125 children (18–34 months of age) with Down syndrome were selected for this study. The following criteria were necessary: (1) a diagnosis of DS as seen on medical records; (2) age between 18 and 34 months; (3) vision and hearing normal as reported medical records; (4) expressive vocabulary of at least one verbal word and fewer than 10 spoken words or manual signs as determined by parent report on the Assessment, Evaluation, and Programming System for Infants and Children Family Report (Bricker, 1993); (5) ability to imitate hand movements as assessed using the Body Imitations section of the Motor Imitation Scale (Stone, Ousley, & Littleford, 1997); (6) Spanish and English as the two languages spoken in the child’s home; and (7) parental consent for the child’s participation in the study with written agreement that their child would not any language intervention outside of the current study.
All 125 participants were recruited from a university-based demonstration focusing on training parents to use The Hanen Project – It Takes Two to Talk (Manolson, 1992); a program specifically designed for was specifically designed for parents of children with language delays. Each of the parents who were recruited had signed up for this program at various universities in North Carolina. The recruitment of participants occurred in the following participating universities: Appalachian State University, The University of North Carolina at Chapel Hill, North Carolina Central University, and East Carolina University. Children participated in this study before their parents began the Hanen Project. Appalachian State University’s Institutional Review Board approved all study procedures. Participants’ language and cognitive abilities were assessed through use of the Preschool Language Scale – 4 (PLS–4; Zimmerman, Steiner, & Pond, 2002). The PLS–4 was completed before the baseline session was given. The test was also conducted after completion of intervention to note progress.
34 bilingual speech-language pathologists from each of the aforementioned universities studying as doctoral students in the field of communication sciences and disorders volunteered to provide speech therapy to the subjects. Seven speech-language pathologists with specializations in language intervention with >10 years’ experience as an early childhood educators in the field oversaw the procedures for the sessions. All the speech pathologists in the study were fluent in both English and Spanish and had used signs as an intervention technique prior to conducting this study, and also had proficient signing capabilities as tested by The American Sign Language Proficiency Interview (ASLPI).
The participants of this study were well represented across gender, race, and socioeconomic status as they pertain to this specific population. Table 1 contains participant information:
The Preschool Language Scale – 4 was the assessment test used at the onset of this study. The PLS-4 is regarded as an extremely reliable and valid formal assessment for language for children up to 6 years of age. The PLS-4 was estimated using test-retest reliability, and the data for the test shows that it is dependable and stable across repeated administrations over time. The test also has very strong internal consistency, as tasks within the test are cohesive and homogenous through scoring. Additionally, the test was created using exhaustive literature reviews and user surveys that addressed what the PLS-4 should test, thus giving it strong evidence of external validity. During the baseline session as well as the intervention sessions, a wide variety of age-appropriate pretend-play toys were used. Examples of toys used were bouncy balls; foam balls; puzzles; cars, car ramp, and car wash toy; pretend car wash; farm animal set with farm; baby dolls and various baby doll accessories; cardboard blocks; Lego blocks; a zoo animal set with various zoo animals; multiple bilingual board books; three pop-up books; rainbow stacking rings; kitchen play items; and pretend food.
The study conducted was a between participants design. Following recruitment and pre-testing with the PLS-4, children were assigned to a treatment condition (i.e., bilingual therapy intervention strategies with gestural support; bilingual therapy intervention strategies without gestural support). Children were grouped according to university clinic site, because treatments were conducted at the university clinic sites. Group-1 contained 62 children and Group-2 contained 63 children. To ensure external validity, children were assigned randomly to each group. Each group was then randomly assigned to a treatment condition. University clinic sites were also chosen at random. Group-1 participants received therapy intervention from clinicians at Appalachian State University and University of North Carolina at Chapel Hill. Group-2 participants received therapy at North Carolina Central University and East Carolina University. All baseline and intervention sessions were video recorded and observed by at least two interventionists for cohesive and reliable data collection.
For the baseline session, the interventionist introduced a variety of fun and interesting toys and engaged in play with toys selected by the child. During the baseline session, the interventionist provided the first half of the 50-minute session in Spanish, and the second half in English. He/she did not demonstrate models of any signs. When the child initiated communication, the interventionist responded by describing her own play actions (self-talk) and those of the child (parallel talk). If the child did not initiate communication, the interventionist asked questions about toys, directed the child to complete a play act, or commented on the focus of the child’s engagement.
Intervention sessions occurred three times per week for 12 weeks and lasted 30 minutes each session, on the basis of child interest and engagement in the interaction. If the child lost interest in participating in the session, the session would discontinue and be made up within the following two weeks. This method was used to ensure that each session was conducted with the child’s full attention. The interventionists provided therapy using the Enhanced Milieu Teaching approach. The EMT strategies implemented by the interventionist included (a) following the child’s lead; (b) responding to the child’s communication; (c) describing shared actions with target words and signs; (d) expanding the child’s communication; (e) modeling target signs; (f) using time delay to promote child communication; and (g) incorporating milieu prompting episodes to promote child practice of communication targets (Kasier and Wright, 2013).
For Group-1, the interventionists engaged in (EMT) therapy strategies to teach 40 signs paired with spoken words, both in English and Spanish. The interventionist provided a gestural sign for at least 75% of her verbal communication with the child in order to model sign-infused verbal communication. The signs that were modeled were selected from the list of early occurring words as listed by the MacArthur Communicative Development Inventory (MCDI) and were appropriate for the toys and the actions with toys occurring during the play interactions (Fenson, 2003). The 40 signs included 20 words for objects, 15 words for actions, and 5 request words. The first half of the Group-1’s sessions was spoken in Spanish, and the second half were spoken in English. The interventionist introduced the same gestural signs for both the English and Spanish words in American Sign Language.
For Group-2, the interventionists provided the same EMT therapy strategies as Group-1 received, both in English and Spanish, but without the implementation of introducing of gestural models. The first half of the Group-2’s sessions was spoken in Spanish, and the second half were spoken in English
All intervention sessions video-recorded with audio included. To account for internal validity, post – treatment assessments were coded and scored by research assistants (speech-language pathologists) who were blinded to the group’s treatment conditions. The scoring was done through transcriptions using Type-Token Ratio Scores. To ensure external reliability, this study implemented inter-rater reliability for coding the final assessments. Observable behavior categories were specified in subjective terms, thus creating strong external reliability. Research assistants were instructed on how to code for each specific circumstance they would observe in the final assessment. For sign coding, the sign was coded as spontaneous if child used the sign independent of a model that had occurred in the previous 3 seconds. Signs were coded as imitated if the child used the sign 5 seconds of the therapist’s model. A prompted sign was coded if the child responded to an interventionist’s questions or if the child responded to a direct prompt from the interventionist (i.e., ‘‘Say hat.’’). These same specifications applied for verbal speech in the same categories. If a child used a sign and a verbal word at the same time, both the word and sign were recorded as separate units.
The study was completed in seven months. The study began with the recruitment of speech pathologists from various universities willing to participate as therapists in a research study. This process of recruiting speech pathologists spanned for about two months, from March 15, 2015 until May 10, 2015. Following this recruitment, participants were recruited from sign-up lists for the Hanen Project across various universities. Participants were contacted within one week of the sign up date. After recruitment was finished and consent were received from participants, the initial formal testing was conducted, spanning over 2 weeks to finish all assessments. Once the initial assessments were completed, therapy for both groups was initiated the following Monday, July 15, 2015. Sessions were held for the following 12 weeks. On the 13th week, the final session was coded and the final formal assessment was completed.