Professional Documents
Culture Documents
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.
Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to The Journal
of Educational Research
This content downloaded from 131.91.169.193 on Sun, 17 Jul 2016 11:17:14 UTC
All use subject to http://about.jstor.org/terms
Presentational Behaviors and Student
Achievement in Mathematics
LYLE R. SMITH
Augusta College
292
This content downloaded from 131.91.169.193 on Sun, 17 Jul 2016 11:17:14 UTC
All use subject to http://about.jstor.org/terms
May/June 1985 [Vol. 78(No. 5)] 293
students. Second, the previous research focused on high a variety of content areas, Smith and Sanders (1981) con
school social studies students and does not necessarily ducted research involving fifth-grade social studies students.
generalize to the secondary school mathematics classroom. The degree of lesson structure significantly affected the
Therefore, one purpose of this study was to examine in a achievement of these students. Smith and Hodgin (1982)
field setting the effect of bluffing and recovery phrases, as reported that high school mathematics students com
well as the effect of other vagueness phrases, on student prehended written material better when the material had a
achievement in high school mathematics. high degree of structure. Therefore, lesson structure appears
Mazes. Hunt (1965, 1968) referred to mazes as units to be a variable that affects student learning, regardless of
of discourse that do not make semantic sense. Smith the grade level or the subject matter area.
(1977) defined mazes to be false starts or halts in speech, Anderson (1969) studied lesson structure in terms of
redundantly spoken words, and tangles of words. Fagan the way in which new concepts are introduced and in
(1982) investigated relationships between the occurrence terms of the frequency with which concepts are repeated
of mazes, sentence complexity, and the use of connec from one sentence of communication to the next.
tives in sentences. Fagan reported a positive relationship Anderson proposed the following equation to define ?,
between the frequency of mazes a person utters and the the degree of structure:
complexity of sentences the person attempts to commu
nicate. Fagan suggested that mazes may be a product of
Alo ? 2rti
language planning during discourse.
Smith and Land (1981) reviewed six studies that where ?i equals the number of concepts repeated in a
reported a negative relationship between frequency of pair of consecutive sentences and n\ equals the number
teacher mazes and student achievement. In four of the six of concepts in one or the other of a contiguous pair of
studies the relationships were significant beyond the .05 sentences that are not repeated from one to the other.
level. One purpose of this study was to determine the rela Examples of the use of this formula are given in the
tionship between mazes and other low-inference discussion concerning Table 1.
behaviors, and to determine the combined effects of these Anderson (1969) defined total lesson structure to be
behaviors on student learning in high school mathematics. the mean of all values of By in a lesson. Anderson (1970)
A review of literature indicates that no previous research suggested that a mean above .50 represents a lesson of
has examined the joint effects on achievement of high structure, whereas a mean below .30 represents a
vagueness terms, mazes, and lesson structure. lesson of low structure.
Lesson structure. The structure of teacher communi Table 1 shows an excerpt of a lesson on direct varia
cation has been addressed by several educational research tion, a topic appropriate for beginning high school
ers. Bruner (1962, 1964) referred to structure in terms of algebra. Key concepts and processes are numbered to
meaningful relationships among concepts. Gagn? (1970) facilitate computation of structure. As in the study by
studied structure as it relates to hierarchical organizations Smith and Sanders (1981), a concept was defined to be a
of knowledge. Ausubel (1960, 1963) related structure to word or phrase that refers to a group of one or more
the use of advance organizers that introduce subsuming things possessing common characteristics. Smith and
concepts involving the content to be learned. Bellack, Sanders pointed out that such a definition includes
Kliebard, Hyman, and Smith (1966) examined structure as phrases such as "function" and "slope" as concepts, but
a series of pedagogical moves that affect subsequent that this definition also includes such phrases as "value"
teacher-student interactions. More recently, Anderson and "multiplication" as concepts. Therefore, as sug
(1967, 1969, 1970, 1974), Anderson and Lee (1975), gested by Anderson (1969), only key concepts (concepts
Browne and Anderson (1974), Mathis and Shrum (1977), that represent ideas directly related to one or more of the
and Trindade (1972) defined another method to determine lesson objectives) were used in the computation of each
a low-inference measure of the degree of structure in class value of ?. In this study, function and slope were
room communication. These researchers indicated that the classified as key concepts, but value and multiplication
degree of structure had a direct effect on student achieve were not classified as key concepts. Vagueness terms and
ment in science. Butterworth (1974) found that structure mazes are shown in italics in the excerpt.
also affected attitudes and perceptions of science students. Referring to Table 1, the procedure for computing B,
Simmons (1977) concluded that structure influenced suc from one segment of communication to the next is
cess in mastering psychomotor skills (such as using a com shown. Segment 1 concerns two key concepts (direct
pound microscope) of science students. Ferraro, Lee, and variation and function), which are concepts 1 and 3.
Anderson (1977) found that structure affects the learning Segment 2 concerns four key concepts (direct variation,
of science students of various mental abilities. slope, function, and constant of variation), concepts 1,
The research of Anderson and his colleagues focused on 4, 3, and 5. Therefore segments 1 and 2 have two key
the teaching and learning of science concepts. To determine concepts in common (concepts 1 and 3) and two key
if lesson structure is a global variable that affects learning in concepts that appear in one segment but not the other
This content downloaded from 131.91.169.193 on Sun, 17 Jul 2016 11:17:14 UTC
All use subject to http://about.jstor.org/terms
294 Journal of Educational Research
Key Concepts B?
1. Direct variation is a little different from idea of linear rela . . . function. 1,3 ?
2. Instead of m for the slope, as in, in, in, . . . function, you know, we will let k be 1,4,3,5 2 + 2(2)= *^7
our constant of variation.
3. Now, on . . ., let's look at this example. It says, "S varies directly as T." Maybe we 2,5,6 5 + 2 = -29
can write an equation involving our constant of variation. Bob? (Student responds, "S
equals KT." Teacher writes equation on board.)
4. The problem says that S equals 4 when T equals 8. Solve for the constant. That's not 2,5 1+2(2)= *^?
very hard to do. (Teacher writes 4 = K?8 on board. Student says, "K equals one half.")
5. OK, now, we need to ... Is everyone with me? We need to find the value of T when S 2,7 2 + 20)= ^
equals . . . when S equals ... 3. So 3 equals one-half T. (Teacher writes 3 = ViT.)
Multiplying both sides by 2 gives T equals 6. OK? (Teacher writes 3x2 = 2 x Vi? = T.)
Note. Vagueness terms and mazes are italicized.
This content downloaded from 131.91.169.193 on Sun, 17 Jul 2016 11:17:14 UTC
All use subject to http://about.jstor.org/terms
May/June 1985 [Vol. 78(No. 5)] 295
bluffing and recovery terms, the remaining eight categories tion. A third block of information involved solving for
of vagueness terms, teacher mazes, and lesson structure. For the value of one variable when the corresponding value
the bluffing and recovery, remaining vagueness, and mazes of a second variable was given. The difficulty arose
variables, coder reliability was checked in two ways. The because a typical lesson involved the teacher asking
percentage of agreement on the total number of identifica questions related to one of the lesson objectives and
tions made in quantifying a variable was checked by (a) sub then having to digress because student responses in
tracting the lowest total of identifications for a particular dicated a lack of understanding concerning a pre
lesson from the highest total of identifications for that requisite concept or process. Because the three coders
lesson, (b) dividing this difference by the highest total of could not consistently agree on when one block of infor
identifications for the lesson, and (c) subtracting the mation was completed and another was begun, each
resulting quotient from one. Also, for each lesson, the lesson was coded for structure by all three coders as a
percentage of agreement on the location of identifications group. When there were discrepancies in perspectives
made in quantifying a variable was checked by dividing the concerning coding of the structure variable, a majority
total number of identifications agreed on by all three coders decision was used as the basis for analysis. In 15 of the
by the arithmetic mean of the three totals of identification 19 lessons, there were no such discrepancies in perspec
made by the coders for that lesson. These reliability coeffi tive. In three of the lessons, one discrepancy was solved
cients were conducted four times for each of the variables of through the majority decision process. In one lesson,
bluffing and recovery, remaining vagueness terms, and three discrepancies were solved through this process.
mazes. If agreement of at least 80% on both the number of
identifications and the location of identifications was not
Results
met for a reliability check, the coders were given practice
sessions before they continued quantifying the variables. Table 2 shows the mean class posttest scores, adjusted
The lesson structure variable was the most difficult by using the CAT mathematics scores as a covariate. The
for coders to analyze. Key concepts and processes were four coded variables, bluffing and recovery terms, remain
identified and agreed upon before the analysis was con ing vagueness terms, mazes, and lesson structure, also are
ducted. Coders then were trained to identify blocks of shown for each of the 19 teachers. The bluffing and
communication based on the objectives for the lesson. recovery variable, the variable concerning remaining
For example, one block of communication consisted of vagueness terms, and the mazes variable all were analyzed
identifying an equation that described a particular direct in terms of their frequencies per minute of teacher talk.
variation. A second block of information involved solv As shown in Table 2, the adjusted posttest mean for
ing for the constant of variation for a given direct varia the entire sample was 10.17 out of a possible 14 points.
Remaining Mazes
"Bluffing" and vagueness terms per min. Mean
Cov.-adj. recovery per min. per min. teacher teacher lesson
Teacher class means teacher talk talk talk structure
This content downloaded from 131.91.169.193 on Sun, 17 Jul 2016 11:17:14 UTC
All use subject to http://about.jstor.org/terms
2% Journal of Educational Research
The split-half reliability of the test was .83 after the lesson structure, and the mazes. The bluffing and
Spearman-Brown correction formula was applied. recovery variable was excluded from this analysis because
A total of 189 bluffing and recovery terms were iden of its small correlation with achievement. Table 5 details
tified by the coders, and 614 vagueness terms from the results of the multiple regression analysis. Unstandard
remaining categories of vagueness were identified. ized regression coefficients, as well as standardized
Therefore, bluffing and recovery phrases represented regression coefficients (beta weights) are shown. The
24% of the total number of vagueness terms identified. total variance in posttest scores explained by the three
The mean values for lesson structure ranged from .48 to variables was 47.2% (multiple R = .69) and the overall
.82, with a mean of .61. value of Fwas significant beyond the .02 level, F(3,15) =
The coder reliability for quantification of mazes was 4.47. As shown in Table 5, the multiple regression equa
higher than reliability reported by Smith (1977). Table 3 tion was: Posttest M = 7.58 + 7.34 S - 1.01 V - 0.03
reports the mean percentage agreements for the coders. M. Elimination of the mazes variable in the analysis
Lesson structure coder reliability is not reported in Table yielded a multiple R of .69, F(2,16) = 7.14, p < .01, and
3 because of the coding procedures previously discussed. equation was: Posttest M = 7.44 + 7.47 S - 1.02 V.
Table 4 is an intercorrelation matrix showing correlations
among the variables investigated. The positive correlation Discussion
between mean lesson structure and mean class achievement
was significant beyond the .05 level, as was the negative cor This study examined four low-inference teacher
relation between mean achievement and the eight categories behaviors and their effect on student learning in high
of vagueness terms excluding the bluffing and recovery school mathematics. Bluffing and recovery phrases were
category. The correlation between bluffing and recovery not significantly correlated with achievement. Mazes
terms and mean achievement was insignificant. Mazes were were negatively correlated with achievement, but the rela
negatively, but not significantly, correlated with mean tionship was not significant. The correlation between
achievement. It should be noted a positive correlation was achievement and vagueness terms other than bluffing and
found between bluffing and recovery terms and the remain recovery terms was significant (r = - Al, p < .05), as
ing vagueness terms, although the correlation did not quite was the correlation between achievement and lesson
reach the .05 significance level. structure (r = .51, p < .05). The results concerning
A multiple regression analysis was performed to deter bluffing and recovery phrases support research reported
mine the proportion of variance in the adjusted mean by Smith (in press) in that there was no significant
posttest scores explained by the vagueness terms, the relationship between bluffing and recovery terms and
achievement. The previous study by Smith concerned
high school social studies students, so this study involving
Table 3.?Coder Reliability: Mean Percentage Agreement secondary mathematics students provides evidence that
these results are generalizable across subject matter areas.
Total Location of Caution should be exercised in interpreting the results
Variable Identifications Identifications concerning lesson structure. The following four remarks
concern analysis of the lesson structure variable. First,
"Bluffing" and recovery terms 94 89
Remaining vagueness terms 92 90 subject matter experts should be involved in identifying
Mazes 86 84 the key concepts and processes to be considered in such
an analysis. Selection of the key concepts and processes
should be related closely to objectives outlined for the
lesson. Second, Anderson (1969) developed the theory of
Table 4.?Intercorrelation Matrix lesson structure based on more fluent presentations,
typically of an informal lecture format. This remark is
12 3 4 5 based on examples Anderson presented and on comments
made in subsequent research reports on structure. This
1. Covariance-adj. 1.00 study analyzed classroom communication that involved
mean achievement
2. "Bluffing" and -.02 1.00 more teacher repetition of definitions, examples, and
recovery terms per min. procedures to be followed, and a high degree of redun
teacher talk
dance was present in the lessons. For this reason, blocks
3. Remaining vagueness -.47 .43 1.00
terms per min. teacher of communication were analyzed as units, rather than
talk single sentences as suggested by Anderson. A typical
4. Mazes per min. teacher -.25 .26 .15 1.00 block of communication would involve the teacher defin
talk
5. Mean lesson structure .51 -.18 -.02 -.31 1.00 ing a term, asking a question to see if students
understood the definition, and then responding to stu
Note: Correlations greater than .46 are significant beyond .05 level. dent questions and comments. Such a procedure of
This content downloaded from 131.91.169.193 on Sun, 17 Jul 2016 11:17:14 UTC
All use subject to http://about.jstor.org/terms
May/June 1985 [Vol. 78(No. 5)1 297
Unstandardized Standard
Regression Error
Coefficients of Estimate Beta Multiple
Variable (b) for b Weights R Significance
This content downloaded from 131.91.169.193 on Sun, 17 Jul 2016 11:17:14 UTC
All use subject to http://about.jstor.org/terms
298 Journal of Educational Research
This content downloaded from 131.91.169.193 on Sun, 17 Jul 2016 11:17:14 UTC
All use subject to http://about.jstor.org/terms