AN EVALUATION arms Us: OF‘

. V coma cuuas To Focus mmmw 7 V
m unscmm-mmauuog ~.
PA'RED’ASSOG'ATE LEARNING - I ” i s - §:; 9’"

| 'TheSis for the Degree ;of_ ,Ph. D; ,
MICHIGAN STATE umvaasm '
' RECHARD L. ALUNGTON
_1373

 

 

«a
. ‘4‘”,

LIBRAR Y

University

This is to certify that the -'-' r.

thesis entitled

An Evaluation of the Use of Color Cues
to Focus Attention in Discrimination and
Paired-Associate Learning

presented by
Richard L. Allington I

,w-n

has been accepted towards fulfillment
of the requirements for

Ph~¢degree in M

Date April 26I 1973

0-7639

 

 

   
  

MlCIll'. -
ngtate ‘

 

 

 

     
     

. 22.. a; ..

, . .u . v
.. ,,. ..Ts..l .f!
.q i . _q ‘Lfivzfi .
.. . _ . .. e _ .
.._ . . .9 . 1..
I... \ Vain 1 .

 

  
  
   
  
 
  
  
 

This :Lunj ~.:

' who IQCEi‘rl one
X

‘:H- techs. 1L4 n
.ubjccro rrnm

public och via

‘udy ‘35

1‘0 one of three t: ca'men: groups.

‘5

‘1nctruhticnal pr0"

the kindergarten pbyulailun. of
7:339 subjncts were randoaly

Pour lotto

. _, " 7' ..£‘S
r v Lfcred xjcfxflvhntly on thse!

:ondu zed uuan 352 randomly

m

r-lLke

'h ll‘ their let: to tight or up to down transformar

 
  
 
 
 
 
 
 
 

" -~ .gugocted'fo; use La m: at«by

(1) no color added tn

”80 mounted to the subunits La 09‘ of three
latter-lit. licence,

0910: added to the letter-like (Scum, ml: k
_Oqa08 added can then veltohnl (to: cl. ‘

mmmmmutnyu .
toann1.1lo verbal fist-lialio at ...,'3%§é§{f5

. 4 ~ ‘
‘ . 1 i. l
. , 1.. y . P
l V .- . \n‘; -
‘ .t ‘ 1 - ‘-_ t x.
4 -'. ‘ \l "‘_ o ‘
u- n! ,7 L‘, r s
-‘ 1 A I '

race. eight

 

 

Ma...

ABSTRACT

AN EVALUATION OF THE USE OF COLOR CUES TO
FOCUS ATTENTION IN DISCRIMINATION AND
PAIRED-ASSOCIATE LEARNING
BY

Richard L. Allington

This study was designed to determine whether sub-
jects who received one of three treatments of color cues
in an instructional program differed significantly on three
learning tasks. The study was conducted using 102 randomly
selected subjects from the kindergarten populations of two
Michigan public schools. These subjects were randomly
assigned to one of three treatment groups. Four letter-like
figures and their left to right or up to down transforma-
tions were selected for use in the study. These eight
figures were presented to the subjects in one of three
treatments: (1) no color added to letter-like figures,
(2) maximum color added to the letter-like figures, and
(3) maximum color added and then vanished from the
letter-like figures.

A program sequence was developed employing an

“audio-f1ashcard reader. The verbal directions of the

 

 

‘V

 

Richard L. Allington

program sequence remained identical for all subjects while
the treatment of color cues varied.

Subjects were pre—tested on their ability to match
the figures to form and to match the figures from memory.
Upon completion of the program sequence, subjects were
post-tested on their ability to match the figures to form,
to match the figures from memory, and to associate a low
meaningfulness C-V-C trigram with each figure.

Significantly better (p < .01 to .05) achievement
on all three tasks was found for the vanished color
treatment over the no-color treatment. From these results
it was concluded that: (a) the vanished color treatment
enhances the learning of visual discrimination, visual
memory, and paired—associate tasks when compared to the
no-color treatment, (b) this enhancing effect seems to
stem from improved attention to the distinctive feature
of a stimulus, and (c) instructional strategies and mater-
ials used in teaching these three basic skills should be
designed to focus attention on salient features of the

stimuli.

 

     
  

AN EVALUATION OF THE USE OF COLOR CUES TO
FOCUS ATTENTION IN DISCRIMINATION AND

PAI RED-ASSOCIATE LEARNING
BY

.J
Richard LUiAllington

A THESIS

Submitted to .
Michigan State University 7 ' ;” 4
; in partial fulfillment of the requirements *r'r_. a
y' for the degree of ‘ i *

DOCTOR OF PHILOSOPHY

Department of Elementary
and Special Education

1973

 

 

  
  
  
  
 
 
  
  
  

l - the st nu. - :.»-: ;. . . .. .y.~ an
“. . I a w}
" “it assistant , . » , . 1 ~ ;,._~ was
”trimmers to .v ‘
'7 'y' Dedicated to:
‘ “ml to emote-.2.- "Jun
"7 . , ~ George C. Allington
"l _: tour. Its-.6;
{ti-qt 1 for his example as a father
itt, oat-Let... and “r .201 ,. s ..- ., _.
1 “it tint, air/jar ~—;-, and 1...». - to rm: itwf ’ ..
affiltion to ‘he pr"‘:"‘a1 A”. has .4- an thi: sun .- 3
\‘ ‘
. ‘Wundinq of the banning of srtv'rlsr;
to Dr. lilliam burr, committee Mm ’: . 2w

. We and lupport;

w 6:. llama Bell and Or. if)“. Multitirw.
“on, for their widest” seas “pg—n2

h It Carl. mm... principle .7! ”‘0‘! ”it
.ud Bx. dong-h m straw oi

 

 

 

 

 

ACKNOWLEDGMENTS

At the time he chooses to pursue a doctoral pro—
gram, the student selects advisors to help in his work.
Their assistance, planning, and encouragement are large
contributors to the success he achieves. The author
wishes to express his appreciation to these people:

to Dr. Gerald Duffy, chairman of the guidance
committee, and Mr. George Sherman, who gave more than
their time, guidance, and support to this study; their
dedication to the profession has given this author an
understanding of the meaning of scholar;

to Dr. William Durr, committee member, for his
suggestions to improve the design of the study and for
his guidance and support;

to Dr. Norman Bell and Dr. Keith Anderson,
committee members, for their guidance and support;

to Mr. Carl Neilsen, principal of the Beach
Elementary School, and Dr. Joseph Kulaga, principal of
the Cody Community School for the opportunity to conduct

this study;

 

     

In
to the kindergarten teachers and pupils who
.wf5fieed their full cooperation: and

to my wife, for her continual encouragement

‘V
, .
’.,- '.- ’
] .
._ :‘ '\
I ._ s
,7. . .~ .
1-. , ,
. .‘.V
. I‘LH a.
‘rx .. , .-
._':nv.> “L‘J'JLC...
Iii 'v.

— —n C' '

 

TABLE OF CONTENTS

Page

      
 

LIST OF TABLES . . . . . . . . . . . . . vii
LIST or FIGURES . . . . . . . . . . . . viii
Chapter

'I.‘ INTRODUCTION . . . . . . . . . . . 1

The Problem . . . . . . . . . . 2
Significance of the Study . . . . . . 10
Definition of Terms . . . . . . 11
Assumptions and Limitations . . . . . 12
Design of Study . . . . . . . . . 13
Hypotheses . . . . . . . . . 14
Organization of the Remainder

of the Dissertation . . . . . . . 15

If} REVIEW OF THE LITERATURE . . . . . . . 17

Color as a Contextual Stimulus . . . . 17
Color as an Attentional Factor . . . . 33
Other Applications of Color .‘ . . . . 31

“ , III. THE EXPERIMENT . . . . . . . . . . 44

Design of the Experiment . . . . . . 44
Materials and Procedures . . . .‘ . . 45
The Program Sequence . . . . . . . . 51
Population of the Study . . . . . . . 59
The Treatment of the Data . . . . . .

 

   

IV.

V.

A.

B.
C.

D.
E.
F.

cc.

Chapter

RESULTS 0 o O I I O I I I I 0

Determining Whether Significant
Differences Exist . . . . .

Identifying Where Significant
Differences Exist . . . . . .

SUMMARY, CONCLUSIONS, IMPLICATIONS
AND RECOMMENDATIONS . . . . . .

Summary . . . .
Conclusions . . .
Implications . .
Recommendations .

BIBLIOGRAPHY . . . . . . . . ; . .
.APPENDICES

Appendix

Method of Color Highlighting for

all Treatments . . . . .
Trigrams and Association Value . . .
Directions for Pre-Tests and

Post-Tests . . .
Introduction to and Verbal Directions
for the Program Sequence . . .

Post-Test Sheets and Pre-Test Sheets .
Group Means of Raw Scores, Raw Scores,
and Within-Group Sample
Correlation Matrix .
Illustration of Audio Flashcard Reader

vi

Page
65

67
71

96
99

100

104
111

113-

119

 

LIST OF TABLES

Table Page
1. Experimental Design . . . . . . . . . 61

2. A Multivariate Analysis of Covariance on
Three Dependent Variables: (l) the
Match to Form Post-Test, (2) the Match
from Memory Post—Test, and (3) the
Naming Task . . . . . . . . . . . 69

 

3. Adjusted Group Means for Each Treatment
Group on the Three Dependent Variables:
’ (1) Match to Form Post-Test, (2) Match
‘ from Memory Post-Test, and
(3) Naming Task . . . . . . . . . 71

L 4. Multiple Correlation Coefficients and

Coefficients of Determination for

Both Covariates with Each Dependent

Variable . . . . . . . . . . . . 72
5. Univariate Analysis of Covariance on the

Dependent Variable Match to Form . . . . 73
{ 6.. Scheffé Post Hoc Comparisons of Treatment

Group Adjusted Means on the Dependent
k Variable Match to Form . . . . . . . 74

7. Univariate Analysis of Covariance on the
Dependent Variable Match from Memory . . 76

8. Scheffé Post Hoc Comparisons of Treatment
Group Adjusted Means on the Dependent
Variable Match from Memory . . . . . . 77

L 9. Univariate Analysis of Covariance on
the Dependent Variable Naming . . . . . 78

 

Scheffé Post Hoc Comparisons of Treatment
Group Adjusted Means on the Dependent
Variable Naming . . . . . . . 79

vii

   

LIST OF FIGURES

Page

  
  
   
   
   
   
   
   
    
   

l. The Four Letter-Like Figures and Their
Transformations Selected for
the Study I I I O I I O I I I O I 4 7

The Letter-Like Stimulus Figures and
Their Paired CVC Trigram . . . . . . . 49

Program Sequence: Order of Teaching
and Testing Trials . . . . . . . . . 54

Styles of Color Highlighting Employed
in the Pilot Study . . . . . . . . . 56

Style of Color Highlighting Used for
Each Treatment Group . . . . . . . . 56

Subjects Participating in Study: By
Sex and School Location . . . . . . . 59

Number of Subjects per Cell fer the
Treatment Effect. . . . . . . . . . 63

Number of Subjects per Cell for the
,L; . $011001 Effect I o e I o I o o a a 63

”VSI"Number of Subjects per Cell for the
> Treatment x School Interaction Effect . . 63

viii

 

 

 

 

CHAP TE R I
INTRODUCTION

Beginning readers must master three basic tasks in-
volving printed letters. The first is that the learner must
visually discriminate between the letters of the ortho-
graphy. A second task is remembering these letters and
those features which make each unique. The final and most
complex task is that of associating the appropriate verbal
response, such as a name or a phoneme, with each symbol.1

Many studies have reported that young children have
considerable difficulty mastering these tasks with letters

‘which are reversals or up-down rotations of each other.2

 

1Jay Samuels, "Formal Intralist Similarity and the
vonRestorff Effect," Journal of Education Psychology, LIX
(1968), 432—37.

 

2John Blair and David Ryckman, "Visual Discrimina-
tion: Lower Case Letter Confusion" (unpublished paper
from the Center for Research on Language and Language
Behavior, University of Michigan); Helen Popp, "Visual
Discrimination of Alphabet Letters," The ReadingiTeacher,
XVII (1968), 221-26; Joanna Williams, "Reactions to Modes
of Word Recognition," Theoretical Models and Processes of
Reading, ed. by H. Singer and R. Ruddell (NewarE, Del.:
International Reading Association, 1970), pp. 38-46; Helen
Davidson, "A Study of the Confusing Letters b, d, p, and

 

 

 

 

 

 

While it is valuable to have identified which upper and
lower case letters present the most difficulty in learning,
it would seem of equal value to develop a teaching tech-
nique which reduces the difficulty of learning these

letters.

THE PROBLEM

Statement of the problem. This study will eval-

 

uate an instructional strategy for teaching children to
discriminate and name letter—like figures and the reversal
or up-down rotation of these figures. This study will
first attempt to determine what effect, if any, adding
color highlights to selected features of the figures has
on the performance of young children in three separate
learning tasks; a paired-associate task, a visual memory
task, and a visual discrimination task. The study will
also attempt to determine whether a gradual vanishing of
these added color cues significantly improves student
performance when compared to students receiving continued
maximum highlights and students receiving no highlighting
on any of the learning trials. Three independent experi-
mental conditions will exist; i.e., (1) figures with no

color highlighting in any presentation, (2) figures with

 

q," Journal of Genetic Psychology, XLVII (1935), 458-68;
and Doreen A330 and Marcia Wyké, "Discrimination of
Spatially Confusable Letters by Young Children," Journal
of Experimental Psychology, XI (1971), 11—20.

 

 

 

f maximum color highlighting throughout all presentations,
and (3) figures with color highlighting gradually
vanished throughout the presentations. The criterion
tasks will present the figures without any color high-

lights for all groups.

Background to the problem. A prime task facing
beginning readers is learning to discriminate and recognize
the symbols of English orthography. While researchers have
identified which letters present the most difficult learn-
ing task for young children, we yet need to know how to
most effectively structure the teaching act in order to
facilitate this learning.

Williams hypothesized that in developing a strategy
for teaching a visual discrimination skill one must con-
sider at least two elements: (1) to what features must
the subject attend in order to solve the task, and (2) how
can we ensure that the subject attends to these features?3
Williams also notes that every letter in English ortho-
graphy differs from every other letter and that these
differences are the distinguishing features.

The use of color to highlight these distinguishing
features of letters may offer assistance in directing the

subjects' attention to the elements necessary to solve the

 

 

3Joanna Williams, "Training Kindergarten Children
to Discriminate Letter—Like Forms," American Educational
Research Journal, VI (1969), 501-13.

 

 

"if

task. Dale, in speaking on the use of color in audio-
visual materials, stated:

. . . color may be used to help us see differences,
to distinguish and emphasize. We may highlight
important parts of a diagram with colored chalk

or trace a route on a map with a line of color.
Color says to the eye: LOOK HEREE4

Otto reviewed five studies employing color and concluded:

. . . the implication of these studies is that . . .
learning should be enhanced by the addition of
color cues for any or all of the following reasons:
aided perception and increased differentiation,

the opportunity for cue selection, and greater
motivation.

Samuels, however, has reacted to the use of

color cues in a different light:

 

One strategy for facilitating word recognition is
to use color cues . . . While this system may increase
rate of initial learning, the critical question is
one of transfer. If the learner focuses his atten-
} tion on color and not letter sha e, what happens
when the color cues are removed?

To answer his own question, Samuels structured a

research project using first graders and college students

 

4Edgar Dale, Audio-Visual Methods in Teachin
(3rd ed.; New York: HoIt, RineHart and WinsEon, I963),
p. 571.

5Wayne Otto, "Color Cues as an Aid to Good and
Poor Readers' Paired-Associate Learning," in Perception

and Readin , ed. by Helen Smith (Newark, Del.: Inter-
nationaI Reading Association, 1967), p. 40.

6S. Jay Samuels, "Modes of Word Recognition,"

in Theoretical Models and Process of Readin , ed. by
Harry Singer and Robert RuddeII (Newark, 5eI.: Inter-
(national Reading Association, 1970), p. 28.

 

 

   

 

-w-m-

 

in a training sequence in which words were presented
either in color or in regular type. He found:

. . . that the rate of learning the words in color
was significantly faster . . . but that on the
transfer tests--when the color cues were removed--
the tables were turned. In comparing recognition
between the words which were always in regular
type to the words . . . in color, recognition

was superior for the words which had always been
in regular type. . . . Apparently, the color cue
was so potent they were unable to focus attention
on the relevant use of letter shape. . . . In
teaching . . . a decision must be made between
speed of initial learning and transfer. The
decision to foster speed . . . at the expense of
transfer may be a false economy.7

Samuels' data provide interesting material for
analysis. One writer interpreted Samuels' data as pro—
viding an empirical base for the following statement:

Efforts to emphasize distinctive features through
the introduction of extraneous cues—-coloring

the diagonal stroke on R, for example, are of
doubtful merit.8

However, Samuels did not employ color to emphasize
distinctive features; rather he printed the total stimulus
in color, thereby emphasizing the stimulus complex. His
subjects did not have their attention focused on a singular
distinctive feature. His data do provide cautions for

the effective use of color. That is, color as applied in

the above studies is not necessarily a useful additional

 

71bid., p. 29.

8J. K. Hemphill, Teachin Readin as Decoding;
Minicourse 18 (Berkeley, a 1 .: Far est Laboratory for
EducationaI Research and Development, 1971), p. 11.

 

1

Alilllll

 

—-——--

 

 

cue and color applied inappropriately may be detrimental
to learning.

What, then, is the role of color cues in teaching
materials or instructional strategies? What cautions must
be observed when employing color? Some answers have been
provided by a number of researchers in the audio-visual
field who offer guidelines for situations in which the
use of color might be helpful. Hoban and Van Ormer con-
cluded that color must not be such a potent cue that it
draws the learner's attention away from important cues.9
McGeoch and Irion found that color can be useful in
assisting in the discrimination between relevant cues.lo
Miller speculated that color cues would be advantageous
if color is one of the most relevant cues or if it can be
used to emphasize relevant cues; it would not be advan-
tageous if it distracts or complicates the subject.11

Black offers a similar finding in the conclusion of a

study on relevant and irrelevant pictorial color cues:

 

9C. F. Hoban and E. B. Van Ormer, Instructional
Film Research 1918-1950 (Port Washington, Long IsIand,
N.Y.: Special Devices Center, 1950).

 

10J. A. McGeoch and A. L. Irion, The Ps cholo
of Human Learning (New York: Longman's Green, I952).

11N. E. Miller et al., "Graphic Communication and
the Crisis in Education," Audio-Visual Communication
Review, V (1957), 3—39.

 

   

 

 

Color cues must be relevant, or help differentiate
relevant cues, or must emphasize relevant cues

but certainly not draw the learner's attention
away from important cues.

In a discussion of the use of cues in general,
Anderson offered the following cautions:

It would seem that the prompting stimulus must

be of such a nature and related to the discrimina—
tive stimulus in such a way that the student
cannot help noticing the discriminative stimulus
during ‘training. Otherwise, when the prompting
stimulus has been withdrawn, it may be found that
the student can no longer respond correctly.

. . . When the prompting stimulus is of such a
nature that it is possible for the subject to
ignore the discriminative stimulus during training,
transfer of control to the discriminative stimulus
may never take place . . .13

As Travers points out, we can currently only
speculate as to the effectiveness this type of training
in visual discrimination, since researchers have

. . . found no adequate analysis of functions which
color might perform in the transmission of infor—
mation. Research tends to make broad comparisons
of color vs. black and white, but no research was
located on the use of color for emphasizing the
crucial aspects of a visual presentation. Another
important use of color of which no mention can be

 

12
Harvey Black, "Relevant and Irrelevant Pictorial

Color Cues in Discrimination Learning,” from Final
nggit: HEW Project 1170, Indiana University (I967),

13
R. C. Anderson, "Educational Psychology," in

Annual Review of P3 cholo , ed. by Paul Farnsworth et
aI. (Pan AIto, CaIiI.: Annual Reviews, Inc., 1967TT

V31. 18, pp. 103- 64.

..—-—_—~— m

 

found in research literature is the separation
of one part of a visual display from another.14

Otto and Askov summarized the use of color in
instructional materials in three points: (1) on the
basis of existing research it is not now possible to
prescribe use of color cues in instruction; (2) color
is presently used not as an additional cue to enhance
learning but as a vehicle for carrying basic information;
(3) the cue value of color appears to be nebulous, being
dependent on the availability of other, more potent
cues.15 Finally, Lumsdaine concluded that:

. . . no really definite studies have been made
on specific ways in which color may contribute
to learning from instructional media.16

Past research provides us with some guidelines
for employing color. Color must be relevant, color must
not draw the learners' attention away from relevant
information and, if the color cues are not present on

the transfer task, the learning environment must be

 

l4Robert Travers, "Research and Theory Related
to Audio-Visual Information Transmission" (unpublished
paper from the Bureau of Educational Research, Univer-
sity of Utah, July, 1964), p. 4.

15Wayne Otto and Eunice Askov, "The Role of Color
in Learning and Instruction" (Theoretical paper #12 from
the Research and Development Center for Cognitive Learn-
ing, University of Wisconsin, March, 1968).

16Arthur Lumsdaine, "Instruments and Media of
Instruction," in Handbook of Research on Teachin , ed.
by N. L. Gage (Ch1cago: Rand McNaIIy, I963), p. 635.

on».

[lilll
.

 

 

 

 

structured to include a procedure which will eliminate
color cues prior to transfer. While research has pro-
vided minimal guidelines for employing color cues, there
have been no studies which provide a solid empirical base
for decisions on the addition of color cues in discrimina-
tion learning.

However, Gibson has summarized research which
demonstrates that children learn to search more efficiently
for the distinguishing features of letters when verbal
instructions were used to focus their attention.17 Haring
has described the role of the teacher as, "ensuring that
relevant dimensions of stimuli gain control over the
child‘s responding."18

While research is inconclusive regarding the role
of color cues in discrimination learning, the importance
of attending to distinguishing features has been stressed.
It appears no research has been conducted which clearly
provides a solution to the second element of the question
Williams posed: How can we ensure that the subject
attends to appropriate features during training? Or more

specifically: Can we effectively use color cues to aid

 

17Eleanor Gibson et al., "A Developmental Study
of the Discrimination of Letter-Like Forms," Journal of
Comparative and Physiological Psychology, LV ,

 

18Norris Haring, Attendin and Responding (San
Rafael, Calif.: Dimension Publis 1ng Company,’1968),
p. 2.

 

 

 

10

subjects in their search for the distinguishing features
of letters? This study, then, will attempt to provide

an answer to that question.

SIGNIFICANCE OF THE STUDY

Many research studies have been conducted in an
attempt to discover how learning is accomplished and how
to facilitate the learning process. Research has also
been conducted on the possible application of color to
learning situations, but while research is available on
the role of color as an aid to learning, it is fragmentary
and research on vanishing color cues is minimal and con-
tradictory. This study is intended to provide data re-
garding the effect of using a single hue vanishing color
cue to emphasize or highlight distinguishing features of
letter—like forms in an attempt to facilitate visual dis-
crimination and recognition of these forms. These data
will provide educators with empirical evidence upon which
to base objective decisions regarding the use of color
in discrimination learning and paired-associate tasks.
The study should also provide insights for further
empirical investigations into the applicability of color

to other instructional techniques.

 

 

 

 

11
DEFINITION OF TERMS

Discrimination: In the presence of two figures,
the subject responds appropriately as to whether the

presented figures are the same or different.

Highlighting: A form of a prompt. Specifically,
the application of color to a distinctive feature of a

stimulus, in an attempt to focus visual attention.

Matching from memory: Following a presentation

 

and removal of a stimulus figure the subject selects

that figure from a group of four figures.

Match to form: In the presence of a stimulus
figure the subject selects that figure from a group of

four figures.

Naming: In the presence of a figure, the subject

makes the appropriate verbal response.

Prompt: A stimulus added to the terminal

stimulus to make a correct response more likely.

Stimulus: In this study, stimulus refers to the
letter-like figures; the terminal stimuli are the letter—
like figures without color highlights; the prompting

stimuli are the color highlights added to the figures.

 

12

 

Vanishing: A method of removing a visual prompt.
Specifically, removing the color highlighting by gradually

reducing the area of color until no color remains.*

ASSUMPT IONS AND LI MI TAT IONS

The following assumptions underlie this study: it
is assumed that
1. it is important for the subject to be able to

match to form;

2. it is important for the subject to be able to

match from memory;

3. it is important for the subject to be able to

name the figures;

4. naming the figures is equivalent to naming or

sounding the letters of the alphabet;

5. the color employed can be generalized to other

colors; and

\

6. learning to discriminate and recognize the
letter-like forms is an equivalent task to
learning to discriminate and recognize letters of

English orthography.

 

 

*Vanishing is a form of the psychological tech-
nique of fading. However, while fading is the reduction
of the intensity or saturation of a color within a desig—
nated area, vanishing is a reduction of the area of color.

 

 

_ W’s.

 

.Amlah;-\

13

The following limitations underlie this study:
findings of this study will be limited to

l. the tasks under investigation, or similar tasks;
2. this population or similar populations;

3. situations in which the same or similar materials

are employed; and

4. situations in which the same or similar learning

environments are employed.

DESIGN OF STUDY

The population of this study consisted of all
kindergarten children enrolled at Beach Elementary School,
Cedar Springs, Michigan; and all the kindergarten children
enrolled at Cody Community School in Flint, Michigan. The
schools in these communities were selected to provide sub-
jects from varied geographic, social, economic and cul—
tural backgrounds.

A total of 102 children participated in the
training program, 51 children from each elementary school.
The children were randomly selected and assigned to one
of three treatment groups. Pretests were administered
for two tasks; visual discrimination (match to form) and
visual memory (match to memory). A training program was

developed in which only the visual stimuli varies in

 

 

 

 

 

14

three treatments: no color highlights added, maximum
color highlights added and continued until transfer, and
maximum color highlights added and vanished prior to
transfer. Each child, upon completion of the training
program, was post-tested for three tasks: naming the
figures, visual discrimination, and visual memory.

The results for all treatment groups were
tabulated and compared to determine whether the null

hypotheses were to be rejected.

HYPOTHESES

The hypotheses, stated as null hypotheses, are

as follows:

Hypothesis 1a

Given the task of Matching to Form, there will be
no significant difference (p < .05) in achievement
between groups upon completion of the program
sequence.

Hypothesis lb

Given the task of Matching to Form, there will be no
significant difference (p < .05) in achievement,
upon completion of the program sequence, between

the samples drawn from the two school populations.

Hypothesis 1c

There will be no significant interaction effect
(p < .05) between treatment and school as measured
by achievement on the Match to Form task.

 

 

 

 

15

Hypothesis 2a

Given the task of Matching from Memory, there will

be no significant difference (p < .05) in achievement
between groups upon completion of the program
sequence.

Hypothesis 2b

Given the task of Matching from Memory, there will

be no significant difference (p < .05) in achievement,
upon completion of the program sequence, between the
samples drawn from the two school populations.

Hypothesis 2c

There will be no significant interaction effect
(p < .05) between treatment and school as measured
by the Match from Memory task.

Hypothesis 3a

Given the task of Naming, there will be no signi-
ficant differences (p < .05) in achievement between
groups upon completion of the program sequence.

Hypothesis 3b

Given the task of Naming, there will be no signi-
ficant difference (p < .05) in achievement, upon
completion of the program sequence, between the
samples drawn from the two school populations.

Hyppthesis 3c

There will be no significant interaction effect
(p < .05) between treatment and school as measured
by the Naming task.

ORGANIZATION OF THE REMAINDER OF THE DISSERTATION

Chapter II is a presentation of the review of

pertinent literature. Particular emphasis is on the

 

16

   
  
    
   
   

‘I lee of color in learning and specifically in discrimina-
,93 tion and paired-associate learning.

‘1 I Chapter III contains a description of all mater-
':"ials and procedures employed in this study. The design of
the study is also presented in detail.

Chapter IV includes presentation of the data
collected, treated and analyzed for the study.

Chapter V provides a summary of the study and
appropriate conclusions. Implications of this study and

suggestions for future research are also included.

'44
.,‘
.

 

 

 

 

CHAPTER II

REVIEW OF THE LITERATURE

Introduction. The purpose of this chapter is to
review existing research associated with the function
of color in learning and to examine the rationale for
the present application of color in this experimental
study.

Relevant research on the function of color in
learning has roughly been divided into two areas; color
as a contextual stimulus to which the response may
become indirectly associated, and color as an atten-
tional cue which brings the visual senses into contact
with distinguishing features of the stimulus object not
previously readily accessible, or of intensifying the
stimulation from these features. The relevant studies

are reviewed below.

COLOR AS A CONTEXTUAL STIMULUS
The most common use of color has been as a con-
textual cue and studies of this use have generally incor-

porated a paired-associate learning model. Color applied

17

 

 

18

as a contextual cue aids the learner in differentiating
visual stimuli; that is, when separate colors are paired
with each visual stimulus, the learner is provided with
an added dimension on which to base his discrimination.
In an early study, Dulsky presented his subjects
with pairs of nonsense words presented on two styles of
backgrounds.19 In one style, labeled homogeneous, he pre—
sented the nonsense words either with a different color
background for each pair or the same color background for
each pair. In the second style, labeled heterogeneous,
he presented the stimulus half of each card in color and
the response half in gray and vice-versa. After learning
, the nonsense words, subjects were tested for recall under
three conditions: repetition of the learning presenta-
tion, shifting background color, and changing to words
presented on gray backgrounds. Results of this research
demonstrated that recall was significantly better when
backgrounds on the test trials remained constant with
backgrounds presented on the learning trials, regardless
of whether stimulus backgrounds in learning were colored
or gray. Learning was severely inhibited when the response
background colors were changed. The subjects used the

color background as an additional cue while learning the

 

198. G. Dulsky, "The Effect of a Change of Back-
ground on Recall and Relearning," Journal of Experimental
Psychology, XVIII (1935), 725-40.

 

 

19

lists. Responses were associated not only with the
printed primary stimulus but also with the contextual
stimulus, the background color of the card on which the
nonsense word was printed.

Dulsky's results seem to support the theory
posited by Guthrie and Smith20 and by Hull21 that re-
sponses can become associated not only with the primary
stimuli but also with context stimuli. The nonsense words
are primary stimuli and background colors contextual
stimuli.

Weiss and Margolius attempted to extend the
findings of Dulsky.

The question now raised is: regardless of the
condition under which retention will occur, might
there be any advantage for learning if certain
context stimuli are presented throughout the
lesson?22 '
To answer this question, they presented subjects with
pairs of nonsense trigrams and simple word responses on
varied color cards. The achievement on the task of

associating the correct responses with the appropriate

trigrams in terms of mean retention scores was as follows;

 

20E. R. Guthrie and S. Smith, General Psychology

(New York: Appleton-Century-Crofts, 192IT.

 

21C. L. Hull, Principles of Behavior (New York:
Appleton-Century-Crofts, 1943).

 

22Walter Weiss and Garry Margolius, "The Effect

of Context Stimuli on Learning and Retention," Journal
of Experimental Psychology, XLVIII, No. 5 (1954), p. 321.

65:-

 

 

20

the highest scores were achieved by the group which had
no change in stimuli and no change in background.
Followed by the group which had a change in stimuli but
no change in background, then the group with change in
stimuli and change in background color. The lowest scores
were recorded by the group which had changes in both
stimuli and background color. Commenting on their find—
ings, Weiss and Margolius report that,

The research confirms the theoretical position that

the representation of context stimuli, which has

been present consistently while learning was occur-

ring, would aid retention of learned responses.

Thus, the findings of Dulsky are confirmed. . . .

a possible explanation [of the findings] presumes

that distinctly discernible context stimuli may

act to reduce intralist generalization [similarity]

and thereby facilitate learning.23

In an attempt to determine whether subjects are

cognizant of added color cues in similar paired-associate
tasks, Birnbaum presented subjects with ten consonant—
vowel—consonant (CVC) trigrams paired with nonsense
syllables, each trigram outlined with a distinctive
color during the study trials. The criterion tasks
indicated that association between color and primary

stimuli had taken place. "S's score significantly better

than chance when required to match stimulus words with

 

23Ibid., p. 321.

 

 

21

the colors which surrounded the words on study trials."24

After intervening tasks, Birnbaum presented half of the
subjects with the same paired-associate task and half with
the same stimulus-response pairs but different color out-
lines. The groups with the same stimulus—response pairs
but different color outlines performed at a lower level
on this post-test than the same paired—associate group
even though the stimulus—response sets remained the same.
The color cues then became a function of the stimulus.
Responses were associated with the colors as well as the
printed stimulus, though to a lesser degree. The color
cues were potent enough to produce interference when the
color was changed and the printed stimulus remained the
same.

Given that color facilitated learning, Saltz
attempted to determine if this facilitative effect
differed under varied conditions.25 To accomplish this,
subjects were presented with stimuli consisting of tri-
grams paired with nonsense syllables. The stimuli was
varied according to one of four conditions: (1) color on

learning trials and color on test trials (C-C), (2) color

 

24Isabel Birnbaum, "Context Stimuli in Verbal
Learning and the Persistence of Associative Factors,"
Journal of Eiperimental Psychology, LXXI, No. 4 (1966),
p. 485.

25Eli Saltz, "Compound Stimuli in Verbal Learning,
Cognitive, and Sensory Differentiation Versus Stimulus
Selection," Journal of Experimental Psychology, LXVI
(1963). 1-5.

 

 

1‘

 

 

 

‘color would differ for good and poor readers.

22

on learning trials and no color on test trials (C-NC);

(3) no color on learning trials and color on test trials,
(NC-C); and (4) no color on either learning or test

trials (NC-NC). His findings confirmed that color facil-
itates learning, with the C-C condition providing the most
significant results. Both C—NC and NC—C conditions pro-
vided more significant gains than the NC-NC condition.
This study, like those previously cited, demonstrates that
paired-associate learning can be facilitated by the ad-
dition of a color cue to the primary stimuli. Beyond
this, the Saltz study indicates that color can facilitate
learning even when the color cue is not readily accessible
as a cue, as in the NC-C and C-NC conditions.

Other researchers have also found that color as a
contextual cue can facilitate learning. Otto reported on
the first of a series of studies he conducted on the role
of color cues in paired-associate learning. This investi-
gation attempted to determine whether the usefulness of
26» Subjects
were selected from grades 2, 4, and 6. At each level,
half of the subjects learned a list of five geometric
forms which had been paired with distinct colors and with
a CVC verbal response. The other subjects learned the

same list printed in black. In this study, the additional

 

26Otto, "Color Cues as an Aid."

 

23

color had no significant effect on learning. Otto con-
cluded, however, that two factors may have inhibited the
usefulness of the color: (1) the geometric figures were
dissimilar and previous research has shown that color
seems most useful when intralist similarity is high, and
(2) the subjects at times seemed unaware of the applica-
tion of color cues in the task.

In an attempt to control the two weaknesses cited
in the first study, Otto designed another study, again

7 In this

using 72 subjects from grades 2, 4, and 6.2
study, the subjects learned a list of six three-letter
words written in the Greek alphabet with the Greek letters
serving as the stimuli and common English words serving

as the responses. These were selected to increase intra-
list similarity and to provide a task more similar to a
reading task. The subjects were informed of the role of
the color. Again, half the subjects learned a list with
color and half with plain black stimuli. Otto concluded,
in this revised study, that color had significantly
enhanced learning and that the second graders seemed to
utilize the color cues to a greater extent than the
others. Otto summarized this research as follows: "With

greater stimulus similarity the need for and use of

 

27Wayne Otto, "Elementary Pupils' Use of Cues in
Paired-Associate Learning," Psychology in the Schools, V,

 

24

further stimulus differentiation on outright cue selec—
tion increases . ."28
In a third study, Otto examined the effect of

three types of facilitative cues in paired—associate
learning.29 The relative value of color, order of pre-
sentation, and intralist similarity were examined. Sub-
jects were asked to learn either a high—similarity or
low similarity list of six three-letter words formed with
Greek letters. The three-letter words were printed in
either black or distinctive colors and the lists were
presented in either serial or scrambled order to second
graders. Order of presentation and the type of printed
stimulus were changed on a post-criterion test in order
to measure interaction of the variables. An analysis of
these scores suggested that color cues seem less potent
than serial order, but that there was evidence of color
cue selection. Commenting on the effects reported in
this study, Otto states:

Color cues appear better than no cues at all, and

it would be sensible to provide them when stimuli

are so similar or so unsystematically presented as
to provide little basis for identification . . .

 

28Wayne Otto and Carin Cooper, "Investigations
of the Role of Selected Cues in Children's Paired-
Associate Learning" (Technical report #53, Wisconsin
Research and Development Center for Cognitive Learning,
University of Wisconsin, May, 1968), p. 6.

29Wayne Otto, "Intralist Similarity, Order of
Presentation and Color in Childrens' Paired-Associate
Learning," Psychonomic Science, IX (1967), 531-32.

 

 

25

One important question that needs a more definitive
answer before sound judgments can be made about the
pragmatic value of secondary color cues in instruc—
tion is whether enhanced learning with color
results simply from cue selection or, at least

in part, from improved differentiation of primary
stimuli. Nevertheless, the essential point seems
clear: in every instance the removal of cues
resulted in poorer performance, and this was true
when color was the only cue removed.30

In the studies reported, it has been shown that
color cues can effect rate of learning but that other
factors may determine the extent of facilitative effect
of those color cues.

For instance, Underwood et al. reported that
adult subjects could recall meaningful trigrams which had
been presented on varied color cards even after the color
backgrounds had been removed but the same subjects failed
to recall the list of low-meaningfulness trigrams when
the color cue was removed. They concluded that color
became a functional stimulus when paired with low meaning
trigrams but that when the adult subjects learned mean—
ingful trigrams their past experience led them to respond

to the words rather than the color.31 Here Otto's state-

ment that "color cues seem to be better than no cues at

 

30Otto and Cooper, "Children's Paired—Associate
Learning," p. 10.

313. J. Underwood, M. Ham, and B. Ekstrand, "Cue
Selection in Paired-Associate Learning," Journal of Ex-
perimgptal Psychology, LXIV (1962), 405-9.

 

.-._. .—o

26

all"32 seems to be supported. The unanswered question

that remains is: When are color cues most effective?
While conducting research to support his reading
series, which he calls "a coloured reading" program, Jones
had young subjects (ages 3.5--5.l) match letters of
English orthography and easily confused words under two
conditions: letters and words in color and letters and
words in black. Only three colors were used to high-
light six letters to eliminate matching on the basis of
color alone. Jones concludes, "matching the black
letters and black words was at least three times as
difficult as the same task in color, even allowing for
color matching."33
Newman and Taylor concluded that color cues are
more effective when primary stimulus intralist similarity
is high.34 Their subjects learned either high or low
similarity lists on cards of varied colors and were given
either identical color cues or no color cues on criterion
trials. The subjects who had been presented with high

similarity lists and then had the color cues removed

 

32Otto, "Intralist Similarity," p. 532.

33J. Kenneth Jones, "Colour as an Aid to Visual

Perception in Early Reading," British Journal of Educa-
tional Psychology, XXXV (1965), 26.

 

34S. E. Newman and R. Taylor, "Context Effects in

Paired-Associate Learning as a Function of Element-
Sharing Among Stimulus Terms," Journal of Verbal Learning
and Verbal Behavior, I (1963), 243-49.

27

performed at the lowest level of the four groups. These
studies appear to support Otto's contention that it seems
sensible to provide color cues when intralist similarity
is high.

If color cues do facilitate learning in this con-
dition, then there are tasks which beginning readers face
where color could be applied. In an attempt to employ
color cues in a functional application, Schutz developed
a training program for discriminating the letters p and

.35 However, he reported considerable difficulty in

2
training four—year-olds to discriminate between these
letters. The letters were presented in different colors
and he found his subjects responding to color rather than
the printed figure. Numerous attempts to fade the color
from red to black were largely unsuccessful, as were
attempts to fade from shades of grey to black. Hall and
Caldwell also attempted to teach this p and g discrimina-
tion to young subjects and encountered similar difficulty

in fading colors from the subjects.36 Again, their sub-

jects often responded to color rather than the printed

 

35R. E. Schutz, "Acquisition of Alphabet Letter
Discrimination With and Without 'Errors'“ (paper read
at the American Psychological Association, Los Angeles,
1964).

36Vernon Hall and Edward Caldwell, "Analysis of
Young S's Performance on a Matching Task" (unpublished
research report from Syracuse Center for Research and
Development in Early Childhood Education, Syracuse
University, April, 1970).

 

28

figure and when the two distinctive colors were elim-
inated the subjects had difficulty in responding
correctly.

While attempting to teach children enrolled in a
Head Start program the names of four alphabet letters,
Washington required the subjects to match stimulus
figures. He varied color on the stimulus figure and the
matching figure under the following conditions: (1)
color on stimulus figure and color on matching figure,
(2) color on stimulus figure and no color on matching
figure, (3) no color on stimulus figure and color on
matching figure, and (4) no color on stimulus figure and
no color on matching figure. Groups with color on the
stimulus figure and no color on the matching figure
achieved at a significantly higher level than the other
groups. The results also seem to indicate that subjects
in the color-color groups did not attend to the letter
stimulus but rather matched colors. Therefore, they
could not match the letters in transfer on the basis of
form alone.37

In the studies reported above, the students often
associated their responses with the color cue rather than

the printed stimulus. This points out an important

 

37Ernest David Washington, "Matching and Naming
Letters of the Alphabet with and without Redundant Color
Cues" (unpublished doctoral dissertation, University of

.Illinois, 1968).

 

29

problem with the use of added cues; that is, how do we
force the subject to attend to each of the stimuli, both
color and printed form. If the subject uses only the
printed stimulus, he is not utilizing maximum information
and if he uses only the color he is basing his response on
irrelevant information since that color is not present on
the transfer task. Further, if he bases his response on
the color cue, color becomes the primary stimulus and
attempts to fade that cue are unsuccessful. These results
were predicted by Sundland and Wickens who advised that
precautions must be take; when supplying multiple cues,
for subjects may attend only to the most salient stimulus
in the stimulus complex. They suggest that when one aspect
of the total stimulus becomes more meaningful or more
discriminable, subjects are not likely to associate re-
sponses with the less meaningful or discriminable compo-
nents of the total stimulus.38
In an attempt to clarify the effects of color cues,
Samuels presented 60 first graders with four artificial
alphabet words which were paired with common three-letter
nouns with the subjects assigned to groups in which the
stimulus words were classed as high, medium, or low simi-

larity. Throughout the learning trials, one stimulus in

 

380. M. Sunderland and D. D. Wickens, "Context
Factors in Paired—Associate Learning and Recall," Journal

of Experimental Psychology, LXIII (1962), 302-6.

 

30

each list was printed in red and the others in black.
Recall was measured after five learning presentations,
with each stimulus word printed in black for the recall
test. Even though during the learning trials the subjects
easily identified the high similarity stimulus in red,
during recall there were fewer correct responses for that
stimulus than for any other. In each of the other groups
(medium similarity, low similarity), there was no signi-
ficant difference between responses to red or black
stimuli. Samuels concludes:
. . . that in high similarity list 4-L, during learn-
ing trials, where discrimination on the basis of
letter form was difficult, S's responded to words on
the basis of color. On transfer tests the color cue
was absent, and S's did poorly with the formerly
isolated (red) words. . . . It is conceivable that
the S's did not have sufficient experience with
the transfer task to realize that color was an 39
irrelevant cue and letter shape the primary cue.
Samuels' study was intended to answer the ques-
tion: "If the learner focuses his attention on color and
not letter shape, what happens when the color cues are

removed?"40

In other words, does color in fact inhibit
learning rather than enhance it? In view of the fact
that his first grade subjects might possibly have lacked
sufficient experience in transfer tasks, he replicated

the study using college students. However, on the

 

39$amuels, "Formal Intralist Similarity," p. 435.

4oSamuels, "Modes of Word Recognition," p. 29.

 

31

transfer tasks the results were again similar to the
experiment using young children.

Samuels found that the rate of learning the words in
color was significantly faster than words in regular
type. But on transfer tests—~when color cues were
removed--the subjects had great difficulty in recog-
nizing the words formerly in color. Thus, on transfer
tests the tables were turned. In comparing recog-
nition between the words which always were in regular
type to the words which had formerly been in color,
recognition was always superior for the words which
had always been in regular type. What makes these
results so surprising is that the college students
knew the color cues were to be removed. Apparently,
the color cue was so potent that they were unable

to focus attention on the relevant cue of letter
shape.41

Even though Samuels' research findings seem to
contradict previously cited research, he has contributed
a valuable criteria for employing color in teaching
strategies. As Williams points out:

His warning that children select the easiest cue
for recognition, and that the easiest cue is
sometimes an "incidental detail" is very well
taken. Moreover, one should expect that other,
more relevant cues, will also be picked up by the
child. One of Samuels' own experiments illustrates
this point. Subjects learned to read words printed
in color more easily than words printed in regular
type; however, when the color cues were removed,
they could not recognize the words that had pre—
viously appeared in color. In this case, the color
cue was so salient that it overrode the relevant
cue.

 

411bid., p. 29.

42Joanna Williams, "Reactions to Modes of Word
Recognition," in Theoretical Models and Processes of
Readin , ed. by He en K. Sm1t ewar , e .: In er-
national Reading Association, 1970), p. 38.

 

I.
f

32

In a further study, in which the results were
highly similar to those of Samuels, Isaacs added separate
colors with high and low discriminability shapes and
paired these shapes with a numeral as the verbal re-
sponse. Reinforcing Samuels' findings, Isaacs reported
that

. . . in the present experiment, the addition of
relevant color cues to the shape cues was associated
with a significant reduction in the number of trials
required to reach the training criterion, but was
associated with a relatively greater number of
errors on the shape transfer test.
Here, too, color facilitated the initial rate of learning
but caused undesirable results at transfer.

A major intent of both the Samuels and the Isaacs
studies was to demonstrate that often a decision must be
made between a "criterion of speed of initial learning
and a criterion in terms of the ability to transfer to

"44 Given that color can facil-

new untrained materials.
itate and increase rate of initial learning, can a
strategy be developed which eliminates the unwanted loss
at transfer? In Williams' opinion, the development of an

effective teaching strategy is based on " . . . good

answers to two questions: (1) what cues must the child

 

43Dan Lee Isaacs, "Cue Selection in Paired-
Associate Learning as a Function of Relevance of Color
Cues and Discriminability of Shape Cues" (unpublished
master's thesis, Indiana University, 1966), p. 24.

44Williams, "Reactions to Modes of Word Recog-
nition," p. 39.

33

attend to, in order to solve the criterion task--which we
assume to be a transfer task; and (2) how can we ensure

that he will attend to those cues?"45 In an attempt to answ
answer the second part of Williams' query, a review of

the literature dealing with color as an attentional

factor follows.

COLOR AS AN ATTENTIONAL FACTOR

Research on color as an attentional factor is
limited. There is, however, a body of research on visual
discrimination which is concerned with the search theory,
attention, distinctive features, and applications of color
in unique instructional settings. In the following sec-
tions, a variety of studies in the above areas are re-
viewed in an attempt to clarify the roles of these
factors.

The following studies consider attention as a
necessary requisite to discrimination. The role of
attention can be viewed from two standpoints: (l) the
teacher or experimenter, and (2) the student or subject.
Considering these roles, Haring has said:

Our task as teachers is to insure that relevant
dimensions of stimuli gain control over the child's
responding. When we are successful in guiding,
attending and reSponding to relevant dimensions of
stimuli, appropriate patterns of responding are
acquired. When we fail, we have problems like

confusion over b and d . . . attending labels of
behavior of "looking at" and noticing certain features

 

451bid., p. 39.

34

which come to be discriminative for the person,
thing or condition being attended to.46

The importance of the attention factor in visual dis-
crimination has also been pointed out by Wheelock, who

notes that in order to make a visual discrimination a

subject must attend to the stimuli to be discriminated.47

The fact that teachers have often failed to ade-
quately direct attention or adequately teach students to
direct attention was reported by McAnnich in an investi-
gation of the visual discrimination abilities of good and
poor third grade readers.

Subjects selected as deficient in reading skills
consistently performed less effectively on all
discrimination tasks than did subjects selected as
demonstrating adequate reading skills . . . In
recognizing word or letter groupings in isolation,
the findings suggest that the disabled readers
attend more to the general configuration of the
form, whereas able readers attend to details
within the configuration. . . .48

Anderson and Samuels drew similar conclusions:
Good readers were superior to poor readers in

Visual Recognition Memory. This superiority . . .
was interpreted to stem from the good reader's

 

46Haring, Attending and Responding, p. 2.

47Warren Wheelock, "An Investigation of Visual
Discrimination Training for Beginning Readers," in
Perception and Reading, ed. by Helen Smith (Newark,
Del.: International Reading Association, 1967),
pp. 101-4.

 

48Myrene McAnnich, "Investigation of Recognition
Variance of Perceptual Stimuli Associated with Reading
Proficiency" (paper presented at International Reading
Association, Missouri, 1969), p. 4.

35

ability to consistently encode distinctive
features of the stimuli. Poor readers then . . .
have difficulty in attending to or identifying
the distinctive features of the stimuli . . .4
Williams states that visual discrimination re-
quires that subjects learn to identify the distinguishing
features of a stimulus; those features which make it

different from other stimuli.50

This is in keeping with
a family of theories in discrimination learning in which
the develOpment of the differential response is said to
occur in two stages. The first stage involves learning
to attend to the distinguishing feature of the stimuli
while the second stage is learning to identify and
remember that feature. According to search theory, only
gross features common to the pair become directly avail-
able upon first presentation. The distinctive features
are not readily differentiated from stimulus. The sub-
ject may respond to the gross features or he may search

for a distinctive feature.51

 

49Roger Anderson and S. J. Samuels, "Visual

Recall Memory, Paired-Associate Learning, and Reading
Achievement" (paper read at American Educational Research
Association, Minnesota, March 2, 1970), p. 10.

50Joanna Williams, "Effects of Discrimination and
Reproductive Training on Ability to Discriminate Letter-
Like Forms" (paper presented at American Educational
Research Association meeting, February, 1968).

51H. M. Jenkins and R. S. Sainsbury, "Dis-
crimination Learning with the Distinctive Feature on
Positive and Negative Trials," in ATTENTION: Contemporary
Theory and Analysis, ed. by David Nostofsky INew York:

 

 

36

It does not often seem that the inability to pick
out these distinctive features is a physical disability.
Smith feels that often a child may have adequate visual
acuity

to identify letters as well as an adult . . . [but]
what the child does not know is where to look

for the distinctive IEEtures of Ietters; he knows
how to look, but not what to look for.52

In Stott's opinion, the necessary dimension of knowing
what to look for may be hindered by lack of attention to
the stimulus.

Nearly all failure to match letters is due to
inattention. The child does not give himself
time to get sensory information owing to his
impulsivity, or else he is not motivated enough
to make the effort of attention . . . Once the
child becomes conditioned . . . to make the
discriminations, the apparent (perceptual) hand-
icap is dissipated.53

This inattention itself may be a manifestation of the lack
of sufficient information about where to look, as des-

cribed by Smith.54

 

Appleton-Century-Crofts, 1970), pp. 239-73; Louise Tighe
and Thomas Tighe, "Discrimination Learning: Two Views in
Historical Perspective," Psychological Bulletin, LXVI,
No. S (1966), 353-70.

 

52Frank Smith, Understanding Reading: A Psycho-
linguistic Analysis of Reading and Learning to Read (New
York: Holt, Rinehart and Winston, 1971), p. 224.

 

53D. H. Stott, "Some Less Obvious Cognitive Aspects
of Learning to Read," The Reading Teacher (January,
1973), 375.

 

54Smith, Understanding Reading, p. 224.

 

37

As mentioned previously, it would seem that color
could serve the function of bringing the visual senses
into contact with distinguishing features of stimulus ob-
jects. Color may serve to augment the influx of informa-
tion from the stimulus, and as reported in Samuels' study,
it may unsuitably lead to an exclusion of some features.
However, if color cues could be used to attenuate features
rather than exclude them, color would help focus attention

by highlighting critical features of the stimulus.

OTHER APPLICATIONS OF COLOR

In this section, attention and color again play an
important role in two quite different instructional situa-
tions. While these studies do not fit neatly into either
the contextual or attentional categories presented
earlier, they offer insights into the application of
color cues.

In the first two studies, color is employed as a
literal cue when teaching sight words. Again, problems
deve10p on the transfer task, reinforcing the need to
carefully structure the learning environment when adding
color cues. In the final study, color is employed as a
highlighting device in college textbooks. The attempt is
made to focus students' attention on the information

highlighted.

38

In the first category, Taber and Glaser paired
Specific colors with color words (e.g., the word "red"
was printed in black and had red lines of color paired
with it) in an instructional program.55 In this study,
first grade children were presented the color words which
had been highlighted using lines of the appropriate color
radiating from the word. Throughout a series of trials,
the radiating lines were removed until on the criterion
trial each subject was responding only to the word in
black print. The authors concluded that the program was
effective in teaching these words and that their simple
paradigm was applicable to many instructional situations
in the primary grades.

Duell and Anderson attempted to replicate the
Taber and Glaser results with kindergarteners. They
failed to achieve the same results and commented as follows
on their lack of success:

An important factor in an effective training
sequence using prompts is forcing the subject

to notice the cue while making the response.

An analysis of the observing behavior required

of the subject in order for him to respond cor-
rectly in the Taber and Glaser study indicates

he only had to notice the shapes of the individual
letters or overall form of the printed word in

order to respond correctly. As a consequence,
our children failed to learn to "read" even

 

55Julian Taber and Robert Glaser, "An Explora-
tory Evaluation of a Discriminative Transfer Learning
Proqram using Literal Prompts," The Journal of Educa-
tional Research, LV, No. 9 (June7JuIy, I962), 599-I2.

 

39

though they completed the training sequence with

few errors. Of course, the child could have

studied the printed word as well as notice the

colors and perhaps voluntarily associate his response

with the cue. Second semester first graders do this

but kindergarteners don't. This sequence worked for

Taber and Glaser because their students were more

sophisticated with reSpect to words and letters.56
Again the students seemed unaware of what the task in—
volved, or what to look at. The color did not direct
attention to useful information.

However, in the second category, the use of color
to cue university students to information in a psychology
textbook proved successful in a recent research study.57
The authors presented three groups of students with text-
book material which had principles, explanation of prin-
ciples or trivial statements underlined. A fourth group
had no color added to their texts. Tests given in the
course had specific questions directed at each of the
three types of knowledge underlined in the texts. An
analysis of the test results showed a slight support for
the prediction that examination performance would depend
on type of material underlined. No evidence of carryover

to unhighlighted material was found. This study found a

useful application of color, concluding that students'

 

56Orpha Duell and R. C. Anderson, "A Failure to
Teach a Sight Vocabulary by Vanishing Literal Prompts"
(paper read at the American Educational Research
Association, 1967).

57Kenneth Leicht and Valjean Cashen, "Type of
Highlighted Material and Examination Performance" (paper
read at the American Educational Research Association,
1971).

4o

attention seemed to be focused primarily on learning
underlined material which they perceived as important.
The color separated certain types of information from the
visual array of the page and directed students' attention

to that information.

Summary. Each study reviewed in the previous
sections of this paper fell generally into one of three
broad categories; color as a contextual cue, color as an
attentional factor, and other applications of color.

The studies reviewed on color as a contextual
one often demonstrated that color facilitated rate of
initial learning but produced undesirable complications
on a transfer task. Samuels was especially concerned
with this characteristic of color cues and his research
provided some necessary cautions for the use of color
cues. Otto's research concentrated on the utility of
color cues for different subjects under varying condi-
tions. His research demonstrated that color facilitated
learning but that other cues, such as order of presenta-
tion, could have a more facilitative effect. In sum,
the research conducted with color as a contextual cue
does appear to demonstrate that color has the potential
to become an additional cue on which to base a dis-
criminative response, but that subjects often reSpond on

the basis of color rather than stimulus form.

41

The research studies focusing on color as an
attentional factor attempted to clarify the roles of
attention, distinctive features, and color cues in dis-
crimination and paired-associate learning. A number of
studies demonstrated that poor readers and children who
had difficulty with the task of visual discrimination
were inferior in their ability to attend to and identify
distinctive features of letters and words. Both Smith
and Stott are of the opinion that the inability to pick
out distinctive features stems not from a physical dis-
ability but, rather, from the fact that the child does
not know where to look or how to focus his attention.
Williams suggested that, to improve a child's ability to
make visual discriminations, it was necessary that the
child learn to identify the distinctive features of the
stimuli. Such research demonstrates that some children
are more successful in identifying distinctive features
of stimuli and that this ability seems to be closely re-
lated to attention to the stimuli; that is, some children
know how and where to focus attention while others ap-
parently do not. The question then raised is: If color
can serve as a discriminative cue, might it not also
serve as an attentional cue to direct the subject's
visual attention to the distinctive feature of a

stimulus?

42

In the final section, studies which did not neatly
fall into the previous categories were reported. Taber
and Glaser reported the results of a program which used
color as a literal cue to teach color words. Their pro-
gram vanished the color cues and was highly successful
in teaching color words to first grade children. Duell
and Anderson attempted to replicate that study with
kindergarten children and found that their children did
not learn because they responded on the basis of color
cue rather than word form. These authors then explain
that when using color cues, one must exercise caution
for if the cue has the ability to become the stimulus on
which the response is based, the intended paired-
association will not take place. These studies again
emphasize the importance of carefully structuring the
addition of color cues. Cues must not override the
primary stimulus; instead they must direct the learner's
attention to useful information.

Two themes seem to emerge from the research
studies reported; in order to make a discriminative
response, subjects must be able to select the distinc-
tive features of a stimuli and color seems to have the
necessary potential to assist the learner in finding
that feature.

However, no study attempted to use color solely

to emphasize or attentuate distinctive features of a

43

visual display. Such a study would seem to be a logical
extension of previous studies for, while it has been
shown that color can enhance learning, its potential as
a cue which aids in focusing the learner's attention on
distinctive features has not been fully realized. By
combining the use of a single hue color cue with the
concept of the search theory, teachers might assist the
learner in his search for the distinctive feature of a
stimulus. Instruction, then, could show the learner
where to look and maximize the use of his attention by
focusing it on the distinctive feature of a stimulus.

The use of multiple hue color cues often allows
the learner to focus his attention on the color rather
than the salient features of the printed stimulus. Using
a single color eliminates the possibility of a learner
identifying a stimulus on the basis of an associative
hook-up with the color cue. A single hue color cue
would serve to highlight the distinctive feature without
becoming a discriminator. When color can no longer pro-
vide the sole basis for making a discriminative response,
vanishing that color cue should be much easier. While it
has been shown that color can sometimes enhance learning,
the role of color cues in learning is nebulous. This
study is designed to clarify the role of color cues and
to evaluate an instructional prOgram employing single

hue color cues.

CHAPTER III

THE EXPERIMENT

Introduction. The study reported herein was de-
signed to determine experimentally whether adding single
hue color cues to distinctive features of low discrimi—
nability stimuli during an instructional sequence affects
the learning of visual discrimination, visual memory, and
paired-associate tasks by young children. In carrying
out the study, the experimenter was concerned with the
design of the experiment, the materials and procedures
involved, the program sequence, the population of the

study, and the treatment of the data.

DESIGN OF-THE EXPERIMENT
In order to determine the effect of the color
hues, this study employed three treatment groups. Group 1
received a program sequence in which the visual stimulus
figures were presented in plain black-on-white without
any added color cues. The program sequence for the
second group employed black stimulus figures with maximum

color cues added to all presentations except the final

44

45

criterion trial. The third treatment group was presented
visual stimuli with added color cues which were grad-
ually vanished throughout the presentations. In the.
criterion trial, the subjects responded to plain black-on-
white stimulus. Verbal instructions remained constant
between groups in an attempt to evaluate the effect of

the treatments of color on the visual stimuli. The only
variable in the program sequence was the treatment of
color between groups.

Fifty-one children from two schools were randomly
assigned to one of the three treatment groups. Each
treatment group contained seventeen subjects from each
school. Thus, thirty-four children formed each of the
three treatment groups and provided a total of 102 sub-
jects for the study.

Pre- and post-program participation measures were
collected from each student, and pre- and post-
participation means and mean gains were computed from the

scores attained.

MATERIALS AND PROCEDURES

Stimuli. The stimulus figures were selected

from a group of figures develOped and used by Gibson

58

et a1. These figures were developed to provide stimuli

 

58Gibson et al., "Discrimination of Letter-Like
Forms," pp. 897-966.

r~--- »- -p

46

which were similar to letters in English orthography but
which would eliminate the influence of previous exper-
ience with the stimuli. A description of the process
used in development of the figures was provided:

Construction of Standard Forms. An analysis was

made of actuaI letters (printed capitals, uppercase,

of the simple type customarily used in primary

texts) in terms of number of strokes, straight

vs. curved lines, angles, open vs. closed forms,

symmetry, etc. This procedure provided a set of

"rules" which describe generally the construction

of letters. New forms were generated which follow

the same constraints.59

Gibson’used these figures and their transforma-

tions in a study of the development of letter differen-
‘ tiation as it related to those features of letters
critical for identification. This study was replicated
using real letters. A high correlation (r + .87) was
found between the confusions of the same transformations
for real letters and the letter-like forms. Gibson con-
cluded: (1) that the effect of a given transformation
is not Specific to any given form, and (2) the equivalence
of the letter-like forms to letters of the English

alphabet.60

While Gibson generated a number of letter-
like figures, as well as twelve transformations of each

figure, only four of her figures were used in this study.

 

59Ibid., p. 897.

60E. J. Gibson, "Learning to Read," Science,
CXLVIII (1965), 1066-72.

47

Each of the four figures were paired with a single
transformation. Two figures were paired with a right to
left transformation and two were paired with an up-to-
down transformation (see Fig. l). The right to left and
up-down transformations were chosen because numerous
research studies have indicated that they are the most

difficult for young children to differentiate.61

 

Right Left Up Down

 

V

 

 

 

 

 

 

 

Fig. l.--The Four Letter-Like Figures and Their
Transformations Selected for the Study.

Plastic lettering guides were manufactured for
each of the figures to ensure comparability of figures
presented in all situations. A black ink medium ball-

point pen was used to draw the stimulus figures for all

 

61Joachim Wohlwill, and Morton Weiner, "Dis-
crimination of Form Orientation in Young Children,"
Child Develo ment, XXXV, No. 4 (DeCember, 1964), 1113-25;
Popp, 'VisuaI Discrimination of Alphabet Letters,"
pp. 221-26; Blair and Ryckman, "Visual Discrimination:
Lower Case Letter Confusion.” ‘

48

tasks. In all cases, the stimulus was in black on a white
background. These figures were printed on 2" x 2" cards
for the program sequences. The pre-test and post-test
sheets were printed on standard white paper with a litho-
graph. Identical, but separate, sheets were used for

both tasks. An orange water base color was used to high-
light figures. Orange was selected because of its bright-
ness and transparent qualities which allowed for easy
identification of the black stimulus figure after the
addition of the color highlighting.

Three sets of seventy-two 2" x 2" cards were pre-
pared for the program sequences. The first set consisted
of the cards with the stimulus figures printed in black.
The second set of cards consisted of the stimulus figures
in black with maximum highlighting on trials 1 through 7,
while trial 8 presented plain black figures. The third
set consisted of plain black figures in trial 8, with
maximum color beginning on trial 1 and subsequently
vanished color throughout the remaining trials (2-7).

(See Appendix A for illustration of highlighting se-
quences.) All subjects were presented plain black figures
in an introductory trial preceding the beginning of the
prOgram sequence to familiarize them with the task re-
quired, the responses, the stimuli and the experimental

setting.

49

The consonant-vowel-consonant (CVC) trigrams
paired with each figure were selected from a group
developed by Archer.62 Each pair of CVC trigrams was
randomly assigned to pairs of stimulus figures (see
Fig. 2). The trigram pairs were taken from two lists and
matched for percentile of association value as determined

by Archer.63

Only trigrams having an association value
of 25 percent or less were used. Appendix B lists the

trigrams and their association value.

 

miv kif boj wuc

 

puv woy yad dax

 

 

 

 

 

 

Fig. 2.--The Letter-Like Stimulus Figures and
Their Paired CVC Trigram.

 

62E. J. Archer, "A Re—evaluation of the Meaning-

fulness of All Possible CVC Trigrams," Psychological
Monographs, LXXIV, No. 497 (1960).

 

 

631bid.

50

Apparatus. In both schools, the experiment took

 

place in small rooms that measured about 8' x 12'. These
rooms were separate from the classroom. The experimenter
sat at a low table which supported the audio-flashcard
reader and supply of pre-tests and post-tests. The sub-
jects sat directly opposite the experimenter with the
flashcard reader directly in front of him. The audio
flashcards were stacked alongside the flashcard reader
and were placed in position by the experimenter.

The flashcard reader employed in this study
was a model 101 produced by Electronic Futures Incor-
porated (see Appendix G for illustration). The audio-
flashcards are products of the same firm and were
develOped for use in the flashcard reader. The flash-
cards are approximately 11" x 5” and, because a strip of
recording tape is affixed to the back of the card, they
have audio capabilities. The EFI Audio Flashcard System
(was chosen for this study because of its simplicity, its
capability of ensuring identical verbal instructions, and

its novelty for the subjects.

Procedures. In each school, the experimenter
spent the first day observing and participating in class-
room activities. This was done to familiarize the
experimenter with both the classroom activities and the
subjects. On the second day, the experimenter again

participated in classroom activities and later began

51

pre-testing children in groups of two until ten subjects
had been pre-tested on ability to match stimulus figures
to form and match stimulus figures from memory. The
third day, and subsequent days, were a mixture of pre-
senting the subjects with the proqram sequence, post-
testing, pre-testing additional subjects, and partici-
pating in classroom activities.

Pre-tests began with the experimenter engaging
each Subject in conversation and then presenting the
directions (see Appendix C) for the tasks. Pre-tests
were administered one day in advance of participation in
the program sequence.

Prior to beginning the program sequence, each
subject was presented with an introductory trial to
familiarize him with the operation of the machine and
also to ensure that the subject could enunciate the CVC
trigram clearly. Immediately upon completion of this
sequence, the subject entered the program sequence. The
four pairs of figures, randomly ordered to reduce serial
learning, were then presented to the subjects and con-

tinued without interruption.

THE PROGRAM SEQUENCE
The total program sequence, including the intro-
ductory trial, consisted of seventy-two audio flashcards.

While total time for program completion varied, average

52

time from beginning to end was approximately 10-12
minutes. (For a complete transcript of verbal directions
on the audio flashcards, see Appendix D.)

The development of the program was influenced
by the results of a pilot study which provided pertinent
data relevant to the following areas of concern: (1)
number of stimulus figures presented, (2) type of stimulus
figures presented, (3) number of presentations per figure,
(4) length of program, (5) type of highlighting to be
employed; and (6) reinforcement procedures. These areas

of development are discussed in the following sections.

Number and type of stimulus figures presented.
Prior to the pilot study, eight figures and their trans-
formations were selected for the study and a program
sequence was develOped. This sixteen figure program
took nearly 30 minutes to complete and, because of short
attention spans of young children, the experimenter de-
cided to shorten the program to twelve figures. These
programs incorporated numerous figures in an attempt to
reduce the possibility that any student would achieve a
maximum score on the pre-tests. During the pilot study,
the twelve figure prOgram sequence was used and few
problems with this "ceiling effect" developed. However,
the program sequence still took nearly 20 minutes to
complete. This led to the decision to eliminate two more

figures and their transformations, or four figures.

53

Each time figures were eliminated, one left to
right and one up-down transformation was eliminated.
Figures which had been most often correctly identified
in the pilot study were eliminated. This was done to
reduce the ”ceiling effect." By using figures which
were more difficult to identify, subjects were less
likely to achieve high scores on pre-achievement

measures 0

Number of presentations per figure. The number
of presentations per figure was initially chosen on the
basis of other similar research. However, an introductory
trial was added when the pilot study demonstrated that,
either because of the audio capabilities of the machine
or the auditory discrimination and memory of the sub-
jects, the subjects often failed to properly enunciate
the trigram after several presentations. The introduc-
tory trial was designed to control this by presenting
the subjects with the unhighlighted stimulus and its
paired trigram and asking each subject to repeat the
trigram. If they did so incorrectly, they were corrected
by the experimenter. This introductory trial also served
to present the subjects with the tasks involved prior to
program entry.

Each figure was presented eight times following
the introductory trial. The pairs of figures were

randomly ordered throughout these presentations. The

54

sequence remained the same for all treatment groups. The
first trial following the introductory trial was a learn-
ing trial. These trials provided the trigram paired with
the figure and directed the subject to look at a parti-
cular feature of the stimulus figure. These verbal direc-
tions also remained constant for all treatment groups.

The second trial following the introductory trial
was a test trial. These trials asked the subject if he
could recall the "name" of the figure. Each subject was
then directed to push another button on the machine to
check the correctness of his answer or to find out the
"name" of the figure if he had failed to respond with a
trigram. A second sound track on the tape affixed to the
audio flashcard provided the subject with the appropriate
"name" of each figure. Throughout the remainder of the
program sequence, learning and test trials were alternated,
with the final test trial serving as the post-achievement

measure for the naming task. Figure 3 illustrates the

sequence.

 

 

I l 2 3 4 5 6 7 8
E 0' U‘ 00 U! 0'0 U‘ I:
a 3‘ .5 5 a .9. :3 s. .2
=3 2 U .8 :3 I: «U £2 H“
'U U m U] U U) U 00)
0 ll 0 III 0 1U a) CU +30
:3 [g [-4 E2 E4 [2: B a 115-0
a U

 

 

 

 

 

 

 

 

 

 

 

Fig. 3.--Program Sequence: Order of Teaching and Testing
Trials.

55

Length of programs. It was decided that approxi-
mately 10-15 minutes was the maximum length of time that
young children could maintain interest in the task. One
other factor also had to be considered. Each subject was
post-tested following completion of the program sequence
and this testing took an additional 5-7 minutes. Thus,
the young subjects employed in this study were spending
approximately 15 minutes away from the classroom.

In summary, the program was shortened based on
data collected in a pilot study. This was done in an
attempt to make the program length appropriate to the
attention Span of the young subjects. After initial
development of the program sequence employing eight pre-
sentations per figure, the only change was to add the
introductory trial for the reason given above. In all

other ways, the eight presentations were satisfactory.

Type of highlighting. Three styles of vanishing
highlighting were employed in the pilot study. In each
style, the features selected as distinctive remained
constant but the method of highlighting differed.

Figure 4 presents the three styles of highlighting com-
pared in the pilot study. One letter-like figure

is presented as an example of each style. On the basis
of the results of the pilot study, style 3 was chosen for
use in this study. Figure 5 presents the treatment of

color for each treatment group throughout the complete

56

 

Style 1 Style 2 Style 3

 

 

 

 

Fig. 4.--Styles of Color Highlighting Employed in
the Pilot Study.

 

1T11==Color Highlighting

 

Trial:
Introductory

Tl/V ANNA/NANA

TZ/V AAA/\A/‘A/N
WV AAA/\AA/N/V

Fig. 5.--Style of Color Highlighting Used for
Each Treatment Group.

2 3 4 5 6 7 8

 

 

 

 

 

Notes: T1 no color

T2 maximum color

T3 vanished color.

 

57

program sequence. Again, one letter-like figure is pre-
sented as an example (see Appendix A for complete high-
lighting sequence).

Three sets of 2" x 2" stimulus figure cards were
prepared, one set for each treatment of color. When all
subjects from any treatment group had finished the program
sequence, the 2" x 2" stimulus figure cards for the next
treatment group were placed on the audio flash cards. To
facilitate changing the visual stimuli, the 2" x 2" stimu-
lus figure cards were attached to the audio flashcard with
a small paper clip. Thus, changing the stimulus figure
cards was a relatively simple task while having the addi-
tional advantage of insuring that each subject, regardless
of treatment group, received identical verbal instructions
throughout the program sequence. As each subject began
the final trial, his verbal responses to the visual stimuli
were recorded on a separate score sheet by the experimenter.
Upon completion of the program sequence, each subject was
post-tested on ability to match the stimulus figures to
form and to match the stimulus figures from memory. The
final trial of the program sequence served as the criter-
ion test for ability to name the figures. Upon completion
of the program sequence and post-tests, subjects returned

to scheduled classroom activities.

Reinforcement procedures. Throughout the experi-

ment reinforcement procedures were constant for all

58

participants. Small candies served as reinforcers. Each
subject was presented with two candies for each correct
verbal response on test trials throughout the program. In
addition, each subject received one small candy on the
third card of each learning trial, together with the com-
ment, "You're really working hard." This method was
developed as a result of a pilot study which demonstrated
that some subjects tended to become disinterested and dis-
couraged when they were unable to correctly respond to any
figure. Including reinforcement for only maintaining

attention to the task eliminated the problem of disinterest.

Test forms. Identical forms were used in admin-
istering the pre-test and the post-test. These forms
were constructed following a model proposed by Smith.
He suggested that
. . . foils be selected so that recognition of the unit
must be based upon unique characteristics. Thus,
other members of its class are used. They have some
features in common (class features), thus are maximally
confusable, and they differ from the unit in other
characteristics (unique features).64
Each form then contained the correct figure, its trans-
formation, an incomplete figure and/or a figure with an

added feature and/or a transformation of the figure other

than one used in the program sequence. Position of

 

64Donald E. P. Smith, "On Discrimination Pro-

gramming" (unpublished paper from the School of Educa-
tion, Bureau of Psychological Services, and Center for
Research on Language and Language of Behavior, Univer-
sity of Michigan), p. 21.

59

correct answer was randomly determined. The pre- and

post-tests used in this study can be found in Appendix E.

POPULATION OF THE STUDY

Subjects. The subjects were 102 children enrolled
in kindergarten. Samples were randomly drawn from the
kindergarten populations of the Beach Elementary School
in the Cedar Springs Public Schools District, Cedar
Springs, Michigan and the Cody Community School in the
Flint Public Schools Districts, Flint, Michigan. These
subjects were randomly assigned to one of three treatment
groups and no subject was disqualified.

Subjects in the study ranged in age from 5 years
3 months to 6 years 1 month. Fifty-four males and forty-

eight females participated in the experiment, as illus-

trated in Figure 6.

 

 

 

School Males Females Total N
Flint 31 20 51
Cedar Springs 23 28 51

Total n = 54' 48 102

 

 

 

Fig. 6.--Subjects Participating in Study: By Sex
and School Location.

The schools. The Beach Elementary School in Cedar

Springs, Michigan houses the primary grades (K-4) for the

60

Cedar Springs School District. Three kindergarten class-
rooms are housed in the building. The student pepulation
of Beach Elementary School is drawn from the town of Cedar
Springs, and from the surrounding rural areas. The Cody
Community School in Flint, Michigan houses students in

all elementary grades (K-6). Cody Community School has
two kindergarten classrooms and draws its students from

an urban working class pOpulation in Flint.

THE TREATMENT OF THE DATA

Experimental design. A two-way multivariate and

 

three univariate analyses of covariance were employed
as the primary statistical procedure. The analysis of
covariance was used because:
Simply stated, the analysis of covariance is a
statistical technique which tests the significant
differences between two or more groups after
initial differences between the groups are
statistically eliminated. The advantage of the
analysis of covariance is that . . . should there
be any initial random error between groups, this
can be eliminated statistically.65
Thus by reducing the error variance, this procedure
provides a more sensitive test of the between-group
differences.
The design, shown in Table 1, included two

independent variables: the treatments of color used in

 

65Gilbert Sax, Eppirical Foundations of Educa-
tional Research (Englewoo s, . .: Prentice-
HaII, Inc., I968), p. 35.

 

 

61

the program sequence, and the school location. The co-

variates were the achievement scores on the Match to Form
and the Match from Memory pre-tests. The dependent vari—
ables were the achievement scores on the three post-test

measures: Match to Form, Match from Memory, and Naming.

TABLE 1.--Experimental Design.

 

 

 

 

 

 

 

 

l 2 3
F FNC FMC FVC
CS CSNC CSMC CSVC
Notes:
School
Flint (F) = sample drawn from all kindergarten

children enrolled at Cody Com-
munity School.

Cedar Springs (CS) = sample drawn from all
kindergarten children enrolled at
Beach Elementary School.

Treatment of Color
No Color (NC)

 

figures receive no added
color.

Maximum Color (MC) figures receive maximum
added color throughout

program sequence.

Vanished Color (VC) figures receive maximum
added color, which is
gradually vanished
throughout the program

sequence.

Statistical procedures. For each main effect
factor (treatment and school) and for the interaction

effect (treatment X school), a multivariate analysis of

62

covariance provided an over-all test based on all three
dependent variables (see Figures 7, 8 and 9). If a source
of variation was found to be significant in the multivariate
analysis of covariance, univariate analyses of covariance
were then applied to the data to identify on which dependent
variables the significant effect occurred. This procedure
was then followed, when appropriate, by a post hoc analysis
employing the Scheffé technique.

The Match to Form and the Match from Memory pre-
tests were used as the covariates in an attempt to gain
greater statistical precision in the analysis. While
random assignment of subjects statistically equates groups,
the use of the covariate allows for reduction of the mean
square within (MS error) by eliminating that portion of

the post-test variance which is predicted by the covariates.

Summa y. This study was conducted to evaluate the
effect of added color cues in discrimination and paired-
associate learning. To this end, the study utilized 102
children enrolled in kindergarten classrooms in,two public
school districts in Michigan.

Subjects were randomly selected and assigned to
one of three treatment groups, no color added, maximum
color added, and vanished color. After the administration
of two pre-test instruments, each subject completed an
audio-visual pregram sequence designed to train him to

visually discriminate and name four pairs of high

63

 

 

34 n 34 34

:3
ll
:3
ll

 

 

 

 

 

Fig. 7.--Number of Subjects per Cell for the
Treatment Effect.

 

 

Flint n = 51
Cedar
Springs n = 51

 

 

 

 

Fig. 8.--Number of Subjects per Cell for the
School Effect.

 

 

 

 

T1 T2 T3
Flint n = 17 n = 17 n = 17
Cedar
Springs n = 17 n = 17 n s 17

 

 

 

 

Fig. 9.--Number of Subjects per Cell for the
Treatment X School Interaction Effect.

I

 

64

similarity letter-like figures. Upon completion of the
program, each subject was tested on the ability to perform
three tasks; name each figure, select each figure from
memory, and match each figure to form.

Multivariate and univariate analyses of covariance
were used to determine whether significant differences
existed between treatment groups. In an attempt to elimi-
nate the possibility of differences between treatment
groups prior to exposure to the program sequence, the pre-
achievement measures were employed as the covariates. The

results of these analyses are reported in Chapter IV.

CHAPTER IV

RESULTS

Introduction. This study was designed to evaluate
the effects of adding color highlighting to distinctive
features of high similarity letter-like figures. To this
end, an audio-visual program sequence was developed.

This program sequence was employed to measure the effect
of the color highlighting in three treatment conditions.
These conditions were: (1) no color added to the visual
stimuli, (2) maximum color added to the visual stimuli,
and (3) maximum color added to the visual stimuli and
subsequently vanished through the program sequence.

The subjects for this study were 102 randomly
selected and assigned students from the kindergarten
populations of the Beach Elementary School, Cedar Springs,
Michigan and the Cody Community School, Flint, Michigan.

Each subject was administered a Match to Form
and a Match from Memory pre-test prior to entry into the

prOgram sequence. These measures served as covariates in

65

66

the analysis. Three post-tests served as the dependent
variables and were administered upon completion of the
program sequence. These were: (1) the Match to Form
post-test, (2) the Match from Memory post-test, and
(3) the Naming task. For all measures a score of one
was given for each correct choice, with a maximum score
of 8 for each measure.

Specifically this study was designed to test the

following hypotheses:

Hypothesis 1a

Given the task of Matching to Form, there will be
no significant difference in achievement between
groups upon completion of the program sequence.

Hypothesis 1b

Given the task of Matching to Form, there will be no
significant difference in achievement upon com-
pletion of the program sequence, between the

samples drawn from the two school populations.

Hypothesis 1c

There will be no significant interaction effect
between treatment and school as measured by
achievement on the Match to Form task.

Hypothesis 2a

Given the task of Matching from Memory, there will
be no significant difference in achievement
between groups upon completion of the program
sequence.

67

Hypothesis 2b

 

Given the task of Matching from Memory, there
will be no significant difference in achievement
upon completion of the program sequence, between
the samples drawn from the two school populations.

Hypothesis 2c

There will be no significant interaction effect
between treatment and school as measured by the
Match from Memory task.

Hypothesis 3a

Given the task of Naming, there will be no signi-
ficant differences in achievement between groups
upon completion of the program sequence.
Hypothesis 3b

Given the task of Naming, there will be no signi-
ficant difference in achievement upon completion
of the program sequence, between the samples
drawn from the two school populations.

Hypothesis Be

There will be no significant interaction effect

between treatment and school as measured by the
Naming task.

DETERMINING WHETHER SIGNIFICANT
DIFFERENCES EXIST

Rationale. The multivariate and univariate

 

analyses of covariance are statistical techniques for

testing whether or not two or more population means for

68

an effect differ at a specified significance level.66

Thus, the first step in this analysis was to conduct a
two-way multivariate analysis of covariance in order to
test the above hypotheses. The results of the multi-
variate analysis of covariance identified a significant
main effect for the treatment variable. However, while
this analysis indicated a significant treatment effect,
it does not identify on which of the three dependent
variables that significant effect occurred. Therefore,
in order to identify on which dependent variables a
significant treatment effect existed, a two-way univariate
analysis of covariance was conducted on each dependent
variable.

The univariate analysis of covariance identified
on which of the three dependent variables significant
treatment effects occurred.

Having obtained a significant treatment effect
for each univariate analysis of covariance, it was
necessary to identify between which treatment groups the

significant difference occurred. To this end, the post

hoc Scheffé was then applied.

 

66Linda Glendenning, "Posthoc: A Fortran IV
Program for Generating Confidence Intervals Using Either
Tukey or Scheffé Multiple Comparison Procedures” (un-
published occasional paper #20, from the Office of
Research Consultation, School for Advanced Studies, Col-
lege of Education, Michigan State University, January,
1973).

69

The following sections report the results of the multi-
variate analysis of covariance, the univariate analyses

of covariance, and the post-hoc analyses.

Results of the multivariate analysis of co-
variance. The first statistical analysis applied to the
data was a two-way multivariate analysis of covariance.
In this analysis, the two independent variables were
treatment and school. There were two covariates: (l) the
Match to Form, and (2) the Match from Memory pre-tests.
Three dependent variables existed: (1) the Match to Form
post-test, (2) the Match from Memory post-test, and
(3) the Naming task. Table 2 reports the results of the
multivariate analysis of covariance as applied to the
treatment and school variables.

TABLE 2.--A Multivariate Analysis of Covariance on Three
Dependent Variables: (1) the Match to Form

Post-test, (2) the Match from Memory Post-test,
and (3) the Naming Task.

 

 

Source of Variation df F
(Treatment 6,184 8.4561*
School 3,92 1.7031
Treatment X School 6,184 1.3561

 

*p < .0001.

70

For the treatment effect, the table illustrates
an F-ratio of 8.4561 was attained which, with 6 and 184
degrees of freedom, had a significance level of p less
than .0001. This provides strong evidence that signi-
ficant differences exist between groups as the result of
the treatment. Thus, on the basis of this multivariate
analysis of covariance test, hypotheses 1a, 2a, and 3a
were rejected.

In contrast, for the school variable, an F-ratio
of 1.7031, with 3 and 92 degrees of freedom, had a signi-
ficance level of p less than .1719, indicating no main
effect for school. Thus hypotheses lb, 2b, and 3b were
not rejected; that is, no significant differences in
achievement existed between the samples drawn from the two
school pOpulations on any of the three dependent
variables.

When considering the interaction between treat-
ment and school, an F-ratio of 1.3561 was obtained which,
with 6 and 184 degrees of freedom, had a significance
level of p less than .2347. Thus, hypotheses 1c, 2c, and
3c were not rejected; that is, there was no significant
interaction effect between the variables treatment and
(school on any of the three dependent variables.

In summary, the multivariate analysis of co-
variance provided strong evidence that significant dif-

ferences did exist and that these differences were not

71

attributable to the differences in school populations.
Rather, the differences were attributable to the effects

of the treatments.

IDENTIFYING WHERE SIGNIFICANT
DIFFERENCES EXIST

Rationale. In order to determine whether the
significance noted above was in Match to Form, Match from
Memory, or Naming, it was necessary to apply a two-way
univariate analysis of covariance to each dependent
variable. In each univariate analysis of covariance, the
independent variables were treatment and school.

In the multivariate analysis of covariance, group
means are adjusted statistically in order to provide
precision.’ These means are adjusted from the original
means of the raw scores in relation to the amount of vari-
ance accounted for by the covariates. The adjusted means
are used as a basis for comparing groups. Table 3 presents
TABLE 3.--Adjusted Group Means for Each Treatment Group

on the Three Dependent Variables: (1) Match

to Form Post-Test, (2) Match from Memory Post-
Test, and (3) Naming Task.

 

 

 

Treatment
Variable
1* 24* 3***
Match to Form 4.9949 5.6083 6.4621
Match from Memory 3.6438 3.9049 5.2817
Naming 0.6942 1.0218 1.4159

 

*No color. **Maximum color. ***Vanished color.

72

the adjusted group means on each dependent variable.
(Group means and raw scores for each subject on all tasks

can be found in Appendix F.)

Multiple correlation coefficients. Table 4
presents the coefficients of determination between the
two covariates and each dependent variable. The co-
efficients of determination is that proportion of the
variance of a dependent variable which may be predicted

or accounted for by another variable or variables.67

TABLE 4.—-Mu1tip1e Correlation Coefficients and Coeffi-
cients of Determination for Both Covariates
with Each Dependent Variable.

 

Variable df R I R2 F

 

Match to Form 2,94 0.6933 0.4807 43.5071*
Match from Memory 2,94 0.6011 0.3614 26.5956**
Naming 2,94 0.3435 0.1180 6.2868***

 

*p < .0001 **p < .0001 ***p < .0028

As indicated by Table 4, the multiple correlation
coefficients for the two covariates with each of the
three dependent variables were statistically significant.

As the R2 coefficients indicate, the two covariates

 

67John T. Roscoe, Fundamental Research Statistics

for the Behavioral Sciences (New York: HoIt, Rinehart
& Winston, Inc., 1969), p. 80.

73

account for 48 percent, 36 percent, and 11 percent of the
variances on the dependent variables (Match to Form, Match
from Memory, and Naming, respectively).

The results for the univariate analyses of co-
variance and post hoc comparisons are reported separately

for each dependent variable in the sections that follow.

Analysis on the Match to Form variable. This

 

section reports the results of the two-way univariate
analysis of covariance and the Scheffé post hoc analysis
as applied to the Match to Form dependent variable. Both
the Match to Form and the Match from Memory pre-tests

were employed as covariates. Table 5 presents the results
of the univariate analysis of covariance on the dependent
variable Match to Form.

TABLE 5.--Univariate Analysis of Covariance on the
Dependent Variable Match to Form.

 

 

Sources of Variation df MS F
Treatment 2 22.7561 16.2927*
School 1 0.9262 0.6631
Treatment X School 2 4.4859 3.2117
Within Cell 94

 

*p < .0001

74

'This table illustrates that the main effect for
treatment had an F-ratio of 16.2927 which, with 2 degrees
of freedom, had a significance level of p less than .0001.
Since this was less than .05 (the p value specified),
hypothesis la for the Match to Form was rejected. This
means that significant differences exist between treat-
ment groups on the Match to Form variable.

Given a significant F-ratio for the treatment
effect in the univariate analysis of covariance, it was
necessary to conduct a post hoc analysis in order to
determine where significant differences between treatment
groups occurred.

The Scheffé post hOc technique was applied to
test for differences between the adjusted means (Y’) of
the three treatment groups: (1) no color (Y1),

(2) maximum color (Yé), and (3) vanished color (Yg).
Table 6 reports the results of that analysis.

TABLE 6.--Scheffé Post Hoc Comparisons of Treatment Group
Adjusted Means on the Dependent Variable Match

 

 

to Form.
. 99% Confidence 95% Confidence
Compar1sons Interval Interval
_l _:
Y3 — Y = .8538 i .8976 r .7176*
I; - I; = 1.4672 : .8923" --
_: _I
Y2 - Y1 = .6134 11.1159 31.0387

 

*Significant at p < .05 **Significant at p < .01

75

An inspection of Table 6 reveals that there was a
significant difference at the .01 level between the
treatment group receiving vanished color highlighting and
the group receiving no color highlighting. This differ-
ence favored the vanished color treatment. A comparison
of the vanished color treatment with the maximum color
treatment yielded another significant difference at the
.05 level, again favoring the vanished color treatment.
No significant difference was found between the maximum
color treatment group and the no color treatment group

on the Match to Form variable.

Analyses on the Match from MemoryAvariable. This

 

section reports the results of the univariate analysis

of covariance and the Scheffé post hoc analysis, applied

to the Match from Memory dependent variable. The Match

to Form and the Match from MemOry pre-tests were employed
as covariates.

Table 7 presents the relevant data for the uni-
variate analysis of covariance on the dependent variable
Match from Memory.

Table 7 illustrates that the main effect for the
variable treatment yielded an F-ratio of 4.5445 which,
with 2 degrees of freedom, had a significance level of
p < .0001. Since this was less than .05 (the p value

specified), hypothesis 2 for the Match from Memory

76

dependent variable was rejected. This means that signi-
ficant differences exist between treatments on the Match
to Form variable.

TABLE 7.—-Univariate Analysis of Covariance on the
Dependent Variable Match from Memory.

 

 

 

Sources of Variation df MS F
Treatment 2 24.6935 4.5445*
School 1 3.5706 2.9352
Treatment X School 2 1.0330 .8492
Within Cell 94

*p < .0001

Given a significant F—ratio for the treatment
effect in the univariate analysis of covariance, it was
necessary to conduct a post hoc analysis in order to
determine where significant differences between treatment
groups occurred.

The Scheffé post hoc technique was applied to
test for differences between the adjusted group means.
Table 8 presents the results of this analysis for the
three treatment groups: (91) no chor, (Yé) maximum
color, (Y3) vanished color.

An inspection of Table 8 reveals that there were
significant differences at the .01 level between the

treatment group receiving vanished color and the groups

‘77

receiving maximum color and no color highlighting. In
both comparisons, the difference favored the vanished
color group. No significant difference was found between
the maximum color treatment group and the no color treat-
ment group on the Match from Memory variable.

TABLE 8.--Scheffé Post Hoc Comparisons of Treatment Group

Adjusted Means on the Dependent Variable Match
from Memory.

 

 

. 99% Confidence 95% Confidence
Comparison Interval Interval
_I _I *
— = + -—
Y3 Y2 1.3768 - .8378
I I
._ _ _. = * __
Y3 Y1 1-6379 1 .8325
I I
I? - Ii - .2611 :1.0411 1.8324

 

*Significant at p < .01

Analyses on the Naming variable. This section

 

reports the results of the two-way univariate analysis of
covariance and the Scheffé post hoc analysis, as applied
to the Naming dependent variable. Both the Match to Form
and the Match from Memory pre-tests were employed as
covariates.

Table 9 presents the results for the univariate
analysis of covariance on the Naming dependent variable.

Table 9 illustrates that the main effect for the
variable treatment had an F-ratio of 4.5445 which, with

2 degrees of freedom, was significant at the .0131 level.

78

Since this was less than .05 (the p level specified),
hypothesis 3 for the Naming dependent variable was re-
jected. This means that significant differences exist
between treatments on the Naming variable.

TABLE 9.--Univariate Analysis of Covariance on the
Dependent Variable Naming.

 

 

 

Sources of Variation df MS F
Treatment 2 4.8009 4.5445*
school 1 .0657 ' .06222
Treatment X School 2 1.3926 1.3182
Within Cell 94

*p < .0131

Given a significant F-ratio for the treatment
effect in the univariate analysis of covariance, it was
necessary to conduct a post hoc analysis in order to
determine where significant differences between treatment
groups occurred.

The Scheffé post hoc technique was applied to
test for differences between the adjusted group means.
Table 10 presents the results of this analysis for the
three treatment groups: (Y1) no color, (Y5) maximum

_I
color, and (Y3) vanished color.

79

TABLE 10.--Scheffé Post Hoc Comparisons of Treatment Group
Adjusted Means on the Dependent Variable Naming.

 

 

Comparison 95% Confidence Interval
I; - I; = .3841 1.6242
3?; - I; — .7217 1.6202*
F; - I; = .3276 4.7759

 

*Significant at p < .05

An inspection of Table 10 reveals that there was
a significant difference at the .05 level between the
treatment group receiving vanished color and the treatment
group receiving no color highlighting. This difference
favored the vanished color treatment. No significant
differences were found between the vanished color and the
maximum color treatments nor between the maximum color

and the no color treatments on the Naming variable.

Summa y. Having found a significant main effect
for the treatment variable in the multivariate analysis
of covariance, a univariate analysis of covariance was
then applied to each dependent variable.

The univariate analysis of covariance on the
dependent variable Match to Form indicated a significant
treatment effect at the p < .0001 level. The Scheffé post

hoc comparisons identified significant difference

80

(p < .01) between the vanished color and the no color
treatments, favoring the vanished color treatment. A
significant difference (p < .05) was also found between
the vanished color and maximum color treatments, again
favoring the vanished color treatment.

The univariate analysis of covariance on the
dependent variable Match from Memory indicated a signi-
ficant treatment effect at the p < .0001 level. The
Scheffe post hoc comparisons identified a significant
difference (p < .01) between the vanished color and the
no color treatment and between the vanished color and the
maximum color treatment. Both significant differences
favored the vanished color treatment.

The univariate analysis of covariance on the
dependent variable Naming indicated a significant treatment
effect at the p < .0131 level. The Scheffe post hoc
comparisons identified a significant difference (p < .05)
between the vanished color and the no color treatments,

favoring the vanished color treatment.

CHAPTER V

SUMMARY, CONCLUSIONS, IMPLICATIONS

AND RECOMMENDATIONS

Introduction. This study was designed to experi-
mentally evaluate whether adding single hue color cues
effects the learning of visual discrimination, visual
memory, and paired-associate tasks.

A review of the research concerned with adding
color cues found that color can have a facilitative
effect upon the rate of initial learning. However,
eliminating the color cues presented difficulty. Often
the color cue became the primary stimulus upon which the
discrimination was based. In other words, the color cue
was often so potent that it overrode the relevant cue
of stimulus form. In these research studies, two or more
colors were used, with separate colors being paired with
each stimulus figure. The study reported in this paper
was designed to test the effect of employing a single
color as a cue for a variety of stimuli. It was felt that

the use of a single color might eliminate the problems

81

82

previously associated with the use of multiple color cues
while still focusing subjects' attention on the distinc-

tive features of the stimuli.

S UMMARY

The study. To evaluate the usefulness of a

 

single hue color cue, a program sequence was developed.
This program was prepared for use with an audio flashcard
reader. An audio track was developed and these verbal
directions remained constant for all treatment groups.
Four letter—like figures and their transformations were
selected as visual stimuli. The transformations of left
to right and up to down, were chosen because of their
difficulty. Each letter-like figure was randomly assigned
a low meaningfulness CVC trigram as its name.

Three sets of the figures were prepared for use
in the program sequence. The first set had no color cues
added to the stimulis figures, the second set had maximum
color cues added to the stimulus figures, and the third
set had maximum color cues added and subsequently
vanished from the stimulus figures.

This program sequence was used with 102 children
enrolled in the kindergarten classrooms of two Michigan
school districts. Each subject was pre-tested on the
ability to match the letter-like figures to form, and the

ability to match the letter-like figures from memory.

83

Following the administration of the pre-tests, subjects
participated individually in the program sequence.
Immediately following completion of the program sequence,
each subject was post-tested on the ability to name the
letter—like figures, to match the letter-like figures
from memory, and to match the letter-like figures to
form.

In order to determine whether significant dif-
ferences existed between the three treatment groups, a
multivariate and three univariate analyses of covariance
were applied to the data. The Match to Form and the
Match from Memory pre-tests were employed as covariates

in these analyses.

The results. On the basis of the results of the
multivariate analysis of covariance, no significant dif-
ferences in achievement between school populations were
found on the three variables: Match to Form, Match from
Memory, and Naming. Similarly, no significant inter-
action effect existed for the treatment X school vari-
able. This means that the treatments of color had
similar effects for subjects in both schools.

The results of the multivariate analysis of
covariance did, however, reveal significant differences
for the treatment effect. A univariate analysis of

covariance was then applied to each dependent variable

84

in order to identify where the significant differences
occurred.

For the Match to Form variable, significant
differences (p < .0001) were found in the univariate
analysis of covariance. Given these differences, a
Scheffé post hoc comparison was applied in order to
identify between which treatment groups these signi-
ficant differences existed on the Match to Form variable.
The results of the Scheffé analysis revealed that the
vanished color treatment was significantly better
(p < .01) than the no color treatment. The vanished
color treatment was also found to be significantly
better (p < .05) than the maximum color treatment on
the Match to Form measure.

For the Match from Memory variable, significant
differences (p < .0001) were found in the univariate
analysis of covariance. Given these differences, a
Scheffe post hoc analysis was applied in order to
identify between which treatment groups the significant
differences existed on the Match from Memory variable.
The results of the Scheffé analysis revealed that the
vanished color treatment was significantly better
(p < .01) than either the no color or the maximum color
treatment.

For the Naming variable significant differences

(p < .0131) were also found in the univariate analysis of

85

covariance. Given these differences, a Scheffé post hoc
comparison was again applied in order to identify between
which treatment groups the significant differences occurred
on the Naming variable. The results of the Scheffé analy-
sis revealed that the vanished color treatment was signi-
ficantly better (p < .05) than the no color treatment.

The results of the post hoc analyses indicate that
the vanished color treatment was superior to the no color
treatment on each of the three learning tasks. Further,
the vanished color treatment was superior to the maximum
color treatment on the Match to Form and the Match from

Memory tasks.

CONCLUSIONS

The major conclusion of this study is that Children
who receive instruction which utilizes vanished color cues
learn the tasks of viSual discrimination, visual memory,
and association of a verbal response at a significantly
higher level of achievement than children who receive in-
struction without vanished color cues. Further, it is con-
cluded that the vanished color cues serve to focus atten-
tion on the distinctive feature of the figure to be learned

without producing interference at transfer.

IMPLICATIONS
The results of this study indicate that attending

to appropriate features of visual stimuli improves

86

performance on visual discrimination, visual memory,
and paired-associate tasks.

More specifically, this study demonstrates that
vanished color cues facilitate attention in learning these
three basic tasks. In contrast to previous studies using
multiple hue color cues, a single hue color cue which is
vanished seems to facilitate attention to the distinctive
features of letter-like figures without producing inter-
ference on a transfer task. These color cues were useful
because they directed the learner's attention to salient
information. A single hue eliminates the tendency to
reSpond on the basis of the color cue alone. By employing
a bright transparent color, attention can be focused on
the distinctive feature of the letter—like figure while
allowing the printed form of that figure to remain visible.

Gradually vanishing the color cue in subsequent
presentations increases the effectiveness of the cue by
forcing the subject to attend more to the distinctive
feature of the stimulus and less to the color cue. The
vanished color cues have the ability, when combined with
verbal directions, to increase stimulus differentiation
by directing attention to the salient features of the
stimulus rather than to the color itself.

The crucial point is that even after a limited
amount of training at an early age (i.e., kindergarten),

there occurred significant differences in achievement

87

favoring the vanished color treatment over the no color
treatment. These data indicate that the effectiveness of
visual discrimination training depends upon the technique
employed. While visual discrimination training is often
provided in kindergarten classrooms, it is not unusual for
a number of children to experience difficulty with these
tasks, especially with discriminations involving reversals
or rotations of letters. The systematic use of vanishing
single hue color cues, however, should assist these young
children in mastering the discrimination tasks more effi-
ciently. Hence, the use of vanishing color cues to focus
attention on distinctive features of visual stimuli should
help minimize many of the relatively common perceptual

confusions currently found in many elementary classrooms.

RECOMMENDATIONS
The recommendations offered here fall into two
separate categories: recommendations for teachers of

reading and recommendations for further research.

Recommendations for teachers of readigg. Teachers

 

of reading should use vanished color cues to assist
Children in
1. learning to visually discriminate between letters;

and in

2. learning to recognize easily confused letters.

88

Recommendations for further research. Further

study is needed regarding the application of vanished

color cues to other instructional tasks and the refine-

ment of the technique developed in this study. The

effect of vanished color cues need to be examined under

the following conditions:

1.

Present only two conditions, vanished color and
no color. For each treatment develop a set of
limited verbal directions (i.e., ”This is a
puv.") and compare the treatments under these

conditions.

Compare the vanished color and the no color
treatments employing limited verbal directions
and expanded verbal directions (i.e., as used in
this study). This would provide needed informa-
tion on the relative potency of the vanished

color cues.

Present the vanished color and the no color
treatments to subjects identified as making the
common error of letter reversals. Highlight
the confused letters in the same manner as

described in this study.

Present the vanished color and the no color treat-
ments to subjects identified as making common

word reversal errors or confusing similar words.

89

Again use the color to highlight the distinctive

features of the words.

5. Test the applicability of vanished color cues
to learning situations outside the area of

reading skills.

6. Test the effect of the use of vanished color
cues over an extended period when compared with

instruction without color cues.

7. Editors for publishing houses may do well to
employ the technique developed herein when de-
signing texts and materials for teaching or

practicing discrimination or association skills.

In summary, while this study has demonstrated
that vanished color cues can enhance the learning of
three basic tasks, still more research is needed into the
relative effectiveness of employing similar techniques
in other learning environs and for other learning tasks.
It has been demonstrated that attention can be appro-
priately focused with the use of vanished color cues but
further research is needed to provide an empirical base

for extending the use of color as an attentional device.

BIBLIOGRAPHY

BIBLIOGRAPHY

Anderson, R. C. "Educational Psychology." Annual
Review of Psychology. Edited by PauI Farnsworth
et a1. Palo Alto, Calif.: Annual Reviews,
Inc., 1967. Vol. XVIII, pp. 103-64.

Anderson, Roger, and Samuels, S. J. "Visual Recall Memory,
Paired-Associate Learning, and Reading Achieve-
ment." Paper read at American Educational
Research Association, Minnesota, March 2, 1970.
(ERIC ED090011)

Archer, E. J. "A Re-evaluation of the Meaningfulness of
All Possible CVC Trigrams." Psychological
Monographs, LXXIV, No. 497 (1960).

 

 

Asso, Doreen, and Wyke, Marcia. "Discrimination of
Spatially Confusable Letters by Young Children."
Journal of Experimental Psychology, XI (1971),
11-20.

Birnbaum, Isabel. "Context Stimuli in Verbal Learning and
the Persistence of Associative Factors."
Journal of Experimental Psychology, LXXI, No. 4
(1966), 483-872

Black, Harvey. "Relevant and Irrelevant Pictorial Color
Cues in Discrimination Learning." Final Re ort:
HEW Project 1170. Indiana University (I967).

Blair, John, and Ryckman, David. "Visual Discrimination:
Lower Case Letter Confusion." Unpublished paper
from the Center for Research on Language and

Language Behavior, University of Michigan.
(ERIC EDO 29770).

 

Dale, Edgar. Audio-Visual Methods in Teachipg. 3rd ed.
' New York: Holt, RInehart and Winston, 1969.

 

90

91

Davidson, Helen. "A Study of the Confusing Letters b, d,
p, and q." Journal of Genetic Psychology, XLVII
(1935), 458-68.

 

Duell, Orpha, and Anderson, R. C. "A Failure to Teach a
Sight Vocabulary by Vanishing Literal Prompts."
Paper read at the American Educational Research
Association, 1967. (ERIC ED 011817)

Dulsky, S. G. "The Effect of a Change of Background on
Recall and Relearning." Journal of Experimental
Psychology, XVIII (1935), 725-40.

 

 

Gibson, Eleanor. "Learning to Read.” Science, CXLVIII
(1965). 1066-72.

, Gibson, James; Peck, Anne; and Osser, Harry.
"A Development Study of the Discrimination of
Letter-Like Forms." Journal of Comprehensive
and Physiological PsyChdIOgy, LV, No. 6 (1962),
897-906.

 

Glendenning, Linda. "Post Hoc: A Fortran IV Program for
Generating Confidence Intervals Using Either Tukey
or Scheffe Multiple Comparison Procedures."
Unpublished Occasional Paper No. 20. East
Lansing, Mich.: Office of Research Consultation,
School for Advanced Graduate Studies, College of
Education, January, 1973.

Guthrie, E. R., and Smith, S. General Ps cholo (New
York: Appleton-Century-Cro s, .

Hall, Vernon, and Caldwell, Edward. "Analysis of Young
S's Performance on a Matching Task." Unpublished
research report from Syracuse Center for Research
and Development in Early Childhood Education,
Syracuse University, April, 1970. (ERIC ED
046644)

Haring, Norris. fittending and Res ndin . San Rafael,
Calif.: Dimension PuEIisEing 50., 1968.
Hemphill, J. K. Teachinngeading_as Decoding: Minicourse

18. BerkeIey, Calif.: Far West Laboratory for
fiancational Research and Development, 1971.

 

 

Hoban, C. F., and Van Ormer, E. B. Instructional Film
Researchy_l918-l950. Port Washington, Long
Island, N.Y.: Special Devices Center, 1950.

 

 

92

Hull, C. L. Principles of Behavior. New York: Appleton-
Century-Crofts, 1943.

Isaacs, Dan Lee. "Cue Selection in Paired-Associate
Learning as a Function of Relevance of Color
Cues and Discriminability of Shape Cues."
Unpublished master's thesis at Indiana University,
1966.

Jenkins, H. M., and Sainsbury, R. S. "Discrimination
Learning with the Distinctive Feature on Positive
and Negative Trials." ATTENTION: Contemporary
Theoryyand Analysis. Edited by DavId’NostofSky.
New York: Appleton-Century-Crofts, 1970.

 

Jones, J. Kenneth. "Colour as an Aid to Visual Perception
in Early Reading." British Journal of Educa-
tional Psychology, XXXV (1965),‘21:27.

Leicht, Kenneth, and Cashen, Valjean. "Type of High-
lighted Material and Examination Performance."
Paper read at American Educational Research
Association, February, 1971. (ERIC ED 050155)

Lumsdaine, Arthur. "Instruments and Media of Instruc-
tion." Handbook of Research on Teaching. Edited
by N. L. Gage. Chicago: Rand McNalIy, 1963.

McAnnich, Myrene. PInvestigation of Recognition Variance
of Perceptual Stimuli Associated with Reading
Proficiency.” Paper presented at International
Reading Association, Missouri, 1969. (ERIC
ED 030552)

McGeoch, J. A., and Irion, A. L. The Psychology of Human
Learning. New York: Longman‘s Green, I952.

Miller, N. E.,et a1. "Graphic Communication and the
Crisis 1n Education.” Audio-Visual Communication
Review, V (1957), 3-39.

Newman, S. E., and Taylor, R. "Context Effects in Paired-
Associate Learning as a Function of Element-
Sharing Among Stimulus Terms." Journal of Verbal
Learning and Verbal Behavior, I (I963), 243-49.

93

Otto, Wayne. "Color Cues as an Aid to Good and Poor
Readers' Paired-Associate Learning." Perception
39d Reading. Edited by Helen Smith. Newark,
Del.: International Reading Association, 1967.

 

. "Elementary Pupils' Use of Cues in Paired-
Associate Learning." Psychology in the Schools,
V, No. 2 (1968), 179-82.

. "Intralist Similarity, Order of Presentation
and Color in Childrens' Paired-Associate
Learning." Psychonomic Science, IX (1967),
531-32.

 

, and Askov, Eunice. "The Role of Color in
Learning and Instruction.” Theoretical Paper
No. 12 from the Research and Development Center
for COgnitive Learning, University of Wisconsin,
March, 1968. (ERIC ED 021697)

, and COOper, Calvin. "Investigations of the
Role of Selected Cues in Childrens' Paired-
Associate Learning." Technical Report No. 53,
Wisconsin Research and Development Center for
Cognitive Learning, University of Wisconsin,
May, 1968. (ERIC ED 036315)

POpp, Helen. "Visual Discrimination of Alphabet Letters."
The Reading Teacher, XVII (1968), 221-26.

 

Roscoe, John T. Fundamental Research Statistics for the
Behavioral Soiences. New York: HoIt, Rinehart
and Winston, Inc., 1969.

 

Saltz, Eli. "Compound Stimuli in Verbal Learning,
Cognitive, and Sensory Differentiation Versus
Stimulus Selection." Journal of Experimental
Psycholggy, LXVI (1963), 1-5.

 

Samuels, S. Jay. "Formal Intralist Similarity and the
vonRestorff Effect." Journal of Educational
Psychology, LIX (1968), 432-37.

 

. "Modes of Word Recognition." Theoretical
Models and Process of Reading. Edited By Harry
Singer and Robert Ruddell. Newark, Del.:
International Reading Association, 1970.

 

Sax, Gilbert. Empigical Foundations of Educational
Research. Englewood’Cliffs, N.J.: Prentice-

HaII, Inc., 1968.

 

94

Schutz, R. E. "Acquisition of Alphabet Letter Dis-
crimination with and Without 'Errors'." Paper
read at the American Psychological Association,
Los Angeles, 1964.

Smith, Donald E. P. "On Discrimination PrOgramming."
Unpublished paper from the School of Education,
Bureau of Psychological Services, and Center
for Research on Language and Language of Behavior,
University of Michigan.

Smith, Frank. Understanding Reading: A Psycholinguistic
Analysis of Reading andeearning to Read. New
York: Holt, Rinéhart and Winston, 1971.

 

 

Stott, D. H. "Some Less Obvious COgnitive Aspects of
Learning to Read." The Reading Teacher (January,
1973), 374-83.

 

Sunderland, D. M., and Wickens, D. D. "Context Factors
in Paired-Associate Learning and Recall." Journal
of Experimental Psychology, LXIII (1962), 302-6.

 

Taber, Julian, and Glaser, Robert. "An Exploratory
Evaluation of a Discriminative Transfer Learning
Program using Literal Prompts." The Journal of
Educational Research, LV, No. 9 (June/JuIy,
1962), 509-12.

 

 

Tighe, Louise, and Tighe, Thomas. "Discrimination
Learning: Two Views in Historical Perspective."
Psychological Bulletin, LXVI, No. 5 (1966),
353-70.

 

Travers, Robert. "Research and Theory Related to
Audio-Visual Information Transmission." Un-
published paper from the Bureau of Educational
Research, University of Utah, July, 1964.

Underwood, B. J., Ham, M., and Ekstrand, B. "Cue
Selection in Paired-Associate Learning." Journal
of Experimental Psychology, LXIV (1962), 405-9.

 

Washington, Ernest David. "Matching and Naming Letters
of the Alphabet with and without Redundant
Color Cues." Unpublished doctoral dissertation,
University of Illinois, 1968.

95

Weiss, Walter, and Margolius, Garry. "The Effect of
Context Stimuli on Learning and Retention."
Journal of Experimental Psychology, XLVIII, No. 5
(1954) , 3I8-22 .

Wheelock, Warren. "An Investigation of Visual Dis-
crimination Training for Beginning Readers."
Perception and Reading. Edited by Helen Smith.

Newark, Del.: International Reading Association,
1967.

Williams, Joanna. ”Effects of Discriminative and Re-
productive Training on Ability to Discriminate
Letter-Like Forms." Paper presented at American
Educational Research Association meeting,
February, 1968. (ERIC ED 017418)

. "Reactions to Modes of Word Recognition."
Theoretical Models and Processes of Reading.
Edited by H. Singer and R. RuddeIIl Newark,
Del.: International Reading Association, 1970.

"Training Kindergarten Children to Dis-
criminate Letter-Like Forms." American Educa-
tional Research Journal, VI (1969), 501-13.

 

Wohlwill, Joachim, and Weiner, Morton. "Discrimination
of Form Orientation in Young Children." Child

Development, XXXV, No. 4 (December, 1964),

 

APPENDICES

APPENDIX A

METHOD OF COLOR HIGHLIGHTING FOR ALL TREATMENTS

 

 

A

 

x. ”‘0‘

 

i
2:;
i

 

 

 

 

 

1‘

 

(fiffifii

 

 

 

 

 

 

‘wwwwwmsw

 

 

“mammmmmm

1an xxx

MMMKMMMM

 

 

 

4 4 4 .14 H H. ‘4-

 

 

98

VANISHED COLOR TREATMENT FOR LETTER-

LIKE FIGURES BY TRIAL

G
N
I
T
H
G
I
L
H
G
I
H
E
G
W
O
.=

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l...

 

XXXXYYYX

F~~ -_____ -,_.--.-,’... +-

DIDH>H>H>H>H>H>i

 

 

 

 

 

N764 .4 4 4 .4 4

 

 

APPENDIX B

TRIGRAMS AND ASSOCIATION VALUE

APPENDIX B

TRIGRAMS AND ASSOCIATION VALUE

 

List 1 Percentile Association Value List 2
yad 25 dax
puv 22 woy
miv l7 kif
boj 16 wuc

99

APPENDIX C

DIRECTIONS FOR PRE-TESTS AND POST-TESTS

APPENDIX C

DIRECTIONS FOR THE MATCH TO FORM PRE-TEST

Instructions to Subjects:

"Look carefully at the figure I'm pointing to. Can you find
this figure on your paper? Put your finger on it. (Check to make
sure finger is on correct figure.) Now, look carefully at each of
the other four figures in the row. Does this one look just exactly
like the first figure? Or this one? Or this one? Or this one?
Only one of these figures is just exactly the same as the first
one, put a mark on it."

"Good. Now look at this figure (point to key figure in
second row). Look carefully at each of the other figures in the
row and mark the one that looks just exactly like the first one.
Remember only one is exactly the same."

"Good. Now put your finger on the first figure in the next
row and find the other figure that looks just exactly like the
first one."

"Good. Do the rest of the rows just like that. Look at the
first figure and then mark the one that looks just exactly like it."

If necessary continue repeating instructions. If subject
marks more than one figure say, "Remember only one figure looks
just exactly like the first one." If subjects asks, "Is this

right?", or something similar, reply, ”Good, mark it."

100

101

DIRECTIONS FOR THE MATCH FROM MEMORY PRE-TEST

Instructions to subjects:

Provide subject with test sheet and cover properly aligned
over first row of figures.

"This sheet is different from the other one. This time I will
show you a figure on a card like this (hold up card). You must look
carefully at the figure on this card and try to remember what it
looks like. After you have seen the figure slide the cover down like
this and look carefully at each figure, then put a mark on the
figure that looks just like the one you saw on the card. You can't
look back because I'm going to put the card down after you've looked
at it. This is tricky, so look carefully at the figure when I show
it to you."

"Ready? Here is the first figure. Look carefully at it.
Now, slide the cover down and look at all the figures, then mark
the one that looks just like the one that was on the card. Remember
only one is just exactly like the one on the card."

"Good. Now look carefully at th is figure. Now slide the
cover down, look at all four figures and mark the one just exactly
like the one on the card."

"Good. Now here's the next one. Look carefully and mark
the figure that looks exactly like it."

"Good. Now here's the next figure." (Continue through
cards.) If subject marks more than one figure say, "Remember, only one
figure looks just exactly like the one on the card." If subject asks,

"Is this right?", or something similar reply, "Good, mark it."

102

DIRECTIONS FOR THE MATCH TO FORM POST-TEST

Instructions to Subjects:

"Remember how we did this paper before? You put your
finger on this figure (demonstrate) and look carefully at each of
the other figures, then you put a mark on the figure that is
just exactly like this first one. Remember that only one figure
looks just exactly like the first one."

Again if any questions arise concerning correctness of

subject's choice reply, "Good, mark it."

103
DIRECTIONS FOR THE MATCH FROM MEMORY POST-TEST

Instructions to Subjects:

"Do you remember how we did this paper? I'll hold up a
card like this, you look carefully at the figure on the card, then
slide the cover down and find the one that looks just exactly
like the figure on the card I held up. Remember only one figure
is just exactly like the one on the card."

"O.K., now look carefully at this figure." (Hold up card.)
"Slide the cover down and mark the figure just exactly like the
one that was on the card."

"Good, now look at this figure." (Continue through cards.)

If questions arise concerning correctness of subject's

choice reply, "Good, mark it."

‘

APPENDIX D

INTRODUCTION TO AND VERBAL DIRECTIONS

FOR THE PROGRAM SEQUENCE

APPENDIX D

INTRODUCTION TO PROGRAM

Instructions to Subjects:

"I'm trying to find out how kids learn. So what
I've done is put some figures on these cards. I want to
see if you can learn the names of these figures. This is
really hard because no one has ever learned all the names.
But every time you do remember even one name you'll get
some of these candies. OK? Let me show you how this (the
Audio Flashcard Reader) works. I'll put the card in and
then push this button. Listen (to audiotrack of card).
When you push the button that voice will tell you the name
of the figure. Later the voice will ask if you remember
the figure's name. If you do, say it, then push this
button to see if you were right. If you don't remember

the name push this button and it will tell you."

104

105

.N

.H OH
.N

.H ma
.N

.H NH
.H 0H
.H ma
.H «H
.H ma
.H NH
.H Ha
.H OH
.H m

.H w

.H m

OH 0

.H m

.H q

.H m

OH N

.H H

 

>wE

053

has

as:
«on

wax

>HE
>03

>5o
Koo

to»
mos
>so
933
“on
was
.mmx
mam
>Hz

 

Bumps oucs< noossz oucosoom sumo enough

v—l

Far:

PIP'PIPCFIPIPIPIFI

 

Howhfi

zasoe Loom H was .>He a as none:
=.usmsu
on.so% we own Ou mourn scuusn band can» use use: ago
has one sex mH woco menu «0 same can nooEosou so» can:
:deh xouuu H two so 033 Emu so» can .033 m so many:
:.unmau mums so» we mom on manna souuso base cosy
new as hem wousmfiw menu «0 same sou uOOEOEOO so» ago:
~n0b mom so» can .nOb m as book
.unwwu some 50% we own cu manna causes band soon one
mama mug hem wouswwm many we seen onu pooeosou 90% one
:.o:3 how «do usaoo omega Ono 30: wow .033 n ma many:
8.
5mm enact ucHOQ nosed on» Son mom .nOb a ma wasp:
=.Lax saw
one as ass: was was. noses mom .LHR may ma page:
=.>as Ram ago ass: was as was. Bose: .>Ha . as Ease:

hum ewswucuoo mug Ass roars 00m .%03 onu no easy:
:okfdwa %.Wm NQHUCWOQ .QUH ##Uflfis 00m 05.5“ .W a.“ OWE:

mam .xmo sou mo Eoqub Ono us 3004 .xmo
:.wmm hem .omh on» we do» moo us 3004 .mm%
:.503 how .%03

:.>sd mom .>so

:.u:s mum .o>3

.28. .8... . .3

:.xmo hum .xmo

:.om> ham .omh

=.max Ram .LHB

:.>«B ham. .>Hs

dddddddddd
a:
0:4
Q
E

 

Macauoouwo Hmouo>

106

.H on
.H om
.H mm
.H NN
.H mm
.H mm

.H ¢N

.H mm

.H mm

.H HN

.H on

 

xomuH 0Hw5¢ Honabz mon¢auom vuao

033
no;
>03

>Dm

“fix

Umz

mos

>=m

MHz

mummflm HSHH

mum .ov 30% MH
:wmo3 ham no» an

 

wwcHunHom mum menu mHs has nuHss mom
wwawunHom mum menu «nu has 30H53 mom
haw wwaHufiwom muH mw3 30H53 mom
mam wwsHunHon m.uH mm3.:UHg3 mom
wzoc do mH @855 mnu mem 30Hgs mom

wao mH mas: «nu mnHm 30Hpa mom

:.o:3.mmm
.u:3.d mH mHga:
=.Hop mum
.Hon a a“ ”Hay:

:.h93
.503.“ a“ many:
...za
.>=m m mH mHshz
=.max hum
.MHH m mH mHnH:
:.>HE hum
.32: u 3" 32h...
:.u£w.mu

mu.:om NH mum on amusu nouuan swam nonu wad uH hum
ow 50% MH .Qaua muH umnamaou so» on

.mao mHsu hay:
:NUSwHH

uH uuw do» an vam man do» fine .vmm m mH mHSH:

.. Jam?

«p.50» HH mum ou wmusu aouudn Swan fionu cam uH

Mano mHsu mo uema m£u nunsuamu so» on:
.H u.aUHv so» quOHuu H nun HH.H:

:.unw«u

ou.:o% MH mam ou mupnu aouunn nmam soap cam uH ham

«an: muH nonsuauu 50% «H .mno can: nusuoaw n.0uum:
:.uanu m.umne .>bn hum 30% an:
:.ustu 09.50% MH mom ou manna nouuan swam

fiuny .uH mum .udo mHnu mo «and «nu nonauauu :0» NH:
:wuwnu Hunaaamu 50% UHQ .MHx nH mHSH .maM:

:.uanu 09.50%
HH mum on amusu nouudn Swan nonu and «sun m.uH hum

.oao mwsu yo mama «nu umnamamu 30% «H mom m.u0H:

 

mcoHuumuHo Hmnu0>

107

.N
.H 00
.N
.H an
.N
.H mm
.N
.H nm
.N
.H mm
.N
.H mm
.N
.H 0m
.N
.H mm
.H «m
.H Hm

 

00059 05050 50055z‘0050m00m 0500 05mw5m H0558

%03

>50

x00

00%

0’3

5...

«50

>55

0500

00%

m

m

.3002
%H055H0000 05.50% 050 50% «H 5%03 .%00 50% 055:
:.00505 505550 0050 5005 .55 %00

00 50% «H ~055w5« 0505 50« 0505 005 50080805 50% 00:
:.050550H5505w500 .050 50% «H .550 %00 50% 055:
:.00505 505550 0055 500B .55 %00

.0805 0.55 50080805 50% «H .050 0500 5005050 0.0500:
.555 500 8.. 30 .5... 2.5 .5 525..
:.00505 505550 0050 5005 .0505 055

%00 .00 50% «H 0050 %00555 0505 50080805 50% 00:
:.50w55 %H055H0000 05.50% .00% 0500 50% «H .005:
:.00505 505550 0050 5005 050

55 %0m .05555« 0505 «0 0505 005 50050805 50% 500:
:.503050 50055 005 0.5009 .053_%00 50% 055 .033:
:.00H£u
505550 0050 5005 050 0505 055 %00 .050 0505

50080805 50% «H .%00555 w55550w %HH005 05.%00H:
:.50m55 %H055H0000 05.50% .n00 0500 50% «H:
:.00505 505550 0050 5005 .55

%0m ~0505 055 50080805 50% on .050 0505 55000 30::
:.00 0000
5 $0055 55 500 .55 30 .03 0 0.5005 .00...
:.00505 505550 0050 5005 .55 %00 500 50% «H .050

0500 0 0.55 .05505« 0505 «0 0805 005 50050505 50% 00:
.50055 050 50% 50 50% «H ~>55.%00 50% 050:
:.50m55 05.50%
«5 000 05 00505 505550 0050 5009 .55 %00 500

50% «H «055w5« 0505 «0 0505 005 50050805 50% 500:
:.§

%00 5500 005 «0 505500 005 000 .500 0 05 050B:

=.005 500 50.5 005 mo 005 as» «00 .005 0 .5 .505:

 

0505500550 50050>

108

.N

.H mm
.N

.H um
.N

.H Hm
.N

.H on
.N

.H 00
.H m0
.H 00
.H 00
.H m0
.H 00
.H m0
.H N0
.H H0

00%

«50

>55

033

500
500
00%

033
52.

«50
>55
%03

>50

:.50w55 05.50% .00% .0500 50% «H

now

5005

.55

.005 a 05 5005:

:.00505 505550 0050 5005 .0505

055 %00 ~050«00.505« 050 0505 50050505 50% 00:
:.055 50. no» 050 .055:
:.00505 505550 0050 5005 .55 %00

50% «H 0050 0505 HH00 03 5003 50050505 50% 00:
:~>58_%00 50% 050 .>05 0 0.5009:

:.00505 505550 0050

.55 %00 .500 50% «H .055w5« 0505 0505 50% 500:
:.053 %00 50% 050 .053:
:.00505 505550 0050 5005 .0505

%00 ~055w5« 0505 «0 0505 005 50050805 50% 500:
:.50w55 05.50% .050 50% «H «H00 %00 50% 050:
:.00505 505550 0050 5009 .55 %00

.50050505 50% «H ~055w5« 0505 «0 0505 005 0.5003:

:.500 %00 0500

005 «0 505500 005 000500 055H 005 00m .500 0 05 0509:
:.00% %00 ~55 000500 055H

0 000 00% 005 «0 005 005 300 000 .00% 0 05 0509:

:.053 %00

005 55500 0850 005 300 000 .053 0 05 0509:

:.5o0 500

505555500 050 0550 005 %03 00503 00m .n00 0 05 0509:

%00

50050 005 50 05 0550 009

:.«50 %0m .0050

.«50 0 05 0509 .«5M:

:.>55

050 05 0550 005 0050 00503 000 .>55 0 05 0509:

:.%03

%00 005555500 05 55 %03 00503 000 .%03 0 05 0509:

%0m

 

 

00059 05000 500552 00505000 0500 055050 H0559

:.550

~50 05 500 005 0050 00503 000 .>50 0 05 0509:

 

WGOHNUGHHQ HQQHN>

109

.N

.5 00
.N

.5 no
.5 00
.5 m0
.5 N0
.5 50
.5 oo
.5 mm
.5 mm
.5 0n
.N

.5 0m
.N

.5 mm
.N

.5 0m

%03
>50
500
00%
«50

>55
053

500
%03
>50

%03

>50

500

:555 50050505 .5.A 050 .503 05 050050 .505:

:.00505 505550 0050 5005 050 050 0505 «0 0505 005 000:
:.00055 0.525 5>00 500 =05 050:
:.00505 505550 0050 5005 .0505 5005 55 000

.00 500 «H .050 0505 «0 0505 005 50050505 505 500:
:.500 000

.500 005 «0 505500 005 05 55500 .500 0 05 0509:
:.000 000

.005 005 0° 005 000 cu 505o0 .005 0 m5 0505 .005:

:.055 500 .0500 005 cu 055o0 .055 0 m5 0505:

=.>5e 500 .0500 005 cu 555cm .>55 0 m5 0505 .>5z=
:.053 000

~05555500 050 0550 005 003 00503 000 .053 0 05 0509:
: . HOD %dm wwfiflufiHOQ

050 0550 005 003 00503 000 .000 0 0.5009 .500:

:.%03 000 :05555500 055 003 00503 000 .003 0 0.5009:
:.>50 %00

~50 05 500 005 0050 00503 000 .>50 05 5009 .>50:
:~50m55 55 500 500 050 .003 0 0.5009 .003:
:.00505 505550

0050 5005 050 0505 055 000 .050 0500 5005050 0.0500:
:~>50 000 500 050 .>50 0 05 5009:
:.00H5u
505550 0050 5005 050 0505 0.55 000 ~050 0505

«0 0505 005 50050505 500 on ~550w0 500 00555 H 500:
~500 050 0505 5500 03 50050505 500 050:
:.00505 505550 0050 5009 .0505 055 000

5.50>00 500 «H .050 0505 50550050« 0>.500 500 559:

 

00059 05050 500552 00505000 0500 055050 50559

 

0505500555 50050>

110

.m
.H mm
.N
.H as
.N
.H on
.N
.H mm
.N
.H we
.N
.H no

ku

max

>HE

.uswqu %Housaomnm ou.50% ass wamm :0» MH .033:

m:.mousu nouusn swam none .ousmam mfinu mo «Ema ecu hum:

zafion mum so» was .uoa a“ «so many:

m:.omu:u :ouuap gang cone .muswwm mafia mo «awn «nu hum:

w

zwuswfiu a“ new no» age .xun:
:.mounu couusn Swan amnu .uw hum

.ov so» MH .musmqm mafia mo «and «nu umnfiufiou do» on:
zwcum ham :0» can .uuw:
:.mounu nauuan

Swan cunu .uawa mufi hum Nona mHSu umnauamu no» on:
:.unwau
mHou=H0mnw mu.=om .mwx vaum :05 «H .mwx m m.uusa:
:.00H30

aouuan swam nonu .oaua mug mum .uao vuun mgnu %uH:
:.unmfiu m.uuna ~>aa.mmm so» can:
:.uounu nouusn Sash nose

.uauc m.u« mum ow 30% MH Nona manu nuaauauu no» on:

 

 

xowuH owwa< Honesz moaondom cuwo ouawam awHuH

mcoauumuaa amnum>

APPENDIX E

POST-TEST SHEETS AND PRE-TEST SHEETS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Nx A Nun»
K VA A %LXV N M \f
WVYV%K<WV
WM Xfl )7 Y3 A 4m
N Y A /F VA mm A?

 

MATCH 'ROM MEMORY
NA‘flZ

LA /\ m \/

 

 

 

 

D >4. K) E

/\ /\ \/ /\

 

 

 

APPENDIX F

GROUP MEANS OF RAW SCORES, RAW SCORES, AND

WITHIN-GROUP SAMPLE CORRELATION MATRIX

GROUP MEANS OF RAW SCORES, RAW SCORES, AND

APPENDIX F

WITHIN-GROUP SAMPLE CORRELATION MATRIX

Raw Group Means

 

T

T

Total Yb Variance

 

 

 

 

1 2 3
X' VM Pre-test 3.5000 3.2647 3.5588 3.4411 2.0353
2' VM Post-test 3.7941 3.8823 5.3529 4.3431 1.8874
‘i MTF Pre-test 4.1176 4.3235 4.0588 4.1666 3.5641
8' MTF Post—test 4.9705 5.5294 6.5294 5.6764 2.6253
3? Naming . 0.7058 1.0588 1.4411 1.0686 1.1648
within-Group Sample Correlation Matrix
on the Two Covariates and the
Three Dependent Variables
1 2 3 4 5
Naming VM Pre VM Post MTF Pre MTF Post
1 Naming 1.000 -- -- -- --
2 VM Pre-test 0.103 1.000 -- -- --
3 VM Post-test 0.345 0.535 1.000 -- --
4 MTF Pre-test 0.343 0.357 0.446 1.000 ~-
5 MTF Post-test 0.370 0.287 0.502 0.692 1.000

 

113

114

PARTICIPANT TEST DATA

No Color

 

 

 

 

 

 

 

 

 

 

 

 

530 mu: 0 , 3 1 1 1 0 0 O 2 1.. 0 2 2 0 2 d. 1.. 1 0
umoh
numom PE 4 5 6 6 7 7 4 4 5 4 3 5 5 0 5 4 3 6 5
uwououm PH: 4 2 5 5 6 7 4 4 3 5 3 3 3 0 3 5 4 5 5
Show E2 1 1 0 0 1 2 1“. 1. O 0 1.. 1.. 0 2 1 an. 1... 5 2
T
quA
numom 2h: 4 3 6 4 4 6 2 3 2 4 3 3 2 3 3 4 3 6 4
umuuoum E2 5 4 6 4 3 4 3 2 2 4 4 4 2 5 2 6 4 1 2
wan—"82+ O 0 0 0 1 1 0 1 1 O 2 0 1 0 1 0 0 2 2
Hon—Ed 1 2 3 4 5 6 7 8 9 O 1 2 3 4 5 6 7 8 9
32569...— 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1
cowuwfll F F F F F A F F F F F F F F F F F F C C
acme—amount: 1. 1. 1 1 1 1 1 1 1 1.. 1 1 1 1 1 1. 1. 1 1.

 

 

.115

No Color (continued)

 

dado m

 

unou
-uuom m

 

7

“.6398 EL

 

£8 E:
T

 

 

unu
.2. a.“

 

3395 FE

 

 

wnaeaz#

0

2

0

O

2

2

 

none:
u.unov=u

 

nowumoafl

 

 

noun—uncha—

1 C 2 0

1 C 2 1

1 C 2 2

1 C 2 3

1 C 2 4

1 C 2 5

1 C 2 6

1 C 2 7

1 C 2 8

1 C 2 9

1 C 3 0

1 C 3 1

1 C 3 2

1 C 3 3

1 C 3 4

 

Maximum Color

 

 

 

 

 

 

0

1

0

3

 

 

 

 

2 F 0 1

2 F O 2

28.03

2 F 0 4

2 F 0 5

 

116

Maximum Color (continued)

 

 

umuu

 

 

 

Hum“

 

 

1

4

O

0

1

1

2

0

0

0

1

2

0

2

3

 

Hogabz
m.u:uvoum

 

nodumuon

 

 

 

uaoeuwmma

2 F 0 6

2 F 0 7

2 F 0 8

2109

2 F 1 0

2811'

2 F 1 2

2 F 1 3

2 F 1 4

2 F 1 5

2 F 1 6

2 F 1 7

2 C 1 8

2 C 1 9

2 C 2 0

2 C 2 1

2 C 2 2

2 C 2 3

2 C 2 4

2 C 2 5

2 C 2 6

2 C 2 7

 

 

117

Maximum Color (continued)

 

£8 E:

 

ummu
numom .mHZ

 

umoumum has

 

58 E:

 

ummu
.88.. E2

 

$30.5 2.9

 

wzwamz

2

0

3

0

 

nunEbz
m.ucovaun

 

nowumooq

 

 

unusumuua

 

2 C 2 8

2029

2 C 3 0

2 C 3 l

2 C 3 2

2 C 3 3

2 C 3 4

 

Vanished Color

 

 

 

 

 

 

1

1

l

2

0

2

 

 

 

3 F 0 1

3 F 0 2

3 F 0 3

3 F 0 4

3 F 0 S

3 F 0 6

3 F 0 7

3 F 0 8

3 F 0 9

3 F 1 0

3 F 1 l

3 F 1 2

3 F 1 3

 

 

118

Vanished Color (continued)

 

 

flwwo he:
umou
uumom MHZ

 

z umuuoum my:

 

nwuo 2m:

 

noun
1 umom 2m:

 

umouuum 2m:

 

_ wcHEmz

1

5

0

1

0

0

3

2

2

1

4

1

0

0

2

2

1

 

Honadz
m.unmvoum

 

 

 

= acqumoog

3 F 1 4

3 F 1 5

3 F 1 6

3 F 1 7

3 C 1 8

3 C 1 9

3 C 2 0

3 C 2 1

3 C 2 2

3 C 2 3

3 C 2 4

B C 2 5

3 C 2 6

3 C 2 7

B C 2 8

P029

C 3 0

8&31

3 C 3 2

3 C 3 3

Haas.

 

= unoauuoue

 

P

 

 

APPENDIX G

ILLUSTRATION OF AUDIO FLASHCARD READER

APPENDIX G

ILLUSTRATION OF AUDIO FLASHCARD READER

Card Holder Frame
Talk Control

  
  
  
 

Card Slot
Track 1 Control
End 0! Track Flag
Track 2 Control __
Track 3 Control

Track 4 Contra

Master Record Receptacle \

(Turn Master Key Clockwise

And Hold To Record)

Talk Control Lock Receptacle

(Turn Master Key Clockwise To Lock) Volume Con tro I
Receptacles (color coded)

For Model 305 Headset

Model 101M Audio Flashcard Reader

   

lllilllllljljijlllfllljllllIljlljllI