w .. LIBRARY

M ' gun Sistc
University

 

This is to certify that the

thesis entitled

THE EFFECTS OF SCHOOLING

0N EQUIVALENCE GROUPING PREFERENCES
0F TANGALE ADULTS

presented by

James Edwin Plueddemann

has been accepted towards fulfillment
of the requirements for

Ph.D . Secondary Education

degree 1n

 

and Curriculum

3Q

Major professor

 

 

Date June 2, 1978

0-7 639

 

 

 

 

 

 

THE EFFECTS OF SCHOOLING
ON EQUIVALENCE GROUPING PREFERENCES

OF TANGALE ADULTS

by

James Edwin Plueddemann

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Secondary Education and Curriculum

1978

ABSTRACT

EQUIVALENCE GROUPING PREFERENCES
OF TANGALE ADULTS

’ {iCa-Rw
C \QL‘ THE EFFECTS OF SCHOOLING ON
/\

by

James Edwin Plueddemann

The study investigated the effects of schooling
on the way adult males of northeastern Nigeria prefer to
sort dissimilar stimuli into equivalence groupings.

As western patterns of education spread throughout
Africa it is important for educators to be aware of the
thinking—process preferences of the learner. It is likely
that a student will learn more easily if the structures
of the curriculum are similar to the student's preferred
and normal manner of structuring dissimilar stimuli.

The data are also useful for continuing development
of a theory for causes of cognitive growth in adults.

Previous studies showed differences among cultures
in both the rate and the limits of cognitive development.
The amount of schooling was often seen to be related to
differences in cognitive functioning. The studies did
not assume that the differences necessarily indicated cog-
nitive deficits. Differences may indicate appropriate
responses to different cultural needs.

Evidence seems to indicate that as children in the

western world grow, they shift from using complementary,

James Edwin Plueddemann

thematic and relational criteria to using similar, analytic
and superordinate criteria in equivalence grouping tasks.
Studies of older adults indicate that once they are

freed from the pressures of schooling and careers they

do not lose the capacity for using superordinate rules.
However, they prefer to use the more natural relational

and complementary strategies.

One way to discover how adults prefer to sort dis—
similar stimuli into equivalence groupings is to ask them
to sort different pictures into groups of pictures that
are related. A picture-grouping instrument was constructed
and after field-testing, was administered to 130 Tangale
males near Biliri in northeastern Nigeria. The sample
was composed of illiterate farmers, pastors and teachers.

Independent variables included the amount of schooling,
the recency of the schooling, the amount of non-formal
education and occupation.

Dependent variables included the number of super-
ordinate equivalence rules used, the number of internal
attributes used (i.e., function, material or being
attributes), and the number of impersonal linguistic
structures used.

Results of one-way analysis of variance showed that
schooling affected the number of superordinate grouping
strategies used, but the amount of schooling did not account

for a difference. Pastors and teachers tended to prefer

James Edwin Plueddemann

superordinate groupings more than did illiterate farmers.
The results did not show a relationship between the recency
of schooling and the amount of non-formal education for

any of the dependent variables.

Fulani language ability was found to be positively
related to superordinate rules, external attributes and
impersonal linguistic structures. Also, the number of
languages spoken correlated positively with superordinate
rules and impersonal linguistic structures.

It was concluded that schooling does affect the pref-
erence for using superordinate rules. However, the amount
of schooling, the recency of schooling, and the amount of
non-formal education does not affect a preference for use
of superordinate rules, internal attributes or impersonal
linguistic structures. Teachers and pastors preferred
superordinate rules more than did illiterate farmers.
Language ability, especially the ability to speak Fulani,
correlated positively with the number of superordinate

rules used.

To the Church
in Africa
with whom and for whom

this study was undertaken

ii

ACKNOWLEDGMENTS

Many people and groups have helped to make this
research project possible. i

I am grateful to the leaders of the association of
Evangelical Churches of West Africa. Rev. Simon Ibrahim
and Rev. David Olusiyi have given kind and enthusiastic
cooperation. The Sudan Interior Mission has been most
generous in giving permission to take two extended fur-
loughs for study purposes and has allowed me to take time
while in Nigeria to gather data. Rev. Harold Fuller,

Dr. Paul Bowers, and Mr. Larry Dick were especially helpful
in their insights and encouragement.

Mr. Samuila Kure and Mr. Toma Dantaro did an excellent
job of data-gathering. They deserve special credit for
their persistence in searching out difficult-to-find sub-
jects. They travelled many miles on their motorcycles over
muddy "bush” paths. Thanks also to Mr. Sani Bature for
his help with data analysis.

I appreciate the helpfulness of the students at Igbaja
Seminary, Kagoro Bible College, Tofa Bible Training School
and Zabolo Advanced Bible School for their part in pilot

studies for the project. Thanks are due to the 130 Tangale

iii

men who willingly gave of their time to make the data-
gathering possible.

Grateful thanks go to Miss Barbara Forster who drew
the pictures for the research instrument. She willingly
took time to draw and re-draw many pictures.

Fellow—students have provided stimulating interaction
and have helped with inter-rater reliability tests. Thank
you, Ed Francis, Dorothy Tweddell, Kathy Wilson, Narendra
John and Bob Drovdahl. Geneva Speas has had a greater
influence than she realizes. Sincere thanks also to Dr.
Donald Miller for his helpful contributions and
encouragement.

Carol Wilson cheerfully helped to edit the manuscript
and Dr. George Sargent gave valuable advice on data analysis.
Carol Cole has shown special concern in the typing of the
manuscript.

I am grateful for the encouragement of my parents,
Dr. and Mrs. Ed Plueddemann, and my parents—in-law, Rev.
and Mrs. Bob Savage.

Sincere appreciation is given to the members of my
guidance committee, Dr. Bill Herzog, Dr. Dick Gardner,

Dr. Walter Scott and Dr. Ted Ward. Ted Ward deserves
special thanks, not only for his guidance in the educational
experience, but also for his concern for the individual
growth of his students. Ted has been academic advisor,

committee chairman, dissertation director and friend.

iv

Life has never been the same since our paths crossed in
Nigeria in 1970!

Finally, my children, Shari and Danny, have been
patient during the long evenings of study away from home
and my wife has given continual encouragement and help.

Thank you, Carol.

Chapter

1.

2.

3.

IDENTIFYING THE PROBLEM

Purpose of the Study

Importance

TABLE OF CONTENTS

Definition of Important Terms .

Research Questions and Hypotheses

Assumptions .

Limitations .

Generalizability

Overview

REVIEW OF THE LITERATURE

Historical Perspective

Culture
Culture
Picture

Summary .

RESEARCH METHODOLOGY

Description of Methodology

Design over Time

Sorting Tasks .

Design over Variables .

Hypotheses

vi

and Equivalence Grouping

and Cognitive Development

Page

11
l3
14
14
15
17
17
22
32
34
39
41
41
42
42
45

Chapter

3.
Population . . . . . . .
Sample . . . . . . . . .
Sampling Procedures . . .
Instrumentation . . . .
Instrument Development .
Hausa Translation . . . .
Hausa Field Testing . . .

Instrument Reproduction .

Training of Research Assistants

Administration Procedures .

Scoring Procedures . . .
Validity Concerns . . . .
Reliability Concerns . .
Data Analysis . . . . . .
Summary . . . . . . . . .
4. FINDINGS . . . . . . . . .
The Research Question . .
Hypothesis Testing . . .
Hypothesis 1 . . . . .
Hypothesis 2 . . . . .
Hypothesis 3 . . . . .
Hypothesis 4 . . . . .
Hypothesis 5 . . . . .

Hypothesis 6 . . . . .

vii

Page

47
48
49
49
49
51
52
52
53
54
56
57
62
63
63
66
66
67
68
69
70
7O
71
72

Chapter

4.

5.

Hypothesis 7 . . . . . . . .
Hypothesis 8 . . . . . . . .
Hypothesis 9 . . . . . . . .
Hypothesis lO . . . . . . . .
Hypothesis ll . . . . . . . .
Hypothesis 12 . . . . . . . .
Observations . . . . . . . . .

Effects of the Fulani Language

Effects of Occupation on the Number of

Pictures 0 O O O O O O I 0

Effect of the Number of Languages Spoken .

A Superordinate and Internal Variable . . .

Effects of Age . . . . . . .
Effects of Literacy Classes .
Effects of TEE . . . . . . .

The Relationship of Schooling
Education . . . . . . . . .

Summary . . . . . . . . . . . .
SUMMARY AND CONCLUSIONS . . . . .
Summary . . . . . . . . . . . .
Discussion of the Findings . .

Conclusions . . . . . . . . . .

viii

to Non-Formal

Page

73
74
75
76
77
77
80
80

80
81
81
82
82
82

82
84
87
87
91
98

APPENDIX A DATA COLLECTION FORMS
APPENDIX B PICTURES USED IN THE RESEARCH INSTRUMENT

APPENDIX C EXAMPLES OF RESPONSES

BIBLIOGRAPHY

ix

Page
104
109
110

120

LIST OF TABLES

Table
4.1 Means for Amount of Schooling and Superordinate
RUleS O O O O I O O O O O O O O O O O O O O O O
4.2 Means for the Amount of Schooling and Internal
Attributes O O O O O O O O O O O O O O O O O O
4.3 Means for the Amount of Schooling and Impersonal
Pronouns O O O O O O O O O O O O O O O O O O O
4.4 Means for the Amount of Schooling and Super-
ordinate Rules . . . . . . . . . . . . . . . .
4.5 Means for Recency of Schooling and Internal
Attributes O O O O O O O O O O O O O O O O O O
4.6 Means for the Recency of Schooling and Impersonal
Pronouns O O O O O O O O O O O O O O O O O O O O
4.7 Means for the Amount of NFE and Superordinate
RUleS O O O O O O O O O O O O O O O O O O O O O O
4.8 Means for the Amount of NFE and Internal
Attributes O O O O O O O O O O O O I O O O O O O
4.9 Means for the Amount of NFE and Impersonal
Pronouns O O O O O O O O O O O O O O O O O O O O
4.10 Means for Occupation and Internal Attributes . .
4.11 Means for Occupation and Internal Attributes . .
4.12 Means for Occupation and Impersonal Pronouns . .
4.13 Analysis of Variance . . . . . . . . . . . . . .
4.14 Summary of Main Findings . . . . . . . . . . . .

Page

68

69

70

71

72

73

74

75

75
76
77
78
79
85

LIST OF FIGURES

Figure Page

4.1 Superordinate Attributes by Amount of Schooling 83

xi

Chapter 1

IDENTIFYING THE PROBLEM

As western patterns of education spread throughout
the world, differences between the thought processes of
western and non-western people become an important study
for educators.

One way of studying the thinking process is to inves-
tigate how people group the thousands of stimuli with which
their minds are continuously bombarded. These stimuli
provide more information than any mind could possibly pro-
cess. Grouping strategies select the most relevant stimuli,
preventing the individual from becoming overwhelmed by the
diversity in the environment. Such strategies for grouping
relevant stimuli seem to be a learned achievement, and
they may be expected to vary with internal growth and with
environmental changes. The process by which people group
different stimuli into meaningful categories is the focus

of the study.

Purpose of the Study

 

The purpose of the study is to investigate how schooling
affects the way adults prefer to group dissimilar stimuli
into equivalence groupings. Specifically, the study tests
hypotheses about the relationship between the amount and
recency of schooling and equivalence grouping preferences
as measured by a picture-sorting task administered to Tangale
men of northeastern Nigeria. Other hypotheses about the
relationship of occupation and non—formal education to

equivalence grouping preferences are also tested.

Importance of the Study

 

The study is important in three ways: 1. it has
practical implications for the structure and content of
adult education in non-technical societies; 2. it is a
continuation of important research methodology by Bruner,
Olver and Hornsby (1966), and fits into a growing body
of cross-cultural research pertaining to equivalence group-
ings; and 3. it has significant theoretical implications,
because it fills a need for data regarding the cognitive

development of adults in a non-western society.

Practical Implications. The entry behavior of the
student affects every other curriculum consideration:
subject matter, type of instruction, place in the instruc-

tional sequence and the objectives of the instruction

3

(Schulman, 1970). Equivalence grouping preferences are

an important aspect of the entry behavior of the learner.
They indicate how the learner normally organizes and inter-
prets stimuli in the real world. Thus it is quite possible
that a student would learn better if the structure of the
subject matter were similar to the way he normally organizes
stimuli. For example, if a person usually groups objects
using complementary criteria (e.g., grouping things that
are normally found together), he may do better in a course
(such as history) that emphasizes complementary criteria
than in a course (e.g., geometry) which emphasizes super—
ordinate criteria.

Understanding equivalence grouping preferences of
adults is especially relevant for Nigeria. Nigerian adults
in almost every profession sense the need to be ”upgraded“
educationally. Twenty years ago, a village pastor may
have felt adequately prepared with only four years of formal
training. Today he feels threatened and isolated by young
people who have more formal education than he has. The
pastor is probably unable to discuss world affairs, and
he may not be able to speak English, the official language,
with the young people in his church. He therefore desires
to continue his education. School teachers, factory workers
and government employees have similar needs. The desire
for continuing education in Nigeria is even greater than
it is in many Western countries, where educational standards

are not rising as rapidly.

4

There is little evidence that equivalence grouping
preferences of Nigerian adults are the same as those of
American adults. Thus a study that describes equivalence
grouping preferences of Nigerian adults is important to

curriculum designers in Nigeria.

Methodological Implications. Bruner and his asso-

 

ciates (1966) developed instruments using pictures, words
and objects to measure the development of enactive, ikonic,
and symbolic representations of reality in children and
adolescents. They used similar instruments in cross-
cultural studies. Since the present study uses instrumen-
tation and procedures similar to those described in Olver
and Hornsby (1966), it contributes to the process of

construct validation.

Theoretical Implications. This section considers
the theoretical relationships between cognitive development,
cognitive preference and the environment, including the
school. The long—term effects of the school on cognitive
preferences are discussed.

Cognitive preferences are closely tied with environ-
mental needs. That is, people prefer the cognitive struc-
tures that are most useful for their specific environment.

Theoretical assumptions indicate that cognitive development

 

is also a result of environmental pressures. Cole (1971)
writes that people tend to develop those cognitive abilities

which are most useful in a particular culture. Thus both

cognitive preferences and cognitive abilities are directly
influenced by cultural and environmental pressures.

Studies done by Bruner and his associates assume that
differences in equivalence formations in a picture-sorting
task measure various levels of cognitive development.

The present study measures cognitive preferences, not
cognitive competencies. Yet the presence of a specific
cognitive preference would indicate a specific level of
cognitive development. In contrast, the lack of a specific
preference does not prove that the cognitive development

is also lacking. However, because of the relationship

of environmental pressures to both cognitive development
and cognitive preferences, it is likely that there is a
close interface between the two.

Schooling, as part of the environment, has an influence
on cognitive preferences. Previous studies have shown
that schooling, while not essential, facilitates the
development of concrete operational thinking. On the other
hand, schooling is necessary, though not sufficient, for
the development of formal operations (Laurendeau, 1977).

It is appropriate to ask if schooling produces permanent
changes in internal mental structures, or if it merely
teaches the skills necessary to pass Piaget-type tests.
There may be little long-term change or transfer of actual
thinking—process preferences. The school is a powerful
environment and seems to have a strong effect on cognitive

development. But if the school environment is greatly

6

different from the out-of—school environment, the long-
term effects of schooling may 'wear off,“ particularly
if the school is facilitating the development of specific
cognitive abilities that are not as useful for normal,
out-of—school thinking needs. Possibly the long-term
effects of schooling on cognition decrease in a non-
technological society, where there is a greater difference
between the school and out-of—school environments.

In summary, the study has important implications for
designing curriculum for adult education, for continuing
the construct validation of the methodology, and for finding
answers to theoretical questions concerning the relationship

of schooling to cognitive development.

Definition of Important Terms

 

Terms to be defined include dependent variables, inde-

pendent variables, and other terms used in the study.

Dependent Variables. The following outline shows
the relationship between the categories and sub-categories.
1. Equivalence rule
a. Superordinate
b. Complexive
2. Equivalence attributes
a. External attributes
b. Internal attributes
c. Nominal attributes

d. Subjective attributes

3. Person indicator
a. Personal
b. Impersonal

The equivalence rule is the logical reasoning derived

 

from the verbal explanation of the subject as to why the
pictures in a group are alike. It is the grouping structure.
Two general classes of equivalence rules or structures

are analyzed: superordinate and complexive.

Superordinate groupings are made on the basis of a

 

common feature or attribute that is equally appropriate
for each picture in the group, (e.g., ”The shirt and the
tree are both green").

Superordinate groups can be either general superordinate
or itemized superordinate. General superordinate construc-
tion consists of ”stating a common characteristic of the
items in the group” (Bruner, 1966, p. 75) (e.g., “The cow
and the goat are animals").

Itemized superordinate structures have a generalized
property joining the pictures together and explicitly stated
in an itemized manner (e.g., "We get meat from the cow
and we get meat from the goat").

Complexive groupings are made on the basis of attributes
that are individual to each object. There is no one general
rule or universal attribute that is common to all the
objects in the group (e.g., "The shirt and the tree are

green, and the tree and the cactus are living things").

8

There are other subheadings in the complexive rule
structure, but they are not separately analyzed in this
study. Instead, this study groups together under complexive
structures several subheadings based on the terms in the
Olver and Hornby study (1966): hyperordinate, edge matching,
key ring, association, collection, multiple grouping and
thematic.

Complexive rule structures make equivalence groupings
either by grouping several superordinate structures together
in one sort or by grouping pictures that complement each
other in the real world (e.g., table and chair).

Equivalence attributes are the actual attributes which
the men in the study saw when they made the grouping.

If the equivalence rule is the structure, then the equiva-
lence attribute is the content of the sort. The equivalence
attributes are divided into four general categories:
external, internal, nominal and subjective.

External attributes are the characteristics that are
perceptible in the picture itself. These are the immediate,
non—interpretive qualities such as color or shape (e.g.,
“The clock and the coin are round” or "The bicycle and
the teapot are both black").

Internal attributes are characteristics that are some-
what more interpretive. These include functional properties
(e.g., ”The banana and the corn are both good to eat“),

materials (e.g., “The lantern and the hoe are both made

9

of metal”), or being (e.g., "The chicken and the snake
both breathe“).

Nominal attributes are linguistic characteristics

 

of the objects (e.g., ”The taxi and the truck are both

vehicles”). Pictures are grouped according to a name or

a word which the objects have in common in that language.
Subjective attributes are characteristics that express

the subject's personal interpretations of the objects.

The subcategory includes affective interpretations (e.g.,

 

”I don't like snakes and cigarettes"), fiat interpretations
(the objects are alike because the subject says they are),

or symbolic interpretations (e.g., ”The cross and the candle

 

are the same because Jesus is the light of the world”).

The third dependent variable is the person indicator

 

in the linguistic structure of the response. When the
subject uses the pronouns ”I," ”we," or ”you,” the grouping

is scored as a personal sort (e.g., ”I ride my bicycle

 

to the market to buy a new shirt"). A sort is impersonal
when the subject uses impersonal pronouns (e.g., with
reference to a hoe and a cow, ”They are useful in farming").
The impersonal sort looks at properties that are intrinsic
to the object (e.g., ”They are...‘I or ”It is...'). The
personal sort looks at properties extrinsic to the object
(e.g., ”You eat them").

The three dependent variables of rule, attribute,

 

and person indicator are considered for each sort. For

example, if the subject chose pictures of a hoe, corn and

10

a cock and then explained, "The cock wakes me up in the
morning and I pick up my hoe and farm corn,“ the rule would
be complexive, the attribute would be internal, and the

person indicator would be personal.

Independent Variables. Independent variables which

 

require specific definition in this study are ”schooling”
and "non-formal learning."

Schooling in this study refers to formal schooling,

 

usually based on the western colonial system. Informal
and non-formal learning are not called "schooling" in this
study.

Non-formal education is sometimes defined as a planned
and specific kind of learning (as distinguished from informal
learning) which takes place outside of the formal school
system. But to merely define non-formal learning in terms
of its difference from other forms of education is of little
use where an operationalized variable is needed for specific
investigation. Therefore, for this study, non-formal
education has been given an arbitrary but appropriate
definition. It is a variable created from:

1. The number of books owned
. Experience in theological education by extension

Experience in literacy classes

Radio ownership

The number of languages spoken

m U1 :5 to N
O

. The number of towns visited (from a specific list)

11

Other Terms. Cognitive development is defined from

 

the conceptual framework of Piaget.

Cognitive preference is not intended as a measure
of cognitive development. Cognitive preference is defined
as the way the subject actually makes a specific picture
sort. The fact that a subject does not make superordinate
groupings does not necessarily mean that he would not be
able to do so if thus instructed. It is likely that all
the subjects could use superordinate structures if they
were given enough guidance (e.g. “Sort out all the pictures
of animals"). Cognitive preference is probably closer
to the subjects' normal thinking practice, and thus more
important for curriculum design purposes than would be
a test of cognitive competencies.

Pastor is a technical term used in Nigeria to denote
hierarchical status. For this study, the term “pastor”
will be used to include anyone in charge of a local church
in the Tangale District Church Council (DCC). This includes
evangelists, unlicensed pastors, licensed pastors, and

ordained pastors.

Research Questions and Hypotheses

The purpose of this study is to inquire into the
relationship between schooling and equivalence grouping
preferences. Does schooling lead to changes in the thinking
habits of the student, and do these new thinking habits

continue after the student finishes school? Previous

12

studies have shown that schooling does make a difference
in how children make equivalence groupings (Greenfield,
1966). Studies also show that older adults in non-technolo-
gical occupations prefer to use more complementary groupings,
i.e., to group things that are normally used together in
everyday life (Denny, 1974).

The general hypothesis is that schooling encourages
people to use superordinate and hierarchical categories,
but these category preferences are not as useful as comple-
mentary or complexive groupings for everyday thinking needs;
thus, they tend to be replaced by more useful complexive
thinking habits soon after the person has finished school.
This is especially true in a rural, non-technological
society.

This general hypothesis, however, may be confounded
by conditions that in this study are labeled ”non-formal
education". Since little previous research has been done
on the effects of non-formal education, the present study
investigates the relationship between specific out-of—school
learning situations and equivalence grouping preferences.
Hypotheses have been generated from the findings. In order
to investigate the general hypothesis, the following specific
hypotheses were formulated:

First, the amount of schooling has an effect on the
equivalence grouping task. The more schooling a person
has, the more likely he will be to use superordinate struc-

tures, internal attributes and impersonal person indicators.

13

Second, the recency of school has an effect on the
equivalence grouping task. The less recent the schooling
experience, the fewer superordinate structures, internal
attributes and impersonal person indicators will be used.

Third, occupation has an effect on the grouping
strategy. Since school teachers are in a continued schooling
experience, they will have a greater tendency to use super-
ordinate structures, internal attributes and impersonal
person indicators.

Fourth, although there is little theoretical evidence
as to the effect of non-formal learning on grouping prefer-
ences, it was hypothesized that if non—formal education
does facilitate cognitive development, responses of persons
with a wide non-formal experience will tend to be similar
to those of the highly schooled. However, if schooling
in the classroom merely teaches skills which have temporary
usefulness, responses of the two groups are not expected

to be similar.

Assumptions
Two main assumptions influence this study. First,
the responses of the test subjects do have a relationship
to the way a Tangale adult would normally organize different
stimuli from the real world environment.
Second, it is important for those who design curriculum

for Nigerian adults to be aware of how these adults typically

14
make equivalence groupings in their out-of—school experi-
ences. If the preferences are taken into account, the
learning experience will not seem so "foreign.” Such an
awareness will facilitiate learning and will increase the

possibility for greater transfer of learning.

Limitations

There are several limitations to the study because
of its exploratory and non-experimental nature.

First, cause-and-effect relationships are explored
but are not proven.

Second, the construct validity of the instrument needs
further investigation. Presently, the instrument is relied
upon only to explore trends and possible relationships.

Third, since the instrument is not intended to measure
competencies, the relationship of the findings to cognitive
development theory is indirect and preliminary.

Fourth, since the study is cross-sectional, the

variables of age and generation are confounded.

Generalizability

While the purpose of the study does not necessitate
wide generalizability, the findings concerning illiterate
farmers and primary school teachers could be generalized
to other farmers and teachers who are male Tangale speakers
living in the Tangale tribal area. There is no reason

to believe that the pastors in the Tangale DCC are different

15

from pastors in the two adjacent DCC's who speak Tangale
as a first language.

The findings are not directly generalizable to females,
to non—Tangale people or to Tangale people with occupations
other than farming, teaching or pastoring.

Potential generalizability is increased as the study
relates to similar studies in other cultural settings and
as findings of other studies contribute to the building

of theory.

Overview

 

In this chapter, the research problem was stated.
Three general reasons for the importance of the study were
given with operational definitions of the variables.
Research questions, assumptions, limitations and generaliza-
bility were discussed.

In Chapter Two the review of significant literature
is presented under the two main headings of theoretical
literature and methodological literature. The theoretical
literature includes literature from the fields of anthro-
pology, linguistics and developmental psychology.

In Chapter Three the research methodology is discussed
under the two main headings of a general description of
the research structure and the research procedures used
in gathering the data.

In Chapter Four the findings are presented as they

relate to the specific hypotheses of the study.

16

In Chapter Five the findings are discussed, with

special attention given to implications and conclusions.

Chapter 2

REVI EW OF LI TERATURE

The purpose of the study is to investigate the effects
of schooling on equivalence grouping strategies of Tangale
adults. The broader issue includes the question of how
environmental and cultural factors influence cognitive
development.

The review of literature encompasses both theoretical
and methodological literature. Historical perspectives
on the issue of culture and thinking are reviewed briefly,
focusing on the fields of anthropology and linguistics.

A discussion of the general theoretical literature on culture
and cognitive development follows. The chapter closes
with a look at the more specific literature treating the

effects of culture on classification strategies.

Historical Perspective

Anthropologists have long been interested in cross-
cultural cognitive differences. In 1899 a study team went
to the Torres Straits Islands to conduct comparative studies

on the perception and memory of primitive people. Some

17

18

of the findings showed that the ”natives" had better sensory
functions than did their European counterparts, but had
less facility in abstract functions. In attempting to
explain the differences, the researchers suggested that
highly developed sensory functions focus predominant atten-
tion on concrete things, and may act as an obstacle to
higher abstract mental development (Rivers, 1901, p. 45).

The issue was, "Why do non-western people seem to think
differently from western people?” The debate centered
around causes of differences and similarities between human
beings in different cultural settings.

Boas (1911) wrote that the laws of mental activity
are the same for all people, but the manifestations of
the mind depend on individual experience. Both western
and non-western thought patterns are useful for understanding
the world. Both systems can find unity in a world of seeming
diversity, and both use an interplay of theory and common
sense.

Wundt (1912) wrote that though intellectual potential
is the same for all mankind, the intellectual capacities
of primitive people have remained on a low plane because
of isolation, few environmental pressures, and the limited
nature of their wants.

Levy-Bruhl, in his books, How Natives Think (1910)
and Primitive Mentality (1926), argued that both mental
preferences and mental competencies of primitive people

were different from those of people in more advanced

19

cultures. He gathered data by looking at ”collective
representations” or systems of belief among primitive
societies. He concluded that "primitives perceive nothing
in the same way as we do” (1926, p. 43). He further stated
that primitives are pre-logical, mystical and entwined

with emotional life. They are at a lower evolutionary
stage of cultural development. Boas (1911) disagreed with
Levy—Bruhl, both in the methodology and in the findings

of his study.

Levi-Strauss (1966) studied indigenous categorizing
systems of non-western societies, and concluded that dif-
ferences in systems were the result of different problem-
solving needs. He found that primitive test subjects used
both complexive and superordinate rules in classifying
animals, and that they often classified people into hier-
archical totem groups and sub-groups. He discovered that
primitive classification systems were based on character-
istics that were easily seen and experienced, in contrast
to the classification systems of modern science which focus
on necessary relationships in the structure of objects.
Levi-Strauss assumed a basic human unity, with differences
explained in terms of different needs in problem solving.

Linguists are interested in the effects of language
on thought patterns. All cultures use language, but there
are vast differences in semantics and syntax. Sapir and
Whorf (1956) concluded that the semantics of a language

determines the structure of thought. Whorf wrote that

20
the Hopi language "shows a much higher plane of thinking"
(p, 84), thus implying that Hopi-speaking people are "better"
thinkers because they have a "better" language.

Fishman (1960) wrote that language not only reflects
the soul and mind of a people, it shapes the mind and soul.
To the extent that a society's "world” is built by its
language habits, people with different languages live in
different worlds.

A great deal of empirical research has been generated
by the hypothesis that language shapes the mind and soul
of people. One frequently used type of research is color
recognition. The hypotheses of Sapir and Whorf suggest
that people who speak languages with few color words would
have difficulty differentiating or remembering certain
colors. However, most studies on the influence of language
on color recognition reject this hypothesis. American
Indians are able to distinguish slight color differences
even when they do not have the color name in their language
(Hoijer, 1956). Greenfield (1966) found that Wolof children
tended to classify objects by color more often than by
other attributes, even though they used few color words
and had to borrow words from the French language.

Harris (1966) found that Zulu children were able to
make fine color discriminations, despite the imprecise
nature of color words in the Zulu language. But the discri-
minations were often made through associations rather than

by name. A color card would be labeled as "the color of

21

my father's hut” or "the color of my uncle's largest cow"
(p, 12). Harris concluded that color classification was
made according to cultural demands and cultural needs,
and was not limited by linguistic deficiencies.

Maclay (1958) found that the presence or absence of
linguistic categories was not a reliable predictor of non-
linguistic behavior among the Navaho.

Brown (1956) concluded that language neither limits
nor promotes the availability of possible concepts, but
linguistic concept names are tied to the needs of the
specific culture. Language is “an inventory of all the
ideas, interests and occupations that take up the attention
of the community" (p. 311). Language is related to the
preferences called for by cultural needs and interests:
it is not related to abilities.

Hall (1976) called for further study on the question
raised by Whorf. He argued that language should be redefined
to include non-verbal communication, since verbal language
is limited by its linear structure, making it difficult
to predict its effects on complex thinking.

Studies have shown that language as used in reading
and writing appears to be powerful in shaping cognitive
development (Bruner, 1966b). Whereas spoken language tends
to be context-dependent (i.e., the things one talks about
are usually in the immediate situation), the language of

reading and writing is independent of the immediate context

(i.e., words are separated from their object). Schooling

22

may encourage one to use a word independently of its
immediate context and may thus lead to the development

of new cognitive structures. Symbolic modes of representa-
tion replace ikonic modes when language becomes independent
of the immediate context.

In summary, anthropologists and linguists have studied
the nature of cognitive similarities and differences between
people of different cultures. The question is, ”Are some
cultures lower on an evolutionary scale, or are cultures
equally adapted to meet the differing cognitive problems
of their different environments?" If cognitive growth
is simply a biological unfolding, one would expect little
cross-cultural variation. But if cognitive growth is
influenced only by cultural experiences, one would expect
little cross-cultural similarity (Berry, 1974, p. 12).

The present study contributes to the subject of

culture's role in cognitive development.

Culture and Cognitive Development

In the following section, a review of the significant
literature pertaining to the theory of culture and cognitive
development is discussed. The study emphasizes the theoret-
ical contributions of Piaget (1966), Bruner (1966), Witkin
(1967) and Cole (1975).

Piaget stressed the need for and significance of cross-
cultural studies to test his theories of cognitive develop—

ment (1966). According to Piaget, four important factors

23

affect cognitive development: 1. biological or epigenetic
systems, 2. equilibration or autoregulation factors,
3. social and interpersonal factors and 4. educational
and cultural factors. It is impossible to test Piaget's
theory of development without cross-cultural studies.
In 1966, when Piaget summarized the findings of cross-
cultural studies, he concluded that the order of stages
in cognitive development was the same in all cultures and
was thus biologically controlled. The rate of development
appeared to be dependent on social, educational or cultural
factors. He commented that there was a need for cross-
cultural studies which would measure formal operational
thinking, since in many cultures there does not seem to
be much progress beyond the level of concrete operations.

Dasen (1972) gave a later summary of cross-cultural
findings from Piaget-type research. His research appeared
to varify the sequential order of the stages of development.
He identified several factors that seem to interact to
influence the rate of development: test materials, urban-
rural differences, schooling, European contact and nutrition.
The studies also showed that one cannot assume that adults
of all societies reach concrete operational thinking or
formal operational thinking.

Okonji (1971) and Price-Williams (1962) found that
if relevant test materials were used there was little lag

between African children and European children.

24

Greenfield (1966) and Price-Williams (1962) found a
lag in unschooled children compared with schooled children.
Schooled children among the Wolof were more like schooled
children in Boston than they were like their unschooled
neighbors in the same rural village.

Bovet (1974) hypothesized that logical structures
are closely linked to concepts used in regular, everyday
activities. Studies among illiterate women in Algeria
showed a relationship between experience in bread-baking
and skills in conservation of weight. The men were more
skilled at conservation of length, supposedly because they
walked long distances and could choose shortcuts. Both
men and women had difficulty with the conservation of time
and speed. These notions did not appear to be common in
their daily activities.

Za'Rour and Khuri (1977) found that in Jordanian
children the concept of speed was most closely related
to school achievement. Al-Fakhir (1977) found a four-year
lag in the development of the concept of speed in Iraqi
children. He blamed the lag on a poor school system which
emphasized rote learning from an authoritatian teacher.

Social interaction between a mother and her child
related to the results of conservation tasks in a study
by Adjei (1977). He observed mother-child interactions
in a puzzle task and then gave the children a conservation

task. Dominant mothers who used negative imperatives and

25
physical interventions tended to discourage self-initiated
actions and hindered later results in conservation tasks.

Both Adjei (1977) and Price-Williams (1969) found
that children with experience in pottery making did better
with conservation of volume skills, but in other tasks
they did no better than control children did. Conserva-
tion skills appear to be task-specific, with no transfer to
other areas of development.

Opper (1977) studied differences between urban and
rural school children. She hypothesized that if cognitive
development were the result of interaction with the environ-
ment, both urban and rural children would have equal inter-
action with the environment, even though the content of
the environment differed. But the results of the study
indicated that urban children attained conservation abilities
faster than rural children. The content of their inter-
actions with the environment appeared to be significant.

Schooling appears to be an important factor in cognitive
development. Kelly (1977) conducted an eight-year study
in Papua New Guinea. Schooling facilitated success in
concrete operational tasks. However, not even schooled
subjects succeeded with the formal operations pendulum
task.

Kiminyo (1977) hypothesized that rural Kamba children
would develop conservation abilities sooner than urban

children because of more active opportunities in the environ-

ment. He found age to be the only significant variable,

26

with no difference resulting from schooling, location or
sex.

Goodnow (1966) found that a lack of schooling made
no difference in conservation of weight, volume or surface,
but schooling did contribute to the development of combina-
torial reasoning. She concluded that schooling is not
necessary for the development of tasks using perception,
but that schooling is helpful in acquiring skills for tasks
that must be worked out in the head without concrete
referents.

Heron and Simonsson (1969), in a study in Zambia,
found that school, by itself, was not enough to promote
concrete operations. Schools which evidenced lack of inter-
action between teachers and students and which encouraged
rote learning did little to promote cognitive learning.

The interaction of sociometric status and culture
was found to be significant in Australia (de-Lacey, 1970).
Europeans achieved concrete operations earlier than abori-
ginals, and in both cultures persons at higher sociometric
levels achieved concrete operations earlier.

The amount of schooling is important. Laurendeau
(1977) found that less than six years of schooling made
no difference in the development of concrete operations.
None of the subjects in Rawanda achieved formal operations
unless they had been in school more than six years. The
study showed that schooling in excess of six years facili-

tated the development of concrete operations, and that

27

schooling is necessary but not sufficient for the attainment
of formal operations.

Piaget (1972) recognized the importance of relevant
experience when he admitted that adults may reach the formal
operations stage only in the specialized areas of their
aptitudes and professional specialization. The development
of formal structures may be more closely related to cultural
factors than to biological factors.

In summary, Piagetian cross-cultural research indicates
that whatever stages are achieved are usually developed
in the same sequence and may be tied to genetic patterns
or potentialities for development. But environmental
factors appear to limit or promote both the rate and the
limits of the developmental process. No single environmental
factor has consistently been shown to determine cognitive
development.

A large and growing body of cross-cultural literature
is emerging from studies on cognitive style by Witkin
(1967). Results of his studies on developmental factors
in cognitive style support some of the Piagetian cross-
cultural studies. Witkin stated several hypotheses and
verified them through research conducted around the world,
including Africa. He described cognitive style as a stable,
self-consistent mode of intellectual functioning that is
organismic, developmental and related to age and the social-

ization process.

28

Variables which interact to either promote or hinder
the developmental process are parental, societal, ecological
and nutritional. For example, mothers in a Temne society
are strict and authoritarian and societal structures are
tight and well-defined, thus not encouraging the development
of the child's articulated cognitive style. Factors which
promote development of articulated cognitive style in a
child are: mothers who have developed an articulated body
concept, mothers who punish consistently and less harshly,
tribal structures that are less dominating, a balanced
diet, and an occupation such as hunting, cattle herding
or sailing (in that these encourage individual mental
resourcefulness). Schooling is often an important variable
in the development of an articulated and differentiated
cognitive style.

Witkin wrote that both global and articulated cognitive
styles may be adaptations to social needs. Thus one cannot
make a value judgment as to which is ”better." Global,
or field-dependent, styles have a social orientation that
is more sensitive to the needs of group members, while
the orientation of articulated, or field-independent, styles
is more impersonal. Possibly there is a need for the global
style when one lives in close contact with many people
and a need for an articulated style when one must make
independent decisions in order to survive.

Bruner (1966b) has developed hypotheses on cognitive

development which have been tested in many parts of the

29

world. The stages of development are seen in the manner

in which the child represents reality, or events. Young
children represent the world through actions, (i.e., an
enactive mode). Next they progress to an ikonic or picture
representation stage, and finally to the symbolic stage.
There are two kinds of pushes toward growth, the internal
and biological unfoldings and the external or cultural
amplifiers.

Language is the most important tool for promoting
cognitive development. Language provides the I'temptation"
to form concepts with objects. Grammatical structures
encourage hierarchical structures of experiences. Words
can be used to hypothetically change the real world, which
encourages symbolic modes of representation.

Words used in reading and writing provide a more
powerful push for cognitive development. Schooling does
much to promote the use of words in reference to abstrac-
tions. The things one talks about or the things one writes
about do not need to be in the immediate situation.

Typically, anthropologists do descriptive studies
which provide insights into important events in a society,
but they do little to investigate internal cognitive struc-
tures. Psychologists, on the other hand, provide insights
into internal cognitive structures but pay little attention
to everyday manifestations of cognitive development in

a given culture. Michael Cole (1974) and his associates

30

did much to bridge the gap between psychology and anthro-
pology. After ethnographic and psychological studies among
the erlle of Liberia, Cole (1975) concluded that 1. all
people can carry out complex mental tasks, 2. people are
good at the kinds of tasks they do most often and 3. cross-
cultural measurement techniques do not measure cognitive
process, only cultural skills.

For example, Cole contrasted the effective, though
subtle, interpersonal communications among the erlle with
their poor, seemingly egocentric, performance on a stick
description task. He concluded that the test was not
measuring the cognitive process but was merely measuring
the cultural skill of stick description. Cole is eager
to disprove the deficit theory of cognitive development,
which resulted from the difficulty researchers have
encountered in attempting to design instruments. Differences
do not mean deficits (Cole and Bruner, 1971). Cole believes
that persons in all cultures have equal potential for cogni-
tive capacity, whereas differences in cognitive preferences
are explainable in terms of the needs, interests and values
of the particular culture.

Traditional people apply complex cognitive skills
in social situations but fail to use them in an experimental
task. Studies of cognitive development need to take into
account the situational variations in which the cognitive

process is manifested.

31

Berry (1976), a cultural relativist and a developmen-
talist, emphasized the appropriateness of adaptation. The
goal of cognitive development is not to reach the ”highest“
stage of development, but to achieve the most appropriate
stage for the natural and cultural environment. Empirical
data may show a consistent order of development, but that
does not mean that each individual in each society should
always move in that direction.

In summary, the theoretical literature of Piaget,
Witkin, Bruner and Cole suggests a tendency toward the
progressive development of cognitive structures in the
individual. Similarities in these developmental structures
have been demonstrated in various cultures, leading one
to accept a biological unity in all mankind. The differences
evidenced in cognitive development result from cultural
factors. Piaget, Witkin, Bruner and Cole state that these
differences do not necessarily demonstrate deficits, but
are probably the most appropriate cognitive structures
for the specific environment.

In the next section, theoretical implications are
applied to the more specific area of equivalence grouping
studies. Since the instrument used in the present study
is an equivalence grouping instrument, the following

literature review is appropriate.

32

Culture and Equivalence Grouping

 

A common way of measuring cognitive differences is
to observe how people group various objects into like or
similar categories. Bruner (1963) asks the question, "What
do people do when they relate one thing to another?” He
says that grouping is determined by gradually learned
systems. Associations are governed by certain rules or
transformations that are imposed on the data by the one
making the association.

The rationale behind a particular grouping is signifi-
cant. Given a set of pictures or objects, there is no
reason to believe all persons will automatically make iden-
tical groupings. One needs a purpose before identifying
an associated cluster. For example, a builder may cluster
materials for putting up a shelter, a traveler may associate
items for packing a suitcase, and a journalist may group
events which are present at the same time. But according
to Bruner, the basic purpose of grouping objects or ideas
is to reduce the cognitive load. A grouping rule is simpler
than all the attributes of all the objects. Grouping rules
reflect past experiences in grouping strategies. "When
people associate things with each other, they most often
do it by the extension or combination of groupings pre-
viously formed" (p. 354).

The theory of cognitive development indicates that as

children move from ikonic to symbolic models of representa-

tion, they tend to group more by superordinate rules and

33

by functional rather than perceptual attributes. Thus
the older child is able to use abstract rules to make hier-
archical categories.

Much of Bruner's work is an extension of the studies
of Vygotsky (1934) investigating the effects of language
on thought. Using wooden blocks of different colors and
shapes, together with nonsense words, Vygotsky measured
the ability to associate nonsense words with certain
concepts. For example, a small flat block was labeled
with the concept 331. He discovered three progressive
phases of development in children. First the child put
the blocks in piles for no reason other than subjective
connections. Vygotsky called these subjective connections
"heaps." Next the child progressed to the ability to think
in complexes. Though the bonds between the objects were
concrete and factual, they were not consistent or logical.
Complex thinking was found to be a preliminary means for
forming a true concept. Finally, the child was able to
abstract a single element out of the collection of objects
and form a superordinate concept. This stage took place
during early adolescence.

Inhelder and Piaget (1964) conducted free classifica-
tion studies in which they found that younger children
made groupings based on inter-relationships (e.g., the

baby goes with the crib). Older children tended to classify

34

on the basis of similar abstracted and superordinate attri-
butes (e.g., the crib goes with the chair because both
are furniture).

In summary, the theory of equivalence groupings indi-
cates that complexive or complementary grouping criteria
of perceptible attributes indicates less cognitive develop-
ment than do groupings with superordinate or similar criteria

with function attributes.

Picture Sorting Tasks

The present study is adapted from an experiment called
”Equivalence Formation With Pictures" by Olver and Hornsby
(1966). That study was conducted with ninety Boston school
children. The children were given an array of forty-two
pictures of familiar objects.

The children were first asked to identify each picture.
If a child could not identify a picture, he was told what
it was. The children were told to select from the pictures
things that were alike in some way. They could select
as many pictures as they wished. When the children had
completed the grouping, they were asked to tell how the
pictures they selected were alike. The pictures were then
replaced to their original positions, and the children
were asked to "form another group." The task was repeated

ten times.

35

Analysis emphasized the percentage of perceptual and
functional groupings, the number of pairs per grouping
and the model of grouping (superordinate or complexive).

Findings included the following:

1. Younger children tended to group more by perceptual
attributes than by functional attributes.

2. Younger children tended to use more complexive
groupings.

4. Younger children tended to group objects by taking
themselves as reference points.

Olver and Hornsby (1966) also used a word sorting
instrument in Boston. They gave children of different
ages a list of words such as banana, peach, potato, meat,
milk, water, air and germs. The children were asked to
tell how these objects were similar and how they were
different. Findings from the study have been used as a
basis for several important comparative studies in less
developed countries.

Maccoby and Modiano (1969) used an equivalence study
in Mexico to compare rural and urban children with the
children of Boston. They presented objects including banana,
orange, bean, meat, milk and water. Their instrument also
included three concepts that were represented verbally:
air, germs and a stone. Results showed that urban children
used more impersonal functional groupings than did rural

children.

36

Greenfield used an adaptation of the Maccoby instrument
in Senegal (1966). She used three sets of pictures that
could be grouped by color, shape and function. She reported
problems among illiterate subjects in understanding the
pictures. She pre-tested the pictures, but in the actual
study she did not ask the children what the pictures were
until the end of the experiment. Illiterates showed a
preference for using color as the common attribute, which
may indicate that they could not identify the pictures.

Reich used a similar instrument among the Eskimos.
Eskimo children differed from Boston children in using
a smaller number of superordinate responses. The instrument
included gloves, mukluks, sweater, parka, blanket, stove,
fire, sun and ice as objects for classification.

Two things can be noted concerning the studies using
picture sorting. The tasks were culturally adapted, and
the results tended to distinguish between schooled or urban
children, who preferred to use superordinate strategies,
and unschooled or rural children, who preferred to use
complexive strategies.

Another important study which included picture sorting
tasks was that of Kagen, Moss and Sigel (1963). Using
an instrument of twenty-two human figures, they found that
adults made three kinds of classifications: 1. analytic
or descriptive, 2. inferential and 3. relational or

thematic.

37

A correlation was found between picture-sorting pre-
ferences and personal characteristics. Analytic adults
tended to be more independent in family relationships,
were striving for social recognition, had higher I.Q.'s
and were persistent in solving problems.

Adults who tended to use more relational strategies
in sorting pictures also tended to be more dependent at
home, less concerned with social recognition, more anxious
in social situations and more passive in problem—solving
situations.

As children grew older, they tended to become more
analytic and to make groupings by similarity or super-
ordinate rules. Younger children tended to use relational
categories.

Winters and Brzoska (1976) used a labeling instrument
of projected slides to measure differences in grouping
strategies between normal and retarded adults. Normal
adults used more efficient superordinate groupings than
did the retarded adults.

Another related area of study has to do with develop-
mental differences among elderly adults.

Annett (1959) found in a study of adults that seventy
percent of the subjects over forty years of age preferred
to group with relational categories, and eight-five percent
of the subjects under forty years of age used similar or

nominal categories.

38

Denny (1972 and 1975) and Kogan (1974) obtained similar
results when older adults were presented with sorting tasks.
Older adults tended to use complementary criteria (e.g.,

a nail goes with a hammer), as do younger children. Younger
adults tended to use similarity criteria (e.g., the nail
goes with the pin), as do older children.

There are two possible explanations for the seeming
regression in older adults: 1) there is a deterioration
caused by neurophysical degeneration, or 2) there is no
environmental pressure for the more advanced forms of
thought.

Other studies indicate that similarities between elderly
adults and young children are more likely a function of
similar environmental pressures than a result of physical
deterioration. Studies by Denny (1975) and Hornblum (1976)
show that both young children and elderly adults can be
taught to use the more sophisticated thinking styles, which
would seem to indicate that similarities between young
children and older adults are not because of neurophysical
degeneration in older adults. Denny concluded that schooling
and occupational pressures place unnatural demands on the
individual to classify with similarity criteria.

There seems to be more evidence for the

position that all individuals are capable of

grouping according to either complementary or

similarity criteria and that the criterion they

use depends upon environmental factors...It may

be argued that both young children and elderly

adults are capable of grouping according to simi-

larity, but that they simply prefer to use comple-
mentary criteria. Neither the young child nor

39

the elderly adult experiences much pressure for
categorizing in any particular way; under such
circumstances, complementary categorization would
seem to be the most natural. After all, comple-
mentary items are grouped naturally in time and
space. One often sees cars in garages and base-
balls with bats. Therefore, the relationship
between such complementary items is readilly
apparent for all individuals and may predominate,
unless the individuals are confronted with external
pressures for organizing their experience in
another way. The transition appears to begin

in the young child at about age six, at the initia-
tion of the child's formal education with its
attendant emphasis upon similarity relations

and more abstract forms of thought. Such demands
would certainly continue through one's education
and into many occupations. In fact it is only
after retirement that many individuals are freed
from demands to categorize according to similarity,
and it is at this time that there appears to

be a return to the use of complementary criteria
for categorization (1975, p. 48).

Kogan (1974) concludes that older, healthy, educated
adults have not lost the capacity for superordinate modes
of classification; "rather, they appear more willing to
indulge in an alternate mode when circumstances permit

it” (P. 228).

Summary

Anthropologists and linguists have studied cognitive
similarities and differences in people of different cultures.
While Levy-Bruhl explains differences as deficits due to
slower evolutionary progress in certain primitive societies,
most researchers explain differences as appropriate responses
to specific environments.

The contributions of Piaget, Witkin, Bruner and Cole

indicate a basic cognitive unity in all mankind. Differences

40

in cognitive development result from cultural differences
and are not necessarily an indication of cognitive deficits.

Evidence seems to indicate that as children develop,
they move from using complementary, thematic and relational
criteria to using similar, analytic and superordinate cri-
teria in equivalence grouping tasks. Schooling is an impor-
tant factor in this developmental process. Elderly adults
do not lose the capacity for using superordinate criteria,
but they prefer to use the more natural and practical rela-
tional, thematic and complementary criteria.

The hypotheses and research questions of the present
study continue the investigation by looking at specific
cultural factors that influence equivalence grouping pre-
ferences among Tangale adults. The research methodology
used to investigate the relationship between schooling
and equivalence grouping preferences is discussed in the

next chapter.

Chapter 3

RESEARCH METHODOLOGY

In this chapter the research methodology is outlined.
The research design, including variables, hypotheses and
research questions, is described. Instrumentation, sampling,
data collection procedures and data analysis are discussed,
along with validity and reliability concerns.

The purpose of the research methodology is to test
hypotheses concerning the relationship between amount of
schooling, non-formal education, occupation and equivalence

grouping preferences.

Description of Methodology

The methodology can be described as descriptive, 35
£225 £3252, and cross-sectional, designed to explore the
effects of schooling and occupation upon picture-sorting
preferences in an equivalence grouping task. Since the
independent variable, schooling, has not been manipulated
except through the sampling of potentially different sub-

jects, the study is not experimental. In a correlational,

41

42
ex post facto study, data are collected after events have

occurred.

Design over Time

 

The instrument was administered to each subject ten
times in one sitting, making it a "one-shot case study"
(Campbell and Stanley, 1963). Since all the pastors in
the Tangale church district were tested, random sampling
of pastors was not appropriate. Teachers and illiterate
farmers were selected by convenience sampling, but were
chosen from each of the seven local church council locations

to ensure geographical spread.

Design over Variables

In order to limit interpretation of differences in
equivalence grouping results, the blocking variables of
age, sex, mother tongue and residence were used. Adult
males who spoke Tangale as a first language and who were
currently living in the Tangale District Church Council
(DCC) area were the only ones sampled.

The subject characteristics are the independent
variables (amount of schooling, recency of schooling, amount
of non-formal education, and occupation). The dependent
variables, described as sorting characteristics, are as

follows:

43

1. The rule or reasoning as to why the pictures are
alike.

2. The attribute observed in the sort.

 

3. The personalness or impersonalness of the pronouns
used to describe the equivalence.

The three main equivalence rules are superordinate,
complexive and thematic. These data were collected from
the verbal explanation given by the subjects.

The attributes of the sorting task are those which

 

the subjects described in their verbal responses as to
why the pictures were alike. Four main classes of attributes

are analyzed: external, internal, nominal and subjective.

 

 

External attributes are color and shape. These are
the obvious attributes based on perception of the pictures
alone with no interpretation needed (e.g., "The cow and
the bicycle are both black”).

Internal attributes are more interpretive and include
function (e.g., "They are used for travel"); place (e.g.,
"The table and chair go together”); material (e.g., "They
are both made of wood" ; or being (e.g., ”They breathe").

Nominal attributes use a hierarchical categorizing
term to say why pictures are alike (e.g., "The cow and
goat are animals”).

Subjective attributes relate to a personal experience
or interpretation of the pictures by the subject. These

can be affective (e.g., "I don't like snakes"): symbolic

44

(e.g., “This candle reminds me that Jesus is the light
of the world"); or fiat (e.g., "Because I say they are
alike").

The personal pronoun variable is a linguistic measure
of the personal or impersonal nature of the sort. A sort
is considered personal or impersonal based on the pronouns
that are used in the verbal description of why the pictures
are alike. An impersonal sort describes in the third person
the attributes intrinsic to the pictures (e.g., "They are
food"). A personal sort describes, in the first or second
person, attributes that are extrinsic to the picture itself
(e.g., "I travel by bicycle and taxi").

The subject characteristics were collected in an inter-
view before the sorting task was administered.

Amount pf schooling was measured in years of formal

 

schooling. Subjects reported which kinds of schools they
had finished, and the interviewer calculated the number
of years involved.

Recency pg schooling was measured from the data given

 

by the subject as the year he finished his last schooling.
Amount of non-formal education is a created variable
based on data concerning the number of books owned, theolog-
ical education by extension courses taken, literacy classes
attended, radio ownership, number of languages spoken and

the number of towns visited from a specific list. Each

45

subject was given one non-formal education (NFE) point for
each of the following criteria:

1. If he owned forty books or more

2. If he had attended theological education by
extension classes

3. If he had attended literacy classes

4. If he owned a radio

5. If he spoke three or more languages

6. If he had visited four or more towns from a
specific list of towns

7. If he read a newspaper more than four times a
month.

Each subject was rated on an NFE scale from zero to
seven based on the above criteria.

The three main occupations sampled were farmers,

pastors, and teachers in local government primary schools.

Hypptheses and Research gpestions

Hypotheses of relationship are stated, based on the
theoretical foundation of previous research.

Previous studies (Olver and Hornsby, 1966) show that
as the amount of schooling increases, there is a greater
tendency to use superordinate rules, internal attributes,
and impersonal pronouns. Theoretical implications (Bruner,
1966) are that schooling increases the ability to use the
more efficient superordinate rule structures. Internal

attributes and impersonal pronouns indicate that school

46

has taught a person to think about an object without its

being physically present.

Denney (1974) suggests that the effects of schooling

may reverse or "wear off" as schooling becomes less recent.

Hypotheses

H1

There is a relationship between
schooling and the percentage of
sorting rules used.

There is a relationship between
schooling and the percentage of
used in the sorting task.

There is a relationship between
schooling and the percentage of
used in the sorting task.

There is a relationship between

schooling and the percentage of

the amount of

superordinate

the amount of

internal attributes

the amount of

impersonal pronouns

the recency of

superordinate

categories used in the sorting task.

There is a relationship between
schooling and the percentage of

used in the sorting task.

the recency of

internal attributes

There is a relationship between the recency of

schooling and the percentage of

used in the sorting task.

impersonal pronouns

There is a relationship between non-formal education

and the percentage of superordinate sorting rules

used.

47

8 There is a relationship between non-formal educa-
tion and the percentage of internal attributes
used in the sorting task.

H9 There is a relationship between non-formal educa-
tion and the percentage of impersonal pronouns
used in the sorting task.

Hlo There is a relationship between occupation and
the percentage of superordinate sorting rules used.

Hll There is a relationship between occupation and
the percentage of internal attributes used in
the sorting task.

H12 There is a relationship between occupation and

the percentage of impersonal pronouns used in

the sorting task.

Population

 

The population of the study was defined as the area
of the Tangale District Church Council (DCC) of the
Evangelical Churches of West Africa (ECWA). The area is
about fifty kilometers from north to south and thirty kilo-
meters from east to west. The area lies about sixty kilo-
meters south of Gombe, in the northwestern part of Nigeria.
Kumo and Biliri are the two largest towns in the district.
The area is predominantly rural, with the only industry
being a cotton mill in Kumo. A hard-surfaced road runs
along one side of the DCC. The larger churches are connected

by dirt roads.

48

Oxen are used for plowing, with no mechanized farming
observed. Most of the actual farm work is done with a
hand hoe. Cotton is the main cash crop, and grain crops
of millet and guinea corn are the food staple. Nomadic
Fulani cattle herders are common in the area, making milk
and meat available in the market.

Missionaries of the Sudan Interior Mission (SIM) have
worked in the area for almost forty years--establishing
primary schools, a secondary school, a teachers' college,

a hospital, dispensaries, many churches, and a Bible Training
School for training pastors.

There are sixty-six local ECWA churches in the DCC,
with 2,000 members and 15,000 people in attendance. There
are about twenty preaching points in addition to the official
churches.

Non-Tangale people in the area include many nomadic
Fulani cattle herders, some Hausa traders in the market
centers, and a Yoruba bakery owner along the main road.

In summary, the Tangale area is rural, but not isolated,
with a high percentage of Christians. Its educational

and health facilities are above average for rural Nigeria.

Sample

The sample was made up of twenty—eight primary school
teachers, forty-six pastors and fifty-six illiterate or

semi-literate farmers. The mean age was forty with SD = 13.

49

For the teachers and pastors, the mean amount of schooling

was nine years with SD 3. The mean recency of schooling

was 10.7 years with SD 9.

Sampling Procedures

Each pastor in the DCC was interviewed. All fit the
sampling criteria for age, sex, mother tongue and residence,
and were given the picture—sorting task. Due to an error
in task administration, one pastor made only nine picture
sorts, and his data were eliminated.

The DCC is divided into seven local church councils
(LCC's), with about ten churches in each LCC. Teachers
and illiterate farmers were selected by a convenience sample,
but efforts were made to sample equal numbers from each
LCC in order to avoid over-sampling farmers and teachers
near the main roads and market towns. Thus the sample

is spread throughout the geographical area.

Instrumentation

The instrument is a set of thirty-five individual
water color pictures on cards six centimeters square.
Each picture represents a common object in the Tangale
environment, such as a bottle, a tree, a hat, a table

and a bicycle (see Appendix B).

Instrument Development
The following procedure was used to develop the instru-

ment. First, the author compiled a list of sixty objects

50

thought to be common in the Tangale area. Some items were
removed from the list in the interest of balancing the
number of objects in each category. For example, there
were too many animals and too many methods of travel and
too few articles of clothing and round objects. To further
refine the list, a Tangale informant gave advice on which
objects would be common.

The researcher then gave the artist a written descrip-
tion of each picture, specifying colors and shapes. For
example, the artist was asked to draw the coin, clock and
ball so that they would all be round and the same size.

Volunteers in the ECWA office tested the pictures
informally, and as a result some pictures were eliminated
or redrawn. Ultimately, the author chose the thirty—five
pictures in the testing instrument.

The author developed the instructions and interview
forms first in simple English. Twenty English-speaking
students at Kagoro Bible College and Igbaja Seminary then
field-tested the instructions and pictures. They were
asked to identify each picture before the instructions
were read to them. This test revealed that some pictures
were difficult to interpret. For example, the mango was
too green, and a church building had a southern Yoruba
architectural style not familiar to northern Nigerians.

As a result, the pictures were modified. The field study

also lead to rewriting the instructions. For example,

51

it was found that the students thought they were to do
all ten sorts at once. The instructions had to be clarified

to show that they were to do the sorts one at a time.

Hausa Translation

 

The instrument was translated into the Hausa language
and then back-translated into English to assure accuracy.
A committee met to discuss the wording of the instrument.
This committee included an English-speaking Nigerian whose
first language was Hausa and a Tangale informant who spoke
Hausa and English as second languages. Criteria for the
instrument included an accurate translation from English
in proper Hausa which would be understood by Tangale people.

In Hausa there is no word for ”similar." If the
instructions asked for pictures that were "alike,” the
Hausa translation would mean ”exactly alike” (921.921)'
Since no two pictures were exactly the same, this would
have caused confusion. The committee did not wish to tell
the subjects to group pictures that might be "found
together," thinking that such instructions might give cues
which would encourage complexive rules. Therefore, special
attention was given to the Hausa phrase 52235 92.1222.E2
kowace 22223 (”alike in any way").

Another careful Hausa construction was the instruction
telling the student to make a grouping ten times. The com-

mittee did not want to imply that the subject must try

52

to use ten different sorting rules and ten different attri-
butes. (The test was to measure grouping preferences,

not grouping abilities.) Instructions needed to be open-
ended enough for the subject to feel free to use different
strategies if he preferred, yet not feel forced to try

to find different strategies each time. The Hausa sentence
read 52 asks warewa EE.!2£2.222X2.9222§ ("Continue to group
them in another way"). The word BEBXE ("way") is as vague
in Hausa as it is in English, meaning anything from a "path”
to a "strategy." Thus the term 23222 £3223 (”another way”)
seemed specific enough, and yet open enough, to encourage

but not force differences in grouping preferences.

Hausa Field-Testipg

 

The Hausa version was field-tested at Tofa Bible
Training School where there were some Tangale students.
This test resulted in a few changes in the instructions,

but no changes were felt necessary in the pictures.

Instrument Reproduction

The author attempted to reproduce the pictures by
means of photography, but the first attempt resulted in
poor color reproduction. Because his time in Nigeria was
limited, and because color prints had to be sent out of
the country for processing, he decided to have a second
set of pictures painted. Each card was then covered with

Saran Wrap to protect it from dirt and water.

53

Interview booklets were mimeographed, with a set of
instructions and ten scoring forms included in each booklet

(see Appendix A).

Training of Research Assistants

When the author field-tested the instrument himself,
he found that at times the subjects used the pictures as
a means of giving a "lecture" on the local culture to the
foreign researcher. For example, one grouping included
a ten-kobo coin and a mortar. The subject explained that
their tribal wedding customs included the practice of
putting a coin in the bottom of the mortar to insure plenty
of food for the new family. To avoid the possibility that
groupings would be made to impress the foreign researcher,
it was decided that it would be better to use local Tangale
men for the administration of the instrument.

Two capable research assistants were employed to
administer the instrument. One was a graduate of Igbaja
Seminary, and the other was a man who travelled extensively
in the area holding literacy classes and Sunday School
seminars. Both men were known by the local pastors.

Training sessions lasted three days. On the first
day procedures were demonstrated and theoretical aspects
were explained. On the afternoon of the first day they
watched the author administer the instrument. On the second
and third days, the two of them observed each other adminis-

ter the instrument to pastors in Kaltungo, a town five

54
kilometers from the Tangale DCC, but not in the target
area. Debriefing of interview techniques took place both

during and after each practice interview.

Administration Procedures

 

The interviewing took place during the months of July

and August, 1976.

Language. The interviews were administered in Hausa,
the trade language of northern Nigeria. Tangale informants
said that Hausa would be easily understood and preferred
by the subjects. (Local church services are held in Hausa.)
Since the language of administration was Hausa, the study
could be replicated in dozens of other tribal groups in

northern Nigeria.

21323. The place of the interview varied from the
shade of a tree to a large church building. Subjects could
at times see other subjects from a distance as they were
taking the instrument, but they were always out of hearing
distance. To avoid contamination, subjects were asked
not to discuss the experience with those who had not been

interviewed.

Time. Letters were sent to the leaders of the seven
LCC's in the district, setting up an interviewing schedule.
All pastors in the LCC would come on a certain day to be

interviewed individually. Then in the afternoon, the

55

research assistants would interview local school teachers
and illiterate farmers.
The length of the interviews varied in length from

about twenty minutes to an hour.

Procedure. First the interviewer visited for a few
minutes with the subject. He then read the instructions,
each sentence twice, and asked if there were any questions.
The interviewer was told to be informal and to discuss
the instructions in the Tangale language if necessary.
Usually there was a short friendly discussion about the
task.

The interviewer then asked the subject to touch each
picture and to name it. If there were any question as
to what the picture was, the interviewer was instructed
to identify it in Hausa or Tangale and to explain it if
necessary. Seldom were pictures not understood; and if
there was a problem, once the picture was explained it
was usually used in the sorting task.

Then the subject was told to ”pick up pictures that
are alike and put them here" (a spot pointed to on the
table). The subject would make a grouping and the researcher
would record which pictures were chosen and in what order.
When the sort was finished, the subject would be asked
to tell why the pictures were alike. The response was

written down verbatim.

56

The subject was asked to continue to group the pictures
another time, or another way (S352 warewa 22.!EEEHEEBXS
.QEEEE)' (He had been told in the initial instructions
that he would be asked to make ten groups (H 5 gap ggma).
Thus each subject completed the equivalence grouping task
ten times.

The author was present periodically during the inter-
views to insure consistency between interviewers, but he

stayed in the background.

Scoring Procedures

 

Scoring was done from a literal English translation
of the data and from the original Hausa interview booklet.
A pre-scoring procedure was used as follows: Ten subjects
(or one hundred sorts) were scored with a preliminary scoring
form and then the scoring form was changed to adapt to
actual categorizing strategies. Forty more subjects were
scored (four hundred sorts) and the scoring forms were
adapted again (see Appendix.A). The one hundred thirty
subjects were then rescored. This procedure was designed
to insure consistency in scoring.

At times, the subjects named more than one rule or
attribute in explaining the similarity of a single sort.
In such cases, only the first rule or attribute mentioned
in the sort was analyzed.

Since there are few pre-conceived groupings, the task

is an open-ended, free-association instrument. While the

57

model measures a more natural thinking process, it also
makes for more difficult scoring.

After a period of instruction, a group of five judges
was given a random sample of interview forms to score.
At least three out of five judges agreed on eighty percent
of the sorts, and there was a ninety-eight percent agreement
with at least two of the judges.

The data were transferred from the scoring forms to
key-punch coding sheets, and then key punched and verified.

Random checks were made to insure accuracy at each stage.

Validity Concerns

The primary validity concern in this study is construct

 

validity. One aspect of the study has been to investigate
the relationship between the instrument and grouping prefer-
ences. Construct validation has come from inferences about
the data based on theories of cognitive development, using
both logical and empirical approaches.

In considering construct validity, one may ask, "What
other factors might account for the findings?” The claim
is that the instrument measures differences in grouping
preferences which relate to normal thinking patterns and
not to grouping competence. There are other possible ways
to account for the findings.

1. Could the task be differentiating on the basis
of the subject's willingness to cooperate? Observed

responses do not warrant such a conclusion. There was

58

full cooperation of the District Church Council and Local
Church Council. The Tangale research assistants took time
to visit with the subjects and to answer questions about
the project. When the senior researcher suggested that
gifts might be appropriate for the adults who took part,
the local Tangale leaders seemed insulted. Their attitude
was that if the results of the study would be useful in
developing a continuing education program for Tangale
people, gifts to the subjects for cooperation would be
inappropriate. Informal observation also indicated that
willingness to cooperate was uniformly high. When news

of the interviews spread to neighboring DCC's, those pastors
urged the interviewers to come and interview them as well.

2. Is the task differentiating on the basis of the
ability to attach meaning to pictures. Because of the
care with which the pilot studies were evaluated, resulting
in the redrawing of all pictures that caused confusion,
it is unlikely that recognition was a significant factor.
In fact, almost all the subjects in the study were able
to identify all the pictures.

3. Is the task differentiating on the basis of the
ability to verbalize reasons for grouping preferences?
Verbal skills may have caused some differentiation. However,
subjects who used the most efficient kinds of strategies

(superordinate, nominal) gave the shortest answers. At

59

no time was a subject unable to give a reason for a parti-
cular grouping. Nevertheless, it would be useful, in a
followup study, to compare statistically the frequency of
certain picture clusters without taking into account the
verbal reasoning behind each grouping.

4. Is the instrument differentiating on the basis
of fear of not following instructions or lack of ability
to understand the instructions? One could argue that such
fear might lock the subject into the same grouping pre-
ferences for all ten sorts, especially if no negative feed-
back was offered after the first sort. If the phenomenon
were more common among certain sub-samples, such as younger
men or less-educated men, then differences could not be
explained as real differences in preferences, thus making
the last nine sorts invalid. Several research factors
emerge to satisfy this question.

a. Each interview began with an informal visit of
five to ten minutes, during which the research assis-
tant attempted to put the subject at ease.

b. The instructions stated that the task was not
an examination, and that there were no right or wrong
answers. Subjects could make groups in any way they
wished.

c. If a subject asked after his first sorting

exercise whether he had done it "right," the research

60
assistant was to repeat in a friendly voice that
any grouping was all right.
d. While the Hausa vocabulary might seem open-
ended, the wording clearly provided for differences

in groupings. The Hausa instruction Ci gaba 93 wani

 

tari daban could be interpreted to mean, ”Continue

 

with another pile,“ or it could also be interpreted
as "Continue with a different pile.” These instruc-
tions were designed to be a compromise between forcing
the subject to use a different strategy for each
sort and the less directive suggestion that the task
was to simply be repeated. The research interest
was whether or not the subject would tend to use
different strategies if it was clear that he was
free to do so.
1. The data suggest that the subjects did
in fact understand the instructions to mean they
could do the task differently in subsequent group-
ings if they wanted to. In fact, no subject
used the same grouping of pictures in any two
of the ten sorts and no subject used only one
grouping strategy throughout all his ten sorts.
2. Since the purpose was to measure preference
in terms of tendency, not competence, the purpose
would have been defeated if the instructions had

been "loaded" to elicit a variety of approaches.

61

3. If the instructions had clearly stated

that subjects were to try as much as possible
to use different strategies, the instructions
would have been so complicated that many of the
subjects would have been afraid to try even the
first sort. For example, the Hausa language
would have required a statement that for each
sort the subject was "to try to choose different
pictures with a different number of pictures in
each sort, using a different grouping strategy
rule each time, using different categories of
attributes and using different pronouns in each
explanation." It is possible that even university
professors in America would be threatened by such
instructions, let alone illiterate Hausa farmers!

Other threats to internal validity were minimized

as follows:

1. Instrumentation Decay - The research assistants

 

rested often, and did not work more than five hours a
day, so that testing accuracy would not decline as the
research assistant became tired. Since the study measured
each subject only one time, problems of invalidity of
history, maturation and mortality can be ruled out.

2. Content validity - Evaluation of the three pilot

studies in Kagoro, Igbaja and Tofa supports content validity.

62

Moreover, the design of the instrument was subject to the
advice of the two Tangale research assistants.

Threats to external validity were minimized in the
following ways:

1. Experimenter effect - The test used only research

 

assistants who were locally known and trusted.

2. Hawthorne effects - The instructions were purposely

 

open-ended so that the subject did not know what the
researchers considered to be the "right" responses. The
research assistants tried to make the subject feel at ease
and to accept any answers.

It remains for future studies to explore further
preferences of equivalence groupings in the picture sorting
task and equivalence grouping preferences in the everyday

"real" world.

Reliability Concerns

 

Since the instrument is not claiming to measure compe-
tence, but preference, the typical reliability coefficients
are not appropriate. The consistency of the responses
is not the primary concern.

Scoring reliability is a concern. Because of the
open-endedness of the answers, a judgment is needed to
evaluate the categorization strategy. Inconsistency in
scoring was minimized by scoring and rescoring procedures

described in the section on scoring.

63

Data Analysis

 

The researcher employed both descriptive and inferential
statistical analysis. Figures are used to show trends,
using the percentage of each type of response for each
grouped independent variable.

It is assumed that superordinate rules, internal attri-
butes, and impersonal pronouns can be totaled for the ten
sorts of each individual, giving each subject a score some-
where between zero and ten. It seems logical to assume
that a subject who uses nine superordinate responses out
of a possible ten has a greater preference for superordinate
responses than does a subject who uses only two superordinate
responses.

The significance level adopted is alpha = .05 .

Summary
In this chapter, the research methodology of the study

has been identified. It can be described as descriptive,
35 ppgp Eggpg, and cross-sectional. The study is not
intended to be experimental. Theoretical relationships
were investigated by testing relational hypotheses. I

The picture-sorting task was administered ten times
in one sitting to each of the subjects.

The variables included the blocking variables of age,

sex, first language and residence to better control for

64

rival hypotheses other than amount of schooling, recency
of schooling, non-formal schooling and occupation. Dependent
variables were derived from the picture-sorting task to

include the equivalence rule used, the attributes described

 

and the personal pronouns used. The measure included fre-
quency counts on all ten sorts.

The equivalence rules were listed as superordinate,

 

complexive and thematic. The categories of attributes

 

included internal, external, nominal and subjective.

 

The personal pronouns describe the linguistic distance
or closeness of the subject to the objects sorted.

Hypotheses were stated.

The population and sample were described as being
adult males who spoke Tangale as a first language and who
lived in the Tangale District Church Council in the north-
eastern part of Nigeria. The sample included one hundred
thirty men from three occupations: illiterate farmers,
pastors and primary school teachers.

The instrument was described as thirty-five water
color pictures of objects familiar to the Tangale population.
The instrument was developed and field-tested for this
study. Two Tangale research assistants were trained to
administer the instrument.

Subjects were told to group any number of the pictures

for ten consecutive groupings of pictures which were alike

65

in some way. The subjects were asked to explain after
each grouping why the pictures they chose were alike.

Validity concerns were discussed. Construct validity
was the primary concern. Rival hypotheses were discussed
as being possible but unlikely. However, the relationship
between the grouping preferences using the instrument and
grouping preferences in the "real" world will need to be
further investigated in future studies.

Reliability concerns included scoring reliability
with an inter-rater reliability. Typical reliability coef-
ficients that measure consistency of response were not
considered appropriate for this task.

Data analysis was descriptive with one-way analysis

of variance used for inferential analysis.

Chapter 4
FINDINGS

The data collected through the research instrument
are reported and analyzed in this chapter. The research
question is reviewed. Hypotheses intended to answer the
research question have been tested, and the test results
are reported. Also reported are other findings which aid
in the exploratory nature of the study. Finally, the

findings are summarized.

The Research_gpestion

The purpose of the study was to investigate the effect
of schooling on the way adults prefer to group dissimilar
stimuli into equivalence groupings. The study looked at
the relationship between the independent variables of amount
of schooling, recency of schooling, amount of non-formal
education and occupation to the dependent variables of
superordinate rules, internal attributes and impersonal
pronouns. A picture—grouping task was repeated ten times
by 131 Tangale men of northeastern Nigeria.

Superordinate rules were defined as picture groupings

made on the basis of a common feature or attribute that

66

67
was equally true for each picture in the group (e.g., "The
shirt and the tree are both green").

Internal attributes were defined as interpretive of
the properties of the pictures, including functional attri—
butes (e.g., "The banana and the corn are both good to eat"),
materials (e.g., "The lantern and the hoe are both made
of metal"), or being (e.g., "The chicken and the snake
both breathe").

The impersonal pronouns were defined in terms of the
linguistic structure of the response. A sort was called
impersonal when the subject used impersonal pronouns (e.g.,

"They are all tools for farming").

Hypothesis Testing

 

Twelve research hypotheses, identified in chapter
three, have guided this study. To test the hypotheses,
a one-way analysis of variance test was used. The depen-
dent variables included the picture grouping characteristics
of the grouping rule, the grouping attribute and the pro-
nouns used in explaining the similarities among pictures
in a chosen grouping. More specifically, the superordinate
grouping rules, the internal attributes and the impersonal
pronouns were measured. The independent variables included
amount of schooling, recency of schooling, amount of non-
formal education and occupation.

One-way analysis of variance was used with the combined

data from all ten sorting trials for each subject. The

68

analysis tested for differences in the means of frequencies
of the dependent variables for groups from the independent
variables.

The Pearson Product Moment correlation technique was
used to explore relationships for variables on nominal
scales, and z-tests were made for variables with only two

means.

Hypothesis 1

The amount of schooling will be positively related
to the percentage of superordinate grouping rules used.

The result of the analysis of variance was an F-ratio
of 2.88 (df 3, 126) and is significant at the a = .05 level
lending support to the hypothesis. The means for the four

levels of schooling are shown in Table 4.1.

Table 4.1 Means for Amount of Schooling and Superordinate

 

 

 

Rules.
Level of Schooling M N
No school 3.67 46
1—4 years 5.90 10
5-8 years 5.00 18

9+ years 5.34 56

 

69

Analysis of the means showed the greatest difference
was between those who have had at least one year of school-
ing and those who have had no schooling. Thus, the data
lend support to a major portion of the hypothesis, but
not to all of it. Schooling does make a difference in
the use of superordinate groupings, but the amount of

schooling does not appear to be significant.

Hypgthesis 2

The amount of schooling will be positively related
to the percentage of internal attributes used.

The result of the analysis of variance was an F-ratio
of 1.40 (df 3, 126) and was not significant at the a = .05
level. The means for the four levels of schooling are

shown in Table 4.2.

Table 4.2 Means for the Amount of Schooling and Internal

 

 

 

Attributes.
Level of Schooling M N
No school 6.41 46
1-4 years ' 5.50 10
5—8 years 6.72 18
9+ years 7.25 56

 

Analysis of the means showed no trends. Therefore

it was concluded that the amount of schooling has no effect

70

on the preference for use of internal attributes. The

data did not support the hypothesis.

Hypothesisp3

 

The amount of schooling will be positively related
to the percentage of impersonal pronouns used.

The result of the analysis of variance was an F-ratio
of .93 (df 3, 126) and was not significant at the a = .05
level. The means for the four levels of schooling are

shown in Table 4.3.

Table 4.3 Means for the Amount of Schooling and Impersonal

 

 

 

Pronouns

Level of Schooling M N
No school 5.43 46
1-4 years 6.00 10
5-8 years 5.67 18
9+ years 5.39 56

 

It was concluded that amount of schooling has no effect
on the use of impersonal pronouns. The data did not support

the hypothesis.

Hypothesis 4
The recency of schooling will be positively related

to the percentage of superordinate rules used.

71
The result of the analysis of variance was an F—ratio
of 2.53 (df 3, 126) which was not significant at the a = .05
level. The means for the four levels of recency of school-
ing are shown in Table 4.4.

Table 4.4 Means for the Recency of Schooling and Super-
ordinate Rules.

 

 

 

Levels of Recency M N
No school 3.77 47
1-5 years ago 5.18 34
6-15 years ago 5.08 26
16+ years ago 5.74 23

 

Analysis of the means showed that the greatest dif-
ference was between those who have had no schooling and
those who have had schooling. For those who have attended
school, the means are almost the same. The preference
for superordinate sorting rules does not appear to "wear
off" as schooling becomes less recent. Thus the hypothesis

was rejected.

Hypothesis 5

The recency of schooling will be positively related
to the percentage of internal attributes used.

The result of the analysis was an F-ratio of .66

(df 3, 126) which was not significant at the a = .05 level.

The means for the four levels of recency of schooling are

shown in Table 4.5.

72

Table 4.5 Means for Recency of Schooling and Internal

Attributes.

 

 

 

Level of Recency M N
No school 6.47 47
1-5 years ago 6.91 34
6-15 years ago 7.35 26
16+ years ago 6.39 23

 

Analysis of the means showed no trends.

for using internal attributes is not related to the recency

of schooling. The hypothesi

Hypothesis 6

 

The recency of schooling will be positively related

3 was rejected.

to the percentage of impersonal pronouns used.

The result of the analysis of variance was an F-ratio
of .92 (df 3, 126) which was not significant at the a = .05

level. The means for the four levels of recency of school-

ing are shown in Table 4.6.

The preference

73

Table 4.6 Means for the Recency of Schooling and Imper-

sonal Pronouns.

 

 

 

Level of Recency M N
No school 5.53 47
1-5 years ago 4.91 34
6—15 years ago 5.54 26
16+ years ago 6.22 23

 

The preference for using impersonal pronouns was not

related to recency of education. The hypothesis was not

supported.

Hypothesis 7

 

The amount of non—formal education will be positively

related to the percentage of superordinate rules used.

The results of the analysis of variance was an F-ratio
of .78 (df 2, 127) which was not significant at the a = .05

level. The means for the three levels of non-formal educa-

tion (NFE) are shown in Table 4.7.

74

Table 4.7 Means for the Amount of NFE and Superordinate

 

 

 

Rules.

Amount of NFE M N
Low 4.34 44
Medium 4.82 68
High 5.44 18

 

Analysis of the means showed a trend toward super-
ordinate preferences as non-formal education increases,
but the amount of difference was too small to support the

hypothesis.

Hypothesis 8

 

The amount of non—formal education will be positively
related to the percentage of internal attributes used.

The result of the analysis of variance was an F-ratio
of .37 (df 2, 127) which was not significant at the a = .05
level. The means for the three levels of non-formal educa-

tion (NFE) are shown in Table 4.8.

75
Table 4.8 Means for Amount of NFE and Internal Attributes.

 

 

 

Amount of NFE M N
Low 6.48 44
Medium 7.07 68
High 6.17 18

 

The preference for using more internal attributes
by subjects with higher NFE was not supported by the data.

The hypothesis was rejected.

Hypothesis 9

 

The amount of non-formal education will be positively
related to the percentage of impersonal pronouns used.

The result of the analysis of variance was an F-ratio
of .57 (df 2, 127) which was not significant at the .05
level. The means for the three levels of non-formal educa-

tion (NFE) are shown in Table 4.9.

Table 4.9 Means for the Amount of NFE and Impersonal

 

 

 

Pronouns.

Amount of NFE M N
Low 5.61 44
Medium 5.26 68

High 6.05 18

 

76

The data did not support a preference for using imper-
sonal pronouns among subjects with high non-formal educa-

tion. The hypothesis was rejected.

Hypothesis 10

There will be a relationship between occupation and
the percentage of superordinate grouping rules used.

The result of the analysis of variance was an F-ratio
of 3.23(df 2, 127) which was significant at the a = .05
level. The means for the three occupations are shown in

Table 4.10.

Table 4.10 Means for Occupation and Superordinate Rules.

 

 

 

Occupation M N
Farmers 3.87 53
Pastors 5.93 46
Teachers 5.93 28

 

Analysis of the means showed that pastors and teachers
had a similar tendency to prefer superordinate rules.
Farmers had a lesser tendency to use superordinate rules.
No direction was hypothesized. The data supported the

hypothesis.

77

Hypothesis ll

 

There will be a relationship between occupation and
the percentage of internal attributes used.

The result of the analysis of variance was an F-ratio
of .16 (df 2, 127) which was not significant at the a = .05
level. The means for the three occupations are shown in

Table 4.11.

Table 4.11 Means for Occupation and Internal Attributes.

 

 

 

Occupation M N
Farmers 6.64 53
Pastors 6.69 46
Teachers 7.07 28

 

The data did not show a difference in the percentage
of internal attributes used by different occupations.

The hypothesis was rejected.

Hypothesis 12

There will be a relationship between occupation and
the percentage of impersonal pronouns used.

The result of the analysis of variance was an F—ratio
of .12 (df 2, 127) which was not significant at the a = .05
level. The means for the three occupations are shown in

Table 4.12.

78

Table 4.12 Means for Occupations and Impersonal Pronouns.

 

 

 

Occupation M N
Farmers 5.36 53
Pastors 5.54 46
Teachers 5.71 28

 

The data did not show a difference in the percentage
of impersonal pronouns used by different occupations.
The hypothesis was rejected.

Table 4.13 summarizes the results of the testing of

the hypotheses by one-way analysis of variance.

79

Table 4.13 Analysis of Variance Table.

 

 

 

Hypothesis Source df F Significance
No. Level, p=
1 Amount of schooling 3, 126 2.88 > .05

& superordinate rules

2 Amount of schooling 3, 126 1.40 n.s.
& internal attributes

3 Amount of schooling 3, 126 .15 n.s.
& impersonal structures

4 Recency of schooling 3, 126 2.53 n.s.
& superordinate rules

5 Recency of schooling 3, 126 .66 n.s.
& internal attributes

6 Recency of schooling 3, 126 .92 n.s.
& impersonal structures

7 Amount of non-formal 2, 127 .78 n.s.
education & superordi-
nate sorting rules

8 Amount of non-formal 2, 127 .99 n.s.
education & internal
attributes

9 Amount of non-formal 2, 127 .57 n.s.
education & impersonal
structures

10 Occupation & superordi— 3, 126 3.23 .05
nate sorting rules

11 Occupation & internal 3, 126 .16 n.s.
attributes

12 Occupation & imper- 3, 126 .12 n.s.

sonal structures

 

80

Observations

 

Other relationships were observed in the study. Some
of these relationships suggest hypotheses for further
studies and are discussed in observations 1-4. Observa-
tions 5-8 disconfirm some speculations concerning other
relationships.

1. Effects of the Fulani Language. Twenty-seven

 

of the Tangale subjects claimed to speak Fulani as one
of their second languages. The Fulani are nomadic cattle
herders who live in the Tangale area. Those who claimed
to speak Fulani as a second language used more impersonal
linguistic structures. The mean for Fulani speakers was
6.78 while the mean for non-Fulani speakers was 5.03
Z = 3.10 and p = .001). Fulani-speaking men also preferred
external attributes (M = 3.00) more than non-Fulani-speaking
men (M = 1.04), (z = 2.76, p = .001). Fulani—speaking
men preferred superordinate groupings (M = 5.85) more than
non-Fulani-speaking men (M = 4.39), z = 2.47, p = .001).

2. Effects of Occupation on Number of Pictures Used.
The use of only two pictures in an equivalence grouping
is seen to decrease in children as they get older (Olver
and Hornsby, 1966). Occupation made a difference in the
mean number of pairs per ten trials: 1. for farmers,

M = 4.69: 2. for pastors, M = 3.66; 3. for primary school

81
teachers, M = 2.93. The results of analysis of variance
show an F—ratio of 3.39 with 2, 128 degrees of freedom
and p = .05.
The relationship between occupation and the total
number of pictures in ten trials was not significant.

3. Effect of Number of Languages Spoken. There was

 

a weak but significant correlation between the number of
languages a subject spoke and the preference for more super-
ordinate rules and impersonal linguistic structures (r =
.227, p = .05 and r = .254, p = .05).

4. A Superordinate and Internal Variable. Informal

 

observation seemed to indicate that illiterate farmers
used more superordinate grouping rules with 22125 as the
attribute. Thus, a new dependent variable was created
that looked only at internal attributes which also used
a superordinate rule. The new variable which used super-
ordinate criteria for internal attributes was called super-
ordinate/internal.

There was a relationship between the amount of educa-
tion and the means for superordinate/internal groupings.
(F = 10.67 with df = 2, 128 and p = .01). There was also
a relationship between occupations and the means for super-
ordinate/internal groupings (F = 7.85 with df = 2, 128
and p = .01). The relationship of age, recency of education
and non-formal education to superordinate/internal groupings

was not significant.

82

Figure 4.1 shows the relationship between superordinate
attributes and the amount of schooling. The superordinate
external attributes (e.g., ”They are all the same color")
decrease from 53% to 3% as the amount of schooling increases.
Superordinate internal attributes (e.g., "They are all
made of wood”) increase from 30% to 56% as the amount of
schooling increases. Nominal attributes (e.g., ”They are
all animals") do not increase until the amount of schooling
reaches twelve or more years.

5. Effects of Age. Data on age were collected.

 

Ages were grouped into four categories: Age 20-29, N = 29;
Age 30-39, N = 39; Age 40-49, N = 32; Age 50 or higher,

N = 31. For the three dependent variables, age did not
have a significant effect as an independent variable.

6. Effects of Literacy Classes. There is no dif-

 

ference as measured by superordinate grouping preferences
between farmers who have had literacy classes and farmers
who have not had literacy classes (F = .01, df = l, 44).
7. Effects of Theological Education by Extension.
There was no difference in superordinate/internal grouping
preferences between pastors who had studied Theological
Education by Extension and those who had not (F = .01,
df = l, 45).

8. Relationship of Schoolingyto Non-Formal Education.

 

There was a relationship between the amount of schooling

and the amount of non-formal education. The means are

83

Figure 4.1 Superordinate Attributes by Amount of Schooling.

Average % of Superordinate Attributes

60%

50

40

30

20

10

 

 

 

External
Internal
/
Nominal
0 1-6 7-11 12+

Years of Schooling

84

as follows: 1. for those with low NFE, the mean years

of schooling were 1.92: 2. for those with medium NFE, the
mean years of schooling were 5.00; 3. for those with high
NFE, the mean years of schooling were 10.14 with an F-ratio
of 26.02 and p = .001. There was also a significant cor-
relation between the number of towns visited and the amount

of schooling (r = .380, p = .01).

Summary
In this chapter the results of the study were pre-
sented. The twelve main hypotheses were analyzed and other
observations about relevant relationships were presented.
One-way analysis of variance was used to test the
differences in the mean number of frequencies in the ten
sorting trials of the dependent variable with groups of
subject characteristics in the independent variable.
A summary of the results is as follows:
H1 The amount of schooling was related to super-
ordinate rules used.
H2 The amount of schooling was not related to internal
attributes used.
H3 The amount of schooling was not related to imper-
sonal structures used.
H4 The recency of schooling was not related to super-
ordinate rules used.
H5 The recency of schooling was not related to inter-

nal attributes used.

85

H The recency of schooling was not related to imper-

6
sonal structures used.
H7 The amount of non—formal education was not related
to superordinate rules used.
H8 The amount of non-formal education was not related

to internal attributes used.

H The amount of non-formal education was not related
to impersonal structures used.

10 Occupation was related to superordinate rules
used.

11 Occupation was not related to internal attributes
used.

12 Occupation was not related to impersonal structures
used.

The results are summarized in Table 4.14.

Table 4.14 Summary of Main Findings.

 

 

 

Superordinate Internal Impersonal
Rules Attributes Structures
H a H a H 0
Amount of H1 >.05 H2 N.S. H3 N.S.
Schooling
Recency of H4 N.S. H5 N.S. H6 N.S.
Schooling
Amount of H7 N.S. H8 N.S. H9 N.S.
Non-Formal
Education

Occupation H10 >.05 H11 N.S. H12 N.S.

 

86

Other findings showed that age was not related to
any dependent variable.

A superordinate and internal variable showed a signi-
ficant relationship for amount of schooling and occupation,
but not for recency of education and non-formal education.

There was no relationship between literacy classes
or theological education by extension (TEE) classes and
superordinate grouping preferences.

Fulani language ability was found to be related to
a preference for superordinate rules, external attributes
and impersonal linguistic structures.

The amount of schooling was found to be related to
the amount of non-formal education.

The number of languages spoken correlated positively
with the mean number of superordinate rules and impersonal

linguistic structures used.

Chapter 5
SUMMARY AND CONCLUSIONS

This chapter presents a summary of the investigation
and a discussion of research findings. Conclusions are
drawn from the discussion, and implications for teaching

and for further research are suggested.

Summary
The study has investigated the effects of schooling

on the way adults in northeastern Nigeria prefer to sort
dissimilar stimuli into equivalence groupings.

As western patterns of education spread throughout
Africa, it is important for educators to be aware of the
thinking-process preferences of the learner. It is likely
that a student will learn more easily if the structures
of the curriculum are similar to the student's preferred
and normal manner of structuring dissimilar stimuli.

The study provides data on the adult learner in a
non-western, rural society. The data are needed for con-
tinuing development of a theory of causes of cognitive

growth in adults.

87

88

Studies which looked at the relationship between
culture and cognitive growth were reviewed. The studies
showed differences between cultures in both the rate and
the limits of cognitive development. The amount of school-
ing, also, was often seen to be related to differences
in cognitive functioning. It was not assumed, however,
that the differences necessarily indicate cognitive deficits.
Instead, they may reflect appropriate responses to different
cultural needs.

Evidence seems to indicate that as children in the
western world develop they move from using complementary,
thematic, and relational criteria to using similar, analytic
and superordinate criteria in equivalence grouping tasks.
Studies of older adults seem to show that once they are
freed from the pressures of schooling and careers, though
they do not lose the capacity for using superordinate rules,
they prefer to use the more natural relational and complemen-
tary strategies.

One way to discover how adults prefer to sort dissimilar
stimuli into equivalence groupings is to ask them to sort
different pictures into groups of pictures that are related.
A picture-grouping instrument was constructed and, after
field-testing, was administered to 130 Tangale males near
Biliri in northeastern Nigeria. The sample was composed

of illiterate farmers, pastors and teachers.

89

Independent variables included the amount of schooling,
the recency of the schooling, the amount of non-formal
education and occupation. Occupation was considered an
appropriate effect of schooling in that one of the occupa-
tions selected was that of the school teacher, who would
likely continue to prefer a ”schooling” thinking style
even though his or her actual schooling had finished.

Dependent variables included the number of superordi-
nate equivalence rules used, the number of internal attri-
butes used (i.e., function, material or being attributes),
and the number of impersonal pronouns used.

In order to answer the research question concerning
the effects of schooling on categorization preferences,
the following hypotheses were tested:

H1 The amount of schooling will be positively related
to the percentage of superordinate sorting rules
used in the sorting task.

H2 The amount of schooling will be positively related
to the percentage of internal attributes used
in the sorting task.

H3 The amount of schooling will be positively related

to the percentage of impersonal pronouns used
in the sorting task.

4 The recency of schooling will be positively related
to the percentage of superordinate categories

used in the sorting task.

10

11

12

90

The recency of schooling will be positively related
to the percentage of internal attributes used

in the sorting task.

The recency of schooling will be positively related
to the percentage of impersonal pronouns used

in the sorting task.

The amount of non-formal education will be posi-
tively related to the percentage of superordinate
sorting rules used in the sorting task.

The amount of non-formal education will be posi-
tively related to the percentage of internal
attributes used in the sorting task.

The amount of non-formal education will be posi-
tively related to the percentage of impersonal
pronouns used in the sorting task.

There is a relationship between occupation and
the percentage of superordinate sorting rules
used in the sorting task.

There is a relationship between occupation and
the percentage of internal attributes used in

the sorting task.

There is a relationship between occupation and
the percentage of impersonal pronouns used in

the sorting task.

Results of one-way analysis of variance showed that

the amount of schooling and the occupation affected the

91

number of superordinate rule grouping strategies used.
The results did not show that recency of schooling or the
amount of non-formal education affected any of the dependent
variables. Hypothesis one and hypothesis ten were supported.

Fulani language ability was found to be positively
related to the use of superordinate rules, external attri-
butes and impersonal pronouns. Also, the number of lan-
guages spoken correlated positively with superordinate

rules and impersonal pronouns used.

Discussion of Findings

Results of the analysis of the twelve hypotheses were
reported in chapter four. Each hypothesis was designed
to help answer the research question concerning effects
of schooling on equivalence grouping preferences of Tangale
adults. In this section interpretations of the findings
are presented.

Hypothesis 1. "The amount of schooling will be posi-
tively related to the percentage of superordinate sorting
rules used.“ Findings partially support the hypothesis.

The amount of schooling does relate to preferences
for superordinate sorting rules over complexive and thematic
options. Tangale men with schooling experience preferred
to use more superordinate grouping strategies. The greatest
difference in mean scores was between those who had no

schooling and those who had at least one year of schooling.

92

There was little difference and no trend apparent between
those who had little schooling and those who had more than
nine years of schooling. This lack of a trend seems to
suggest that while schooling may encourage superordinate
sorting preferences, the actual amount of schooling makes
little difference. It could be that those who prefer to
use superordinate grouping strategies are the ones most
likely to go to school. But if schooling does have a
causal effect on superordinate sorting preferences, this
effect seems to take place in a short period of time and
does not appear to increase with more schooling.

Hypothesis 2. "The amount of schooling is positively

 

related to the percentage of impersonal pronouns used in
the sorting task." The findings do not support the
hypothesis.

At each level of schooling there was a tendency to
use impersonal pronouns. The idea that schooling is respon-
sible for the ability to look at objects from an impersonal
and more abstract viewpoint is not suggested by the data.
The preference for impersonal pronouns (e.g., "They are
used for farming") over personal linguistic structures
(e.g., ”We use them on our farms") does not seem to be
an effect of schooling.

Hypothesis 4. ”The recency of schooling will be posi-

 

tively related to the percentage of superordinate categories

93

used in the sorting task." The findings do not support
the hypothesis.

While the presence of the schooling experience does
make a difference in superordinate sorting preferences,
the recency of that schooling is not significant. It
appears that the effect of schooling is not significant.
It appears that the effect of schooling is long-lasting
and does not wear off with time. An alternate explanation
is that those with a natural tendency for superordinate
sorting preferences are the ones more likely to attend
school and that this natural tendency does not wear off.

Hypothesis 5. "The recency of schooling will be

 

positively related to the percentage of internal attributes
used in the sorting task.” The findings do not support
the hypothesis.

Since the amount of schooling does not affect the
percentage of internal attributes used, it is not surprising
that the recency of schooling is not related either.

If internal attributes are the more useful or normal for
all groups, one may assume that findings of differences
in the variable are measurements of differences in pref-
erences rather than measurements of differences in compe-
tencies as taught through schooling.

Hypothesis 6. "The recency of schooling will be

 

positively related to the percentage of impersonal pronouns

94

used in the sorting task." The findings do not support
the hypothesis.

Since schooling does not seem to promote the use
of impersonal pronouns, it is not surprising that the
recency of schooling has no relationship either. All
groups tend to prefer impersonal linguistic structures.

Hypothesis 7. "The amount of non-formal education

 

will be positively related to the percentage of super-
ordinate sorting rules used.” The findings do not support
the hypothesis.

It appears that the factors which make up the non-
formal education variable (literacy classes, extension
courses, newspapers read, radio ownership, books owned,
languages spoken and towns visited) do not have the same
effect on thinking styles that formal schooling has.

The result could be interpreted in at least two ways.

It could be that non-formal experiences, or ”awareness-
creating“ experiences, do not do as "good” a job of pro-
moting cognitive growth as does formal schooling. On

the other hand, it could be that superordinate preferences
are more useful for school experiences than they are for
the out-of-school world. Thus it may be possible that

the non-formal learning experiences do less to hinder

the more practical complementary and thematic preferences.
The data are not sufficient to support either of the

interpretations.

95

Hypothesis 8. "The amount of non-formal education
will be positively related to the percentage of internal
attributes used in the sorting task." The findings do
not support the hypothesis.

The preference for the use of internal attributes
(about 65% in all groups) cannot be attributed to the
amount of non-formal education a person has.

Hypothesis 9. "The amount of non-formal education
is positively related to impersonal pronouns used in the
sorting task." The findings do not support the hypothesis.

Impersonal pronouns were preferred by all groups,
even though amounts of non-formal education differed.

Hypothesis 10. “There is a relationship between
occupation and the percentage of superordinate sorting
rules used.” The findings support the hypothesis.

Both teachers and pastors preferred to use super-
ordinate rules more than did illiterate farmers. This
result is not surprising, in that sampling procedures
were used which intentionally selected farmers with no
formal schooling. The hypothesis was included to see
if school teachers who spend much time each day in a school
setting are different from pastors and farmers in their
sorting preferences. It could be that pastors see their
church ministry as a schooling task and thus are not dif-
ferent from teachers in their sorting preferences. It

is possible that the effects of the schooling background

96
for both teachers and pastors account for the similarity.
Or it could be that farming experiences encourage a more
pragmatic thematic or complexive grouping rule.

Hypothesis 11. "There is a relationship between
occupation and the percentage of internal attributes used
in the sorting task.” The findings do not support the
hypothesis.

All occupational groups tend to prefer the more useful,
but more interpretive, internal attributes.

Hypothesis 12. "There is a relationship between
occupation and the percentage of impersonal pronouns used
in the sorting task." The findings do not support the
hypothesis.

All occupational groups tend to use impersonal lin-

guistic structures.

Other Significant Observations. Age did not have
a relationship to any of the dependent variables. Thus
the findings do not support the idea that the tendency
toward complexive preferences indicates merely a develop-
mental lag in age.

The combined superordinate and internal sorting pref-
erence showed effects of amount of schooling and occupation,
but still did not show a relationship to non-formal educa-
tion or recency of schooling. Those with less schooling
tended to use external attributes when they used super-

ordinate rules (e.g., ”They are all the same color").

97

Possibly schooling encourages competency as well as pre-
ference for the more efficient, yet more difficult, super-
ordinate rule for internal attributes.

Neither literacy classes or theological extension
courses made a difference in preferences for superordinate
rules. It could be that these educational experiences
are not intensive enough or prolonged enough to promote
the use of superordinate strategies as does formal school-
ing experience.

The influence of language in the sorting task was
interesting. The ability to speak English was related
only to the combined superordinate/internal preferences,
with no relationship to superordinate or impersonal pronoun
preferences.

However, the ability to speak Fulani was related
to superordinate, impersonal pronouns, and external attri-
bute sorting preferences. It would be useful to investigate
the daily relationships Tangale men have with Fulani cattle
herders, to see if there is a transfer of cultural thinking
styles as suggested by the influence of use of the Fulani
language. Studies in cognitive style have suggested that
a nomadic life style fosters analytic and independent
thinking. Findings from this study indicate similar assump-
tions, in that Tangale men who speak Fulani prefer to

use the more analytic superordinate grouping rules and

98
have a greater tendency to use impersonal pronouns in
explaining the sort.

There was a similar correlation between the number
of languages spoken and the tendency to use superordinate
and impersonal grouping modes. Schooling has no relation-
ship to the linguistic structures, nor did use of English;
but the use of Fulani and the number of languages spoken
did have a relationship to impersonal modes. It is noted
that all the subjects spoke Tangale as a first language.
The possibility that second, third and fourth languages

can affect sorting preferences should be studied further.

Conclusions

Based on the findings of the study and in answer to
the research questions, the following conclusions can be
drawn.

1. Persons who have attended school prefer to use
superordinate rules more than those who have not attended
school.

2. Schooling does not affect preferences for sorting
with internal attributes.

3. The number of languages spoken has more effect
than schooling has on the use of impersonal linguistic
structures.

4. The recency of schooling is not related to any
of the dependent variables. Apparently the effects of

schooling are stable.

99

5. Non-formal education is not related to any of
the dependent variables. Apparently non-formal education
affects a person differently than does formal education.

6. Impersonal linguistic structures are equally
preferred by all groups in the study. Apparently schooling
does not affect a greater tendency to see objects from
an impersonal and objective viewpoint.

7. Pastors and teachers have a greater preference
for using superordinate sorting rules than do illiterate
farmers. The finding could reflect either the effect
of schooling or the effect of occupation.

8. Those who speak Fulani as a second language have
a greater tendency to prefer superordinate, external and
impersonal modes than do non-Fulani speaking people.

The reason for the tendency is unclear, but may be related
to cultural differences in the nomadic Fulani. Other
studies report that a nomadic life style is related to

a greater tendency to use analytic cognitive styles.

Implications

The above conclusions suggest a number of implications
for adult education practice and for further research.

In chapter one the need for an appropriate match
between the thinking style preference of the learner and
the structure of the curriculum was emphasized. The find-

ings of this study show differences in thinking preferences

100
in the equivalence grouping task. The differences are
related to schooling and language. People with schooling
have a greater tendency to prefer superordinate groupings.
Those who speak Fulani, or who speak more languages, tend
to prefer superordinate groupings.

In education practice, it cannot be assumed that
illiterate adults are accustomed to using superordinate
grouping strategies. Yet in current education models,
some subject matter areas and some instructional techniques
require the learner to use superordinate strategies.

To facilitate learning, subject matter areas could be
rearranged in order to put the more thematic-type studies
earlier in the shoooling experience and the more systematic
subjects later. Instructional techniques which give many
examples and do not depend on hierarchical logical struc-
tures should be used. These suggestions are most relevant
for the earliest years of schooling.

It cannot be assumed either that adults from different
language backgrounds have the same thinking preferences.
Subject matter structure and instructional techniques
may need to vary depending on the linguistic background
of the learner. Linguistic influences may be broader
than the learner's mother tongue alone.

The findings show another important factor for the
educator to consider. Most of the factors that reflect

the thinking preferences of a person have no relationship

101

to the amount of schooling he or she has had. Non-school
influences are great. The curriculum developer should

be aware of the fact that analysis of adult entry behavior
must take into account more than past schooling experiences.
Other culturally related experiences must be considered

as well.

The study shows that the preference for superordinate
modes of thinking is related to schooling and cannot be
assumed for newly entering adult students. Linguistic
background may also have a relationship to entry behavior.
Finally, the educator should consider out-of—school factors
in determining the thinking preferences of the learner.

Implications for further research. The findings of
the study suggest further areas of research needed in order
to answer the research question on factors which affect
thinking preferences.

The cause-and-effect relationship of schooling and
preference for superordinate thinking modes was not clari-
fied by the findings of the study. One way to further
investigate the relationship would be through the use
of a longitudinal study, which would indicate if adults
with an inclination toward superordinate thinking modes
have a greater tendency to go to school or if schooling
causes the inclination toward superordinate thinking

preferences.

102

The study has not determined whether the actual equiva-
lence grouping strategies were preferences or competencies.
To modify the technique by asking the subject to use a
different sorting strategy for each of the ten trials
would probably be confusing to the subject. One way to
modify the instrument would be to use the same pictures
to pose a problem which would stimulate the subject to
use different strategies or reward superordinate strategies.
One such modification would be the game of "Twenty
Questions," in which the subject would try to discover
which picture the experimenter had mentally selected.

The subject could ask superordinate questions which at
once eliminate several incorrect pictures (e.g., "Is it

a picture of an animal?“). Or the subject could make
inefficient and ineffective guesses (e.g., ”Is it the
cow?'). The instrument would pose a problem that would
reward superordinate responses and thus measure competen-
cies more than preferences.

The generalizability of the study was intentionally
limited in order to eliminate as many confounding variables
as possible. The study would have a wider generalizability
if it were expanded to include rural Tangale women and
children, urbanized Tangale, and people of other tribal
backgrounds.

The study did not attempt to look at cultural influences

(other than schooling) on equivalence grouping strategies.

103

Yet there are implications that language is a significant
variable underlying thinking differences. The influence
of the language of the nomadic Fulani seems to be signifi—
cant and suggests further research.

The findings of the study are not clear as to the
influence of non-formal education on equivalence grouping
strategies. Further studies could include other out-of-
school learning experiences. More sophisticated statistical
procedures such as linear regression equations could be
used to measure the relative effects of each of the factors
and to suggest further hypotheses.

The above suggestions for the practice of adult educa-
tion and for further research could do much to enhance
the learning experience of adults in Africa, where schooling
is often valued merely for the certificates achieved.

If thinking styles were considered in the development
of curriculum for adults, the educational experience could
become less frustrating and more relevant to life needs,

thus facilitating the personal growth of the individual.

APPENDICES

APPENDIX A

DATA COLLECTION FORMS

The following forms were filled in by the research

assistants as the subjects were interviewed.

104

kai.

105

GABATARWA

 

Dalilin wannan ziyara, shi ne ina so ne in gana da
Zan yi maka wa'dansu tambayoyi. In ka yarda sai

ka bani amsoshinsu. Wadannan tambayoyi ba gwadawa ba ce.
Amsoshin da za ka ba ni sai taimake mu shirya koyarwa irin
ta zamani a cikin Bible School da aikin Ilmin Kirista a
cikin ECWA nan gaba.

1.
2.

10.

Menene sunanka

 

L.C.C.

 

L.C.B.

 

Menene harshenka?

 

Shekarunka na haifuwa nawa ne?

 

Ko ka yi aure?

 

Littattafai nawa kake da su na kanka?

 

Mece ce sana'arka?

 

Ta yaya ka koyi wannan sana'a?

 

 

Wadanne irin makarantu ka yi?

Primary School

 

Reshen Bible School

 

BOTOS.

 

Advanced Bible School

 

Bridge Course

 

T.C.

 

BOC.

 

Secondary School

Yaki da jahilci

 

Correspondence School

T.E.E.

11.
12.
13.
14.
15.
16.
17.
18.

19.

20.

106

Craft School
Seminary
Wadansu dabam
Wace shekara ce ka gama makaranta takarshe?

Ka kan saurari rediyo?

 

Kana da rediyo na kanka?

 

Ka kan karanta jarida

 

Sau nawa a mako kakan karanta jarida?

 

Wace jarida ce kakan karanta

 

Shekaru nawa babanka ya yi a makaranta?

 

Shekaru nawa mamarka ta yi a makaranta

 

Wadanne harsuna ne kuma kake ji

 

 

Ka taba zuwa wadannan wurare?
Gombe
Bauchi
Maiduguri
Jos
Kano
Enugu

Lagos

10.
11.

12.

107

BAYYANI
Ka saurari dukan bayyani kafin ka fara.
Dubi dukan hotunan da ke gabanka.
Ka fadi sunan kowanne hoto bi da bi
Ka nemi hotunan da suka yi kama da juna. Suna iya
yin kama da juna ta kowace hanya.
Sai ka zabi hotunan da suka yi kama da juna. Za ka
tara su wuri 'daya.
Bayan ka zabi hotunan da suka yi kama da juna sai
ka fadi dalilin da ya sa hotunan suka yi kama da juna.
San 'an nan ka ci gaba da zaben wadansu.
Ka iya yin amfani da hotuna da yawa bisa ga sonka
a cikin tari dabam dabam.
Ka iya yin amfani da hoto 'daya a cikin tari dabam
dabam.
Za ka sami dama ka yi zabe sau 10.
Ka iya ka hada hotunan ta hanyarda kake so.
Wannan ba gwadawa ba ce.
Babu batun amsoshin da suka yi daidai da wadanda ba

su yi daidai ba.

giya zakara tebur
itace sa santali
buka riga takalmi
hula akuya taba

littafi maciji soro

108
kujera
kobo

itacen
masara

fanta

giciye

masara taxi
agogo ball

lorry fitila

ayaba matsefi

garma keke

ME YA SA KA SA WA'DANNAN A WURI 'DAYA?

kandir
almakashi

'kirana

majami'a

turmi

 

 

 

 

 

 

 

 

 

 

APPENDIX B

PICTURES US. IN THE RESEARCH INSTRUMENT

 

109

APPENDIX C

EXAMPLES OF RESPONSES

The first line of each example lists pictures chosen
as being alike. A literal translation from Hausa follows,
giving the explanation given by the subject and recorded
by the researcher as to why the pictures were alike. Each
adult made ten sorts and gave ten explanations as to why

the pictures were alike.

110

111

Example 1

10.

A 40 year old pastor

Total school - 4 years

Recency — 9 years ago

NFE Score - Medium

Cross, Bible: The Bible explains to us how Jesus
died on the cross.

Cap, shirt, show: Man wears a shirt and cap and puts
on shoes.

Hoe, corn: You cannot get corn until farming is done
with a hoe; then corn will appear.

Lamp, church: If a man has a house, he puts a lamp
in it to give light.

Table, chair: Man will sit on a chair and write on
the table.

Bicycle, lorry (truck): Bicycle and motor are articles
for traveling.

Hut, tree: If a man has a house, he needs a tree

in front of the house so that he can be resting.

Cow, goat, cock: We eat the meat of these things.

We sell them to get money.

Coin, clock: If a man has money he could buy a clock
so that he can watch the time for working.

Scissors, comb: Scissors can cut the hair of the

head. Comb is to comb the head with.

112

Example 2

A 49 year old pastor

Total school — 12 years

Recency - 9 years ago

NFE score - High

Candle, lamp, tobacco: These things use fire; they
bring smoke. They can all bring light in the darkness.
They can burn a person.

Church, house, hut: They are all buildings of mud.
Man can go into all of them and take shelter from
wind and rain. Man can sleep in them. Their doors
could be closed with no one to go into any of them.
Tree, cornstalk, cactus: They grow by themselves,
they have lives. They all die. They have leaves.
They eat food from the ground.

Taxi, lorry (truck), bicycle: All of them can run. All of
them can fall. They have engines. Man can ride on
them. They can deteriorate.

Table, chair, cross, mortar: They have been made

of dry wood. They can be burned by fire. They could
be cut with a metal knife.

Ball, shoe: All of the two are made of leather.

Man has made them. They could be burned. They could
be loosed and made to another thing different.

Fanta, wine, teapot: All of them could contain water.

They could get broken. All of them are built.

10.

113

Corn, banana: They are fruits of plants. All of
them get ripe. They are all eaten. They all soil.
Cow, goat, cock, snake: They are all living things.
They all eat food. They all run. They are all meat.
They could all die.

Bible: I have put the Bible alone. The Bible alone
among these pictures because it is the only thing
that has the word of life. Its word does not change.
It gives wisdom to man. The Bible is the inspired

word of God. It is not made by us.

114

Example 3

10.

A 23 year old teacher

Total school - 11 years

Recency - 1 year ago

NFE Score — Medium

Candle, lamp, cross, Bible: Because they are the

way of our salvation. The follower should have light
like the candle and lamp. The Bible and cross are
the way of life.

Beer, tobacco: Because they cause man to be drunk.
Coin: Because it is the root of all evils. It is
the one that brings different evils.

Church: Because it is where Christians should worship.
Shirt, cap, shoe, taxi: Because they are articles

of the enjoyment of man. Cap, shirt and shoe for

the closing of the secret of man, taxi also for the
traveling.

Cornstalk, corn, banana: Because they are the food
that brings blood to the body.

Cock, clock: Because all of them tell us time.
House: Because it is the place that Moslems do their
worship.

Cow, goat: Because they are animals that we tame for
getting meat and milk and skin for shoes.

Hut: Because it is the place where we sleep in for

the covering of our secret.

115

Example 4

10.

A 22 year old teacher

Total school - 12 years

Recency - 1 year ago

NFE score - Medium

Beer, Fanta: These are bottles

Bible, cross: These are related because the church
has a picture of the cross on its body.

Clock, ball, kobo: These three have the same shape.
Cornstalk, corn: These all are corn. One is the
cornstalk and the other was broken from a cornstalk.
Tree, cactus, cornstalk: These are plants. They
grow. They bear fruit.

Candle, lamp: These are used for giving light by
lighting fire on them.

Table, chair: They are useful in the work of teaching
and writing.

Bicycle, lorry (truck), taxi: These help man in
traveling.

Taxi, lorry (truck): These two, unless oil (gasoline)
is put in them they cannot work.

Cap, shirt, shoe: These are clothes, they are very

useful to man.

116

Example 5

A 55 year old farmer

Total school - 0 years

Recency - not appropriate

NFE score — Medium

Church, comb, banana: Because the color is the same
and they are beautiful.

House, tree: Because they are the same through their
color and they are beautiful.

Table, hut: Because their color and beauty have become
one.

Taxi, cornstalk, corn, cactus: Because their color
is the same, that is green. Taxi for traveling, corn
for eating and cactus for shade.

Cap, truck, Fanta: Because they have beautiful color
and they are useful. Truck for traveling, cap we

put on the head. There is also something good and
sweet to drink.

House, cock, cross, chair: Because their color is
beautiful. The chair is something for sitting. Here
is the cross of Jesus. Here is the European cock.
Here is a house for staying.

Bible, bicycle, scissors: Because their color is
beautiful. Scissors are for cutting cloth, a bicycle
is a thing for riding for traveling, but the Bible

is greater than them all for it is the word of God.

10.

117

If man reads it he will know the greatness of God.

May God give me knowledge to read his book.

Shirt, snake, mortar: Because they are good. Mortar
is for pounding corn for porridge which I will eat.
The snake is only beautiful, but is not useful. The
shirt is beautiful, I wear it, I will look handsome.
Cow, teapot, shoe, lamp: Because of seeing them and
their colors are good for me. If I put fire in the
lamp it will give light to my room. The owner of

the lamp should give light as the lamp has light.

The shoe is for putting on the foot to prevent the
foot from dirt and thorns and stumbling. The teapot
is for drinking water. Cows are good wealth. It helps
man in farming food for him so that he can help others.
Clock, coin, ball, goat: Because all of them are
useful to man. Clock is for looking at the time.

With money I can buy a shirt or my food and the book
of God to study. If I play with the ball I will exer—
cize my blood and I will rest from the heaviness of
body. The goat is good riches to sell and pay school
fees of my children. The rest I will buy my cloth

with.

118

Other examples from several adults

1.

Tree, cock, house, hut, chair, snake: This tree,

a man could sit under it on a chair. The one who

sits on this chair has a house and hut. In this house
there is a cock. In this tree there is a snake which
stays in it.

Bible, goat scissors, candle, taxi, hoe: This goat

is very proud. The pastor has come in his car. He
will talk to the goat. Pastor is talking to the goat
from the Bible. This scissors is the symbol of God's
power, how He cuts the one that is proud, so that

he can have light in his heart. The proud man will
take a hoe, that is to take the word of God and work
with it.

Snake, goat, cock, cow: These, all of them breathe.
Church, comb, candle, ball, clock, coin: These are
human handmade for the enjoyment of his living, but
the candle is more useful because it brightens the
room.

Cock, cow, goat, snake: The cock has life as man does.
The cow also has life and she moves as man. The

goat also has life: the snake also has life.

Banana, cactus, ball. Banana is not a tree, its fruit
is food. Its tree does not make fire. Cactus is
poison, it kills. The ball brings reward if it is

played between country and a country.

10.

119

Cactus, cornstalk, corn, taxi, church, comb: The
church and taxi are one in color, and the corn and
cactus and cornstalk look the same in their color.
Candle, scissors: I need them because they are useful
to me. The candle will brighten my house in the time
of darkness. It is the medicine for dripping if it

is applied. The scissors, because its color looks
nice; also because I barb with it and tailors cut
cloth with it.

Lamp, table, chair, clock, truck: I have seen that
these improve our living inside the house. They
beautify the inside of the house.

Fanta, teapot, candle, banana: These are useful for
the building of the family at home. They bring health

and the rest of mind. Every human being likes them.

BIBLIOGRAPHY

120

BIBLIOGRAPHY

Adjei, Kwabena. "Influence of specific maternal occupation
and behavior on Piagetian cognitive development,"
in P.R. Dasen (Ed.), Piagetian Psychology: Cross
Cultural Contributions. New York: Garden Press,
I977.

 

 

Al—Fakhir, Salima. ”The development of the concept of
speed among Iraqi children," in P.R. Dasen (Ed.),
Piagetian Psychology: Cross Cultural Contributions.
NEw York: Garden Press, 1977.

Annett, Marion. "The classification of instances of four
common class concepts by children and adults,” The
British Journal of Educational Psychology. Vol{_29,
pp. 223-236, 1959.

Berry, J.W. Human Ecology and Cognitive Style. New York:
John Wiley and Sons, 1976.

Berry, J.W. and P.R. Dasen (Eds.). Culturalrand Cognition:
Readings in Cross Cultural PsychEIogy. LondOn: Methuen
and Co. Ltd., 1974.

Boas, F. The Mind of Primitive Man. New York: Macmillan,
1911.

 

Bovet, M.C. ”Cognitive process among illiterate children
and adults," in J.W. Berry and P.R. Dasen (Eds.),
Culture and Cognition. London: Methuen and Co. Ltd.,
I974.

Brown, Roger. "Language and categories,’I in J.S. Bruner
et al., A Study of Thinking. New York: John Wiley
and Sons, 1956.

Bruner, J.S., R.R. Olver, and P.M. Greenfield. Studies
in Cognitive Growth. New York: John Wiley and Sons,

Bruner, J.S. and P.M. Greenfield. "Culture and cognitive
growth,“ International Journal of Psychology. Vol. 1,
NO. 2’ Pp. 89-107, 1966b.

121

Bruner, J.S. and R.R. Olver. "Development of equivalence
transformations in children," Mono raphs of Social
Research in Child Development. Vo . 28, No. 2, pp. 125-
143,41963.

 

 

Campbell, D.T. and J.C. Stanley, Experimental and Quasi-
Experimental Designs for Research. Chicago: an
MENélly, I966.

 

 

Cole, Michael. "An ethnographic psychology of cognition,"
in R.W. Brislin, S. Bochner, and W.J. Lonner (Eds.),
Cross Cultural Perspectives on Learning. New York:
JOhn Wiley and Sons, 1975.

Cole, Michael, and Sylvia Schribner. Culture and Thought:
A Psychological Introduction. New-York? *John Wiley
andTSOns, 1974.

 

 

Cole, M., J. Gay, J. Glick and D.W. Sharp. The Cultural
Context of Learning and Thinking. New York: Basic
Bocks, 1971.

 

Cole, M, and J.S. Bruner. "Cultural differences and
influences about psychological process," American
Psychologist. Vol. 26, pp. 867—876, 1971.

 

Dasen, P.R. "Cross-cultural Piagetian research: a summary,"
Journal of Cross Cultural Psychology. Vol. 3, No. 1,
pp. 2j-39' 1972.

 

de-Lacey, P.R. "A cross cultural study of classificatory
ability in Australia," Journal of Cross-Cultural
Psychology. Vol. 1, No. 4, pp. 293-304, I970.

Denny, Nancy Wadsworth, and Stephen W. Cornelius. "Class
inclusion and multiple classification in middle and
old age," Developmental Psychology. Vol. 11, No. 4,
pp. 521-5223 I975.

 

Denny, Nancy Wadsworth, and Martin Zobrowski. "Develop-
mental changes in clustering criteria," Journal of
Experimental Child Psychology. Vol. 13, pp. 275-282,
1972.

 

Denny, Nancy Wadsworth. "Evidence for developmental changes
in categorization criteria for children and adults,"
Human Development. Vol. 17, pp. 41-53, 1974.

Fishman, J.A. "A systematization of the Worfian hypothesis,"
Behavioral Science. Vol. 5, No. 4, pp. 323-339, 1960.

122

Goodnow, Jacqueline J., and Gloria Bethon. “Piaget tasks:
the effects of schooling and intelligence," Child
Development. Vol. 37, No. 3, pp. 573-582, 1966.

 

Greenfield, P.M. ”On culture and conservation," in J.S.
Bruner, R.R. Olver, and P.M. Greenfield (Eds.), Studies
In Cognitive Growth. New York: John Wiley and Sons,
1966. '—

 

Hall, Edward T. Beyond Culture. New York: Anchor Press/
Doubleday, 1977.

 

Harris, Verona. "Color perception and language,“ Report
of Seminar On the Perogption of Pictures and Its
Rélation tofiEducation. Kampala, Uganda, Makere Univer-
sity college, 1966.

Hoijer, J. Language in Culture. American Anthropological
AssociatIEn MemOIr, No. 79. Chicago: University
of Chicago Press, 1954.

 

Heron, A., and M. Simonsson. ”Weight conservation in
Zambian children: a nonverbal approach,” International
Journal of Psychology. Vol. 4, No. 4, pp. 281-292,
I969.

 

 

Hornblum, Judith Newman, and Willis F. Overton. "Area
and volume conservation among the elderly: assessment
and training," Developmental Psychology. Vol. 12,
No. 1, pp. 68-747 1976.

 

Inhelder, B., and J. Piaget. The Early Growth of Logic
I2 the Child. New York: Harper and Row, 1964.

Kagen, J., H. Moss and I. Sigel. "Psychological signifi-
cance of styles of conceptualization." Monographs
of the Society for Research in Child Development.
V61. 28, pp.73:112, 1963.

Kelly, Max. "Papua New Guinea and Piaget: an eight year
study," in P.R. Dasen (Ed.), Piagetian Psychology:
Cross Cultural Contributions. New York: Gar en ress,
Ific., 1977. '"T

 

Kiminyo, David M. "A cross cultural study of the develop-
ment of conservation of mass, weight and volume among
Kamba children." in P.R. Dasen (Ed.), Pia etian
Psychology: Cross Cultural Contributions. New York:
Garden Press,—Inc., 1977.

123

Kogan, Nathan. "Categorizing and conceptualizing styles
in younger and older adults.” Human Development.
Vol. 17, pp. 218-230, 1974.

Laurendeau-Bendarid, Monique (trans., H. Feider). "Culture,
schooling and cognitive development: a comparative
study of children in French Canada and Rawanda," in
P.R. Dasen (Ed.), Piagetian Psychology: Cross Cultural
Contributions. NeW—York: John Wiley and Sons, 1977.

 

Levi-Strauss, C. The Savage Mind. Chicago: University
of Chicago Press, 1966 (La pensee savage, originally
published Paris, 1962).

 

 

Levy-Bruhl, L. (trans. L.A. Clare). How Natives Think.
London: Allen and Unwin, 1910.

Levy-Bruhl, L. Primitive Mentality. London: Allen and
Unwin, 1923.

 

Maccoby, M. and N. Modiano, "Cognitive style in rural
and urban Mexico," Human Developemnt. Vol. 12,
pp. 22-33, 1969.

 

Maclay, H. “An experimental study of language and non-
linguistic behavior," Southwestern Journal of Anthro—
pology. Vol. 14, pp. 220-229, 1958.

Okonji, O.M. ”A cross-cultural study of the effects of
familiarity on classification behavior," Journal
of Cross-Cultural Psychology. Vol. 2, No. I, pp. 39-
'49, 1971. _'

 

Olver, R.R. and J.R. Hornsby. "On equivalence," in J.S.
Bruner et al. (Eds.), Studies in C nitive Growth.
New YorE: John Wiley 36d Sons{fil96 .

Opper, Sylvia. "Concept development in Thai urban and
rural children,” in P.R. Dasen (Ed.), Pia etian Ps chol-
pgy: Cgoss Cultural Contributions. New York: Garden
Press, Inc.,‘1977.

 

Piaget, J. "Intellectual evolution from adolescence to
adulthood,” Human Development. Vol. 15, pp. 1-12,
1972.

 

Piaget, J. "Need and significance of cross-cultural studies
in genetic psychology," International Journal of
Psychology. Vol. 1, No. I] pp. 3-13, 1966.

 

 

124

Price-Williams, D.W.G., and M. Ramirez III. "Skill and
conservation: a study of pottery making children,"
Developmental Psychology. Vol. 1, No. 6, p. 769,
1969.

 

Price-Williams, D.R. "Abstract and concrete modes of
classification in a primitive society," British Journal
of Psychology. Vol. 32, pp. 50-61, 1962.

Rivers, W.H.R. "Introduction and vision," in A.C. Haddon
(Ed.), Report of the Cambridge Anthropological Expedi-
tion to_the Torres Straights,*Vol. II. Cambridge:
Cambridge University Press, 1901.

 

Whorf, B.L. Lan uage, Thought and Reality. Boston:
M.I.T. Press, I956. (Also New York: John Wiley and
Sons)

Schulman, Lee S. "Psychology and Mathematics Education,"
in Edward G. Begle (Ed.), The Sixty-ninth Yearbook
of the National Society for the Study of EducaEion.
Chicago: University of ChicagoTPress, 1970.

 

 

Vygotsky, L.S. Thought and Langua e (E. Hanfmann and
G. Vaker Trans.). New York: gohn Wiley and Sons
(first published 1934), 1962.

Winters, John J., and Brzoska, Mary Anne. "Development
of the Formation of categories by normal and retarded
persons," Developmental Psychology. Vol. 12, No. 2,
pp. 125-13I,‘1976.

 

Witkin, H.A. "Cognitive styles across cultures," Inter-
national Journal of Psychology. Vol. 2, pp. 233:250,
1967.

Wundt, W. Elements of Folk Psychology. London: Allen
and Unwin, 1916.

Za'Rour, George J. and Shada A. Khuri. "The development
of the concept of speed by Jordanian school children
in Amman," in P.R. Dasen (Ed.), Piagetian Psycholo :
Cross Cultural Contributions. New York: Garden Press,
1977.

"lllll'lllllllllllllll“