Longitudinal comparisons of dental arch changes in

normal and untreated Class II, Division I subjects and

their clinical implications
Samir E. Bishara, BDS, DDS, D. Ortho., MS, a Paymun Bayati, DDS, b and Jane R. Jakobsen, BA, MS °
Iowa City, Iowa

The purpose of this study was to determine on a longitudinal basis whether the growth trends in
maxillary and mandibular dental arch widths and lengths in persons with Class II, Division 1
malocclusions were different from those of normal subjects. Dental casts were available on a total
of 37 subjects with Class II, Division 1 malocclusions (15 males and 22 females) and 55 normal
subjects (28 males and 27 females). For each subject, three sets of casts were evaluated at the
following developmental stages: (1) deciduous dentition (X age = 5.0 years); (2) mixed dentition
(X age = 8.0 years); and (3) permanent dentition (X age = 12.5 years). Eighteen dental arch lengths
and widths parameters were measured on both the maxillary and the mandibular arches.
Interexaminer and intraexaminer reliability were predetermined at 0.5 mm. The Analysis of Variance
general linear models procedure was used to compare the various dental arch parameters on both
a cross-sectional and a longitudinal basis. Curve parallelism and magnitude were evaluated. The
results of the analysis of variance comparing the growth curves of the various dental arch
parameters in the normal subjects and untreated subjects with Class II, Division 1 malocclusions
indicated the absence of significant differences between the groups, i.e., the curves were parallel
(/9 < 0.01). This finding indicates that the growth trends in the various dental arch parameters
evaluated were similar in the normal and Class II, Division 1 malocclusion groups in both male and
female subjects. In addition, the differences between the measurements of maxillary and
mandibular intermolar arch widths were greater in the normal male subjects than in the male
subjects with Class II, Division 1 ma occ us on s Female subjects had similar trends but the
differences were not statistically significant. In conclusion, the present findings indicate that the
clinician should assume that the changes in arch lengths and widths in both the subjects with Class
II, Division 1 malocclusions and the normal subjects follow the same general patterns. On the other
hand, the relative constriction of the maxillary arch as related to the mandibular arch, in Class II
malocclusions is expressed from the earlier stages of dental arch development. These trends
continue in the mixed and early permanent dentitions and do not self correct without treatment.
Therefore when such trends are diagnosed, early correction of the transverse discrepancy should
be attempted together with that of the anteroposterior discrepancy. (Am J Orthod Dentofac Orthop
1996;110:483-9.)

T h e growth potential of persons with Class II
malocclusions is of interest to the practicing orthodon-
tist because these malocclusions constitute a significant
percentage of the cases that are treated. Ast et al. ~
examined 1413 high school students from up-state
New York, aged 15 to 18 years, and found that 23.8%
had Class II malocclusions, whereas 69.9% had
From the College of Dentistry, University of Iowa.
aProfessor of Orthodontics.
bPrivate practice.
CAssistant professor, Department of Preventive and Community Dentistry.
Reprint requests to: Dr. Samir Bishara, Department of Orthodontics, Univer-
sity of Iowa College of Dentistry, 220 Dental Science S, Iowa City, IA
52242-1001.
Copyright © 1996 by the American Association of Orthodontists.
0889-5406/96/$5.00 + 0 8/1/63!25
Class I malocclusions. This ratio, approximately 1:3,
was similar to that reported by Goldstein and Stanton 2
for white American children and by Massler and
Frankel 3 for children aged 14 through 18 years. How-
ever, in a group of American blacks evaluated by
Altemus, 4 the ratio of Class iI to Class I malocclusions
was about 1 to 6.
Clinicians involved in the treatment of Class II
malocclusions in growing persons are interested in
evaluating the growth trends of the dental and skeletal
structures.
Class II malocclusions often express vertical and
transverse discrepancies in addition to its anteroposte-
rior component. Adams and Kerr5 examined a total of
41 subjects with a Class I, Class II, Division 1, or
Class II, Division 2 malocclusions and observed that
483
484 Bishara, Bayati, and Jakobsen American Journal of Orthodontics and Dentofacial Orthopedics
November 1996

Table I. Descriptive statistics and age comparisons (in years) for the normal subjects and the subjects with Class II,
Division 1 malocclusions at three stages of dental development
Males Females

Normal Class II Normal Class H

Y~ SD 2 ] SD P .~ SD g SD P

Stage 1 4.9 0.6 5.1 0.6 NS 5.0 0.4 5.0 0.4 NS

Stage 2 7.8 0.6 8.2 1.2 NS 8.1 0.8 8.5 1.1 NS
Stage 3 13.0 0.8 12.7 1.5 NS 13. i 0.8 12.8 1.5 NS

Y, Mean; SD, standard deviation; P, probability; NS, not significant (p < 0.05).

Table II. Descriptive statistics and the results of cross-sectional and longitudinalcomparisons of maxillary and
mandibular arch lengths measurements (in mm) for the normal subjects and subjects with
Class II, Division 1 malocclusions
Males Females

Normal Class 1I Curve Normal Class II Curve

Y. SD £ SD P Parallelism Magnitude x SD ~ SD P Parallelism Magnitude

Cross-sectional comparisons
Maxillary
Stage l 74.6 4.2 73.4 4.2 NS 72.9 2.9 71.7 2.6 NS
Stage 2 77.4 4.7 75.1 5.3 NS 74.0 3.5 74.5 4.0 NS
Stage 3 77.9 3.9 75.8 5.3 NS 74.7 3.2 74.4 4.4 NS
Mandibular
Stage 1 68.3 2.7 67.8 2.8 NS 65.7 3.2 64.8 3.1 NS
Stage 2 69.5 3.0 70.0 3.4 NS 67.0 2.1 67.5 5.0 NS
Stage 3 66.9 3.8 67.6 4.4 NS 63.7 3.3 64.5 5.3 NS
Longitudinal comparisons
Maxillary arch length NS N > Class II NS NS
Mandibular arch length NS NS NS NS

the overbite was largest in the subjects with Class II,
Division 2 malocclusions and smallest in the subjects
with Class I malocclusions.
Burstone, 6 DeKock, 7 Knott, 8 Meredith and Hopp, 9
Moorrees, I° and Sillman 11 evaluated the changes in
dental arch widths in the molar and canine regions in
different populations and concluded that there were
little or no change in these dimensions by 13 years in
girls and 15 years in boys. They suggested that the
significant changes in arch width in both the maxillary
and the mandibular arches occur at the earlier stages of
dental development.
F(irhlich 12a3 examined persons with Class II, Divi-
sion 1 malocclusions and divided them into three
subgroups; Class II, Division 1 with a V-shaped max-
illary arch; Class II, Division 1 with flaring; and
spacing of the incisors and Class II, Division 1, with
well-aligned maxillary incisors. F6rhlich found that
when compared with normative data, only the subjects
with Class II, Division 1 malocclusions with V-shaped
arches had narrower maxillary intercanine widths, but
all had narrower intermolar widths.
Staley et al.14 compared dental and cephalometric
pretreatment measurements of adult subjects with Class
II, Division 1 malocclusions with matched normal sub-
jects. They observed no differences in intercanine
widths but narrower maxillary molar arch widths in the
Class II group and no differences in the mandibular arch
widths. They concluded that these differences revealed a
posterior crossbite tendency in the Class II group.
The literature review indicates that the width of the
dental arches in subjects with Class II, Division 1 mal-
occlusions was found to be either normal or narrower
than the corresponding widths of normal subjects. Such
a discrepancy may be attributed to differences in the
absolute size of the dental arches in the various Class II
samples compared. A more relevant approach is to cal-
culate and compare the differences between the maxil-
lary and the mandibular arch widths in subjects with
Class II, Division 1 malocclusions and normal sub-
jects.14 Furthermore, it would be of interest to determine
whether the tendency for a transverse discrepancy found
in the adult Class II dentition is also expressed in the
earlier stages of dental arch development.
American Journal of Orthodontics and Dentofacial Orthopedics Bishara, Bayati, and Jakobsen 485
Volume ll0, No. 5

Table III. Descriptive statistics and the results of cross-sectional and longitudinal comparisons of the differences
between maxillary and mandibular intercanine and intermolar widths measurements (in mm) for the normal
subjects and subjects with Class II, Division 1 malocclusions
Males Females

Normal Class lI Curve Normal Class li Curve

Y SD Y~ I SD P Parallelism Magnitude ~ [ SD Y SD P Parallelism[Magnitude

I
Cross-sectional comparisons
Intercanine widths differences
Stage I 6.9 1.8 6.4 2.1 NS 6.2 1.3 6.0 1.8 NS
Stage 2 7.0 2.2 5.4 2.2 NS 6.3 1.4 6.0 1.8 NS
Stage 3 9.3 1.8 9.2 1.9 NS 7.6 1.2 7.4 t.5 NS
Intermolars widths differences
Stage t 6.5 2.1 5.1 2.1 NS 6.4 1.1 5.8 1.6 NS
Stage 2 6.4 2.0 3.6 2.9 NS 6.0 1.5 6.6 4.7 NS
Stage 3 7.5 1.5 4.7 2.5 NS 7.1 1.3 6.2 2.1 NS
Longitudinal comparisons
Intercanine widths differences NS 0.0357 NS NS
Intermolar widths differences NS 0.0001 NS NS

The purpose of this study was to determine on a
longitudinal basis whether the growth trends in max-
illary and mandibular dental arch widths and lengths in
persons with Class II, Division 1 malocclusions were
different from those of normal subjects.
MATERIALS AND METHODS
A total of 92 subjects, 43 males and 49 females, were
evaluated in this study. All subjects were participants in the
Iowa Facial Growth Study.7-9
Normal Subjects: Dental casts on 55 subjects, 28 males
and 27 females were identified. Each subject had a clinically
acceptable occlusion, that is, a Class I molar and canine
relationship, anterior crowding of less than 2 mm at the time
of eruption of the second permanent molars, and no apparent
facial disharmony. None of these subjects had congenitally
missing teeth and none had undergone orthodontic therapy.
In addition, each subject had a complete set of data at three
stages of dental development, namely, at the time of eruption
of the deciduous dentition, at the mixed dentition with all
four permanent incisors erupted, and at the early permanent
dentition when all second molars have erupted. These selec-
tion criteria limited the number of normal subjects in this
investigation to 55.
Class II, Division 1 Subjects: Thirty-seven subjects (15
males and 22 females) with a Class II, Division 1 dental
malocclusions were evaluated at the same dental develop-
mental stages described earlier. None of the subjects under-
went orthodontic treatment.
The ages of the subjects at the three stages evaluated are
detailed in Table I.
Parameters evaluated at the three stages of dental
development:
Arch width measurements: Maximum rectilinear distance
between the cusp tips of the canines, and the mesiobuccal
cusp tips of the second deciduous molars and succedaneous
premolars7,s in both the maxillary and mandibular arches.
The differences between maxillary and mandibular intermo-
lar and intercanine widths were also calculated.
Arch length measurements: The mesiodistal length of the
following arch segments were measured for the right and left
sides: "'~2 (1) Anterior segments, between the mesial contact
point of the central incisors and the point between the canine
and the first premolars; (2) posterior segments, between the
contact point of the canine and the first premolar and that
between the second premolar and the first molar. The total
arch lengths in both the maxillary and mandibular dental
arches were then calculated.
STATISTICAL ANALYSIS
Longitudinal Comparisons. The first step in the
statistical analysis was to determine whether signifi-
cant differences are present between male and female
subjects. The presence of differences necessitated that
male and female growth curves be examined sepa-
rately. The growth profile for each dental arch param-
eter in the four groups were compared by using the
Analysis of Variance General Linear Models proce-
dure. Male and female subjects were evaluated inde-
pendently. The sum of squares of the variation was
partitioned into those attributable to age, sex, param-
eter measured, malocclusion, and their interactions.
In the statistical analysis of the growth curves,
there were two aspects to be evaluated - the shape or
profile of the curves and the magnitude of the curves.
The shape or profile is the slope that describes growth
direction. In this respect, the curves might show a
486 Bishara, Bayati, and Jakobsen American Journal of Orthodontics and Dentofacial Orthopedics
November 1 9 9 6

Maxillary Intercanine Width: FEMALES Mandibular Intercanine Width: FEMALES

34- 29
- e - Class 2
33- 28
- M- Normal
27
32-
26
) 31-
25 ..... . .... .-- ........ •
30-
24
29-
23
Z8
22
27 21

26 , = 20 , i
Stage 1 Stage 2 Stage 3 Stage 1 Stage 2 Stage 3
A Deciduous Mixed Dentition Early Permanent a Deciduous Mixed Dentition Early Permanent

Maxillary Intercanine Width: MALES Mandibular Intercanine Width: MALES

37.
29
Class 2
- O - Class 2
3s .m. Normal
28 -II-
3S. , , .i
27
~1 3 4
2e
/ ................. ,
32. 251
o o""
31 ! .... o • - 24 .'"

30! 23 I

29 , ,
22 , i
Stage I Stage 2 Stage 3
Stage 1 Stage 2 Stage 3
C Deciduous Mixed Dentition Early Permanent
D Deciduous Mixed Dentition Early Permanent

Fig. 1. A, Changes in maxillary intercanine widths in female subjects with Class II, Division 1
malocclusions and normal female subjects. B, Changes in mandibular intercanine widths in female
subjects with Class II, Division 1 malocclusions and normal female subjects. C, Changes in maxillary
intercanine widths in male subjects with Class II, Division 1 malocclusions and normal male subjects.
D, Changes in mandibular intercanine widths in male subjects with Class II, Division 1 malocclusions
and normal male subjects.

parallel relationship indicating that the growth trends
are the same. On the other hand, lack of parallelism
among curve profiles indicates differences in growth
trends. The magnitude of the curves is the height of the
curves with age held constant and describe differences
in the size of the parameter. The method of analysis
used to compare the growth curves was described in
detail by Kleinbaum and Kupper. '~
The level of statistical significance was predeter-
mined at the 0.01 level of confidence for the compari-
sons of the curve parallelism and at the 0.05 level of
confidence for the comparisons of curve magnitude.
This variation in the level of significance is suggested
by Bonferroni. The Bonferroni method 16 takes into
consideration all tests of significance to be examined in
one analysis. Because the test for parallelism was part
of an analysis that included three or more variables, the
0.01 level of significance was chosen to keep the
overall level of significance relatively high.
Cross-sectional Comparisons. The analysis of vari-
ance was used to compare the dental arch parameters
between the subjects with Class II, Division I maloc-
clusions and the normal subjects at each of the three
developmental stages.
FINDINGS
Comparisons of Arch Length Measurements
(Table II)
Cross-sectional Comparisons. The results of the
analysis of variance for the maxillary and mandibular
arch lengths indicate that there were no significant
differences (p < 0.05) between the subjects with Class
American Journal of Orthodontics and Dentofacial Orthopedics Bishara, Bayati, and J a k o b s e n 487
Volume 110, No. 5

Maxillary Intermolar Width: FEMALES Mandibular Intermolar Width: FEMALES

51 43-
50 - 0 - Class 2 .... •
42- -~-Class 2 , JIB" " " ~
49 • II- Normal ...... "" - •- Normal . , , / ~
45 41
47 40-
46
39-
45
.~ 44 38-

37-
42 36-
4t
35,
4O
39 3,1
38 33
Stage 1 Stage 2 Stage 3 Stage I Stage 2 Stage 3
A Deciduous Mixed Dentition Early Permanent B Deciduous Mixed Dentition Early Permanent

Maxillary Intermolar Width: MALES Mandibular Intermolar Width: MALES

55. so 1
54 -I~ Class 2 49 t - O - - Class 2
..........lib
53
52 :t -'-"o='
51 .. ..... . ..... -•
50 ,M ........
49.
43 , ""
.~ 48
42 ,"
47
41 , -"
45 4o ,,"
451 39 ,'"
38 ,,""
"i
43 37 "
42 35 E
Stage 1 Stage 2 Stage 3 Stage I Stage 2 Stage 3
C Deciduous Mixed Dentition Early Permanent D Deciduous Mixed Dentition Early Permanent

Fig. 2. A, Changes in maxillary intermolar widths in female subjects with Class I[, Division 1
malocclusions and normal female subjects. B, Changes in mandibular intermolar widths in female
subjects with Class II, Division 1 malocclusions and normal female subjects. C, Changes in maxillary
intermolar widths in male subjects with Class II, Division 1 malocclusions and normal male subjects.
D, Changes in mandibular intermolar widths in male subjects with Class II, Division 1 malocclusions
and normal male subjects.

II, Division 1 malocclusions and the normal subjects at
the three different stages of dental arch development in
either male or female subjects.
Longitudinal Comparisons. The results indicate
that, in general, there were no significant differences
between the normal subjects and the subjects with Class
lI, Division 1 malocclusions in either the curve profiles
or curve magnitudes of the maxillary and mandibular
arch lengths in both the male and female subjects
(p < 0.05). The only exception was in the male subjects,
where the curve magnitude of maxillary arch length was
significantly larger in the normal group (p > 0.01).
C o m p a r i s o n s of A r c h W i d t h M e a s u r e m e n t s
(Table III)
Cross-sectional Comparisons. The results of the
analysis of variance comparing the differences between
maxillary and mandibular intermolar widths indicate a
significantly larger difference in the normal subjects
than in the male subjects with Class II, Division 1
malocclusions at each of the three stages. No signifi-
cant differences were present in the female compari-
sons of the intermolar and the intercanine widths
measurements at any of the three stages.
Longitudinal Comparisons, Comparisons of the
growth curves for maxillary-mandibular intermolar and
intercanine width measurements indicated the absence
of significant differences between the curves. In other
words, the growth trends were similar in the normal
subjects and the subjects with Class II, Division 1
malocclusions.
On the other hand, the comparisons of the curve
magnitudes indicated that the differences in maxillary-
mandibular intermolar and intercanine arch widths
488 Bishara, Bayati, and Jakobsen American Journal of Orthodontics and Dentofacial Orthopedics
November 1996

Differences between Maxillary-Mandibular Differences between Maxillary-Mandibular

Intercanine Widths in Females Intercanine Widths in Males
1o
-Q- Class 2 1 -O- Class 2
"U" Normal sl'l Normal ,.,'~
=.~ .''

~S | °-" " ' " "

ra~ -"
~= 7 . . . . . . . . . . . . . . . . . . . . . l"
.................... !"

Stage 1 Stage 2 Stage 3 Stage I Stage 2 Stage 3

A Deciduous Mixed Dentition Early Permanent B Deciduous Mixed Dentition Early Permanent

Differences between Maxillary-Mandibular Differences between Maxillary-Mandibular

Intermolar Widths in Females Intermolar Widths in Males
s
--e- Class 2 --t)- Class 2
• Normal ..-ll - m Normal .. 8
7 ..--''''"
.
. ° . - ° ..................... . ........
°-
I. . . . . . . . . . . . . . . . . . . . . 1..° ~*.
v

li'
4

3 3 , ,

Stage 1 Stage 2 Stage 3 Stage I Stage 2 Stage 3

C Deciduous Mixed Dentition Early Permanent D Deciduous Mixed D e n t i t i o n EarlyPermanent

Fig. 3. A, Changes in differences between maxillary and mandibular intercanine width in female
subjects with Class II, Division 1 malocclusions and normal female subjects. B, Changes in
differences between maxillary and mandibular intercanine widths in male subjects with Class II,
Division 1 malocclusions and normal male subjects. C, Changes in differences between maxillary and
mandibular intermolar widths in female subjects with Class II, Division 1 malocclusions and normal
female subjects. D, Changes in differences between maxillary and mandibular intermolar widths in
male subjects with Class II, Division 1 malocclusions and normal male subjects.

were larger in normal male subjects when compared conclusions. The contradicting findings could be
T M

with male subjects with Class I1, Division 1 malocclu- attributed, in part, to the variation in the absolute size
sions (p > 0.05). No significant differences in curve of the persons who comprised either of the two groups
magnitudes were present in the female subjects. compared and the stage of dental arch development at
which they are measured. Such variation in the growth
DISCUSSION trends can be better observed on a longitudinal basis. A
Clinicians and researchers have attempted to evalu- point that can be clearly illustrated in the present
ate the different dental arch parameters of the Class II, investigation where the normal subjects were larger
Division 1 malocclusion. In addition to the obvious than the subjects with Class II, Division 1 malocclu-
anteroposterior discrepancy, the findings on the vertical sions at one stage of dental arch development but
and more important the transverse relationships were smaller at other stages (Fig. 1, A and B). Furthermore,
marked with a certain degree of controversy. Specifi- opposite trends were observed in male and female
cally, the cross-sectional comparisons of arch widths subjects (Fig. 1, C and D and Fig. 2). Staley et al. ~4
between subjects with Class II, Division 1 malocclu- suggested that a more pertinent approach to evaluate
sions and normal subjects did not produce uniform the transverse discrepancies in the dental arches is to
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 110, No. 5
calculate the differences between the maxillary and the
mandibular arch widths. By using this method, it
became obvious that more consistent and interpretable
results can be obtained (Fig. 3 and Table III).
The present findings indicate that, in general, the
differences between the measurements of maxillary
and mandibular intermolar width are larger in the
normal subjects than in subjects with Class II, Division
1 malocclusions. These trends were present in both
male and female subjects but were statistically signif-
icant only in male subjects.
Another important observation is that these trends
in intermolar arch widths were present from the earlier
stages of dental arch development and did not self-
correct. As a result, when evaluating Class II, Division
1 malocclusions, a thorough diagnosis of the transverse
relationship of the dental arches should be performed.
When a discrepancy is present, its correction should be
part of the overall treatment plan and should be
performed at the earlier stages of development. If such
correction is not attempted, the clinician will be pro-
longing the treatment time, because the transverse
problem will be clinically apparent after the correction
of the anteroposterior occlusal discrepancy.
Regardless of the techniques used to measure arch
length, width or shape, in either a two-dimensional or
three-dimensional approach, it is important to remember
that the findings on the maxillary and the mandibular
arches should be related to each other to obtain a more
accurate diagnosis of the extent of the malocclusion.
CONCLUSIONS
From the present findings, the following can be
concluded:
1. The general growth trends for the changes in dental
arch dimensions from the deciduous to the mixed and
permanent dentitions were similar in subjects with
Class II, Division 1 malocclusions and normal sub-
jects. This was true for both male and female subjects
in both the maxillary and the mandibular arches.
Bishara, Bayati, and Jakobsen 489
2. The differences between the measurements of maxil-
lary and mandibular intermolar arch widths were
greater in the normal subjects than in the male
subjects with Class II, Division 1 malocclusions.
Female subjects had similar trends but the differences
were not statistically significant.
3. The relative constriction of the maxillary intermolar
width in the subjects with Class II, Division 1 real-
occlusions was present at the three stages of dental
development. Therefore special attention should be
paid to the transverse relationship during the initial
diagnosis of persons with Class II, Division 1 maloc-
clusions.
4. The transverse discrepancy did not seem to self-
correct from the deciduous to the permanent den-
titions.
REFERENCES
1. Ast DB, Carlos JP, Cons DC. Prevalence and characteristics of malocclusion among
senior high school students in up-state New York. Am J Orthod 1965;51:43%45,
2. Goldstein MS, Stanton FL. Various types of occlusion and amounts of overbite in
normal and abnormal occlusion between two and twelve years. Int J Orthod 1936;22:
549-69.
3. Massler M, Fraukel JM. Prevalence of malocclusion in children aged 14-18 years. Am
J Orthod 1951;37:751-68.
4. Ahemus LA. Frequency of the incidence of malocclusion in American Negro children
aged twelve to sixteen. Angle Orthod 1959;29:189-200.
5. Adams CR Kerr WJS. Overbite and face height in 44 male subjects with Class I. Class
Ill1 and Class II/2 occlusion. Eur J Orthod 1981;3:125-9.
6. Burstone CE. A study of individual variation in mandibular bicanine dimension during
growth. Am J Orthod 1952;38:848-65.
7. DeKock WH, Dental arch depth and width studied longitudinally from 12 years of age
to adulthood. Am J Orthod 1972;62:56-66.
8. Knott VB, Longitudinal study of dental arch widths at four stages of dentition. Angle
Orthod i972;42:387-94.
9. Meredith HV, nopp WM. A longitudinal study of dental arch width of the deciduous
second molars in children 4-8 years of age. J Dent Res 1956;35:879-89.
I0. Moorrees CFA, The dentition of the growing child. Cambridge, Massachusetts: Harvard
University Press, I959.
11, Sillman JH. Dimensional changes of the dental arches: longitudinal study from birth to
25 years. Am J Orthod 1964;50:824-41.
12. Frohlich FL A longitudinal study of untreated Class II type malocclusion, Fr Eur
Orthod Soc 1961:137-51.
i3. Frohlich FJ. Changes in untreated Class iI type malocclusions. Angle Orthod 1962;32:
167-79.
14. Staley RN, Sluntz WR, Peterson LC. A comparison of arch widths in adults with
normal occlusion and adults with Class II, division 1 malocclusion. Am J Orthod
1985;88:163-9.
15. Kleinbanm DG, Kupper LL. Applied regression analysis and other multivariate
methods. Boston: Danbury Press, 1978:181.
16. Green PE. Analyzing multivariate data. Hinsdale, Illinois: Dryden Press, 1978.