. Other researchers,

for example, have documented the effects of age of implantation, age

of diagnosis and intervention, and the quality of early intervention (

Dunn et al., 2013;Geers et al., 2009;Kyle & Harris, 2010;Moeller, 2000).

For any single child, variables impacting language and reading will also

include those related to the home (e.g., parental access to professional

support, home literacy environments) and to school (e.g., quality and

consistency of classroom instruction). Given the low incidence of

childhood deafness, and the difficulty of examining large samples, it is

likely that no single study can examine simultaneously all possible

variables. It is important that readers keep in mind that the variables

examined in this study are only some of the many variables that will

impact children’s language and reading.

Unlike some researchers, we included children whom teachers

reported as having mild disabilities. Because it is possible that the

presence of these disabilities may have inf luenced the results, we ran

post hoc analyses to examine the effects of disability. Children with

mild disabilities scored, on average, lower than children without

disabilities. While there were main effects, disability status did not

interact with time, grade, or group. Excluding children with mild

disabilities changed minimally the overall means and analyses reported

in this paper.

Conclusion

The positive news is that the DHH children in our sample made

progress in vocabulary learning and were acquiring the syntax of either

English (spoken-only and bimodal children) or ASL (sign-only and

bimodal children). Thus, despite delays, children were acquiring

language to which they had access in their environment. They were

also developing phonological processing skills either auditorily or

visually. However, for all groups, delays in reading comprehension

increased in higher grades. Future research might clarify how and why

classroom instruction inf luences these outcomes. The challenge, as

always, will be to develop evidence-based instructional strategies that

will lead to improvement in literacy for all children while taking

advantage of their opportunities for visual and auditory access to

language.