Sie sind auf Seite 1von 36


WATKINS, California State University, Stanislaus

and TIMOTHY A. SLOCUM, Utah State University

Instruction is designed to serve this purpose.

The Components Accomplishing this goal requires keen atten-
tion to all aspects of teaching. It would be
of Direct Instruction* much easier if we could focus on one or two
“key issues” and produce measurably superior
instruction, but this is not the case. Producing
highly effective teaching requires that we
Objectives attend to a wide variety of details concerning
After studying this chapter you should be able the design, organization, and delivery of
to instruction. If any one element of instruction
is not done well, high-quality instruction in
1. Identify the three major elements of Direct other areas may not compensate for it. For
Instruction. example, superior instructional delivery cannot
make up for poorly designed instructional
2. Explain what it means to teach a general materials. Likewise, well-designed programs
case. cannot compensate for poor organization.
3. Describe each of the five juxtaposition prin-
Three main components enable Direct
ciples and explain how they contribute to
Instruction to accomplish the goal of teaching
clear communication.
all children effectively and efficiently: (a) pro-
gram design that identifies concepts, rules,
4. Explain the shifts that occur in formats over
strategies, and “big ideas” to be taught and
clear communication through carefully con-
5. Explain what tracks are and how track design structed instructional programs to teach these;
differs from more traditional instruction. (b) organization of instruction, including
scheduling, grouping, and ongoing progress
6. Explain the guidelines for sequencing tasks. monitoring to assure that each student
receives appropriate and sufficient instruction;
7. Describe effective student–teacher interac- and (c) student–teacher interaction tech-
tion techniques. niques that assure that each student is actively

8. Summarize the results of Project Follow

Journal of Direct Instruction, Vol. 3, No. 2, pp. 75–110. From
The purpose of Direct Instruction is to teach Nancy Marchand-Martella, Timothy Slocum, and Ronald
Martella, Introduction to Direct Instruction. Published by Allyn
subject matter efficiently so that all the stu- and Bacon, Boston, MA. Copyright (c) 2004 by Pearson
dents learn all the material in the minimum Education. Reprinted by permission of the publisher.
amount of time. Every strategy, tactic, and * This article is a re-publication of Chapter 2 from
specific technique employed in Direct Introduction to Direct Instruction.

Journal of Direct Instruction 75

engaged with instruction and masters the Content Analysis. The goal of Direct
objectives of each lesson. Instruction is to teach generalized skills; thus,
the first step in developing a Direct
Direct Instruction has been the focus of a vast Instruction program is analysis of the content
amount of research and has been shown to be and identification of concepts, rules, strate-
highly effective for a wide range of content gies, and “big ideas” (i.e., those concepts that
and with diverse learners—from those identi- provide strategies that students can use to fur-
fied as gifted, to students who require special ther develop their expertise in a subject mat-
education services. Studies have shown excel- ter) to be taught. The content area, such as
lent outcomes in basic skills, complex cogni- reading or earth science, is carefully analyzed
tive tasks, and affective areas such as to find key big ideas that can be taught to stu-
students’ self-concepts and confidence. This dents to enable them to exhibit generalized
chapter will describe the three main compo- performance to the widest possible range of
nents of Direct Instruction, and briefly review examples and situations. Identification of
the research base on the effectiveness of these generalizations is the foundation of
Direct Instruction. Direct Instruction.

Becker (1971) illustrated the power and effi-

Main Components ciency of strategy-based instruction with an
example from the area of basic reading. A non-
of Direct Instruction strategic or rote teaching approach would
teach students to recognize whole words. In
In this section, we describe the three main this rote approach, each word would be taught
components of Direct Instruction: the pro- as a separate entity with no system for teach-
gram design, organization of instruction, and ing generalizable strategies for decoding new
student–teacher interactions that make Direct words. In the rote approach, after the teacher
Instruction effective. has taught 10 words, students should be able
to read (at best) 10 useful words. In contrast,
Program Design a strategic approach would be to teach 10 let-
Program design includes five main elements. ter–sound relations and the skill of sounding
First, program design begins by carefully ana- out words. When students have learned these
lyzing the content matter and identifying 10 sounds and the sounding-out skill, they can
central organizing ideas and generalizable read 720 words made up of 3 sounds (e.g., cat),
strategies that enable students to learn more 4,320 words of 4 sounds (e.g., cram), and
in less time. Second, clear communication is 21,600 words of 5 sounds (e.g., scram) for a
designed to minimize ambiguity for stu- total of over 25,000 words. Not all of these
dents. Third, instructional formats are words would be real words, some would be
designed to structure the dialogue between pseudowords (e.g., blums), but the example
teachers and students. Fourth, skills are illustrates the power of strategic instruction.
sequenced to maximize student success and (This strategy and other reading strategies are
minimize points of confusion. Fifth, instruc- described in more detail in Chapter 4.) The
tional topics and objectives are organized efficiency that results from teaching generaliz-
into tracks that allow for systematic skill able big ideas is the goal of the content analy-
development across the length of a program sis that underlies Direct Instruction. This
and support cumulative review and applica- example also illustrates that even in difficult
tion. Together, these elements result in content areas that are fraught with exceptions,
instructional programs that are highly effec- such as reading in English, powerful general-
tive for a wide range of learners. izations are possible.

76 Summer 2003
Spelling is often taught by rote memoriza- A common and persistent misunderstanding is
tion of whole words resulting in little or no that Direct Instruction teaches students to
generalization. However, wide generaliza- memorize simple responses to specific stimuli,
tions are possible. Teaching the skill of commonly referred to as rote learning. In reality,
detecting individual sounds in a spoken word Direct Instruction programs enable students
and matching sounds to written letters is a to learn more in less time for the very reason
very efficient beginning point. In addition, if that they are not learning isolated, unrelated
students learn to spell the parts of words bits of information by rote, but are learning
called morphographs (prefixes, base words, strategies that can be broadly applied across
and suffixes) and rules for combining them, numerous examples, problems, and situations.
they can correctly spell many new words that
they have never encountered. Table 2.1 This mistaken notion that Direct Instruction
shows seven morphographs and some of the is a rote learning approach not only reflects a
words that can be correctly spelled by using fundamental misunderstanding of the
rules to combine them. The Direct approach but also fails to recognize that so-
Instruction program, Spelling Mastery, teaches called higher order thinking depends on the
750 morphographs that can be combined to mastery of more basic skills and involves the
form over 12,000 words. (This program is integration of concepts, rules, and strategies.
described in detail in Chapter 6.) Virtually all Direct Instruction programs con-
cern higher order thinking skills: classifying,
These examples from reading and spelling learning rules, making inferences, testing
illustrate the goal and importance of content generalizations, analyzing arguments, and
analysis to Direct Instruction. Direct solving problems. Carnine and Kameenui
Instruction is about teaching strategies that (1992) have described how the principles of
enable students to go beyond the particular design have been applied to teach sophisti-
items that are taught and to apply their learn- cated problem-solving skills to a variety of
ing to new items or situations. learners and across various domains. As the

Table 2.1
Seven Morphographs and Some of the Words Derived From Them

Prefixes Bases Suffixes

re cover ed
dis pute able

Words Formed

recover, recoverable, recovered, unrecoverable, unrecovered, repute, reputable, reputed,

disreputable, disrepute, coverable, covered, uncover, uncoverable, uncovered, discover,
discoverable, discovered, undiscoverable, undiscovered, dispute, disputable, disputed,
undisputable, undisputed, etc.

Journal of Direct Instruction 77

American Federation of Teachers (1998a) friendly” instruction. In order to teach a gen-
noted, although the early mastery of basic eral case, it is necessary to show students a
skills is a key element, Direct Instruction set of items that includes examples and
programs also address students’ general com- nonexamples arranged so that similarities and
prehension and analytic skills. differences are readily apparent. Irrelevant
aspects of the teaching must be held constant
Clear Communication. Identification of gen- to minimize confusion, and relevant aspects
eralizable strategies that students can use to must be carefully manipulated to demonstrate
solve a wide variety of problems is the founda- important differences. Engelmann and
tion of Direct Instruction. The first step of Carnine (1982) developed five principles for
building on this foundation is designing a sequencing and ordering examples to commu-
sequence of instruction that communicates
nicate clearly:
these strategies and enables students to dis-
play generalized skills to the full range of
1. The wording principle. To make the sequence
appropriate situations. Becker, Engelmann, and
as clear as possible, we should use the
Thomas (1975) and Engelmann and Becker
same wording on all items (or wording that
(1978) called this “general case programming”
is as similar as possible). This wording
because the goal is to teach the general case
helps focus students’ attention on the
rather than to teach a set of discrete specific
cases. General case programming is the design details of the examples by reducing dis-
of instruction that clearly communicates one traction or confusion that may be caused
and only one meaning and enables students to by variations in teacher language. Figure
exhibit generalized responding. 2.1 shows a pair of items that follow the
wording principle; teachers use nearly the
General case programming is based on princi- same wording for the two items. The fig-
ples for the logical design of teaching ure also shows a pair of items that does not
sequences (Engelmann & Carnine, 1982). It follow the wording principle; teachers add
enables Direct Instruction program develop- potential confusion by excessive variation
ers to design effective and efficient “learner in their wording.

Figure 2.1
The wording principle.

Following the wording principle Not following the wording principle

3 2 3 2
— — — —
2 3 2 3

The larger number The smaller number The larger number In this ratio state-
is on top. is on top. is on top. ment, the denomina-
tor is greater than the

78 Summer 2003
2. The setup principle. Examples and nonexam- lessons, we would demonstrate that the
ples selected for the initial teaching of a con- concept on holds for all objects and surfaces.
cept should share the greatest possible
number of irrelevant features. In Figure 2.2 3. The difference principle. In order to illustrate
the pair of items on the right does not follow the limits or boundaries of a concept, we
the setup principle. The two items differ in should show examples and nonexamples
several ways, so there are many possible that are similar to one another except in the
interpretations. Naive students might think critical feature and indicate that they are
that the label on means ‘rectangle’ or ‘things different. The difference principle is most
with corners.’ It might mean ‘gray.’ It might effective when the items are juxtaposed—
mean ‘horizontal.’ Or, it could mean ‘on.’ that is, they are shown next to each other or
Any of these interpretations is possible, and consecutively in a series—making the simi-
larities and differences most obvious. In
there is no way of determining which inter-
Figure 2.3, the juxtaposed items on the left
pretation students will make. From a Direct
side follow the difference principle. The
Instruction perspective, this ambiguity is
nonexample (not horizontal) is highly simi-
considered poor communication.
lar; it is just different enough to change a
positive example of the concept (horizon-
The pair on the left of Figure 2.2 follows
tal) into a negative example of the concept
the setup principle. The items are exactly
(not horizontal). In the pair that does not
alike except in the critical aspect of being follow the difference principle, the item
(or not being) on. The other interpretations that is not horizontal is quite different.
(rectangle, having corners, gray, horizontal) Failing to follow the difference principle
are eliminated because these features are leaves students with limited information
shared by both the positive and negative about the point at which an example is no
examples. This pair of positive and negative longer horizontal. Students might assume
examples differs in a single feature, so only that an object must be quite tilted in order
one interpretation is possible. In later les- to be not horizontal.
sons, additional examples would be used to
further expand the range of the concept. 4. The sameness principle. To show the range of
For example, by changing the setup (that is, variation of the concept, we should juxta-
by using different materials) in subsequent pose examples of the concept that differ

Figure 2.2
The setup principle.

Following the setup principle Not following the setup principle

This is on. This is not on. This is on. This is not on.

Journal of Direct Instruction 79

Figure 2.3
The difference principle.

Following the difference principle Not following the difference principle

The line is The line is not The line is The line is not
horizontal. horizontal. horizontal. horizontal.

from one another as much as possible yet applies to examples that are very similar to
still illustrate the concept and indicate that those shown. Thus, students may not show
they are the same. This sequence is generalized responding to the full range of
intended to foster generalization to unfa- possible examples.
miliar concept examples that fall within
the demonstrated range. In Figure 2.4, the 5. The testing principle. To test for acquisition,
set of examples on the left demonstrates we should juxtapose new, untaught exam-
the sameness principle by presenting a ples and nonexamples in random order. The
sequence of examples that are greatly dif- left side of Figure 2.5 shows an unpre-
ferent from one another, but are treated dictable order that provides a good test of
the same; that is they are all called dog. students’ understanding of the concept of
The set of examples on the right does not improper fraction. The right side of the fig-
show the possible range of variation. ure shows an alternating order. This order
Presenting students with a set of examples could be predictable; it is possible for stu-
that are very similar to one another may dents to get all answers correct simply by
suggest to them that the label dog only responding yes or no in accordance with the

Figure 2.4
The sameness principle.

Following the sameness principle Not following the sameness principle

example shown example shown

“This is a dog.” Chihuahua “This is a dog.” Cocker Spaniel
“This is a dog.” Irish Wolfhound “This is a dog.” Beagle
“This is a dog.” Cocker Spaniel “This is a dog.” Fox Terrier

80 Summer 2003
pattern. Therefore, it is not a good test For example, suppose that a group of students
because teachers could receive inaccurate is learning the strategy for reading words that
information about students’ understanding. end with the pattern of a vowel followed by a
consonant, followed by the letter “e” (VCe
Instructional Formats. After the concepts, words) such as rate, note, and slope. The main
rules, and strategies have been identified and difficulty of reading these words is to say the
sequences for clear communication of the gen- long sound for the medial (middle) vowel. In
eral case have been outlined, then instructional order to know when to say the long sound for
formats are constructed. A format specifies the the vowel, students must distinguish these
way that teachers will present each example, words from words that end with the pattern of
explanations that they will give, questions that a vowel followed by a consonant (VC words)
they will ask, and corrections that they will use. such as rat, not, and slop. The reading program
Formats are carefully designed to be clear and could use a format like the one shown in
concise, to help students focus on the impor- Figure 2.6 (format 1). This format would be
tant aspects of items, to provide appropriate used with many examples of words that end
support for students’ developing skills, and, with a VCe pattern (e.g., rate, slope) and a VC
above all, to communicate clearly with stu- pattern (rat, slop).
dents. The consistency of wording helps stu-
dents focus on the content to be learned rather Formats change as students become more pro-
than on irrelevancies such as how teachers are ficient. Initially, formats include a great deal of
asking for a response. This consistency is also structure and support for students’ use of
very helpful to teachers as it allows them to use skills. Format 1 in Figure 2.6, for example,
very effective, well-designed, and precise lan- gives students strong support in use of the
guage to communicate clearly with all students. VCe rule. This support is important to ensure

Figure 2.5
The testing principle.

Following the testing principle Not following the testing principle

2/4 Is this an improper fraction? 4/3 Is this an improper fraction?

3/5 Is this an improper fraction? 3/5 Is this an improper fraction?
8/5 Is this an improper fraction? 8/5 Is this an improper fraction?
48/32 Is this an improper fraction? 15/32 Is this an improper fraction?
18/12 Is this an improper fraction? 18/12 Is this an improper fraction?
6/7 Is this an improper fraction? 6/7 Is this an improper fraction?
9/3 Is this an improper fraction? 9/3 Is this an improper fraction?

Note the alternating order:

yes, no, yes, no, yes, no, yes

Journal of Direct Instruction 81

Figure 2.6
A series of formats for teaching students to read words that end VCe.

Format 1.

1. Teacher: Remember, when there is an ‘e’ on the end, this letter (point to it) says its name.

2. Teacher: Is there an ‘e’ on the end? Students: Yes.

3. Teacher: Will this letter (point) say its name. Students: Yes.

4. Teacher: What is its name? (Or what sound will it make?) Students: a.

5. Teacher: So what is the word? Students: rate.

Repeat Steps 2 through 4 for each of the following words: name, not, vote, rat, him, fine.

Format 2.

1. Teacher: Is there an ‘e’ on the end? Students: Yes.

2. Teacher: What sound will this letter make? Students: a.

3. Teacher: So what is the word? Students: rate.

Repeat Steps 1 through 3 for each of the following words: name, not, vote, rat, him, fine.

Format 3.

1. Teacher: What sound will this letter make? Students: a.

2. Teacher: So what is the word? Students: rate.

Repeat Steps 1 and 2 for each of the following words: name, not, vote, rat, him, fine.

Format 4.

1. Teacher: What is the word? Students: rate.

Repeat Step 1 for each of the following words: name, not, meat, first, boy, turn.

Format 5.

Students encounter VCe words in story reading with no additional assistance.

82 Summer 2003
a high level of success when strategies are ini- the end of the instruction, students apply the
tially introduced. However, formats must grad- skill without any prompts.
ually be modified so that the students learn to
apply the skills independently. If teachers con- Shift from massed practice to distributed practice.
tinued to use this format indefinitely, some Initially, students learn a new skill best when
students would come to depend on the they have many practice opportunities in a
sequence of questions to apply the rule and short period of time. In later learning, reten-
would falter when they encountered new tion is enhanced by practice opportunities
examples of VCe words in story reading. conducted over a long period of time. Thus,
formats begin with massed practice and
The support that is so important during initial progress to distributed practice.
instruction must be gradually reduced until
students are using the skill independently, Shift from immediate feedback to delayed feedback.
with no teacher assistance. The process of fad- Early in an instructional sequence, teachers
ing the format from highly supportive to highly provide immediate feedback to encourage stu-
independent is shown in the series of five for- dents and to provide them with immediate
mats in Figure 2.6. In the early stages of information about the accuracy of their
instruction of a particular strategy, teaching is responses. As students become more capable
highly teacher directed. However, by the com- and confident, feedback is increasingly
pletion of the instructional program the stu- delayed to create a more natural situation.
dents’ performance is independent, widely
generalized, and applied in various contexts Shift from an emphasis on the teacher’s role as a source
and situations. Becker and Carnine (1980) of information to an emphasis on the learner’s role as a
described six “shifts” that should occur in any source of information. Initially, teachers model
well-designed teaching program to facilitate new skills and provide very explicit instruction
this transition. in concepts, then later they fade out as the stu-
dents themselves become the source of infor-
Shift from overtized to covertized problem-solving mation on how to solve a problem.
strategies. Initially, formats assist students by
leading them through the steps of a strategy Taken together, these six shifts in instruction
out loud (overtly). Later, formats gradually constitute a coherent system for providing suf-
shift to allow students to complete the strat- ficient support to ensure initial success with
egy “in their head” (covertly). learning and applying complex strategies and
skills, then maintaining a high level of success
Shift from simplified contexts to complex contexts. as students systematically move to independ-
Formats for introducing each skill use a simpli- ent, generalized, real-world application of
fied context so students can focus on the criti- strategies and skills.
cal new learning. Later, formats include
increasing complexity. By the end of instruc- Sequencing of Skills. The sequence in which
tion on a skill, students should be applying it skills are taught in an instructional program is
in a natural and complex context. another important contributor to its success.
Learning can be made more or less difficult for
Shift from prompted to unprompted formats. In the students depending on the order in which
early stages of instruction, formats include skills are taught. The key principle is that stu-
prompts to help focus students’ attention on dents should be well prepared for each step of
important aspects of the item and to increase the program to maintain a high rate of success.
their success. These prompts are later system- That is, instructional programs should set stu-
atically removed as students gain a skill. By dents up for success. Direct Instruction uses

Journal of Direct Instruction 83

four main guidelines for deciding the order, or 27, while introduction of /b/ is delayed until
sequence, of skills. Lesson 121. Thus, 94 lessons separate the
introduction of these two very similar
First, prerequisite skills for a strategy should sound–symbol correspondences.
be taught before the strategy itself. Students
learn strategies most easily when they have Track Organization. Traditional programs
already mastered the components or prerequi- are typically organized in units where skills
sites of that strategy. For example, students and strategies are introduced, practiced, and
will learn column addition most easily if they tested within a limited period of time. For
have already mastered basic math facts. example, a math program may have a unit on
adding fractions with different denominators.
Second, instances consistent with a strategy In this unit, there may be a great deal of work
should be taught before exceptions to that on finding common denominators and adding
strategy. Students learn a strategy best when the numerators. But after this, when students
they do not have to deal with exceptions. go on to the next unit (perhaps on multiplying
Once students have mastered the basic strat- fractions), practice on adding with different
egy, they should be introduced to exceptions. denominators often ends suddenly.
For example, when the VCe rule is first intro- Information in one unit is seldom incorporated
duced, students apply the rule to many exam- into subsequent units. This lack of incorpo-
ples (e.g., note) and nonexamples (e.g., not). rated information results in predictable errors
Only when they are proficient with these when students (a) forget to watch for different
kinds of words will they be introduced to denominators when adding fractions, (b) for-
exception words (e.g., done). get how to find common denominators, and
(c) confuse the multiplication procedure with
Third, easy skills should be taught before the addition procedure. In contrast, tracks
more difficult ones. Students are more likely rather than units, provide the organizational
to experience success if they begin with tasks framework for all Direct Instruction programs.
that are easier to accomplish. For example, Tracks are sequences of activities that teach a
some sounds are easier to produce than others. skill across multiple lessons. Each lesson con-
Easy sounds (such as /a/, /m/, and /s/) are tains activities from several tracks. This way,
taught before more difficult sounds (such as Direct Instruction can extend teaching and
/t/, /d/, /p/) are introduced. (Note: When a let- practice of a skill across many lessons and
ter is enclosed in slashes [e.g., /a/] it refers to weave prerequisite skill tracks into the tracks
the sound of the letter. Thus, /a/ refers to the that integrate these skills into more complex
first sound in at.) strategies.

Finally, strategies and information that are Figure 2.7 shows the scope and sequence chart
likely to be confused should be separated in for Connecting Math Concepts Level C. The hori-
the sequence. The more similar things are, the zontal rows show skill development tracks and
more likely it is that students will confuse the vertical lines show lessons. For example,
them; therefore, items that are most confus- Lesson 1 includes activities from the tracks on
able should not be introduced together. For Addition and Subtraction Number Families,
example, the symbols b and d look very similar, Addition Facts, Place Value, and Column
and they make sounds that are very similar. Addition. Lesson 30 includes Addition and
Therefore, students are likely to confuse these Subtraction Number Families but does not
two letters. In the Direct Instruction begin- include Addition Facts. The Addition Facts
ning reading program, Reading Mastery Plus track has been completed at this point and is
Level 1, the sound /d/ is introduced in Lesson folded into the tracks on Column Addition,

84 Summer 2003
Estimation, and applications such as Analyzing several instructional tracks, so each lesson pro-
Data: Tables. As shown in this scope and vides instruction and practice on multiple con-
sequence chart, no Direct Instruction lesson is cepts, rules, and strategies. In essence then,
devoted to a single topic. Instead, each lesson each lesson is composed of a collection of
consists of activities that develop skills from mini-lessons on a variety of objectives.

Figure 2.7
Scope and sequence chart from Connecting Math Concepts Level C.

Lessons 1 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120

Addition and Subtraction
Number Families
Addition Facts
Subtraction Facts
Multiplication Facts
Division Facts
Mental Arithmetic
Number Relationships
Place Value
Column Addition
Column Subtraction
Column Multiplication
Division With Remainders
Calculator Skills
Equation Concepts
Problem Solving
Analyzing Date: Tables
Coordinate System
Statistics: Range

Connecting Math Concepts, Level C places a strong emphasis on higher-order thinking. Students learn a variety of
mapping techniques for relating problem solving to real life situations. With word problems, measurement, money, time,
and various projects, students graphically represent information before they attempt to calculate an answer. The
detailed instruction leads both teachers and students to develop positive feelings about problem solving.

In addition, instruction covers place value, geometry, estimation, calculator use, and statistics. Concepts and computa-
tion skills are also taught for regrouping, multiplication, division, and fractions. The Scope and Sequence Chart shows
where each track or major topic begins and where it ends.

Journal of Direct Instruction 85

There are numerous advantages to designing Instruction teachers organize students into
programs in tracks. First, student attention is groups to best meet the needs of each individ-
better maintained because they do not work ual. Second, Direct Instruction teachers allo-
on a single skill for an extended period. cate sufficient time for teaching and assure
Instead, lessons are made up of relatively short that the time is used well. Third, Direct
exercises that call on a variety of skills. Instruction teachers implement precise and
Difficult tasks are interspersed among easier careful plans for instruction through the use of
ones. Newly introduced skills are mixed with a scripted presentation. Fourth, Direct
well-practiced ones. Each lesson includes a Instruction teachers engage in continuous
variety of skills, formats, and difficulty levels. assessment of student performance.
This variety can be seen in the scope and
sequence chart by scanning down a line and Instructional Grouping. Every teacher faces
noting how many different tracks are touched choices about how to group students for
in a single lesson. Second, skills can be intro- instruction. Teachers may teach to the entire
duced and developed gradually over a number class or may arrange the class into smaller
of lessons. Each lesson can include a very small instructional groups. If teachers use smaller
step in the development of the skill because groups they must decide how many groups to
skills may be developed across many lessons. create and which students should be in each
Note, for example, that the track on Analyzing group. The principle that guides grouping in
Data: Tables extends across 105 lessons. This Direct Instruction is that each student should
track development provides the time neces- receive instruction that is appropriate to his or
sary to elaborate the strategies gradually. her individual needs. That is, students should
Third, practice can be massed within a lesson be placed at a level where they have the nec-
to promote learning and distributed across les- essary prerequisite skills and have not yet mas-
sons to promote retention. Students receive a tered the objectives. The skills that are to be
sufficient number of examples in each exercise taught are close to those that students have
so that they can master each step in the already learned but somewhat beyond their
sequence. In addition, practice is distributed current competence. Psychologists refer to this
over a substantial period of time. Organizing as the student’s “zone of proximal develop-
programs in tracks also makes it possible to ment” (Vygotsky, 1997). To enable all students
integrate information. In Direct Instruction to participate in instruction that is well suited
programs, no skill is ever introduced and then to their individual needs, Direct Instruction
dropped. However, some tracks are discontin- teachers organize their class into groups of stu-
ued as the skills in that track are incorporated dents who have similar zones of proximal
into other tracks. For example, when the track development. This type of grouping enables
on Addition Facts ends, other tracks such as teachers to present instruction to the group, to
Column Addition provide ongoing practice in interact with the group, and to address the
these facts. needs of all the individuals in that group.

Of course, each student has individual

Organization of Instruction strengths and needs. Therefore, students who
are in the same group for reading may not be in
In addition to program design whereby the the same group for math; their placement in
characteristics are embodied in the written each subject depends on their needs in each
program, we turn to the second major compo- subject. In constructing groups, we are less
nent of Direct Instruction: how the teacher concerned with students’ general achievement
organizes instruction. There are four key ele- or broad cognitive skills than we are with their
ments to organizing instruction. First, Direct mastery of the specific skills that are prerequi-

86 Summer 2003
site to a given lesson and the particular skills (e.g., Rosenshine & Berliner, 1978). Thus,
that are taught in that lesson. A placement test Direct Instruction teachers must allocate suffi-
within each Direct Instruction program (or cient time in their schedule for teaching the
other placement guidelines) aids teachers in most important academic areas. Of course, it is
forming appropriate groups. The placement not sufficient to allocate or schedule time for
tests are designed specifically to identify stu- instruction; this allocated time must also be
dents’ performance on the key skills that are used efficiently. Direct Instruction teachers
important for them to be successful in the pro- will organize smooth transitions, have materi-
gram. The results of these tests indicate the als at hand, and develop efficient routines to
program level and lesson that is an appropriate maximize the time that is actually available for
starting place for students. However, no test is instruction. Teachers must ensure that stu-
perfect. Therefore, when teachers form groups dents are actually engaged in instruction dur-
based on placement test scores, they should ing the designated time. However, it is even
anticipate making adjustments when they see more important that students are engaged in
how students respond to the first several les- tasks they can perform with high levels of suc-
sons. Students who make no errors and appear cess. The time that students are engaged and
to be bored should be moved to more advanced have high success rates is called academic learn-
groups, and students who make many errors ing time and is one of the strongest predictors
should be moved to less advanced groups. of student achievement. In other words, we
must allocate sufficient time, then make sure
Even if students are placed into ideal groups that we use that time efficiently and make
at the beginning of the year, we expect stu- certain that students are involved in learning
dents to progress at different rates. Some stu- activities that they can perform successfully.
dents who were appropriately placed into less
advanced groups learn quickly and show that Scripted Presentation. When we attempt to
they would now be better served in a more create performances of great complexity and
advanced group. Conversely, other students we want consistently successful outcomes, we
may struggle to learn the material at the rate generally plan very carefully. For example, crit-
of other members of their group. Direct ical parts of space missions such as liftoff, dif-
Instruction grouping should be flexible to ficult operations in space, and reentry are
accommodate students’ changing needs. scripted in detail and practiced extensively. In
the theater, sophisticated drama with complex
This flexible skill grouping based on students’ characters and multiple levels of meaning are
instructional needs is very different from the scripted in advance and practiced thoroughly.
practice of “tracking” in which students are Casual planning and dependence on extensive
assigned to rigid, inflexible groups based on improvisation are simply not successful ways of
general characteristics such as “intelligence.” producing these complex results. Similarly,
Tracking is absolutely incompatible with Direct from a Direct Instruction perspective, teach-
Instruction because it does not allow for adjust- ing important and complicated skills such as
ment according to students’ changing needs. reading, math, and language arts requires care-
ful planning and precise implementation.
Instructional Time. An important factor in Therefore, Direct Instruction programs
determining how much students learn is the employ detailed scripts with carefully devel-
amount of time students are directly engaged oped explanations, examples, and wording.
with the material. Of course, this makes logical
sense to most people, but anyone who needs Scripts are tools designed to accomplish two
to be convinced can refer to a large amount of goals: (a) to assure that students access
research that demonstrates this simple fact instruction that is extremely well designed

Journal of Direct Instruction 87

from the analysis of the content to the specific Armed with these Direct Instruction scripts,
wording of explanations, and (b) to relieve the teacher’s role becomes similar to an
teachers of the responsibility for designing, actor’s. Actors have a critical role in delivering
field-testing, and refining instruction in every planned performances. They are the ones who
subject that they teach. One of the main breathe life into the words that are written on
premises that leads to the use of scripts is that the page. Without their skill and art in convey-
students deserve precisely planned instruc- ing the script, there is no drama. However,
tion. Some might argue that the level of depending on the actors’ inspiration to per-
thought about the content of instruction, the form is far different from asking them to go on
details of clear communication, and the careful stage and spontaneously create drama. Like
actors, Direct Instruction teachers are per-
sequencing of tasks that are embodied in
formers who put life into scripts. They relate
Direct Instruction scripts are not really impor-
to the students through the words in the
tant and that instruction that is largely impro-
scripts. These teachers are the source of
vised is sufficient. It is certainly true that warmth, excitement, and life in the presenta-
some students will master some of the objec- tion. They make the expected adjustments for
tives with instruction that is casually planned individual differences among students.
and loosely organized. But it is also true that Teachers are the only ones who can motivate
many students will fail to master many objec- students with praise and other feedback on
tives with casually planned instruction. their work. Teachers are also the only ones
Students who, for whatever reason, are most at who can adjust the pace to the needs of the
risk of learning difficulties are disadvantaged group, allowing more time for tasks that are
by instruction that is not carefully planned and difficult for a particular group and moving
well implemented. Flaws in instruction are more quickly through tasks that are easier.
reflected in students who have poor skills. Teachers must also play the critical roles of
Even those students who are capable of learn- problem solver and decision maker, identifying
ing from weak instruction can learn more if problems with student learning and adjusting
instruction is well planned. If we aspire to the instruction accordingly. These jobs are
reach all the students and teach all the objec- extremely demanding but are made much eas-
tives, we must plan instruction very carefully. ier if teachers are given excellent tools includ-
Careful and detailed planning of instruction is ing a well-designed scripted curriculum.
important to the degree that we value excel- Relieved of the instructional design role,
teachers can focus on the critical job of deliv-
lent educational outcomes.
ering instruction, adjusting it to the unique
needs of individual students, and solving unex-
As we might guess, planning lessons is
pected problems.
extremely time-consuming. Even for a team of
expert instructional designers, developing a
Continuous Assessment. It is important to
new math or language arts program is a daunt- monitor students’ progress toward program
ing task. For an individual teacher who has objectives continuously. All Direct Instruction
several hours of planning time each day after programs include various types of ongoing in-
school and must prepare lessons for several program assessments. These assessments pro-
subject areas, it is simply impossible to pro- vide teachers with feedback on the
duce instructional plans that meet these high effectiveness of their teaching and allow them
standards. Scripts can provide carefully devel- to evaluate the adequacy of their students’
oped, field-tested, and detailed plans for skill development. Data provided by these
teachers. However, as with any plan, there are assessments can be used to make critical
limits to what scripts can accomplish. instructional decisions. If progress is inade-

88 Summer 2003
quate, teachers need to adjust instruction. There are seven components for promoting
They may determine that some students are effective teacher–student interactions: active
inappropriately placed and regroup students student participation, group unison respond-
accordingly. They may develop additional ing, signals, pacing, teaching to mastery, cor-
instruction and practice for students who have rection procedures, and motivation.
not yet mastered a particular skill.
Active Student Participation. Students learn
On the other hand, students may perform best when they are actively engaged with the
above the specified criterion on these meas- material. Active engagement is important for
three reasons. First, and most obviously, stu-
ures. Teachers may elect to skip lessons when
dents learn when they interact with the
data indicate that a group of students is capa-
instructional material and receive relevant
ble of moving at a faster pace. Or, teachers
feedback. More interaction and more feedback
may find that some students in a group are result in more learning. A student who reads
able to move at a faster pace and may elect to 20 words and receives feedback on each will
change those students’ group placement. All tend to learn more than a similar student who
decisions—from initial placement and group- reads only 5 words. Thus, actively responding
ing, to acceleration of instruction—are made to a large number of relevant items would be
based on students’ assessment performance expected to increase learning directly.
not on “hunches.”
The second reason for maximizing engage-
ment has to do with the pragmatics of the
Teacher–Student Interactions classroom. When students are engaged, they
are less likely to become distracted and to dis-
Direct Instruction defines the teacher’s role tract others. Therefore, active engagement
more clearly and explicitly than most other reduces time that would otherwise be devoted
forms of instruction. Scripted programs relieve to management of behavior problems. In this
teachers of the role of instructional designer. way, active engagement can actually increase
Instead, their role is to deliver instruction in a the time available for teaching. (Martella and
way that is effective and motivating to the par- Nelson, in press, provide further information
ticular group of students and to make the criti- on behavior management issues related to the
cal decisions about how the program should be teaching of Direct Instruction.)
adapted to the needs of the particular group.
This role emphasizes (a) knowing the stu- The third reason to maximize active engage-
dents as individuals and creatively motivating ment involves knowledge of student skill lev-
them through presentation of the script and els. When teachers have an excellent
understanding of each student’s current level
by adding motivational systems that are appro-
of mastery, they can make the best decisions
priate to the particular group, and (b) knowing
about instruction. Ideally, they would have
the students’ ever-changing skills and adjust-
very rich information on students’ skills in
ing the pacing of lessons, amount of practice, order to make well-informed decisions. When
and other factors according to their needs. we consider these three reasons for active
These roles emphasize problem solving and engagement, it becomes clear why active
creativity. However, this creativity is not engagement is one of the centerpieces of
unstructured and undirected. It is creativity Direct Instruction.
within the context of well-conceived lessons
and with the clear goal of enhancing learning Group Unison Responses. There are many
and motivation. ways to organize active student engagement.

Journal of Direct Instruction 89

One of the most common is to call on individ- dents and have them all write the answer. This
ual students to answer questions orally. If the technique can be very useful if written answers
items are relevant and the questions well are appropriate and students have strong writ-
designed, oral responses can give individual ing skills. However, many students we teach do
students practice, keep them attentive, and not yet have strong and fluent writing skills,
give teachers immediate information on indi- and much of the content does not lend itself to
vidual student skill levels. However, individual written answers. In addition, teachers must cir-
oral responses also have several limitations. culate around the group very quickly to assess
While the teacher is interacting with one stu- the skills of all the students. If they do not cir-
dent, other students may not be paying atten- culate quickly and check answers as students
tion. Each question that is directed to a single write them, then they will not have the high-
student may constitute down time for the quality assessment information that is one of
other students and not promote active engage- the goals of student engagement.
ment. In addition, with individual questions,
the teacher receives information about only Another alternative, one that is often (though
one student at a time. It is possible that the not always) employed in Direct Instruction, is
student who answered the question is the only to ask all the students to answer orally in uni-
one who understood the material. This possi- son. This responding is often described as
bility is even greater if the teacher calls on vol- choral responding because it is similar to a choir
unteers. Students who do not know the singing in unison. If students answer in uni-
answer are the least likely to volunteer. Calling son, then (a) all students get high-quality
on volunteers may give the teacher a distorted practice on every item because they provide
picture of the group’s performance. their own response and cannot echo other stu-
Additionally, it directs the response opportuni- dents, (b) all students are busy learning the
ties to students who are the most skilled and material and are less likely to become dis-
away from those who are most in need of tracted, and (c) teachers can assess the skills
active engagement. of all the students in an instant and be well
informed about their skills. If teachers can
In order to provide practice and to assess many orchestrate group unison responses, they can
students, teachers can provide numerous indi- greatly increase students’ opportunities to be
vidual questions. However, this questioning engaged with the content and greatly increase
leads to a great deal of down time for stu- their understanding of each student’s skill
dents. If there are 10 students in a group, each level. Group unison responses are highly effi-
student may waste nine-tenths of the time cient. Suppose a teacher has a group of 10 stu-
devoted to individual questions. In addition, if dents and he can ask 10 questions per minute.
teachers repeatedly ask a small set of ques- If he asks all individual questions, each stu-
tions, then the first students to answer the dent makes one oral response per minute. In
question get a high-quality opportunity to fig- contrast, if he asks group questions and gets
ure out the answer. Other students who unison responses, each student can make 10
respond to a question after having heard sev- responses per minute.
eral other students answer that same question
get lower-quality opportunities because they Group unison responses have some substantial
may simply be repeating what they heard from advantages; however, they should not be the
the other students. They may not have had a only form of response. Group unison responses
chance to figure it out for themselves. are useful when the answer is relatively short
and when all students would be expected to
An alternative way to organize student provide the same answer. For example, a
responses is to pose a question to all the stu- teacher might show students a fraction and

90 Summer 2003
ask, “Is this a proper fraction?” All students a single error in a group. However, if answers
would respond “No.” The teacher might then are dragged out or unsynchronized, it is
ask, “How do you know?” and the group would much more difficult to detect errors. As a
respond, “The top number is larger.” In both result, it is very important that teachers use
of these instances, the use of a group unison some system to enable all students to
response would be appropriate. On the other answer simultaneously.
hand, if the teacher gave a request such as,
“Give an example of an improper fraction,” we In many noneducational pursuits, people want
would expect students to give a variety of to coordinate their actions. In an orchestra or
answers so a group unison response would not choir, musicians watch the conductor for visual
be appropriate. For this item it would be best signals and listen to each other for auditory
to call on an individual or to ask all students to cues about when to begin. In football, a quar-
write an answer. terback usually gives auditory signals by
yelling. Coordination of our driving in traffic is
In addition, group unison responses should be mostly arranged by visual signals of traffic
followed by individual questions. Individual lights, signs, and (depending on where you are
turns provide information about the skill levels driving) other drivers’ turn signals. The com-
of different students to respond to a task mon element among these diverse examples is
when there is no support from the group. that we use various kinds of signals to coordi-
Individual turns are generally presented after
nate groups of people.
the group has been brought to mastery. When
teachers provide individual turns (usually sig-
The goal in signaling for a group unison
naled by teachers as “time for turns”), they
response is to enable all students to initiate
provide an instruction to all students and then
the answer at exactly the same time. In this
place an individual student’s name at the end
way, teachers gain precise information about
of this instruction. For example, “Read the
student performance that one-to-one instruc-
second row of words, James” or “Say these
tion permits, while still achieving the effi-
sounds, Sally” as compared to “James, read the
ciency of group instruction. Teachers hear a
second row of words” or “Sally, say these
sounds.” In this way, all students are prepared single response, are able to evaluate the
to answer the teacher until a specific student’s response, and can proceed accordingly. In
name is called. order for students to initiate a response at the
same time, we must use some kind of signal to
Signals. The group unison oral response is a coordinate their answers. Direct Instruction
very useful tool. However, if answers are not teachers use various signals, depending on the
quite in unison, if some students answer circumstances. For example, when students
slightly earlier or slightly later than the oth- are reading words from a textbook, they are
ers, or if students drone their answers, then not looking at the teacher. Therefore, an audi-
these group responses become much less tory signal such as a snap, tap, or clap is useful
powerful and may even be counterproduc- because it does not require students to look
tive. The problem is that if responses are away from their books. On the other hand, if
not crisp and in unison, then students who students are reading words that are written on
answer later may simply be echoing those a chalkboard, teachers may use a visual signal,
who answered earlier. Thus, they may not be such as pointing to the word, because they are
practicing the academic task but, rather, already looking at that word. In each Direct
practicing the skill of chiming in after other Instruction program, the teacher’s guide and
students. In addition, when responses are teacher presentation book specifies how to sig-
crisp and in unison, teachers can easily hear nal for each task.

Journal of Direct Instruction 91

Figure 2.8 illustrates the parts of a basic signal. makes, and then to sound out the word overtly
To signal a unison response during group with the appropriate vowel sound. This task
instruction, teachers provide (a) a focus cue to obviously will take several seconds when the
gain students’ attention to the task, (b) think skill is being acquired. If think time is too
time that varies depending on the skills of the short, students will not answer on signal or
students, (c) a verbal cue followed by a pause will make errors. As students become more
(interval), and (d) a signal. To focus the stu- proficient in application of a particular skill,
dents’ attention, teachers may point to an teachers reduce this think time.
item on the board or in a teacher presentation
book or may direct students to point to an Following think time teachers provide a verbal
item in their book. In giving directions, teach- cue such as “get ready” or “what word.” This
ers tell the students the type of response they verbal cue has teacher voice inflection built in
will make. For example, they may say, “Spell
(e.g., get READY) as illustrated by the curved
each word after I say it” or “Read these words
line under the verbal cue column in Figure 2.8;
the fast way.” Next, think time is provided.
this cue is immediately followed by a short
The length of think time depends on the diffi-
pause (interval). Right after this short pause,
culty of the task. If the task is relatively sim-
ple for students in the group, the think time teachers provide a signal. Teachers may indi-
may be very brief. For instance, after sufficient cate this signal by making a gesture (such as
practice with letter–sound correspondence, touching the board or dropping their hand) or
most students would need little think time to an audible cue (such as a clap or tap). It is crit-
respond to the question, “What sound does ical that teachers do not “talk and move” at the
this letter make?” However, for more difficult same time; that is, teachers should not overlap
tasks, teachers need to provide more think what they say with their signal. The interval
time. When asked to read words that follow after the verbal cue and before the signal
the VCe pattern, students need sufficient should be maintained. Students are to come in
time to determine if the word ends in an e to “on signal”; that is, they respond when teach-
think about what sound the medial vowel ers snap their fingers or tap the board.

Figure 2.8
Parts of a basic signal.

Focus Cue Think Time Verbal Cue Interval Signal

Point to task “Get ready” Snap

Ask question “What word”
Give direction “What sound”
Hand drop

Note. Illustrated by Tracey Hall.

92 Summer 2003
Signals tell students when to answer and pro- prepared they will find a task easy, even “natu-
vide all students the opportunity to partici- ral.” However, those same students would find
pate. The signal is a tool that enables all that same task extremely difficult if they were
students to be actively engaged with instruc- not well prepared for it. This simple, even
tion and gives teachers the opportunity to self-evident, logic is the basis of the Direct
monitor student responses and adjust their Instruction principle of teaching to mastery.
instruction accordingly. Mastery involves performing skills at high lev-
els. Engelmann (1999) likens mastery to a
Pacing. Active student engagement is further stairway: “Mastery is the guarantee that stu-
enhanced when teachers maintain a brisk pace dents are able to reach each stair without
in their teaching. Brisk pacing of instruction is falling” (p. 4). Effective teachers carefully
important for several reasons. First, a rapid design instruction around this goal.
pace allows teachers to cover, and students to
learn, more material (Brophy & Good, 1986). Direct Instruction programs are designed to
Second, a brisk pace holds student attention prepare students for each new challenge and
and reduces time between related information, set the students up for success. If students
thereby enhancing student learning. When we have mastered the skills taught in Lessons
speak too slowly, we can actually be harder to 1–80 of a Direct Instruction program, they will
understand, especially by distractible children. be well prepared for Lesson 81. However, if
Third, well-paced instruction keeps students students are weak on the tasks from the previ-
engaged and, in turn, reduces behavior prob- ous lessons, then Lesson 81 will be more diffi-
lems. Inappropriate behavior often occurs dur- cult. Therefore, we can make the program
ing down time when students are not easiest for students and enhance their success
occupied with productive tasks. Engelmann by bringing them to mastery on every lesson.
and Becker (1978) reported that when teach- Some teachers are tempted to reduce their
ers maintained a pace of about 12 responses mastery standards for groups of students who
per minute, students answered correctly about are struggling. We often hear that a group’s
80% of the time and were off-task only 10% of performance is, “pretty good for the low
the time. However, when teachers asked only group.” The problem is that students who
four questions per minute, the students’ accu- tend to struggle are the very students who
racy dropped to 30% and they were off-task most benefit from mastery and are most disad-
about 70% of the time. Clearly, a brisk pace vantaged when lessons are made more diffi-
contributes to the effectiveness of instruction. cult. Thus, from a Direct Instruction
perspective, it is very important to assure that
Proper pacing is a difficult teaching technique every group reaches mastery on every lesson;
to master. The pace should be relatively quick, mastery is particularly crucial for students who
but must give students sufficient think time. struggle with the material.
Experienced Direct Instruction teachers
become very sensitive to the demands of the Mastery should generally be gauged by the
task and the skills of the individual students performance of the lowest performing student
and adjust their pace accordingly. Finding an in the group. If that student has mastered the
appropriate pace is an important and subtle material, we can assume that others in the
skill, one that is learned from experience and group have as well. It also means that the stu-
close observation of students’ learning. dent who is most in need of support will not
be further disadvantaged on the next lesson.
Teaching to Mastery. The difficulty of a Engelmann (1969) advised that we “seek flat-
learning task depends on how well students tery from the lowest-performing children.
are prepared for it. When students are well When they succeed, the teacher should indeed

Journal of Direct Instruction 93

feel that she has received the highest form of information that enables students to answer
professional compliment” (p. 46). questions correctly. In addition, during group
responses, corrections are presented to the
Four criteria allow precise interpretation of entire group. The individual student who
how students respond during lessons as noted makes an error is never singled out. All stu-
by Engelmann (1999): dents can benefit from the additional prac-
tice provided by the correction procedure. If
1. Students should be at least 70% correct on the errors occur in individual responses, cor-
information that is being introduced for the rections are typically directed to the student
first time. (If they are only at 50%, they are who responded.
at chance levels and are probably guessing.)
The basic correction procedure for student
2. Students should be at least 90% correct on response errors in Direct Instruction programs
skills taught earlier in the program (assum- consists of reteaching and retesting students.
ing previous skill mastery). Figure 2.9 shows the basic Direct Instruction
correction procedure of model–test–retest.
3. At the end of a lesson, all students should
be “virtually 100% firm on all tasks and Immediately after the error, teachers (a)
activities” (p. 6). demonstrate the correct answer (model), (b)
ask the students to respond to the original
4. Student error rates should be low enough to item (test), and (c) give several other items,
ensure that teachers have sufficient time to then retest the item that was missed (retest).
complete a lesson. Each step in this correction procedure has a
specific purpose. The model step clearly com-
Teaching to mastery involves closely monitor- municates what students should do. The test
ing student performance and making appropri- step assesses whether the model was effective
ate adjustments. and students can now respond correctly to
that item. However, because the test immedi-
Correction Procedures. Effective instruction ately follows the teacher’s model, it does not
requires effective correction procedures. While show whether students can answer correctly
Direct Instruction is designed to minimize stu- and independently. Thus, the retest step (also
dent errors, mistakes are inevitable when stu- called “starting over”) is crucial. The retest
dents are acquiring new information. Teachers comes after several other items. If students
must notice every error, determine the type of can make the correct response after these
error that was made, provide an appropriate other items, then teachers have greater confi-
correction, and arrange for additional practice dence that students have learned the
on items of that type. Without effective correc- response. In addition to these basic steps,
tions, learning is difficult or impossible. teachers may decide to use delayed tests of
missed items after completing the exercise or
Many different types of correction procedures lesson. Delayed tests may be given at varying
are used in Direct Instruction. The particular intervals throughout the day and on subse-
correction procedure used depends on the quent days to ensure that students remember
teachers’ diagnosis of errors. However, two important information.
features characterize all corrections in Direct
Instruction: (a) They are immediate, and (b) In some situations, teachers may add steps to
They are direct. Teachers should correct mis- the basic correction procedure. Chapters 3–8
takes immediately when they occur. provide specific correction procedures for the
Corrections explicitly and directly provide various academic programs (presented under

94 Summer 2003
Teaching Techniques in the chapters). For response by asking students to, “Say it with
example, if students err in applying an explicit me,” and respond with students on a lead
rule, teachers can replace the model step with step. Assume that students made a mistake
a rule. If students misread the word note, (say- when asked to say the sounds in sat without
ing not) teachers could assist students in stopping. Figure 2.10 illustrates an appropriate
applying the rule for reading VCe words by correction procedure that includes a lead step.
asking, “Is there an ‘e’ on the end of the Teachers may lead (respond with the stu-
word?” followed by, “So what do you say for dents) several times and must monitor the
this letter (pointing to the letter ‘o’).” They students’ responses closely by watching and
then proceed with the test and retest. listening. Only when teachers are confident
that students can produce the response with-
Students may make errors because they have out their assistance are students asked to
trouble producing the response. For example, respond independently.
students may have difficulty making the sound
for a particular letter, saying the sounds in a The teacher’s guide and teacher’s presenta-
word without pausing between sounds, or tion books for all Direct Instruction programs
reciting a list such as days of the week. In provide very detailed guidelines for effective
these situations, teachers should add a lead correction procedures of common types of mis-
step after the model. They initiate the takes in that particular program. All of these

Figure 2.9
Steps in a basic correction procedure.

Step Teacher says Student says

Model This word is “eventually.”

Clear communication of what

students should do.

Test What word is this? eventually

Opportunity for students to

perform skill correctly.

Retest What word is this? treatments

Teacher intersperses several
other items before retesting What word is this? submarine
“eventually.” (submarine)

Gives students opportunity to What word is this? eventually

perform skill independently. (eventually)

Journal of Direct Instruction 95

corrections are variations on the basic correc- placement is appropriate and instruction is well
tion procedure of model–test–retest. designed and well delivered, students experi-
ence a high level of success. Classroom experi-
In addition to correcting student response ences that produce success are one of the
errors, teachers should also correct signal foundations for motivation. Thus, to maximize
errors. When signal errors occur it means that student motivation, we refer to the same
students did not answer together on signal. To instructional issues we have been concerned
correct this error, teachers might say, “I need with in maximizing student learning.
to hear everyone together” or “Everyone
should respond right at my signal” and repeat In a well-designed program, day-to-day success
the task (starting over). will result in continual learning and improve-
ment of skills. For students, the reward of see-
Motivation. In Direct Instruction, learning ing their own improvement can powerfully
and motivation are seen to be closely related. support motivation. Learning, of course, has
Motivation begins with success, and success other natural rewards. Learning basic language
requires motivation. The experience of success skills results in communicating more effec-
is one of the most important bases of motiva- tively, opening vast possibilities. Learning to
tion in the classroom. Thus, motivation begins, read offers the great reward of access to litera-
as instruction does, by appropriate placement. ture as well as the social rewards of reading
Placement in excessively difficult material such as being complimented by one’s parents.
results in failure and reduced motivation.
Placement in excessively easy material results Teachers play a key role in motivation. They
in boredom and reduced motivation. When arrange a classroom environment that results

Figure 2.10
Correction with lead step.

Step Teacher says Student says

Model My turn to say the sounds in

sat. sssaaat.

Lead Say it with me, sssaaat. sssaaat

Teacher and students say

the response together. The
lead may be repeated several
times if necessary.

Test Say the sounds in sat all by sssaaat


Delayed Test Say the sounds in sat. sssaaat

96 Summer 2003
in success for all students. They recognize with math facts. They may also make a point
that success and make it more apparent to stu- of recognizing student effort and persistence
dents. By frequently commenting on success in this area. Teachers can make student
and praising students for their efforts, teachers progress more obvious. For example, they may
amplify the effects of the success and add a teach students to graph their performance on
positive social element. Teacher recognition is certain skills or activities such as each day’s
a strong motivator for most students, but the math assignment. Beating one’s own best
effects of praise depend on the relationship score is often a powerful motivator.
between teachers and students as well as the
way in which teachers deliver praise. When These relatively simple techniques used con-
teachers have a warm and positive relationship sistently and thoughtfully are sufficient for
with their students, their praise will be more creating a positive, motivated, and productive
powerful. Also, if they are sincere, specific, classroom. However, this is not to claim that
and age-appropriate in their praise, the effect these techniques will eliminate all behavior
will be most powerful. management problems. When problems arise,
the first question for teachers should be
Admonishments, reprimands, nagging, and whether these basic motivation systems are in
other forms of attention given to undesirable place. They should ask whether students are
behavior should be minimized. Reprimands placed at an appropriate level and are experi-
and other forms of attention given to undesir- encing success in the program, and they
able behavior are generally ineffective. should ask whether students are aware of their
Madsen, Becker, Thomas, Koser, and Plager successes and are receiving sufficient recogni-
(1968) compared the effects of reprimanding tion for their efforts. This simple analysis can
students for being out of their seats versus unravel the reasons, and suggest solutions, for
ignoring them and praising students who were many behavior challenges. However, there will
in their seats and on-task. The authors con- still be challenges that require even more
cluded that, if the primary way that children focused analysis and intervention. Martella
get attention is by misbehaving, they actually and Nelson (in press) describe strategies for
misbehave more often. That is, teacher atten- working with a wider variety of classroom man-
tion, even though intended to reduce the agement techniques.
undesired behavior, may actually make it more
frequent. Thus, one of the basic slogans of
motivation in Direct Instruction is, “Catch
them being good.”
Direct Instruction
Much of the time the immediate rewards of
and Effective Teaching
success, learning, and recognition from the The practices that have been identified by the
teacher are sufficient to produce strong moti- “effective teaching” literature (described in
vation. However, when learning is harder, more Chapter 1 as effective instruction) are inte-
rewards may be required. Also, for a wide vari- grated into Direct Instruction. The organiza-
ety of reasons, some children are not suffi- tion of instruction in Direct Instruction
ciently motivated by these simple motivational includes a general academic focus with an
techniques. Thus, additional strategies are emphasis on maximizing engaged time and
necessary. These additional strategies may instruction in small interactive groups—all
include more focused praise. For example, if characteristics of effective instruction
teachers know that particular students are (Rosenshine & Stevens, 1986). Direct
struggling with math facts, they may be alert Instruction includes organizational elements
for situations in which those students succeed beyond those described in the effective teach-

Journal of Direct Instruction 97

ing literature. These elements include group- In a recent popular educational psychology
ing students with similar instructional needs text, Slavin (2003) states that “the research on
and scripted presentations. Direct Instruction direct instruction models has had mixed con-
student–teacher interaction practices such as clusions . . .” However, he also points out,
brisk pacing, high success rates, and explicit “Studies of Direct Instruction . . . a program
instruction, followed by guided practice and built around specific teaching materials and
independent practice with emphasis on mas- structured methods, have found strong posi-
tery of content, are all prominent recommen- tive effects” (p. 239).
dations from the effective teaching literature
(Rosenshine & Stevens). Direct Instruction
builds on these techniques by adding specific Students for Whom Direct
practices such as unison responding to further
increase active participation by students, sig- Instruction is Appropriate
nals to coordinate student answers, and spe-
cific recommendations for error corrections. Research has confirmed that Direct
Instruction has been effective for students
The most important way that Direct with diverse learning needs (including stu-
Instruction extends effective teaching is in dents in special education and general educa-
program design. Effective teaching does not tion), students with diverse language
deal with program design—it takes the pro- backgrounds, and students of all ages from
gram as a given and focuses on effective meth- preschool through adult.
ods for delivering the content. Direct
Instruction, on the other hand, is built on the Students With Diverse
foundation of instructional programs that Learning Needs
embody efficient strategies and carefully Students who are receiving special education
crafted explanations. This attention to what is services are particularly at-risk for academic
taught takes Direct Instruction beyond the failure. If these students are to be successful,
recommendations of effective instruction. they often require careful instruction in which
Thus, Direct Instruction is consistent with the details are carefully planned and well imple-
recommendations of the effective teaching lit- mented. Direct Instruction has been successful
erature and goes beyond it by further specify- in accelerating the achievement of students
ing teaching techniques and attending to the who receive special education services.
design of programs.
Even students who would be predicted to
Direct Instruction is often confused with the have low levels of achievement benefit greatly
more general techniques described in the from Direct Instruction. Gersten, Becker,
effective teaching literature. In fact, the term Heiry, and White (1984) examined the yearly
direct instruction (note the lack of capital let- achievement test profiles of students in
ters) is often used to refer to any form of Direct Instruction classrooms to determine
instruction involving direct interactions whether annual gains made by students with
between teachers and students. Many profes- low IQ scores differed significantly from the
sional educators and professional publications gains made by students with average or supe-
fail to distinguish between direct instruction, rior IQ scores.
which is a set of teacher practices for organiz-
ing instruction and interacting with students, Figure 2.11 shows the yearly gains made by
and Direct Instruction, which is an integrated students in reading as measured by the Wide
system of curriculum and instruction Range Achievement Test. As shown in this fig-
(Schaefer, 2000). ure, students with higher IQ test scores

98 Summer 2003
started at higher achievement levels and These results provide evidence that Direct
ended with higher levels than their peers with Instruction is appropriate for, and effective
lower scores. However, the pattern of growth with, a wide variety of individuals including
of students with low IQ scores is remarkably those with low IQ scores, those with IQ scores
similar to that of other students. The group in the average range, and those with high IQ
with the lowest scores (under 70) gained scores. In addition, because children in this
nearly as much each year in reading as stu- study were taught in small homogeneous
dents with much higher scores. By the end of groups (having students with relatively the
third grade, those students with the lowest IQ same skill levels), the gains of students with
scores were performing at the 70th percentile, lower IQ scores were not made at the expense
or a grade equivalent of 4.3. of other students nor the other way around.

The results are even more pronounced in Several reviews of research focusing on the use
math as seen in Figure 2.12. This figure shows of Direct Instruction with special education
the students’ performance on the populations have all converged on the finding
Metropolitan Achievement Test. The growth that Direct Instruction is measurably effective
rate for all groups of students corresponds to with these students. White (1988) reviewed
one grade equivalent for each year in school. 25 such studies and found that all comparisons

Figure 2.11
Results of Direct Instruction on reading as measured by the Wide Range Achievement Test
for students with diverse IQ scores. Adapted from Gersten et al. (1984).

100 n n n n
l n l l
90 n

80 n H
70 H n
H u u
n n n

50 ..............................................................................................................
n Average Score
10 H
Entering K K 1 2 3

IQ under 71 IQ 71–90 IQ 91–100 IQ 101–110 IQ 111–130 IQ over 131

n u H n l n

All testing was performed at end of academic year, except EK.

Journal of Direct Instruction 99

favored the Direct Instruction group. Forness, Engelmann and Carnine (1989) found that
Kavale, Blum, and Lloyd (1997) conducted an typical second graders who had received 2
analysis of various intervention programs for years of Direct Instruction scored an average
special education and determined Direct 4.6 grade equivalent in reading on a standard-
Instruction to be one of only seven interven- ized achievement test. The children’s average
tions with strong evidence of effectiveness. scores in science and math were 4.0 and 3.4,
respectively. Other researchers have arrived at
Perhaps because Direct Instruction programs similar findings. Tarver and Jung (1995) inves-
have been so successful with students who tigated the effects of a Direct Instruction
have failed in other instructional programs, math program (Connecting Math Concepts) and a
their use is commonly associated with children discovery learning math program on the math
who are behind, who are failing, or who are at- achievement and attitudes of general educa-
risk for failure. And some have questioned tion students in the primary grades. They
their appropriateness for general education. found that, at the end of second grade, the
However, Figures 2.11 and 2.12 provide direct children in the Direct Instruction program
evidence of the effectiveness of Direct scored higher on measures of math computa-
Instruction for students with IQ scores in the tion and math concepts than children in the
middle range and those in the upper range. comparison group. In addition, children in the

Figure 2.12
Results of Direct Instruction on math as measured by the Metropolitan Achievement Test
for students with diverse IQ scores. Adapted from Gersten et al. (1984).

80 l
Mean Standard Scores

70 n H
l n
n H
50 l n
40 H
30 n
1 2 3

IQ under 71 IQ 71–90 IQ 91–100 IQ 101–110 IQ 111–130 IQ over 131

n u H n l n

All testing was performed at end of academic year.

100 Summer 2003

Direct Instruction program had significantly Gersten, Taylor, Woodward, and White (1997)
higher scores on a survey of attitudes about described the evaluation of a 14-year imple-
math. Finally, Tarver and Jung reported that mentation of Direct Instruction in Uvalde,
the Direct Instruction program was equally Texas, whose population is 98% Hispanic. The
effective for lower and higher performing chil- authors concluded that the approach had a
dren who participated in the study. Other consistent, positive effect on the achievement
studies provide additional evidence that of language minority students. They reported
Direct Instruction programs accelerate the that achievement levels were at or near grade
learning of high-performing students in lan- level in math, reading, and written language
guage (Robinson & Hesse, 1981), reading for more than a decade. Scores in reading com-
(Schaefer, 1989; Sexton, 1989), and science prehension and vocabulary were at the 28th to
(Vitale & Romance, 1992). 31st percentiles. These scores are, according
to Gersten et al., “appreciably above typical
Students With Diverse Language levels for low-income Hispanic students” (p.
Backgrounds 37). Perhaps more importantly, follow-up stud-
Children who have no English oral language are ies conducted 2 and 3 years after students left
not ready to start in a Direct Instruction pro- the program indicated that the achievement
gram any more than they are in any other pro- effects had been maintained.
gram that delivers instruction in English.
However, Direct Instruction programs are Children With Various
appropriate for students who have very basic “Learning Styles”
English language skills (Grossen & Kelly, Many educators believe that students have
1992). More generally, they are appropriate for different “learning styles” and that learning
those students who demonstrate the specific can be maximized by matching instruction to
prerequisite skills necessary for success in the individual students’ learning style. However,
program based on performance on the place- despite its common appeal and widespread
ment test that accompanies every program. acceptance, reviews of controlled research
Gersten (1997) suggested that, because of the studies have consistently failed to find any
careful sequencing of prerequisite skills, con- relationship between instruction and learning
trolled vocabulary, and ongoing assessment of styles (Snider, 1992; Stahl, 1999; Stahl &
mastery, Direct Instruction seems to provide “a Kuhn, 1995). That is, there is no empirical evi-
workable basis for establishing a structured dence that matching instruction to a student’s
immersion program for limited- and non- so-called learning style results in better out-
English-speaking students” (p. 22). Gersten comes for the student than instruction that is
also suggested that the design of Direct not “matched.” The idea is simply not sup-
Instruction programs “allow[s] for one of the ported by research findings.
cardinal principles of structured immersion—
that new material be introduced in English but Attempts to prescribe specific teaching
at a level understood by the children” (p. 28). approaches based on measures of learning
styles have systematically failed. However, it is
Duran (1982) showed that more rapid acquisi- clear that effective teaching does depend on a
tion of math concepts was found with Hispanic much more focused approach to adjusting
students with limited English proficiency using instruction to the needs of individual stu-
instructional materials developed according to dents. Students’ instructional needs are based
Engelmann and Carnine’s (1982) instructional on the skills that they currently possess.
design principles (discussed earlier in this Direct Instruction places a high value on con-
chapter) than with traditional math programs. tinually adjusting students’ placement in pro-

Journal of Direct Instruction 101

grams, pace of lesson coverage, and amount of possible for high-school students to make
repetition on each activity based on students’ achievement gains of more than 2 years in only
performance. This approach eschews the 9 months of instruction (Campbell, 1988).
hypothetical and elusive characteristics of (See Chapter 10 for further information on
learning styles and instead focuses on stu- studies involving high-school students in reme-
dents’ needs that are clearly seen in their per- dial reading programs.) Herr (1989) showed
formance and are directly relevant to making that even adult learners with a long history of
specific adjustments in instruction. failure and severe skill deficits can be success-
ful when taught with Direct Instruction.
Students of Different Ages
When educators discuss whether a particular
instructional program is appropriate to a spe- Research on Direct
cific child or group of children, they often use
the term “developmentally appropriate.” Instruction
According to Church (2002), developmentally More than any other commercially available
appropriate practice is an approach that instructional programs, Direct Instruction is
involves providing children with programs that supported by research. Numerous studies pro-
fit their age and needs. The principles of Direct vide empirical support for the specific Direct
Instruction are entirely consistent with this Instruction design principles and teaching
position (Kozloff & Bessellieu, 2000). Each practices that were discussed previously
Direct Instruction program includes extensive (Engelmann & Carnine, 1982; Kameenui,
techniques for assessing the individual needs of Simmons, Chard, & Dickson, 1997). We have
children and responding to those needs. already seen a number of examples of research
on Direct Instruction with diverse learners.
Studies have shown Direct Instruction to be Several summaries are available providing
effective in teaching learners of all ages, from additional research with a range of learners, in
preschool to adult. The origins of Direct various settings, and in different content areas
Instruction are in the Engelmann-Bereiter pre- (e.g., Adams & Engelmann, 1996; Becker,
school where children demonstrated a substan- 1978; Kameenui et al.; MacIver & Kemper,
tial increase in language skills as well as IQ 2002). In addition, current research and evalu-
scores (Bereiter & Engelmann, 1966). Later, ation of Direct Instruction may be found in
Weisberg (1988) reported that preschool chil- The Journal of Direct Instruction. In the follow-
dren who received 2 years of Direct ing sections, we describe Project Follow
Instruction consistently performed above the Through, a large-scale research project that
98th percentile on measures of reading. More included Direct Instruction, independent
recently, research has demonstrated significant reviews of research and evaluation literature
improvements in language and social interac- related to Direct Instruction, and several stud-
tions of preschool children (Waldron-Soler et ies of long-term outcomes from early experi-
al., 2002). Chapter 3 discusses further research ences with Direct Instruction.
conducted with preschoolers.

At the other end of the age spectrum are older

learners. It is not surprising that Direct
Project Follow Through
Instruction is also effective in teaching older Project Follow Through was originally con-
students. Effective programs are not differen- ceived as a large-scale comprehensive service
tially effective; they are effective for learners of program for economically disadvantaged chil-
all ages. Research has demonstrated that it is dren that would essentially extend Head Start

102 Summer 2003

into the primary grades. However, because the and a control group, the other made direct
funds needed for such an ambitious undertak- comparisons among the models.
ing were not appropriated, the United States
Office of Education (now the U.S. In the first type of analysis, the performance
Department of Education) decided to imple- of students at each Follow Through (FT) site
ment Follow Through as an educational was compared to the performance of a Non-
research program. Follow Through provided an Follow Through control group (NFT) in the
opportunity to compare different educational same community with similar economic and
approaches in order to accumulate evidence social circumstances. If the difference on a
about their effectiveness in teaching children given outcome favored the Follow Through
who are economically disadvantaged. Follow group, that is, if the scores of the Follow
Through is the largest educational experiment Through group were significantly higher than
in history, costing close to one billion dollars, the scores of the control group, the outcome
and involving nearly 100,000 children from 170 was considered positive. Conversely, when the
communities throughout the United States. performance of the control group surpassed
The experimental phase of Follow Through that of students in a particular Follow Through
lasted from 1968 to 1976. Follow Through con- model, the outcome was considered negative.
tinued as a service program until funding was An index of significant outcomes (Stebbins,
discontinued in 1995. St. Pierre, Proper, Anderson, & Cerra, 1977)
for each model is shown in Figure 2.13.
Follow Through created a sort of national learn-
On this graph, a score of zero (represented by
ing laboratory, and the design, called planned
the vertical dividing line) would indicate that
variation, provided a unique opportunity to
there was no difference on that measure
implement various instructional approaches (or
between the Follow Through group and the
models) in classrooms and then evaluate their
control group. Bars extending to the right of
effects (Watkins, 1997). Developers of the dif-
the vertical line indicate positive outcomes for
ferent approaches acted as “sponsors” of their
the Follow Through model. Bars extending to
model. The models fell into three categories: the left of the center line indicate negative
those that emphasized instruction of academic outcomes for the Follow Through model
skills, those that emphasized cognitive growth, (Stebbins et al.).
and those that stressed affective (i.e., self-
esteem) development. The major models are As can be seen, the Direct Instruction model
described in Table 2.2. was the only model to demonstrate significant
positive outcomes on basic skills measures,
The study measured three kinds of outcomes: cognitive–conceptual measures, and affective
basic skills (word recognition, spelling, language, measures. The majority of the other models
and math computation), cognitive–conceptual had negative outcomes, which means that the
skills (reading comprehension, math concepts, performance of students who participated in
and problem solving) and affective (self-con- those models was lower than that of the con-
cept). Children were tested with these meas- trol group.
ures when they entered the program (in
kindergarten or first grade) and at the end of It is particularly important to observe that the
each school year until they completed third Direct Instruction model was more effective on
grade. The evaluation data were collected and cognitive–conceptual measures than any other
analyzed by researchers from two independent model, including those whose explicit goal was
agencies. Two main analyses were conducted. cognitive–conceptual development (Parent
One made comparisons between each model Education, TEEM, Cognitively-Oriented

Journal of Direct Instruction 103

Table 2.2
Follow Through Models

Model Sponsor Description

Direct University of The curriculum emphasis was reading, arithmetic, and language.
Instruction Oregon Behavioral methods were used in conjunction with sponsor-developed
College of teaching materials. Carefully sequenced lessons specified teacher
Education behaviors (scripted presentation). Instruction took place in small,
homogenous groups. Children’s progress was assessed frequently.

Behavior University of Primary objective was mastery of reading, writing, spelling, and math
Analysis Kansas skills. A token economy was implemented and programmed instruc-
tional materials were used. Three or four adults staffed classrooms.
Children’s progress was continuously monitored.
Parent University of Curriculum objectives varied depending on the assessed needs of indi-
Education Florida vidual children. No particular curriculum or teaching strategies were
recommended. Focus was on motivating and training parents to serve
as teaching aides in the classroom and to visit the parents of children
in the class and teach them how to teach their children.
Tucson Early University of Emphasis was development of broad intellectual skills and positive
Educational Arizona attitudes toward school. Language was emphasized as the medium of
Model skill development. Children’s interests determined the curriculum.
Cognitively High Scope This developmental model was based in part on Piagetian theory. The
Oriented Educational focus was on developing children’s reasoning abilities. Children sched-
Curriculum Research uled their own activities. Teachers were trained to function as cata-
Foundation lysts rather than providers of information. Science, math, and reading
were emphasized.

Responsive Far West Instruction was self-paced and self-determined. The primary objective
Education Laboratory was the development of problem solving skills, sensory discrimination,
and self-confidence. A basic assumption was that given self-esteem
and an appropriate learning environment, acquisition of academic
skills would follow.
Bank Street The Bank The curriculum objectives of this model included the development
Street of positive self-image, creativity, coping skills, and the use of lan-
College of guage to formulate and express ideas. Instructional procedures were
Education not described.

Open Education The primary objectives were development of self-respect, imagination, and
Education Development openness to change. The schedule was flexible with children initiating and
Center terminating activities. The open classroom approach stressed a stimulating
environment. The model assumed basic academic skills would be
more readily acquired if they were not treated as academic exercises.

The Language Southwest This model stressed bilingual language development for Spanish
Development Educational speaking children. Positive emphasis on the child’s native language
(Bilingual Development and culture was emphasized. Spanish and English were taught simul-
Education) Laboratory taneously; teaching procedures were not specified.

104 Summer 2003

Curriculum). These findings are important It is also noteworthy that the Direct
because one common misunderstanding is that Instruction model produced positive results on
Direct Instruction promotes only rote learning. affective (self-esteem) measures. The children
In fact, the children in the Direct Instruction in the Direct Instruction model had higher
model demonstrated higher scores on cogni- scores on this set of outcome measures than
tive–conceptual measures (problem solving and the control group. It is striking to note that
thinking skills) than students in the control those models that focused on affective devel-
group. Without exception, the other models opment (Bank Street, Responsive Education,
were unable to demonstrate significant positive Open Education) had negative effects on
results on cognitive–conceptual measures. those measures. This finding means that stu-

Figure 2.13
Follow Through results: Index of significant outcomes for all models.
Adapted from Stebbins et al. (1977).





Basic Skills
Cognitive Cognitive
Bank Street



-500 -400 -300 -200 -100 0 100 200 300 400 500

Journal of Direct Instruction 105

dents who experienced these models demon- The purpose of the Follow Through evaluation
strated lower self-esteem than students in the was to study instructional methods that were
control group. The results of the independent intended to reduce the disparity between eco-
evaluation of Project Follow Through support nomically disadvantaged children and their
the conclusion that young children who peers. The Direct Instruction model was the
acquire the skills that enable them to be suc- sole model that succeeded in raising student
cessful in school feel more positive about performance to a level on a par with national
themselves and their school experiences. norms by the end of third grade. At the end of
third grade, children in the Direct Instruction
The second type of analysis provides informa- model were performing at or near the national
tion about the achievement level of students norm on each measure. These data provide
in each of the models. This comparison uses clear evidence of the measurable effectiveness
results from the reading, math, spelling, and of Direct Instruction. The independent evalua-
language subtests of the Metropolitan tors (Stebbins et al., 1977) summarized the
Achievement Test. Figure 2.14 shows the results as follows, “When all Direct Instruction
results of the major models in these four areas. sites are grouped and compared with the
Metropolitan Achievement Test norms, stu-
To fully appreciate these data, we must under- dents on the average are performing at grade
stand that, although the national norm is the level in Reading, Math, and Spelling” (p. A-
50th percentile, disadvantaged students (as a 168). Stebbins concluded that the Direct
group) typically score in the 20th percentile. Instruction model was generally effective in
Thus, the 20th percentile can be used as a raising the achievement of Follow Through chil-
standard for measuring the benefits of receiv- dren to a level comparable with national norms.
ing instruction according to the various Follow
Through models (Becker, 1978). That is, if
Independent Reviews of Research
students who participated in a Follow Through
on Direct Instruction
model were expected to be performing at the
20th percentile at the end of third grade with- It has been argued (e.g., Allington, 2002) that,
out intervention, then an outcome above the because the Follow Through evaluation was
20th percentile would be judged to be an completed 30 years ago, the data are no longer
improvement over that prediction. Conversely, relevant. However, the findings of the Follow
if the children who participated in a particular Through evaluation have not been contra-
Follow Through model scored below the 20th dicted by more recent research findings. In
percentile, we could conclude that their per- fact, recent evaluations have affirmed the find-
formance was actually worse than it would ings of Project Follow Through. The American
have been without participation in that Follow Federation of Teachers (AFT) (1998a) identi-
Though model. fied Direct Instruction as one of seven promis-
ing programs for teaching reading and language
We see that only the Direct Instruction model arts. The AFT report summarized the research
demonstrated substantial improvement over on Direct Instruction saying, “when this pro-
the 20th percentile on all measures of aca- gram is faithfully implemented, the results are
demic achievement. At the end of third grade, stunning” (p. 17). In a separate report the
the average of students in the Direct AFT (1998b) also identified Direct
Instruction model was the 41st percentile in Instruction as one of six school reform pro-
reading and the 48th percentile in math. The grams. In the third report AFT (1999) named
children in the Direct Instruction model Direct Instruction as one of five remedial
scored, on average, at the 54th percentile in reading intervention programs that are backed
spelling and at the 50th percentile in language. by strong research results.

106 Summer 2003

The American Institutes of Research (AIR) at Risk analyzed the research related to 29 of
was commissioned to provide an independent the most widely implemented comprehensive
review of literature on 24 prominent school- school reform models. This review found that
wide reform approaches. After an extensive Direct Instruction was one of only three mod-
review of research reports, AIR concluded that els that could be rated as having the strongest
Direct Instruction was one of only three evidence of effectiveness. The review con-
approaches that could show strong evidence of cluded that Direct Instruction had “statisti-
positive outcomes on student achievement cally significant and positive achievement
(Herman et al., 1999). effects based on evidence from studies using
comparison groups or from third-party compar-
In a fifth independent review, the Center for ison designs” (Borman, Hewes, Overman, &
Research on the Education of Students Placed Brown, 2002, p. 29).

Figure 2.14
Follow Through results: Metropolitan Achievement Test scores for all models.
Adapted from Stebbins et al. (1977).

Comparison of Third Grade Follow Through Children

on the Metropolitan Achievement Test














Total Reading Total Math Spelling Language

Journal of Direct Instruction 107

Long-Term Follow-Up Research American Federation of Teachers. (1998a). Seven promis-
ing schoolwide programs for raising student achieve-
A small, but widely publicized, research study ment. Retrieved November 10, 2002, from
followed up on graduates from several pre-
school programs when they were 15 years old American Federation of Teachers. (1998b). Six promising
(Schweinhart, Weikart, & Larner, 1986). In schoolwide reform programs. Retrieved November 10,
this study participants were asked to provide a 2002, from
self-report (i.e., complete a questionnaire)
American Federation of Teachers. (1999). Five promising
about their antisocial acts. The 18 students remedial reading intervention programs. Retrieved
who had graduated from a Direct Instruction November 14, 2002, from
preschool program reported more antisocial whatworks/wwreading.htm
acts than those who had completed other Becker, W. C. (1971). An empirical basis for change in educa-
kinds of preschools. This single study has been tion. Chicago: Science Research Associates.
Becker, W. C. (1978). The national evaluation of Follow
widely cited and, in some circles, the idea that
Through: Behavioral-theory-based programs come out
participation in Direct Instruction can have on top. Education and Urban Society, 10(4), 431–458.
negative effects measured 10 years later has Becker, W. C., & Carnine, D. W. (1980). Direct
been accepted as a proven fact. Instruction: An effective approach to educational
intervention with disadvantaged and low performers.
Recently, however, other researchers con- In B. B. Lahey & A. E. Kazdin (Eds.), Advances in clini-
ducted a similar study with many more partici- cal child psychology (Vol. 3, pp. 429–473). New York:
pants (at least 77 per group compared to only
Becker, W. C., Engelmann, S., & Thomas, D. R. (1975).
18 in the Schweinhart et al. study) and sub- Teaching 2: Cognitive learning and instruction. Chicago:
stantially stronger experimental methods Science Research Associates.
(Mills, Cole, Jenkins, & Dale, 2002). This Bereiter, C., & Engelmann, S. (1966). Teaching disadvan-
recent research also contacted 15-year-olds taged children in the preschool. Englewood Cliffs, NJ:
and used the same survey as in the earlier Prentice-Hall.
Borman, G. D., Hewes, G. M., Overman, L. T., & Brown,
study. The authors found no substantial differ-
S. (2002). Comprehensive school reform and student achieve-
ences between graduates of a Direct ment: A meta-analysis (Report No. 59). Retrieved
Instruction program and graduates of a “child- November 15, 2002, from
centered” program. In fact, the very small dif- Brophy, J., & Good, T. (1986). Teacher behavior and stu-
ferences that did exist actually favored the dent achievement. In M. C. Wittrock (Ed.), Third
Direct Instruction program. In a careful com- handbook of research on teaching (3rd ed., pp. 328–375).
New York: Macmillan.
parison of the two studies, Mills et al. con-
Campbell, M. L. (1988). Corrective Reading Program evalu-
cluded that the differences found in the ated with secondary students in San Diego. Direct
Schweinhart study were most likely due to the Instruction News, 7(4), 15–17.
fact that the Direct Instruction group in that Carnine, D., & Kameenui, E. (Eds.). (1992). Higher order
study included a higher ratio of boys than did thinking: Designing curriculum for mainstreamed students.
the other groups, and boys are known to par- Austin, TX: Pro-Ed.
ticipate in unlawful behavior at a much higher Church, E. B. Defining developmentally appropriate. (n.d.).
Retrieved November 18, 2002, from
rate than girls.
References Duran, E. (1982). Hispanic children can be taught: Or
Adams, G. L., & Engelmann, S. (1996). Research on Direct which teaching method is most effective. Teaching and
Instruction: 25 years beyond DISTAR. Seattle, WA: Learning Review, 2, 4–6.
Educational Achievement Systems. Engelmann, S. (1969). Conceptual learning. Sioux Falls, SD:
Allington, R. (2002). What do we know about the effects ADAPT Press.
of Direct Instruction on student reading achievement? Engelmann, S. (1997, July). Student-program alignment and
Retrieved September 15, 2002, from teaching to mastery. Paper presented at the 23rd National Direct Instruction Conference, Eugene, OR.

108 Summer 2003

Engelmann, S. (1999, July). Student-program alignment and Mills, P. E., Cole, K. N., Jenkins, J. R., & Dale, P. S.
teaching to mastery. Paper presented at the 25th National (2002). Early exposure to Direct Instruction and sub-
Direct Instruction Conference, Eugene, OR. sequent juvenile delinquency: A prospective examina-
Engelmann, S., & Becker, W. C. (1978). Systems for basic tion. Exceptional Children, 69, 85–96.
instruction: Theory and applications. In A. C. Catania Robinson, J. W., & Hesse, K. (1981). Morphemically based
& T. A. Brigham (Eds.), Handbook of applied behavior spelling program’s effect on spelling skills and spelling
analysis (pp. 325–377). New York: Irvington. performance of seventh-grade students. Journal of
Engelmann S., & Carnine, D. W. (1982). Theory of instruc- Educational Research, 75, 56–62.
tion: Principles and applications. New York: Irvington. Rosenshine, B. V., & Berliner, D. C. (1978). Academic
Engelmann, S., & Carnine, D. W. (1989). DI outcomes engaged time. British Journal of Teacher Education, 4,
with middle-class second graders. Direct Instruction 3–16.
News, 8(2), 2–5. Rosenshine, B., & Stevens, R. (1986). Teaching functions.
Forness, S. R., Kavale, K. A., Blum, I. M., & Lloyd, J. W. In M. C. Whittrock (Ed.), Third handbook of research on
(1997). Mega-analysis of meta-analysis: What works in teaching (3rd ed., pp. 376–391). New York: Macmillan.
special education. Teaching Exceptional Children, 19(6), Schaefer, E. (1989). Is DI only for low achievers? Direct
4–9. Instruction News, 8(2), 6–9.
Gersten, R. (1997). Structured immersion of language Schaefer, E. (2000, July). Creating world class schools: Peak
minority students: Results of a longitudinal evaluation. performance through Direct Instruction. Presented at the
Effective School Practices, 16(3), 21–29. annual conference of the Association for Direct
Gersten, R., Becker, W., Heiry, T., & White, W. A. T. Instruction, Eugene, OR.
(1984). Entry IQ and yearly academic growth in chil- Schweinhart, L., Weikart, D., & Larner, M. (1986).
dren in Direct Instruction programs: A longitudinal Consequences of three preschool curriculum models
study of low SES children. Educational Evaluation and through age 15. Early Childhood Research Quarterly, 1,
Policy Analysis, 6(2), 109–121. 15–45.
Gersten, R., Taylor, R., Woodward, J., & White, W. A. T. Sexton, C. W. (1989). Effectiveness of the DISTAR
(1997). Structured English immersion for Hispanic Reading I program in developing first graders’ language
students in the U.S.: Findings from the 14-year evalua- skills. Journal of Educational Research, 82, 289–293.
tion of the Uvalde, Texas program. Effective School Slavin, R. (2003). Educational psychology: Theory and practice
Practices, 16(3) 30–38. (7th ed.). Boston: Allyn & Bacon.
Grossen, B., & Kelly, B. (1992). Effectiveness of Direct Snider, V. E. (1992). Learning styles and learning to read:
Instruction in a third-world context. Direct Instruction A critique. Remedial and Special Education, 13, 6–8.
News, 9(4), 4–11. Stahl, S. (1999). Different strokes for different folks? A
Herman, R., Aladjem, D., McMahon, P., Masem, E., critique of learning styles. American Educator, 23(3),
Mulligan, I., O’Malley, A., et al. (1999). An educators’ 27–31.
guide to schoolwide reform. Retrieved November 15, 2002, Stahl, S. A., & Kuhn, M. R. (1995). Current trends in
from reading: Does whole language or instruction matched
zation/Reform to learning styles help children to read? School
Herr, C. (1989). Using Corrective Reading with adults. Direct Psychology Review, 24, 393–404.
Instruction News, 8(2), 18–21. Stebbins, L. B., St. Pierre, R. G., Proper, E. C., Anderson,
Kameenui, E. J., Simmons, D. C., Chard, D., & Dickson, R. B., & Cerva, T. R. (1977). Education as experimenta-
S. (1997). Direct Instruction reading. In S. A. Stahl & tion: A planned variation model (Volume IV-A: An evaluation
D. A. Hayes (Eds.), Instructional models in reading (pp. of Follow Through). Cambridge, MA: Abt Associates.
59–84). Mahwah, NJ: Erlbaum. Tarver, S. C., & Jung, J. S. (1995). A comparison of mathe-
Kozloff, M. A., & Bessellieu, F. B. (2000). Direct Instruction matics achievement and mathematics attitudes of first
is developmentally appropriate. Retrieved July, 9, 2002, and second graders instructed with either a discovery-
from learning mathematics curriculum or a Direct
MacIver, M. A., & Kemper, E. (2002). Guest editors’ intro- Instruction curriculum. Effective School Practices, 14(1),
duction: Research on Direct Instruction reading. 49–57.
Journal of Education for Students Placed At-Risk, 7, Vitale, M., & Romance, N. (1992). Using videodisc
107–116. instruction in an elementary science methods course:
Madsen, C. H., Becker, W. C., Thomas, D. R., Koser, L., & Remediating science knowledge deficiencies and facili-
Plager, E. (1968). An analysis of the reinforcing func- tating science teaching. Journal of Research in Science
tion of “sit down” commands. In R. K. Parker (Ed.), Teaching, 29, 915–928.
Readings in educational psychology (pp. 265–278). Boston: Vygotsky, L. S. (1997). Educational psychology. Boca Raton,
Allyn & Bacon. FL: St. Lucie Press.
Martella, R. C., & Nelson, J. R. (in press). Managing class- Waldron-Soler, K. M., Martella, R. C., Marchand-Martella,
room behavior. Journal of Direct Instruction. N. E., Tso, M. E., Warner, D. A., & Miller, D. E.

Journal of Direct Instruction 109

(2002). Effects of a 15-week Language for Learning
implementation with children in an integrated pre-
school. Journal of Direct Instruction, 2(2), 75–86.
Watkins, C. L. (1997). Project Follow Through: A case study of
the contingencies influencing instructional practices of the educa-
tional establishment. (Monograph). Concord, MA:
Cambridge Center for Behavioral Studies.
Weisberg, P. (1988). Reading instruction for poverty level
preschoolers. Direct Instruction News, 7(4), 25–30.
White, W. A. T. (1988). Meta-analysis of the effects of
Direct Instruction in special education. Education and
Treatment of Children, 11, 364–374.

110 Summer 2003