394;. .

4 .1 ..
4H .In
"(3532‘

'3:‘ ‘8': .
g t.‘ ‘3?” -3:
9e 1?

‘2

n

. " ¢ 2 , 2 u
‘ -\- '~"r 2' .- ' 2:2.“v x'l’vW-‘C.
If: “'7‘. _ . _ . .I‘ ~ 1 _ , - ‘0: ‘1. "'2‘ . ' >t
. 'W‘ «1.1"?» . $22”: : - -.= "v 1- .~. . .- we?
. ’ r
f.

z. 2 ‘ . -
, . 2 .4 , ."r
2' figfiflé’f" "“5“; 3:"
.,.. 1,11" f4“ may
‘ ,figxéfi ’22..ng ‘
“1H - 2- .a- ' . '-
23: #2; ;_fi:§,

1.4-1! _ - 7'
Yr ‘ k ”fir,” ‘
3”“ 1:! ,r.',";}' .

v - I: ' "2 2'1" ’2
. , -

,5 Q, 44$. 1": - .
f /' .“r. l - ‘
5/

L45,

. 'f I, " .'
"V/~""’"19 .I '0. . ‘
VJ" MJ'” “4." l 2
, :f. .
3. :15! Is: a i‘gwu‘fi &;
"H .J- ‘J {awn-:1 -'|

u
I ' ,
F . 2’22”
‘1’ .

“55“ "’
:6
fl):

"‘3

1 C‘ V.
.1,

' "W

n . 1 -'r
KGTX’Z‘K‘.
.1 {3| .

. I ‘2

‘M

' "' -'... . ‘
‘. ’, '1‘ |§,€X-¢
'~i‘v

»
.a

4.42: 2-. . 2 2 . u-’~x;~
. ‘17 . > .1“ fig..:% "'3’
» -- 2 — , 222 2222,22,.
7%! ‘2'.“ “3:11;. 7"
' @‘é’fifirfirl
-~ u mewi“
‘ imagm 'hvfliéfi;
, .. '1 Maggy”;
f.>-.~¢:wc,-- /' ,‘ ‘3; ‘2. -' ' " 1353‘“;

Ju-‘w. X‘- ‘
245;? ":3 -- *f;
$22,22W‘“

:
z
‘w

r"
g.
1-3“
' I

L '2'

i- - » 4."
<"‘=*£*-‘2££<~é:f~~

,‘
'J A

1.- y .

«1.2/2.

“I: ‘/-."'
.9 '4‘ A?

.4.

 

\‘I
s.
8'2. '
.. ,

”if! .l.}|'|1(3'".I.NH‘§I-vuv A;

aims

.nw‘oiu. ...

«Minna».

.3;

7'. '

._“.

 

 

miniiiMiqui‘iiiii

1(9qu Oqo

”-1.”?

 

LIBRARY
Michigan State
University

 

 

 

: This is to certify that the
dissertation entitled

. Effect Of Group Size, Gender
And Ability Grouping 0n Learning Science
Process Skills Using Microcomputers.

presented by

Zane Lee Berge

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Educational Systems
Development

 

 

, 1' (/7 C‘,? r
flzév/ //;74 4

 

7/ . M‘fjor professor

Dateé/g/ 27):)

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

_ 7 _ _ 7 7 ,fi,,_

 

RASL)

 

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from

 

LIBRARIES
__‘—. your record. FINES will
be charged if book is
returned after the date
stamped below.
“‘“VM.
1.30 .
Nmi " "
{5-1 a. .2. .3; "Va-"4 2; 0 200?
' ’ i ~ -' APR 0 6 2009

f'JI‘L" final
uqixt U 2 Lvu

,iii . '7 r

r ' '- '

e'sé 6 if“
ere 1_ 1"”?
,.. t..—' -—

 

031809

 

 

EFFECT OF GROUP SIZE, GENDER, AND ABILITY GROUPING ON
LEARNING SCIENCE PROCESS SKILLS USING MICROCOMPUTERS.

BY

Zane Lee Berge

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Counseling, Educational Psychology,
and Special Education

1988

ABSTRACT

EFFECT OF GROUP SIZE, GENDER, AND ABILITY GROUPING ON
LEARNING SCIENCE PROCESS SKILLS USING MICROCOMPUTERS.

BY

Zane Lee Berge

This study investigated the effect of group size,
(individuals, pairs and quads of students), gender, and
ability grouping, (high, medium, and low), on student
achievement within an environment that uses microcomputers
as tools in learning science process skills. A split—plot,
multivariate factorial design was used to analyze the above
factors and interactions among the factors. Two response
variables, the Test of Integrated Process Skills and a
researcher developed test that more closely matched the
style and format of the practice that the students had
during the intervention, were measured using a pre-test and
post-test.

Science education was chosen as the context for the
study as it is an important school subject, yet one in which
the learning of problem-solving skills, commonly referred to
as process skills, is difficult. Three separate areas of
educational research are relevant: 1) student acquisition of
problem-solving skills, 2) appropriate use of computer
technology in school learning, and 3) students learning in

cooperative groups.

(n

(n

(W

Two hundred and forty-five, seventh and eight grade
student subjects were the focus of this study. They were
selected from twelve classrooms in three different school
districts. Selection of classrooms was based upon the
number of computers available and the teachers’ willingness
to participate in this research.

Analyses indicated that the only statistically
significant result was a main effect on ability for both
response measures. However, the two post-test measures
showed opposite trends in gain scores by low, middle, and
high ability students. Speculation concerning this ability
by post-test interaction was discussed. Other major
conclusions included: 1) teams of two and four members
working together solved problems as effectively as
individuals, 2) the lessons and procedures implemented in
the manner described, generated gender-neutral activities in
science, and 3) microcomputers, using a file management
program and structured activities, can be used as a tool to

promote student learning of science process skills.

To my parents, Iva and Mark Berge,

who gave me the education most worth having.

iv

i

t‘

< J

L)

‘fiv‘ I
¢~~
s .

ACKNOWLEDGEMENTS

The completion of a dissertation is usually the result of
support from many people. I would like to thank a few of the
people who have made the most significant contributions to my
efforts:

To my committee Chair, Richard McLeod, for your continued
encouragement and support editing multiple drafts of
the manuscript; and for your help in general throughout
my years in the ESD program.

To the members of my committee, Norman Bell, James Rainey,
and Stephen Yelon, for your support, encouragement,
suggestions, and patience throughout this research
effort.

To Andrew Porter, Co-director of the Institute for Research
on Teaching, for your suggestions concerning the
research design of this project; and for recruiting me
into the IRT fellowship program, plus your continuing
help throughout my years at Michigan State University.

To Kris Morrissey, a friend and fellow ESD student, for your
editorial comments on the final draft of this study;
and for making several collaborative projects fun.

To each of the classroom teachers, school administrators,
and students who participated in this study.

To Nancy, my wife, for your enduring encouragement and
support; certainly without which this dissertation
would not have been possible.

TABLE OF CONTENTS

PRELIMINARIES
List of Tables. viii
List of Figures. ix

CHAPTER ONE. INTRODUCTION.

Introduction to the problem. 1

Central Questions. 9

Research questions. 10
Purpose of the study. 11
The delimitations. 11
Significance of study. 12
Assumptions. 13
Chapter summary. 16

CHAPTER TWO. REVIEW OF LITERATURE.

Introduction. 17
Science education related to problem solving. 18
Computer technology in school learning. 23
Microcomputer database research. 25
Summary of computer related literature. 29
Gender related differences. 30

Gender related differences in
science education. 30

Gender related differences in
computer education. 33
Summary of gender related literature. 35
Learning in groups. 37
Group size. 38
Group composition. 40
Summary of learning in groups. 42
Chapter summary. 44

CHAPTER THREE. METHODS.

Population and sample. 47
Training. 47
Instrument. 48
Materials. 50
Procedures and design. 55
Assigning students to groups. 55
Data Analysis. 58
Chapter Summary. 59

vi

CHAPTER FOUR. RESULTS.

Description of the subjects. 61
Description of the data collected. 63
Research hypotheses. 68
Design. 70
Results and discussion. 73
Summary. 78

CHAPTER FIVE. SUMMARY, CONCLUSIONS,
AND RECOMMENDATIONS

Purpose. 80
Procedure. 81
Hypotheses. 82
Discussion. 83
Conclusions. 87
Limitations. 89
Recommendations for further research. 91
Other considerations/recommendations. 94
APPENDIXES
Appendix A. Teacher checklists for students
working in groups and individually. 97
Appendix B. The Processes of Analysis Lessons 99
Appendix C. Human Subjects Form 120
Appendix D. Researcher Developed Pre-test and
Post-test Questions (Part II) 122
LIST OF REFERENCES 130

vii

Table

Table

Table

Table

MANOVA

MANOVA

T-test

T-test

LIST OF TABLES

results for Part 1 of post-test.
results for Part 2 of post-test.

for Post-test Part 1 vs. Pre-test
Part 1

for Post-test Part 1 vs. Pre—test
Part 2

viii

71

72

73

74

Figure
Figure
Figure
Figure

Figure

LIST OF FIGURES .

Gradient Scale.

Plan to form groups with 2 students.

Plan to form groups with 4 students.

Within-subjects design.

Ability x Post-test Gain Scores.

ix

57
57
58
59

76

In

CHAPTER ONE. INTRODUCTION.

Introduction to the Problem.

This research focuses on students learning certain
science inquiry skills considered to be an important subset
of many problem-solving skills, within a microcomputer
environment. Science process skills are defined in this
study to mean those skills associated with planning,
conducting, and interpreting results from scientific
investigations. The phrases "process skills", "inquiry
skills", or "higher—order cognitive skills", are used
synonymously throughout this discussion, unless otherwise
stated. The perspective taken here is that science process
skills are a subset of problem solving. Problem solving is
a complex process that is recognized by many educators as an
important, necessary part of schooling and to an
individual’s life outside of school (Doerr, 1979; McGuire,
1973; Yinger and Eckland, 1975). Therefore, the teaching of
problem solving is frequently incorporated into various
school curriculum plans.

Definitions of problem solving appear to vary greatly
from simply a description of problem solving as a thinking
process, to relatively complex sequences of events, (e.g.,
active phase, evaluation). One list of problem solving

steps which may be useful for thinking about the focus of

this study was developed by Stapp and Cox (1979): I)
recognize the problem, 2) define the problem, 3) listen with
comprehension, 4) collect information, 5) organize
information, 6) analyze information, 7) generate alternative
solutions, 8) develop a plan of action, 9) implement a plan
of action, and 10) evaluate the plan of action. Any given
activity and/or instructional objective involving the
integrated process skills may concentrate on some subset of
these ten steps. However, it appears that the ten steps in
the Stapp and Cox model are not mutually exclusive and it is
assumed all steps are used to some extent by the student in
learning science process skills.

Before deciding hgw to teach problem solving, the
question of whether general problem solving can be taught
(e.g., Greeno, 1980) should be considered. In practice,
this question involves domain-independence of problem
solving, (i.e., general problem-solving skills that cut
across many disciplines) vs. domain specificity of problem
solving (i.e., problem solving depending on knowledge within
a given subject matter area). This debate is not new, of
course; and at one point the issue was whether students
should be taught Latin to "improve their minds", (i.e. to
give them general problem-solving skills). Some educators
believed that one trains the mind to discipline; others
thought that this type of discipline did little or nothing
to promote the transfer of problem-solving skills to other

domains (see e.g., Dewey, 1964; Whitehead, 1929). To date,

3

research has failed for the most part, to find significant
transfer from training on one task to another. This is
consistent with the generally accepted modern, experimental
psychological position that the learning of problem-solving
skills is generally unique to a given task (Newell, 1980).

Over the last few decades, science education as a
school subject has taken on the orientation of teaching
problem solving within the science domain and educators have
often concentrated on teaching problem solving using an
inquiry based curriculum. This type of curriculum utilizes
an inductive approach, (i.e., the reasoning from particular
to general, the inference of laws from observation). The
materials used in the classroom for this research focus
primarily on an inductive learning approach.

Even though few educators agree on the specific
characteristics or steps in general problem solving, most
authors agree that inquiry skills are a part of the problem-
solving process (Bruner, 1961). One instructional strategy
that is thought by some educators to help in the teaching of
higher-order thinking skills is to have students focus on
asking questions, instead of seeking one right answer; to
focus on gathering, organizing and analyzing relevant data,
rather than only on results (see e.g., Linn, 1986; Tobin and
Capie, 1980). New technologies, including microcomputers,
have emerged in the past decade which have promise as
valuable tools in the teaching of science process skills.

Used in laboratory type instruction, these tools may aid in

 

'71

(J.

(7

4

the development of scientific skills and positive scientific
attitudes in students (White, 1985; Thompson, 1986; Cox and
Berger, 1985).

In this study, students used database management
software to develop certain problem-solving skills. Unlike
computer-assisted instruction (CAI) software, which commonly
drills students on particular content matter (e.g.,
vocabulary), database management software (DBM) can be used
as a tool for a diversity of problem—solving activities.
Rather than presenting specific material or information,
this software has the potential for facilitating the process
of organizing, manipulating, and accessing information
(Freeman, Hawkins, and Char, 1984).

As microcomputers become available in schools, many
questions arise as to how they can best be used. The
computer environment provides many opportunities for
students to develop and test plans. Newell (1980) suggested
that the problem solver first constructs a plan in some
abstract or simplified "problem space" and then uses that
plan to guide the solution to the problem. The interactive
nature of computer learning often allows students to quickly
discover whether their plans work (Linn, 1985). The
computer need not be viewed as a discipline per se (i.e.,
computer science and computer programming), but as a unique
tool for encouraging activities and skills already stressed
in the established subject areas. This concept of using the

computer as a tool to promote inquiry skills already

5

stressed in science education is central in the current
research.

To date, there has been one published study identified
which investigated students’ achievement of process skills
using microcomputers and databases. In his dissertation,
White (1985) found that students using a computer-based file
management program, along with structured activities,
achieved higher scores on an information processing test
than did students in a control group. The subject area was
social studies, but the current study parallels that work in
many substantial respects as will be described in Chapter
Two.

Because of the rather rapid introduction of computers
into schools throughout the United States, it has become
important to examine ways in which computers can be
integrated smoothly and effectively into existing curricula.
Since computers are expensive and may be somewhat limited in
number in the schools at least in the near term, it is also
important to consider the advantages and disadvantages of
having students use computers in small groups (Trowbridge
and Durnin, (1984). Given the current conditions in schools
of far more students than computers (Bitter and Gore, 1986),
students often do their work in groups. One of the major
lines of investigation in the current study involved
comparing the student achievement of process skills of
students working individually at computers with those of

students working in groups at computers.

6

There is also a growing concern about the effects
(i.e., cognitive, social, and attitudinal) that computers
have on students compared to other forms of delivering
instruction, especially on lower ability and female
students. In the field of science, there is evidence of
differential success of males and females that might be
attributed at least in part to schooling. The fear is that
computer use will create or widen the schism between the
"haves" and "have nots" (Lipkin, 1983; MacGregor, 1985;
Walker, 1983). By age thirteen, females have begun to slip
significantly behind males in science achievement (National
Assessment of Educational Progress, 1978). The gap
continues to significantly increase through high school and
into adulthood.

Although warnings of gender related differences in
computer education have received considerable press (Walker,
1983; McPhail, 1985; Macgregor, 1985) there is a paucity of
systematic research (Anderson, et al., 1983) on this topic.
The research conducted to date involving gender related
differences in computer education examines differences in
computer access or usage but not achievement related to
problem solving directly. However, to the extent inequity
in access to computers and computer usage results in
achievement differences, these become important issues to
study. It is a common concern that all students have equal
opportunity and appropriate support for acquiring skills and

literacy with new technology (Linn, 1985; Walker, 1983).

(T

(n

[—4

(I)

(f)

(I)

(I)

3!

"fi

7

The research literature regarding group learning points
to strong evidence that cooperative learning strategies can
be used to decrease certain equity problems (Slavin, 1983).
There is also evidence showing that students working
together in cooperation are far more successful in problem-
solving achievement than students in competitive or
individualistic conditions (Johnson, Johnson, and Stanne,
1986). A cooperative learning activity is defined in this
study as a task performed by two or more individuals
"employing common means in a coordinated manner to attain
individual goals" (Bar-Tal and Geser, 1980, p. 214).
Learning cooperatively is directly related to this study,
since it investigated the effects on achievement of process
skills in science of individual students vs. groups of
students using microcomputers as tools in learning, with
special consideration given to the male/female equity issue.

There is very little previous research on cooperative
learning in microcomputer environments, especially in
naturalistic, (i.e. classroom), settings. Cox and Berger
(1985) in a laboratory study, concluded from their research
as it relates to learning problem-solving skills within a
microcomputer environment that "students work better in
teams than alone" (p. 467). They concluded that groups with
two to four students were most effective in solving the
types of problems used in their study. The Cox and Berger
(1985) study provides much of the basis in the current study

for determining the number of students per treatment group.

I).

' h

C4

(I)

m

In

nu

In

In

,4

8

Group size is only one factor that needs investigation,
however. The cooperative learning literature suggests an
aptitude by treatment interaction based on the ability
distribution of learners within groups. Some researchers
found a significant interaction favoring high achievers in a
small group (Peterson and Janicki, 1979). Others (Peterson,
Janicki, and Swing, 1981; Webb, 1977) found a curvilinear
interaction where high and low achievers learned best in
small groups while average achievers learned best working
independently. Therefore, the current research compared the
gains in achievement of low, middle, and high ability
students (as measured by achievement on the pretest), while
working either individually or in small groups.

When students learn in groups, individual
characteristics and behaviors are important to learning, as
are group characteristics (Allen and Feldman, 1976). The
current study investigated three variables that may effect
learning. The input characteristics focused on in this
study were: one individual characteristic (i.e., gender),
and two group level characteristics (i.e., group ability
composition and group size). Two output variables involving
science process achievement, were the outcome measure used
in this study. When a limited number of variables is
studied, many other variables (e.g., age, teacher expertise,
locus of control) must be held as constant as possible or
controlled for in some way, if the study is to be valid.

Two important methods utilized in this research to control

fly

. nu

AL
a

9

potentially confounding variables were: 1) randomization
(i.e., of classrooms to treatment; of students to within
classroom grouping), and 2) use of a pretest/post-test
design. The use of both of these controls is described
later in Chapter Three.

Central Questions.

Central to this study was the question, can
microcomputers be used to promote the learning of science
process skills? More specifically, what are some factors
which effect such learning within a microcomputer
environment? Three factors were chosen for this research:
1) group size, 2) gender, and 3) ability grouping.

With achievement in learning inquiry skills as the
dependent variable, the central questions related to this
study are: 1) With the limited number of computers currently
in classrooms, can more than one student effectively use a
microcomputer at the same time? 2) As microcomputers are
used as tools in solving more and more sophisticated
problems, and groups of students learn together, does the
ability level of a student within any group have an effect?
3) Can cooperative groups be used to decrease inequity in
achievement in schools regarding gender? Out of these
questions emerges a need for systematic investigations of
the cognitive consequences of process skills learning within

a microcomputer environment.

10

Research Questions.

The central questions translated into three research
questions examined in this study. The research questions
and their corresponding null hypotheses are:

1) Are cooperative learning groups more effective than
individuals working alone in learning science
process skills within a microcomputer environment?

H01: There is no significant difference in the
learning of science process skills between two-
member cooperative learning groups, four-member
cooperative learning groups and individuals who
work alone using microcomputers.

2) What is the interaction between high, medium, and
low ability students and group size in learning
science process skills within a microcomputer
environment?

H02: There is no interaction between high, medium, and
low ability students and group size on learning
science process skills within a microcomputer
environment.

3) What is the interaction between female and male
students and group size in learning science
process skills within a microcomputer environment?

H03: There is no interaction between the gender of the
student and group size on learning science process

skills within a microcomputer environment.

a:
his

ll

Purpose of the Study.

This study examined the effects of group size, gender,
and ability grouping on learning (i.e. achievement gain).
Students used a microcomputer and a database application
program as a tool when practicing certain science problem—
solving skills, (described previously as integrated science
process skills). This study also investigated the
interactive effects of gender on group size and ability
composition of the group on group size. The independent
variables were: 1) group size 2) gender, and, 3) ability
level. The dependent variables were two outcomes that were
a measure of student achievement of selected integrated
process skills.

The Delimitations.

This study involved only the output performance
variables described previously as integrated science process
skills. It did not involve other performance outcomes, nor
social outcomes, of learning. Therefore, this study cannot
be used to make inferences beyond the specific set of
problem-solving skills measured as the outcome variables
(i.e., relating to the outcome objectives which collectively
involve planning, conducting, and interpreting results from
investigations as stated in Chapter Three).

Since classrooms were selected based on teachers having
access to microcomputers and who volunteered to participate
in this research, the sample studied cannot be considered a

random sample. Students in the classrooms selected may have

Pr!

(1

12

more access to computers and/or other characteristics which
may not be representative of the middle-school population.
Therefore, Care should be exercised when making inferences
beyond the classrooms studied.

All students studied were middle school age (i.e.,
grades 7 and 8). Since students of other ages may have
different experiences and other characteristics (e.g.,
maturation) which may substantially change the population,
this study cannot be used to make inferences beyond the
population of middle-school age students.

Significance of the Study.

An individualistic assumption has been dominant in the
instructional use of computers in education (Johnson,
Johnson, and Stanne, 1986). The assumption by software
designers of one learner to one computer has gone largely
unchallenged. The results of this study may have
implications for software designers relevant to inclusion of
cooperative design for some learning objectives. It may
also have implications for policy makers regarding a method
for organizing group learning; and how many computers are
needed for the particular type of learning described and
used in this project.

This research extends to microcomputer environments,
the work already completed concerning cooperative learning
and group composition. It extends the work done by Cox and
Berger (1985) on group size in a microcomputer environment

to a naturalistic (i.e., regular classroom) setting. Thus,

13

the study may have implications for teachers wishing to
structure learning groups within their classroom in which
microcomputers are used. It also extends the cooperative
learning literature to a non-mathematics subject area (i.e.
science achievement), as suggested by Webb (1980); and
extends to science education the work completed by White
(1985) in social studies.

There has been concern in the computer research
literature that use of computers in the schools will
increase inequity in education (e.g., poor children
receiving qualitatively different kinds of instruction than
rich children, such as only drill and practice vs.
simulations; males dominating programming classes with only
a few females enrolling in computer classes). Cooperative
learning, on the other hand, has been effective in some
cases in decreasing student inequities concerning
achievement. Part of this study was designed to investigate
the effects of gender and ability composition of groups of
students on achievement (i.e. with regard to the development
of certain science process skills) within a cooperative
learning, microcomputer environment. Therefore, there may
be implications for classroom teachers wishing to group
their students with regard to ability and/or gender
considerations.

Assumptions.
This study assumes that teaching and learning in

schools do not usually take place within a one-to-one

14

interaction between teacher and student. That is, students
learn within a network of relationships with peers. It is
generally accepted that these student-student interactions
are important to the development of, among other things,
social competencies, socialization of sex roles, and
achievement. It is assumed here that constructive peer
relationships are not always formed automatically; and that
formation of learning groups can be planned and administered
effectively in a classroom setting.

Another assumption concerns Vygotsky’s distinction
between the social plane of a child’s development and the
individual plane. In the social plane, a child’s knowledge
is guided by the instruction of others; whereas in the
individual plane, the child's learning is under his/her own
guidance (Stein & Yussen, 1985). A "zone of proximal
development" refers to the difference in functioning between
these two planes. It is assumed that the child’s
development is accelerated when social agents promote
cognitive activity in excess of that attainable by the child
learning in the individual plane.

It is also assumed that peer interaction can, under
certain conditions exhibited in the procedures of this
study, be a catalyst in accelerating an individual student's
learning. Forman and Kraker (1985) suggest that peer
collaboration may enhance problem solving when students must
define and revise their understandings of the task, and

monitor and critique their problem-solving strategies in

15

social interactions. Peers may define problems differently
resulting in disagreements which may be resolved in a manner
leading to constructive learning.

It is assumed that the interactions among assigned
groups are not significantly different across the various
treatments.

The lessons developed for this research (McLeod, 1987)
were not field tested with students of middle school age
prior to the current research. Therefore, it was not known
that learning would indeed occur as a result of delivering
these five lessons that constitute the instructional
activities in the main study. The lessons were based on
similar materials by the McLeod, Hunter, and Finkel (1987)
which were tested with middle school students. However, the
tested materials were designed for more teacher involvement
than was desired in the current study. The lessons for this
study were modified to reduce the amount of teacher
involvement, and it was expected that these more self-
directed lessons would control to a large extent teacher
variations across schools and classrooms. Face validity of
the lessons for this study, (i.e., the evaluation by experts
regarding whether there is a match between the lesson
content and the author’s stated objectives for the lessons),
was established by independent evaluation and agreement of

three science educators.

16

Chapter Summary.

Problem solving continues to be an important part of
the general goals of education. Science educators have
chosen to direct some of their curriculum development
efforts toward teaching inquiry skills as part of problem-
solving skills within their subject domain.

Unfortunately, recent major reports and articles have
criticized the effectiveness of students’ learning of
inquiry skills; claiming this training is not very
successful. One approach that may help in the teaching of
these skills is to include the use of computerized databases
as tools to promote practice in what appears to be inquiry
skills. When integrating computers into the curriculum,
however, educators must be concerned with many issues.
Among these concerns are equity issues (e.g., equal use by
males and females) and, given that there are far more
students than computers in the schools, educators should
consider investigation of group learning with computers.
These considerations lead to the current research project
which investigated the effects of group size, gender, and
ability grouping on student achievement of science process

skills within a microcomputer environment.

v.—

AC

CHAPTER TWO. REVIEW OF LITERATURE.

Introduction.

In reviewing the literature relevant to the stated
research problem, this chapter focuses on three areas of
educational research: 1) acquisition of science skills, 2)
learning in small groups, and 3) the appropriate use of
computers in education. Particular attention will be
focused on learning of higher-order skills (e.g., integrated
process skills) and studies within computer environments in
education. Higher- order skills, for this discussion, are
related to Bloom’s taxonomy (Bloom, 1981; Bloom, Englehart,
Furst, Walker, and Kathwohl, 1956, 1972). This taxonomy is
organized using six major classes, which are: 1) knowledge,
2) comprehension, 3) application, 4) analysis, 5) synthesis,
and 6) evaluation. These behavioral classes are in a
relative hierarchical order, and move from simple to complex
and concrete to more abstract levels. This study was
focused on the higher-order skills of analysis, synthesis,
and evaluation. Of course, the lower order skills are
subsumed in these skills as well. Of these skills, the most
central to this research are the skills of analyses.
Analysis is often used in scientific inquiry, and is

believed by some educators to be one of the most directly

17

18

related skills practiced by students using databases (e.g.,
Bommarito, 1986; Pon, 1984).

The chapter begins by identifying the difficulty in
learning process skills within the larger context of problem
solving in science education; and proposes support for why
this type of learning is important to students. Secondly,
the literature concerning learning within microcomputer
environments is investigated, emphasizing applications use.
Thirdly, two areas that are related to the learning of
science in schools are discussed - they are: 1) the problem
of gender-related differences, and 2) learning in groups.
Science education related to problem solving.

Problem solving has long been an important part of
school learning. /The unprecedented pace of scientific and
technological innovation since the middle of the century has
made insistent demands on science education) The quality of
science training beginning as early as the pre—school years
affects how well citizens understand their increasingly
complex world and how effectively they cope with change (see
e.g., National Science Teachers’ Association Position Paper,
1975; Linn, 1986). Pogrow (1983) states that voters will be
asked to express preferences on a variety of issues related
to scientific issues (e.g., acid rain) and the use of
technology in society. Pogrow (1983) goes on to state that
there are implications for curricula decisions if a country
is to have a well informed national population. He suggests

five major curricula implications inherent in having a

19

population with the flexibility to meet employment and voter
demands in an "information society". Two of these are: 1)
increase the distribution of higher-order skills among
students, and 2) strengthen the mathematics and science
curriculum.

As noted earlier, few authors agree on the specific
procedures in general problem solving (i.e., generic
problem-solving skills that cut across all or many subject
matter domains), or whether generic problem-solving skills
can be taught. Some definitions describe problem solving
only as a thinking process. Others include recognition and
thinking processes, and still others include the action and
evaluation steps (see e.g., Bruner, 1961a, 1961b; Taba,
1962; Dewey, 1910; Wallas, 1926; Polya, 1957; Simon &
Newall, 1971). However,/in the past two decades, science
educators have focused their curriculum on developing the
ability of students to use specific components of a process
of problem solving. Instead of attempting to develop
generic skills of problem solving, science educators work
within the domain of science to develop skills related to
components that many educators believe to be an important
part of the problem-solving process. One well defined list
of these components has been developed by the American
Association for the Advancement of Science (AAAS, 1967;
AAAS, 1976). The AAAS has emphasized the development of the
skills of students to employ these processes of science to

learn about every day phenomena. In the "Science... A

20

Process Approach" program (SAPA), AAAS identified eight

basic processes and five integrated processes as follows:

 

 

Basic Processes Integrated Processes

1. Observing, 1. Formulating hypotheses

2. Inferring 2. Controlling variables

3. Using time and space 3. Defining operationally
relationships 4. Interpreting data

4. Using numbers 5. Experimenting

5. Measuring

6. Communicating

7. Classifying

8. Predicting

/"
The developers of SAPA believed that thegsystematic teaching

of science process skills would result in the acquisition of
scientific literacy for all citizens>(Baird, 1985)./ This
line of curriculum development leads to the idea of teaching
science as a process, i.e. what scientists do, rather than
teaching science as a body of knowledge, i.e. what

scientists know.fi

/

Linn (1986) expressed the need for integrated process
skills in science education:

The information explosion changes the
nature of knowing from the ability to
recall information to the ability to
define problems, retrieve information
selectively, and solve problems
flexibly. Rapid advance changes the
nature of learning from the need to
master topics in class to the need to
learn autonomously. Educated citizens
need to know how to revise their ideas
and how to locate and synthesize
information (page 13).

/
(;Is there reason to believe a process approach to
teac ing science is effective? Shymansky, Kyle and Alport

(1983) reviewed results from 45 studies regarding the

21

effectiveness of SAPA. Their review suggests that student
achievement, process skills, perceptions of science,
analytic skills, and related science skills increased an
average of 0.27 standard deviations in elementary school
learning. Furthermore, process skills improved by 1.08
standard deviations over students using other curricula?)
Problem-solving skills are thought of as important in
school learning. Both the "Educating Americans for the let
Century" (1983) and "A Nation at Risk" (1983) emphasize the
need for instruction which fosters problem solving, prepares
learners to deal with naturally occurring problems, and

encourages students to think critically.) Although general

\

problem solving is not well defined, there are skills which
have been identified as contributing to problem solving in

\\

science. Some of those skills are the focus of this study.)
More specifically, the integrated process skills that were/
addressed in this research are: 1) formulating hypotheses,
2) controlling variables, 3) defining operationally, 4)
interpreting data, and 5) experimenting.

In the rapidly changing world, it may be assumed that
teaching inquiry skills prepares students to meet challenges
in the future world in which knowledge and "facts" will have
changed from what we believe today (Finley, 1983; Walsh,
1985). In particular, it has been argued that an inquiry
orientation in science education empowers a person for

autonomous, life-long learning (Joyce, 1985; Streibel and

Garhart, 1985; Thelen, 1972). Yet, currently, science

22

students are not learning higher order cognitive skills (see
e.g., Mitman, Mergendoller, Packer, and Marchman, 1984;
Doyle, 1983). In the words of Tobin, (1986):

Stake and Easley (1978) found that
teachers emphasized learning of facts about
science and provided students with few
opportunities to develop the higher level
thinking skills that most of the courses
purported to develop. The courses were very
textbook oriented and students tended to lack
motivation to learn about applications of
science to the world outside of the classroom.

(p. 1)

Several important studies document the serious problems
in science education today (e.g., A Nation At Risk;
Educating Americans for the let Century). The National
Commission on Excellence in Education reported in "A
National at Risk" (1983) a pressing need for educational
reform to create a "learning society." The National Science
Board in "Educating Americans for the let Century" has

state a need for "new basics" or the thinking skills

,_':—o
”‘0.” .._ _, . .4..—r'--—---"

required to cope Wlth rapid technological and soientific

""“..I-r

changes. These studies also often point out that,

...__ J..-

 

unfortunately, our schools may not be adequately preparing

«MW

students for problem solving in our rapidly changing world.
/in light of this, it may be that the existing
curriculum can be changed to include some of the new
technologies that may help in promoting practice of
desirable skills in many of the subject areas. The next
section reviews the literature concerning one of the

emerging technologies - the microcomputer - as it relates to

school learning and this research.

23

Computer technology in school learning.

Within any given classroom of students, there are
differences among individuals. These differences may
include motivation, ability, attitudes, personality, sex,
ethnicity, and socioeconomic class (Peterson, 1981). This
diversity presents a challenge to the classroom teacher
attempting to meet the needs of individually different
students. It has been suggested by many educators that the
microcomputer has the potential to increase learning on an
individual learner basis. The computer can be a tool used
in educating students or an object to be programmed and
learned about. In the past, public schooling has been more
interested in the latterfg To date, very little empirical
research has been done on learning when and how the computer
can be used as a tool and integrated throughout the
curriculum:)

Businesses have for years enjoyed the advantages of
"high tech" tools. These advantages include computers that
help organize and speed transactions, reduce labor costs,
and improve competition. Most of us, in the normal course
of our daily lives, are affected in some manner by computers
each day. Zinn (1979) proposed that educators need to
realize the pervasiveness of computing in a person's
everyday life when planning education (Cox, 1980).

9):: [The explosion of new technologies and new information

|
a , I
’ 1

has challenged science educators to revise traditional

approaches and set new priorities. This challenge, for the

24

most part, has not yet been satisfactorily met (Linn, 1986).
Computer technologies enable students to perform tasks that
are very different from those done in the past. These tasks
require new skills such as planning solutions using software
instead of manipulating equations. The new technologies
often reduce the workload for students much as they have
done in the business community, by relieving students of the
need to focus attention on technical details; and thereby
permitting them to concentrate on the problems they are
solving.

/ There is a shortage of empirical research in many areas
of computers in education (e.g., simulation). The research
on the use of microcomputers in education shows few
controlled studies which provide any evidence of more
effective learning using computers vs. other instructional
delivery methods. (Effectiveness should not be confused
with efficiency here. Research on computer assisted
instruction has fairly well established a clear image of
learning taking less time using computers under many
learning conditions than with more traditional instructional
delivery methods). As Clark (1983) has suggested, the lack
of evidence showing CAI as superior in effectiveness
compared with other delivery systems may be because
computers are no more or less effective than the other
methods per se; but rather it is the underlying
instructional design that differs from one treatment to

another.

25

Of course, as with any instruction, there are
conditions which when met, utilize certain instructional
delivery systems and methods more than others. Fisher
(1983) summarized the research concerning computer assisted
instruction and indicates effectiveness when the following
conditions are met: 1) when it is aimed at specific student-
body groups (i.e most effective in raising achievement among
low-achieving and high-achieving students regardless of
whether the "disadvantage" causing the low achievement is
physical or social), 2) when it is fully integrated into the
regular classroom curriculum (i.e., CAI was found most
effective as a supplement to regular classroom instruction),
3) when certain subject areas are selected (CAI was shown to
be almost always ffective in the areas of science and
foreign language) The current study meets these
conditions. I
Microcomputer Database Research.

Database programs (i.e. programs for data management)
refers here to software that allows the user to manage
electronically filed data using a computer. Data management
involves performing tasks that are similar to recording and
manipulating information on index cards; planning how to
record the information, recording it, organizing the cards
in some order, seeking relationships among variables,
finding a particular card in the filebox, reorganizing the
cards, updating a card when the information is obsolete, and

developing a report based on the information on the cards.

26

Much of what has been written concerning computer—based
information technology supports the belief that survival in
the modern world will require citizens to access and solve
problems using information which has been stored and/or
transmitted electronically. This belief follows earlier
rationales for teaching processing skills, but focuses on an
additional concern about the amount and rate of information
flow that is accessible by the citizen (Becker, 1982; White,
1985).

The research base concerning database use in education
is nearly nonexistent. The only study found to date that
investigated certain processing skills in social studies
within a microcomputer database environment was a
dissertation by White (1985). He used Scholastic Publishing
Company’s PFS: Curriculum Data Bases for U.S. History and
for U.S. Government (Hunter and Furlong, 1985a and 1985b),
PFS: File and PFS: Report, a general file-management program
and its associated report generator (PFS: File, 1984 and
PFS: Report, 1984), and printed support materials as the
curriculum base for the research. (PFS: File is similar to,
and functions much like, the Appleworks database used in the
current research; the U.S. History and U.S. Government
databases are similar to the Climate and weather databases
in the current study.) These database management systems
allow the user to store, manipulate, and retrieve data

stored in related files. (In the case of the current study,

27

weather and Climate database files were used by students for
these activities.)

White (1985) used the database and materials for a two
week time period, in which the software was introduced, and
then activities where presented to students in grades 8-12.
He hypothesized why these type databases, used according to
his design, promote learning of process skills. Both the
print materials that were used by the students to lead them
through a step—by-step group of activities, and the computer
program itself, provide practice using the same structure.
He states:

"The operation of the software imposes
structure on the manner in which students
enter, organize and retrieve data. In
specifying the criteria for data searches,
students are required to consider explicitly
their information needs with respect to both
relevance and sufficiency. Students must
also specify in detail how they wish data to
be displayed as output; again, explicit
consideration of alternative organizations is
necessary as part of interacting with the
software." (p. 10).

White (1985) further comments that information-
processing theory, and the related research, is the basis
the instructional methods and materials used in his
research. He states that computerized databases "are well
suited to serve as memory sources for student problem
solvers, not only as repositories of information but as
models for information storage, retrieval and organization"
(p. 32). He goes on to point out that the materials

accompanying the database management software must provide a

balance between explicit, direct instruction and discovery.

28

The former guides students through efficient search and use
of information for specific problems; and the later enhances
the possibility of transfer to new problems of the kind most
frequently encounter in social sciences.

Results from the White (1985) study found students
receiving the treatment involving the use of a computer-
based file management program, in concert with structured
activities, "achieved higher scores on an information
processing instrument than their (control group)
counterparts" (p. 91). White’s (1985) study supports the
view that the microcomputer can be used as a tool in
teaching processing skills in social studies.

A number of science educators believe the use of
databases help students develop and practice the process
skills needed by a scientist. McLeod and Hunter (1987)
state that scientists often ask questions and construct
hypotheses based on their prior knowledge. With the help of
research assistants, scientists often design experiments,
gather relevant data, organize that data in ways that will
help support their hypotheses, reorganized the data, analyze
the data, draw inferences, and modify their inference(s).
These are the processes a student using a database can
utilize when experimenting and seeking patterns or

relationships among the data.

29

Summary of computer related literature.

The new technologies challenge science educators to
find ways to revise curriculum that will utilize the
capabilities of those new tools in student learning.
Research points to some conditions (e.g., science curricula;
low achieving and high achieving students) that seem to be
more suited for learning with computers than other school
situations. In addition, the White (1985) study supports
the notion that microcomputers used as tools can positively
affect achievement under condition similar to those used in
this research.

However, no instructional delivery system to date has
been shown as best in all settings with all students - and
computers are no exception. As might be expected,
computers are not seen by every educator as a panacea. The
growing use of microcomputers for public school instruction
has raised both pedagogical and social policy concerns. On
the pedagogical side are questions about the instructional
capabilities of computer hardware/software, and how teachers
can use the potential of computers successfully. On the
social policy side, a concern is for equity of access to
microcomputers, particularly for minorities and females.

One reason equal access to computers may be important
throughout schooling is that often early education
influences later educational and occupational opportunities
and choices (Stasz, Shavelson, and Stasz, 1985). Along with

the promise of better deliver of well designed instruction

>1

If‘\

m In ’1‘

r—J

30

using computers comes the concerns of equity (e.g., how to
insure computer use narrows the male/female science
achievement gap rather than widening it), and considerations
regarding how to group students when using the limited
number of computers currently found in schools. These
issues will be reviewed in the last two sections of this
chapter; and it will become clear that they are of direct
interest to this research project.

The next section reviews gender related differences;
first in science education, and then in computer education.
Gender Related Differences.

Gender Related Differences in Science Education.

There is a long history of reported differences between
boys and girls in interest and achievement in mathematics,
science, and related disciplines. Historically, careers in
science have been dominated by men (Burton, 1979; Fennema,
1980; Burlin, 1976). There has been much concern expressed
recently regarding the unequal numbers of male vs. female
students in the various science content areas; and later in
life the lack of female scientists.

Steinkamp and Maehr (1983) state:

An ongoing argument in educational
circles concerns whether one should
stress the development of proficiency in
the hope that motivation will follow or
stress the development of positive
feelings in the hope that this will
encourage the development of
proficiency. This argument takes on a
special form in the case of observed
male/female differences in science
achievement. There is little question

that women have not achieved in the area
of science to the same degree men have

31

(cf., e.g., Steinkamp & Maehr, in

press). A major cause is thought to be

attitudinal: Females simply do not like

science as well. The implication of

this conclusion is that science

instruction ought to focus especially on

affective outcomes (p. 369).
Before settling on this instructional strategy, however,
Steinkamp and Maehr (1983) suggest examining the research
literature regarding gender-related differences and science
achievement. Unfortunately, that body of literature does
not speak with one voice. Using meta-analysis, they
conclude that, in accord with other reviews, boys do better
in school science than girls do. The differences are
slight, but they appear to be reliable.

In a re-analysis of the studies in the above cited
Steinkamp and Maehr study, Becker and Chang (1986) found
much of the gender differences within subsets of the
original studies could be explained in part by the science
subject matter being tested and also on the type of measure
used in the studies. Gender differences for all subject-
matter groupings except for studies of general science were
found to be consistent, and the average gender differences
are less than one half of a standard deviation. In physics
and biology, it was concluded that males tend to do
significantly better than females by about one-third of a
standard deviation for physics and about one sixth of a
standard deviation for biology. There was no significant

differences between males and females on either geology

(e.g., weather and climate) or chemistry. The authors

In

In

32

suggest that, since the degree of gender differences in
achievement varies significantly across subject-matter areas
in science, "care should be taken to distinguish between
content areas when discussing or researching science
achievement and gender" (p. 17).

Becker and Chang (1986) subdivided the general—science
studies according to the school grade of the subjects.
Subgroups of elementary schoolers (grades 1 through 6),
junior-high students (grades 7 through 9), senior-high
students (grades 10 through 12, and college students, showed
only studies of junior—high groups had a common population
effect size. It indicated more than a quarter of a standard
deviation advantage for males.

Other research suggests sex-related differences are
dependent on what type of science performance is being
measured. When types of science performance were analyzed
separately by Kahl, Malone & Fleming, (1982) modest effect
sizes were found in favor of middle school boys for
application problems only; effect sizes for knowledge,
comprehension and higher order processes were minimal for
these students. Meehan (1984) also reports modest effect
sizes in favor of boys in proportional reasoning, but no sex
effects for either propositional logic or combinations
tasks.

Studies of a 1980 Science in the Schools survey of 13
year old students in England (Schofield, Murphy, Johnson, &

Black, 1982); and a 1981 Science in the Schools survey of 11

e
m ..

k .
«L

J ‘
(is

b.

33

year olds in England, Wales, and Northern Ireland (Harlen,
Black, Johnson, & Palacio, 1983) reported few sex
differences. However; girls were superior to boys in
planning investigations, while boys performed better than
girls in the application of knowledge of scientific facts
and principles in the physical science area.‘

Gender Related Differences in Computer Education.

Gender differences in computer education have received
much attention but little systematic research (Anderson,
1983). Studies by Lockheed, Nielsen, and Stone (1983) and
Anderson, Klassen, Krohn, and Smith-Cunnien (1982) report
that young women in secondary schools are less likely than
men to spend time with computers and to enroll in computer
classes. In addition, the 1981-82 National Assessment in
Science provides data showing a substantial gap between
females and males in signing up for computer programming
classes. Females are less likely to take these courses than
are males; 8 percent of the females and 14 percent of the
males have enrolled in a programming course for at least one
semester. A gap has been evident since 1978.

Computers, sometimes referred to as "numbers
crunchers", are often associated with mathematics. Research
supports the notion that sex-stereotyping leads to females
having less confidence in their ability to learn mathematics
than their male counterparts. Mathematics serves as a
filter into the sciences; therefore, math avoidance

prohibits female entrance into these fields (Mathews and

PM

In

5d-

(1)

34

Winkle, 1982). This filtering out of technological and
scientific options poses a major barrier to women’s
occupational and economic equity.

The pattern of avoiding mathematics and science can
lead to computer anxiety; e.g., the feeling that computers
are too complex to be understood by the average woman. Too
frequently, women do not take advantage of opportunities for
learning about computers as evidenced by the fact that
computer courses at all levels remain predominantly male.

One area in which there is some evidence that this
pattern of avoidance does not relate to gender is in using
computers as a tool to accomplish other tasks. Tool uses of
computers at schools or in places of business do not seem to
show sex differences (Lockheed, 1985). Some reasons
Lockheed (1985) suggests for this are: 1) girls are more
interested in tool use than are boys, 2) activities
supported by tool software, (i.e., drawing; sending and
receiving mail), are not sex stereotyped, and 3) tool use is
perceived as more relevant to future activities and
occupations than other instructional uses for computers,
(e.g., programming).

Finally, a study by Webb (1984) investigated the
effects of the gender composition of a group on achievement
and interaction patterns. The gender composition factor
varied the ratio of females to males in the group. Three
kinds of mixed-gender groups were studied. The types were:

1) groups with two females and two males, 2) majority-female

35

groups (typically, three females and one male), and, 3)
majority-male (typically, one female and three males).
Achievement in the Webb (1984) study depended on the ratio
of females to males in a group. _The achievement of females
and males was nearly identical in the groups with two
females and two males. In the majority-female groups and
the majority-male groups, however, the males showed higher
achievement than the females.

Summary of gender-related literature.

With regard to gender-related differences, Becker and
Chang (1986) were able to explain much of these differences
normally found in studies involving gender and science
achievement using more sophisticated meta-analysis
techniques than earlier reviews used. Gender-related
differences favoring males were still found to be
significant in middle school students. Therefore, gender
was selected as one of the independent variables in the
current study, given the pervasive nature of gender—related
differences in overall science achievement research favoring
boys, especially in middle school students.

(Furthermore, it is important to note that in science
boys are superior to girls on tasks that require knowledge
of or familiarity with stereotypically male objects or
apparatus, but no sex differences are observed when the task
is more gender neutral.j There seems a need for research to

report treatments showing no sex-related differences in

36

hopes of identifying more gender-neutral activities in
science.

Even though the opportunities for computer learning in
schools are increasing, inequities continue. Low-income,
female, and rural students are especially disadvantaged in
receiving computer experiences in school. To the extent
that computer literacy and computer expertise are necessary
for success in getting employment, computer inequity is a
serious problem (Anderson, et al., 1983). If males and
females participate differentially in computer learning
environments, this could lead to corresponding differences
in cognitive attainments and career access.

Hawkins (1985) has hypothesized that because computers
are often incorporated into math or science curricula (e.g.,
Saunders, 1978), there are serious consequences for girls.
Since computers are often linked with
science/math/technology into educational environments long
dominated by males, computers typically enter the classroom
with an aura of sex-related inequities that have an impact
on learners. The current study is designed to compare male
vs. female achievement. One implication of the analysis of
this comparison is to discover whether or not support is
found for the type of instruction used in this research
being gender-neutral.

The other study that had a direct impact on the design
of this study is the Webb (1984) study showing that when

females outnumbered the males or were outnumbered by the

.r\

37

males in a group learning experience, the female's
achievement was lower than the males. Furthermoreivwhen
females and males were of equal number in a learning group,
gender-related achievement differences were not significant
in the Webb (1984) study. These findings support the use of
equal numbers of males and females in the groups with two
and four students in the current study.

Learning in groups.

While research findings are not consistent concerning
group learning, overall the evidence points to higher
achievement by most students in many school situations
resulting from students placed in heterogeneous learning
groups; especially when students learn within cooperative
groups (Johnson & Johnson, 1978b; Johnson, Skon, & Johnson,
1980; Skon, Johnson, & Johnson, 1981). Conversely, there is
evidence to support the position that homogeneous groups
seem to be less effective in certain high-level tasks (Lorge
and Soloman, 1959).

Research does suggest, however, that within
heterogeneous small groups, learning may be differentially
effective for students with different skills and
backgrounds. When learning is done in small groups, a
number of interrelated issues need explored. Two important
variables are group size and group composition. The
following subsections contain a review of relevant
literature on these two variables and relates that

literature to the current study.

l(\

in

if)

*(3

(I)

The.

38

Group Size.

The number of students within a learning group has
several important implications for academic achievement,
cognitive development, and socialization. Optimum group
size depends on the group’s task, composition of members,
time available, level of social skills of students, and many
other factors.

A number of research efforts have focused on the
effects of instructional group size on learning in non—
computer environments. The results are not consistent.
Students working in groups have shown greater gains than
individuals working alone in some cases but not in others
(Trowbridge and Durnin, 1984). Klausmeier, Wiersma and
Harris (1963) concluded from their work with groups of
various sizes that students working in pairs and quads
grasped concepts faster than individuals. However, on
transfer tests, individuals learning alone generally showed
greater concept retention than students who had learned in
groups.

There is some evidence that the computer may promote
peer teaching and collaborative learning. Bracey (1984)
cites a study by Chernick and White which found fifth grade
students working with computers collaborated three times
more often than students using traditional instructional
materials.

Trowbridge (1982) and Trowbridge and Durnin (1984)

reported gains among adolescents working in groups of 2-4

f‘1

FY

(V)

f).

39

with computers on scientific activities. Trowbridge and
Durnin (1984) found that pairs performed more interactions
than individuals, triads or quads among undergraduates using
physics simulations. Trowbridge concluded that small groups
working with computers constitute a unique system which
requires further study before the potential of computer
based education is realized. They failed to find evidence
of achievement differences between subjects working alone or
in any size groups.

Okey and Majer (1975) also found no significant
difference in achievement when comparing students studying
alone with students in pairs, or students in groups of 3 or
4. Students worked together at the PLATO IV computer
assisted instruction terminal and then completed criterion-
based tests individually on materials covered. The
researchers did conclude that time of learning the materials
was reduced by students working in groups of 3 or 4.

The only direct evidence linking group size with
achievement within a mipppcomputer environment identified in
the review of literature was a study by Cox and Berger
(1985). In their study of seventh and eighth grade students
solving problems, they found children who worked on
computers stayed on task longer and reached a correct
solution quicker than students working without computers.
The five skills focused on in their study were: 1)
collecting data, 2) organizing information, 3) analyzing

information, 4) developing alternatives, and, 5) selecting

a:

DA

40

the most appropriate solution. Further, their findings
suggested groups of two or three members produced more
correct problem solutions than individuals working alone or
in groups of five. They concluded "teams of two to four
would seem best suited to work together to solve problems
similar to those in this study" (p. 467).

Group composition.

Whenever more than two persons form an instructional
group, overall diversity increases. One question facing
teachers concerning group composition is whether students
should be placed in homogeneous or heterogeneous groups.
Traditionally, students have been grouped according to
ability into separate classrooms or within the classrooms.
The rationale for this practice is that narrowing the
ability range in the classroom, or within a group of
students, facilitates the provision of more appropriate
learning tasks, makes more teacher time available to
students of a given ability level, and stimulates teachers
to gear their teaching to the level of the group (Goldberg,
Passow, & Justman, 1966).

However, many research findings suggest higher
achievement will result when students are placed in
heterogeneous, cooperative groups (Johnson & Johnson, 1978a;
Johnson, Skon, & Johnson, 1978; Skon, Johnson, & Johnson,
1981; Wodarski, Hamblin, Buckholdt, and Ferritor, 1973).
Hoffman and Maier (1961) found four-member heterogeneous

groups consistently scored as high or higher than

41

homogeneous groups in a study investigating problem-solving
tasks. Furthermore, there may be a variety of experiences
important for socialization and cognitive development in
classrooms where students of various characteristics
interact and learn together. In a series of studies,
Amaria, Biran & Leith (1969) found learning in small mixed-
ability groups was better than individual learning,
especially for low ability students. They hypothesized that
small mixed-ability groups permitted teacher-learner
relationships to develop between high ability and low
ability students in the groups.

Along this strategy of heterogeneous ability grouping,
research has suggested learning may depend on student
characteristics or aptitudes. Aptitudes refers here to any
characteristic of a student that predicts his/her
probability of success in a given instructional approach,
(e.g., abilities, motivation, attitudes, references). That
is, differences in a student aptitude may interact with an
instructional approach to produce differential achievement.
This phenomena has been called aptitude-treatment
interactions (ATI) (see, for example, Cronbach & Snow,
1977). Results from a study by Webb (1977) found overall
learning in small mixed-ability groups was more effective
compared to individual learning. Webb’s study also found an
ATI which can be summarized as: High ability students did
equally well after learning in small mixed-ability groups or

individually; medium ability students did better after

42

learning individually; and low ability students did better
after learning in small mixed-ability groups.

Peterson (1981) also concluded from results of two
studies involving mathematics learning that there did seem
to be evidence for the existence of a curvilinear ATI for
ability when the treatments involved having a student
working either alone or in a small mixed ability group.
High ability and low ability students benefited from the
small-group learning and medium ability students did
slightly better working alone.

Summary of learning in groups.

Several reviews of cooperative learning indicate a
growing research interest in instructional uses of peer work
groups (Sharan 1980; Slavin 1980). In many educational
settings, however, peer interaction is eliminated through
the use of competitive and individualistic instructional
procedures (Skon, Johnson, and Johnson, 1981), (e.g., such
individual instruction is often suggested for learning with
computers). There is evidence that cooperative learning
among peer work groups promotes higher achievement than do
other type efforts, (Aronson, Bridgeman, & Geffner, 1978;
Buckholdt & Wodarski, 1978; DeVries & Slavin, 1978; Johnson
& Johnson, 1978a, 1978b; Slavin, 1978), especially in
context of higher order skills learning (Skon, Johnson, and
Johnson, 1981). There are issues that need clarified. A
number of variables suggested in the literature may mediate

the relationship between cooperation and achievement. Three

43

of these variables that are hypothesized to be important to
the current study are: 1) group size, 2) group ability
composition, and, 3) gender.

The experimental literature suggests that group
investigation methods are particularly appropriate for
pursuit of the type outcomes investigated in this study,
(i.e. higher cognitive learning goals). Sharan, Ackerman
and Hertz-Lazarowitz (1980) found no significant differences
between those who worked individually and those who worked
in small groups when learning low level information. When
learning of higher level concepts was examined, however,
they found groups did better.

All things considered, the evidence concerning group
size indicates that the optimal size of learning groups
within the classroom might be from 4 to 6 members. It was
hypothesized that such a group is large enough that enough
resources are present for achievement gains, and is small
enough that everyone’s resources are utilized and that
everyone can participate (McMillan, 1980). However, when
students are very young, or when there is a serious lack of
social skills necessary for working with others, pairs and
triads may be more productive.

The review of literature concerning group composition
suggested there seems to be a curvilinear ATI for ability
when the treatments involved having a student working either
alone or in a small mixed ability group. High ability and

low ability students benefited from the small-group learning

44

and medium ability students did slightly better working
alone. One purpose of this research was to explore further
the possibility of aptitude-treatment interactions when
different numbers of students form groups and work
individually or together in small mixed-ability groups.
Chapter Summary.

Science teaching seems well suited for study, since it
emphasizes problem solving and inductive reasoning within a
subject domain. Rubinstein (1975) describes problems as one
of two kinds: 1) problems requiring synthesis to solve, and,
2) problems requiring analysis to solve. The problems
requiring analysis focus on the application of known
transformation processes to achieve the obscure or hidden
solution. The emphasis in this type problem solving is on a
set process from a known initial condition (i.e. initial
state). This is the kind of problem most often presented to
the scientist. It is also the type of problem hypothesized
to be well suited for practice using a microcomputer and
database management application program like the one used in
the current study.

As computers enter the school systems, one challenge
for educators is to design instruction which utilizes these
powerful tools in effective ways to meet their instructional
goals. Given that one important goal in science education
is the teaching of inquiry skills, evidence exists which

shows computers can be used to meet these goals.

45

However, as with any innovation, care must be taken to
investigate possible harmful (albeit unintended) results
(e.g., widening gender equity problems), as well as the
intended benefits. In addition, given the limited number of
computers currently in our schools, it is important to
consider issues related to groups of students learning at
computers. The current study is designed to investigate a
number of questions with regard to achievement and how it is
affected by group size, gender of the students, and the
ability of the students within groups.

The evidence for computer learning in groups is
somewhat conflicting and difficult to interpret. In
computers, boys were superior to girls in the few studies
where performance was assessed; sex differences were small,
however (Lockheed, Thorpe, Brooks-Gunn, Casserly, & McAloon,
1985). Although most researchers report anecdotal or
intuitive support for positive outcomes resulting from
learning in microcomputer environments, it is difficult to
find hard data in support of such claims. There is,
however, some evidence that microcomputers can provide an
environment for the learning of science process skills
(Berger, 1982). In general, the perspective used in this
study was that the computer is a vehicle used to deliver
instruction (Clark, 1983) and viewed as a tool (Taylor,
1980) has no significant effect on science achievement in

and of itself.

46

With regard to gender differences, it may be unfair to
ask girls and women to change their math/science/computer
avoidance patterns without changing some of the influences
operating on them (Mathews and Winkle, 1982). There exists
a need to design and develop materials that will promote
opportunities for all students - and especially females — to
learn processes of science. This study investigated one way
to intervene using microcomputers as tools that are
integrated in the science curriculum, and reports on any
sex-related differences found in achievement.

Finally, regarding group size and the current study, it
is not as important that students working in groups do
better than students working alone, as it is important for

them to do equally well.

CHAPTER THREE . METHODS .

This chapter describes the selection of students in the
sample, and the method of assigning students to treatment.
It also describes the training students and teachers
received, the measurement instruments and materials used,
and the research design in the study.

Population and Sample.

There were twelve science classrooms from three schools
in three different school districts involving 306 students
taken from middle school grades (i.e. 7th and 8th grades).
These classes were selected on the basis of the number of
computers available in the classroom and the teachers'
volunteering to participate in the research. The twelve
classrooms were assigned randomly to the three treatments
described below. Each of the treatments received four
classrooms each.

Training.

Before the science instruction began, teachers and
students were given instruction and practice in using the
electronic database, (McLeod, Hunter, and Finkel, 1987),
necessary to complete the lessons that followed. Teachers
were given a checklist as well, on some "do’s and don’t’s"

concerning implementation of the research lessons, (e.g.,

47

48

encourage students in a group to discuss the activities;
encourage students to make sure all members of their group
understand and participate in each activity) (See Appendix A
for checklists). There was a separate checklist to be used
by the teacher for classes in which students work in groups
and for classes in which students work individually.
Instrument.

Assessing student ability in science process skills can
be difficult and time consuming through observation of
laboratory situations (Burns, Okey, and Wise, 1985). Due to
time requirements in public schools, quality tests are
needed to measure pupil performance without always making
observations of process skills in the laboratory.

Initially, efforts to design tests to measure process
skills were tied to the inquiry oriented curricula prominent
in classrooms in the 19605 and early 19705 (e.g., Walbesser,
1965; McLeod, Berkheimer, Fyffe, and Robinson, 1975;
Ludeman, 1975; Riley, 1972). Later, tests were developed
that were not curricular specific, but were targeted toward
upper elementary and middle school students (e.g., Molitor
and George, 1976; Tannenbaum, 1968). The Test of Integrated
Process Skills (TIPS) was developed by Dillashaw and Okey
(1980) to respond to the need for a non—curricular specific
process skills test for middle grade and secondary students.

The Test of Integrated Process Skills (TIPS); and the
Test of Integrated Process Skills II (TIPS II) (Burns, Okey

and Wise, 1985; Dillashaw and Okey, 1980; Tobin and Capie,

49

1982) were used to measure the outcome variable (i.e.
integrated science process skills score). Together these
two paper and pencil examinations serve as criterion
referenced, alternative test forms to measure integrated
process skills. Total test reliability using Cronbach’s
alpha was reported by Burns et al. (1985) for TIPS as .82
and TIPS II as .86. The researchers summed up their
findings concerning the comparisons of the two tests as two
highly equivalent tests which related to the same objectives
and produced highly similar mean scores.

One form of the test was selected randomly to be used
as a pretest in all classrooms and the remaining form of the
test was used for the post-test. The pre/post-test design
was used for two reasons: 1) the pretest served as the
measure to rank order students within classrooms according
to ability levels (i.e. high, medium, and low ability in
science process skills), and 2) the pretest was used to test
the overall effectiveness of learning during the
intervention. Both forms of the test have been content
validated against nine objectives. The developers believe
that collectively the tests measure skills involved with
planning, conducting, and interpreting results from
investigations. The lessons developed for this study
(McLeod, 1987) (See Appendix B) gave the student practice in
the four objectives listed below. Therefore, pretest and
post-test items identified by the tests developers as those

which measure these four objectives become the focus of

50

analysis in this research. These four objectives are
(Dillashaw and Okey, 1980):

1) Given a description of an investigation, identify
the independent, dependent, and controlled
variables and the hypothesis being tested.

2) Given a problem with dependent variables specified
and a list of possible independent variables,
identify a testable hypothesis.

3) Given a problem with a dependent variable
specified, identify a testable hypothesis.

4) Given a hypothesis, select a suitable design for an
investigation to test it.

These four objectives, as opposed to the entire nine
objectives, were selected in part because in the two week
time period of this study, only a certain number of
objectives are likely to be learned by the students.
Materials.

Curriculum. Lessons were developed (McLeod, 1987) to
match the four objectives described above for the outcome
tests. These lessons were designed to be used by the
student with little help from the teacher, thereby reducing
teacher variation as much as possible.

The lessons were developed to be used with Appleworks
(Lissner and Apple Computer, Inc., 1983) and the Climate and
weather databases (McLeod et al., 1987). Appleworks is an
integrated computer program that includes three functions:

1) word-processing program (e.g., for typing and editing

51

letters and reports), 2) database program (e.g., for keeping
track of information like mailing lists or inventories),
and, 3) spreadsheet program (e.g., for creating mechanized
worksheets for accounting, or doing other tasks that arrange
information into rows and columns). Of primary interest to
this study was the Appleworks database program which allows
the user to enter, in an organized way, information such as
customer files, elevation of world cities, inventories, or
stock portfolios. The user can retrieve that information
sorted and arranged in whatever way needed. For instance,
lists can be arranged in alphabetical order or numerical
order, and in either ascending or descending order.
Searching and sorting can be done using multiple criteria.
For example, if trying to determine whether there is a
pattern of temperature shifts that are associated with
approaching pressure centers, one of the activities the
experimenter might wish to conduct is to select only those
days when the pressure change is decreasing gpg the
temperature is increasing. If investigating the
relationship between wind direction shifts and approaching
pressure centers, the user might choose only those days when
the pressure change is decreasing apd the wind is from the
northerly direction. (See the section later in this chapter
on the format used in each lesson for more on how databases
can be used to teach process skills.)

Weather is defined in the weather database as the

condition of the atmosphere over a short period of time, as

52

opposed to climate, which is the average weather conditions
over a prolonged period of time. Students use weather data
for one location, Grand Rapids, MI, USA. The data were
obtained from the National Oceanic and Atmospheric
Administration (NOAA) and include weather data collected
every three hours for the months of July and January. The
authors stated that they selected these two months to
illustrate the differences in weather between them. A
sample of the kinds of weather data collected for the two
data files used in lessons prepared for this research are:
Each record in the Local database file is for the same
location, Grand Rapids, MI, but for a different time and
date. For each day, there are eight records (1 am, 4am,
7am, 10am, 1pm, 4pm, 7pm, and 10pm). Each record contains

the measurements for that time period regarding: Sky Cover,

 

sky cover Ceiling, Wind directions, average wind 8 eed, and
precipitation.

Each record of the Skytemp database file is for the
same location, Grand Rapids, MI, and include temperature and
sky cover readings for a day, the minimum and maximum sky
cover for the day, and a calculation to show the temperature
variance during the day.

Climate in the Climate database is defined as the
average weather conditions over a prolonged period of time
as opposed to weather, which is the condition of the
atmosphere over a short period of time. The author of the

database states the locations for the Climate database were

53

first selected so as to represent general latitude
considerations. Locations were added to insure that water
and mountain effects would be sufficiently represented to
illustrate their contribution to climate. Finally,
locations were added to insure that every state was
represented. A sample of the kinds of climate data
collected for the two data files used in lessons prepared
for this research are: geographic information on the
latitude and longitude, the normal temperature each month,
the precipitation, the wind direction, the yearly average
temperature, the minimum temperature, maximum temperature,
and a calculated difference between the minimum and maximum
temperatures.

Based on his experience with developing and testing
similar instructional materials, the author believed that
the curriculum activities accompanying the databases
described above allowed students to practice integrated
process skills. The author stated that these lessons were
specifically designed to have students practice problem
solving through tasks which include: 1) determination of the
data needed, 2) organization of the data, 3) performance of
mathematical operations on the data if necessary, and, 4)
analyzing the data.

Lessons follow a general format that includes the
following sections:

1) Purpose - States the purpose of the experiment.

Generally, this is what the experimenter expects

 

54

to find. For example, "To determine if there is a
relationship between temperature and latitude.

2) Hypothesis - A statement of the purpose in a way
that suggests what to look for. For example, "As
the latitude decreases, (as we go South), the
temperature will increase."

3) Controlling Variables — Lists (or asks the student
to list and question why) variables which may also
affect the results in the experiment besides the
variables of interest stated in the hypothesis.
For example, when attempting to determine if there
is a relationship between temperature and
latitude, altitude (or mpppp) must be controlled
for.

4) Determine the Information Needed - Asks the student
to list the information needed to test the
hypothesis. For example, when attempting to
determine if there is a relationship between
temperature and latitude, the experimenter needs
at least the following information, a) temperature
and b) latitude.

5) Arrange the information - Asks the student to list
how the data is to be arranged, or reported. For
instance, the way the information is arranged on
paper is very important in helping interpret the

data. Assuming the data is arranged in columns,

it becomes easier to find relationships. If

 

55

latitude is arranged in decreasing order in a
column, and the experimenter asks for temperature
in an adjacent column, if the hypothesis is
correct, the data in the temperature column should
fall into a pattern showing an increase.

6) Analyzing the Results - The experimenter includes a

statement of a) whether a relationship was found,
b) if possible, whether the relationship seems to
be a direct or inverse one, and c) if the
relationship is not perfect, note some of the
exceptions and state some of the inferences for
them. For example, in the experiment on latitude
vs. temperature, this section would cue the
student to analyze the data reports to determine
items a through c above, and focus on the
anomalies (e.g., larger due to water effects).
Procedures and Design.

Assigning Students to Groups.

Each classroom of students was rank ordered within
gender, from low to high on their ability concerning
integrated process skills based on their pretest scores.
The research design that follows called for three levels of
treatment using individual students, pairs of students and
groups of four students working at microcomputers. The
twelve classrooms used in the study were randomly assigned
to treatments with four classrooms each for individual

learning, learning in pairs, and learning in quads.

56

However, each of the three different schools had at least
one classroom assigned to each of the three treatments.
This random assignment to treatment was designed as one
control for teacher, school, and classroom effects. In all
cases, a student's sex and ability level was recorded.
Groups of two and four were formed using random stratified
sampling on ability.

Since the overall weight of the evidence indicates that
higher achievement is achieved by high, medium, and low
achieving learners when they are placed in heterogeneous,
cooperative learning groups (Johnson & Johnson, 1978;
Johnson, Skon, & Johnson, 1980; Skon, Johnson, & Johnson,
1980; Wodarski et al., 1973), no homogeneous ability groups
are intentionally investigated in this study.

Furthermore, the review of literature suggested using
an equal number of males and females within a cooperative
group. Therefore, groups were formed based on equal numbers
of boys and girls, randomly chosen as permitted by the
natural classroom demographics and the rules for
heterogeneous ability grouping discussed below. The
following outlines the grouping procedures in more detail

(L=low ability; M=medium ability; H=high ability):

57

M H (student group)

_____ |____—————|--——---—| achievement

gradient

50 75 100 (percentile)

Figure 1. Gradient Scale

Groups with 2 and 4 members were formed using the

method shown in Figures 2 and 3 below. In all cases,

G=girl, B=boy,

L=low ability, M=medium ability, and H=high

ability.
Student Ability L M H

Group

Number
1 G B
2 G B
3 B G
4 B G
5 B G
6 G B

Figure 2.

Plan to form groups with 2 students

58

Student Ability L M H
Group
Number

1 G BG B

2 B BG G

3 G BB G

4 B GG B

Figure 3. Plan to form groups with 4 students.

Data Analysis.
A split-plot, multivariate factorial design was planned
to analyze the data for main and interaction effects. The

design can be represented by Figure 4.

59

T3 I

C9 C10 C11 C12 |

 

 

 

 

(S=sex; A=ability; C=classroom; T=treatment)

Figure 4. Within-subjects design.

Chapter Summary.

A total of 306 middle school students made up the
sample. Training in the use of the software and database
was completed prior to the administration of the five
lessons and accompanying material that were the focus of the
comparative study. The Test of Integrated Process Skills

was used as a pre-test and post-test response variable.

60

Lessons, based on the Scholastic Weather and Climate (McLeod
et al., 1987) databases were developed for this research to
control for teacher variability across the 12 classrooms.
Students in the treatment classes that were designed to have
pairs and quads working at microcomputers were assigned
according to ability and gender as outlined in the
procedures section of this Chapter. A split-plot,
multivariate factorial analysis was planned to analyze the

data.

CHAPTER FOUR. RESULTS.

This chapter describes the subjects in detail, the data
collected, and the methodology used to analyze the data. It
also reports the results of data analyses.

Description of the subjects.

There were twelve science classrooms selected from
three schools in three different school districts at the
middle school grades (i.e., 7th and 8th grades) used in this
research. The total number of students involved was 306.
Selection was based on the number of computers available in
the classroom, and the teachers volunteering to participate
in the research. The twelve classrooms were assigned
randomly to each of the three treatments described below;
however, each of the three schools had at least one
classroom assigned to each of the three treatments. Of
course, a different teacher was found in each of the three
different schools. It was assumed that by assigning at
least one classroom to each of the three treatments for each
school, teacher effects may have been reduced. All three
treatments (i.e., individuals, or pairs, or quads of
students working at the computers) received four classrooms

each.

61

62

Of the 306 total students, 61 were dropped from the
study for various reasons. A summary of the sixty-one
dropped students is as follows: Seven students failed to
return their completed human subjects form (See Appendix C),
which was given to all students. Test scores of those
students failing to return the completed form were dropped
from the study. There was missing data (i.e., pre-test or
post—test scores), for four students; and three students
were absent from school for an extended period of time,
either during the intervention or during the time the post-
test was administered. These students’ scores were dropped
from this study. Therefore, a total of 14 students out of
the 306, (i.e., 4.6%) were dropped for reasons out of the
control of the researcher.

In addition, 43 (forty-three) students, due to the
demographics of individual classrooms (e.g., there were more
girls than boys in the class), could not be grouped
according to the rules of grouping described in Chapter 3;
and their scores were not included in this study. Four
students and/or their parents declined to allow their test
scores to be used in the research, as indicated on the human
subjects form. Since the identity of the four students
declining to participate in the research was known to the
researcher prior to grouping of students, these persons had
no detrimental effect on the design of the research. A
total of 245 students were the focus of the data and results

presented in this chapter.

63

Description of the Data Collected
The data used in this study were collected during

individual student pre- and post-treatment tests. There

were two parts to each of these tests; one part (I) was the

Test of Integrated Process Skills (TIPS), and the other part

(II) was a researcher developed test (See Appendix D).

Prior to collecting the data in the 12 classrooms which
were the focus of this research, a number of other
classrooms in school districts outside the main research
area were selected to test the implementation of lessons,
procedures, and evaluation design. From this pilot study, a
number of suggestions were made by the classroom teachers
and building administrators, the researcher, and the
researcher’s guidance committee members, to enhance the
procedures before implementing the lessons and evaluation in
the actual twelve classrooms that became the focus of this
research. There were a number of minor changes that may
have contributed to the student learning which took place at
the twelve research sites. Four suggestions that appear to
be significant are reported here:

1) Grading of the students in the pilot study was not done
by the classroom teachers. Because of this, it
appeared that a significant number of the students
involved in the pilot study did not take the lessons
and/or evaluation seriously. It was suggested that

students’ work at subsequent times (i.e., during the

64

actual research) would be graded as a regular unit in
their science curriculum.

2) The Test of Integrated Process Skills is a multiple
choice paper and pencil test, which measures the
objectives stated in Chapter 3. However, this test
does not follow the form or style of the practice the
students received when using the computer during the
research intervention. It was suggested an additional
response measure (i.e., besides the TIPS test) be
included in the pre- post-treatment testing that would
be a closer match, at least in form and style, to the
practice students would have during the intervention.

3) At more than one of the pilot study sites, the
intervention was interrupted by the students absence
from the classroom for scheduled school events (e.g.,
spelling bees, theatrical plays), or on vacation from
school. Through observations by the classroom teachers
affected by these interruptions and by the researcher,
it appeared that those type interruptions, or
anticipation of them by students, caused students to
take the lessons and/or evaluation process less
seriously than students not burdened by these
interruptions. It was suggested that the total
research intervention, including both pre- and post-
tests, be conducted during a period of the schools’
calendars that was not interrupted by days off (e.g.,

in-service training, parent conferences, Easter or

65

Christmas vacations), or the day or two adjacent to
vacations, unless absolutely necessary to schedule
during those times.

4) Of the 36 items in the Test of Integrated Process
Skills, twelve items dealt with objectives not included
in the focus of this study (e.g., graphing), as
discussed in Chapter 3. It was suggested that the 12
items that were irrelevant to this study be deleted
from the test to shorten it and save the students’ time
and energy.

In response to suggestions one and four above, students

were graded as part of their regular school works and the 12

items on the TIPS test that were not the focus of this

research were deleted from the pre- and post-tests.

Concerning suggestion two, subsequent to the pilot study and

prior to the formal study, it was believed that the TIPS

test (referred to in this Chapter as Part I of the pre- and

post-tests), may not promote the transfer of skills, (i.e.,

integrated process skills as practiced by the students

during the intervention), necessary to show significant
learning taking place. Therefore, an additional test

(herein called Part II) was developed by the researcher in

response to the suggestion for a closer match between

practice and testing. Four problem cases were identified by
teachers and the researcher from middle school science
textbooks and/or teaching experience. One group of problems

(i.e., two of the cases), in general, involved

66

weather/climate content. The other group of problems (i.e.,
the other two cases) involved topics often found in
textbooks directed toward middle school age students, but
were not in earth science. The four cases were designed to
follow the format and style of the lessons the students
practiced during the intervention. Face validity of this
part of the test (i.e., the evaluation by experts regarding
whether there is a match between the practice students
received during the lessons and the test’s objectives), was
established by independent evaluation and agreement of one
instructional designer and two science educators. One
question from each of these two groups was randomly assigned
as Part II of the pre-test, and the remaining two questions
formed Part II of the post-test. While still a paper and
pencil test, Part II of the pre- and post-tests matched the
practice the students had during the intervention period in
format and style and was believed to require less transfer
of training than Part One required.

In response to the third suggestion from the pilot
study, every attempt was made to schedule this research at
times in the schools’ calendars that were relatively
uninterrupted. This goal was achieved as much as possible,
with intervention and tests being conducted on continuous
days in each school.

Although the addition of the second response variable
(i.e., part two) is consistent with suggestions made during

the pilot study of these materials, it may be helpful to the

67

reader if more on the perspective taken concerning transfer
is briefly stated here.

Basically, the distinction Kelly (1967) made, between
prerequisite knowledge and beneficial knowledge is useful.
Prerequisite knowledge is essential for a certain piece of
new learning. Beneficial knowledge may be helpful for the
new learning but is not essential. Gagne (1974) makes a
similar distinction between vertical transfer and lateral
transfer for intellectual skills. Vertical transfer is
dependent upon prior learning of simpler skills. Lateral
transfer, or near transfer as Mayer (1974) called it, refers
to skills that are similar to those taught. It was believed
that part one (i.e., TIPS test) may require the student to
show integrated process skills regarding essential
knowledge. On the other hand, part two may require the
learner to show near transfer skills, or those skills that
are more similar to those practiced during the intervention
than those skills required by the TIPS. In addition, the
assumption is made that all learning involves transfer from
prior learning to a greater or lesser degree (Ausubel, 1963;
Voss, 1978).

As described in Chapter 3, each of the twelve
classrooms were assigned randomly (within school) to one of
three treatment groups, so that each treatment had four
classrooms total. Treatment One was individuals working at
the computer, Treatment Two was pairs of students working

cooperatively at the computer, and Treatment Three was quads

68

of students working cooperatively at the computer. Students
and groups of students will be referred to by these
treatment numbers throughout this chapter.

Students in all treatments were divided within
classroom by their sex, and within each sex students were
rank ordered by their achievement on the pre-test part one
scores. High achievement and low achievement were the upper
and lower 25% of students respectively, within each
classroom and within the appropriate gender. The middle
achievement group consisted of those students scoring in the
middle 50% on the pre-test, within each classroom and within
gender. Students in treatment groups two and three were
assigned to their learning groups based on the rules
described in Chapter 3, Methods. Low achievement was
assigned Ability One, middle ability was classified Ability
Two, and high achievement was assigned Ability Three; the
results are reported using those classifications in this
chapter. Males were coded one and females were coded two
for analysis purposes.

Research Hypotheses.

The null forms of the research questions of interest
are:

1) There is no significant difference in the learning
of science process skills between two-member
cooperative learning groups, four-member
cooperative learning groups and individuals who

work alone using microcomputers.

69

2) There is no interaction between high, medium, and
low ability students and group size on learning
science process skills within a microcomputer
environment.

3) There is no interaction between the gender of the
student and group size on learning science process
skills within a microcomputer environment.

Since this study was exploratory in nature and not
definitive (i.e., not an exact replication of any prior
study, or attempting to draw definitive conclusions),
attempts were made to discover relationships beyond these
proposed hypothesis that were plausible based on the
literature review. Two additional hypotheses seemed
promising and were tested for significance:

4) There is no significant difference in the learning
of science process skills between males or
females.

5) There is no significant difference in the learning
of science process skills by those students
showing low, middle, or high ability (i.e., based
on pretest part one rankings).

Concerning these two additional hypotheses, both
factors (i.e., gender and ability) were mentioned in chapter
one as important. Therefore, the two additional hypotheses
are of interest to the researcher and supported in earlier

chapters.

70

Design.
Kirk (1982, Chapter 11.11) supplies a split-plot

factorial model for viewing the research design. Since the
classes were assigned to treatment, the data was analyzed
using class mean scores as the raw data (i.e., classroom was
the unit of analysis), rather than individual scores on the
criterion measures (Page, 1965; Herron, Luce and Neie,
1976). Within each classroom, the various conditions
reflecting two levels of gender and three levels of ability
permits a multivariate treatment of the data. A MANOVA
procedure, using the multivariate set-up, was used to
analyze the scores on both parts of the post-test. SPSSx
MANOVA is a generalized multivariate analysis of variance
and covariance program which can be used to analyze designs
such as the split-plot design used in this study (SPSSX,
1986). This procedure performs univariate and multivariate
linear estimation and tests of hypotheses for any crossed
and/or nested designs, with or without covariates.

Using classrooms1 as the unit of analysis, treatments
(i.e., group size) is a between-subjects factor, and ability
and sex are within-subjects factors (Figure 4). A breakdown
of the raw data involving treatment x classroom x gender x
ability was used to aggregate the mean scores for the within
subjects factors (i.e., sex x ability) for this procedure.
Individual cells contained the mean scores across students.

1 There is not enough data in this study to separate the
classrooms within treatment source of variance from the
treatment alone (i.e., use a nested design with classrooms
Within treatment).

71

A MANOVA procedure using SPSSx was used, without the pretest
as a covariatez, to test the significance of the three
factors listed above, and interactions among those factors,

for the two parts of the post-test (see Tables 1 and 2).

TABLE 1. MANOVA results for Part 1 of post-test.

 

Source of Sign of
Variation SS DF MS F F
Test of significance of between-subjects effects.
Within Cells 104.90 9 11.66
Treatment 3.09 2 1.55 .13 .877
Tests involving 'Sex’ within-subject effect.
Within Cells 25.23 9 2.80
Sex 1.70 1 1.70 .61 .457
Treatment by Sex 6.65 2 3.32 1.19 .349
Tests involving ’ability’ within subjects effects.
Within cells 140.61 18 7.81
Ability 492.50 2 246.25 31.52 .000*
Treatment by

Ability 14.06 4 3.52 .45 .771
Tests involving 'sex x ability’ within subject effects.3
Within cells 123.35 18 6.85
Sex by

Ability 10.90 2 5.45 .80 .467
Treatment by Sex by

Ability 17.63 4 4.41 .64 .639

* significant p <.001

 

2 Keppel (1982) describes the case of using analysis of
covariance as being questionable in educational research
when intact classrooms are assigned to different treatment
conditions (See Keppel, 1982, Chapter 20).

3 Gender x ability is confounded by the grouping practice
used. Since males and females were rank ordered separately
within each classroom, there is no absolute criteria for
high, medium, or low ability across classrooms or within
classrooms. This design does not allow for meaningful
interpretation of gender x ability or gender x ability x
treatment interaction effects.

72

TABLE 2. MANOVA results for Part 2 of post-test.

 

Source of Sign of
Variation SS DF MS F F
Test of significance of between-subjects effects.
Within Cells 171.45 9 19.05
Treatment 10.18 2 5.09 .27 .771
Tests involving 'Sex’ within-subject effect.
Within Cells 193.52 9 21.50
Sex 19.64 1 19.64 .91 .364
Treatment by Sex 6.40 2 3.20 .15 .864
Tests involving ’ability’ within-subject effects.4
Within cells 253.29 18 14.07
Ability 575.18 2 287.59 20.44 .000*
Treatment by

Ability 20.54 4 5.13 .36 .830
Tests involving ’sex x ability’ within-subject effects.3'4
Within cells 144.11 18 8.01
Sex by

Ability 7.39 2 3.70 .46 .637
Treatment by Sex by

Ability 26.77 4 6.69 .84 .520

*

Significant p < .001

 

4 The univariate results of the analysis of repeated
measures designs have greater statistical power, and can be
used if the conditions of symmetry are met (SPSSX, 1986).
If these necessary and sufficient conditions are not met,
the multivariate test results should be used in assessing
effects.

In analyzing the results of the Mauchly sphericity
test, the conditions of symmetry appear to be violated in
both the ’ability’ and the ’sex x ability' within-subjects
effects on Part 2. However, the multivariate tests show the
same significant results as the corresponding univariate
tests. (i.e., p < .001 for ability as the only significant
result).

73

Results and Discussion.

To determine if learning occurred during the
intervention, a t-test for both parts of the pre- and post—
test design was conducted. Each part of the pre- and post-
tests was scored with 24 points being the highest. Results
for part one (i.e., the TIPS test) showed an average gain
between the pre—test and the post-test of 2.2612, which is
significant at the .001 level (See Table 3). Results for
part two (i.e., part designed for "near transfer" of
skills), showed an average gain between the pre-test and the
post-test of 4.6531, which again is significant at the .001

level (See Table 4).

Table 3. T-Test for Post-test Part 1 vs. Pre-test Part 1.

 

Standard Standard
Means Deviation Error
Post-test
Part 1 14.4286 4.516 0.289
Pre-test
Part 1 12.1673 4.647 0.297

Number of cases = 245

(Difference) Means = 2.2612

Standard Deviation = 4.275

Standard Error = 0.273

Correlation = 0.565

T Value = 8.28

Degrees of Freedom = 244

2-Tailed Probability (Difference in means) = 0.000

74

Table 4. T-Test for Post-test Part 2 vs. Pre-test Part 2.

 

Standard Standard
Means Deviation Error
Post—test
Part 2 8.3959 5.475 0.350
Pre-test
Part 2 3.7429 3.982 0.254

Number of cases = 245

(Difference) Means = 4.6531

Standard Deviation = 4.905

Standard Error = 0.313

Correlation = 0.499

T Value = 14.85

Degrees of Freedom = 244

2-Tailed Probability (Difference in means) = 0.000

It appears there is a main effect on ability, and no
other highly significant differences (See Tables 1 and 2
above). Since the literature on effects of group size
regarding learning is inconsistent, it was not known whether
a main effect of treatment would be found for either
response variable in this comparative study. Very little
research has been conducted within microcomputer
environments which reports on group size effects. This
research supports the position that groups of two and four
students learn certain problem-solving skills equally as
well as individuals working at computers under regular

classroom conditions.

75

Other results that were hypothesized included no
significant differences between males and females (i.e., no
main effect on gender), and no interaction effects of gender
by treatment. In fact, a number of procedures (e.g., an
equal number of females and males assigned to groups) were
deliberately used to promote this intervention as a gender-
neutral activity.

It was expected that high ability and low ability
students would be superior to middle ability students in the
small-group learning conditions; and that the middle ability
students would do significantly better when they worked
alone rather than in pairs or quads of students (i.e., an
aptitude by treatment interaction was expected). Contrary
to expectations, the results of this study showed no ability
by treatment interaction.

There were significant main effects on ability for both
parts of the post-test. Looking at the ability main effect
on Part one of the post-test, the gain scores were highest
for the lower ability group (4.63), less for the middle
ability group (1.96), and least for the highest ability
group (0.53). On the other hand, raw mean gain scores
across all individuals on Part two of the pre— post-test
resulted in the opposite trend. The mean gain scores were
3.72, 4.63, and 5.98 for the low, middle, and high ability

groups respectively (See Figure 5).

76

 

 

PART 1 PART 2
| I
G 6 I 7|
a I l
i 5 | 6|
n I I
4 | 5 |
S l |
c 3 I 4|
o I I
r 2 I 3 I
e l I
1 l 2 |
l =
0 | |

A1 A2 A3 A1 A2 A3
Ability Ability

Figure 5. Ability x Post-test Gain Scores

Perhaps the most unexpected finding in this study is
that in part one of the post-test, the trend was for low
ability students to gain the most and the high ability
students to gain the least, compared to the pre-test; while
in part two that trend was reverse. This might be described
as a post-test by ability interaction. Usually the term
"interaction effect" is used to describe the relationship
between two independent variables. However, here the term
is used in reporting a phenomena between an independent
variable and two different dependent variables. Since this
study was not designed to explore this interaction, the
researcher can only speculate about why this interaction

occurred.

77

It may be that low ability students gained the most
between part one of the pre- and post-tests because they
concentrated on the essential knowledge, (i.e., general
terms and concepts such as independent and dependent
variables, controlling variables), while the high ability
students already understood those general terms and
concepts. By definition, the highest ability students were
those scoring in the top 25% on Part One of the pretest,
within each classroom. Therefore, it is not surprising that
not much gain was made by high ability students on part one.
Even though there was no ceiling effect on either part of
the pre-test or post-test, it is still expected that the
high ability students are those who had already shown their
competency in the more general skills on the pretest.

On the other hand, given that the high ability students
had prior knowledge concerning the essential knowledge
regarding science process skills, they were able to
concentrate their efforts on learning the skills tested on
part two of the pre- and post—tests. The low ability
students, struggling with the concepts measured on part one,
were not able to concentrate on the specific tasks in the
lessons given them during the intervention. These low
ability students did not yet have the prerequisite
knowledge, (i.e., basic concepts) needed to understand the

more task specific skills presented in the lessons.

78

Summary.

This research could be described as using a split-plot,
multivariate factorial design for analyzing the data. The
independent factors included treatment, (i.e., individuals
vs. pairs vs. quads of students working at the computer), as
the one between-subjects factor, plus gender and ability
grouping as two within-subjects factors. Two response
variables were measured in both a pre-test and a post-test.
One of these response variables was the Test of Integrated
Process Skills (i.e., Part One), and the second response
measure was a researcher developed paper and pencil test
that more closely matched the practice in style and format
that the students had during the intervention.

The three null hypotheses that were developed prior to
the study were tested for significance. Two additional null
hypotheses were stated in this chapter, based on the review
of literature. These two hypotheses were also tested for
significance. The within-subjects factors were collapsed
for the analysis; and a multivariate F test statistic was
chosen to test these hypotheses using the MANOVA procedure
with SPSSX. This procedure benefited from the balanced
design afforded when within-subjects factors were collapsed,
and also was chosen since it has fewer assumptions
associated with its use than does other test statistics
which were considered.

When testing for main and interaction effects as

hypothesized, the only statistically significant result that

79

was shown in the analyses was a main effect on ability for
both response measures. Part one of the gain scores between
post-test and pre-test showed results that were highest for
the lower ability group (4.63), less for the middle ability
group (1.96), and least for the highest ability group
(0.53). On the other hand, analysis of the raw mean gain
scores across all individuals on Part two of the pre- post-
test resulted in the opposite trend. The mean scores were
3.72, 4.63, and 5.98 for the low, middle, and high ability
groups respectively. Speculation concerning this ability x

post-test interaction was discussed.

CHAPTER FIVE. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS.

This chapter summarizes the perspective taken by the
researcher and the procedures used during the research. The
findings reported in Chapter Four and the limitations of
this study are discussed. Finally, conclusions and
recommendations based on this study are made.

Purpose

The diffusion of new technology, including
microcomputers, into the public school system brings with it
questions concerning the appropriate use of this technology.
Since the acquisition of problem-solving skills, in
particular integrated process skills, is an important part
of science learning, many educators have spent time
considering how to best utilize new technology to help in
the teaching of these important skills. The purpose of this
study was to empirically explore some of the variables which
may influence educators in their practice of science
teaching using microcomputers. An emphasis in the design of
this experimental research was ecological validity; that is,
a naturalistic study characterized by conditions of the
treatments matching the standard classroom environment and

conditions.

80

81

A review of literature suggested a number of variables
which may play an important role in both organizing teaching
methods and delivery of instruction in science classrooms.
This study examined three factors as suggested in the
literature:

1) individual students working alone at a computer vs.
cooperative groups of two and cooperative groups
of four students working at a microcomputer,

2) the gender of the students,

3) the grouping of the students based on their ability
to solve problems using the integrated processes
skills prior to the beginning of the lessons given

to them during this research.

Procedure

Two hundred and forty—five seventh and eighth grade
student subjects were the focus of this study. They were
selected from twelve classrooms in three different school
districts. Selection of classrooms was based upon the
number of computers available and the teachers' willingness
to participate in this research.

Within the three schools, classrooms were randomly
assigned to treatment conditions, (i.e., individuals vs.
pairs vs. four member groups of students working at the
computer). A total of four classrooms across all schools
were assigned to each of the three treatment groups. Within

classroom and gender, students were randomly assigned to

82

groups based upon their ability (i.e., high, middle, low),
according to the grouping rules described in Chapter Three.
A pre-test consisting of two parts, (i.e., the TIPS and a
researcher developed test that more closely matched the
practice students received during the intervention), was
graded as a regular science unit, with each part receiving a
maximum of twenty-four points. After the approximately two
week intervention, students were given a post-test
consisting of a different form of the two part pretest.
Data was collected on the two response variables in each of
pre-test and post-test, plus the students gender, ability
level and treatment assignment.
Hypotheses.
A total of five hypotheses were tested for significance
with regard to each of the two response variables (i.e.,
part one and part two of the tests). The multivariate F
test statistic and a within-subjects factorial design was
used for these analyses. The results are summarized below.
Null Hypothesis 1) There is no significant difference in
the learning of science process skills between two-
member cooperative learning groups, four—member
cooperative learning groups and individuals who work
alone using microcomputers.
Null hypothesis 1 was not rejected for either response

variable (p = .877 for part one, and p = .771 for part two).

83

Null Hypothesis 2) There is no interaction between high,
medium, and low ability students and group size on
learning science process skills within a microcomputer
environment.

Null hypothesis 2 was not rejected for either response

variable (p = .829 for part one, and p = .934 for part two).

Null Hypothesis 31 There is no interaction between the
gender of the student and group size on learning
science process skills within a microcomputer
environment.

Null hypothesis 3 was not rejected for either response

variable (p = .349 for part one, and p = .864 for part two).

Null Hypothesis 4) There is no significant difference in
the learning of science process skills between males or
females.

Null hypothesis 4 was not rejected for either response

variable (p = .457 for part one, and p = .364 for part two).

Null Hypothesis 51 There is no significant difference in
the learning of science process skills by those
students showing low, middle, or high ability (i.e.,
based on pretest part one rankings).

Null hypothesis 5 gas rejected for both response variable (p

= .001 for part one, and p = .000 for part two).

Discussion.
As stated in the review of literature, the research

efforts which have focused on the effects of instructional

84

group size regarding learning in non-computer and computer
environments do not show consistent results. Some studies
conclude individuals are superior to small groups, other
studies conclude the opposite or show no significant
difference. The Cox and Berger (1985) study was the only
research linking group size and achievement within a
microcomputer environment identified in the literature
review. In a laboratory setting, they found seventh and
eighth grade students who worked in groups of two or three
members solved more problems correctly than individuals
working alone or in groups of five. In the current study,
group size (i.e., treatment) effects were insignificant.
Although this research is not consistent with the Cox and
Berger (1985) conclusion that teams of two to four members
would seem best suited to work together to solve problems,
the research conditions and activities between the two
studies appear significantly different.

Conversely, this research does not support the
hypothesis that individuals are superior to small groups
when working on this type of problem-solving within a
microcomputer environment. This may have implications for
educators/administrators wishing to design environments in
which students are involved in problem-solving. Also, it
may lend support to those educators who believe that, even
if a school does not have one computer for each student,
microcomputers can be valuable tools in teaching certain

problem-solving skills. If group learning is as effective

85

as individual learning, it may be valuable information in
policy decisions when considering cost efficiency. Further,
effective group learning may have implications concerning
how software and courseware designers develop programs which
include what is known about collaborative learning of
science process skills using microcomputers.

There is a growing body of research on ability by
treatment interaction. Differences in a student’s aptitude
may interact with an instructional approach to produce
differential achievement. The current study did not find
the curvilinear ATI that researchers such as Webb (1977) and
Peterson (1981) have found, (i.e., high ability and low
ability students benefit from small-group learning and
medium ability students do slightly better working alone).
However, differences between this study and others (e.g.,
setting; content; operational definitions of variables), may
have been significant, thus making direct comparisons
difficult in the area of ATI.

In Chapter Two, the focus of the review of gender-
related literature was that there exists a need to develop
and report treatments showing no sex-related differences in
hopes of identifying more gender-neutral activities in
science. It was anticipated that the current research would
prove to be such an activity. This study found no
significant differences between males and females in
learning science process skills in a microcomputer

environment. In addition, no interaction effects were found

86

on gender by treatment, or gender by ability, or gender by
ability by treatment. Therefore, the results of this study
indicate that the lessons and procedures implemented in the
manner described, generated gender-neutral activities in
science.

Concerning the significant results of ability in both
parts of the post-test, at least one study (Mayer, 1974) has
investigated the phenomena of different instructional
methods resulting in learning outcomes which differ. This
was indicated by a pattern of post-test performance in which
subjects in one instructional group excelled on one kind of
transfer post-test item and subjects in another group
excelled on another kind; producing what Mayer (1974) called
a Treatment X Post-test interaction (TPI).

The phenomena in the current research differs from that
in the Mayer (1974) study; instead of differential learning
of transfer items between instructional groups, the current
research is concerned with different transfer of learning
wiphin instructional method, but between students of
different ability levels - an Ability x Post-test
interaction (API). At least one of the characterizations
Mayer (1974) proposed to describe the difference in TPI may
be useful in describing API as well. Mayer proposed that
"different kinds of learning outcomes are due to acquisition
processes in which the same content material is encoded
within different assimilative sets by different subjects"

(p. 644). Viewed from this perspective, which appears to be

87

consistent with the distinction between prerequisite
knowledge and beneficial knowledge described in Chapter
Four, speculation can be made regarding the API found in
this study. It may be that high ability students, who
already are knowledgeable about the more general skills
tested in part one of the post-test, concentrate their
learning on the task specific skills tested in part two. On
the other hand, the lower ability students concentrate their
learning on the more general, essential knowledge tested in
part one of the pre- and post-tests, and do not have time
during the two week intervention to engage in the more task
specific skills tested in part two of the tests.
Conclusions.

The following conclusions are made concerning this research
study:

1) This study found that students working in a
microcomputer environment, in teams of two and four members
were as effective in learning integrated science process
skills as were students working alone.

2) No significant interaction between ability (i.e.,
high, medium, and low) and group size (i.e., individuals,
pairs, and quads), was found in this study of students
learning science process skills while working in a
microcomputer environment.

3) Results of this study indicate that the lessons and
procedures implemented in the manner described, generated

gender-neutral activities in science.

88

4) Results of this study indicate that in part one of
the post-test the trend was for low ability students to gain
the most and the high ability students to gain the least,
compared to the pre-test; while in part two that trend was
reverse. This might be described as an ability by post-test
interaction.

For subjects who do not have a well integrated set of
general experiences in the science process skills, testing
on tasks related to specific problems may not measure the
learning that may have occurred. Conversely, for students
who dp have a relatively high ability in science process
skills, testing in general abilities after they have
practiced solving specific problems might not detect the
skills they may have learned. In summary, the common
practice of evaluating all students in a classroom using a
post-test composed of only one type of items after the
students have practiced problem solving in science may not
detect the learning that has taken place by any given
individual, depending upon that individual’s prior skills
level.

5) The results from this study in science are
consistent with those of White (1985) in the area of social
studies, indicating microcomputers using a file management
program, along with structured activities, can be used as a

tool to promote student learning of process skills.

89

Limitations.

It was not possible to assign students randomly to
treatments. Although no significant differences between the
three treatments in either part of the post-test were found,
the use of intact classrooms provides no control for certain
effects (e.g., teacher, overall differences in mean IQ of
classroom members). Therefore, any differences in the
between subject factor, (i.e., treatments), which may have
been found in this study might have been from either the
differential effectiveness of group size, from differences
in the classrooms in general, or from both sources of
differences. However, in an attempt to reduce classroom
differences, four classrooms were assigned to each
treatment, and at least one classroom from each school was
assigned to each treatment. There was no significant
differences between the treatment means on the pre-pesp.
Given the threat to internal validity posed by the use of
intact classes in research (Borg and Gall, 1983), there was
no evidence that the results would have been different if
random assignment of individuals to treatment within
classrooms had been possible.

The lessons used to give students practice in science
process skills were developed for this research. No effort
was made to study the effect of using this courseware under
different conditions (e.g., time of intervention, amount of
teacher directed instruction), than those described for this

study. Therefore, care should be exercised not to assume

90

the conditions in which these materials were used are
optimal. Results of this study may not be applicable to
other lessons and/or conditions, even if the other
conditions and/or courseware are designed for the same
learning objectives.

The effectiveness of the cooperative groups was not
formally assessed. Therefore, the emphasis placed on
cooperation within assigned groups by the classroom teacher
while instructing and interacting with students may not
accurately reflect the level or type of cooperation designed
in other cooperative learning group research studies.

Treatments were limited to the problem-solving
objectives and conditions described in this study. It would
be inappropriate to generalize results to other problem-
solving situations. The acquisition of science process
skills using microcomputers and cooperative learning groups
might be more or less effective if other problem-solving
conditions are used.

The three schools used in this research, while from
three different school districts, did not represent a cross-
section of different size or socio-economic schools. For
instance, no urban school was included. Mainly, the three
schools were taken from cities within 15 miles of Lansing,
MI; however, all were outside of Lansing. Therefore, given
this and the method of selection of the schools used (i.e.,
based on the number of computers and teacher willingness to

participate in the study), generalizations of these results

91

can not be made to other schools in all types of school
districts.

While significant learning was detected on both parts
of the pre- and post-tests, problem solving with
microcomputer databases probably cannot be effectively
learned in a one-time activity or within one subject area.
Using only a one to two week time block, in one subject area
at one grade level, does not seem sufficient to teach
problem-solving as discussed in this study.

Recommendations for further research.

This research study has shown that microcomputers can
be used as a tool in promoting learning of science process
skills. During the research process, different aspects of
individuals and small groups of students were explored
interacting with microcomputers. This exploration and the
resulting analyses may have implications and suggest some
recommendations for future research and use.

1) The lessons prepared for this research were
successful in promoting student learning of science process
skills. However, other lessons and procedures for
implementing them may be more successful. Consideration
should be given to designing lessons that are more optimal.

2) In this research design, teacher interaction with
students was intentionally held to a low level to reduce the
teacher effects across different schools as much as
possible. Consideration should be given to designing

research which explores the mix of teacher-directed vs.

92

student-directed activities in this type of microcomputer
environment.

3) Given the ability by post-test interaction effect
described in this research, a high priority may be to design
research which specifically investigates this phenomena,
and/or developing test(s) to use for near transfer that are
more rigorously measured for reliability and validity.

4) This research explored only effectiveness as a
response variable. Informal questionnaires given to
students in this study resulted in some students liking some
components of the lessons and procedures used (i.e., their
group members, science work using microcomputers) while
other students disliked these same and/or other aspects of
the study. The cooperative learning literature suggests
responses other than effectiveness (e.g., affective
responses) as important outcomes. Consideration should be
given to designing research which explores outcomes in
addition to achievement.

5) Critical outcomes measured in this study, and as
far as is known in all cited studies, were by individuals
taking paper and pencil tests. Given this, and considering
recommendations 3 and 4 above, research should be considered
that measures dependent variables in groups, perhaps using
the microcomputer as a testing medium.

6) Teams of subjects were formed randomly, given the
rules described in Chapter Three, and friction among members

developed in some cases. The students whose scores were

93

dropped from this study solely because they were unable to
be grouped according to the rules showed significant overall
learning on both parts of the pre- post-test. What would be
the result if students where allowed to form their own
groups? Would they choose to work alone, or in varying size
groups? Would they form mixed-gender groups or same-gender
groups? Friendship groups may be superior to some other
grouping method because of the greater willingness of
friends to work together toward a common goal.

Consideration should be given to studying the differences
between groups formed by student preference and groups
formed in other ways.

7) The results reported here indicate that, overall,
the activities were gender-neutral. It is not enough to
ignore gender-related issues when developing lessons and
implementation procedures for science curriculum. Research
in the areas of science and microcomputer use should report
results of gender in an attempt to build knowledge in the
area of activities which are gender-neutral.

8) Cooperative group learning with microcomputers is a
more efficient use of a scacre resource, at least at this
time. Since this research found no strong advantage
favoring achievement by individual students working alone,
assignment of learners to small groups may be indicated for
this type learning activity. However, more research is

needed to replicate these findings, especially across

94

different school districts, before settling on this
strategy.

9) All subjects in this study were 7th and 8th
graders. Research should be considered which investigates
age differences in science process learning. It is
interesting to conjecture what might be the outcome of this
research if older (or younger) students, who have different
perspectives on mixed-gender cooperative group and a greater
or lesser overall science knowledge base, had been used for
this study on achievement. Also, what would be the results
if affective outcomes were formally measured?

10) This research used only a quasi-experimental
design. Consideration should be given to other
methodologies beyond those employed in this study (e.g.,
observation and analysis of students’ verbal accounts of
their own thought processes while working alone or in
groups).

11) The use of intact classrooms being assigned to
treatment completely confounded these two variables.
Research should be considered that uses random assignment of
students to treatments, if possible, to directly control for
differences among classrooms.

Other Considerations/Recommendations.

12) It became clear to the researcher during this
intervention, that use of microcomputers by classroom
teachers is not an automatic process. Incorporation of new

technologies requires teacher in-service and pre-service

95

training. Science teachers need to know when and how it is
appropriate to use technology, such as computerized
databases, to promote their instructional objectives.
Teacher training that is being referred to here is not the
usual "computer literacy" (e.g., how a computer works,
programming courses in BASIC). Instead, it is the training
in how and when to incorporate computers into existing or
developing curriculum - methods driven training.

13) Simply because microcomputers can be used as a
tool to promote certain kinds of problem-solving skills,
does not mean they will be used for those goals. It appears
that schools are investing considerable time and money to
develop programs to teach students how computers are put
together and how to program computers, with little attention
to how the computer can serve students as a tool.
Consideration should be given by teachers and school
administrators on how to promote computers as tools as well
as these other uses. Computers used as tools may be one way
to integrate new technology into various subject areas and
curriculum. Problem solving is an important part of many
school subjects, and all methods, including microcomputers,
should be exploited to promote learning of these skills.

14) It is clear that courseware designers should
explicitly consider how their programs and lesson plans can
be used in small group and whole group instruction, and not
simply assume one student to one computer as the way their

courseware will be implemented. As suggested by White

96

(1985), curriculum developers must use the best of what is
known about problem solving and information processing

skills when designing their materials. The advantages of
new technology may be lost if the problems we ask science

students to solve are trivial.

APPENDICES

APPENDIX A

APPENDIX A

TEACHER CHECKLIST FOR STUDENTS WORKING IN GROUPS

Tell the students that their grade will be based partly
upon their individual improvement between the pretest
and the post test. Also, the other part of their
individual grade will be based upon how much
improvement each member of their group does on the
pretest and post test. So, it will help each of them
to make sure all members of their group learns the
content and practices the lessons.

Tell the students, at least each day, that they should
work with each other within their group.

Tell the students, at least each day, that they should
ask each other for help. Let the students know it is
each of their jobs to make sure everyone in the group
understands the lessons.

Tell the students, whenever it is appropriate, that they
should consult with the instructor only if no one in
their group knows how to proceed.

Remind the students, whenever appropriate, to refer to
the AIMS if they have a question concerning the use of
the database.

Instruct the students to take turns at the keyboard.
Each student in each group should have approximately
equal time entering commands at the computer for the
group. It is up to them to see that everyone uses the
keyboard.

Given you are convinced every one in a group has no
reasonable suggestion on how to proceed while they are
working on a lesson, direct the students in that group
to the part of that lesson, or a previous lesson, that
will help them. Give the students encouragement and
every opportunity to work through the lessons without
direction from you. (Note: If the question involves
how to use the database, answer the question directly
or direct the group of students to the AIM that will
help them. How to use the database is pp; a
substantive part of this research.)

97

98

TEACHER CHECKLIST FOR STUDENTS WORKING INDIVIDUALLY

Tell the students that their grade will be based upon
the individual improvement between their pretest score
and their post test score.

Tell the students, whenever it is, that they should
consult with the instructor only if they can not
determine how to proceed on their own.

Remind a student, whenever appropriate, to refer to the
AIMs if he/she has a question concerning the use of the
database.

Given you are convinced a student has no reasonable
suggestion on how to proceed while he/she is working on
a lesson, direct that student to the part of that
lesson, or a previous lesson, that will help him/her.
Give all students encouragement and every opportunity
to work through the lessons without direction from you.
(Note: If the question involves how to use the
database, answer the question directly or direct the
individual student to the AIM that will help him/her.
How to use the database is pp; a substantive part of
this research.)

APPENDIX B

APPENDIX B

THE PROCESSES OF ANALYSIS
R. J McLeod

When scientists conduct an experiment, they perform
several processes. These processes may not be performed in
the order that we will describe them, but they are all part
of the general process of experimenting. You will use these
same processes as you do the experiments on Climate and
Weather. The processes that you will learn to use in this
lesson are:

Stating the purpose of an experiment

Stating a hypothesis

Determining the information needed

Controlling variables

.Arranging the information in a report so that the
hypothesis is easy to determine

* Analyzing the results

iii-I'l-

Each of these processes will be explained in this
lesson and you will be given opportunities to practice them.

Stating the Purpose: The purpose of an experiment is a
general statement of what you expect to find. Many times,
the purpose of an experiment is given to you.

For example, in Experiment One, the purpose is:

To determine if there is a relationship between temperature
and latitude.

The purpose is almost always a general statement of
something that we believe to be true. We may believe it to
be true because of some experience or because of other
scientific theory. In this case, the purpose seems
reasonable because of lots of experiences (in the northern
hemisphere, we go south for vacations in the winter, birds
fly south for the winter, the weather person on TV usually
points to warmer temperatures in the south than in the
north, etc). There is also scientific reasons to expect a
relationship between temperature and latitude.

What is meant by "a relationship between temperature
and latitude"?

99

(D

(W (\l n) _-_

i—de

H (D (I O
L-‘ ,A .140"

A A- iu..n 1"

In

100

In science, a relationship between two variables (for
example temperature and latitude) means that when one of the
variables changes, the other changes also. This is most
easily seen by arranging one variable (either from high to
low or from low to high) and observing how the other
changes. The variable that we choose to arrange is called
the independenp variablg, while the variable that we observe
as a result of arranging the independent variable is called

the dependent variable.

Example 1

* As one variable increases, the other also increases.

 

 

Variable A Variable B
20 500
30 600
40 700

* In this example, variable A is ordered from the smallest
value (20) to the largest (40). Notice that as variable A

gets larger, variable B also gets larger. Variable A is the
independent variable in this case because we chose to order
it and look for an effect on variable B. Variable B is the

dgpendenp variable.

Exam le 2

* As one variable decreases, the other also decreases.

 

 

Variable B Variable A
40 700
30 600
20 500

* In this example, variable B is ordered from the largest
value (40) to the smallest (20). Notice that as variable B
gets smaller, variable A also gets smaller. In this
example, Variable B is the independent variable.

1. Can you explain why Variable B is called the independent
variable?

 

2. What is Variable A called?

 

3. Why?

 

(Answers are on page 8 and 9 of this lesson)

101

When both variables change in the same direction
(increasing g decreasing) , there is a
relationship between the variables and it is
called a dirgzt relationship. Examples 1 and 2
are both direct relationships.

Example 3

* As one variable increases, the other decreases.

 

 

 

Variable A Variable B
20 700
30 600
40 500

* In this example, variable A is ordered from the smallest
value (20) to the largest (40). Notice that as variable A
gets larger, variable B gets smaller. They go in opposite
directions.

4. What is Variable A called?

 

5. Why?

 

6. What is Variable B called?

 

\)

Why?

 

(Answers are on page 8 and 9 of this lesson.)

Examplg 4

* As one variable decreases, the other increases.

 

 

Variable A Variable B
40 500
30 600
20 700

* In this example, variable A is ordered from the largest
value (40) to the smallest (20). Notice that as variable A
gets smaller, variable B gets larger.

8. What is Variable A called?

 

9 Why?

 

10. What is Variable B called?

 

11. Why?

 

(Answers are on page 8 and 9 of this lesson.)

102

When the variables change in the ppposipe
direction (one increases while the other
decreases), there is a relationship between the
variables and it is called an inverse
relationship. Examples 3 and 4 are both inverse
relationships.

Relationships do not need to be perfect. Look at the
following:

 

 

Variable A Variable B
50 600
40 500
30 700
20 200
10 100

12. What is Variable A called?

 

13. Why?

 

14. What is Variable B called?

 

15. Why?

 

16. What kind of relationship does this example show?

17. Why?

 

(Answers are on page 8 and 9 of this lesson.)

* In this example, variable A is ordered from the largest
value (50) to the smallest (10). In general, as Variable A
decreases, so does variable B. However, there is an
exception. Can you find it? In real life, there are almost
always exceptions. For example, adult males are usually
taller than adult females. If we were to arrange a table of
height vs. sex so that the height was arranged from shortest
to tallest, we would expect to find that most of the short
adults are females and most of the tall adults are males.
However, we all know short males and tall females. These
are the exceptions, but the relationship between height and
sex still exists. In this case, since sex is not numerical,
the relationship is neither direct nor inverse. It is just
a relationship.

103

If one variable is arranged so that it
increases (or decreases), and the other
variable seems to be quite random
(neither increases or decreases, even in
a general way) then there is said to be
no relationship between the two
variables.

All of this means that if we want to find out if
there is a relationship between two variables, we
should choose one of them to be the independent;
variable and arrange it so that it increases (or
decreases) and then look at the other (dependent;
variable) to see if it shows a direct or inverse
relationship, (or no relationship at all). A
computer can do this for us and do it very easily
and quickly.

Stating a hypothesis. In the northern hemisphere, as the
latitude decreases, (as we go South), the temperature will
increase. (Inverse relationship)

Notice that the hypothesis states the purpose in a way
that suggests what to look for. Other hypotheses that are
equally good may be:

1. As the latitude increases in the northern
hemisphere, (as we go North), the temperature will
decrease. (Inverse relationship)

We could also state the following hypotheses and
they would be OK for the purpose of our
experiment. However, our experiences suggest that
these hypotheses are pp; true for the northern
hemisphere, but might be true for the southern
hemisphere.

2. As the latitude decreases, (as we go South), the
temperature will decrease. (Direct relationship)

3. As the latitude increases, (as we go North), the
temperature will increase. (Direct relationship).

104

Determine the information needed: In order to test the
hypothesis, we need:

a) Latitude
b) July Temperature

c) Location’s Name (The hypothesis doesn’t require that
we have this information, but it is nice to have
additional information like this at times).

Controlling variable: Whenever an experiment is conducted
to determine the effect of one variable on another (in this
case, the effect of latitude on temperature), other
variables may also affect the results. In this experiment,
an obvious variable is the time of year. We know that the
temperature also gets warmer in the summer and colder in the
winter for most places in the United Sates. Therefore, if
we look at a July for one location and a January for
another, we may get very strange results. The answer is to
control all of the other variables that we can. In this
case, we will control month by selecting the same month for
all of our data. Let’s control on month by selecting July.

Arranging the Information: The way that the information is
arranged on paper is very important in helping us interpret
the data.

We will assume that the information will be arranged in
columns. This makes it easy to find relationships. The
hypothesis states that as the latitude decreases, the
temperature should increase. This tells us that one column
should be the latitude and it would be nice if the one next
to it is the temperature for that latitude.

It also tells us that it would be very helpful if the
data were arranged so that the latitude decreases (or
increases). We will then look at the temperature column to
see if it also increases or decreases. Since we are using
real data, we should expect to find some exceptions, even if
there is a relationship.

Finally, we might print the name of the location in the
third column. The report headings might look something like
the following:

Latitude Temperature Location

105

Important information about arrangement:
.Arrangement includes which data is placed in which
columns ppg ordering one of the columns (arranging
the data in the column from high to low or low to
high) to see how the data in the other column is
then arranged.

Analyzing the Results: Once you have stated the hypothesis,
determined the information needed, the organization of the
information, and printed this information out, you must then
analyze it. Many times, you can just look at the printout
and be able to see if a relationship exists. Other times,
it may be necessary to go back one or two steps (or
sometimes back to the beginning) and produce another
printout. For example, we suggested that you control on
month by selecting only the month of July. You may want to
try another printout by controlling on a different month.
You may decide that you need to arrange the information in a
different way to make your analysis easier. With the
computer, this is very quick and easy to do.

Your analysis should include a statement of:
1. Whether you found a relationship to exist.

2. If possible, whether the relationship seems to be a
direct relationship or an inverse one.

3. If the relationship is not perfect, note some of the
exceptions and, if you can, state some inferences for
them.

Practice Exercises

Now let’s see if you understand the processes of
experimenting. See if you can help conduct the following
experiment.

The students in a certain school believed that boys
were taller than girls. Their teacher asked them to take a
survey of their class to see if this was true. The purpose
of gathering the data was to see if there is a relationship
between sex and height. Their hypothesis was:

Hypothesis: If we compare the heights of boys to
girls, the heights of
boys .

 

106

The following data were obtained:

 

NAME AGE SEX HEIGHT (inches)
Alma 11 female 52
Andrea 13 female 59
Arthur 11 male 53
Bernard 11 male 57
Bill 13 male 57
Byron 13 male 62
Catherine 13 female 60
Cecil 13 male 57
Daisy 11 female 51
Debbie 13 female 59
Harold 11 male 52
Helga 11 female 56
Irma 13 female 56
Jackie 11 female 53
James 13 male 60
Jean 11 female 63
Jim 11 male 54
Joan 11 female 57
Luke 11 male 55
Nancy 13 female 58
Richard 13 male 59
Rick 11 male 51
Ronald 13 male 58
Sue 11 female 54
Trudy 11 female 56
Victor 13 male 56
Violet 13 female 61
Walter 13 male 61

Although all of the data are collected, it is very difficult
to test their hypothesis! Look at the data and answer the
following questions:

18. Which variables do you need to test the hypothesis that
boys are taller than girls?

 

 

19. Which variable should be controlled so that it does not
effect the results?

 

20. Which variable is really not needed for this
experiment?

 

21. How might you organize the data so that the answer is
easier to see?

 

22.

107

What arithmetic operations could be performed on the

data to help in the analysis?

 

10.

ll.

12.

l3.

14.

15.

16.

(Answers are on page 8 and 9 of this lesson).
Answers to Questions on The Processes of Analysis
Variable B is the independent variable because we chose
to order (arrange) this variable and observe changes in
another.

Variable A is called the dependent variable.

Because its values depend on the way we arrange the
values of variable B.

Variable A is the independent variable.

Because we chose to order (arrange) this variable and
observe changes in another.

Variable B is called the dependent variable.

Because its values depend on the way we arrange the
values of variable A.

Variable A is the independent variable.

Because we chose to order (arrange) this variable and
observe changes in another.

Variable B is called the dependent variable.

Because its values depend on the way we arrange the
values of variable A.

Variable A is the independent variable.

Because it is obvious that the data for this variable
are ordered from high to low. At first glance, you
might think that is true of Variable B also, but look
closely.

Variable B is the dependent variable.

Because its values depend on the arrangement of
Variable A. It is clear that the data are not ordered
on Variable B because of the value of 700. This is an
exception to the general decrease of Variable B. If
the data would have been ordered on Variable B, the
value of 700 would have been first.

This is a direct relationship.

17.

18.

19.

20.

21.

22.

108

Because both variables change in the same direction.
In this case, when Variable A was ordered so that it
decreased, Variable B also decreased.

You need both the sex and the height.

The age might affect the results. It is possible that
at certain ages, girls are taller than boys. At any
rate, since we are not sure, we should obtain data for
the same ages and, in this way, control for age.

The name of the student is of little interest for this
experiment.

It would be much easier to answer the question if all
of the boys were grouped together and then all of the
girls. In other words, we should order on sex. It
would also be helpful to have sex and height printed in
columns next to each other.

An average of all of the boys' heights and an average

of all of the girls’ heights would make comparison much
easier.

The following table is the same data except the age has

been controlled at 13 years, the names of students have been
left out, and the averages for the heights have been
computed. Can you use this table to answer the question,
"are boys taller than girls"?

AGE SEX HEIGHT (inches)
13 female 59
13 female 60
13 female 59
13 female 62
13 female 63
13 female 58
13 female 61
Average height of females 60.3
13 male 57
13 male 62
13 male 57
13 male 60
13 male 59
13 male 58
13 male 61

Average height of males 59.1

109

Which variable is the dependent variable and which is
the independent? Why? If you have trouble with this one,
think about which variable was chosen to be ordered.

Analysis might be even easier if the same data were
organized so that the height is first and it is ordered for
each sex like the following table. Does this make the
picture any clearer? Note that there are some males taller
than some females, but on the average, females in this class
are taller than the boys. Notice also how much easier it is
to identify those males that are taller than females because
the data are ordered on height.

HEIGHT SEX AGE
63 female 13
62 female 13
61 female 13
60 female 13
59 female 13
59 female 13
58 female 13
Average height of females 60.3
62 male 13
61 male 13
60 male 13
59 male 13
58 male 13
57 male 13
57 male 13

Average height of males 59.1

110

EXPERIMENT ONE
The Relationship between Temperature and Latitude
Note: All data for these experiments are for a location in
the northern hemisphere. In the northern hemisphere,
latitude increases as you move north.

Purpose: To determine if there is a relationship between
temperature and latitude.

Hypothesis: As the latitude decreases, (as we go South),
the temperature will increase.

Can you suggest other hypotheses that would
help us conduct this experiment?

Controlling variables: Which variables will you control?
Why?
Determine the Information Needed: List the information that

you will need in order to test the hypothesis:

a)

 

b)

 

c) Location’s Name (The hypothesis doesn’t
require that we have this information, but it
is nice to have additional information like
this at times)

 

Arrange the Information:

Which column will be
first?

 

Which column will be
second?

 

Which column will be
third?

 

Which column will you have the computer
order?

 

Which variable will be the independent
variable?
Why?

 

111

Which variable will be the dependent
variable?
Why?

 

Will you order from high to low or low to
high?
Why?

 

Analyzing the Results:
Your analysis should include a statement of:

1. Whether you found a relationship to
exist. If so, what makes you believe there
is a relationship? If there is no
relationship, what is your evidence?

2. If possible, whether the relationship
seems to be a direct relationship or an
inverse one.

3. If the relationship is not perfect, note
some of the exceptions and if you can, state
some inferences for them.

4. Will this same relationship (or lack of
it) be true for another month?

Materials Needed:
* AppleWorks program disk
* Climate data disk. Use Data Base File,
CLIMATETEMP. You may use Lab Report Format,

Lab 1 MonthlyA, Lab 1 MonthlyB, or you may
format your own.

112

EXPERIMENT TWO

Climate Experiment Two - The Relationship Between Latitude
and the Temperature Difference Between July and January

Purpose of the Experiment: In some locations, the
temperature is nearly the same all year, while in other
locations, there is a great difference between the summer
temperature and the winter temperature. In most places,
July is the hottest month and January the coldest.
Therefore, for this experiment, we will use the difference
between the July temperature and the January temperature as
one variable. This difference has already been computed and
is in the database.

The purpose of this experiment is to determine if this
difference is related to latitude. In other words, do you
expect the difference between January and July temperatures
to be greater in the North, the South, or is there no reason
to believe that there is a relationship?

Hypothesis: As the latitude decreases, (as we go South),
the temperature difference (between January and July) will

 

Can you suggest other hypotheses that would
help us conduct this experiment?

Controlling variables: Which variables will you control?
Why? In this experiment, the variables have already been
controlled by the selection of temperature difference
between July and January. If you had reason to believe that
East-West location was related to this temperature
difference, you might later control on latitude. How would
you do this?

Determine the Information Needed: List the information that
you will need in order to test the hypothesis:

a)

 

b)

 

c) Location’s Name (The hypothesis doesn’t require that
we have this information, but it is nice to have
additional information like this at times.)

.J ..

1 l'\- _ mini-pr i‘ a Fun..-

.gJ

113

Arrange the Information:

Which variable will be the independent
variable?

 

Which variable will be the dependent
variable?

 

Which column will be
first?

 

Which column will be
second?

 

Which column will be
third?

 

Which column will you have the computer
order?

 

Will you order from high to low or low to high?
Why?

 

Analyzing the Results:
Your analysis should include a statement of:

1. Whether you found a relationship to
exist. If so, what makes you believe there
is a relationship? If there is no
relationship, what is your evidence?

2. If possible, whether the relationship
seems to be a direct relationship or an
inverse one.

3. If the relationship is not perfect, note
some of the exceptions and, if you can, state
some inferences for them.
4. Will this same relationship (or lack of
it) be true for another month?

Materials Needed:
* AppleWorks program disk
* Climate data disk. Use Data Base File,

CLIMATETEMP. You may use Lab Report Format,
Lab 2, or you may format your own.

114

EXPERIMENT THREE
The Relationship Between Precipitation and Latitude
Purpose: To determine if there is a relationship between
precipitation (rain, snow, etc) and latitude.

Hypothesis: As the latitude , (as we go South),
the precipitation will .

 

 

Can you suggest other hypotheses that would
help us conduct this experiment?

Controlling variables: Which variables will you control?
Why?

Determine the Information Needed: List the information that
you will need in order to test the hypothesis:

a)

 

b)

 

c) Location's Name (The hypothesis doesn’t
require that we have this information, but it
is nice to have additional information like
this at

times)

 

Arrange the Information:

How will you arrange your
columns?

 

 

Which variable will be the independent
variable?

 

Which variable will be the dependent
variable?

 

Which column will you have the computer
order?

 

115

Will you order from high to low or low to
high?
Why?

 

Analyzing the Results:
Your analysis should include a statement of:

1. Whether you found a relationship to
exist. If so, what makes you believe there
is a relationship? If there is no
relationship, what is your evidence?

2. If possible, whether the relationship
seems to be a direct relationship or an
inverse one.

3. If the relationship is not perfect, note
some of the exceptions and, if you can, state
some inferences for them.

4. Will this same relationship (or lack of
it) be true for another month?

Materials Needed:
* AppleWorks program disk

* Climate data disk. Use Data Base File,
CLIMATEPRECIP. You may use Lab Report
Format, Lab 3 MonthlyA and Lab 3 MonthlyB, or
you may format your own.

116

EXPERIMENT FOUR

The Relationship Between Relative Humidity, Dry Bulb
Temperature, and Wet Bulb Temperature

Purpose: The purpose of this experiment is to determine the
relationship between relative humidity and wet and dry bulb
temperature.

Dry bulb temperature is found by the normal means of
taking the temperature reading from a thermometer that is
dry (and they usually are unless it's raining). Wet bulb
temperature is found by first wrapping some cotton around
the bulb of a thermometer, soaking it in water, and then
causing air to blow past it (for example using a fan to blow
on it). Just as you feel cooler when you are wet and a wind
is blowing, a wet bulb thermometer will usually have a lower
temperature than a dry bulb thermometer. Scientists use the
differencg between thg dry bulb and the wet bulp
temperatures to determine relative humidity. Relative
humidity, in turn, is an indication of how much water is in
the air. For example, when it is raining, the relative
humidity is 100%. In this experiment, we want to find out
what relative humidity is related to. That is, consider
relative humidity the dependent variable and determine what
other variable is related to it and whether the relationship
is direct or inverse.

Hypothesis #1: As the , the
relative humidity will also

 

 

Hypothesis #2:

 

 

Hypothesis #3:

 

 

Controlling variables: Which variables will you control?
Why? Note that the database you will be using for this
experiment is all for the same location including
temperature and relative humidity readings every three hours
for the months of July and January. Do you think that the
month may affect your results? How can you control for
this?

117

Determine the Information Needed: List the information that
you will need in order to test the hypothesis:

a)

 

b)

 

Arrange the Information:

For each hypothesis, you must produce a report that
will test the hypothesis (help you to answer it). As in the
other experiments, consider:

The arrangement of the columns

Which will be the independent and the dependent
variables.

The need to control variables.

How you will order the data.

Analyzing the Results:
Your analysis should include a statement of:

1. Whether you found a relationship to
exist. If so, what makes you believe there
is a relationship? If there is no
relationship, what is your evidence?

2. If possible, whether the relationship
seems to be a direct relationship or an
inverse one.

3. If the relationship is not perfect, note
some of the exceptions and, if you can, state
some inferences for them.

4. Will this same relationship (or lack of
it) be true for another month?

Materials Needed:
* AppleWorks program disk
* Weather data disk. Use Data Base File,

LOCAL. You may use Lab Report Format, Lab 4
Humid, or you may format your own.

118

EXPERIMENT FIVE

Are Winter Months Mere Cloudy than Summer Months?

Purpose: The purpose of this experiment is to determine if
there is a relationship between the amount of cloud cover
and the two seasons, winter and summer. That is, is winter
more cloudy than summer or is summer more cloudy than
winter, or is there no difference. Remember, the data you
are using is only for 1985 for the city of Grand Rapids,
Michigan. Without more data, you panno; make general
statements for all years in Grand Rapids, nor can you say
that these same conditions exist in other communities. The
investigation is intended to show you how this analysis is
done so that, if you want to, you could obtain data for your
community and for other years and do a similar report.

Hypothesis: The average cloud cover for all of the days of
July will be less than

 

Can you suggest other hypotheses that would help
us conduct this experiment?

Controlling variables: Which variables will you control?
Why? Remember that the database you will be using for this
experiment is all for the same location. However, you have
very detailed information for this location including
readings every three hours for the months of July and
January. Do you think that the time of day may affect your
results? How can you control for this? How will you
compare January cloud cover to July cloud cover?

Determine the Information Needed: List the information that
you will need in order to test the hypothesis:

a)

 

b)

 

c) In order to do this experiment, you will

need to get the total of the cloud cover for

the period you select. If you don’t remember
how, review the AIMS modules.

Arrange the Information:

What will you consider as you arrange the
information?

119

Analyzing the Results:
Your analysis should include a statement of:

1. Whether you found a relationship to
exist. If so, what makes you believe there
is a relationship? If there is no
relationship, what is your evidence?

2. If possible, whether the relationship
seems to be a direct relationship or an
inverse one.

3. If the relationship is not perfect, note
some of the exceptions and, if you can, state
some inferences for them.
4. Will this same relationship (or lack of
it) be true for another month?

Materials Needed:
* AppleWorks program disk
* Weather data disk. Use Data Base File,

SKYTEMP. You may use Lab Report Format, Lab
5 Cloud, or you may format your own.

APPENDIX C

APPENDIX C

Dear Parent or Guardian,

I am preparing to study the effects of the number of
students in a learning group on the achievement of certain
science process skills as part of the requirements for my
degree at Michigan State University. This study involves
student lessons that we have prepared, which will be used
during the regular school day. It also involves the
administration to each student of a 36 item test before and
after these lessons which will permit measurement of the
effectiveness of the lessons. Your child's teacher and
school principal have approved this project, and now we are
requesting you permit the teacher to release your child’s
grade to us for research purposes.

Would you help us in learning more about science
instruction by giving us permission to include your child in
this study? If so, please sign the reverse side of this
sheet. Your child’s identity will not be revealed in any of
our written reports. If you are interested, I would be
happy to make the overall results of this research available
to you if you contact me with such a request.

Thank you for helping me with my research, and to
improve our knowledge of science instruction for the
schools.

Sincerely,

Zane L. Berge
Graduate Student

120

121

Michigan State University
Department of Educational Systems Development

As the legal parent/guardian of ,
I give my consent to one of the following options:

 

I give my permission for the above named student to

participate in the study as has been described.
I do not give my permission for the above student’s
scores to be released for research purposed in the
study as has been described.

I am indicating that the research project being conducted by
Michigan State University, has been explained to me and that
I have been informed about my child’s involvement in this
project. I recognize that I have the right to withdraw my
permission for my child’s participation at any time prior to
the study without penalty.

Signed

 

(parent/guardian)

Date

 

I am indicating my willingness to participate as
indicated above.

Signed

 

(student)

PLEASE RETURN TO YOUR TEACHER!

APPENDIX D

APPENDIX D

What variables Effect the Strength of a Magnetic Field?

The following describes a question that lends itself to
research, and asks you to answer some questions about how
you might plan for an experiment to help answer that
question. Write your answers directly on these pages in the
spaces provided.

A classroom discussion on magnetism lead to the
production of a simple electromagnet. The class used
insulated bell wire connected to a dry cell battery(s), and
coiled the wire around a nail, (used as the electromagnet’s
core in this case), to make an electromagnet. They found
this magnet could be used to attract tacks. The idea was
presented that the strength of the magnetic field, measured
by how many tacks could be attracted, can be controlled by
how the electromagnet is constructed.

For instance, one class member suggested that, (given
more than one battery is used in the construction of the
magnet), whether the batteries are connected in parallel or
series effects the strength of the resulting magnet. Other
suggestions about what affected the strength of the magnet
involved the number of turns of the coil (wire) around the
core, the number of batteries, and the kinds of materials
making up the core of the electromagnet.

Purpose: To determine what variables effect the strength of
a magnetic field when constructing a simple electromagnet.

Hypothesis: The greater the number of turns of the coil,
the more tacks the electromagnet will pick up in a string.

1. What type relationship is hypothesized (direct,
indirect, no relationship)?

 

Controlling variables: Given the hypothesis above, name
two variables you would control so they do not effect the
results when testing that hypothesis?

2a.

 

b.

 

122

a".

123

Determine the Information Needed: List the information (two
variables) that is absolutely needed in order to test the
hypothesis that the number of turns of a coil increase the
strength of the magnet:

3a.

 

b.

 

Arrange the Information:

4a. Which variable will be the independent variable?

 

b. Which variable will be the dependent variable?

 

c. Which variable should be ordered (either from highest
value to lowest or visa versa), to test the hypothesis
given above?

 

Can you suggest two other hypotheses that may help the
students answer the question of which variables effect the
strength of a magnetic field?

 

 

5a. As the is increased,
the will increase.
b.

 

 

Analyzing the Results: The analysis of this data should
include a statement of whether a relationship exists between
the independent and dependent variables, and what evidence
there is for such an inference. Can you name at least two
other statements the analysis should include?

6a.

 

 

 

 

124

Effect of Sunlight on Heating various Earth Surfaces

The following describes a question that lends itself to
research, and asks you to answer some questions about how
you might plan for an experiment to help answer that
question. Write your answers directly on these pages in the
spaces provided.

During a classroom discussion about wind, a number of
issues were noted by students in a 9th grade class. It was
pointed out that as cold air moves in to replace warm air
that is rising, winds are created. The discussion turned to
how might the heating of the earth’s surface cause
convection currents (wind). The experience of some students
caused them to guess that sunlight causes materials such as
white sand to become hotter than materials such as water.
They believed the different temperatures of the earth’s
surfaces caused convection currents. Other class members
were not so sure these suspicions were justified.

To start with, a couple class members recalled that
since the earth is tilted, different parts of the earth
receive more or less direct rays of the sun. Further, they
suspected that direct rays of the sun produce a greater
heating effect than slanted rays. However, further
discussion lead to ideas that heating of the various
surfaces on earth may be affected by such things as: the
total amount of materials available to absorb the sun’s
radiant heat, the area of the surface exposed to the sun,
the length of time of exposure of the surface to the sun,
and the color of the surface. The students wanted to
scientifically test some of the ideas they suspected to be
true.

To test one of the notions, the students took two equal
size bread pans filled with an equal amount of soil. They
stirred the soil thoroughly to ensure equal starting
temperatures, and placed one container flat on a window sill
in the sunlight. This pan received slanted sunlight. They
propped up one end of the second container so that it
received the sunlight at right angles (that is, direct
rays). They used a thermometer to record the temperature of
each soil sample at fifteen-minute intervals for one hour.

Purpose: To determine what variables effect the sun’s
heating of various surfaces.

Hypothesis: As the angle of the earth’s surface and the
rays of the sun decreases from 90 degrees (that is, becomes
less direct), the rate at which the surface will absorb heat
energy will decrease.

7. What type relationship is hypothesized (direct,
indirect, no relationship)?

 

125

Controlling variables: Given the hypothesis above, name
two variables you would control so they do not effect the
results when testing that hypothesis.

8a.

 

b.

 

Determine the Information Needed: List the information (two
variables) that is absolutely needed in order to test the
hypothesis that the angle the sun’s rays strike a material
has an effect on the amount of heat absorbed:

9a.

 

b.

 

Arrange the Information:

10a. Which variable will be the independent variable?

 

b. Which variable will be the dependent variable?

 

c. Which variable should be ordered (either from highest

value to lowest or visa versa), to test the hypothesis
given above?

 

Can you suggest two other hypotheses that may help the
students answer the question of the effect sunlight has on
heating various surfaces on earth?

11a. AS the

 

increased, the will increase.

 

b.

 

 

Analyzing the Results: The analysis of this data should
include a statement of whether a relationship exists between
the independent and dependent variables, and what evidence
there is for such an inference. Can you name at least two
other statements the analysis should include?

12a.

 

 

res
YOL
que
spa

126

What variables Effect the Period of a Pendulum?

The following describes a question that lends itself to
research, and asks you to answer some questions about how
you might plan for an experiment to help answer that
question. Write your answers directly on these pages in the
spaces provided.

Students in a 9th grade classroom have noticed that not
all pendulums swing at the same speed. For example,
pendulums of different clocks swing at different speeds.

The students want to find out what causes this phenomena.
Using a thumb tack at the edge of a desk, some lengths of
string for the pendulum arm, and different size iron rings
for bobs, they fashion simple pendulums and decide to
conduct an experiment to discover what effects the number of
swings (per unit of time) of a pendulum. These students
gathered data on the mass of the bob (they use either 10
grams, 25 grams, or 50 grams weights), the actual time for
different numbers of swings of the pendulum, the number of
swings (they timed either 50 swings, 100 swings, or 200
swings), and the length (they used either 15 centimeters, 30
centimeters, or 45 centimeters), of the string to the center
of the bob from the thumb tack.

The students worked in pairs and gathered dozens of
records, each record contained the data described above.
Following are examples (not the actual data) of the
information in three records (each record is the four pieces
of information on one line about one experimental trial):

 

 

TIME MASS NUMBER OF LENGTH OF THE
Alseconds) (grams) SWINGS STRING (cm)
145 10 50 15
192 25 200 30
45 50 50 30

The students gathered the information and now wish to
plan and organize the data in a way that will let them
discover what variables effect the period of a pendulum.

Purpose: To determine what variables effect the period of a
pendulum.

Hypothesis: As the mass of the bob is increased, the period
of the pendulum will increase.

1. What type relationship is hypothesized (direct,
indirect, no relationship)?

 

)0. «\\ '1‘ -l 8 PC

127

Controlling Variables: Given the hypothesis above and the
four pieces of information collected for each trial, name
two variables you would control so they do not effect the
results, when testing this hypothesis?

2a.

 

b.

 

Determine the Information Needed: List the information (two
variables) that is absolutely needed in order to test the
hypothesis that the mass of the bob is related to the period
of a pendulum:

3a.

 

b.

 

Arrange the Information:

4a. Which variable will be the independent variable?

 

b. Which variable will be the dependent variable?

 

c. Which variable should be ordered (either from highest
value to lowest or visa versa), to test the hypothesis given
above?

 

Can you suggest two other hypotheses that may help the
students answer the question of which variables effect the
period of a pendulum?

5a. As the is increased, the

 

will increase.

 

 

 

Analyzing the Results: The analysis should include a
statement of whether a relationship exists between the
independent and dependent variables, and what evidence there
is for such an inference. Can you name at least two other
statements the analysis should include?

6a.

 

b.

 

13(1Q

1 I I 1 .

(Will J‘I

128

Effect of Sunlight on Heating various Kinds of Earth
Surfaces

The following describes a question that lends itself to
research, and asks you to answer some questions about how
you might plan for an experiment to help answer that
question. Write your answers directly on these pages in the
spaces provided.

During a classroom discussion about weather and the
heating effect of sunlight on the world’s oceans and land
masses, a number of issues were noted. A couple class
members pointed out that surfaces such as roadways and
beaches become very warm in sunlight, whereas water does not
seem to become so warm. The experiences of some students
caused them to guess that sunlight causes white sand to
become hotter than water; and that sunlight causes topsoil
to become hotter than water. Other students were not so
sure about these suspicions.

Further discussion lead to ideas that heating of the
various surfaces on earth may be affected by such things as:
the total amount of materials available to absorb the sun’s
heat, the area of the surface exposed to the sun, the length
of time of exposure of the surface to the sun, whether or
not the surface is exposed to direct rays of the sun or
slanted rays, and the color of the surface. The students
wanted to scientifically test some of the ideas they suspect
to be true.

Purpose: To determine what variables effect the sun’s
heating of various surfaces.

Hypothesis: As the area of the surface exposed to the
sunlight is decreased, the temperature of the surface will
increase.

7. What type relationship is hypothesized (direct,
indirect, no relationship)?

 

Controlling variables: Given the hypothesis above, name
two variables you would control so they would not effect the
results when testing that hypothesis.

8a.

 

b.

 

Determine the Information Needed: List the information (two
variables) that is absolutely needed in order to test the

129

hypothesis that the size of the area of the surface exposed
to sunlight is related to the temperature of that surface:

9a.

 

b.

 

Arrange the Information:

10a. Which variable will be the independent variable?

 

b. Which variable will be the dependent variable?

 

c. Which variable should be ordered (either from highest

value to lowest or visa versa), to test the hypothesis
given above?

 

Can you suggest two other hypotheses that may help the

students answer the question of which variables effect the
sun’s heating of various surfaces?

11a. As the is

 

increased, the will increase.

 

b.

 

 

Analyzing the Results: The analysis should include a
statement of whether a relationship exists between the
independent and dependent variables, and what evidence there

is for such an inference. Can you name at least two other
statements the analysis should include?

12a.
b.

 

 

LIST OF REFERENCES

AAAS. (1967). Science - A Process Approach. Washington, D.C.
American Association for the Advancement of Science.

AAAS. (1976). Curriculum Catalog. (AE9505-10-85). Lexington,
MA: Ginn and Company (Xerox Corp.).

Allen, V., and Feldman, R.S. (1976). Research on children
tutoring children: A critical review. Review of
educational research. A§(3): pp. 355-385.

 

Amaria, R.P., Biran, L.A. & Leith, G.O. (1969). Individual
vs. cooperative learning: Influence of intelligence and
sex. Educational Research (British). 11, 95-103.

Anderson, Ronald E.; And Others. (1983). Computer
inequities in opportunities for computer literacy.
Minnesota University, Minneapolis. Minnesota Research
and Evaluation Center. (ED249 293).

Anderson, R.E., Klassen, D.., Krohn, K.R., Smith-Cunnien, P.
(1982). Assessing Computer Literacy. (Publication
#503). St. Paul, MN Minnesota Educational Computer
Consortium.

Aronson, E., Bridgeman, D., Y Geffner, R. (1978).
Interdependent interactions and prosocial behavior.
Journal of Research and Development in Education. 1;:
16-27.

Ausubel, D.P. (1963). The psychology of meaningful verbal
learning. NY: Grune & Stratton.

Baird, William Erwin. (1985). Changes in preservice
elementary teachers' hypothesizing skills following
qroup cooperative or individual study with computer-
presented text or computer simulations. unpublished
doctoral dissertation. The University of Texas at
Austin.

Bar-Tal, Daniel and Saxe, Leonard, (Eds.). (1978). Social
psychology of education: Theorv and research. New York:
John Wiley and Sons.

Bar-Tal, Daniel, and Geser, Devorah. (1980). Observing
cooperation in the classroom group. In Cooperation in
education, edited by Sholomo Sharan, Pau; Hare, Clark
D. Webb, and Rachel Hertz-Lazarowitz. Provo, Utah:
Brigham Young University Press.

130

Bc

131

Barker, R., & Gaump, P. (1964). Big school, small school:
High school size and student behavior. Stanford, CA:
Stanford University Press.

Barker, R. & Wright, H. (1955). Midwest and its children.
New York: Harper and Row.

Becker, Betsy Jan. and Chang, Lin. (1986). Measurement of
science achievement and its role in gender differences.
Paper presented at the Annual Meeting of the American
Educational Research Association. San Francisco, CA.
April 16-20.

Becker, H.J. (1982). Microcomputers in the classroom -
Dreams and realities. (Report No. 319). Baltimore, MD:
Center for Social Organization of Schools, The Johns
Hopkins University.

Berger, Carl F. (1982). Attainment of skill in using science
processes. I. instrumentation, methodology and
analysis. Journal of Research in Science Teaching.
19(3): 249-260.

Bitter, Gary G., and Gore, Kay. (1986). Curricular
implications of a "computer for every student".
Computers in the Schools. 13(2). Summer. The Haworth
Press.

Bloom, Benjamin. (1981). All our children learning: A primer
for parents. teachers and other educators. New York:
McGraw Hill.

Bloom, B.S., Englehart, M.D., Furst, E.J., Walker, W.H., &
Kathwohl, D.R. (1956). Taxonomv of education
objectives. The classification of educational goals.
Handbook I: Cognitive domain. New York: Longmans Green.

Bloom, B.S., Englehart, M.D., Furst, E.J., Walker, W.H., &
Kathwohl, D.R. (1972). Taxonomv of behavioral
objectives: the classification of educational goals.
New York: David McKay.

Bommarito, Frank. (1986). Central TIME - Applying effective
social studies software. Micro Lines. Kalamazoo, MI.
Jan/Feb.

Borg, W.R. (1964). An evaluation of abilitv grouping.
Cooperative Research Project, No. 577. Salt Lake City:
Utah State University.

Boyer, Ernest L. (1983). High School. New York: Harper &
Row, Publishers.

C1

132

Bracey, Gerald W. (1984). Research column in Phi Delta
Kappan. 65(9):645.

 

 

Brown, John Seely. (1985a). Idea amplifiers - new kinds of
electronic learning environments. Educational Horizons.
Spring.

Brown, John Seely. (1985b). Process versus product: A
perspective on tools for communal and informal
electronic learning. Journal of educational computing
research. 1(2). pp. 179-201.

Bruner, Jerome S. (1961). The1process of education.
Cambridge, MA: Harvard University Press.

Bruner, Jerome S. (1961). The act of discovery. Harvard
Educational Review. 31:pp. 21-32.

Buckholdt, D., & Wodarski, J. (1978). Effects of different
reinforcement systems on cooperative behaviors
exhibited by children in classroom contexts. Journal of
Research and Development in Education. 12:50-68.

 

Burlin, F. (1976). Sex-role stereotyping: Occupational
aspirations of female high school students. School
Counselor. 24: pp. 102-108.

Burns, Joseph C., Okey, James R., and Wise, Kevin C. (1985).
Development of an integrated process skill test: TIPS
II. Journal of research in science teaching. 22(2): pp
169-177.

Burton, G. (1979). Regardless of sex. Mathematics Teacher.
lgzpp. 251-270.

Clark, R. (1983). Reconsidering research on learning from
media. Review of educational research. 53, pp. 445-
459.

Cox, Dorothy A.H. (1980). Early adolescent use of selected
problem-solving skills using microcomputers. Ph.D.
Dissertation. University of Michigan.

Cox, Dorothy and Berger, Carl F. (1981). Microcomputers are
motivating. Science and Children. September: pp. 28-29.

Cox, Dorothy A., and Berger, Carl F. (1985). The importance
of group size in the use of problem-solving skills on a
microcomputer. Journal of educational computing
research. 1(4). pp. 459-68.

Cronbach, L.J., & Snow, R.E. (1977). Aptitudes and
instructional methods. New York: Irvington Publishers.

Dev

08‘»

DEV

131'

n}

n}

133

DeVries, D., & Slavin, R. Teams-games-tournaments (TGT):
Review of ten classroom experiments. Journal of
Research and Development in Education. 12:28-38.

Dewey, John. (1910). How we think. Boston: Heath.

Dewey, John. (1964). School conditions and the training of
thought. In Reginald D. Archambault (Ed.) John Dewey
on Education. New York: The Modern Library.

Dillashaw, Gerald F. and Okey, James R. (1980). Test of the
integrated science process skills for secondary science
students. Science Education 63(5): pp 601-608.

Doyle, W. (1983). Academic work. Review of educational
research. 33(2), pp. 159-199.

Deutsch, M. (1962). Cooperation and trust: Some theoretical
notes. In M.R. Jones (Ed.), Nebraska symposium on
motivation (pp. 275-319. Lincoln: University of
Nebraska Press.

Deutsch, M. (1969). Productive and Destructive. Journal of
Social Issues. 33 pp. 7-43.

Doerr, Christine. (1979). Microcomputers and the 3 R’s: A
guide for teachers. New Jersey: Hayden Book Company.

Douglas, J.W.B. (1964). The home and the school. London:
MacGibbon & Kee.

Eash, M. J. (1961). What have we learned? Educational
Leadership. 13, 429-434.

Fennema, E. (1980). Teachers and sex bias in mathematics.
Mathematics Teacher. 13: pp. 169-173.

Freeman, C., Hawkins, J. and Char, C. (1984). Information
management tools for classrooms: Exploring database
management systems. Technical Report No 28. Center for
Children and Technology. New York: Bank Street College
of Education.

Finley, Fred N. (1983). Science processes. Journal of
research in science teaching. 33(1), pp. 47-54.

Fisher, Glenn. (1983). Where CAI is effective: A summary of
the research. Electronic Learning. 3(3): pp. 82-84.

Forman, E., & Kraker, M. (1985). The cognitive benefits of
collaborative problem solving: Piagetian and vaotskian
perspectives. Paper presented at the American
Educational Research Association, Chicago.

Ge

134

Gagne, R.M. (1974). Essentials of learning for instruction.
Hinsdale, IL: Dryden.

Gallini, Joan K. (1986). Cognitive outcomes in a logo
environment. Paper presented at the American
Educational Research Association Conference, San
Francisco.

Gibb, J. (1951) The effects of group size and of threat
reduction upon certainty in a problem-solving
situation. American Psychology. 6, pp: 324.

Goldberg, M.L., Passow, A.H., & Justman, J. (1966). The
effects of ability grouping. New York: Teachers College
Press.

Graybeal, S. S., and Stodolsky, 5.3. (1985). Peer work
groups in elementary schools. American Journal of
Education. May.

Greenfield, Patricia M. (1985). Multimedia education: why
print isn’t always best. American Educator. Fall.

Greeno, J.G. (1980). Trends in the theory of knowledge for
problem solving. In D.T. Tuma & F. Reif (Eds.),
Problem solving and education: Issues in teaching and
researgh. Hillsdale, NJ: Erlbaum.

Harlen, W., Black, P., Johnson, 3., & Palacio, D. (1983).
Science in Schools. Age 11: Report No. 2. London,
England: DES.

Hawkins, Jan. (1985). Computers and girls: Rethinking the
issues. Sex Roles. 13(3/4): pp. 165-180.

Herron, J.D., Luce, T.G., and Neie, V.E. (1976). The proper
experimental unit: Comparative analyses of empirical
data. Journal of Research in Science Teaching. 13(1):
pp. 19-27. .

Hoffman, L. Richard, and Maier, Norman R.F. (1961). Quality
and acceptance of problem solutions by members of
homogeneous and heterogeneous groups. Journal of
Abnormal and Social Psychology. 63(2):40l—407.

Hofstein, Avi, and Lunetta, Vincent N. (1982). The role of
the laboratory in science teaching; Neglected aspects
of research. Review of educational research. 33(2), pp.
201-217.

Hunter, B. & Furlong, M. (1985a). Scholastic PFS: Curriculum
data bases for U.S. history. New York: Scholastic, Inc.

:4
‘09“

135

Hunter, B. & Furlong, M. (1985b). Scholastic PFS: Curriculum
data bases for U.S. government. New York: Scholastic,
Inc.

Husen, T., & Svensson, N. (1960). Pedagogic milieu and
development of intellectual skills. Science Review. 33,
36-51.

Jernstedt, G. Christian. (1983). Computer enhanced
collaborative learning: A new technology for education.
T.H.E11Journal. 10(7): 96-101.

Johnson, D.W. (1970). The social psychology of education.
New York: Holt, Rinehart & Winston.

Johnson, D.W. (1979). Educational psychology. Englewood
Cliffs, NJ: Prentice-Hall.

Johnson, D., and Johnson, R. (1974). Instructional goal
structure: Cooperative, competitive or individualistic?
Review of Educational Research. 33(2): 213-240.

Johnson, D.W., & Johnson, R.T. (1975). Learning together and
alone: Cooperation. competitionL and individualization.
Englewood Cliffs, NJ; Prentice-Hall.

Johnson, D., and Johnson, R. (1978). Cooperative,
competitive, and individualistic learning. Journal of
Research and Development in Education. 13:pp. 3-15.

Johnson, D.W., & Johnson, R. (Eds.) (1978). Social
interdependence within instruction. Journal of Research
and Development in Education. 13(1).

Johnson, D.W., & Johnson, R. (1983). The socialization and
achievement crisis: Are cooperative learning
experiences the solution? In L. Bickman (Ed.), Applied
social psychology annual 4. Beverly Hills, CA: Sage.

Johnson, R.T., Johnson, D.W. and Stanne, Mary Beth. (1986).
Comparison of computer-assisted cooperative,
competitive, and individualistic learning. American
educational research iournal. 33(3), pp. 382-392.

Johnson, D.W., Maruyama, G., Johnson, R., Nelson, C. & Skon,
L. (1981). The effects of cooperative, competitive, and
individualistic goal structures on achievement: A meta-
analysis. Psychological Bulletin. 33, 47-62.

Johnson, Roger T., Ryan Frank L., And Schroeder, Helen.
(1974). Inquiry and the development of positive
attitudes. Science education. 33(1): 51-56.

136

Johnson, D.W., Skon, L., & Johnson, R. (1980). Effects of
cooperative, competitive, and individualistic
conditions on students’ problem-solving performance.
American Educational Research Journal. 11, 83-94.

Joyce, Bruce. (1985). Models for teaching thinking.
Educational leadership. May.

Kahl, S. R., Malone, M.R., Fleming, M.L. (1982). Sex-related
differences in pre-college science: Findings of the
science meta-analysis project. Paper presented at the
Annual Meeting of the American Educational Research
Association, New York. (Eric Document Reproduction
Service #ED216 909).

Kantowski, Mary Grace. (1983). The microcomputer and problem
solving. Arithmetic teacher. February.

Kelly, E.L. (1967). Transfer of training: An analytic
study. In B.P. Komisar and C. J. Macmillan (Eds.),
Psychological concepts in education. Chicago: Rand
McNally.

Keppel, Geoffrey. (1982). Design and analysis: A
Researcher’s Handbook (2nd). Englewood Cliff, NJ:
Prentice-Hall, Inc.

 

 

 

Kirk, Roger E. (1982). Experimental design: Procedures for
the behavioral sciences (2nd). Brooks/Cole Publishing
Company.

Klausmeier, H., Wiersma, W. and Harris, C. (1963).
Efficiency of initial learning and transfer by
individuals, pairs, and quads. Journal of Educational
Psychology. 33(3), 160-164.

Leentz, Linda. (1986). Variety - The key to success. Micro
Lines. 1(2). Michigan Special Discretionary
Microcomputer Literacy Grants. Kalamazoo, MI.

Lipkin, John P. (1983). Equity in computer education.
Educational Leadership. 31(1): p. 26.

Linn, Marcia C. (1985). Gender equity in computer learning
environments. Computers and the Social Sciences. 1:
pp. 19-27.

Linn, Marcia C. (1986). Establishing a research base for
science education: Challengengtrends, and
recommendations. Lawrence Hall of science and Graduate
School of Education, University of California,
Berkeley.

137

Lockheed, Marlaine E. (1985). Women, girls, and computers:
A first look at the evidence. Sex Roles. 13 (3/4) pp.
115-121.

Lockheed, M.E., Thorpe, M., Brooks-Gunn, J., Casserly, P.,
and McAloon, A. (1985). Sex 3 ethnic differences in
middle school mathematicsL science and computer
science: What do we know? A report submitted to The
Ford Foundation.

Lochhead, J. & Clement, J. (Eds). (1979). Cognitive process
instruction: Research on teaching thinking skills.
Philadelphia: Franklin Institute Press.

Lockheed, Marlaine E., Gulovsen, Joan P., and Morrison,
Donald. (1985). Student use of applications software.
Technical report. Educational Technology Center;
Harvard Graduate School of Education, Cambridge, MA.

Lockheed, M.E., and Harris, A.M. (1982). Classrooms
interaction and opportunities for cross-sex peer
learning in science. Journal of Early Adolescence.
3(2): pp. 135-143.

Lockheed, Marlaine E., and Mandinach, Ellen B. (1986).
Trends in educational computing: Decreasing interest
and the changing focus of instruction. Educational
researcher. 13(5), pp. 21-26.

Lockheed, M., Nielsen, A., Stone, M. (1983). Sex
differences in microcomputer literacy. In Proceedings
of the National Educational Computer Conference.
baltimore, MN.

Lorge, Irving and Solomon, Herbert. (1959). Individual
performance and group performance in problem solving
related to group size and previous exposure to the
problem. Journal of Psychology. 33: 107-114.

Ludeman, R.R. (1975). Development of the science processes
test (TSPT). Dissertation Abstracts International. 33.
203-A (Doctoral dissertation, Michigan State
university, 1974). (University Microfilms No. 75-14,
783).

MacGregor, S. Kim. (1985). Research issues in computer-
assisted learning environments. Paper presented at the
Annual Meeting of the American Educational Research
Association Chicago, IL. March 31-April 4.

Mathews, Walter, M. and Winkle, Linda Wyrick. (1982).
Computer equity for young women in rural schools.
Research in Rural Education. 1(1): pp. 37-41.

138

Mayer, R.E. (1974). Acquisition processes and resilience
under varying testing conditions for structurally
different problem-solving procedures. Journal of
Education Psychology, 33(5), pp. 644-656.

McGuire, Christine. (1973). Simulation technique in the
teaching and testing of problem-solving skills.
Occasional paper series -- Science Paper 8. Presented
at the 46th Annual Meeting of the National Association
for Research in Science Teaching, March.

McLeod, Richard. (1985). Using data bases to teach science
processes - even if you only have one microcomputer in
the classroom. NSTA.

McLeod, R.J., Berkheimer, G.D., Fyffe, D.W., & Robinson,
R.W. (1975). The development of criterion-validated
test items for four integrated science processes.
Journal of Research in Science Teaching. 13, 415-421.

McLeod, Richard. (1987). Lesson plans for weather and
climate using Appleworks database management software.
Unpublished lesson plans. East Lansing, MI.

McLeod, R. and Hunter, B. (1987). The data base in the
laboratory. Science and Children. 33(4). pp. 28-30,
155.

McLeod, R., Hunter, 8., and Finkel, L. (Computer software).
(1987). Weather and Climate Databases. New York:
Scholastic, Inc.

McMillan, James H. (Ed.) (1980). The social psychology of
school learning. New York: Academic Press.

McPhail, I.P. (1985). Computer inequities in school uses of
microcomputers: Policy implications. Journal of Negro
Education. 33(1):pp. 3-13.

Meehan, A.M. (1984). A meta-analysis of sex differences in
formal operational thought. Child Development, 33(3),
1110-1124.

Mitman, A.L., Mergendoller, J.R., Packer, M.J., and
Marchman, V.A. (1984). Scientific literacy in seventh-
grade life science: A study of instructional processl
task completion. student perceptions. and learning
outcomes. San Francisco, CA: Far West Laboratory for
Educational Research and Development.

Molitor, L.L., & George, K.D. (1976). Development of a test
of science process skills. Journal of Research in
Science Teaching. 13, 405-412.

139

National Assessment of Educational Progress. (1978). Science
achievement in the schoo1s. (Science Report No. 08-S-
01). Denver, CO: Educational Commission of the States.

National Commission on Excellence in Education. (1983). A
nation at risk: The imperative for reform. Washington
D.C.: U.S. Government Printing Office.

National Science Board Commission on Precollege Education in
Mathematics, Science and Technology. (1983). Educating
americans for the let century. September.

 

National Science Teachers’ Association Position Statement on
School Science Education for the 19705. (1975).
Readings in Science Education for the Elementary
School. Edited by Victor and Learner. New York:
Macmillian Publishing Company.

Newell, Allen. (1980). One final word. In D.T. Tuma and F.
Reif (Eds.). Problem solving and education: Issues in
teaching and research. Hillsdale, NJ: Lawrence Erlbaum
Associates, Publishers.

Norton, Pricsilla. Problem-solving activities in a computer
environment: A different angle of vision. Educational
technology. October.

Okebukola, Peter A. and Ogunniyi, Meshach B. (1984).
Cooperative, competitive, and individualistic science
laboratory interaction patterns - effects on students’
achievement and acquisition of practical skills.
Journal of research in science teaching. 31(9): 875-
884.

Okey, James R., and Majer, K. (1975). Individual and small
group learning with computer assisted instruction.
Paper presented at the Annual Meeting of the American
Educational Research Association, Washington, D.C.

Olds, Jr. Henry F., and Dickenson, Anne. (1985). Move over,
word processors - here come the databases. Classroom
computer learning. October.

Page, E.B. (1965). Recapturing the richness within the
classroom. Paper presented at the American Educational
Research Association Annual Meeting. February, 1965.

Perkins, D.N. (1985). The fingertip effect: How
information-processing technology shapes thinking.
Educational Researcher. 13(7): 11-17.

if)?

a. .-

140

Peterson, Penelope L. (1981). Ability X treatment
interaction effects in studies of small-group learning.
University of Wisconsin-Madison. Paper presented at the
annual meeting of the American Educational Research
Association, Los Angeles, CA.

Peterson, Penelope L. and Janicki, Terence C. (1979).
Individual characteristics and children’s learning in
large-group and small-group approaches. Journal of
educational psychology. 11(5), pp. 677-687.

Peterson, P.L., Janicki, T.C., & Swing, S.R. (1981). Ability
X treatment interaction effects on children’s learning
in large-group and small-group approaches. American
Educational Research Journal. 13: 453-473.

Peterson, P.L., Wilkinson, L.C., and Hallnan, M. (1984). The
social context of instruction: Group organization and
group processes. New York: Academic Press, Inc.

"U
'71
(1)

File. (Computer program). (1984). Mountain View, CA:
Software Publishing Company.

PFS: Repgrt. (Computer program). (1984). Mountain View, CA:
Software Publishing Company.

Ploeger, Floyd D. (1984). Instructional microcomputing: A
survey of research. Capitol-izing on computers in
education. Proceedings of the Association for
educational data systems annual convention, Computer
Science Press, 267-273.

Pogrow, Stanley. (1983). Education in the computer age:
Issues ofspolicy, practice. and reform. London: Sage
Publications Ltd.

Polya, G. (1957). How to Solve It. New York: Doubleday
Anchor.

Pon, Kathy. (1984). Databasing in the elementary (and
secondary) classroom. The computing teacher. November.

Riley, J.W. (1972). The development and use of a group
process test for selected processes of the Science
Curriculum Improvement Study. Dissertation Abstracts
International. 33, 6200A-6201-A (Doctoral dissertation,
Michigan State University, 1972). (University
Microfilms No. 73-12).

Resnik, Hank. (1984). From social studies to social science.
Learning. October.

Rubinstein, Moshe F. (1975). Patterns of problem solving.
Englewood Cliffs, NJ: Prentice-Hall Inc.

11.413
1 I.

141

Saunders, J. (1979). What are the real problems involved in
getting computers into the high school? Mathematics
Teacher. May. pp. 443-447.

Schofield, 8., Murphy, P., Johnson, 3., & Black, P. (1982).
Science in schools. Age 13: Report No. 1. London,
England: DES.

Sharan, S., Ackerman, z. and Hertz-Lazarowitz, R. (1980).
Academic achievement of elementary school children in
small-group versus whole-class instruction. Journal of
Experimental Education. 33(2): 125-129.

Sharan, S. (1980). Cooperative learning in small groups:
Recent methods and effects on achievement, attitudes,
and ethnic relations. Review of Educational Research.
33 (2): 241-271.

Shaw, Terry J. (1983). The effect of a process - oriented
science curriculum upon problem-solving ability.
Science education. 31(5): pp. 615-623.

Sheingold, K., Hawkins, J., Kurland, D.M. (1983). Software
for the information age. Unpublished manuscript.

Shymansky, James A., Kyle, William G., and Alport, Jennifer
M. (1983). The effects of new science curricula on
student performance. Journal of research in science
teaching. 33(5): 387-404.

Simon, Herbert A. and Newell, Allen. (1971). Human problem
solving: The state of the theory in 1970. American
Psychologist. 33: 145-159.

Skon, L., Johnson, D.W., and Johnson, R.T. (1981).
Cooperative peer interaction versus individual
competition and individualistic efforts: Effects on the
acquisition of cognitive reasoning strategies. Journal
of Educational Psychology. 13(1): pp. 83-92.

Slavin, R. (1978). Student teams and achievement divisions.
Journal of Research and Development in Education.
13:39-49.

Slavin, R. (1980). Cooperative learning. Review of
Educational Research. 33(2): 315-342.

Slavin, Robert E. (1983). Cooperative learning. New York:
Longman.

 

SPSSx (1986). SPSSK User’s Guide, (2nd Edition).
Chicago, IL.

Ci 41‘
‘(“J'

142

Stake, R.E. & Easley, J.A. (1978). Case studies in science
education (Vols 1 3 2). Urbana: Center for
Instructional Research and Curriculum Evaluation and
Committee on Culture and Cognition, University of
Illinois at Urbana-Champagne.

Stasz, Cathleen, Shavelson, Richard J., and Stasz, Clarice.
(1985). Teachers as role models: Are there gender
differences in microcomputer-based mathematics and
science instruction? Sex Roles. 13(3/4): pp. 149-164.

Stein, N., & Yussen, S.R. (1985). Review of Wertsch’s
analysis. In S.R. Yussen (Ed.), The growth of
reflection in children. (pp. 99-101). New York:
Academic Press.

Steinkamp, M.W., and Maehr, M. (1983). Affect, ability, and
science achievement: A quantitative synthesis of
correlational research. Review of Educational Research.
33(3): pp. 369-396.

Sternberg, Robert J. (1985). Teaching critical thinking,
part 1: Are we making critical mistakes? Phi Delta
Kappan. November.

Stapp, William B, and Cox, Dorothy, A. (Eds.). (1979).
Environmental Education Activities Manual. Farmington
Hills, MI.

Streibel, Michael J., and Garhart, Casey. (1985). Beyond
computer literacy. T.H.E. Journal. June.

Taba, Hilda. (1962). Curriculum Development. New York:
Harcourt, Brace, and World, Inc.

Tamir, Pinchas. (1985). Content analysis focusing on
inquiry. Journal of curriculum studies. 11(1), pp. 87—
94.

Tannenbaum, R.S. (1968). The development of the Test of
Science Processes. Dissertation Abstracts
International, 33, 2159-A (Doctoral dissertation,
Columbia University, 1968). (University Microfilms No.
69-677).

Taylor, Robert. (1980). The computer in the school: Tutor;
tool. tutee. New York: Teachers College Press.

Thomas, E.J., & Fink, C. (1963). Effects of group size.
Psychological Bulletin. 33. pp. 371-384.

Thelen, H.A. (1972). Education and the human question.
Chicago, IL: University of Chicago Press.

143

Thompson, Charles L. (1986). It’s no tool if it makes work
harder. Electronic Learning. September.

Tobin, Kenneth. (1986). Exemplary practice in science
classroom. Paper presented as part of a symposium
entitled "Exemplary Practice in Science Classrooms" at
the Annual Meeting of the National Association for
Research in Science Teaching. San Francisco, March.

Tobin, Kenneth G., and Capie, William. (1980). Teaching
process skills in the middle school. School Science and
Mathematics. 33: pp. 590-600.

Tobin, Kenneth G., and Capie, William. (1982). Development
and validation of a group test of integrated science
processes. Journal of research in science teaching.
13(2): pp. 133-141.

Tobin, Kenneth, and Fraser, Barry J. (1986). Investigations
of exemplary practice in silence and mathematics.
Paper presented at the annual meeting of the American
Educational Research Association, San Francisco, April.

Trowbridge, David. (1982). Groups and computer based
learning materials. Proceedings of National Educational
Computing Conference, pp. 11-14.

Trowbridge, David, and Durnin, Robin. (1984). Results from
an investigation of groups working at the computer.
ERIC Document ED 238 724.

Trowbridge, David and Durnin, Robin. (1984). Individual vs.
group usage of computer based learning materials.
Proceedings of the National Educational Computing
Conference. pp. 168-173.

Vargas, Julie S. (1986). Instructional design flaws in
computer based learning materials. Phi Delta Kappan.
June.

Voss, J.F. (1978). Cognition and instruction: Toward a
cognitive theory of learning. In A.M. Lesgold, J.W.
Pellegrine, S.D. Fokkema and R.Glaser (Eds.), Cognitive
psychology and instruction. NY: Plenum Press.

Walbesser, H.H. (1965). An evaluation model and its
application. Washington, D.C.: The American Association
for the Advancement of Science, AAAS Miscellaneous
Publication No. 65-9.

Walker, Decker F. (1983). Reflections on the educational
potential and limitations of microcomputers. Phi Delta
Kappan. October.

e113! «I...» v

144

Wallas, G. (1926). The art of thought. New York: Harcourt
and Brace.

 

Walsh, Debbie. (1985). Socrates in the classroom. American
Educator. 3(2): pp. 20-25.

Webb, Noreen M. (1977). Learning in individual and small
group settings. Technical Report No. 7. Office of
Naval Research, Arlington, VA. Personnel and Training
REsearch Progress Office.

Webb, Noreen M. (1980). A process-outcome analysis of
learning in group and individual settings. Educational
Psychologist. 13(2), pp. 69-83.

Webb, Noreen M. (1984). Sex differences in interaction and
achievement in cooperative small groups. Journal of
Educational Psychology. 13: 33-44.

Westbury, Ian. (1973). Conventional classrooms, "open’
classrooms and the technology of teaching. Journal of
curriculum studies. 3(2): pp 99-121.

White, Charles S. (1985). The impact of structured
activities with a computer-based file-management
program on selected information-processing skills.
unpublished doctoral dissertation. Indian University.

Whitehead, Alfred North. (1929). The rhythmic claims of
freedom and discipline. In The Aims of Education and
Other Essays. New York: The Macmillan Company.

Wodarski, J.S., Hamblin, R.L., Buckholdt, D.R., & Ferritor,
D.E. (1973). Individual consequences versus different
shared consequences contingent on the performance of
low-achieving group members. Journal of Applied Social
Psychology. 3: 276-290.

Yates, A. (Ed.). (1966). Grouping in education. New York:
Wiley.

Yinger, Joanne, and Eckland, Ruth. (1975). Problem solving
with children. San Francisco, CA: Far West Laboratory
for Educational Research and Development.

Zinn, Karl L. (1979). Computers in science teaching: Today
and tomorrow. In What research says to the science
teacher, Vol. 2. Edited by Mary Budd Rowe. NSTA.

9.!

11311 1' 41

    

RIES

"111111111111

1