Title of your poster can be placed here a second line can be added or leave it blank Your name can g

1
Childrens Oral Narratives Are There Connections
to Reading Achievement? Joan E. Foley, Barbara A.
Wasik, Laura M. Justice Presented at the
American Speech-Language-Hearing Association 2007
Annual Convention
Correspondence joan.foley_at_temple.edu
Introduction
Results
Results
Childrens oral narrative production
provides an integrated view of multiple language
systems simultaneously, allowing comparisons of
words, sentences, and narrative structure (Liles,
1993 Liles, Duffy, Merritt, Purcell, 1995
Miller, 1981 Miller Chapman, 1981 Miller,
Heilmann, Nockerts, Iglesias, Fabiano, Francis,
2006). Oral narratives represent an important
bridge to literacy, however, few studies have
adequately investigated their connections to
reading or provided the direction for empirically
derived classroom models of assessment and
intervention (Hedburg Westby, 1993 Roth,
Speece, Cooper, 1996). Studies have
shown that oral narratives produced by learning
disabled children are shorter in length (Leadholm
Miller, 1995), use fewer different words
(Miller, 1991) and less complex syntax (Gillam
Johnson, 1992), and lack a cohesive structure
(Catts, 1993 Catts, Fey, Tomblin, Zhang,
2002). Miller et al. (2006) used a narrative
elicitation task to evaluate the production
features of oral narratives and their
relationship to word/letter identification and
reading comprehension in kindergarten through
third grade students. Evaluating the oral
narratives through lexical, syntactic, and
narrative structure measures, the researchers
found a significant contribution of oral language
skills to both word reading efficiency and
reading comprehension beyond the general
developmental progress accounted for by grade.
Given the extant literature regarding oral
language and literacy connections, this study
investigated whether the results of a narrative
elicitation task administered to kindergarteners
enrolled in Headstart would significantly
differentiate low and high passage comprehenders
at the end of Grade 1.

• High passage comprehenders scored significantly
  higher (p lt .05) compared to low passage
  comprehenders in measures of mean length of
  t-units in morphemes, number of different words,
  total number of words, number of propositions,
  number of nouns, number of evaluative comments,
  number of episodic events identified, and
  reference cohesion at the sentence level.
• 2. Results for reference cohesion at the text
  level for high passage comprehenders were highly
  significant (p lt .01) compared to the scores of
  low passage comprehenders.
• 3. No significant differences existed between
  high and low passage comprehenders in total
  number of t-units, mean length of t-units in
  words, type-token ratio, mazes, repetitions,
  morphological errors, or number of pronouns used.

Table 1 Means, Standard Deviations, and p Values
for Oral Narrative Measures of Low and High
Passage Comprehenders
Measures
Low Comprehenders High
Comprehenders
(n 16)
(n 20)

M SD
M SD
?² p
Discussion

• While the number of t-units did not differ
  significantly between high and low passage
  comprehenders, high comprehenders used more words
  and a greater number of different words compared
  to low comprehenders.
• High comprehenders described significantly more
  events, or propositions (Reilly et al. (1998),
  using more nouns than low comprehenders.
• The use of evaluation, or references to frames
  of mind such as emotion, mental state, or causal
  intentions, were significantly greater for high
  comprehenders. While the majority of evaluative
  references from both groups referred to emotion,
  high comprehenders used more mental state verbs
  and causal references.
• A measure of narrative structure, story
  components (Reilly et al., 1998), consisted of
  eight possible events. Low comprehenders more
  often failed to acknowledge the event involving
  the gopher biting. A lack of vocabulary
  knowledge could account for this difference as
  some children stated their lack of familiarity or
  uncertainty with the type of animal. Some low
  comprehenders also failed to adequately identify
  the resolution of the story (the boy finds the
  frog).
• The most significant finding relates to
  referential cohesion. High comprehenders were
  more likely to use cohesive ties for maintaining
  character reference. Ambiguous pronouns, in which
  the antecedent is unidentifiable, were used
  significantly more by low comprehenders. In many
  instances, the lack of cohesion between nouns and
  pronouns made it difficult to follow the childs
  narrative. In previous studies of children with
  typical language development and those with
  language impairment, (Baltaxe DAngiola, 1992
  Finestack, Fey, Catts, 2006 Liles, Duffy,
  Merritt, Purcell, 1995 Norbury Bishop, 2003
  Paul Smith, 1993 Roth, Spekman, Fye, 1995),
  referential cohesion consistently differentiated
  the two groups of children. Due to the
  developmental nature of referential cohesion,
  some children may lack the linguistic maturity to
  adequately apply cohesive ties at this stage
  (Haslett, 1993 Hickman, 1980 Stahl, 1990).
• Reference cohesion results mirrored those of
  Norris and Brunings (1988) investigation of
  cohesion in low and high achieving readers. They
  concluded that reference cohesion differences
  were tied to the relationship between reading and
  the use of decontextualized language. They state
  it must be determined whether beginning readers
  are sufficiently able to use decontextualized
  language as a means of sharing experiences. Until
  this status is achieved, learning to read with
  fluency and comprehension would seem to be beyond
  the childs linguistic capabilities.
• Reference cohesion as an assessment measure
  presents opportunities for clinicians to go
  beyond assessment batteries that evaluate the
  structure of language. Inclusion of tasks that
  assess whether the child understands how pronouns
  refer anaphorically and categorically across
  sentences may highlight the potential connection
  between oral language and reading comprehension
  difficulties.
• Instruction that incorporates communicative
  opportunities, such as storytelling at various
  levels of decontextualization, provides children
  with challenge, practice, and encouragement in
  developing more mature forms of reference
  cohesion.

Research Question
Which oral narrative measures, collected in
kindergarten, best differentiate low from high
passage comprehenders at the end of Grade 1?
Method
Participants 139 kindergarten students including
66 females (47.5) and 73 males (52.5) who
participated in a larger study, the Preschool
Curriculum Evaluation Research Project (PCER),
conducted at University of Virginia
(UVA). Materials At the end of kindergarten,
students completed a narrative elicitation task,
Frog, where are you? (Mayer, 1969). Follow-up
assessment occurred at the end of first grade
using measures of reading achievement including
the Passage Comprehension subtest of the
Woodcock-Johnson III Tests of Achievement (WJ
III). Procedure A comparative analysis of oral
narrative skills was conducted for students who
scored less than 1 SD below the mean and greater
than 1 SD above the mean of the Passage
Comprehension subtest. Microstructure and
macrostructure measures of oral narrative
production were evaluated using the Systematic
Analysis of Language Transcripts (SALT Miller
Iglesias, 2003-2004) and procedures from prior
research of childrens oral narratives (Bamberg
Damrad-Frye, 1991 Halliday Hasan, 1976
Justice, Bowles, Kaderavek, Ukrainetz, Eisenberg,
Gillam, 2006 Norbury Bishop, 2003 Reilly,
Bates, Marchman, 1998 van der Lely, 1997).

• p lt .01
• p lt .05
• Kruskal Wallis Test
• Grouping Variable SD for WJ-III Passage
  Comprehension Standard Score