The Pedagogical Goals of Soliloquy
Learning
Dr. Marilyn Jager Adams, Chief
Scientist, Soliloquy Learning
A. Introduction
Soliloquy Learning's electronic
reading tutor is designed to present stories and graphics on the screen
of a desktop or laptop computer and, by means of speech recognition-based
technology, to assess, guide, and respond to students in real-time as they
read aloud.
B. Background and Significance
Over the past several decades, researchers
have made great strides toward understanding the nature of reading and
its acquisition. Built from hundreds of separate scientific studies,
and repeatedly re-examined and endorsed through formal reviews policy efforts
across the country, the message of this work is compelling: Given
proper instruction and appropriate activities and support, every healthy
child can learn to read.
Yet, the National Assessment of
Educational Progress (NAEP) continues to find that as many as 40% of our
nation’s fourth-graders read too poorly to understand or learn from grade-level
texts; in high poverty neighborhoods, this statistic rises above 60% (see
http://nces.ed.gov/nationsreportcard/reading/
and C is for
Consequences of Reading Failure).
In view of the lessons from research,
the results of our national assessments have become a call to action.
The Reading First section of the most recent authorization of the Elementary
and Special Education Act ("No Child Left Behind") is specifically intended
to provide assistance to schools and districts toward implementing five
essential components of reading instruction:
• Phonemic Awareness
• Phonics
• Fluency
• Vocabulary Development
• Comprehension Strategies
Again, the importance of each
of these components of literacy growth has been amply documented by research.
Moreover, these are the five key components called out as most critical
by the National Academy of Sciences Committee on Preventing Reading Difficulties
(National Research Council, 1998). In addition, a review by the federally
appointed National Reading Panel (2000) of the instructional research has
affirmed the responsiveness of each to well-designed instruction.
Yet, the National Reading Panel
also found that the maturity and availability of instructional options
ranged broadly across these five components. In the areas of Phonemic Awareness
and Phonics, the Panel was able to identify instructional techniques that
had been proven effective in experiments including large-scale field studies
involving the full range of student populations and implementation by actual
teachers. In contrast, the instructional literature on Vocabulary
Development and Comprehension Strategies was limited to small studies,
most often conducted or tightly supervised by the experimenters themselves,
and focused on methods, goals, or student populations that were, each in
themselves, too specific for ready generalization to the larger challenge.
The Panel’s findings on Fluency
lay in between. Through review of the available instructional research,
the Panel identified as particularly effective a single class of activities,
Guided Oral Reading. The study pool showed this general procedure to be
beneficial at least through grade 5 with normal readers and well beyond
that with struggling readers. Its implementation was shown to require
no special materials other than the texts normally available to students
in their homes or classrooms. Nor did its benefits depend on any
special or sophisticated training on the part of the helpers which included
teachers, parents, adult volunteers and, in a few cases, supervised student
peers. Under such names as neurological impress, repeated reading, peer
tutoring, shared reading, and assisted reading, the Panel identified a
number of variants of the procedure. However, it found no indication of
differences in the effectiveness of these different approaches.
Instead, the key and common attribute
of the effective techniques seemed to be that they provided students with
the opportunity to read aloud with on-going guidance and feedback.
Most importantly and across all such differences, the National Reading
Panel reported that the benefits of Guided Oral Reading are strong and
broad, significantly impacting students' growth in word recognition, fluency,
and comprehension, and full-scale reading scores.
In summarizing these findings, the
National Reading Panel reminds us that "Children who do not develop reading
fluency, no matter how bright they are, will continue to read slowly and
with great effort" (p. 3-3). On the basis of its findings, the Panel
therefore urges that classroom teachers find ways to engage their students
in guided oral reading on a regular basis and, additionally, that they
make a practice of regularly assessing the students' oral reading accuracy
and rate.
Yet here lies a problem. If
the amount of supported oral reading is of such paramount importance to
young readers, then how do we make it happen? The assessment and
guidance of oral reading requires one-on-one attention to each student.
In the classroom, however, even 5 minutes a day of supported reading time
is far more than any teacher can deliver to each of her 20 or more students.
Nor does relegating this task to parents offer a fair or hopeful solution,
especially for children of poverty (see, e.g., Heath, 1983).
C. The Reading System
The primary goal of Soliloquy Learning's
reading system is to address this core educational problem. As such
the system is designed with two purposes in mind: (1) to offer
students the oral reading opportunities and guidance they need to learn
through reading, and (2) to provide teachers with the assistance and information
they need to monitor and guide their students' growth.
The current version of Soliloquy's
reading tutor is targeted at "novice" readers -- readers who essentially
know how to decode but lack the automaticity, vocabulary, and comprehension
control to read with adequate confidence, speed, and productivity on their
own. Presenting texts and supporting graphics to the student on the
computer monitor, the system is designed to offer supported opportunity
for orally reading and rereading level-appropriate text. Whether
in the course of her or his own reading or during preview or review of
a text, the student can ask the computer for the pronunciation of any word.
Alternatively, if the student requests that a word be explained, the computer
presents the word's contextually appropriate meaning, a sentence to illustrate
its usage and, wherever possible, a graphic as well. Students can
also request that the text, or any segment of it, be read aloud to them.
Soliloquy's narrators are professional voice models whose readings are
expressive and carefully enunciated. So as to make it easier for
the student to read along during such modeling, the text is highlighted
in time to the narrator's voice and the pause between sentences has been
computationally altered to ensure enough time for the student to understand
and keep up.
The key attribute of Soliloquy's
tutoring system lies in the support it provides for students themselves
to read. Specifically, by virtue of a speech-recognition layer, the
software listens to the student's performance as she or he reads aloud.
Monitoring for signs of difficulty, it autonomously provides assistance
when the student stumbles or gets stuck. In the background, meanwhile,
it builds ongoing records of what the student has read and reread, of her
or his fluency on each reading, and of the specific words and segments
of text with which the student had difficulty. The system is designed
to let students review all such information at will so that they can practice
and repair their readings to their own satisfaction.
Currently, each text offered by
the Soliloquy tutor is followed by a set of "quiz" questions. These quiz
items are not designed to "test" the students' memory of the texts.
Rather, their purpose is to cause the students to recognize and refine
their grasp of key concepts, events, and vocabulary from the passage.
In this way, the primary goal of the quizzes is to prepare students to
read the text with more focus and purpose the next time through.
Even so, the potential of Soliloquy Learning's system for assisting students'
reading comprehension extends beyond such after-reading activities.
By virtue of the speech recognition layer, the system continually knows
where a student is while reading. Thus, in the course of reading,
at the very moment when most appropriate, it can enrich, guide, or query
comprehension by means of any of its multimedia capabilities. Developing
such interactive comprehension supports is among Soliloquy Learning’s priority
projects as it moves forward.
In complement to such direct assistance
and opportunity for students, Soliloquy Learning is equally committed to
maximizing the utility of the system for teachers. For each student,
therefore, the system maintains cumulative records of what she or he has
read and how well. For each reading of each text, the system creates
reports of students' accuracy, reading speed, and comprehension.
It also maintains digital recordings of the student’s reading, enabling
teachers to build performance portfolios for monitoring growth.
Because these recordings are digital and are tied to the text and the performance
reports, teachers can listen to them selectively, quickly and easily identifying
and accessing segments of special concern or lesson-relevance. Further,
as the system lets teachers review exactly what each child has read and
how well, it lets them knowingly choose books and challenges on which to
focus during the one-on-one time they have together.
By working with educators and in
classrooms, Soliloquy Learning's goal is to continue refining and extending
the system's modes of use, assessment and reporting options, and educational
activities as most useful and usable for teachers. By creating partnerships
with trade and schoolbook publishers, Soliloquy Learning's goal is to ensure
that the system can provide such help across the ability levels, curricular
topics, and language and literary challenges on which students' literacy
growth depend. By continually refining the software's dynamics through
empirical research, Soliloquy Learning's goal is to maximize the sensitivity
and helpfulness with which the system can provide intervention and assistance
to students, regardless of their reading characteristics, or specific difficulties,
or linguistic background.
You can view selected story summaries
(in PDF format) to get an idea of the types of text used in Soliloquy's
Reading Assistant.
View
the first summary
View
the second summary
View
the third summary
View
the fourth summary
To learn more about this topic,
see F is for Fluency
and V is for Volume.
References:
Heath, S. B. (1983). Ways with Words:
Language, Life, and Work in Communities and Classrooms. New York: Cambridge
University Press.
National Reading Panel (2000). Teaching
children to read: An evidenced-based assessment of the scientific research
literature on reading and its implications for reading instruction. NIH
Publication No. 00-4754. Washington, D.C.: National Institute of Child
Health and Human Development.
Snow, C. E., M. S. Burns, and P.
Griffin, eds. (1998). Preventing Reading Difficulties in Young Children.
Washington, DC: National Academy Press.
