Date post: | 30-Aug-2018 |
Category: |
Documents |
Upload: | nguyenkien |
View: | 213 times |
Download: | 0 times |
DOCUMENT RESUME
ED 062 019 PS 005 629
AUTHOR Hines, Brainard W.
TITLE Analysis of Visual Perception of Chilren in theAppalachia Preschool Education Program.
INSTITUTION Appalachia Educational Lab., Charleston, W. Va.
REPORT NO TR-16BUREAU NO BR-6-2909PUB DATE Dec 71NOTE 27p,
EDRS PRICE MF-$0.65 HC-$ .29DESCRIPTORS *Data Analysis; *Educational Programs; Educational
Research; Educational Television; ExperimentalPrograms; Measurement Instruments; NonprofessionalPersonnel; Pilot Projects; *Preschool Children;Program Evaluation; Psychomotor Skills; TaskPerformance; *Testing; Test Results; *VisualPerception
IDENTIFIERS Appalachia Preschool Education Program; *FrostigDevelopmental Test of Visual Perception; WestVirginia
ABSTRACTA description and application of the Frostig
Developmental Test of Visual Perception are presented. The reportincludes a description of the Frostig as a total instrument and anoverall analysis, as well as separate analyses and descriptions ofeach subtest. A brief summary of the experimental design andsamplingplan also is included. A'pilot study included administration of theFrostig to a sample of children, ages 3 to 6 years, in northern westVirginia. The results of this testing indicated overall deficits in
the areas of figure7ground perception.and form-constancy tasks.Results of the study indicate that paraprofessiOnal personnel made asignificant contribution in the_area of same-different discriminationifi terms of spacial rotation and that the television-program had itsmajor effect on eye-motor coordination, shape'constancy, and theability to conserve patterns after-spatial rotaticin. (AuthoriCK)
U.S. DEPARTMENT OE HEALTH.EDUCATION SI WELFAREOFFICEpF EnucATIoN
THIS DOCUMENT HAS SEEN REPRO-DUCED EXACTLY AS RECEIVED FROMTHE PERSON OR ORGANIZATION ORIG-INATING IT. POINTS OF VIEW OR OPIN.IONS STATED DO NOT NECESSARILYREPRESENT OFFICIAL OFFICE OF EDU-CATION POSITION OR POLICY.
ual Perceptione Appaiachia
cation Program
t
Dete e 971
?
TABLE OF CONTENTS
Chapter 1
ANALYSIS OF VISUAL PERCEPTION OF CHILDREN IN THE
6
..
6 ..
. .
............
.
.
..
6 6
1
1
3
3
6
9
13
16
1922
23
APPALACHIA PRESCHOOL EDUCATION PROGRAMIntroductionMethodDescription of Instrument and Results of Analysi
Frostig Subtest 1 (Eye Motor Coordination)Frostig Subtest 2 (Figure-Ground)Frostig Subtest 3 (Constancy of Shape)Frostig Subtest 4 (Position in Space )Frostig Subtest 5 (Spatial Relationship ..........
Frostig Total Raw ScoreSummary and Implications 6WP 0.*0 ...........................
REFERENCES .. . . f 6 .. .................List of Tables
Table
16.1 Frostig Subtest 1 (Eye Mbtor Coordination) Pre and Post-testMean Raw Scores, Standard DeviatiOns, and Numbers of Sub-jects by Age and Sex within Treatment Groups . ..... 4
16.2 Analysis of Covariance Table for Fro t4g qflhfst 1 ff. . .o0o. 6
16.3 Frostig Subtest 2 (Figure7-Ground) Pre and Post-testMean Raw Scores, Standard Deviations, and Numbersof Subjects by Age and Sex within Treatment Groups 7
16.4 Analysis of Covariance Table for Frostig Subtest 2
16.5 Adjusted Post-test Means for Males and Females onFrpstig Subtest 2
16.6 Frostig Subtest 3 (Constancy of Shape) Pre and Post-testMean Raw Scores,. Standareviationsi and Numbers ofSubjects by Age and Sex within Treatment Groups 10
16.7 Analysis of Covariance Table,for Frostig Subtest 3 12
16.8 Adjusted Post-testMeans for Males and Femaleson Frostig Subtest 3 12
Table
16.9 Frostig Subtest 4 (Position in Space) Pre and Post-test
Mean Raw Scores, Standard Deviations, and Numbers ofSubjects by Age and Sex within Treatment Groups 14
16.10 Analysis of Covariance Table for Frostig Subtest 4 15
16.11 Frostig Subtest 5 (Spatial Relationships) Pre and Post-testMean Raw Scores, Standard Deviations, and Numbers of Sub-jects by Age and Sex within Treatment Groups ........... .. 17
16.12 Analysis of Covariance Table for Frostig Subtest S 18
16.13 Frostig Total Pre and Post-test Mean Raw ScoresStandard Deviations, and Numbers of Subjectsby Age and Sex within Treatment Groups- 20
16.14 Analysis of Covariance Table for Frostig Total Score 71
16.15 Adjusted Post-test Total Frostig Mean Scores forMales and Females ............ .. . ...... ............. . .. 21
List of Figures
Figure
16.1 Adjusted Post-test Group Scores and Normative Meansfor Comparable Age Groups--Frostig SubteSt 1 5
16.2 Adjusted Post-test Group Scores and Normative Meansfor Comparable Age Groups--Frostig Subtest 2 .... ....
16.3 Adjusted Post-test Group Scores and Normative Meansfor Comparable Age Groups--Frostig Subtest 3
16.4 Adjusted Post-test Group Scores and Normative Meansfor Comparable Age Groups--Frostig Subtest 4 15
16.5 Adjusted Post-test Group Scores and Normative Means
for Comparable Age Groups--Frostig Subtest 5 18
16.6 Adjusted Post-test Group Scores and Normative *ansfor Comparable Age Groups--Frostig Total 19
Analysis of Visual Perception of Children in the
Appalachia Preschool Education Program
Introduction
A large proportion of the initial Preschool Education Program curriculm
(Hooper and Marshall, 1968) and the subsequent program materials are devoted
to teaching motor coordination and perceptual learning tasks. Because of
the difficulty of developing specific measures for these cbjectives and
because of its previous use in evaluating preschool programs, the Marianne
Frostig Developmental Test of Visual Perception was used to measure behavioral
change in these areas.
As a program evaluation instrument, the Frostig shares several problems
with the ITPA. Neither was designed to be a specific measure of program
effects; both are administered individually, and thus are susceptible to
examiner bias. In addition, both instruments were designed to be diagnostic
of perceptual, expressive and motor deficits rather than to measure overall
treatment effects. However, the Frostig test does provide a potentially
valid and reliable measure of perceptual development in the preschool-age
child, and most important, it provides national norms for comparison with a
sample outside of the AEL region.
This report includes a description of the Frostig as a.total.instrumeit
and an overall analysis, as well as separate analyses and descriptions of
each subtest. A. brief summary.of the e.xperimental design and sampling plan .
also is included.
.MethOd
A pilot -stpdy_by Hooper and.--:MArshall'f968 moltded:ubinistr4tion Of
the,-Frostig to a sample of children, ages 3 o years in'horthernWet--
Virginia. .The results-Of:ths'Aestingjfidicated-olferall defidts
areas of figure-ground perception and form-constancy tasks as measured by
Subtests 2 and 3. Both of those skills were cited by tlle test suthors as
being relevant to reading readiness. In order to investigate the Pooper-
Marshall data, and as a method of measuring program effects on motor learning
and perceptual development, the Frostig was included as a major evrduation
instrument for the Appalachia Preschool Education Program.
Data were collected in June and September of 1970 from a s,mple of chil-
dren in three treatment groups and a control group. The three treatment
groups were:
A group which received an instructional television program in
their homes (TV only).
A group which received the TV program plus weekly visits by a para-
professional home visitor (TV-HV).
A group which received the TV program, a home visit, and which also
visited a mobile classroom CTV-HV-MC).
These pretest data wers compared to results of a post-test administered
in June of 1971, Details of the sampling procedure and a description of
each group are included in the introduction to this report.
The primary data analysis technique consisted of analysis of covariance
using post-test data, with age in months as of June, 1971, and the PPVT post-
test raw score as covaristes. A preliminary three-way analysis of variance
on nost-test scores revealed treatment differences in the covariates. Similar
analyses were performed on gain scores, and these results will be included
where appropriate.
The nature of the statistical methods used Precludes definite cause and
ffect reasoning; however, the existence of significant differences between
group means when these differences were not evident on-the pretest was assumed
to reflect genuine .treatMent effects.
It is assumed that the use of chronological age As d ceVariate adjusts
for differences iP moan age:among treatment groups.- It should be noted,
however, that the group with.the highest mean age (TV's:Inly) produced the'
lowest mean scores on most subtests of the evaluatien-battery. Thu$, the
use of coVariance analysis was a.conservative procedure and may haveobscured
actual program effects.
The eta2 values,pressources: treatment sex
term. They represent the
nted in the ANCOVA tables-are derived only-for four
the interaction.of these variables,, and the error
proportion- of variance accounted for by each source.
Description of Instrument and Results of Analysis
The Frost g is designed primarily as a method of assessing perceptual
development in the visual area. However, hand-eye coordination and overallmotor skills are involved in the child's responses to the visual configura-
tions on the test. With the Frostig, unlike ITPA, it is difficult to
separate the receptive processes (visual acuity) from the mediating activity
(figure recognition) and the expressive act (drawing, outlining, etc.). For
this reason the instrument is considered to be an indicator of perceptual-
motor development in the program evaluation plan, and no effort is made to
-eparate the two areas of functioning which may cause variance on the total
test scores. Each of the following suhtests has fairly high "face" validity,
and it is assumed that each measures the general area to which it is attri-
buted. Technical Report No. 17 further explores the factorial validity of
this test.
Frostig Subtest Eye Motor Coordination)
The authors state that this subtest is "... a
tion involving the drawing of continuous straight,between boundaries of various width, or from point
lines." (Frostig, 1966.)
test of eye-hand coordina-curved, or angled linesto point w thout guide-
Mean raw se res, standard deviations, and numbers of subjects according
to age, sex, and treatment group for Subtest 1 are presented in Table 16.1along with overall treatment group means for pre and post-test measures.
The adjusted means from the ANCOVA are graphically represented in Figure 16.1
along with the mean scores of the Frostig normative sample. Note that the
adjusted means in Figure 16.1 differ slightly from the raw score total means
in Table 16.1.
As shown in Figure 16.1, two treatment groups and the control groupscored slightly below the ribrm, while one group (TV-HV-MC) slightly exceeded
the norm. The analysis of covariance summary table is presented in Table 16.2.
The treatment effect was significant at:the .01 level of confidence.
Ta Else 16.1
Frostig Stbtest 1 (tye Motor
Coordination) Pre and Post-test
Mean Raw
Sdores, Standard Deviations,
and Nuthers of Subjects by Age,
and Sex within Treatment Groups
Age
'
TV-HV-MC
TV-HV
TV onl
Control
1
Pretest
R=
4.83
N= 10
Post-test
R=
5.50
N= 12
Pretest
ii=
6.58
N. 11
Post-test
R=
5,13
N= 16
Pretest
-5.13
N= 11
Pos -test
i=
5.50
N=
8
Pretest
i=
2.63
N.
8
Post-test
i=
5.24
N.
17
0=
5 61
0=
1.85
0=
5.88
,
O.
2.72
0.
6,68
=2.96
a=
2.60
,
a=
1.66
=6.67
=838
=3.30
.4.61
x=36O
=5.00
'
x=
2.50
x=
4.58
N= 11
N= 13
N= 10
N= 18
N5
=5
-5.20
a=
3.58
a=
3.98
o=
2.26
H=
2.24
H.
4.20
'
=2.12
0=
3.31
1
-9.67
.2.25
i
x= 10.20
-.
7.96
--
8.15
x=
10.79
=-47-37-
y=
6.74
N=
9N= 20
N=
9N=25
N= 13
N= 14
'
N. 13
N= 17
=6
40 -
3.15
q =
3.87
H
-3.09
-6.04
0=
3.14
G=
2.14
=3.42
8.7
=9.0
-= 15.00
-.
9.39
x.
9.60
- 10.39
5.54
1
x=
6.95
N=
9N= 19
N= 10
N= 22
1
N.
101
N=
8N= 13
N= 17
-7.92
0=
3.14
o=
4.35
0=
3.44
0.
7.06
g=
5.17
g2.40
1
0_
2 77
x. 14.50
- 12.31
x= 19.60
H= 11.22
R= 13.60
-.
9.83
-7.78
ii=
9.67
- 13
= 16
II
N=
8N= 27
N=
8,
. 12
N=
9N= 18
a=
8.11
0=
3.69
' a=
4 37
a=
2.90
a=
8.15
'
g=
2.76
a=
4.09
1
0=
3.42
x= 10.40
-. 11.40
2= 10.401
= 10.77
R= 13.54
I
R= 11.11
-=
01
FN= 11
N=
15
N=
10
N= 22
= 13
N= 19
N= 10
N= 15
0=
5.32
1
0=
2.85
0=
2.95
0=
3.04
a=748
a=
3.32
0 =
3.78
0=
2.63
x=
9.37
x=
9.39
.
10.48
R=
8.52
=9.35
ii=
9.58
i=
5.57
ii=
7.16
TOTAL
N=
63
N= 95
N=
58
N= 130
N= 60
N= 66
N= 61
N.105
-6.55
G=
3.13
o=
4.36
o=
4.36
a=
6.77
cf=
3.51
g=
2.91
a=
2.95
,
12
6
Treatment Mean
Normative Mean
9.39 9.35 8.58 9.35 9.42 1001 7.71 8.75TVHV-NC TV-HV TV only Control62 mos. 62 mos. 65 mos. 59 mos.
F1gure 16.1
Adju- ed Po -test Group Scores and Norma iveMeans for. Comparable Age Groupsr
F-ostig Subtest 1
Table 16.2
Analysis of Covariance Table for Frostig Subtest 1
Source n2 d.f. Nea.n Square_
Trt. .023 3 62.34 5.94 p < .01
Sex .03 1 23.15 2.20
Trt. by Sex .01 3 3.10 0.30
Covs. 2 14976.81 1426.18Cov. 1 1 386.27 36.78Cov. 2 1 227.91 21.70
Error .73 392 10.50
A Dunnett's post-hoc covarison indicated that the contrasts betweenthe three treatment groups and the control group contributed to the signifi-
cant F-ratio. Differences between treatment group means were not signifi-cant, indicating that the television component probably was responsible forthe higher level of performance of children in treatment groups. A similarpattern was evident in the previous year's data. The relatively smalldifferences between the treatment group means and the normative means do
not seem to be indicative of an overall trend to lower scores for the Appa-
lachian sample.
The television program stresses objectives whiCh involve the child withthe use of paint brushes, crayons, and other art materials which may welladd to the overall level of perceptual-motor functioning measured by this
subtest.
Frostig Subtest 2 ( igure-Ground)
Frostig Subtest 2, according to the authors, is a measure of abilityto perceive .. shifts in perception of figures against increasingly complexgrounds. Intersecting and 'hidden' geometric forms are used." (Frostig,
1966.) Subtest 2 is thought to correlate highly with reading readiness, in
that it measures the ability to discriminate shapes and configurations from
their context. The same ability is necessary in recognition of letters andnumbers in written mat,i'.Tial.
Table 16.3 shows mean raw scores, standard deviations, and numbers ofsubjects.for each age7by-sex cell within:the four treatments. It also .shows
the same parameters collapsed acrosS treatment groups'.
The adjusted overall means from the ANCOVA are presented graphically.
in Figure 16.2 along with the national norms for tho ooMparable ages in
each treatment condition.
Table 16.3
Ftostig Subtest 2 (Figure-Ground) Pre and Post-test Mean Raw
Scores', Standard Deviations, and Numbers of Subjects
by Age and Sex within Treatment Groups
....ge
,
Sex
TV-HV-MC
,
TV-HV
TV only
Control
'Prete,st , ,!
)=..
0, 50 ....
I,.
...,,: N= lo
:.
:.
!.
....
.-
,_
.,.
.161,1517, ..
,,
'Post-tes
.e est
R* 3,50
x=
.1.5O
, N=. 12
.
N= 11
..
0=
2.96
a* 2,06
Post-tes
.
.. R=,
5.81
N= 16
0 = 3.81
P .etest
k=
0... go
7.1= 11
G=
1.00
Pos ...tes-
i=
2.00
N=
8
C=
2.40
,
Pretest
*=
1.50
N=
8
a= '2.55
Post-test
i=
3.82
N= 17
0=
3,01.
.:=... 1,81 :
1.
F.
N= IL-.
-.
,=
.1.75,-
x=
5.92. '
,..N,,, 13
....Ia.:- 3.97
R=
-1'. 20
N= 10
a =
1.46
R=
5,2'
N= 18
0 -
4.59
i= 2.60
N=
5
.0
3.56
x=
2.60
N=
51
0 -
3.77"
i=
2.1
1N=
8
C =
2.52
.
=4.53
N= 19.
0.--::
3,.9.5,
.15=
3,25
:',.
N7,
9.H '',.
d=
..3469::H
'7-
9,75
1= 2.89
,..14720.
N=
9.
.,
:::OF:
4,...77.
--.1,:r..0=.
5,84
=9.56
N=, 25,
0=. .4.82
'R.=
5.00
N=- 13
a=-H. 5.44
,
R.= 11-11.
NF .14.
0=
4.6.
R=
4.15
N=13,
0=
5.01,
R=,
7.32
N= 17
0=,
5.58.
,x=. ,:2,15:
..
.N= ,..9.
.
-a= ...:41..A7 :
,...x*, '10.14:
-, x=
5..10
.,
..
1.
..
.
'N=191'..!..
1! N= 10..
.,
-6,318..
:'
o''.=
5.70
.x= 11,7
N= 22
0=
4.80
-x=
1.1.
N=10
c=
1.60.
x=
8.75
N=
8
c=
6.12
x=
4.62
N= 13,
1=.
5.72
i=
9.55
N= 17'
0=
5.01
.."'.5:-'
.,
..
.'.''.
.,:7,92'.:..:.
'N=..13.....',.'..
H.'
.:0* 6.30.:
::!:1=:..15,00.
g=12,.40
11=:-...16.
..........
,=
8.
.
.T.'07=.- .8.01.-.!
4.10
= 15.04:
:= 27
a = 4.48
M=. .4.50
N=
8
,G=
5,31
R= 11.42
N=. 12.
0=.
5.51'
,=,
8,61'
.......,
.9-
=6.14
x= 72.50
,
N= 1,8
0=
5.10
'(..7.*.-1L7,2' -HT-R=13,27:-
N= 11'.'
,,.
.,
..
.
! .0= H6.42:
=9,10
* N= j. 5- .'.
',
.N= 10
..
4.86"
-2=
6,19
,
R= 15,64
1N= 22-
03.81
R.=
6,84
Nr-- 13
0=
6,00
x= 13.68
N= 19.
0=- 5,05
= 11.00
I= 10
G=
6.18
g= 13.75
'N= 15
0=
4.89
..
,
-..:,TpTAL!:
.
1.......
...
x=':4,41..
....x=,..10.09..
N=-,163-H
T'll':..N=
Ici.-
4.41.
,1:0=1
,
_....
..
95.. :
14.91
....
k= .. 5,10
1 N= 58
-7=
4.66
.
it= 11.02
,N=130
:=
4.45
R=
3.73
N= 60
7.1= -.4,44
1
i7 10.00
N= 66
0=
4.98
)::=..
51.43,
N= 61
-7=
5.08.
g=
8,62
N=1.05
G=
4.77
3 2
6
0
Treatment Mean
ormative Mean
10.06 11.30 11.01 11.20 9.98 12.80 8.59 10.30
TvHy-nc TV-HV TV only Control62 mos. 62 mos. 65 mos: 59 mos.
Figure 16.2
Adjus -d Post-test Group Scores and NoatiieMeans for cotparable Age Groups
Frostig Subtest 2
A three-way ANOVA on post-test scores showed a statistically signifi-cant (p < .005) treatment effect for Frostig Subtest 2. Since the ANCOVAdid not show this effect, it is assumed that the significant ANOVA treat-ment effect was attributable to a factor which was controlled in the ANCOVA.Similar results were observed in tilt_ previous year's data.
The summary of the analysis Of covariance is presented in Table 16.
Table 16.4
Analysis of Covariance Table for Frostig Subtest 2
Sou_ ce n2
d.f. Mean S-uare
2.036.280.15
805.710.54
53.30
.05
Trt.SexTrt. by SexCovs.Cov. 1
Cov. 2
Error
.02
.02
.001
.97
3
1
3
2
1
1
392
52.50162.40
3.7920825.35
13.891377.61
25.85
Table 16.5 shows adjusted means by sex for each treatment collapsedacross age. As can be seen from these data, femaJes outscored males in allfour treatment groups. This result is unusual in the light of the findingsof an earlier study (Hooper and Marshall, 1968, p. 77) which found signifi-cani male superiority on this measure in the 51/2-year-old group.
Table 16.5
Adjusted Post-test.Means for Males and Females o- Frostig Subtest 2
TV-F1V-MC TV-HV TV
9.94 10.85 9.32 8.23
10.21 11.19 10.72 9.00
An ANCOVA on gain scores, using age and PPVT raw score as covariates,did not show significant sex effects, although it did identify a signifi-cant treatment effect. It seems likely that the differential treatmenteffects served to equalize post-test scores across groups.
Frostig Sub e t 3 (Constancy, of Shape
This section is designed to measure "...-recognition of certain geo-metric figures presented in a variety of sizes, shadings, textures, andpositions in spaces, and their discrimination from similar geometric figu es.Circles, squares, rectangles, ellipses, and parallelograms are used."(Frostig, 1966, p. 5.)
Tab
le 1
6.6
Fros
tig ,S
ubte
st 3
(C
onst
ancy
of
Shap
e) P
re a
nd P
ost-
test
Mea
n R
aw.
Scor
es, S
tand
aid,
Dev
iatio
ns ,
and
Num
bers
of
Subj
ects
by
Age
and
Sex
With
in T
reat
men
t Gro
ups
.e
Sex
TV
-HV
-MC
ITV-HV
TV only
Control
Pretest
Post-test
Pretest
Post-test
Pretest
Post-test
Pretest
!Post-test
5E=
0.20
Tc=
2.50
i= 1.25
R= 4.00
k= 0.90
R= 0.38
k= 1.75
i= 2.94
MN
.= 1
0N
= 1
2N
= 1
1N
= 1
6N
= 1
1N
=8
N=
8N
= 1
7,
a= 0.40
0= 2.60
0= 1.92
a= 1.73
a= 1.24
0=0.48
a= 1.56
o= 1.86
X=
2.0
0=
4.8
5R
= 1
.10
x= 3
.06
k= 1
.40
=2.
20x=
1.3
81.
89=
11
M13
:N=
10
N=
18
N=
5N
=5
N=
8N
= 1
9
a=
1.90
0= 3
.21
0= 1
.87
a= 2
.56
0= 1
.20
0- 2
.04
o= 1
.49
= 2
.48
3.50
x= 6
.05
Fc=
3.78
!3-
r-t
4.60
5.i=
3.23
Fc!=
6.43
x= 3
.15
k= 3
.00
1N
=9
N=
20
N-
911
N=
25
N=
13
N=
14
N=
13
N=
17
40=' 3.20
0 = 3.25
.'.
=4.00
a= 3.22
o= 3.40
c7=
4.58
!=2.
470=
2.2
71
r--
4.50
=8.
58x=
4.0
0=
'7.4
8.-
=1.
40x-
5.0
0X
=2.
92x=
3.50
1!"
1/4
=9
!N
= 1
9!
Nr=
10
!N
= 2
2N
= 1
0N
=8
N=
13
!N
= 1
71
u="
3:16
0= 3
.69
10=
2.30
!0=
3.7
30=
2.0
6cs
=4.
18u=
2.4
00=
2.1
41.
51=
3.54
k= 10.13
5-c=
5.38
R=
7.33
it=
3.25
R= 6.17
x=
2.01 1
k= 4.56
i1
N=
13
I= 1
6N
=8
!
1
N=
27
N=
8N
= 1
2N
=9
N=
18
0= 3
.20
0 =
3.6
20-
110=
3.21
0= 2
.87
o= 3
.95
G.=
1.73
0= 3
.50
5t=
' 5.8
1ii=
7.80
:Tc=
6.20
ii=8.
45R
= 3
.69
=6.
84x=
4.8
0)1
=5.
56F
N=
11N
= 1
5N
= '1
0N
= 2
2N
= 1
3N
= 1
9t!
= 1
0N
= 1
5
0=.
4.00
0=3.
780
= 4
.54
a=4.
410=
4.4
43=
4.02
0=4.
390=
3.21
x=3.
27x=
6.91
Si=
3.51
5c=
6.02
R=
2.4
5x-
5.27
'1.i=
2.79
i=3.
7'.-
TO
TA
L_
N=
" 63
N=
95
N=
58
N=
130
N=
60
N=
66
N=61
I.,7
=10
5
=2.
900
r=3.
450=
2.87
0 =
3.32
tI=
3.07
-=3.
79.1
=2.
600=
2.6
4
Since these figures are relative common, it would be expected thatthe child's familiarity with the shapes would correlate positively with hisperformance on the test. Many of the program's curriculum objectives doinvolve recognition and labeling of these and similar geometric figures andthus could be expected to influence scores on this subtest.
Table 16.6 presents means, standard deviations, and numbers of subjectsfor each age-by-sex-cell within the four treatment groups, as well as for
groups on both pre and post-test measures. Overall means, adjusted for ageand PPVT raw score from the analysis of covariance, are presented in Figure16.3, along with normative-means for equivalent age groups.
1
Treatment Mean
auNormative Mean
6.90 5,40. 6.01 5 40 5,27 59.0 3.74 4.90
TV ',only .Control'mos . 62 ino. 65 mOs= 59 mos.
Figure 16.
Adjusted Post-test Group Scores and Normativemeans for Comparable Age Groups - -
Frostig Subtest 3
Table 16.7
Analysis of Covariance Table fox Frostig Subtest 3
Source n2 d.f. Mean Square r E
Trt. .07 3 110.84
_
9.84 p < .005
Sex .02 1 100.87 8.95 p < .005
Trt. by Sex .00 3 2.08 0.18
Covs. 2 6282.82 557.69
Cov. 1 1 0.76 0.07
Cov. 2 1 540.61 47.99
Error .91 392 11.27
Table 16.7 summarizes the results of the ANCOVA on post-test scores for
this subtest. Treatment effects for Subtest 5 seem to stem from the linear
increments apparent in Figure 16.3. A Dunnett's post-hoc comparison failedto reveal any single pair of means whose difference was great enough to be
significant. Insi-Jection of Figure 16.3 suggetAs that both the televiSionprogram and home visitor contributed to children's performance on tasksmeasured by Subtest 3, a finding which was also evident in the previous
year's data. Also, it is of interest that the two groups which were visitedby a paraprofessional scored above the national normative means for comparable
age groups while the other groups scored below the norms.
The significant sex effect on this subtest is illustrated in Table 16.8which gives adjusted means for boys and girls of all ages over the four treat-
ment conditions. As was the case for Subtest 2, females outscored males in
all groups. The test's author cites Subtests 2 and 3 as being related toreading readiness in first grade children. It may well be that girls aremore receptive to this type of learning experience. In the pilot assess-ment, Hooper found female superiority only in the 41/2-yearTold group, whilemales scored significantly higher in the 51/2-year-old group. The lattergroup most c osely resembles the-mean age in the four treatments.
16.8
Adjusted Po t test, Means for Males and Femaleson-FrOstig Subtest' 3.
TITHV7NIC TVEIV
Since it is not possible to show that the mobile facility and the homevisitor added significantly to the learning the children received from thetelevision program, it may well be that the identification and labelingwhich took place as the Child was exposed to the visual stimuli wereresponsible for the differences found. It also is possible that the para-professional and the teacher did not stress recognition of geometric shapesas the television program did.
Frostig Subtest 4 (Position in Space)
This particular measure involves "... the discrimination of reversalsand rotating of figures presented in series. Schematic drawing representingcommon objects are used." (Frostig, 1966, p. 5.)
The ability to follow explicit directions and the ability to comprehendthe meaning of same and different probably are involved in making correctresponses to this subtest. Table 16.9 presents the raw score means andnumbers of subjects for each age by sex and treatment cell for FrostigSubtest 4.
For purposes of comparison, the adjusted post-test means are presentedin Figure 16.4 along with national norns for each treatment group. Averageages in months are given below each group.
As is apparent from Figure 16.4, differences exist between treatmentgroups as Well as between treatment and:normatiVe groups. The implicationsof these differences are clarified in Table.16-.10 whielLpresentS aisutmaryof the ANCOVA results for Subtest:4.
Table 16.9
Frostig Subtest 4 (Position
in Space) Pre and Post-test Mean Paw
"
Scores
Standard Deviations ,
and Numbers of Subjects by Age
and Sex within Treatment Grouns
A e
.,TV-HV-MC
TV-HV
TV only
Control
1
,Pretest
Post-test
Pretest
Post-test
Pretest
Pos
- est
Pretest
Post-test
5.(=
1.75
2=
.
2.75
=2.67
1
5(=
*
3.19
i=
1..82
i=
1.75
X'=.
2.00
i=
2.12
N=
10
N=
12
11=
11
N=
16
N= 11
-
N=
8N=
8N= 17
'=
1.6
l'-=
1.30
0=
1.11
a=
1.18
0=
1.19
a=
0.97
0=
1.12
a=
1.23
,
=2.33
i=
3.100
=2.00
R=
2.28
X=
1.80
x=
1.40
x=
1.88
=1.89
= .11
.N= 13
.=
10
N=
18
N=
5N=
5N=8
N= 19
0=
1.11
0=
O. 96
=1.18
a=
2.41
0-
0.40
0=
1.02
0=
1.17
=O. 91
3.00
.
=4.55
X=
3.22
=4.24
=3.69
R=
4.07
x=
2.62
x=
2.42
-N=
8N=
20
N=
9'
N=
25
N= 13
N=
14
N= 13
N= 17
' 4
1-
0.70
=1.47
0=
01.83
a=
1.82
0=
1.43
0=
1.83
a=
1.12
a=
1.33
-=
2.44
=3.719
-3.20
R=
6.43
=3.50
cc=
2.50
=2.00
=2.80
, ,
'
. 9
,
.
N=
1
=19 1.,36
N= 10
-=
0=
22
12.82
N= 10
0=
N8
0 =
1.12
N= 13
0 =
N= 17
00.99
4.62
.R=
5.69
i=
5.13
4.78
X=
3.25
=4.58
i=
3,33
x=
3.69
N=
1= 16
N=
8N=
27
N=
8.
,
N=
12
N=
9N= 18
50.
.0,96
-1.57
a=
1.64
g=
1.81
g=
1.58
0=
1.93
=1.58
0=
1.37
'
=4.00
=5.20
x=
4.40
R=
5.05
Fc=
3.92
x=
3.84
X=
2.60
X=
3.31
11
N=
15
.N=
10
11N=
22
N=
13
N=
/9
N= 10
N= 15
,
,1.95
=1.42
0=
1. 3
-1 66
g=
1.66
G=
2.01
0=
1.17
-=
0.6
=3.11
R=
4.25
R=
3.37
-
4.48
R=
3.15
R=
3.42
i=
2.41
x=
2.73
-' T"OTAL
'
,
=63
N=
5N= 58
N= 130
N=
60
'
N=
66
N= 61
N=105
-=
1.25
a= 1.38
j= '1.24
O=
5.53
0=
1.28
a=
1.68
0=
1.10
0=
1.11
6
5
4
3
2
0
Treatment M anNormative Mean
4.25 4.20 4.47 4.20TV -RV-- MC TV-I-1V62 mos. 62 mos.
'-"Figure
.42.5.20TV only65 mos
2.73 3.90Control59 mos.
Adjusted Post-test Group Scores and NormativeMeans for 'Comparable Age Groups-
Frostig Subtest 4
Table 16.10
Analysis of Covariance Table for Irost i g Sub test 4Sou rce
Trt.SexTrt . by SexCovsCoy- 1.
Coy= 2Erro
2r1- .d.f.
.03 3.00 1. 004 3
1. 97 392
Mean.Square
45.42 3.930.07 0.017.09 0.61
2773.76. .240.00
8 -.66 0.75
143-64 -12. 43li-.. S6
.05
(-15)
A Dunnett's post-hoc analysis indicates that the two groups visited bya paraprofessional scored significantly better than those which did notreceive visits from a paraprofessional. Since no effects were evidencedfor the television program, it is likely that the home visitor was respon-sible for the learning which took place in this area of perceptual develop-ment. It is also of interest to note that the two groups visited by theparaprofessional scored at Or near the national norms, while the other twogroups scored below the norms. A similar pattern also appeared in thesecond year's data.
Frostig Subtest 5 (Spatial:RelatiOnShips)
This section of the Frostig is said to measure theability to make... the analysis of simple forms and patterns. These consist of lines of
various lengths aad angles which the child is required to copy,.using dotsas guide points (Frostig, 1966, p. 6,)
Essentially, this task involves transposing a specific configurationof lines onto a set of dots, a task which involves both motor coordinationand short-term memory for visual designs. Scores for each age-by-sex cellwithin the four treatments are reflected in the data presented in Table16.11. Figure 16.5 presents a graphic comparison of the adjusted overallgroup mean scores on Subtest 5 with normative means for children of similarages.
Table 16.12 summarizes the res lts of the analysis of covariance onpost-test raw scores for Subtest 5. A Eunnett's post-hoc comparison ofthe four means showed that the two groups receiving visits from the para-professional (TV-HV and TV-HV-MC) outscored their counterparts in two othergroups who did not receive these visits. The logical conclusion is thatthe paraprofessional effectively, teaches objectives measured by Subtest 5,which deal with rotation in space and discrimination of "same" and "different"figures. Neither the television program nor the mobile classroom alone hada measurable effect on the area of visual-motor development.
(16)
Table
1Frostig Slubtest 5 (Spatial Relationships) Pre and
Post-test Mean Raw
; Scores, Standard Deviations, and Nunbers
of Subjects by Age
and. Sex within Treatment
Groups
Age
Sex
TV-NV-MC
T -HV
TV only
Control
Pretest
ost-test
Pretest
post-test
Pretest
Post-test
Pretest
Post-test
R=
0.30
067
x=
0.18
k=
1.19
k=
0.316
x=
0.25
11
*=
0.00
R=
0 .47
,N= 10
2= 11
N= 16
-11
I
N=
8N=
8N=
17
a=
0.46
0=
0.39
=1.33
G=
0.48
0=
0.43
c=
0.00
10.=
0.70
3i.=
0.41
1 .31
-=-
101.14
5 -
0.61
x=
0.201
R=
0.20
R=
0.13
k=
0.68
N= 11
N1
=10
N18
N=
5N=
5N=
8N= 19
0=
0.50
a=
1.26
0.35
G =
0.91
0=
0.40
0 =
0.40
0=
0.33
a=
0.73
R7
0.89
X=
2.65
X =
101.78
X -=
1.96
R=
0.92
R=
1.93
R=
0.62
R=
0.56
MN=
. 9
N=
20
i=
9N=
25
N=
13
t@14
N=
13
N= 17
,a=
0.93
0=
1.71
=0.97
0=
1.71
0=
0.86
0 =
1.39
G=
0.96
0=
1.08
k=
0.75
R=
279
X -
1.00
1R=
2.57
x=
0.50
R=
1.50
x=
1.08
R=
1.010
-
N=
9N
=19
N = 10
N=
22
N= 10
N =
8N=
13
N= 17
, a=
1.03
a=
- 1.96
=1.41
O =
1.85
O =
0.70
O =
1.12
0=
1.60
a=
1.15
2.00
R=
4.25
R=
2 00
R=
3.74
X =
01.63
R=
2.67
=1.67
x=
2.76
'
N=
13
II
N=
16
N=:.8
N= 27
N=
8N=
12
N=
9N= 18
I
a=
1.73
I
a=
1.56
-7=
1.85
0 =
1.88
a=
0.74
G=
1.80
0=
1.87
3 =
2.30
h2.09
R=
30
R=
2.00
=3.50
-k=
1.62
R=
2.26
R=
1.10
-=
2.81
F'
N=
11 '
rq=
15
10
=22
lq=
13
N=
19
N=
10
N=
15
-1.51
0=
1.94
1.70
0 =
2.15
a=
1.61
0.=
1.65
0=
1.45
0=
1.93
R=
'
1.13
X2.69
R=
0 97
-=
2'39
x=
0.80
ii=
1 . 77
x=
1
0.81
=1. 38
T0TAL
N=
63
95
N=
58
N= 130
, N= 60
N= 66
N.=
61
N=-105
a=
1 .16
1 64
=1 .23
o=
1 .73
c=
0 .97
0 =
1 .40
1=
1 .27
G=
1.43
2
0
Treatment Mean
NOrmatiVe mean
2.68 2.20 2.39 2.20 1.77_2.50 1.37 1,50TITHV-Dic TV-HV TV only -Control62 mos. 62 mos. 65 rrupp.' '59 mos.
Figure 16,5
AdjuL§ ed Pos -test Group Score_ and Norma_iveMeans for Cpmporah1,0 Ag,HGroups--
.-Frostig,Subt st:5:
Source
Trt.SexTrt. byCovs.Cov. 1Cov. 2Error.
(18)
Table 16.12
nalysis of Covariance Table for Frostig Subtest
Sex
d.f. Mean Square
.05 3 22 08 7.38
.00 1 1.29 0.43
.00 3 0.66 0.222 933.81 312.261 0.70 0.241 89.39 29.89
.94 392 2.99
.005
Frostig Total Raw Sce e
As a total instrument, the Frostig measures overall perceptual levelas well as the ability to recall and transform visual configurations. Italso gives an indication of the dhild's motor development as reflected byhis hand-e_e coordination.
Total Frostig raw scores are presented in Table 16.13 for each age-by-sex-cell within the four treatments. Overall raw scores also,are presented,collapsing the individual cells for age and sex groups.
Figure 16.6 illustrates adjusted mean totals by group for 'all five sub-tests along with a normative total mean score derived by summing the normative'means for all subtests.
40
30
20
10
1111111 Treatment Mean
Normative Mean
.25 31.13 31.81 31.13 29.76 34.62 3.73 28.18TV-HV-MC TV-HV TV only Control62 mos. 62 mos. 65 mos. 59 mos.
Figure 16 .6
Adjusted Post-test Group ScOres and NormativeMeans for Comparable Age Groups-.-
Frostig Total
(19)
Table 16.13
Ftostig Total Pre and Post-test Mean Raw Scores,
Standard Deviations, and Numbers of Subjects
by Age and Sex within Treatmemt Groups
Age
. ex
TV -HV -MC
TV-HV
TV only
Cont
ol
,
,
pretest
H
x=
'5.91.
=10
=5.04
Post-test
k=
14.92
N=
12 .6.161
.
Pretes
,
R=
12.33
1
N=
11
0=
.7.96
Post-test
k=
19.31
N=
18
0 =
8.15
Pretest
i=
8.27
N=
11 7.82
1
Post-test
1
R=
9.88
N=
8
=5.25
Pretest
i=
7.88
N=
8
a=
4.46
Pos -test
k=
14.24
N=
17
0=
5,67
F
1
R=
12.18-
N= .
11
.
,.
=8.73
=
1
N=
-
=
23.38
13 9.83
=7.70
N=
10
-
0 =
6. 7
x=
15.44
N=
18
a=
8.64
R=
9.60
N=
5
0=
5.95
N=
a=
11.40
1
5 9,22
=
N=
0=
8.00
8 4.36
k=
14.53
N=
19
0=
9.22
NI
' s
-20.88
=9 ,
0 =
9.25
N=20
0
31.85
,
10:46
= -
.19.78
=9
11 =
11.66
R=
27.80 .
N=
25
0 =
11.82
R=
21.00
1N=
13
0 =
14.14
x =
N=
0 =
34.93
14
11.65
R=
N=
0=
14.85
11
13 9.14
x=
20.05
N=
17
0 =
1.016
x= 18,25
=-
9
a= 13.60
-= N19
0= 7
34.68
.
13,12
R=
28.30
,N=
10
a=
9.94
R=
35.04
I
N=
22
5=
11.71
"
-16.10
N=10
a=
8.94
R=
N=
0=
25.8
8
16.37
N=
0=
15.38
13 9.60
x=
23.80
N=
17
0=
10.09
,
=32,31
,
=13
=,
11.18
= ,
N=16
a=
47.38
1
. .
11.36
k=
44.63
.. =
81
C =
11.75
R=
42.15
N=
27
0 =
8.11
=25.38
N=
8
0=
14.14
=
N=
0=
34.67
12
13.43
=
N=
0 =
23.44
9
12.87
R=
33.17'
N=
18
0 -
10.82
1
='
31.18
p ti= J1 -
-
i:
'
0= 12,89
x =
N= .
..
.
- 0=
.
41.47
15 -8.85
R=
32.10
: N=
10
'
0=
14.30
.
=43.41
N=
22
0-
10.91
R=
29.92
=13
I
cY =
14.25
k=
N=
10=
37.74
19
11.22
-=
N=
0=
29.60
10
11.86
R=
36.00
'=
15
G=
11.84
-
,
- TOTAL
1
201.77
1
N=
63
=10.16
x=
N=
' =
33.25
'
95 10,47
x- .23.31
N=
58
. 0=
10.61
R=
31.82
N= 130
0=
9.78
k=
19.42
N=
60
1=
11.89
R=
N=
a=
29,77
66
1
11.
84
R=
N= 'i=
16.84
61 9.49
i=
23.76
N= 105
:3 =
8. 8
8
The differences between treatment and sex means are significant at the.01 level as shown in the ANCOVA simmnary in Table 16;14. Differences betweenpairs of means which contributed to the overall treatment effect were sub-jected to a Dunnett's post-hoc analysis. All three treatment groups weresignificantly higher than the control group, while no meaningful differenceswere found between any of the treatment groups, a pattern which was evidentin the previous year's data. This finding points to the overall effective-ness of the television program in promoting -visual-motor development.
Table 16 14
Analysis of Covariance Table for Frostig Total Score
Source n2 d.f. Mean Square F
Trt. .05 3 874.73 7.06 p < .01
Sex .01 1 655.89 5.29 p < .01
Trt. by Sex .00 3 3.43 0.03Covs. 2 181452.77 1463.73 --Cov. 1 1 587.72 4.74Coy. 2 1 9313.36 75.13Error .94 392 123.97
The significant sex effect for the Frostig total score favored thefemales. Table 16.15 shows scores for both sexes derived from the ANCOVAadjusted means.
Table 16.15
Adjusted Post-test Total Frostig Me Scores for Males and Females
Although the differences are small, girls consistently outscore theirmale counterparts in terms of overall perceptualmotor development. Thispattern is not modified by the program treatment effects.
Summary and Implications
The following tabulation lists the probable effects of various programcomponents on the areas measured by the Frostig:
Mobile facility - no effects.
Paraprofessional - significant contribution in the area of same-different discrimination in terms of spatial rotation.
Television program - major effect on eye-motor 'oordination, shapecc:instancy, and the ability to conserve patterns after spatial rota-tion. Additionally, the television .program apparently caused asignificant contribUtion to the total raw score.
Mhere sex differences existed, they constantly favored females overmales, eVen when treatment effects also were present. This pattern indicatesan overall developmental difference between males and females in terms ofperceptual-motor development as measured by this test.
In summary, the television program.seemed to have a broad effect onchildren's perceptual motor development, and this:effect probably involved'perceptual learning from viewing the program and motor learning from activeinvolvement in the drawing, cutting and other manual:taSks taughton Aroundthe Bend.
Home visitors evidently made a significant contribution in the area ofdiscrimination of spatial rotation, which may well have been related to theiremphasis on same-different relations.
Ranking the treatment means across subtests results in the followingtable:
Frostig 1Frostig 2Frostig 3FrostigFrostig 5Frostig Total
Tv-Hy-mc TV-HV Jv on y Control
2 4
This illustrates the dichotomy between the scores of the children whowere visited by the paraprofessional and those who were not. It clearlyshows the impact of the home visitor on the developMent of child's visual-
-.
motor skills.
(22)
Re exenc es
Frostig, Marianne, Adminigt rat ion And S do±ing Nanual, FroStig 'DéVelop-mental Test of .VigUal Porte Consulting Psydhologists Press, 1966).
Hooper, Frank H. and William H. Marshall, The'Initial 'Phase Of'a"PreschoolCurriculum Development Projett - Final _poport Charleston, W.Va.: Appa-lachia Educational Laboratory, Inc. , 1968).