TECSE 23:3 124–136 (2003) 124

Early Parent–Child Interactions and

Early Literacy Development

T
he relationship between parent–child interactions and early literacy skills for
27 families living in low-income households was examined. Parent–child inter-
Beverly J. Dodici actions in “simulated” daily experiences were videotaped when the children were
14, 24, and 36 months old. These tapes were coded with the Parent-Infant/Toddler
Dianne C. Draper Interaction Coding System (PICS; Dodici & Draper, 2001), a scale that rates child lan-
and guage, parent language, emotional tone, joint attention, parental guidance, and parental
Carla A. Peterson responsivity. These behaviors were related strongly to early literacy skills in the areas
Iowa State University of receptive vocabulary, symbolic representation, and phonemic analysis, which were
measured in the spring prior to kindergarten entry. In addition, observed parent–child
interactions were better predictors of early literacy skills than was a parent report of
home literacy experiences.

A sociocultural perspective of child development empha- sociated with the child’s emerging competencies in social,
sizes that infants, toddlers, and young children learn many cognitive, and linguistic domains throughout early and
skills through adult–child interactions. Vygotsky (1986) middle childhood” (p. 1399). Scientists have demonstrated
posited that social interactions with an adult within the consistently that these interactions related to positive lan-
cultural contexts of society are fundamental to cognitive guage and cognitive development contain elements of joint
development. Building on Vygotsky’s theories, Rogoff attention (Tomasello & Farrar, 1986), parental responsiv-
(1990) proposed the idea that skill development re- ity (Bornstein & Tamis-LeMonda, 1989; Landry, Smith,
quires the interaction of two parties, a teacher (typically, Swank, Assel, & Vellet, 2001), positive emotional tone
an adult) and a learner (typically, a child). She called (Estrada, Arsenio, Hess, & Holloway, 1987; NICHD,
this an “apprenticeship-type relationship” that involved 1999), and appropriate parental guidance or scaffolding
“guided participation” occurring during daily experiences. (Adamson & Bakeman, 1984). Furthermore, the amount
The central tenet thus is that children learn via daily in- of talk and the guidance style that parents use with their
teractions with their parents and other adults during play, children have also been related to later language and
teaching, and routine situations—some of which are frus- cognitive development (Hart & Risley, 1995; Walker,
trating. This work has laid the theoretical groundwork Greenwood, Hart, & Carta, 1994). Specific characteris-
for empirical examination of how specific characteristics tics of these interactions, which are discussed next, are
of parent–child interactions influence skill development. important contributors to young children’s skill develop-
ment.

RELATIONSHIP OF PARENT–CHILD
Responsivity/Sensitivity
INTERACTIONS TO LATER SKILL DEVELOPMENT
Although defined slightly differently across studies, both
Researchers funded by the National Institute of Child responsivity and sensitivity have been related consis-
Health and Human Development (NICHD, Early Child tently to positive child outcomes. In general, researchers
Care Research Network, 1999) have stated that “support- have shown that higher levels of maternal responsivity
ive, warm, and engaged parent-child interactions are as- and sensitivity are associated with better outcomes in the

Address: Beverly J. Dodici, 2309 Merritt Drive, Northfield, NJ 08225; e-mail: bjdodici@yahoo.com
Early Parent–Child Interactions and Early Literacy 125

areas of social, cognitive, and language skills (Barnard, onstrated that joint attention plays a significant role in
1997; Lamb-Parker, Boak, Griffin, Ripple, & Peay, 1999; language and skill development (Harris, Jones, Brookes,
Landry, Smith, Miller-Loncar, & Swank, 1997; Landry et & Grant, 1986; Landry et al., 1997; Saxon, 1997;
al., 2001). Maternal responsivity is defined as a mother’s Tomasello & Farrar, 1986; Tomasello & Todd, 1983). In
prompt, contingent, and appropriate behaviors in re- general, these researchers have revealed that children are
sponse to a child’s actions (Bornstein & Tamis-LeMonda, more likely to acquire novel words and skills when their
1989). Bornstein and Tamis-LeMonda’s cross-cultural parents demonstrate joint attention than when children
study, completed with families from the United States change their attention to a different, adult-selected object
and Japan, revealed that maternal responsivity toward 4- not currently in their focus.
to 5-month-old infants was related to children’s compe-
tencies at age 4 years.
Parental Talk
Parental sensitivity is the degree to which parents
adapt to children’s needs and abilities (Beckwith & Rod- Children from low-income families are less likely to have
ning, 1996). Beckwith and Rodning observed maternal conversations with adults and are exposed to fewer words
sensitivity toward preterm infants during parent–infant than children from families with higher socioeconomic
interactions in a laboratory setting when the infants were backgrounds (Hart & Risley, 1995; Snow, Tabors, &
13 and 20 months old to examine the relationship be- Dickinson, 2001). An intensive review of the literature
tween these interactions and later social, language, and revealed a strong, consistent, and positive relationship
cognitive development. They found statistically signifi- between the amounts of time a child talked with a par-
cant correlations between later child language and social ent or adult and later literacy skills (Snow, Burns, & Grif-
skills and maternal sensitivity, child engagement, and fin, 1998). Furthermore, Hart and Risley (1995) found
dyadic fit. These authors demonstrated that a relation- that the number of words parents said to their children
ship existed between parental responsiveness and later per hour was related to the children’s language and cog-
language and social skills; however, measures of early lit- nitive skills at age 3 years and expressive language at age
eracy skills were not included in their study. 9 years. Walker and her colleagues (1994) found that chil-
dren’s vocabulary at age 3 years predicted their school
Emotional Tone achievement levels (reading and spelling) in kindergarten
through third grade. Other researchers have linked the
The emotional tone, or affective aspect, of parent–child
amount of time children engage in conversations with
interactions has also been found to be related to child de-
adults to later skill development (DeTemple, 2001; Rush,
velopment (Barnard, 1997; Estrada et al., 1987; Hart &
1999).
Risley, 1995; Lamb-Parker et al., 1999; Pianta & Ege-
land, 1994). Positive statements, comments, praise; smiles
and laughter; nurturing embraces or touches; and limited Parental Guidance
negative comments or yelling are some of the parental
Parental guidance style has been defined as the relative
behaviors that have been related to positive child outcomes
amount of prompting that a child experiences or how
(Barnard, 1997). Researchers have shown that the level
often the child is asked—rather than told—what to do
of positive affect present in parent–child interactions is
(Hart & Risley, 1995). In parent–child interactions, this
influenced by a family’s socioeconomic status as well. In
may include various strategies in which the parent pro-
their longitudinal study, Hart and Risley (1995) found
vides more control and structure through increased in-
that children living in low-income households heard twice
formation and less choice (Landry et al., 1997; Landry et
as many prohibitions as affirmative statements from their
al., 2001), depending on the child’s skill levels. The de-
parents compared to children living in middle- and upper-
gree of guidance is based on the utterances parents use
income households. These negative comments made by
and can be categorized by the responses their utterances
parents to their children in their first years of life appear
prompt. Directive statements demand prompt action (e.g.,
to have had lasting effects because they were related neg-
“Get your coat on”), whereas questions or suggestive
atively to the children’s later cognitive and language abili-
statements ask for a response (e.g., “Can you get your
ties.
coat on?”). A third type of utterance used by parents, in-
formative statements (e.g., “It’s cold outside; do you
Engagement think you’ll need your coat?”), give limited or no direc-
Joint attention, the amount of time a parent and child are tion and allow a child to use the information as he or she
mutually focused on a single object or activity (Toma- chooses.
sello & Farrar, 1986), is a main element of parent–child Parental directiveness has been consistently found to
engagement. A good number of researchers have dem- inhibit a child’s vocabulary (Landry et al., 1997; Toma-
126 Topics in Early Childhood Special Education 23:3

sello & Todd, 1983) and would seem to relate to parental the differences are evident as early as kindergarten (Dick-
guidance style. In most instances, however, parental di- inson & Snow, 1987). In the 1960s, the United States be-
rectiveness referred to times during which the parent was gan nationwide intervention programs in an effort to
redirecting the child’s activities versus following the child’s help preschool children from low-income households have
lead. an “even start” when they entered school. It has been
To date, most researchers have evaluated the rela- suggested, however, that interventions implemented after
tionship between parent–child interactions beginning at children are 3 years old, especially for children living in
6 months of age and early developmental skills (e.g., 24- low-income households, may have a limited impact on
month language skills). These researchers have usually later cognitive or other developmental skills due to the
found relatively short-term developmental impacts (Akh- cumulative effects of experiences during the first 3 years
tar, Dunham, & Dunham, et al., 1991; Pappas-Jones & of life (Hart & Risley, 1995).
Adamson, 1987; Saxon, 1997; Tomasello & Farrar, 1986). A growing awareness that skill development during
Other researchers have focused on parent–child interac- the preschool years is related strongly to later academic
tions and later language and cognitive development, but success has brought increased attention to the under-
these studies have generally included samples of children standing and measurement of early predictors of school
over the age of 3 years (Dickinson & Tabors, 2001; Es- achievement (Adams, 1990; Snow et al., 1998; Walker et
trada et al., 1987). Still other researchers have focused on al., 1994). In addition, the level of early academic skills
specific populations, such as preterm infants (Beckwith has been found to predict later academic success or fail-
& Rodning, 1996), or on specific developmental periods ure (Adams, 1990; Cunningham & Stanovich, 1997; Juel,
(e.g., between the ages of 6 months and 40 months); few 1988; Slavin, Karweit, Wasik, Madden, & Dolan, 1994).
have examined later language development (Landry et al., The levels of literacy and language skills that children
1997). Subsequent work by Landry and her colleagues have at kindergarten and first grade strongly predict
provides evidence that not only early maternal respon- school achievement and even completion of high school
siveness but also consistent responsiveness throughout (Cunningham & Stanovich, 1997). Several specific skills
infancy and the preschool years is related to more opti- acquired by many children during the preschool years
mal cognitive and social development at age 54 months, have been found to facilitate literacy development during
especially for preterm infants (Landry et al., 2001). the early elementary school years. These skills include
Another research team focused on the long-term letter naming (Wagner, Torgesen, & Rashotte, 1994), pho-
effects of parent–infant/toddler interactions, beginning nological awareness (Adams, 1990; Snow et al., 1998),
when the children were 7 months to 36 months of age and vocabulary development (Adams, 1990; Dickinson,
and related these skills to language and cognitive skills Cote, & Smith, 1993; Hart & Risley 1995; Walker et al.,
demonstrated in early elementary school (Hart & Risley, 1994).
1995; Walker et al., 1994). However, these researchers The relationship between early academic skills and
focused on how early child language skills predicted read- later school performance, and the academic risks that
ing outcomes and academic success in early elementary children from low-income families face, provide a strong
school (Walker et al., 1994). They did not examine the rationale for effective early literacy activities and inter-
full range of parent–child interaction variables that may ventions designed to enhance overall academic success.
have influenced later outcomes, nor did they measure Given this point, efforts to identify key components of
intermediate outcomes at the early elementary grades early home activities and/or intervention experiences
(Hart & Risley, 1995). Taken together, their findings em- that yield desirable results will be important.
phasize the point that much still needs to be learned about One variable to consider when evaluating factors
how parent–infant/toddler interactions are related to early that influence children’s early literacy skills is the area
literacy skills. of literacy experiences within their homes. Ethnographic
studies have linked early home experiences with later
skill development (Purcell-Gates, 1996; Purcell-Gates &
PARENT–CHILD INTERACTIONS Dahl, 1991); however, these types of ethnographic stud-
IN LOW-SES FAMILIES ies are costly, typically involve a small sample, and may
have limited generalizability. Studies utilizing less inva-
Children who live in low-income households are at risk sive, less costly objective measures (Whitehurst, 1993) have
for (a) consistently scoring lower on measures of intelli- identified links between home literacy experiences and
gence and (b) experiencing academic difficulties (Same- language skills for children from low-income households
roff, Seifer, Baldwin, & Baldwin, 1993). These children (Payne, Whitehurst, & Angell, 1994; Rush, 1999). It
are also more likely to have difficulty learning to read, should be noted that these relationships were demon-
compared to children in middle-income households, and strated with children approximately 54 months old, an
Early Parent–Child Interactions and Early Literacy 127

age after which timely opportunities to intervene in early age. A similar follow-up assessment was completed the
parent–child interactions have passed and possible foun- spring or summer prior to the children’s age eligibility for
dations for early literacy skills have been established. kindergarten. Twenty-seven families met the criteria for
Therefore, broadening understanding of the relation- this study; 13 (48%) of the children who were assessed
ships between early parent–child interactions and early were boys. Maternal age ranged from 15 years to 34 years
literacy skills needs to be done. at the time of the child’s birth, with a median age of 23.5
Overall, researchers have established links between years (SD = 5.03). Education levels of the participants
early parent–child interactions and early language skills, ranged from completing: 5th to 8th grades (3%), com-
as well as between language skills and academic skills. pleting 11th grade (3%), completing high school or the
The connections between parent–child interactions and equivalent (44%), attending some college but no degree
early literacy skills have not been well established, how- (41%), and obtaining a bachelor’s degree (7%). All of the
ever. The current study thus was undertaken to examine participants were Caucasian.
the relationships between early parent–child interactions
and early literacy skills demonstrated at the end of the pre-
Measures
school period. The study addressed the following three
specific research questions: Direct measures of child and parent functioning were
gathered when the children were 14, 24, 36, and approx-
1. What is the relationship between parent– imately 54 months of age. Parent–infant/toddler interac-
infant/toddler interactions and early liter- tions were videotaped during a session structured to elicit
acy skills for children from low-income teaching, play, and frustration behaviors. These tapes
families? were later coded by research assistants using the Parent-
2. Do parent–infant/toddler interactions, as Infant/Toddler Interaction Coding System (PICS; Dodici
measured at 14, 24, and 36 months, or a & Draper, 2001). All predictor and outcome variables, as
parent report about home literacy experi- well as covariates and the times at which these measure
ences, taken at approximately 54 months were gathered, are presented in Table 1, with brief de-
of age, better predict early literacy skills? scriptions provided in the text.
3. Do parent–infant/toddler interactions from
one simulated activity predict early literacy Predictor Variables. The PICS was used to rate six
skills as well as parent–infant/toddler in- items relevant to the quality of parent–child interactions:
teractions from the combination of three
different types of simulated activities? 1. infant/toddler language (the amount of
age-appropriate language demonstrated by
the infant/toddler);
METHOD 2. parent language (the amount of develop-
mentally appropriate language the parent
This study was part of a larger longitudinal study con- used with the infant/toddler);
ducted by a midwestern university, Mathmatica Policy Re- 3. emotional tone (positive or negative,
search, Inc., and the Early Head Start National Research including verbal comments);
Consortium. The primary purpose of the larger study was 4. joint attention (the amount of time the
to investigate the long-term effects of Early Head Start parent and infant/toddler were looking
programs. The focus of the current study was investigat- at/interacting with the same object);
ing the relationship between parent–infant/toddler inter- 5. parental guidance (ratio of informative
actions and early literacy skills. vs. directive statements used by the parent
when interacting with the infant/toddler);
and
Participants 6. parental responsiveness (degree to which
Participants in this study were families who (a) had a child the parent responded to the child’s cues).
born between September 1995 and September 1996,
(b) were income-eligible for Early Head Start (EHS) ser- Each item was rated on a 5-point scale (higher scores rep-
vices at that time, and (c) had enrolled in the EHS Na- resented better quality). These six items were used to rate
tional Evaluation study within a year after their child videotaped parent–infant/toddler interactions during sim-
had been born. All participants in this study completed ulated play, teaching, and frustration activities when the
annual child assessments and parent interviews when their child was 14, 24, and 36 months of age. The ratings were
children were approximately 14, 24, and 36 months of then summed or averaged to create various PICS scores.
128 Topics in Early Childhood Special Education 23:3

TABLE 1. Measures Used in Study

Instrument Purpose Age at testing

Predictor variables
Parent-Infant/Toddler Interaction Coding System Evaluated quality of parent–infant/ 14, 24, 36 months
(PICS; Dodici & Draper, 2001) toddler interactions
Stony Brook Family Reading Survey Evaluated children’s home Spring or summer prior
(SFRS; Whitehurst, 1993) literacy experiences to kindergarten

Outcome variables
Peabody Picture Vocabulary Test–Third Edition Assessed child’s receptive Spring or summer prior
(PPVT-III; Dunn & Dunn, 1997) vocabulary skills to kindergarten
Woodcock-Johnson Tests of Achievement–Revised Assessed child’s prereading skills Spring or summer prior
(WJ-R, Letter-Word Identification; Woodcock & to kindergarten
Johnson, 1989)
Test of Language Development–Primary: Assessed children’s phonemic Spring or summer prior
Third Edition (TOLD-P:3, Phonemic Analysis awareness skills to kindergarten
subtest; Newcomer & Hammill, 1997)

Covariates
MacArthur Communicative Development Assessed children’s expressive Parent interview at
Inventories-II (CDI; Fenson et al., 1993) vocabulary skills 24-month assessment
WJ-R (Picture Vocabulary subtest) Assessed parent’s vocabulary skills Parent interview at
24-month assessment

Four measures were derived from the PICS rating Outcome Variables. The Peabody Picture Vocabu-
system: Item scores, Age-Related scores, Activity scores, lary Test–Third Edition (PPVT-III; Dunn & Dunn, 1997)
and the Overall PICS score. Item scores were simply the presents four black-and-white pictures on an easel paired
rating given to each of the six items (child language, par- with one verbally presented target word. Children are
ent language, emotional tone, joint attention, parental required to point to the target picture in order to assess
guidance, and parental responsivity) observed during the their receptive language skills. The Woodcock-Johnson
simulated activities. Age-Related scores were created by Tests of Achievement–Revised (WJ-R; Woodcock & John-
summing the Item scores at each of the three age points, son, 1989) Letter-Word Identification subtest initially pre-
resulting in 14-, 24-, and 36-month PICS scores. The sents rebus figures with pictures and requires a child to
Age-Related scores from each of the three different sim- match these items, to indicate a beginning understanding
ulated types of parent–children activities were then aver- of symbolic representation. This subtest moves into ba-
aged to create Activity scores for the play, teaching, and sic letter and word identification if these skills are in-
frustration activities. These Activity scores then were av- tact. For the Test of Language Development–Primary:
eraged to provide a summary score for each child, re- Third Edition (TOLD-P:3; Newcomer & Hammill, 1997)
sulting in one Overall PICS score. Averages were used to Phonemic Analysis subtest, a child is verbally presented
create the Activity and Overall PICS scores in an effort with compound words and then asked to say the target
to include all possible videotapes but not penalize for word without the first or last word in an effort to assess
missing data, because it was not possible to code two his or her ability to segment words. The scores for these
tapes due to audio problems. The maximum score for the measures were reported as standard scores: PPVT-III and
play, teaching, frustration, or Overall PICS scores was 30. WJ-R subtest—M = 100, SD = 10; TOLD-P:3 subtest—
The Stony Brook Family Reading Survey (SFRS; M = 10, SD = 3.
Whitehurst, 1993) is a 52-item interview that asks par-
ents to answer nine questions related to their children’s Covariates. Parental vocabulary skills, as measured
reading interests and habits, as well as their own reading by the Picture Vocabulary subtest of the WJ-R, and chil-
habits. An adaptation of the SFRS, which includes the dren’s vocabulary at 24 months of age, measured using
nine home literacy environment questions, had been used the MacArthur Communicative Development Inventories–
previously to measure family literacy activities (Payne et Second Edition (CDI; Fenson et al., 1993), were used as
al., 1994; Rush, 1999) and was used for this study. The covariates in this study. The Picture Vocabulary subtest
maximum SFRS score was 36 in this study. required parents to name pictures; the CDI asked parents
Early Parent–Child Interactions and Early Literacy 129

to indicate if their child used the words from a list of simulate frustrating situations commonly encountered dur-
100 words. The number of correct responses was the re- ing daily activities. There was no frustration task at the
ported score and was used in this study for the WJ-R Pic- 36-month assessment. Instead, the frustration activity was
ture Vocabulary (maximum score = 50) and the CDI replaced with an activity selected by the parent and child.
(maximum score = 100) scales. Maternal age at the time For coding purposes, the first 3 minutes of each frustra-
of enrollment was also included as a covariate. tion activity were rated.
Seven research assistants were trained using practice
tapes, and they attained interrater reliability with the
Data Collection
authors prior to coding any tapes independently. Inter-
Data were collected from each participating family in their rater reliability levels for the PICS were targeted for 85%
home around the children’s 14-, 24-, and 36-month birth- for each item, within 1 point. The 85% reliability level
days, as well as in the spring or summer prior to the fall for this study was computed using ratings for individual
the children were age-eligible for kindergarten. All data items coded at each age level. Because the PICS was used
were collected as part of the EHS National Evaluation to code five interaction activities (three play activities, one
Study and were made available to the first author. The frustration activity, and one teaching activity) at each age
standardized assessments were completed as per proto- level, and 6 items were rated for each, there were 30 items
col at the times listed in Table 1. to be rated at each age level. Therefore, if the two coders
The PICS was used to code the videotapes of parent– failed to rate at least 26 items within at least 1 point,
child interaction activities. As noted previously, these those interaction activities were reviewed and rescored
videotaped interactions consisted of three separate activ- by the entire research team until reliability above 85%
ities: a semistructured play activity, a teaching activity, agreement was met.
and a frustration activity. For the teaching activity, moth- Interrater agreement estimates were calculated for 23
ers were given the choice of teaching their 14-month-old (28%) randomly selected tapes coded for this study. Five
infants or toddlers to stack blocks or point to body parts of the 23 tapes did not meet the 85% reliability target;
in a book, and they were given the choice of having their therefore, the entire research team recoded these tapes,
24-month-old children sort blocks by color or point to discussed disagreements, and reached consensus on an
clothing articles in a book. At the 36-month assessment, item-by-item basis until 85% agreement within at least
the children were given an age-appropriate puzzle, which 1 point on all items was met. Across all tapes, interrater
was followed by a more challenging puzzle after the first reliability was 88%, (exact agreement on 47% of the
puzzle was completed. These activities lasted for 4 minutes, items). Interrater reliability for each of the six predictor
but only the first 3 minutes of interaction were coded. items was as follows: child language–90% (40% exact),
For the play activity at each age level, the parents parent language–86% (38% exact), emotional tone–96%
and children were given three separate cloth bags con- (77% exact), joint attention–89% (54% exact), parental
taining toys. Each bag was marked with the numerals 1, guidance–90% (52% exact), and parental responsivity–
2, or 3. The parents were instructed to play with their 77% (27% exact).
children as they wished but to start with Bag 1, move to The spring or summer prior to the children being el-
Bag 2, and then move to Bag 3. No other instructions were igible for kindergarten, a follow-up assessment and an
given. The bags contained age-appropriate toys. This ac- interview were completed with each family. During this
tivity lasted for 10 minutes. For coding purposes, three follow-up assessment, the children completed various
separate 3-minute intervals were coded during this activ- subtests of standardized assessments, including the PPVT-
ity; the final minute of interaction was not coded. III, WJ-R, and TOLD-P:3. At this time, the parents an-
During the frustration activity, which occurred at the swered questions from the SFRS as well.
14- and 24-month assessments, a scenario was presented.
For the 14-month assessment, the mother was asked to
sit approximately 6 ft. to 8 ft. behind the infant and RESULTS
draw a picture of her family while the infant was secured
in a high chair. The mother was told that she could in- Descriptive statistics for all predictor, outcome, and co-
teract with the child in whatever manner she wanted, but variate variables are presented in Table 2. Findings related
she could not give the infant anything or take the infant to each research question are presented in this section.
out of the seat. During the 24-month assessment, a set of
keys or a toy was placed approximately 2 ft. in front of
the child and the parent. The parent was informed that
Correlating Early Literacy and PICS
the toddler should not touch the keys or toy, which was Research Question 1. What is the relationship between
the only limitation. No additional directions were given parent–infant/toddler interactions and early literacy skills
for either scenario. These two scenarios were selected to for children from low-income families? This question was
130 Topics in Early Childhood Special Education 23:3

TABLE 2. Descriptive Statistics: Means and Standard outcome variables were completed; the results are pre-
Deviations sented in Table 4. The 14-month PICS scores without
Variable M SD Range child language revealed no statistically significant corre-
lations, which was consistent with the initial findings
Predictor when child language was included. The 24-month PICS
Overall PICS 21.80 2.01 17.20–25.97 scores without child language revealed that correlations
dropped from statistically significant levels of .47 on the
14-month PICS 20.72 3.18 15.00–27.50
PPVT-III and .51 on the WJ-R when child language was
24-month PICS 22.42 2.64 15.00–28.00 included to .33 for both the PPVT-III and WJ-R when
36-month PICS 22.15 2.64 16.00–26.00 child language was removed. For the 36-month PICS
scores, statistically significant correlations between the
Frustration activity 19.00 3.11 13.67–24.50
PPVT-III and the TOLD-P:3 remained even after the
Teaching activity 20.93 2.50 13.43–26.50 child language component was removed. Finally, the Over-
Play activity 22.86 2.01 18.78–26.22 all PICS score without child language decreased slightly,
revealing statistically significant correlations between the
SFRS 17.88 4.71 10.00–30.00
PPVT-III and WJ-R and the Overall PICS at a .05 level
Outcomes
versus a .01 level.

WJ-R Letter-Word
identification 93.00 11.53 72.00–115.00 PICS Versus SFRS:
PPVT-III 99.69 11.06 72.00–126.00 Predicting Early Literacy Skills
TOLD-P:3 8.84 2.98 4.00–14.00 Research Question 2. Do parent–infant/toddler interac-
tions at 14, 24, and 36 months or a parent report (SFRS)
Covariates
regarding literacy habits completed when the child is ap-
Maternal age 23.90 5.88 16.00–38.00 proximately 54 months old better predict early literacy
Maternal language skills? was examined using regression. Results are pre-
skills 39.86 3.96 32.00–49.00 sented in Table 5.
CDI 60.13 25.06 9.00–98.00
Analyses of the predictive nature of the Overall PICS
score versus the SFRS for the PPVT-III and the WJ-R re-
Note. PICS = Parent-Infant/Toddler Interaction Coding System (Dodici &
vealed that the Overall PICS score added significant in-
Draper, 2001); SFRS = Stony Brook Family Reading Scale (Whitehurst, crements in the multiple correlation squared (R2) at the
1993); WJ-R = Woodcock-Johnson Tests of Achievement–Revised (Wood- point of entry for each test. The SFRS did not predict the
cock & Johnson, 1989); PPVT-III = Peabody Picture Vocabulary Test–
Third Edition (PPVT-III; Dunn & Dunn, 1997); TOLD-P:3 = Test of
PPVT-III or the WJ-R. Analyses of the predictive nature
Language Development–Primary: Third Edition (Newcomer & Hammill, of the Overall PICS score versus the SFRS for the TOLD-
1997); CDI = MacArthur Communicative Development Inventories-II P:3 yielded results that were not statistically significant.
(Fenson et al., 1993).

Predicting Early Literacy Skills From

examined using zero-order correlations. The Overall PICS
a Single Type of Parent–Child Activity
score, as well as the Age-related 14-, 24-, and 36-month Research Question 3. Do parent–infant/toddler interac-
PICS scores, were correlated with the outcome variables. tions from one simulated activity predict early literacy
The results are presented in Table 3. skills as well as parent–infant/toddler interactions from
These analyses revealed a statistically significant cor- the average of three simulated activities combined? This
relation between the PPVT-III and the 36-month PICS question was examined using regression analyses. When
scores, the Overall PICS score, and the 24-month PICS entered separately into regression equations, none of the
scores. Statistically significant correlations also were found Activity scores (play, teaching, or frustration) predicted
between (a) the WJ-R and the 24-month PICS scores, early literacy skills as measured by the PPVT-R, WJ-R, or
(b) the WJ-R and the Overall PICS score, and (c) the TOLD-P:3. Multicollinearity between the Activity scores
TOLD-P:3 and the 36-month PICS scores. most likely accounts for this; therefore, these results should
Because young children’s language may correlate with be interpreted with caution.
their early literacy skills and possibly inflate these corre- In an effort to evaluate the third research question in
lations, additional analyses were completed. Correlation an alternate manner, zero-order correlations were com-
matrices comparing the Age-Related and Overall PICS pleted for the play, teaching, or frustration Activity scores;
scores, without the component of child language, with Overall PICS score, and the outcome variables. This was
Early Parent–Child Interactions and Early Literacy 131

TABLE 3. Zero-Order Correlations for PICS Scores and Outcomes

Overall PICS 14-month PICS 24-month PICS 36-month PICS

Instrument M n M n M n M n

PPVT-III .58** 26 .23 26 .47* 26 .63** 25

WJ-R .50** 27 .27 27 .51** 25 .30 27
TOLD-P:3 .31 25 .05 25 .13 25 .48* 23

Note. PICS = Parent-Infant/Toddler Interaction Coding System (Dodici & Draper, 2001); PPVT-III = Peabody Picture Vocabulary Test–Third Edition
(Dunn & Dunn, 1997); WJ-R = Woodcock-Johnson Tests of Achievement–Revised (Woodcock & Johnson, 1989); TOLD-P:3 = Test of Language
Development–Primary: Third Edition (Newcomer & Hammill, 1997).
*p < .05. **p < .01.

TABLE 4. Zero-Order Correlations for Outcomes and PICS Scores Without Child Language

Overall PICS 14-month PICS 24-month PICS 36-month PICS

Instrument M n M n M n M n

PPVT-III .40* 26 .25 26 .33 26 .56** 25

WJ-R .40* 27 .29 27 .33 25 .26 27
TOLD-P:3 .10 25 -.00 25 -.04 25 .44* 23

Note. PICS = Parent-Infant/Toddler Interaction Coding System (Dodici & Draper, 2001); PPVT-III = Peabody Picture Vocabulary Test–Third Edition
(Dunn & Dunn, 1997); WJ-R = Woodcock-Johnson Tests of Achievement–Revised (Woodcock & Johnson, 1989); TOLD-P:3 = Test of Language
Development–Primary: Third Edition (Newcomer & Hammill, 1997).
*p < .05. **p < .01.

TABLE 5. Regression of PICS Total and SFRS Scores on Outcome Variables

Instrument/variable R R2 Unstandardized β F change p

PPVT-III
PICS (M) .57 .33 2.83 11.13 .01**
SFRS .58 .33 -0.22 0.23 .64
WJ-R
PICS (M) .49 .24 2.64 7.59 .03*
SFRS .49 .24 0.01 0.04 .84
TOLD-3
PICS (M) .28 .08 0.45 1.80 .16
SFRS .31 .10 0.11 0.53 .48

Note. PICS = Parent-Infant/Toddler Interaction Coding System (Dodici & Draper, 2001); SFRS = Stony Brook Family Reading Survey (Whitehurst, 1993);
PPVT-III = Peabody Picture Vocabulary Test–Third Edition (Dunn & Dunn, 1997); WJ-R = Woodcock-Johnson Tests of Achievement–Revised (Woodcock
& Johnson, 1989); TOLD-P:3 = Test of Language Development–Primary: Third Edition (Newcomer & Hammill, 1997).
*p < .05. **p < .01.

done in an effort to demonstrate which PICS score (play, well as the Overall PICS score. The Overall PICS score
teaching, frustration, or Overall PICS) was related most yielded the strongest relationship with the PPVT-III. Sta-
strongly to the outcome variables. The results are pre- tistically significant correlations were found for the WJ-
sented in Table 6. R and the teaching Activity score, the play Activity score,
Examination of the correlation matrix revealed sta- and the Overall PICS score. The play Activity score
tistically significant correlations between the PPVT-III yielded the strongest relationship with the WJ-R. The
and the play, teaching, and frustration Activity scores, as TOLD-P:3 correlated only modestly with the play, teach-
132 Topics in Early Childhood Special Education 23:3

ing, or frustration Activity scores and with the Overall (see Table 3), indicating that the children in this study per-
PICS score. formed similarly to the overall population on those tasks.
To determine whether including a rating of child’s
language in the PICS inflated the relationships between
DISCUSSION PICS scores and early literacy scores, additional analyses
were completed. The Activity and Overall PICS scores
The current study examined the relationship between without the child language component and outcome var-
parent–infant/toddler interactions and early literacy skills iables, while somewhat depressed, were generally con-
for children from low-income households. It appears sistent with the PICS scores in which the child language
that the quality of parent–child interactions, even at very component was included. This result suggests that the
early child ages, is related to early literacy skills. Fur- component of child language did not significantly inflate
thermore, the PICS seems to be a fairly powerful predic- the relationship between the PICS scores and scores on
tor of early literacy skills. It predicted early literacy skills the outcome variables; therefore, maintaining child lan-
better than did the SFRS. Unfortunately, none of the guage as part of the PICS rating system was supported.
PICS Activity scores predicted early literacy skills better The observational data provided through the PICS rating
than the Overall PICS score. The play, teaching, frustra- system is gathered in a more natural manner than that
tion, and Overall PICS scores instead tended to be re- used in other measures of child language (e.g., the PPVT-
lated to the outcome variables in similar ways and, when III) and may help to provide additional information about
combined, predicted early literacy skills. These findings how children use their language to interact with adults in
will be discussed further in the following sections. simulated daily activities.
The relationships between parent–child interactions
and early literacy skills found in this study extend previ-
Interactions and Early Literacy Skills ous researchers’ findings regarding relationships between
Results of this study showed that parent–infant/toddler early parent–child interactions and later skills develop-
interactions are related to the early literacy skills of re- ment. For example, mothers’ responsiveness has been
ceptive vocabulary, symbolic representation, and phone- shown to correlate with later language development (Born-
mic analysis. The PPVT-III scores were related most stein & Tamis-LeMonda, 1989) and with cognition and
strongly to the 36-month PICS scores, and they were re- social development (Landry et al., 2001), joint attention
lated strongly to the Overall PICS score and the 24-month has been shown to correlate with acquisition of words
PICS scores. The WJ-R Letter-Word Identification subtest (Tomasello & Farrar, 1986), and parental guidance has
scores were related most strongly to the Overall PICS been shown to correlate with social development (Lan-
score and the 24-month PICS scores. The 36-month PICS dry et al., 1997). These researchers demonstrated that
scores were related to phonemic analysis skills as mea- parent–child interactions were related to skills that de-
sured by the Phonemic Analysis subtest of the TOLD-P:3. veloped later, which was consistent with the results in the
Furthermore, the mean scores on the PPVT-III, WJ-R, current study. In each of these previous studies, however,
and TOLD-P:3 for the children in this study were close only one aspect of the parent–child interaction was ana-
to the norm-referenced standard scores and deviations lyzed. This study combined ratings of six aspects of
parent–child interactions into one scale in an effort to
determine how this combination of early parent and
child behaviors, considered globally, influences early lit-
TABLE 6. Zero-Order Correlations for Teaching,
Frustration, Play, and Overall PICS Scores eracy skills.
Furthermore, it should be noted that the studies ref-
PICS score PPVT-III WJ-R TOLD-P:3 erenced above tended to measure skill development over
a very short time. For example, in the study conducted
Overall (M) .58** .50** .31
by Tomasello and Farrar (1986), adult–child interactions
Teaching .44* .43* .20 were analyzed and then language acquisition skills were
Frustration .43* .16 .33 assessed after only 1 week. Consistent with the results of
the current study, some researchers have demonstrated a
Play .54** .52** .30
relationship between early parent–child interactions and
skill development longitudinally. For example, Hart and
Note. PICS = Parent-Infant/Toddler Interaction Coding System (Dodici
& Draper, 2001); PPVT-III = Peabody Picture Vocabulary Test–Third Risley (1995) found that the variety and amount of words
Edition (Dunn & Dunn, 1997); WJ-R = Woodcock-Johnson Tests of parents used with their children prior to age 3 were re-
Achievement– Revised (Woodcock & Johnson, 1989); TOLD-P:3 = Test lated to receptive language skills at age 9. Beckwith and
of Language Development–Primary: Third Edition (Newcomer &
Hammill, 1997). Rodning (1996) found that the parent–child interaction
*p < .05. **p < .01. characteristics of maternal sensitivity, child engagement,
Early Parent–Child Interactions and Early Literacy 133

and dyadic fit measured at 20 months influenced social was slightly larger (39) than in the present study, whereas
development at 60 months. Landry et al. (2001) discovered in the Payne et al. study, the sample size was much larger
that parental responsiveness at 6, 12, and 24 months was (236).
related to social and cognitive development at 48 to 60 The different findings noted between the current
months. study and the Rush study may be influenced by the fact
that in the former study, the parent–infant/toddler inter-
actions were observed and rated over a 3-year period,
Interactions Versus Reported
versus a 1-hour observation session that occurred when
Literacy Experiences the children were between the ages of 48 and 66 months
In this study, the Overall PICS score predicted early lit- in the latter study. The number of observations and the
eracy skills better than the SFRS did. Correlations between ratings that occurred over a 3 years versus 1 hour may in-
the Overall PICS score and all outcome variables were dicate that the cumulative effect of parent–child interac-
stronger than those between the SFRS and all outcome tions over the first 3 years of life has more influence on
variables. Furthermore, the correlations between the Over- early literacy skills than do later parent–child interactions.
all PICS score and the outcome variables in this study This speculation supports the conclusions presented in
were as strong, or stronger, than correlations between the Hart and Risley (1995) study. Hart and Risley pro-
the SFRS and outcome variables found in previous stud- posed that interventions for children introduced after the
ies (Payne et al., 1994; Rush 1999). age of 3 years may have limited impact on developmen-
Specifically, Rush (1999) found significant correla- tal skills due to the cumulative effect of the first 3 years
tions between (a) the SFRS and (b) the PPVT-R and a letter- of parent–child interactions.
naming task that was similar to the WJ-R subscale used
in this study. In the Rush study, which had 39 participants,
Relationships Between PICS Activity
the SFRS correlated with the PPVT-R at r = .61, p < .01,
and with the letter-naming task at r = .48, p < .01. In this Scores and Outcomes
study, the PICS correlation with the PPVT-III was r = .57 A high level of multicollinearity existed between the PICS
and with the WJ-R was r = .49. In a study of 323 partic- Activity scores and the Overall PICS score, making it im-
ipants, Payne and his colleagues (1994) also found a sig- possible to determine if any one individual PICS Activity
nificant correlation between a derived score taken from score better predicted early literacy skills than the Over-
the SFRS, which included the nine questions used in the all PICS score. One might suspect that parents use simi-
Rush study and the current study, and the PPVT-R (r = lar behaviors and techniques with their children when
.42, p < .001). In both the Rush and Payne et al. studies, dealing with frustrating circumstances, teaching novel
the SFRS was completed at approximately the same time tasks, or playing, which certainly was true in this study.
that the PPVT-R and other measures were completed but To further analyze how the Overall PICS and the PICS
revealed similar correlations among the predictor and Activity scores related to early literacy skills, we exam-
outcome variables when compared to the results in the ined the variables via a correlation matrix. The PPVT-III
current study. The correlations reported in the current was strongly related to the three PICS Activity scores of
study, which are as strong or stronger than the Rush and teaching, frustration, and play, as well as to the Overall
Payne et al. results, suggest that the PICS is a good pre- PICS score. The WJ-R was strongly related to all of these
dictor of early literacy skills. The fact that the PICS, variables, except for the frustration Activity score. In
which was based on parent–child interactions prior to a general, it appears that the Overall PICS score is a better
child age of 3 years, was able to predict early literacy skills predictor of early literacy skills than any one PICS Activ-
better than the SFRS, which was based on parent reports ity score.
at a child age of 54 months, holds possible implications
for earlier interventions and assessment related to early
literacy skills.
Limitations
The differences between the results in the current Limitations of this study included the small sample size,
study and those for the Rush (1999) and Payne et al. the homogeneity of the participants, the correlational na-
(1994) studies are puzzling. In both the Rush and the ture of the analyses, and possible intrusiveness of video-
Payne et al. studies, the SFRS was completed by the taping parent–infant interactions. All of the participants
mother when the child was between the ages of 48 and 66 were Caucasian and had similar educational levels. These
months, similar to the time at which the measure was two variables may have restricted the variability and
completed in the current study. Furthermore, the popula- the overall generalizability of this study. Expanding the
tions were similar; all studies included only low-income population to include ethnically diverse families from
families. Sample size, however, may have played a role in middle-income and high-income households may pro-
the conflicting results. In the Rush study, the sample size vide additional information regarding the external valid-
134 Topics in Early Childhood Special Education 23:3

ity of the PICS scores. Because of the small sample size, the PICS has captured early parent–child interactions in
correlation was the most appropriate—but limiting— relatively naturally occurring situations and have related
statistical analysis. Furthermore, the fact that the parent– these interactions to early literacy skills. It is our hope
child interactions were videotaped could have altered that further investigation will provide better, earlier in-
normal parent–child interactions. terventions to parents and children from low-income fam-
ilies in an effort to decrease the problems that children
may face when learning to read.
Implications The longitudinal nature of the current study high-
Understanding how everyday interactions of parents with lights the importance of these findings and argues for the
their infants and toddlers may influence early literacy potential usefulness of the PICS. In this study, the quality
skills, and possibly later school success, is critical, espe- of parent–child interactions, observed when children were
cially for parents from low-income households. Many 24 months old, was related to early literacy skills mea-
parents may believe that because infants and most tod- sured when children were approximately 54 months old.
dlers do not talk, talking to their young children is irrel- This 30-month difference between when the interactions
evant, or they may think that a child should be seen and were observed and language skills were measured repre-
not heard. The findings of this study and other studies sents an important window for development. It is possi-
(Hart & Risley, 1995; Walker et al., 1994) indicate this ble that interventions designed to enhance parent– child
is not true. The first 3 years of parent–child interactions interactions could be implemented during that time and
predict early literacy skills. In fact, previous research has in turn could influence literacy skills at the pre-
shown that early literacy skills predict later academic kindergarten level and beyond. ◆
success (Juel, 1988). It therefore seems imperative that
parents be made aware of the influence their everyday in-
teractions may have on their children’s success in school. AUTHORS’ NOTES
Parents need to be made aware that it is not only lit- 1. The findings reported here are based on research conducted as part
eracy activities, such as reading with their children or of the national Early Head Start Research and Evaluation Project
going to the library, that influence later literacy skills. funded by the Administration on Children, Youth and Families
(ACF), U.S. Department of Health and Human Services under con-
Although the importance of these activities cannot be
tract 105-95-1936 to Mathematica Policy Research, Princeton,
minimized, parents should also be informed that children New Jersey, and Columbia University’s Center for Children and
learn from everyday interactions (Rogoff, 1990). Parents Families, Teachers College, in conjunction with the Early Head
have the opportunity to guide their children’s participa- Start Research Consortium. We also acknowledge funding from the
tion in life and teach them through daily experiences, not Head Start Bureau through Grant 90YF0014/01 to Iowa State Uni-
versity and funding to support the graduate study of the first author
just during reading and literacy activities. Parents need to
from the Iowa State University College of Family and Consumer Sci-
seize teachable moments as they occur all day long, and ences. The research is part of the independent research Iowa State
many parents may need education and/or support to do University conducted with the Mid-Iowa Community Action, Inc.
this effectively. Early Head Start Program, which is one of 17 programs participat-
ing in the national Early Head Start study. The Consortium consists
of representatives from 17 programs participating in the evaluation,
Future Research 15 local research teams, the evaluation contractors, and ACF. Re-
search institutions in the Consortium (and principal researchers) in-
Establishing the reliability and validity of the PICS with clude ACF (Rachel Chazan Cohen, Judith Jerald, Esther Kresh,
additional data from larger and more diverse samples sizes Helen Raikes, and Louisa Tarullo); Catholic University of America
needs to be done. If psychometrically appropriate para- (Michaela Farber, Lynn Milgram Mayer, Harriet Liebow, Nancy
Taylor, Elizabeth Timberlake, and Shavaun Wall); Columbia Univer-
meters for the PICS are established, developing a version sity (Lisa Berlin, Christy Brady-Smith, Jeanne Brooks-Gunn, and
suitable for use by interventionists working in family’s Alison Sidle Fuligni); Harvard University (Catherine Ayoub, Bar-
homes would be useful. For example, developing a PICS bara Alexander Pan, and Catherine Snow); Iowa State University
checklist that would negate the need for videotaping and (Dee Draper, Gayle Luze, Susan McBride, and Carla Peterson);
coding could assist an interventionist to not only moni- Mathematica Policy Research (Kimberly Boller, Jill Constantine,
Ellen Eliason Kisker, John M. Love, Diane Paulsell, Christine Ross,
tor parent–child interactions but also determine the fo- Peter Schochet, Cheri Vogel, and Welmoet van Kammen); Medical
cus of future interventions. University of South Carolina (Richard Faldowski, Gui-Young
Hong, and Susan Pickrel); Michigan State University (Hiram
Fitzgerald, Tom Reischl, and Rachel Schiffman); New York Univer-
Conclusions sity (Mark Spellmann and Catherine Tamis LeMonda); University
of Arkansas (Robert Bradley, Richard Clubb, Andrea Hart, Mark
This study provides support for the theory that a strong Swanson, and Leanne Whiteside-Mansell); University of California
relationship between early parent–child interactions and at Los Angeles (Carollee Howes and Claire Hamilton); University
later skill development exists. We have demonstrated that of Colorado Health Sciences Center (Robert Emde, Jon Korfmacher,
Early Parent–Child Interactions and Early Literacy 135

JoAnn Robinson, Paul Spicer, and Norman Watt); University of guage development. British Journal of Developmental Psychology,
Kansas (Jane Atwater, Judith Carta, and Jean Ann Summers); Uni- 4, 261–268.
versity of Missouri at Columbia (Mark Fine, Jean Ispa, and Kathy Hart, B., & Risley, T. R. (1995). Meaningful differences in the every-
Thornburg); University of Pittsburgh (Beth Green, Carol McAllis- day experiences of young American children. Baltimore: Brookes.
ter, and Robert McCall); University of Washington School of Edu- Juel, C. (1988). Learning to read and write: A longitudinal study of 54
cation (Eduardo Armijo and Joseph Stowitschek); University of children from first through fourth grades. Journal of Educational
Washington School of Nursing (Kathryn Barnard and Susan Spieker), Psychology, 80, 437–447.
and Utah State University (Lisa Boyce, Gina Cook, Catherine Callow- Lamb-Parker, F., Boak, A. Y., Griffin, K. W., Ripple, C., & Peay, L.
Heusser, and Lori Roggman). (1999). Parent-infant/toddler relationship, home learning environ-
2. The content of this publication does not necessarily reflect the views ment, and school readiness. School Psychology Review, 28, 413–425.
or policies of the U.S. Department of Health and Human Services Landry, S. H., Smith, K. E., Miller-Loncar, C. L., & Swank, P. R.
or of Iowa State University, nor does mention of trade names, com- (1997). Predicting cognitive-language and social growth curves
mercial products, or organizations imply endorsement by the U.S. from early maternal behaviors in children and varying degrees of bi-
government. ological risk. Developmental Psychology, 33, 1040–1053.
Landry, S. H., Smith, K. E., Swank, P. R., Assel, M. A., & Vellet, S.
(2001). Does early responsive parenting have a special importance
REFERENCES
for children’s development or is consistency across early childhood
Adams, M. J. (1990). Beginning to read. Cambridge, MA: MIT Press. necessary? Developmental Psychology, 37, 387–403
Adamson, L. B., & Bakeman, R. (1984). Mother’s communicative acts: National Institute of Child Health and Human Development, Early
Changes during infancy. Infant Behavior and Development, 7, 467– Child Care Research Network. (1999). Child-care and mother-child
478. interaction in the first 3 years of life. Developmental Psychology,
Akhtar, N., Dunham, F., & Dunham, P. J. (1991). Directive interactions 35, 1399–1413.
and early vocabulary development: The role of joint attentional fo- Newcomer, P. L., & Hammill, D. D. (1997). Test of language development–
cus. Journal of Child Language, 18, 41–49. Primary (3rd ed., Examiner’s manual). Austin, TX: PRO-ED.
Barnard, K. E. (1997). Influencing parent-infant/toddler interactions Pappas-Jones, C., & Adamson, L.B. (1987). Language use in mother-
for children at risk. In M. J. Guralnick (Ed.), The effectiveness of child and mother-child-sibling interactions. Child Development, 58,
early intervention (pp. 249–270). Baltimore: Brookes. 356–366.
Beckwith, L., & Rodning, C. (1996). Dyadic processes between mother Payne, A. C., Whitehurst, G. J., & Angell, A. L. (1994). The role of
and pre-term infants development at ages 2 to 5 years. Infant Men- home literacy environment in the development of language ability
tal Health Journal, 17, 322–333. in preschool children from low-income families. Early Childhood
Bornstein, M. H., & Tamis-LeMonda, C. S. (1989). Maternal respon- Research Quarterly, 9, 427–440.
siveness and cognitive development in children. In M. H. Bornstein Pianta, R. C., & Egeland, B. (1994). Predictors of instability in chil-
(Ed.), Maternal responsiveness: Characteristics and consequences dren’s mental test performance at 24, 48, and 96 months. Intelli-
(New Directions for Child Development, No. 43, pp. 49–61). San gence, 18, 145–163.
Francisco: Jossey-Bass. Purcell-Gates, V. (1996). Stories, coupons, and the TV Guide: Rela-
Cunningham, A. E., & Stanovich, K. E. (1997). Early reading acquisi- tionships between the home literacy experiences and emergent liter-
tion and its relationship to reading experience and ability 10 years acy knowledge. Reading Research Quarterly, 31, 406–428.
later. Developmental Psychology, 33, 934–945. Purcell-Gates, V., & Dahl, K. (1991). Low-SES children’s success and
DeTemple, J. M. (2001). Parents and children reading books together. failure at early literacy learning in skills-based classrooms. Journal
In D. K. Dickinson & P. O. Patton (Eds.), Beginning literacy with of Reading Behavior, 23, 1–34.
language: Young children learning at home and school (pp. 31–51). Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development
Baltimore: Brookes. in social context. New York: Oxford University Press.
Dickinson, D. K., Cote, L., & Smith, M.W. (1993). Learning vocabu- Rush, K. L. (1999). Caregiver-child interactions and early literacy de-
lary in preschool: Social and discourse context affecting vocabulary velopment of preschool children from low-income environments.
growth. New Directions for Child Development, 61, 67–78. Topics in Early Child Special Education, 19, 3–14.
Dickinson, D. K., & Snow, C. E. (1987). Interrelationships among pre- Sameroff, A. J., Seifer, R., Baldwin, A., & Baldwin, C. (1993). Stability
reading and oral language skills in kindergarteners from two social of intelligence from preschool to adolescence: The influence of so-
classes. Early Childhood Research Quarterly, 2, 1–25. cial risk factors. Child Development, 64, 80–97.
Dickinson, D. K., & Tabors, P. O. (2001). Young children learning at Saxon, T. F. (1997). A longitudinal study of early mother-infant in-
home and school: Beginning literacy with language. Baltimore: teractions and later language competence. First Language, 17, 271–
Brookes. 281.
Dodici, B. J., & Draper, D. C. (2001). Parent-infant/toddler interaction Slavin, R. E., Karweit, N. L., Wasik, B. A., Madden, N. A., & Dolan,
coding system. Ames: Iowa State University. L. J. (1994). Success for all: A comprehensive approach to preven-
Dunn, L. M., & Dunn, L. M. (1997). Peabody picture vocabulary test tion and early intervention. In R. E. Slavin, N. L. Karweit, & B. A.
(3rd ed., Examiner’s manual). Pines Circle, MN: American Guidance Wasik (Eds.), Preventing early school failure (pp. 175–205). Bos-
Service. ton: Allyn & Bacon.
Estrada, P., Arsenio, W. F., Hess, R. D., & Holloway, S. D. (1987). Snow, C. E., Burns, M. S., & Griffin, P. (Eds.). (1998). Preventing read-
Affective quality of the mother-child relationship: Longitudinal ing difficulties in young children. Washington, DC: National Acad-
consequences for children’s school-relevant cognitive functioning. emy Press.
Developmental Psychology, 23, 210–215. Snow, C. E., Tabors, P. O., & Dickinson, D. K. (2001). Language de-
Fenson, L., Dale, P., Reznick, S., Thal, D., Bates, E., Hartung, J., et al. velopment in the preschool years. In D. K. Dickinson & P. O.
(1993). MacArthur communicative development inventories (2nd Tabors (Eds.), Beginning literacy with language: Young children
ed.). San Diego, CA: Singular. learning at home and school (pp. 1–25) Baltimore: Brookes.
Harris, M., Jones, D., Brookes, S., & Grant, J. (1986). Relations be- Tomasello, M., & Farrar, M. J. (1986). Joint attention and early lan-
tween the non-verbal context of maternal speech and rate of lan- guage. Child Development, 57, 1454–1463.
136 Topics in Early Childhood Special Education 23:3

Tomasello, M., & Todd, J. (1983). Joint attention and lexical acquisi- of school outcomes based on early language production and so-
tion style. First Language, 4, 197–212. cioeconomic factors. Child Development, 65, 606–621.
Vygotsky, L. (1986). Thought and language. Cambridge, MA: MIT Whitehurst, G. J. (1993). Stony Brook family reading survey. Stony
Press. (Original work published 1934) Brook, NY: Author.
Wagner, R. K., Torgesen, J. K., & Rashotte, C. A. (1994). Development Woodcock, R. W., & Mather, N. (1990). Woodcock-Johnson tests of
of reading-related phonological processing abilities: New evidence achievement–Revised (Examiner’s manual). In R. W. Woodcock
of bidirectional causality from a latent variable longitudinal study. & M. B. Johnson (Eds.), Woodcock-Johnson psycho-educational
Developmental Psychology, 33, 73–87. battery–Revised. Allen, TX: DLM.
Walker, D., Greenwood, C., Hart, B., & Carta, J. (1994). Predictions