Content uploaded by T. V. Joe Layng
Author content
All content in this area was uploaded by T. V. Joe Layng on Jun 17, 2015
Content may be subject to copyright.
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™: Reliably Teaching
Children to Read
1
T. V. Joe Layng, Ph.D.
2
, Janet S. Twyman, Ph.D., Greg Stikeleather, MA.
Reading proficiency is a crucial foundation for success in all academic
areas, yet we are a nation faced with a reading crisis. Four in ten children have
literacy problems, and over 40% of our nation's fourth graders score below basic
reading levels (National Center for Learning Disabilities, 2001). Learning to read
is a formidable challenge for more than 50% percent of our nation's school
children (Lyon, 1998), and parents spend billions of dollars each year on
extracurricular books, software, tutors, and other reading aids. Teachers and
schools face the challenges of finding the best teaching method, implementing
these methods in large classrooms, and accommodating students' widely varying
abilities and readiness. Despite the time and money spent on solving the reading
difficulties of our nation's children, the problems aren't disappearing. Headsprout,
a Seattle-based applied learning sciences company, has been working on a
solution that bridges the efforts of parents, schools and agencies with the goal of
preparing children for success in any core reading program chosen by a teacher,
school, or school district.
Headsprout spent over four years in a major research and development
effort to build a beginning reading program that incorporates principles derived
from the scientific investigation of early reading (see for example Rayner et al,
2002) with principles derived from the experimental and applied analysis of
behavior. The result is a highly effective beginning reading program available to
all children over the Internet.
Headsprout Early Reading™ is a new engaging, Internet-based reading
program that effectively teaches the essential skills and strategies required for
rapid reading success. Experienced educators and learning scientists leveraged
years of classroom experience and scientific research in developing the product.
The program was extensively tested with thousands of children, who produced
millions of data points that were collected and analyzed at a cost of millions of
dollars in order to research, develop, test, and refine the program (see Twyman
et al, in press). On the surface the program appears to the child as an interactive
cartoon. Children learn essential reading skills through multiple interactions with
engaging, cartoon-based episodes set in the entertaining environs of Space
World, Dinosaur World, Undersea World, and Jungle World. Underneath is a
sophisticated, patented technology, designed by experienced educators and
learning scientists, that systematically teaches the phonics skills and
comprehension strategies necessary to sound–out words and read with
understanding. Every aspect of the program from instructional design, to
performance outcomes, to ease of use and learner motivation has been
1
This manuscript was the basis for a shorter article, “Inside Headsprout Early Reading,” that
appeared in the Fall 2003 edition of the Current Repertoire, Newsletter of the Cambridge Center for
Behavioral Studies; it also draws upon and includes text from an upcoming book chapter by the
authors: Selected for success: How Headsprout Reading Basics™ teaches beginning reading, to
appear in D. J. Moran & R. Malott (Eds.) Empirically supported educational methods, St. Louis, MO:
Elsevier Science/Academic Press. Which is available from Elsevier/Academic Press
(www.elsevier.com).
2
Senior Scientist, Headsprout, 127 Braodway Ave. E., Suite 300, Seattle, WA 98102.
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
2
extensively researched, reviewed and refined (after Neuman & McCormick,
2002; Sidman, 1960). The program has been so rigorously developed that
Headsprout offers schools a full refund of the price of the product for each
Kindergarten or 1
st
grade student who is not at or above grade level upon
completing the Early Reading program.
Key Skills and Strategies: The building blocks of reading success
Although phonics instruction has drifted in and out of favor in the
educational establishment, a large body of research points to its essential role in
the process of teaching children to read. In April 2000, the Congressionally
mandated National Reading Panel reported that early systematic phonics
instruction improves children's reading and spelling abilities (National Institute of
Child Health and Human Development, 2000). Research also suggests that the
absence of explicit instruction in phonemic awareness and phonics can cause
learning problems that put learners at a permanent educational disadvantage
unless they are corrected by the end of the third grade (National Reading Panel,
2000).
The Public Library Association (n.d.) points out that "research has shown
that there is nearly a 90% probability that a child will remain a poor reader at the
end of the fourth grade if the child is a poor reader at the end of first grade." The
research suggests that Headsprout's approach of explicit instruction in phonemic
awareness, phonics, and a strategy for sounding out words can prevent many
children from developing learning problems and can give almost all children an
equal opportunity to become good readers. When Headsprout children arrive at
school, they will be prepared no matter how large or small the role phonics plays
in their classroom. Headsprout Early Reading is a teacher's ally (not a teacher's
replacement), giving students a boost in essential skills and raising the likelihood
of reading success in a busy classroom, or before formal classroom instruction
even begins.
Research has identified five basic, interconnected sub-skills that all
children must master if they are to become proficient readers (The National Right
to Read Foundation, n.d.), all integral to Headsprout Early Reading. First,
beginning readers must develop what is called phonemic awareness-the
recognition that all words are made of separate sounds, called "phonemes."
Second, beginning readers also need to learn phonics, which is the ability to link
these sounds to the specific letters or combinations of letters representing them
in written language. This association between letters and sounds must become
fluent so that learners can decode words almost instantly. Beginning readers
must learn a strategy to sound out the sequence of phonemes in a word and
blend the sounds back together to read whole words. Third, a learner's spoken
vocabulary must be extended to become a reading vocabulary. They must
understand that the words they read have meaning just as do the words they
say. Further, they should come to understand that words they read have
meaning even if they have not yet encountered that meaning. Fourth, reading
fluency is important to reading success. Fifth, comprehension of what is read is
essential. The seeing and saying of words, although essential, is not sufficient to
create a good reader. Children must understand what they read, and be able to
act on that understanding.
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
3
Headsprout Early Reading tackles these five important features in the
following way:
Phonemic Awareness — Phonemic Awareness instruction is integrated
throughout many of Headsprout Early Reading's teaching routines. Learners hear
letter sounds in order to select visual stimuli, and then hear them again as
confirmation of selections. Learners are asked to say the sounds and then listen
to cartoon characters say sounds, and then select the character that "said the
sound just like you did." Learners put the sounds together, hear them slowly
blended, say them slowly blended, and then hear the sounds said fast as whole
words, and eventually say the words fast. They learn to not only identify and say
the sounds letters make, both independently and as blended units, but to listen to
and identify the sounds they say, a critical step in becoming a speaker as own
listener.
Phonics — In Headsprout Early Reading children learn 110 carefully
chosen phonetic elements, most of which maintain a consistent pronunciation in
nearly 85% of the words in which they appear. This early consistency is
extremely important to ensuring the transfer of segmenting and blending skills
learned in the program, to words encountered outside the program. This allows
the natural outcome of reading in a social environment to become the critical
consequence for reading.
Headsprout has addressed learner and teacher concerns about the rule-
filled, exception-filled English language. Too often, learners are expected to
begin reading by memorizing rules that dictate sound/letter associations only to
have to memorize further exceptions to those rules. The English language uses
the 26 letters of the alphabet to represent 44 sounds - sounds that can be written
in over 400 different ways. To untangle this confusing web for the beginning
reader, Headsprout Early Reading begins with very consistent letters and
sounds, such as "ee," "v," "cl" and "an." As noted earlier, the sounds in
Headsprout Early Reading are stable (read the same way), greatly increasing the
likelihood of learners reading the word correctly.
With Headsprout Early Reading, learners gain confidence early in their
ability to sound out without being distracted by the challenge of memorizing the
English language's many vagaries. By using one, two, and three letter
combinations learners find that sounds can be combined to make meaningful
units of phonemic information. Further, learners quickly discover that some
sounds can have other sounds inside them and that sound units can be
combined to make new sounds. Headsprout Early Reading's instructional
strategies result in learners reliably "adducing" these insights in a discovery-
learning environment, rather than having to be directly taught. They learn to use
their phonics knowledge for sounding-out words in isolation, as parts of
sentences, and when reading stories with words they have not been directly
taught.
Vocabulary Development — Headsprout Early Reading provides a critical
foundation for early vocabulary building, particularly as it affects reading. An
essential component of vocabulary growth is the concept that words are made of
sounds, that when put together, have meaning. Headsprout Early Reading
teaches that words have meaning, and that they make sentences that, in turn,
make stories. Learners begin to add words that are likely to be in their spoken
vocabulary to their reading vocabulary as they sound out new words and learn
selected sight words. Through the use of character names, they learn that words
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
4
they may have never before encountered have meaning as well. More phonetic
elements are added as the initial sounding-out strategies are learned; the words
made from the elements are practiced to ensure that they become a permanent
part of the learner's vocabulary. Once the sounding-out skills are firmed and all
110 phonetic elements taught, a typical learner would, n less than 30 hours of
instruction, have a potential reading vocabulary of over 5000 words.
Reading fluency, including oral reading skills — Fluency is a critical
element to all Headsprout Early Reading activities. Often, fluency work is left to
end of the reading process, when a learner practices reading sentences.
Headsprout understands that fluency at the component skill level is critical to
fluency at the composite skill level (Johnson & Layng, 1994; LaBerge & Samuels,
1974; Samuels & Flor, 1997). From as early as lesson one, learners engage in
fluency building activities for finding sounds in words. By lesson 4, learners are
building fluency on words made up of the sounds they have learned in the
previous lessons, and by lesson 5, learners read their first story. Soon, learners
are practicing reading entire passages in carefully designed fluency activities. In
the 80 lessons that comprise Headsprout Early Reading, over 50 fluency-building
opportunities have been specifically designed to build a strong reading repertoire.
In fewer than 30 hours of instruction a learner will have read 80 separate stories.
Most of the stories are designed for learners to read independently. These
stories include narrative and expository text as well as poetry; they begin with as
few as three sentences and grow to include chapter books.
Some early stories are to be read with someone else, such as a parent.
These stories are more complicated, punctuated with sentences learners can
easily read. Learners, thereby, are exposed to fluent reading at a higher level
then they can currently handle, and must pay close attention so they can read
"their" sentences when it is their turn.
Reading comprehension strategies — An article about beginning
reading began with the following observation (paraphrased): If "Look at the
ceiling" is written on a black board, and a person says, 'look at the ceiling,' the
person is decoding, if the person's head tilts back and a glance upward is
observed, the person is comprehending (Goldiamond & Dyrud, 1966). Though
overly simplified, it emphasizes the important point that the evaluation of
comprehension requires indicator responses that are separate from simply
seeing and saying words or sentences. These indicator responses are key to
teaching and evaluating comprehension. Accordingly, Headsprout Early Reading
employs frequent use of comprehension indicators to test whether what is being
decoded is also being understood. Carefully designed indicators are used to
teach self-observation as well as sentence and story comprehension. After each
reading exercise, learners must choose one of three pictures that go with the
sentence. The pictures vary in such a way as to ensure that the words in the
sentences have been read and are understood. From as early as lesson 5,
learners understand that the sentences they read are not simply lists of words,
but units of meaning. Learners then transition to continually more challenging
comprehension activities, including: selecting pictures that represent meaning
derived from whole stories, constructing meaning by building sentences that
result in an animated picture that represents the sentence, expressing meaning
by building sentences that describe a picture, completing sentences that best
describe a picture by selecting a missing word from four alternatives, and reading
a text passage and selecting the best answer to a written question from among
three written alternative answers.
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
5
Rigorous Formative Evaluation: Ensuring the Success of Technology
Transfer
Whereas these essential features of successful reading instruction are
often said to be included in many reading programs, the ways in which the skills,
which ultimately define these features, are taught, are not so well defined, nor
are they tested. Whereas many programs claim to be “based on research,” the
precise meaning is often left to the reader’s interpretation, and many do not
mean it in the sense of a program that is developed and tested through rigorous
research protocols (Layng, Stikeleather, & Twyman, in press; Layng, Twyman, &
Stikeleather, in press; Twyman, Layng, Stikeleather, & Hobbins, in press). Often
it is a statement of content; that is, the program includes practices that earlier
research has shown to be effective, but the program itself has not been so
tested.
In contrast, Headsprout scientists and instructional designers employed
scientifically derived instructional principles drawn from both the basic and
applied learning sciences, and a rigorous, control–analysis formative evaluation
process throughout the development of the product. A description of the
formative evaluation process and the instructional approach taken by Headsprout
has been described elsewhere (Layng, Stikeleather, & Twyman, in press; Markle,
1967; Markle & tiemann, 1967; Twyman et al, in press). Briefly however, the
process followed that described in Figure 1.
Figure 1. Grapic representation of the iterative control–analysis instructional
development process used by Headsprout.
First, a content analysis was performed (after Tiemann & Markle, 1991;
Twyman et al, in press) to determine the precise objectives toward which the
program would be directed, and the instructional strategies needed to achieve
those objectives. Second, as each instructional activity and then lesson was
designed learners tested it in our laboratory. If the learners had difficulty with the
lesson, it was revised until almost every new learner was successful with that
activity. Our standard for most activities was 90% of learners achieving 90%
correct on the activities and meeting rigorous exit criteria. For example, learners
had to pick out a letter based on its sound from other topographically similar
letters that changed from trial-to-trial with 5 consecutive corrects (no errors).
Since mastery criteria are used, no learner can proceed to the next activity until
the mastery criteria are met. Accordingly, percent correct reflects the amount of
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
6
instruction required to meet the final mastery exit criteria rather than an overall
score reflecting level of mastery.
The objective of the control–analysis development strategy was to ensure
that each leaner met the individual mastery criteria, and reduce the variability in
responding between individual learners such that we could state not only what
the criteria were, but that nearly all learners would meet them. For example,
Figures 2, 3, and 4 show the effect of the design, test, revise, test,
control–analysis approach to controlling variability. The figures depict total
opportunities to respond, responses, and correct responses for different learners
completing Episodes 2 during the course of development. Upon initial episode
testing we determined that we needed to increase the response opportunities,
and reduce the variability between learners for all categories. Changes were
made in the activities that comprise the episode and tested. Variability between
learners, as shown in Figure 3, actually increased. Further revisions were
required which ultimately led to the stability depicted in figure 4. Figures 5 and 6
Episode 2, version 1
0
25
50
75
100
125
150
175
200
225
250
275
300
325
350
375
400
Learners
Total Per Episode
Opportunities
Responses
Corrects
Figure 2. Variability in responding between learners in Episode 1 version 1.
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
7
Episode 2, version 2
0
25
50
75
100
125
150
175
200
225
250
275
300
325
350
375
400
Learners
Total Per Episode
Opportunities
Responses
Corrects
Episode 2, version 3
0
25
50
75
100
125
150
175
200
225
250
275
300
325
350
375
400
Learners
Total Per Episode
Opportunities
Responses
Corrects
Figure 3. Variability in responding between learners in Episode 1 version 2.
Figure 4. Variability in responding between learners in Episode 1 version 3.
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
8
show another example of program revisions that led to stable, predictable,
responding by each individual completing Headsprout Early Reading Episode 17.
Episode 17, version 2
0
25
50
75
100
125
150
175
200
225
250
275
300
325
350
375
400
Learners
Total Per Episode
Opportunities
Responses
Corrects
Episode 17, version 1
0
25
50
75
100
125
150
175
200
225
250
275
300
325
350
375
400
Learners
Total Per Episode
Opportunities
Responses
Corrects
Episode 17, version 1
0
25
50
75
100
125
150
175
200
225
250
275
300
325
350
375
400
Learners
Total Per Episode
Opportunities
Responses
Corrects
Figure 5. Variability in responding between learners in Episode 17 version 1.
Figure 6. Variability in responding between learners in Episode 17 version 2.
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
9
Third, once the developmental testing was completed the program was
tested on a larger scale over the Internet to ensure the validity of the
developmental test data. Over 1000 learners participated in this phase. Over
90% of the learners achieved over 90% correct per each Episode for the entire
program. This validation testing set the occasion for further revisions. Fourth,
tests were conducted in schools. Data from those tests confirmed the
developmental test data and are reported in Layng, Twyman, and Stikeleather (in
press). Thousands of data–based program revisions were made over the course
of development. This process culminated in an instructional methodology we call
Generative Learning Technology, which was issued a U.S. Patent in 2003 (No.
6,523,007).
Our goal for the entire 80–lesson Headsprout Early Reading program was
to produce learners who consistently score at or above grade level, and produce
Kindergarten children with reading skills typical of a mid-second grade learner.
Standardized test outcome data from 16 initial completers suggest we have
accomplished this goal (see figures 7, 8, and 9). Eight had completed Pre–K and
were entering kindergarten, the other eight completed kindergarten and were
entering first–grade. Further school–based summative evaluation studies using
randomized–control procedures are currently underway.
Figure 7. Outcome scores for 8 Pre–Kindegarten and 8 Pre–1
st
Grade learners
Headsprout Early Reading (Episodes 1 - 80)
Woodcock-Johnson Word Identification Subtest
Learners Who Completed the Program in Summer 2003
3.0
2.1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
1st Grade
Kindergarten
Grade Level Equivalent - After Headsprout
Scores Prior
to Entering
(Fall 2003)
:
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
10
Headsprout Early Reading (Episodes 1 - 80)
Iowa Test of Basic Skills: Reading Total - Grade Equivalent
Learners Who Completed the Program in Summer 2003
2.3
1.6
0.0 0.5 1.0 1.5 2.0 2.5
1st Grade
Kindergarten
Grade Level Equivalent - After Headsprout
Scores Prior
to Entering
(Fall 2003):
Headsprout Early Reading (Episodes 1 - 80)
Iowa Test of Basic Skills: Word Analysis- Grade Equivalent
Learners Who Completed the Program in Summer 2003
2.5
1.4
0.0 0.5 1.0 1.5 2.0 2.5 3.0
1st Grade
Kindergarten
Grade Level Equivalent - After Headsprout
Scores Prior
to Entering
(Fall 2003):
Figure 8. Outcome scores for 8 Pre–Kindergarten and 8 Pre 1
st
Grade learners.
Figure 9. Outcome scores for 8 Pre–Kindegarten and 8 Pre–1
st
Grade learners
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
11
Headsprout Early Reading offers a truly balanced approach to beginning
reading instruction that shrinks the chasm between phonics traditionalists and
advocates of literature–based reading instruction. While it has its foundation in
teaching learners to identify letter-sound combinations and combine them with
other letter–sound combinations, its emphasis on comprehension incorporates
elements that do not appear in many phonics programs, and prepares children
for lifelong reading success. Headsprout Early Reading represents a
fundamental example of a technology transfer process that combines the
scientific study of reading and the experimental analysis of behavior, with lessons
learned from instructional systems design, the applied analysis of behavior, and
practical classroom teaching. All of which produced an effective and reliable
technology that was shaped into a comprehensive instructional program through
a rigorous formative evaluation, performed in large part by those for whom it was
designed –the children themselves.
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
12
References
Goldiamond, I. & Dyrud, J. (1966). Reading as operant behavior, J. Money (Ed.)
The disabled reader: Education of the dyslexic child. Baltimore, MD: Johns
Hopkins Press.
Johnson, K. R., & Layng, T. V. J. (1992). Breaking the structuralist barrier: Literacy
and numeracy with fluency. American Psychologist, 47, 1475–1490.
LaBerge, Dl, & Samuels, S.J. (1974). Toward a theory of automataic information
processing in reading. Cognitive Psychology, 6, 293-323.
Layng, T. V. J., Stikeleather, G., Twyman, J. S. (in press). Research-Based
instruction: Should we be paying more attention to formative evaluation? In:
Subotnik, R. & Walberg, H. (Eds.) The Scientific Basis of Educational
Productivity, Washington, DC: American Psychological Association.
Layng, T. V. J., Twyman, J. S., & Stikeleather, G. (in press). Selected for success:
How Headsprout Reading Basics™ teaches beginning reading, to appear in D.
J. Moran & R. Malott (Eds.) Empirically supported educational methods, St.
Louis, MO: Elsevier Science/Academic Press.
Lyon, G. R. (1998). Statement of Dr. G. Reid Lyon. ReadbyGrade3. Retrieved
from http://www.readbygrade3.com/lyon.htm
Markle, S. M. (1967). Empirical testing of programs. In P. C. Lange (Ed.),
Programmed instruction: Sixty-sixth yearbook of the National Society for the
Study of Education: Vol. 2 (pp. 104-138). Chicago, IL: University of Chicago
Press.
Markle, S. M. & Tiemann, P. W. (1967). Programming is a process: Slide/tape
interactive program, Chicago, IL: University of Illinois Chicago.
National Center for Learning Disabilities, The (December, 2001). Legislative
update, 107
th
Session of Congress. Retrieved January 15, 2002, from
http://www.ncld.org/advocacy/update.cfm
National Institute of Child Health and Human Development (2000). Report of the
National Reading Panel. Teaching Children to Read: An Evidence-Based
Assessment of the Scientific Research Literature on Reading and Its
Implications for Reading Instruction: Reports of the Subgroups (NIH Publication
No. 00-4754). Washington, DC: U.S. Government Printing Office.
National Right to Read Foundation, The (n.d.). Retrieved January 15, 2002, from
http://www.nrrf.org/nichd.htm
Neuman, S. B. & McCormick, S. (2002). A case for single subject experiments in
literacy research. In M. L. Kamil, P .B. Mosenthal, P. D. Pearson, & R. Barr
(Eds.) Methods of literacy research (PP. 105-118). Mahwah, NJ: Lawrence
Erlbaum Associates.
Behavioral Technology Today, 3, 7-20 (2003). © Cambridge Center for Behavioral Studies
Headsprout Early Reading™
© Copyright, 2003 Headsprout. www.headsprout.com
13
Public Library Association (n.d.). Retrieved October 3, 2001, from
http://www.read+lang.sbs.sunsb.edu/pla2/research.htm
Rayner, K., Foorm, B. R., Perfetti, C. A., Pesetsky, D. & Seidenberg, M. S. (March,
2002). How should reading be taught? Scientific American, 85-91.
Samuels, S. J., & Flor, R. (1997). "The importance of automaticity for developing
expertise in reading." Reading and Writing Quarterly, 13, 107-122.
Sidman, M. (1960). Tactics of scientific research: Evaluating experimental data in
psychology. Boston, MA: Authors Cooperative, Inc.
Tiemann, P. W., & Markle, S. M. (1991). Analyzing instructional content: A guide to
instruction and evaluation. Champagne, IL: Stipes Publishing Company.
Twyman, J. S., Layng, T. V. J., Stikeleather, G. and Hobbins, K. A. (in press). A
non-linear approach to curriculum design: The role of behavior analysis in
building an effective reading program. In W. L. Heward et al. (Eds.), Focus on
behavior analysis in education, Vol. 3. Upper Saddle River, NJ: Merrill/
Prentice–Hall.
