. MI...
$.rfid.“ gang

3:: . .. $.23.
2)- Ji, .
..... .. fmfflw
f: .
- .
t;
q _

3%.... 3.33m
: 3w. 3......
.. 53...

 

. . . . . . . _ 33mm:
. , . . J.
1.1. u... . . . . . . La
- A . . . ‘
. . 34. 3 3 . . 3 r
.3: 3 . . .

.3351...” . , . . . .

can an? . . . . . . .
.133;

3.6.9..

.3 z».

 

 

1!

DH. 5 ; .

 

 

 

m.

(If.

 

«vi.- 1551». .2 6. 5. .1.
3.33.33.33.33“. . 333...»:

5.3:;
1.0 3 u

‘1 {355.14

2,
,, LIBRARY
1/6279 Michigan State
University

This is to certify that the
dissertation entitled

ACADEMIC SUCCESS IN COLLEGE GENERAL
CHEMISTRY

presented by
STEVEN P. POULlOS

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Higher, Adult, and Lifelong
Educafion

g%%fizwi//fllyg X4}

’Majof Professor’s Signature

¥/‘7—-”/; — a; ,5

 

 

 

 

 

 

 

 

 

 

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/ClRC/Date0ue.indd-p.1

 

ACADEMIC SUCCESS IN COLLEGE GENERAL CHEMISTRY
By

Steven P. Poulios

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Higher and Lifelong Education

2005

ACADEMIC SUCCESS IN COLLEGE GENERAL CHEMISTRY
By

Steven P. Poulios

ABSTRACT

Research on the link between student learning and success in higher education is
extensive and includes many perspectives, most of them highly focused and pertinent to
the researcher’s field. However, few, if any, studies have been performed from a more
‘holistic’ or well-rounded approach, that is, incorporating both cognitive and non-
cognitive factors in examining academic success.

Six factors leading up to and continuing through general chemistry courses in
college are likely to affect academic success. These factors include a variety of cognitive
and non-cognitive variables (Astin, 1968, Pascarella & Terenzini, 1991, Tinto, 1986,
Weidman, 1989): pre-college characteristics and experiences, college experiences in the
classroom, college experiences out of the classroom, instructional resources, student-
focused variables, and instructor-focused variables. These factors, ordered sequentially
from high school through college, form the components in the development of a student-
learning model (Pascarella & Terenzini, 1991).

Data from 2,247 students enrolled in general chemistry at Michigan State
University during the 2002-2003 academic year were analyzed using standard multiple
regression analysis. The model significantly predicted that in-class experiences and pre-
college factors explained a large portion of shared variability with the final grade in the

course (R2 = .398).

ACKNOWLEDGEMENTS

I would like to thank my committee members Roger Baldwin, Chuck Ostrom,
Kris Renn and especially the chair, Jim Fairweather for reading and providing feedback
on revision after revision. I would also like to thank the Department of Chemistry at
Michigan State University, Wendy Whitford and Paul Hunter for facilitating the data
collection for this study, and Kristin Anderson for reading many drafts. A special thank
you goes to Joe Brocato, Patty Kenney, Lucy Mailette, Michael Marshall and Eileen
Wilson, who traveled this road before me and were constant reminders that it can be
done. I would also like to thank Sarra Baraily for editing.

Last, but not least, my parents, sister, and brother-in-law, for their unwavering
support during the entire process. Like the other trials in my life you went though this

one with me. I could have never completed it without you.

iii

TABLE OF CONTENTS

LIST OF TABLES .....................................................................................................
LIST OF FIGURES ...................................................................................................

CHAPTER 1

INTRODUCTION .....................................................................................................
The Importance of Chemistry ...........................................................................
Demographics of University General Chemistry Courses ...............................
Statement of the Problem .................................................................................
Research Questions ..........................................................................................

CHAPTER 2
LITERATURE REVIEW
Introduction ......................................................................................................
College Impact Studies ....................................................................................
College Impact Models ....................................................................................
Research on Academic Success in Chemistry and Other Physical Sciences
Active Learning ........................................................................................
Collaborative Learning ...........................................................................
Chapter Summary ............................................................................................

CHAPTER 3
METHOD
Introduction ......................................................................................................
Conceptual Framework ....................................................................................
Components of the College Impact Model ......................................................
Factor Analysis ........................................................................................
Block One: Pre-College Characteristics and Experiences .....................
Block Two: In-Class Experiences ...........................................................
Block Three: Out-of-Class Experiences ..................................................
Block Four: Instructional Resources ......................................................
Black Five: Student Focused Variables ..................................................
Block Six: Instructor-Focused Variables ................................................
Research Questions ..........................................................................................
Study Participants ............................................................................................
Student Sample .........................................................................................
Data Collection ................................................................................................
Pilot of Survey Instrument .......................................................................
Data Collection Procedures ....................................................................
Data Analysis ...........................................................................................
Analysis Used for Research Questions ............................................................

vi

vii

Ofl-B—‘H

1

13
14
16
24
30
31
32

36
36
4o
41
44
46
49
51
52
54
56
57
58
59
61
63
66
67

CHAPTER 4

RESULTS
Introduction ...................................................................................................... 70
Analysis of Collected Data .............................................................................. 71
Research Question I ................................................................................ 71
Research Question 2 ............................................................................... 75
Research Question 3 ............................................................................... 83
Research Question 4 ............................................................................... 89
Summary .......................................................................................................... 90
CHAPTER 5
DISCUSSION AND CONCLUSION ........................................................................ 94
Conclusion 1 ..................................................................................................... 97
Conclusion 2 .................................................................................................... 99
In-Class Experience Block and Variables ............................................... 101
Pre-College Characteristics and Experiences Block and Variables ....... 101
Student-Focused Block and Variables .................................................... 104
Instructor-Focused Block and Variables ................................................. 108
Instructional Resources Block and Variables .......................................... 108
Out-of-C lass Experiences Block and Variables ....................................... 109
Suggestions for Model Revision ....................................................................... 111
Further Research .............................................................................................. 113
Scope and Limitations of this Research ............................................................ 115
APPENDICES .......................................................................................................... 117
Appendix A: MSU General Chemistry Survey Data ....................................... 118
BIBLIOGRAPHY ..................................................................................................... 128

LIST OF TABLES

Table 1: Spring, 2003 — Fall, 2003. Enrollment in CEM 141 and CEM 142 .......... 57
Table 2. Ratio of Male to Female Students enrolled in general chemistry

at Michigan State University .................................................................................... 76
Table 3. Ratio of Male to Female Students Attending College ................................ 76

vi

LIST OF FIGURES

Figure 1. Conceptual Framework of the College Impact Model .............................. 39

vii

CHAPTER 1

INTRODUCTION

The Importance of Chemistry

Chemistry is considered “the central science” (Brown, LeMay & Bursten, 2003,
p, 3). It is the universal language used to describe the understanding of the structure of
matter that is essential to studying biology, agriculture, engineering, geology and other
sciences. The importance of knowing chemistry has led to enormous advances in health
care, conservation of natural resources, energy, increased food production, and many
other environmental concerns.

An excellent example of this is the awarding of the 2005 Nobel Prize in
Chemistry to three scientists (one of whom, Robert Grubbs, was a student at Michigan
State University) for utilizing metathesis in a process for simplifying the synthesis of

complex molecules used to make new compounds for industrial and medical use.

This year's Nobel Prize Laureates in chemistry have made metathesis into
one of organic chemistry's most important reactions. Fantastic
opportunities have been created for producing many new molecules -
pharmaceuticals, for example. Imagination will soon be the only limit to
what molecules can be built!

In 1971 Yves Chauvin was able to explain in detail how metatheses
reactions function and what types of metal compound act as catalysts in
the reactions. Now the "recipe" was known. The next step was, if possible,
to develop the actual catalysts.

Richard Schrock was the first to produce an efficient metal-compound
catalyst for metathesis. This was in 1990. Two years later Robert Grubbs
developed an even better catalyst, stable in air that has found many
applications.

Metathesis is used daily in the chemical industry, mainly in the
development of pharmaceuticals and of advanced plastic materials.
Thanks to the Laureates' contributions, synthesis methods have been
developed that are: more efficient (fewer reaction steps, fewer resources
required, less wastage); simpler to use (stable in air, at normal
temperatures and pressures); and environmentally friendlier (non-injurious
solvents, less hazardous waste products).

This represents a great step forward for "green chemistry," reducing
potentially hazardous waste through smarter production. Metathesis is an
example of how important basic science has been applied for the benefit of
man, society and the environment. (Royal Swedish Academy of Sciences,
October 5, 2005)

Over the years, the study of chemistry has developed from a training ground for
scientists to a complex application for a much more diverse group of students and citizens
in a highly technical economy. Knowledge of chemistry has become essential for a
technically literate society.

Many of the greatest scientific discoveries in the twentieth century resulted from
the work of chemists and chemical principles. The discovery of the double helix
structure of DNA by Watson and Crick in 1953 would not have occurred without the
understanding of the relationship of the oxygen, sugar and phosphate molecules that
make up deoxyribonucleic acid.

The knowledge of chemical principles, in conjunction with biology, was also
instrumental in the discovery of the anti-cancer drug, Cisplatin. In the early 19605,
Barnett Rosenberg, a physicist at Michigan State University, was investigating the
influence of electrical fields on bacterial cell division. Platinum electrodes were used to
introduce electrical current into a medium of E. coli slurry. During the experiment, the
cell division in the E. coli bacteria was arrested to the point of stopping altogether. It was

determined that the platinum leaching from the electrodes stopped the cell division in the

bacteria. The researchers concluded that if platinum compounds could stop the cell
division in E. coli, then perhaps it would have the same influence on cancer cells. The
development of Cisplatin and other metallic chemotherapy drugs has saved the lives of
many cancer victims (Kotz &Treichel, 1996).

The study of general chemistry assists in making essential life decisions. Almost
every day there is a feature on the television news or in the newspaper regarding health
issues. For example, one feature story may report that drinking a moderate amount of
alcohol (especially red wine) is beneficial for cardiac health while another feature will
report on the dangers of alcoholism. Alcoholic beverages have been around for centuries.
Distilling alcohol is one of the oldest chemical procedures used by humans. Learning the
properties of alcohols is an integral part of general chemistry.

Another example is the warning label on common household bleach and ammonia
products instructing users not to mix the two products in the washing machine or cleaning
bucket. The source of the noxious gas produced from the acid-base reaction and concept
of chemical equilibrium becomes obvious after the first course in general chemistry.

Another important aspect of the study of general chemistry impacts public policy
decisions made by politicians and the general citizenry. For example, holes in the Ozone
layer are believed to increase the risk of skin cancer. Automobile exhaust and smoke
from coal fired electric generating facilities are purported to aggravate the problem.
There are agreements and ongoing discussions from the leaders of the largest
industrialized countries to limit these pollutants. In the case of developing countries,
however, there is a different case scenario: They are just beginning to enjoy the benefits

of reliable electric power for manufacturing, internal combustion engines for farming,

and transportation and refrigeration to keep food safe to eat. Knowledge of the
interaction of chemicals and the atmosphere is essential to the development of rational
and equitable policy decisions on these issues.

The study of chemistry is also important in disciplines other than the physical
sciences. Recently, student interest in the study of forensic science has increased. This
may be partially due to television shows such as “C.S.I. Miami” and “C.S.I. Crime Scene
Investigation,” which have popularized the role of the forensic scientist in crime scene
investigation. The life of a forensic scientist looks exciting, but television shows do not
fully explain that crime scene investigation is really about collecting samples and
performing quantitative analysis using procedures and principals learned in general

chemistry.

Demographics of University General Chemistry Courses

Students entering four-year colleges and universities are a diverse group in terms
of race, ethnicity, maturity, intelligence, socioeconomic background, and academic
preparation for study in higher education. However, what they all have in common is
that they have met the minimum requirements for admission to college, including
required coursework, grade point average (GPA), and standardized test scores. These
entrance requirements are believed to indicate ability to succeed in college.

While in college, many students choose, or are required to enroll, in a chemistry
course. Typically, general chemistry is the first chemistry course taken by college
students, as chemistry courses are vertically integrated, moving from basic chemistry to

more advanced levels. As in mathematics, students must have an adequate foundation in

the principles of chemistry to be successful in upper-level chemistry courses. For
students in the natural sciences and other majors, general chemistry is a prerequisite for
organic chemistry, analytical/physical chemistry, and advanced inorganic chemistry. It is
defined as the study of matter and its changes and provides grounding in basic chemical
principles and problem solving (Brady & Holum, 1996).

General chemistry is also an important service course, which means that is it
offered to students who are not majors in the department; it is recommended or required
by a student’s major; it is an introductory course to the discipline of chemistry; or it is a
single course without pre-requisites. Therefore, general chemistry includes students with
varied backgrounds, interests and abilities. At Michigan State University, all students in
the colleges of Natural Science, Engineering, Agriculture, and Natural Resources must
enroll in at least one semester of general chemistry. Successful performance in general
chemistry can also serve as an equalizer for students who enter higher education less
prepared than their peers. This is especially true for students from economically
disadvantaged backgrounds. If these students do well in general chemistry they then have
access to stimulating and financially rewarding careers in the sciences, medicine and
engineering.

Most students enrolled in college general chemistry have successfully completed
high school chemistry. However, despite earning good grades in high school and scoring
well on their college entrance examinations, many students do not do well academically
in college general chemistry. An example of this tendency can be found in the
distribution of final grades in Chemistry 141 for fall semesters 1996-2001 at Michigan

State University: approximately 20% of the students in the first semester of general

chemistry earned less than a 2.0 (or “C”) for their final grade in the course (Hunter,
2003)

Introductory chemistry courses often have the reputation of being difficult to
master. For the science major, general chemistry is often perceived as an essential first
step in the chemistry curriculum. For the non-science major, it is a course to be endured
in order to progress in their chosen major. The concepts and principles learned in general
chemistry, however, are an essential part of future coursework in all of the physical
sciences (Hunter, 2003).

General chemistry also serves as a “gatekeeper” course for many chemistry
majors (Ditzer and Ricci, 1995), which means that they need to take it as a pre-requisite
to upper-level chemistry courses.

Pre-professional students are especially concerned because their entry into
scientific or health-related careers might be impeded by failure to succeed academically
in college general chemistry. Students who apply to medical school must have completed
general chemistry, organic chemistry, and physical chemistry.

In many colleges and universities such as Michigan State University, the
Department of Forensic Science is housed within of the College of Social Science, not in
the College of Natural Science. Students in the College of Social Science, interested in
earning a Master’s degree in Forensic Science, must have a strong background in
chemistry and must complete coursework in general chemistry, organic chemistry, and
physical chemistry. These courses are essential for the understanding of DNA mapping

and quantitative analysis of crime scene evidence.

Whether or not students are science majors or choose other areas of study, they
should have the opportunity to learn about scientific inquiry and develop the critical
thinking skills taught in general chemistry. The scientific method is the basis for inquiry
in general chemistry and other sciences. It allows students to construct a reasonable,
accurate, and non-arbitrary meaning of the world around them.

In all the natural sciences and social sciences, researchers follow some variation
of the scientific method. They observe and describe phenomena, formulate a hypothesis
to explain the phenomena, collect data, and perform some type of analysis to accept or
reject certain hypotheses. If the hypotheses are accepted after repeated tests, a theory

may be proposed.

Statement of the Problem

Colleges and universities usually work hard to recruit and retain students,
especially women and minorities. Many students choose the sciences and engineering
because they can be academically engaging and financially rewarding. However, the
transfer rate from sciences to other majors is very high. Nationally, approximately 35%
of all students majoring in math, science, or engineering change majors or drop out
between their first and second year of college. For minority students the rate is 50 -
65%.

Many of the students who do remain in the sciences express their dissatisfaction
with their experience (Gainen, 1995). One of the reasons for this is that many of the
students enrolled in college general chemistry usually did well in high school chemistry

classes, which led to high expectations for success in college. Many had positive

experiences in high school chemistry after an enthusiastic teacher ignited their interests in
the sciences. Yet, interestingly enough, many of these students perform poorly in college
general chemistry courses and are disturbed by their performance (Deal 1984).

The majority of students enrolled in general chemistry are required to take the
course for another major (Oliver-Hoyo & Allen, 2005). Of the students that major in
chemistry, few go on to become chemists. The BS. chemistry program is intended to
prepare students to go on to graduate school in chemistry. The B.A. chemistry majors
typically apply to a professional program, such as medical school, dental school, or
forensic science program. Still, others go on to receive a teaching certificate for teaching
high school chemistry.

Nearly all of the movement out of the field of chemistry along with the
mathematics, sciences, and engineering fields usually occurs after the first-year of study.
This is primarily due to the gatekeeper effect of these courses (Gainen, 1995). General
chemistry is a difficult course. In some colleges and universities competition can be
fierce due to large numbers of pre-med students and course grade distributions skewed to
the lower end of the grading scale. The curriculum is fast-paced and requires aptitude in
mathematics and logical reasoning. Students must weigh “the profit-to-grief ratio”
(Gainen, 1995, p.12) to determine the future payoff in jobs and material benefits.
(Gainen’s term refers to students who weigh the difficulty in studying the sciences
against the eventual payoff in monetary and career rewards). Many students, especially
women and minorities, have switched to other majors, after determining that the
instructional culture of chemistry (or the other sciences) due to its hostile, alienating, and

impersonal nature, has dissuaded them from remaining the field.

This phenomenon has been evident in past MSU summer academic orientations
where more students declared a major in the College of Natural Science, for example,
than in the College of Communication Arts. The academic advisors for natural science,
especially the pre-professional advisors, were very busy working with first-year students,
while the Department of Communications Arts and Sciences did not usually have an
advisor in attendance. After two years, however, the reverse became true and the
students who failed in the sciences either dropped out of college or migrated to other
major areas of study.

Higher education is a huge public and political investment. In times of shrinking
funding and public support, it is politically and fiscally irresponsible to squander
resources on recruiting students only to have them leave the sciences or do so poorly that
they drop out completely. Moreover, the potential loss of scientists in the United States’
educational system, because of students’ failure in general chemistry is an important
academic and national security issue. Colleges and Universities in the United States may
be “the envy of the planet” (Axtell, 2003, p.1) but they are not keeping up with the rest of
the world in terms of emphasizing and endorsing the sciences. Moreover, higher
education, as a whole, is not keeping up with advances in technology.

Most institutions of higher education in the United States offer a variety of
courses that satisfy the requirement for general chemistry. The required courses usually
depend on the student’s major area of study. In general, there are courses for chemistry
majors and courses for non-chemistry majors. Non-chemistry majors enrolling in general
chemistry typically include students majoring in engineering, agriculture, human ecology,

and pre-professional studies (pre-medica], pre-dental, nursing, teacher education, etc.).

In contrast, chemistry majors, chemical engineering majors, and biochemistry majors are
usually enrolled in a general chemistry course with a higher mathematics prerequisite
under the assumption that these students have all had at least one year of high school
chemistry.

At Michigan State University, enrollment in non-major general chemistry courses
(CEM 141 and CEM 142) is always greater than the courses designed for chemistry
majors. These two courses enroll approximately 4000 students a year. The enrollment in
the courses for Bachelor of Science degree (B.S.) chemistry majors, biochemistry majors,
and chemical engineering majors (CEM 151 and CEM 152) is approximately 400
students.

Very few students enrolled in CEM 141 and CEM 142 are chemistry majors.
Consider the number of chemistry majors for the academic year of 2004-2005. There
were approximately 100 declared B.S. chemistry majors but only 11 spring semester
2005 graduates. For Bachelor of Arts (BA) chemistry majors, there were approximately

125 students with 32 graduating spring semester 2005.

Research Questions
The objective of this study was to investigate the factors that influence academic
success in college general chemistry. Most students are inquisitive about their world and
everyday phenomena and few students inherently dislike studying chemistry and other
sciences.
Each semester, a new group of students enrolls in college general chemistry.

Many of the students enrolled are first-year students with good grades in high school

chemistry, good (high) ACT and/or SAT scores, and a desire to succeed. All have had a
mathematics placement examination, or the necessary mathematics prerequisite course to
enroll in general chemistry. Despite good high school grades, competent standardized
test scores, and tests for math placement criteria, some students do not succeed in college
general chemistry. All of the elements for academic success appear to be in place, yet
some students earn a poor grade in the course.

The subject of chemistry is not new to most students who attend college,
especially those majoring in the natural sciences. Most of the material in college general
chemistry is included in the high school chemistry curriculum. Although the pace of
instruction in college is accelerated, compared to high school, much of the material looks
familiar. Moreover, those students coming from an accelerated high school program may
have experienced a more intensive level of chemistry instruction than is offered in the
first semester of general chemistry in college. Occasionally, students enrolled in college
general chemistry may not have had any high school chemistry, especially if they are
from a small high school and had to choose between taking physics or advanced biology
instead of chemistry.

However, for most students coming into the university for the first time, the large
lecture format of general chemistry is much different from the high school instructional
format. Whereas in high school, teachers have had formal teacher training, in the
university, the instructor may have had extensive training and experience in research, but
usually not a formal course in pedagogy. He or she may have a strong desire to be an

effective teacher, but is often perceived by the students as uninspiring and unimaginative.

ll

Therefore, this study examined the following major research questions:

1. What is the description of the sample and independent variables?

2. What is the relationship between key student demographics and control
variables on the final grade in general chemistry?

3. What is the relative importance of pro-college characteristics, in-class
experiences, out-of-class experiences, instructional resources, student-
focused variables, instructor-focused variables, and the final grade in
general chemistry? (r-square change by block).

4. What is the relationship between individual indicators within pre-college
characteristics, in-class experiences, out-of—class experiences, instructional
resources, student-focused variables, instructor-focused variables, and the
final grade in general chemistry? (The significance of individual

predictors by block)

12

CHAPTER 2

LITERATURE REVIEW

Introduction

The literature and research on student learning and success in higher education is
extensive and encompasses many research perspectives. Learning theorists, investigating
academic success in college, focus primarily on in-classroom and instructional variables.
Student affairs researchers, interested in retention, diversity, and equity issues, examine
admissions criteria and out-of class experiences. Admissions personnel collect data on
student demographics test scores and other pre-college characteristics. However, no
studies have been performed from a more ‘holistic’ approach, that is, to incorporate both
cognitive and non-cognitive factors, such as pre-college characteristics and college
student experiences in and out of the classroom, instructional resources, student-focused
variables, and instructor-focused variables, as they relate to academic success.

More specifically, the purpose of this study has been to use this more ‘holistic’
lens to determine the influences that affect student academic success in general chemistry
courses at Michigan State University, namely CEM 141 and CEM 142.

There are two lines of literature on college success included in this literature
review: One studies the influence and effect of college on students (college impact
studies) and the other looks at the effects of college on students in specific subject areas
(in the case of this study — chemistry and other natural sciences).

College impact studies include literature on student retention (Tinto, 1986),
student socialization (W eidman, 1989), student involvement (Astin, 1984), and student

change through the college experience (Pascarella & Terenzini, 1991, 2005). The

13

literature is important because it suggests generic models for investigating factors that
influence success in college, models which became the basis of the conceptual
framework for this study.

The literature, which looked at the effects of college on students in a variety of
courses, addressed some of the factors that influence academic success in college general
chemistry, but there was little consensus across the studies. In the end, this study’s review
of the literature demonstrates the need for research that casts a wide net into the variables

potentially influencing academic success in college general chemistry.

College Impact Studies

In 1969, Feldman and Newcomb synthesized the results of 45 years of research on
the impact of college on students. They determined that the students’ values orientation
changed considerably during their college years, i.e., students became less authoritarian,
less dogmatic, and more liberal. They discovered, too, that those students who became
more involved in college life tended to do better in their classes than those who did not
become involved. These students were ofien not satisfied with their intellectual and other
experiences in college, such as feeling that they had little contact with faculty members
outside of the classroom and that the faculty had little concern for their welfare.
Situations like these often increased college withdrawal and dropout rates.

Research has also determined that all students learn at some level in college.
Some may not learn as much as others but they still learn (Pace, 1979). Most students
increase their verbal and quantitative skills and become more competent and

knowledgeable in their major area of study (Pascarella & Terenzini, 1991). Using test

scores as an outcome variable, Pace (1979) found that seniors, as a group, know more
than sophomores. Also, Pace discovered that the more credit hours a student had had in a
specific subject area, the higher his/her achievement in that subject: “Students learn what
they study, and the more they study, the more they learn” (Pace, 1979, p. 18).

Astin (1968) identified and studied a variety of college environments. In his
study, Astin collected data on more than 30,000 students at 247 higher education
institutions. He utilized the Inventory of College Activities (ICA) consisting of 275
“stimuli” covering four different areas: the peer environment, the classroom environment,
the administrative environment, and the physical environment. His most salient
conclusion was that undergraduate students’ acquisition of knowledge or philosophical
on life, might be highly dependent on the institution he or she attends. He referred to the
college environment as “an array of potential stimuli” consisting of “any behavior, event,
or other observable characteristic of the institution which is capable of changing the
student’s sensory input, and whose existence or occurrence can be confirmed by
independent observation.” (Astin, 1968, p. 118)

Astin’s 1968 survey reflects many issues affecting today’s college students even
though the research is more than 40 years old. Of course, there were some very outdated
questions in the ICA, such as “danced the twist” or “students were permitted to smoke in
class” (Astin, 1968, p. 148), but other questions, including “drank beer” (p. 148) “felt
homesick” (p. 148) and “the instructor was enthusiastic” (p. 149) are all generic and
pertinent issues for today’s students and are included as variables in this study.

In 1971, Astin completed another comprehensive study and produced a book,

Predicting Academic Performance in College, intended primarily for high school

students moving on to some form of higher education upon graduation. In the chapter
“Who succeeds in college?” Astin illustrated that high grades in high school translated to
high grades in the freshman year of college: “Of all of the information available about the
high school student, his record of academic performance is the best single indicator of
how well he will do in college” (Astin, 1971, p.4).

Also included in Astin’s 1971 study were variables such as gender and academic
ability as measured by standardized tests (American College Test (ACT), the Scholastic
Aptitude Test (SAT), and the National Merit Scholarship Qualifying Test (NMSQT)).

He found that male students were twice as likely to receive borderline or failing grades as
college freshmen, than were female students, and he reported a positive relationship
between how well a student performed on standardized tests of academic ability and his

or her grades as a college fi'eshman.

College Impact Models

As early as 1979, Pace proposed a model for college student learning and
development. He used the term “impress” rather than “impact” to explain the effects of
college because of the characteristic mark, stamp, or impression made on students. Pace
said that the question for evaluating student development is simple, “given the students
who come here, what happens to them and what they are like when they leave” (Pace,
1979, p.125).

His Path for a Student Development and College Impress Model (1979) had three
basic propositions: the first is that students coming to college bring a basic set of

experiences, beliefs, knowledge, personal traits and other pre-college characteristics. The

second is that the college environmental factors and quality/quantity of student effort
influence these pre-college characteristics. Examples of environmental factors include
classroom experiences, interaction with peers and faculty, and participation in clubs or
organizations. The third component of the model is the outcome, which be labeled
student development and college impress.

This component in Pace’s model was seen by him as a result of the combined
influences of the normative effects of the college environment and student effort exerted
on the student’s background and characteristics. Pace refers to development as the
difference between before and after scores on criterion referenced metrics collected at the
beginning and end of college. Impress is the mark or impression left on the student as a
result of the college environment. The impress may be academic achievement or
personal feelings, such as satisfaction with college or progress to desired goals. The
impress is a lasting mark that can be measured after a student leaves college in terms of
interests, successes, outlook, philosophies, or other aspects of the college experience.

In 1991, Pascarella and Terenzini performed an extensive compilation of the
research in the 19805 which investigated how and why students change in college.
When explaining change in college, they differentiated between developmental change
models and college impact models. Developmental models tend to identify what changes
as students progress through college. Examples of developmental issues are socio-
emotional development and identity formation.

College impact models focus on how the changes occur and the sources of the

change. This includes interactions with other students and faculty, programmatic issues

17

and sources of instruction or support. What versus how is the most important distinction
between developmental and college impact models for explaining student change.

Fifteen years later, Pascarella and Terenzini (2005) did a follow-up of their
original work. In their new book, they reviewed the research in the 19903 conducted
since the publication of the initial volume in 1991. They concluded that while there
were hundreds of studies done since their original How College Aflects Students was
published in 1991, their original college impact model continued to explain how students
change in college.

Building on their earlier work with college affect studies and college impact
models, Pascarella, Edison, Nora & Terenzini (1996) investigated the experiences of
first-generation college students. In this study, using much of the recent literature on how
college experiences Shape educational outcomes for college students, they proposed a
“causal sequence” model of student success. The assumption of this model is that student
success in college is dependent on factors that begin pre-college and continue throughout
college, both in and out of the classroom. They maintained that the major impact of
college on students is the “cumulative result of a set of interrelated experiences sustained
over an extended period of time” rather than the result of any single experience
(Pascarella, et al. 1996, p. 749).

Another variation on the college impact model is Astin’s (1991) input-
environment-outcome (I-E-O) model for assessing educational outcomes in higher
education. His I-E-O model was originally used to assess Ph.D. productivity in
educational institutions. He found that by simply measuring the number of Ph.D.s

produced by an institution, or the correlation of student characteristics such as sex, major,

18

or academic ability to PhD. productivity, did not completely explain the phenomenon.
Although student ability was the most important factor for attaining a Ph.D. there was
some discrepancy between institutions. Some of the institutions that should have been
highly productive were not and some with less potential to produce Ph.D.s overproduced.

Astin also determined that outcomes, such as Ph.D. production at educational
institutions, could not be determined by single input measures. He found that simply
having input and outcome data was not enough. One needed to identify the
environmental forces of college acting upon students to influence Ph.D. production.
Input and outcome variables look at students at two different times in their educational
careers. Environmental experiences are the intervening factors that affect change. More
importantly, environmental experiences often can be controlled, changed, or otherwise
manipulated.

Astin (1984, 1993) also proposed a college impact model with his theory of
involvement. He contended that the more students are involved in college the better they
would succeed academically. He referred to involvement as the quality and quantity of
interaction students invest in the college experience. He deduced that two of the most
important factors of involvement are student interaction with peers and with faculty. He
also included place of residence as a factor in college involvement. Students who live on
campus tend to have more student friendships, a better social life with more involvement
in extracurricular activities such as athletics, and also assumed leadership roles more
frequently. They also become more liberal, social, hedonistic, and less religious.

College impact models were also used to explain student withdrawal from

college. Perhaps the most famous is Tinto’s theory of student departure (1986). His

19

longitudinal model of student departure included student pre-entry attributes,
goals/commitments (T1), institutional experiences, integration, goals/commitments (T2)
and outcomes.

Much like Astin’s theory of involvement, Tinto’s underlying theme is integration
into the college environment. The premise of his model is that students enter college
with disparate family backgrounds, academic skills, and abilities. As the student
experiences the college environment, these pro-college attributes are influenced by
interactions with the academic and social systems of the institution. If the interactions
are positive, the student integrates into the system and does not withdraw. If the
interactions are negative and integration does not occur, the student may depart.

The college impact models by Astin, Pascarella, and Tinto dealt primarily with
cognitive outcome variables, or outcomes that affect student academic success and
retention. In contrast, Weidman (1989) proposed a college impact model that can be
applied to non-cognitive variables to explain the theoretical underpinnings of college
socialization, including career choices and lifestyles.

Similar to previous college effect studies, Weidman’s model included student pre-
college or background characteristics. In his model, forces exerted by the social and
academic structure of the institution, influenced student pre-college backgrounds and
experiences. Outside the normative social and academic structure of the institution were
mediating and often conflicting effects of parental background and non-college reference
groups (such as peers, employers, and community organizations). These external
constraints could affect the student’s experiences in the structural and organizational

settings of college and influence socialization outcomes and development.

20

An additional line of research looked at both cognitive and non-cognitive factors
that influence success in college. Ting (1997, 1998) examined low achieving, specially
admitted, first-year college students. He found that data from composite ACT scores,
high school rankings and non-cognitive psychosocial variables collected from a
questionnaire were important in predicting success in college. When both cognitive and
non-cognitive variables were regressed on freshmen year grade point average, ACT
scores, high school rank, successful leadership experiences, and community service were
significant predictors of freshman year academic success. He suggested that cognitive
and non-cognitive variables might better predict academic success than cognitive
variables alone.

Richardson and Sullivan (1984) also investigated non-cognitive factors that
influenced academic success for college freshmen. They looked at 199 students with
high school grade point averages of less than 1.7 who were considered “academically
challenged” (p. 90). They found no relationship between SAT scores and end-of-year
freshmen grade point averages. There was a small positive correlation between high
school grade point average and college freshman grade point average. The other
significant predictors of college grade point average were motivation-related variables as
identified by the College Student Inventory.

Students spend much more time outside of class than in class. Two thirds of
college students’ waking hours are spent on activities other than attending lectures,
recitations, and studying. Yet most studies of student achievement in college focus on the
students’ academic experiences. According to Kuh, Schuh & Whitt (1991), “It seems

reasonable to assume that when out-of-class experiences complement the institution’s

21

educational purposes, they contribute significantly to student learning and personal
development (p. 6).

Much of Kuh’s work on seamless learning and involved colleges was related to
student development issues and not specifically to teaching and learning. His basic
premise was that out-of-class experiences provide more of a complete explanation of the
role of college in students’ lives than classroom experiences alone.

Other studies found relationships between student success and non-cognitive
variables. Milem and Berger (1997) and Berger and Milem (1999), extending the work
of Tinto ( 1986) and Astin (1993), found that early involvement of students with faculty
members is significantly related to student persistence. Pascarella, Edison, Nora,
Hagedorn & Braxton (1996) found a positive relationship between teacher
organization/preparation and specific course achievement. Pascarella, et al. found that
academic success for first-year students was influenced by the type of institution attended
(two-year or four-year), total credit hour load, perceived teacher organization,
preparation, and support, and other non-cognitive factors.

Other studies investigated the factors that predicted success for first-generation
college students using cognitive and non-cognitive variables. Nauman, Bandalos and
Gutkin (2003) looked at the differences in academic achievement between students who
were first in their family to attend college versus second-generation college students.
They found that first-generation college students did not have the same sources of
support as second-generation students and had to rely more on their own motivation to

succeed in college.

22

Nauman et al. used ACT scores, along with a questionnaire, that measured
motivational and learning strategies for identifying the most important variables for
academic success in college. They found that the variable related to expectancy for
success was the most significant predictor of success for first-generation college students,
closely followed by ACT scores. For second-generation college students, ACT scores
were the most significant predictor. They concluded that ACT scores were a valid
predicator of college grade point averages in college but should be used in conjunction
with other variables, especially for first-generation students.

The majority of the studies reviewed used the final grades in courses or overall
grade point average to quantify student learning or learning outcomes in college. Studies
of learning outcomes, however, do not always employ grades as the dependent variable.
A study by Anaya (1999) used self-reports of learning, total college grade point averages
(GPA) and performance on a standardized test (the Graduate Record Exam) as outcome
measures. Using the data from a national study of 2,289 students she regressed within-
college factors, nonacademic activities, and learning environments on these outcome
variables. Although there were mixed results, in general, college environmental
influences and learning activities in the classroom were associated with some aspect of
student learning.

An interesting finding from her study was that the gains in learning were “domain
specific,” (Anaya, 1999, p. 516) in that certain learning environments and activities
promoted specific learning outcomes. For example, students who majored in the physical
sciences and took courses emphasizing mathematics skills showed increased quantitative

gains as measured by both self-reports and GRE scores. Conversely, environments or

23

activities not associated with the learning outcome showed no association or a negative
association. The two examples used in her study were that being a guest in a professor’s
home was positively correlated to verbal learning but not quantitative learning, and the
number of foreign languages taken was negatively associated with quantitative learning.
Overall, environmental factors had some influence on college student learning (although
the influence may be either positive or negative).

In summary, college impact studies that include non-cognitive variables in
addition to the traditional cognitive variables identify better the factors that contribute to
academic success in college. Pre-college characteristics and academic background (such
as standardized test scores and high school grades) may be reliable predictors of the
effects of college on academic success, but are most likely not the only factors.
However, including non-cognitive factors, such as experiences in the classroom and
experiences outside of the classroom, and instructional resources as well as student and

instructor-focused variables, is essential to a comprehensive model of student learning.

Research on Academic Success in Chemistry and Other Physical Sciences

Although many studies on college students and academic success exist, relatively
few focus on students enrolled in college chemistry. Journals such as The Journal of
Chemical Education publish articles on pedagogical issues and practices related to
cherrristry but rarely publish empirical studies on the factors contributing to academic
success. The majority of the studies used standardized test scores in mathematics, high

school grades, and math/chemistry pre-tests as predictors of success in college chemistry.

24

Few looked at other elements, such as course-related factors and student activities out of
the classroom.

Fletcher (1978) collected archival data on 150 students at Tennessee
Technological University. He determined that the best predictor of success in college
chemistry was the composite score on the ACT. The next variable (albeit minor) that had
the most influence on the variance accounting for (R2) the dependent variable, the final
grade in college chemistry course, was a student’s grade in high school physics.

Prior to his 1978 study, Fletcher (1977) produced a summary of the research
literature predicting student success in college chemistry. He found that high school
chemistry has little effect on college chemistry, and sex, as a variable, was not a
Significant predictor of success in college chemistry. Fletcher noted one common theme
throughout the literature, namely that students with strong backgrounds in math and
science were more successful in college chemistry than those with weak backgrounds.
Strong backgrounds included four years of math and a full science sequence (chemistry,
biology, and physics).

Ozsogomonyan and Loftus (1979) looked at four independent variables as
predictors of general chemistry grades. A simple regression model containing math SAT
scores, chemistry pre-test scores, math pre-test scores, and high school chemistry grades
indicated that 35% of the variance in the final grade in general chemistry can be
explained by these four variables.

Other variables, such as the course lecture and major field (or field of study), are
potentially important factors affecting success in college general chemistry. Arrnine Paul

(1978) found that the course instructor was a significant factor for academic success.

25

Teaching assistants, however, had no effect. He also found that students in more highly
selective majors tended to perform better than other students. In his study, pre-medical
technology and pre-pharmacy students were the best performers in college chemistry.

One use of research on factors that influence academic success in college general
chemistry is to make predictions of students’ course grades. Predicting students’
chemistry course grades is a useful tool for instructors and counselors. Students at risk of
failure can be identified early (Deal 1984), and predicting course grades may be used as a
motivating factor for students who are less prepared for college. If a student’s predicted
grade is less than the desired grade, she or he may understand the need to work harder in
the course to achieve the expected or desired grade.

In his study of first-year chemistry students at the University of California at
Riverside, Deal (1984) found that the average grade of students was one grade lower than
her or his high school chemistry course. The best predictor of chemistry grades was a
combination of math SAT scores and high school chemistry grade. He reported that 73%
of students received a grade within 33 percentage points of his or her predicted grade.

He also reported that teaching assistants and supplemental assistance had no effect on
prediction of college chemistry grades.

Much of the research on factors that influence academic success in college
chemistry includes the pre-college variable, high school chemistry. There is some
inconsistency, however, in the literature regarding the variable of high school chemistry
and its effect on college chemistry achievement. Some contend that high school

chemistry has little influence on college chemistry (Fletcher 1977; Willingham, 1985).

26

Others found high school chemistry to be of some value for the study of college
chemistry, although high school mathematics may be a better predictor (Ogden, 1976).

Prior studies on the relationship between high school chemistry, math, and other
high school science courses and success in college were primarily based on correlating
high school grades with college grades. Few non-cognitive variables were investigated.
Yager (1989) found that for high ability students, prior high school courses in chemistry
and physics had little effect on success in college chemistry and physics. He collected
data on 60 students over a two-year period. One student group completed high school
chemistry and physics while the other did not. At the end of the courses in college
chemistry and physics, there was no difference between the two groups of students. He
suggested that students with prior high school coursework might become less attentive
and prone to poor study habits when compared to students without prior coursework who
had to spend more time studying and using tutors.

J. Daniel House in 1995 conducted a study on the predictive relationship between
initial student attitudes, including academic self-concept and expectation of academic
performance, and years of high school math and college admissions test scores with
subsequent success in introductory college chemistry. He concluded that student
attitudes were significant predictors for success in freshman chemistry. Furthermore,
students’ initial attitudes were better predictors of course GPA than ACT composite
scores or years of high school math completed. Background characteristic variables, such
as family characteristics or prior academic achievement of students, were not part of the
study. However, the author acknowledged that including such variables would provide

better guidance for further research.

27

Although most of the literature reviewed for this study was from four-year
institutions, some literature from community colleges was included as well. Boughan
(1996) investigated the correlates of positive performance in Chemistry 101 at a
community college. He found three primary factors for success in the course. The best
predictor was cumulative college grade point average. The second was a scale he termed
socio-educational maturity, which combined age, delayed entry into college, full-time
employment, and whether or not the student transferred from another academic
institution. He posited that students, who were adult learners and typically worked full-
time, were older, had transferred or delayed entry into college, and had the maturity and
real world experience to meet the demands of college. The third significant factor was
whether or not students were placed in developmental reading courses; those that
received reading support performed lower than those who were not in such support
classes. Nevertheless, these three factors accounted for only 35% of the variance to the
final grade in the course. He noted that future studies should include other factors such
as motivation, study habits, and family background.

More of the recent studies on success in college chemistry use both cognitive and
non-cognitive variables in their models. Burdge and Daubenmire (2001) looked at
several variables for increasing success in college chemistry and reducing attrition. They
found that older students tended to get higher grades. Also, students who took chemistry
later in their college careers did better than those taking the course as first-year students.
Higher verbal scores on standardized tests were also correlated with greater success in
college chemistry. The composite test scores were better predictors than math scores

alone, according to this study.

28

Burdge and Daubenmire also found that most students had the necessary math
skills to answer test questions correctly (such as the ability to do ratios or solve for “x”),
but had trouble gleaning the necessary information from the text to answer written test
questions correctly and set up and solve exam problems. In addition, they found that
repeating the course had a negative relationship with success. Students taking the course
more than twice had a higher failure rate than those taking it for the first or second time.
In their study, no student who took the course more than three times passed the class.

General chemistry in large research universities is traditionally taught in the large
lecture format. New chemical compounds and analytic processes are discovered every
year but the classic scene of the chemistry instructor standing in front of the room
muttering to the chalkboard has not changed much in the past 100 years. The instructor
stands in the front of the room solving problems on the chalkboard while the students
feverishly take notes. Words travel fiom the mouth of the instructor to the ears of the
student and on to the paper or note pad. Later, when the student attempts to solve a
similar problem they find they are often unsuccessful. The solution to the chemistry
problem the instructor worked on the chalkboard appeared straightforward while the
student was taking notes during lecture, but a few hours later, the student discovers he or
she is unable to solve the problem on their own.

One of the tenets of the information-processing theory (Gagne & Medsker, 1996)
is that the body’s sensory preceptors are constantly bombarded with stimuli (sights,
sounds, smells, tastes). Immediately, the sensory register determines whether the
information should be moved to the short-term working memory, or filtered out. In

short-term working memory the information must go through rehearsal and then encoded

29

for permanent storage before it is available for search and retrieval later on. The act of
learning is the process of encoding information for later retrieval and use. That is the
primary problem with taking notes and not processing the information. It is not enough
to simply write down on paper what is heard in lecture. The information must be
processed for later retrieval and use.

There are a variety of ways to encode information for storage and retrieval in

long-term memory. Two examples are active learning and collaborative learning.

Active Learning

In active learning, students are engaged learners rather than passive learners.
Active learning takes place when students are involved in the learning process. But more
important than just simple involvement and activity is engaging students in the physical
and mental activities that require students to think about what they’re learning and how it
becomes meaningful to them (Angelo, 1993).

One technique for promoting active learning is class discussions where learners
answer questions for themselves and other students in class. Another is the use of
response cards or finger signals (Paulson, 1999). With either of these two methods, the
instructor can ask a question and get a response from the students. For example, the
instructor can ask a true or false question. The students can hold up a red card for true,
green card for false or a yellow card if they do not know. The same is true for fingers.
One finger for true, two fingers for false, and/or a fist for “I don’t’ know”.

A new and innovative variation on the response card technique is an electronic

response system. This system uses a device similar to a television remote, except that it

30

transmits the answers from the students to a computer screen. These electronic response
systems or “clickers” may also be used for in-class quizzes or to take attendance. Each
student has a personal code assigned to his or her clicker and their responses are
registered as a group or individually.

Another method is the “one minute paper,” (Harwood, 1993) which gives the
students a few minutes at the end of class to reflect on a question. This approach gives
the instructor a starting point for the next class session. In all cases, the instructor can get
a relatively anonymous assessment of whether or not students understand the material
presented and readjust it as necessary.

When using active learning techniques, there is less emphasis on the transmission
method of the teaching. Active learning involves students thinking, discovering,
processing, and applying what they have learned in the classroom. The transmission
method of teaching is instructor-centered. The assumption with the transmission method
is that there is a stable body of knowledge that must be learned by students and there is a
specific and effective way to impart the knowledge (Pratt, 1998). The transmission
method in some courses, such as chemistry, is an effective method of teaching when the

new material is related to prior knowledge that students already possess.

Collaborative Learning

Collaborative learning involves students working together to achieve a common
goal. Formation of study groups is a form of collaborative learning. Students can get
help from other students. Explaining a concept is a good way to enhance learning. You

have to know a concept very well in order to teach it to some one else. Moreover, with

31

collaborative learning, students can ask questions of one another in a non-threatening
environment.

An alternative form of collaborative learning is in-class activities and out-of—class
group homework assignments. Students, who participate in group activities, whether in
or out of the classroom, learn to work together to find a solution to a problem. This
participation provides help for learning the course material as well as experience for the
working world, where collaboration is the norm rather than the exception.

One final collaborative learning technique is to establish a fixed or absolute
grading scale for the course. The competitive nature to earn high grades in gatekeeper
courses is exacerbated in curved grading system. Some students may be motivated by the
competition to earn a high grade. Others may be discouraged or turned off by the notion
of competing for a grade in a course. A fixed grading scale theoretically allows every
student in the class to earn a high grade in the course (or to fail the course). No student

has to worry about raising the curve in the grading scale by helping his or her peers.

Chapter Summary
The literature on academic success in college is extensive. Some of the literature
on factors that influence success in college is subject Specific and includes research on
the factors that influence success in college chemistry.
The earliest research consisted of correlational studies that focused on academic
achievement as measured primarily by scores on standardized tests. Some of the studies
investigating college chemistry found a relationship between standardized test scores and

high school performance with overall academic success in general chemistry. Others

32

looked at high school. grades and coursework in math and science, in addition to overall
high school grades.

Although many of the early studies found some correlation between cognitive
variables and academic success in chemistry, there has never been consensus across the
literature that academic achievement (as measured by standardized test scores) or
performance in high school (as measured by high school GPA) is the best predictor of
academic success in college chemistry. Furthermore, adding non-cognitive variables to
the equation increases the accuracy of predicting academic success in college.

A perfect correlation between math ability, standardized test scores, high school
grade point average, etc., with academic success in college general chemistry, would be
neat and tidy. With prior academic data and a simple regression equation, the final grade
of any given student could be predicted. The literature, however, maintains that there are
other non-cognitive variables that influence academic success in college chemistry.

If cognitive variables were the only predictors for academic success in college
chemistry then there would be little incentive to do further research. Educators and
policymakers can do little to influence prior academic achievement and other pre-college
variables affecting academic success in college without major societal changes over
generations of students. The only variables policymakers can influence are non-cognitive
variables related to college students’ experiences inside and outside of the classroom,
instructional resources, student focus variables, and instructor focus variables. From a
policy perspective, non-cognitive variables are extremely important when investigating

the factors that influence academic success in college general chemistry.

33

To obtain the most precise and general description of the factors that influence
academic success in general chemistry both cognitive and non-cognitive variables were
included in the research design. To date, there are no studies investigating success in
general chemistry using both cognitive and non-cognitive variables in relation to a
student learning model.

There was a large body of literature on academic success in higher education, and
it cut across all areas of higher education from community colleges to public and private
institutions. Some of the literature was subject-specific and some was not. With such a
diverse body of literature it was difficult to limit the number of constructs for the
conceptual model created for this study. The student learning model for this study
borrowed heavily from other research on college student experiences.

The most comprehensive lens to use for viewing student learning is to integrate
the cognitive and non-cognitive elements, including social-emotional behavior, by
linking student in-class and out-of-class experiences (Cove & Love, 1996). This provides
a broader view than viewing student learning from a single perspective.

Assessing student pre-college characteristics was an element to common to most
of the previous studies on academic success in college. In some of the studies, pre-
college characteristics were as simple as using SAT or ACT scores. Others were more
elaborate using high school data and parental SES. I chose to use all of the pro-college
characteristics suggested in the literature when constructing the model for this study.

Most of the studies in this literature review looked at some form of instruction as
a construct. Some of the studies were experiments with various instructional methods.

This study did not incorporate instructional methods as a construct in the model. In class

34

experiences and instructor focused variables assessed student perceptions of their
instructor skills. There was no or experimentation with specific instructional methods.
The focus of this study was the influence of students’ pre-college characteristics,
experiences in the classroom and experiences outside of the classroom on the their final

grade in the general chemistry.

35

CHAPTER 3

METHOD

Introduction
The research methods used for this study are presented in this chapter. They
include the conceptual framework, the variables in this study, and the study design. The
chapter also includes a description of the population and respondent sample, a discussion
of the development and administration of the survey instrument and focus groups, and a
description of the pilot study. There is also a restatement of the research questions and a
description of the data analysis techniques related to each question. Finally, a summary

of the study methods and procedures is presented.

Conceptual Framework

The conceptual framework for this study was derived from a variety of college
impact models. College impact models look at the process and origin of student change
(Pascarella & Terenzini, 1991). This approach is in contrast to other models that examine
internal characteristics of students such as affective, social, and moral development.
Models that focus primarily on internal characteristics of student change omit the
environmental effect when investigating student change (Kuh, 1991).

College impact models have been utilized extensively for investigating a variety
of outcomes related to college students such as learning, retention, and socialization.
Astin’s (1968) investigation of the college environment, Tinto’s (1986) theory of student
departure, Weidman’s (1986) evaluation of college as a socialization process, Pascarella

and Terenzini’s (1991, 2005), How college aflects students, and the investigation of first-

36

generation college students by Pascarella, Edison, Hagedorn, Nora and Terenzini (1996)
are all variations on a generic college impact model. Prior to these studies there was little
research on academic success in college that considered both current academic conditions
and prior life events on academic success.

Pascarella’s (1991, 2005) general model for assessing change consists of five sets
of institutional and environmental variables. These include structural/organizational
characteristics of institutions, student background and pre-college traits, interactions with
agents of socialization (faculty and peers), institutional environment, and quality of
student effort.

The conceptual framework for this study is a college impact model. It is a
variation of Pascarella’s student learning model (Pascarella & Terenzini, 1991, 2005;
Pascarella, et. al., 1996) and Tinto’s longitudinal model of institutional departure (Tinto,
1993)

Pascarella’s model was modified by removing the construct of “Quality of
Student Effort” and “Structural/Organizational Characteristics of Institutions,” such as
enrollment, faculty-student ratio, selectively and percent residential students, which were
not needed for this study. The balance of the constructs is similar except that the college
impact model used for this study includes student out-of-classroom experiences that were
not included in Pascarella’s model. He included “Interactions with agents of
socialization,” such as faculty and peers, which is related to the college experiences
outside of the classroom construct contained in the college impact model used for this

study.

37

Tinto’s model posits that students enter college with a specific set of skills,
attitudes, and family background variables. The decision to stay or leave the institution is
determined through a series of formal and informal interactions with individuals and the
social structure of the institution. Tinto’s model was originally designed to study student
attrition, but was successful in predicting other student outcomes as well (Pascarella &
Terenzini, 2005). Tinto included influences from the external environment, such as
family and other commitments. Because this particular study was primarily concerned
with the internal institutional influence of student academic success, the college impact
model modified for this study omitted the external environmental effects.

The college impact model developed for this study consists of six basic constructs
and an outcome. These are the result of a factor analysis conducted on the data collected

on the General Chemistry Survey, given to a cohort of college students.

1. Pre-college characteristics including high school course/grades and parent
education/ income.

2. In-class experiences, such as course organization and satisfaction.

3. Out-of-classroom experiences including social adjustment and time

management.
Instructional resources, such as usefulness and availability.
Student factors including study habits and time commitments.

Instructor factors, such as skills, attitude, and availability.

S95”?

Outcome, which is the final grade in the course.

This student-leaming model was used as the conceptual framework for
investigating factors that contribute to positive educational outcomes in general
chemistry. It served as a map for guiding the data collection and analysis (Ostrom &

Rogers, 1996). Figure 3.1 presents this conceptual framework.

38

an

338:3 8:35 mwezob 2R 3103:8393 k 3338389 ._.m oSwE

 

 

 

5 2928536..
86a 38 5 .
£3396 68.52. .
8353 .925»... o

2....» 8896a
29: .68ng .
£53 033 .90:me .
0338.38235

 

co=m=Em .mcozoucmuco
$305.00: E3 :0 o>3
co=m=Em co_~m~_cm90
cozooccoo oz. maaEmo
mc_~__m_8w

cote 29:88. 322
mucoE 953.338 5 3oz
“5.53.6.3 cm. 2.950

398.398 3393.50

 

 

 

 

 

L

 

/ baa—EOp—O

 

 

8:28 6.388 .98
3:956:38 oEF
can. seem
EoEgEEoo 3.300
60333035036

3.2.3 5 32a .23.“.

 

 

 

 

5:03ng 02:00
Eon—>023 3.500
22:28 609330
5.83.590 0230

305.3on 329:.

 

 

.\

 

 

 

 

 

£aE dO< .59: ho<

ano .9 .0532 .mco.._.0<
EmEofing Lo PS .82.
oEoo:<:o:ao=uo 9:29.
8:23 9.. s 385

3.3.3quan
32.335

 

63:00 58.52... .
30.58 .28

Lo .6655 E0: 22.. 6

Eco. 90... on: c
89:82 65:0ng

.0 $8.23: .

much—One

3:28.52.

 

 

 

 

39

Components of the College Impact Model

The College Impact Model depicted in Figure 3.1 is a modified version of
Pascarella’s Student Learning Model. This model, which is the conceptual framework for
the study, delineates six primary factors that influence academic success in college
general chemistry. These factors are: students’ pro-college experiences, college
experiences in the classroom and college experiences outside of the classroom,
instructional resources, student-focused variables, and instructor-focused variables. The
six factors in the model are observed according to their relationship to each other and to
the dependent variable, which is the final grade in general chemistry.

The final grade in the course was chosen as the dependent variable for this study
because it was the most frequently used outcome variable in the literature on college
academic success (Bauer & Liang, 2003, Broughan, 1996, Fletcher, 1978, House, 1995,
Pascarella & Terenzini 1991, 2005, Pritchard & Wilson, 2003, Sanchez & Betkouski,
1986, Yager, 1989). One study (Anaya, 1999), advocated using student self-reported
gain scores in lieu of final grades as an outcome measure. Anaya contested the validity
of final course grades as an outcome measure because there were no pretests given for
comparison with the final grade. She also questioned the reliability of grades because
they are a non-standardized measure and 3.8 at one institution may not mean the same at
another.

Pascarella’s causal model for student success, (which was modified for this
study) used grades to explain learning and cognitive development. In the 2005 follow-up
of How college affects students: Findings and insights from twenty Years of Research,

Pascarella and Terenzini (2005) note, “Our 1991 synthesis concluded that academic

4o

achievement indicated by grades, is among the most revealing indicators for students’
successful adjustment to the intellectual and other demands of a course of study” (p. 618).
Using the final grade was also helpful for the data analysis and data collection for
this study. Final grade was a ratio variable with defined zero value that was common to
every case in the dataset. Moreover, it was a variable available to the participants and
non-participants and was the only way to compare the two groups for any response bias.
To operationalize the conceptual model proposed for this study the factors that
affect academic success in general chemistry were first looked at and then used to create
the General Chemistry Survey. The General Chemistry Survey (GCS) was a 91-item
instrument that corresponded to the constructs contained in the college impact model (see
Appendix A). The questions on the survey were constructed from the literature on
academic success in college and personal observations and experiences from the course

structure of CEM 141 and CEM 142 at Michigan State University.

Factor Analysis

There were 91 questions on the GCS. In addition, ACT scores, AOP math scores
and the final grade in CEM 141 and CEM 142 were obtained by signed informed consent
from the Michigan State University Student Information System (SIS). There were other
control variables obtained from the SIS that were not used in the factor analysis. These
included, age, sex, ethnicity, course repeat status, residency, major, level (first-year,
sophomore, junior, senior, graduate student) and Honor’s College status. Some of the
data obtained from the SIS were not used in the data analysis. However, it was easier to

collect all the data available and discard any that were not needed.

41

When the Pascarella Student learning model was first modified for this study
three overarching constructs were included that corresponded to the three possible
influences on academic success in college. The first was pro-college factors, including
academic achievement as measured by standardized test scores. The next two were
relatively straightforward; students spent their time either in the classroom or out of the
classroom.

The General Chemistry Survey (GCS) was then developed with the questions
derived primarily from the literature on academic success in college and items unique to
CEM 41 and CEM 142 at MSU. The GCS was constructed for data collection, which
necessitated keeping three primary categories of questions in mind: pre-college
characteristics and experiences, in-class experiences, and out-of-class experiences.

However, after some thought and consideration, the notion of only having three
constructs in the model appeared too simplistic. The additional constructs of the
instructional resources, student focus, and instructor focus needed to be added. This was
because of noting how other models were differentiated in the literature and how they
logically fit together from my own observations and experiences.

In addition, as a result of the factor analysis, the variables did not cluster neatly
under each construct as intended with the original model. There were variables fi'om
distinct constructs clustered together in the factor analysis. For example, two questions on
the survey related to attendance at lecture and recitation were originally grouped together
under the in-class experiences construct. Another question asked about doing Learning
On-line Computer Assisted Approach homework (LON-CAPA), was an out-of-classroom

question. The factor analysis combined these three questions into one cluster that did not

42

fit under the in-class or out-class construct so a new construct entitled student focus was
created. Additional variables logically also fit under the student focus construct, such as
commitment to the course and study habits.

The literature was very valuable in constructing the GCS. There was a high
degree of consistency in the types of questions asked relative to academic success in
college. The literature was inconsistent, however, regarding the categorization of the
variables. This inconsistency was primarily because of the number and types of
questions asked. Some asked very specific questions, while some were more global. I
borrowed many of the questions contained in The College Student Experiences
Questionnaire (CSEQ) and The National Survey of Student Engagement (N SSE) for GCS
and applied them to the constructs in the conceptual model used for this study.

The 91 questions on the GCS were reduced to the 29 variables on the College
Impact Model using factor analysis. Factor analysis is a data reduction method that looks
at the correlation patterns of the data and attempts to determine the variables that account
for most the variance in the data set. The data were reduced to summated scales that are
an arithmetic average of the variables extracted from the structure matrix in the factor
analysis.

The factor analysis extraction method for the data reduction of the questions on
the GCS was principal component analysis with oblique rotation. The variables were
entered according to each construct in the conceptual model.

The following six blocks contain the factor-derived scales from the data collected
on the General Chemistry Survey. The questions on the General Chemistry Survey were

determined primarily from the literature and The Inventory of College Experiences (ICA)

43

(Astin, 1968), The College Student Experiences Questionnaire (CSEQ) (Pace & Kuh,
1998), National Survey on Student Learning (N SSL, 1995) and The Non-cognitive
Questionnaire (NCQ) (Tracy & Sedlacek, 1984). Some of the questions on the survey
were specific to the chemistry courses at MSU (i.e., questions regarding LON-CAPA).
These blocks of data corresponded to the college impact models in the literature
but were primarily a modified Pascarella College Impact Model. The six blocks of data

collected and analyzed for this study are as follows:

Block One: Pre-College Characteristics and Experiences

0 Grades in high school sciences
- Years high school sciences
0 Parental education/income

o ACT-English/Reading/Science/Comprehensive scores

0 ACT/AOP math

The first block of data represented student characteristics and experiences prior to
enrolling in college general chemistry. These pre-college experiences included data
related to high school coursework, performance in math and science courses, parental
education/income, and standardized test scores.

The block of pre-college characteristics and experiences were included in the
model because they were a part of all the models of student learning and college impact
included in the review of literature. The models that were modified to construct the
College Impact Model (Pascarella,l991, 2005; Tinto, 1996, Weidman, 1989) included

pre-college characteristics and experiences as the first construct in their models.

Pre-college characteristics and experiences are an important construct because
other studies have shown that variables, such as high school data and ACT (or SAT)
scores, account for the largest variance when explaining college outcomes (Astin, 1971;
Bauer & Liang, 2003; Pascarella, et al., 1996, Ting, 1997, 1998).

In addition, many of the variables in pre-college characteristics are relatively
convenient data to collect. Pre-college data are usually archival data included on
student’s application records. Data, such as standardized test scores, are reported to the
institution directly from the testing organization. Because they are not self-reported,
there were fewer errors and missing data. Moreover, because most standardized test
scores such as the ACT used in this study are norm-referenced the validity of the variable
is increased and it is can be used as a control variable in the data analysis.

High school data was included on the survey, which contained self-reported data
regarding the years of mathematics and chemistry taken and if students had completed a
physics course. The data also included the grades from all high school math and science
courses.

Other pre-college characteristics collected on the survey were factors related to
parental education and income. The questions asked the students to choose the highest
educational level attained by their mothers and fathers. The choices ranged from high
school graduate to PhD. It also asked them to estimate their parents’ combined yearly
gross income.

The final group of variables presented in the pre-college characteristics block was
data collected on ACT and AOP math scores from the SIS. Many of the students

participating in the study took both the SAT and ACT tests. The majority submitted their

45

ACT scores (84.5%) so those data were used for this study. The ACT data included sub-
scores in math, English, science, and reading, in addition to the composite score.

Along with ACT scores data were the scores from the math placement tests given
during summer academic orientation at MSU. All students must have a math course as
part of their general education requirement. Students are placed in a math course at
summer academic orientation based on the score on their math placement exam. The
math placement exam consists of a proctored test given at academic orientation, or a non-
proctored test that students can take on the computer at home prior to attending the
academic orientation session. The Michigan State University Department of Mathematics
accepts the results of either test for placement purposes. The majority of students took

the non-proctored math placement test (n= 510, 67.1 .%).

Block Two: In-Class Experiences

0 Course organization
0 Classroom conditions
0 Course enjoyment

0 Course satisfaction

The second component of the student learning model involved the examination of
college experiences in the classroom. Classroom experience data included student
perceptions of course organization, the classroom conditions, and satisfaction and
enjoyment of the course.

The questions on the GCS related to course organization were taken from the

Inventory of College Activities (ICA) (Astin, 1968), College Student Experiences

46

Questionnaire (CSEQ) (Pace & Kuh, 1998), National Survey of Student Engagement
(N SSE) (2005) and other studies related to academic success in college (Hagedorn, et a1,
1999; Pascarella, et a1, 1996).

The ICA was a 400-item questionnaire covering four broad categories including;
the peer environment, the classroom environment, the administrative environment and the
physical environment. The CSEQ asked students about how they spent their time at
college in relation to faculty, friends, classes, extra curricular activities, employment and
use of campus facilities such as the library. The NSSE asked the students to report on
the frequency of their engagement in specific activities. These included hours per week
doing schoolwork, extracurricular activities, employment, family commitments and study
patterns.

Astin, in the ICA listed his course organization variables as part of the classroom
environment. Pace and Kuh in the CSEQ listed the course organization variables (such as
“course goals and requirements are clearly explained”) under Perceived Teacher
Organization and Preparation. Hagedom, et al., using NSSL data, listed the course
organization variables and instructor variables under the Perceptions of Teaching
construct.

In the College Impact Model used for this study, the variables related to course
organization were included under in-class experiences because there was a separate
construct for instructor-focused variables. The items on the GCS related to course
organization asked students if the course goals were clear and the syllabus accurate. They

also indicated on the survey whether or not they understood how their grade was

47

calculated in the course, if the course calendar was strictly followed, and if they
perceived the course as well organized overall.

The classroom conditions variable was constructed of questions that asked
whether or not students could see and hear the instructor in lecture, as well as being able
to see the chalkboard or overhead projector. These questions were not included in any of
the literature reviewed for this study but were added to the GCS, because the capacity of
the lecture hall for CEM 141 and CEM 142 is 432 students, and it was important to know
if seeing or hearing the professor effectively had any influence on academic success in
the two courses.

Pace (1984) concluded, “Students who are the most satisfied with college put the
most into it and get the most out of it” (p. 52). The CSEQ and most of the other
questionnaires that collected data for national data sets included items on satisfaction
with college as a whole but did not address individual college courses.

The course satisfaction variable asked students if they were satisfied with their
performance in the course and if their scores on exams were about what they expected.
There was also a question that asked about their satisfaction with the course overall.

The in-class experience block included a variable related to overall course
enjoyment. This variable was composed of questions that asked if students enjoyed
chemistry and if they would have taken chemistry if it were not a required course for their

major.

48

Block Three: Out-of-Class Experiences

0 Campus life adjustment
0 New or existing friends

0 Knew roommate prior

0 Socializing

0 Campus life connection
0 Organization affiliation

0 Live on party floor/house

The third component of the student learning model was college experiences out of
the classroom. As much as 85% of an undergraduate student’s waking hours are spent
out of the classroom (Terenzini, 1991). Moreover, the transition from high school to
college is a form of “culture shock” (Pascarella & Terenzini, 1991, p. 58). For some
students this is the period in their lives when they begin to re-evaluate their goals and/or
values through encounters with new ideas, new teachers, and new friends. College
allows students to meet individuals with varied backgrounds and beliefs. College also
provides students with opportunities for new freedoms, along with new academic,
personal and social demands.

There were 18 questions on the survey related to out-of-class experiences (the
most of any construct on the conceptual model). The CSEQ and NSSE asked questions
regarding working for pay or volunteer work, and affiliation with clubs or organizations.
The NSSE was a bit more specific and included one global question that asked the

number of hours per week students spent in co-curricular activities such as student

49

organizations, student government, fraternities or sororities, and intercollegiate or
intramural sports.

Astin’s ICA (1968) asked very detailed questions about students’ out-of-class
experiences in his peer environment construct. For example, regarding alcohol use, in
addition to asking if students became intoxicated he also asked whether or not they drank
beer, whiskey, or other hard liquor or drank wine. The GCS used in this study asked if
students drank alcohol and went to parties but did not get specific about the type of
alcohol consumed.

The ICA also included questions on student adjustment, such as homesickness
and college roommates. It asked if the climate or weather of the university attended was
different from that of the student’s hometown. Although an interesting question, since
8% of the students in this study who completed the GCS were from Michigan, it was not
included as a variable.

However, the GCS did include questions related to student adjustment such as
whether or not students ever got homesick, if they would have liked to attend college
closer to home, if they had considered leaving college, or thought it was a mistake to
come to MSU in the first place. Similar to the ICA, there were also questions on
socializing, which included drinking alcohol or going to parties and dating. Another
variable asked if the respondent knew her or his roommate prior to coming to MSU, or if
they were fiiends. The new or existing fi'iends’ variable on the GCS asked the same
about their fiiendships and whether they made new fiiends when they came to college, or

they knew their friends prior to coming to MSU.

50

The CSEQ included questions on living on or off campus and whether or not the
students lived with parents. Pace (1984) concluded that living on campus had some
influence on students’ social and personal development as long as they were involved in
activities while on campus. However, campus residence was neutral rather than positive
with respect to general education and intellectual skills in college. The GCS included
questions on whether or not students lived on campus or off campus, were commuter

students, or lived at home with their parents.

Block Four: Instructional Resources

0 Usefulness of instructional resources
0 Use help room

0 Help fiom friends other sources

0 Instructor contact

Block four of the conceptual model included variables related to instructional
resources for the course. The CSEQ and ICA asked about completing reading
assignments but did not ask specifically about reading course textbooks. The CSEQ,
ICA, and other instruments used to collect data for national data sets were not able to ask
questions about specific courses. When they asked whether or not student do the
assigned reading for the course, it may or may not mean reading the textbook because
many courses use course packs, instructor produced handouts or media, other than
textbooks.

In CEM 141 and CEM 142 there is a textbook and study guide available in the

bookstores. Both are recommended for the course, but not required. However, most

51

students purchase the textbook for the course. The same textbook is used for CEM 141
and CEM 142. The study guide that accompanies the textbook was written by one of the
course professors. The GCS asked students about the usefulness of the textbook and
accompanying study guide.

There were also questions about the help room and how it was used, in addition to
questions about their sources of help with chemistry, such as fiiends or other students in
the class. The literature reviewed for this study did not include information on student
use of help rooms.

The instructor contact related variables were common in the literature. The ICA,
CSEQ and NSSE asked whether or not students met with their instructors outside of
class. The ICA asked if their instructor called the students by their first name. The GCA
used for this study asked the students in CEM 141 and CEM 142 if they took advantage

of their instructor’s office hours and whether or not the instructor knew their names.

Black Five: Student-Focused Variables

0 Course commitment

0 Study habits

0 Time commitments

0 Total credits enrolled

The fifth component of the conceptual model was student-focused variables. The
data collected for these were related to their commitment to the course, study habits, and

time commitments, including the total number of credits enrolled.

52

Questions on the GCS that assessed course commitment included attendance at
lectures and recitations. Interestingly, none of the questionnaires reviewed in the
literature asked if students attended lecture or recitation. Attendance in CEM 141 and
CEM 142 is not required. The question of whether or not students attended lecture and
recitation was included on the GCS to determine whether or not there was a relationship
between attending class and final grade in the course.

There were some questions on the GCS that were specific to students in CEM 141
and CEM 142. For example, CEM 141 and CEM 142 are large lecture courses. The only
way to provide graded practice problems for the course is to use LON-CAPA homework.
(LON-CAPA was developed at MSU to provide individual homework problems for
students, especially in large lecture courses. Courses at MSU in mathematics, chemistry,
physics and biochemistry are some of the departments that utilize LON-CAPA as
homework).

One question common in the literature was relative to study habits. The ICA,
CSEQ, NSSA and others (Hagedorn, et a1, 1999; Kuh, et a1, 1997; Milem & Berger,
1997; Pascarella, et a1, 1996; Whitt, et a1, 2003) included questions relating to when and
how students study. The questions on the GCA that looked at study habits of students
enrolled in CEM 141 and CEM 142 are the number of hours per week they study
chemistry and whether or not they attempted to do practice problems other than the
assigned LON-CAPA homework.

The total credits enrolled during the semester seemed to be an important variable.
Questions on the CSEQ were very specific on the number and type of courses taken.

Pascarella, et a1 (1996), utilizing the CSEQ, asked about the number of pre-

53

professional/technical, arts and humanities, and science and engineering course taken.
The GCS only included questions asking for total physical science credits enrolled and
total non-physical science credits enrolled for the semester, because most of the students
were first-year students and did not yet have the opportunity to take other courses related

to their major.

Block Six: Instructor-Focused Variables

o Instructor basic skills

0 Instructor more advanced skills

0 Instructor attitudes

o Instructor available/office hours adequate

0 Teaching Assistant (TA) skills

The sixth component of the conceptual model was instructor-focused variables,
and there were 17 questions on the survey related to these. This component and out-of-
class experiences accounted for more than one third of the questions on the 91-item
survey.

Classroom instruction is an important part of the college student experience. As
in high school, college students remember their best teachers fondly and worst teachers
less so. Teaching excellence is the goal for most instructors. There are five
characteristics that students attribute to excellent college teachers. These include
enthusiasm, clarity, preparation/organization, stimulation, and love of knowledge

(Sherman, T. M., Armistead, L. P., Fowler, F., Barksdale, M. A. & Reif, G).

54

The items in this construct were taken from the NSSE (in Hagedorn, et a1, 1999;
Whitt, et a1, 2003), CESQ (and in Kuh, et a1, 1997; Pascarella, et a1, 1996) and the ICA.
Questions relative to professors and instruction were numerous and varied. They ran the
gamut from asking for assistance in coursework to discussing personal problems and
career goals. There were also questions about instructor preparedness, skills/clarity,
support, and feedback provided. The ICA asked about teaching assistants.

Questions on the GCS related to the instructor-focused construct included
questions relative to student perception of their instructor’s basic teaching skills, such as
whether or not the instructor was prepared for lecture, if the instructor spoke intelligibly,
and whether or not the instructor made mistakes in the lecture. There were questions
about more advanced teaching skills including asking and answering questions during
lecture, using good examples, and reviewing and summarizing the material at the close of
lecture.

Instructor attitude was also an important part of the instructor focus component of
the conceptual model. These questions asked whether students perceived that their
instructor cared whether or not they learned chemistry, sounded excited, used humor and
liked teaching chemistry overall. Instructor focus variables also included questions on
instructor availability, adequacy of office hours, and availability during office hours.

Graduate students teach all the recitations in CEM 141 and CEM 142. Most of
the teaching assistants are international students. There were two questions on the survey
related to teaching assistants. One was whether or not the students understood their
teaching assistant and the other was whether or not they thought their teaching assistant

was doing a good job in recitation. None of the questionnaires reviewed in the literature

55

for this study asked specific questions about international teaching assistants. Because
issues regarding international teaching assistants were voiced in the focus groups, two
questions were added to the General Chemistry Survey asking student perceptions of
their teaching assistants in recitation.

The dependent variable in the College Impact Model used in this study was the
assessment of academic success in general chemistry at Michigan State University as
measured by the final grade at the end of the semester on a standard 0.0 to 4.0 scale. The
final grade in general chemistry is based on a total number of points earned on exams and
homework according to a fixed grading scale. The grading scale for CEM 141 and CEM
142 is fixed and unchanged for the past 5 years. It is as follows:
>800=4.0, >740=3.5, >680=3.0, >620=2.5, >560=2.0, >500=1.5, >440=1.0, <440=0.0

Other methods of assessment, such as authentic assessment or performance
assessment based on genuine or real examples of students’ work, are not used in general
chemistry at Michigan State University. The only evidence that students have achieved

the goals of the course is through multiple choice examinations and homework.

Research Questions
The objective of this study was to investigate the factors that influence academic
success in college general chemistry. The major research questions of this study were:
1. What is the description of the sample and independent variables?
2. What is the relationship between key student demographics and control

variables on the final grade in general chemistry?

56

3. What is the relative importance of pre-college characteristics, in-class
experiences, out-of-class experiences, instructional resources, student-focused
variables, instructor-focused variables, and the final grade in general
chemistry? (r-square change by block).

4. What is the relationship between individual indicators within pre-college
characteristics, in-class experiences, out-of-class experiences, instructional
resources, student-focused variables, instructor-focused variables, and the
final grade in general chemistry? (The significance of individual predictors by

block)

Study Participants
The participants for this study were students enrolled in CEM 141 and CEM 142 -
the general chemistry courses offered at Michigan State University. CEM 141 and CEM
142 are offered every semester during the year. The enrollment, however, is highest
during the fall semester. For the spring 2003 through fall 2003 the enrollment was as
follows:
Table 1.

Spring, 2003 - Fall, 2003 Enrollment in CEM141 and CEM 142

CEM141 CEM142

 

Spring semester 2003 927 626
Summer semester 2003 106 79
Fall semester 2003 1966 284

57

There are a total of 65 sections of the CEM 141 schedule during the fall semester.
The maximum enrollment per section is 30 students. The lecture hall for CEM 141 has a
capacity of 432 students, thus the 65 sections are divided into five 50-minute lectures per
day, three times per week (Monday, Wednesday and Friday). There are 13 sections per
lecture. In addition, each section has a 50-minute recitation per week for a total of four
contact hours for the course. The lectures meet at various times during the day.
Attendance is not required at lecture or recitation.

CEM 142 meets for two 90-minute lectures per week (Tuesday and Thursday).
There is also a recitation that meets for one 50-miunion session per week. There are 14
sections of a maximum of 30 students per section. Attendance is not required in lecture

or recitation.

Student Sample

This study drew on a convenience sample of students enrolled in CEM 141 or
142, who volunteered to complete the survey. As an incentive to participate, 10 points of
extra credit in the course were given to the students who completed the survey (the
course grade is based on a grading scale of 1,000 possible points).

All the students enrolled in CEM 141 and CEM 142 during the fall semester of
2003 had the option to complete the survey. The students who participated in the survey
were not a random sample but a self-selected group. Because it was not a random sample,
there was no way of sampling a specific stratum (such as female engineering majors) to
reflect the target population. However, like a stratified random sample, pre-established

quotas of the various subgroups could be designated to better reflect the population of

58

students enrolled in general chemistry (such as sex, race, major, etc.). No special
sampling techniques were used, but final grade data was collected on the non-participants

to check for any non-response bias.

Data Collection

This was a cross-sectional study utilizing survey research and archival data. The
survey was developed primarily from the literature on college effect studies. To
determine the validity of items on the survey instrument, there were focus groups
conducted with students enrolled in CEM 141. An email was sent to all students during
the 10th week of the semester inviting them to participate in a focus group to discuss the
factors that influence academic success in general chemistry. The students who
responded were assigned a time and place to meet. They were provided with pizza and
soft drinks because the focus groups met in the early evening. The size of the focus
groups was limited to eight or ten students. This size group is an optimal size for quality
data collection in focus groups (Madhavi & Nelson, 2002)

The survey was comprised of a 91-item questionnaire utilizing a Likert-type scale
for student responses. The five possible responses were: Strongly Agree -Agree -
Neutral- Disagree Strongly Disagree. For data analysis, these interval data were coded as
Strongly Agree=l, Agree=2, Neutral=3, Disagree=4 and Strongly Disagree=5.

To minimize the length of the survey instrument, some of the data were collected
via a query fi'om the Student Information System (SIS) maintained by the Office of the
Registrar at Michigan State University. The SIS is a large database maintained by the

university with information on every student enrolled. The data included are extensive.

59

All of the information on a student’s initial application for admission, as well as all other
pertinent information on students while they are attending Michigan State University are
entered into the system. Through informed consent, overall high school grade point
averages, composite scores, and subject sub-scores on the American College Test (ACT),
scores on the Michigan State University Academic Orientation (AOP), mathematics
placement scores, and final grades in general chemistry were extracted from the SIS. In
addition, data on each student’s major, class-standing (first-year, second year, third year
or fourth year), honors college status, sex, ethnicity, and whether or not they repeated the
course, were gathered from the SIS. Reducing the length of the survey increases the
response rate (Rosenberg & Daly, 1993). Moreover, collecting archival data from
university records eliminates the need for tabulating, coding, and entering data collected
on surveys, thus increasing the validity of the data collected.

The objective portion of the survey instrument was comprised of 91 questions.
The data from the 91 questions on the survey and other data college from the SIS were
reduced to 24 subsets by employing a factor analysis.

A numerical coefficient of reliability of the survey instrument was calculated
using Cronbach’s alpha. Reliability or the internal consistency of an item measures the
stability of the measuring device over time. Internal consistency is especially important
for variables developed fiom summated scales because it is desirable to know whether
identical items if re-constructed, will elicit the same responses if re-administered to the
same respondent sample and to use insure they are measuring the same underlying

construct. (Santos, Lippke, & Pope, 1998). In addition, summated scales, based on a

60

greater number of questions, have higher reliability than single question variables
(Mitchell & Jolley, 1988).

The 29 scales in the College Impact Model are a series of interrelated items
designed to measure the constructs of student pro-college characteristics, college
experiences in the classroom, college experiences out of the classroom, instructional
resources, student-focused variables and instructor-focused variables.

The alpha’s for each scale is included in Appendix A with the factor loadings.
Some of the alphas were very low although the factors loaded on each other in the factor
analysis. For example, variable entitled “time commitments” was as a summated scale
consisting of the average hours per week of work for pay and average hours per week of
volunteer or non-paid work. These two items seemed to hang together in the factor
analysis but the reliability was .084. The same was true for “total years of math and
chemistry”. The reliability coefficient (as calculated with Cronbach’s alpha) of this item
was .124. This low alpha may be because of the high correlation between the number of
years of math and chemistry in high school and the low variance between the items.
Pilot of Survey Instrument

A pilot study of the GCS was conducted during the Spring semester 2003. The
final data collection was completed in late fall semester 2003. The data collection for this
study was originally intended for both semesters of the 2003 academic year (spring 2003
and fall 2003). Unfortunately, there was an undetected error in the Likert scale printed on
the survey instrument. The “Disagree” and Strongly Disagree” responses were reversed

on the survey.

61

The choices were listed as:
Strongly Agree -- Agree -- Neutral -- Strongly Disagree -- Disagree
1 2 3 4 5

instead of

Strongly Agree -- Agree -- Neutral -- Disagre -- Strongly Disagree
1 2 3 4 5

The student participants quickly pointed out this error. It was corrected and the
instrument reprinted. Unfortunately, there was no way to know whether the students
answered the questions according to the headings on the Likert scale, or if they ignored
the text and marked option five as “Strongly Disagree.”

One alternative was to recode the data into a 3-response Likert scale with
responses 1 and 2 representing “Agree,” response 3 representing “Neutral” and collapsing
responses 4 and 5 into “Disagree.” This idea was abandoned since the data collection
was planned for the fall semester and the enrollment was typically sufficient to get and
adequate sample of the CEM 141 and CEM 142 population without the spring semester
data. With the data unusable, this first wave of data collection was used a pilot for the
newly developed GCS.

There were also some questions that needed re-wording. For example, question #
44 originally asked, “I bought the book of old exams from the general chemistry office.”
The response choices were, Strongly Agree — Agree - Neutral - Strongly Disagree -
Disagree,” which did not make sense. The question was amended for the final version of
the survey to read, “The book of previous exams from the general chemistry office was
helpful for studying for exams.”

Other than the errors on the instrument, the pilot of the survey went very well.

The response was better than expected and the students answered the questions

62

thoughtfully. There was some concern that students would mark every question on the
survey with the same answer or leave most of them blank so they could receive the extra
credit, because the informed consent said they could resign from the study at any time
and still receive the 10 points. From a visual inspection of the surveys returned, there
were only five or six with “Neutral” or another single response darkened in for every
question on the survey instrument.

The pilot study was very helpful because it confirmed that five days was
sufficient time to allow for the students to complete and return the surveys. Furthermore,
the surveys were machine scored at the campus scoring office, but never tested in large
numbers on the campus scoring machines. Because there were almost 2000 surveys
scored, printing errors or any other problems that precluded the scanning of the surveys
would have required manual data entry at considerable cost in time and money.

The pilot of the survey instrument was unintended but valuable nonetheless. It
was costly in terms of time, money and unusable data but provided some reassurance that
the actual study would occur without procedural errors. However, it did serve its purpose

of detecting potentially devastating errors on the survey instrument.

Data Collection Procedures

With the approval of the University Committee on Research Involving Human
Subjects (UCRIHS), the 91-item survey instrument was given to the students enrolled in
CEM 141 and CEM 142 during the 13th week of the fall semester of 2003.

An email message was sent to all students enrolled in CEM 141 and CEM 142

during the 13‘h week of the 15-week semester informing them of the opportunity to

63

participate in a study on academic success in college general chemistry. They were asked
to complete a survey that would be distributed in lecture. As an incentive for
participation, they would be able to earn ten points of extra credit upon completion of the
survey and informed consent form. The students who missed lecture on the day the
surveys were distributed and wanted to participate were instructed to go to the General
Chemistry office to get a copy of the survey.

Each student who participated in the study was given a survey form and informed
consent. There was a box in the General Chemistry office where students returned the
completed surveys and consent forms. The survey was distributed on a Monday in all
five lectures and the deadline for return was on Thursday of the same week at 5:00 pm.
No late surveys were accepted.

The students marked their responses to the questions on the survey with a #2
pencil. The completed surveys were scored at the Michigan State University scoring
office. The students were very familiar with Optical Character Reader (OCR) type
answer sheets and there were few errors on the surveys.

The only major problem when administering the survey was that several students
failed to darken their Personal Identification Number (PID) on the top of the page or
entered an incorrect PID even though they were reminded numerous times that the
information was essential. This created a problem, because their PID was the only way
to identify the student and credit the 10 extra credit points toward their grade. Moreover,
the PID was the only way to match the students who completed the survey to their final
grade in the course, which was the outcome variable for this study. Fortunately, there

were fewer than 50 students with erroneous, incomplete or incorrect PIDs. Some of the

surveys with incomplete PID’s were salvaged because the student’s were identified if the
PID was darkened in incorrectly on the grid for scanning but written correctly on top of
the survey form. The instrument was printed double-sided.

The students were reminded to complete all 91 questions because it was easy to
overlook the questions on the back. There were approximately 25 students, however,
who did not complete the back half of the survey instrument, which is a small percentage
of the 1,891 that completed the entire survey

The survey was given to students in CEM 141 and CEM 142 during the fall
semester 2003. There were 2,252 students enrolled in CEM 141 and CEM 142 during the
semester. The response rate to the survey was 83.9 %. This was greater than expected,
which was probably due to the extra credit given for participating, although the ten extra
credit points were only 1% of the total points possible (1,000 total possible points) for the
semester. Although this was a self-selected sample, an 83.9 % response rate was
representative of the population studied.

The timing of the survey during the semester worked out very well. The 13th
week of the 15-week semester was ideal. The students appeared to have enough time in
the class to answer the questions accurately, yet it was not so close to the end of the
semester that they were busy studying for final exams. It also was close enough to the
end of the semester that the prospect of extra credit was very attractive, which may have
increased the response rate.

The students were given four days to complete the survey. The survey was
distributed in lecture on Monday and due by the following Thursday at 5 pm. There

were copies of the survey available in the General Chemistry office for students who

65

missed lecture. The students also were instructed to take and Sign an informed consent
form (and keep a copy if they desired). No late surveys were accepted. There were boxes
outside the lecture hall and in front of the General Chemistry Office where students could
deposit the completed surveys and the signed informed consent forms.

The completed surveys were taken to the university scoring office on a Friday and
the data returned the following Monday. The students did not wait long for their extra
credit. The 10 points extra credit showed up on each student’s total scores on Monday
afternoon on the CEM-SCORES Web site, which is a web-based grade book for students
to check their grades. The only complaints registered by students were from those who
were not allowed the extra credit points because they did not turn in their survey by the

deadline.

Data Analysis
The student learning model has six components that influence the outcome

variable, which is academic success in college general chemistry.

0 Pre-college characteristics and experiences.
0 College experiences in the classroom

0 College experiences out of the classroom

- Student-focused variables

0 Instructor-focused variables

0 Instructional resources

The research questions defined for this study all focused on positive educational

outcomes in general chemistry in relation to a variety of student experiences and

66

background characteristics. All data were analyzed using SPSS Graduate Pack 11.5 for
Windows. Initial analysis included descriptive statistics on the dataset with an overview
of each item in the survey instrument. Moreover, descriptive statistics provided an
overview of the dataset for identifying any obvious errors or omissions.

Multiple regression analysis looks at the joint relationship between a dependent
variable and two or more independent variables (Bohmstedt & Knoke 1994). The
dependent variable in this study was the final grade in CEM 141 and CEM 142. The
independent variables included the demographic information and the 28 summated scales
that made up the constructs of pre-college characteristics, student experiences in the
classroom, student experiences out of the classroom, instructional resources, student

focus variables and instructor focus variables.

Analysis Used for Research Questions
Research Question I . What is the description of the sample and independent variables?
Descriptive Statistics were used to describe the sample for this study. In addition,

the responses to the survey and archival data were summarized.

Research Question 2. What is the relationship between key student demographics and

control variables on the final grade in general chemistry?

0 sex

0 ethnicity

0 participants and non-participants
o first-generation students

0 physics in high school

67

0 class (first-year students)

0 honors student status

0 major county of residence

0 age of student at the time of study
0 repeat status and final grade

0 cohort group (CEM 141 or CEM 142)
Independent sample t-tests and ANOVA were used to compare the group means

of data in relation to the final grade in general chemistry.

Research Question 3. What is the relationship between the blocks of variables related to
pre-college characteristics, in-class experiences, out-of-class experiences, instructional
resources, student-focused variables, instructor-focused variables, and the final grade in
general chemistry? (r-square change by block).

Standard multiple regression analysis was used to determine the relationship
between the blocks of variables labeled in-class experiences, out-of-class experiences,
instructional resources, student-focused variables, instructor-focused variables pre-
college characteristics and final grade in general chemistry. Particular attention was paid
to the unique R2 accounted for by each category of variable in predicting the final grade

in general chemistry.

68

Research Question 4. What is the relationship between individual indicators within pre-
college characteristics, in-class experiences, out-of-class experiences, instructional
resources, student-focused variables, instructor—focused variables, and the final grade in
general chemistry? (The significance of individual predictors by block)

Standard multiple regression was used to determine the relationship of the 29
independent variables contained in the conceptual model on the final grade in general

chemistry.

69

CHAPTER 4

RESULTS

Introduction

This chapter presents the statistical analysis of the model of performance in the
general chemistry courses of CEM141 and CEM 142 at Michigan State University.
There are discussions regarding: data collection procedures; analysis of the data
collected, including descriptive statistics of the data set; details of the factor analysis;
reliability of the derived summated scales and a description of data transformations '
necessary for coding of raw data; and results of the regression analysis testing the
components of the conceptual model. In addition, eleven key demographic and control
variables were analyzed.

This is a cross-sectional study using survey research and archival data. The
conceptual model is a variation of the college impact model by Pascarella and Terenzini
(1991). All data collected were from students enrolled in general chemistry at Michigan
State University during the fall semester of 2003.

The students who volunteered to participate completed a 91-item survey. They
also completed an informed consent that allowed access to their final grade in the course
and demographic data from the Student Information System (SIS) at Michigan State
University. In addition, there were focus groups assembled early in spring 2003 to
validate the items on the survey.

A pilot study was completed in the spring semester 2003 to test the final version

of the survey. The pilot study quickly detected a small number of errors and omissions

70

on the survey instrument. The instrument was corrected and a new version was printed

for the data collected during the fall 2003 semester.

Analysis of Collected Data
The general chemistry survey included 91 questions that correspond to the six

components of the conceptual model for this study. These include:

0 Pre-college characteristics

0 In-classroom experiences

0 Out-of-classroom experiences
0 Instructional resources

0 Student-focused variables

0 Instructor-focused variables

The respondents answered the questions by darkening a circle next to each
question on an OCR answer sheet. The Michigan State University scoring office scored
the answer sheets. The raw data were loaded into SPSS, Version 11.5. The descriptive

statistics for each question are included in table 4.1

Research Question I: What is the description of the sample and independent variables?
The sample was 55.2 % female (n=1,243), 44.8 % male (n=l,009) and
predominantly Caucasian (n=1,685, 74.8 %). There were 144 Black students in the study
(7 %) and 140 Asian/Pacific islanders (6.9 %). 30 students listed their ethnicity as
Hispanic (1.3 %) and 15 as Chicano (.7 %). There were 29 students (1.3 %) listed under

the “Not requested” category. These are usually international students that do not report

71

ethnicity on their application for admission. There were no background data on 182
students (8.1%). Two thirds of the students in this study were first-year students
(n=1440). Less than a quarter of the students were first-generation college students
(n=440).

Pre—college characteristics. The majority of students had algebra II in high
school and earned an “A”. Most also had physics and chemistry and earned an “A” or
“B”. The students parents were well educated. Over two thirds of the students’ fathers
had at least a bachelor’s degree. About 60 % of mothers had a bachelor’s degree or
higher. About half of the students reported their parents combined income was between
$80,000 and $120,000. About 15 % earned greater than $150,000.

In-classroom experiences. Approximately 90% of the students responded that the
course was well organized and the classroom conditions were good (they could see and
hear the instructor). They were generally neutral on whether they enjoyed chemistry but
about 60% indicated that they would not take chemistry if it were not required for their
major. About twice as many students were satisfied overall with the course compared to
those who were not satisfied. This finding is consistent with other studies that found
students were only moderately satisfied with their intellectual experience in college
(F eldman & Newcomb, 1969).

The students’ responses basically responded neutral on the GCS in regards to their
performance in the course. However, they were not happy about their grades on exams
and approximately 60% reported that the scores on the exams were not what they
expected. This finding may reflect that most of the students in this course did well in

high school science and were accustomed to getting good grades. The mean final grade

72

in the course was 2.56 (SD=1.13). This result is consistent with other studies, such as
Deal (1984), who found that “Most students enter college with high grades and high
expectations, but many have their hopes dashed by poor performance.” (p. 154). In his
study the average grade in the freshman chemistry course at the University of California
at Riverside was 2.52.

College experiences outside of the classroom. Seventeen questions on the survey
assessed college experiences outside the classroom. Overall, 90% of the students thought
they made the right decision to come to MSU and only about 7% regretted not attending a
college closer to home. They essentially responded neutral on the GCS about getting
homesick at times and only about 20% thought often about leaving MSU.

Students participating in the survey were a social group; two thirds of the
respondents went to parties and on dates. Less than half belonged to a club or
organization and only about a third attended religious services. The majority (over 64%)
of the respondents drank alcohol even though most of them were minors at the time of the
survey.

Most of the students lived in residence halls on campus. Only about 20% lived
off campus. Very few students lived with their parents (less than 4%) and about 9 %
were commuter students.

When asked whether or not they had made new friends since coming to MSU,
70% said they had made new friends when they came to MSU. About 30% knew their
fi'iends before coming to MSU. About one third of the students said they were friends
with their roommate prior to coming to college and about 36% indicated they knew their

roommate prior to living together on campus.

73

Instructional resources. Purchasing the textbook for the course was optional but
highly recommended. The same was true for the study guide that accompanies the
textbook. The students were neutral on the value of the textbook for the course. More
than 50%, however, perceived the study guide that accompanied the textbook to be
helpful. This may be because one of the instructors for the course was the author of the
study guide and ofien referred to it in lecture (he was not the author of the textbook,
however).

When students sought help with their chemistry, they were inclined to get it from
their fiiends or other students and not from the chemistry help room or private tutor.
Less than 20% of students reported using the help room or other resources such as a
private tutor, but over two thirds sought help fiom friends or other students.

Student-focused variables. The questions on the GCS related to student focus
indicated that over 70 % of students always attended lectures but only about 30%
attended every recitation. An overwhelming majority (over 90%) of students did their
Leaning On-line-Computerized Personal Approach (LON-CAPA) homework.
Attendance at lectures and recitations for CEM 141 and CEM 142 is not monitored or
required. However, LON-CAPA homework is due weekly and constitutes about 25% of
their total grade for the course.

Regarding study habits, about half of the students did practice problems other
than the assigned LON-CAPA homework, but reported studying chemistry only about
two to four hours per week. Approximately 60% of the students studied on weekends.

Many of the students participating in the survey worked at a job or some type of

non-paid work. If a student had a job, the majority (about 60%) worked one to five hours

74

per week for pay and spent approximately one to two hours per week in volunteer or non-
paid work. This result may not be an accurate representation of students’ work
commitments, however. The questions in the survey related to work read: “Average
number of hours per week you work for pay” and “Average number of hours per week of
volunteer or non-paid work.” The first option was “(1-5)” and did not offer a choice for
zero hours worked. Over 25% of the students did not respond to either question so it was
not known whether or not the students that did not work simply skipped the question or
chose the first option (which was 1-5 hours per week) since they did not have the option
to choose zero hours worked.

Instructor—focused variables. Approximately 90% of the students agreed that
their instructor was prepared for lecture, making few errors and speaking intelligibly.
Teaching assistants did not fare as well. Only about 50% said they understood what was
presented by their teaching assistant and that their teaching assistant was doing a good
job in recitation. The students also perceived that their instructors were interested in their
learning (approximately 50%) and asked/answered questions in lecture (70%). Over 90%
perceived that their instructor liked teaching chemistry and most liked their instructor
(almost 80%). Less than half of the students, however, agreed that the instructor had

adequate office hours or was available during office hours.

Research Question 2. What is the relationship between key student demographics and
control variables on the final grade in general chemistry?
Gender and final grade in general chemistry. The mean final grade for all

students in the sample was 2.562 (SD=1.132). The mean final grade for male students in

75

the course was 2.604 (SD=1.154). Females earned slightly lower grades (M=2.527,
SD=1.112). Independent sample t-tests indicted that there was no difference between
male and female students in the sample regarding the final grade in the course (t=1.588,
p= .112).

This finding of no difference in the mean final grades between male and female
students is consistent with previous studies (Fletcher, 1977, House, 1995, Sanchez &
Betkouski, 1986). The ratio of male to female students in this sample (55.2 % female to
44.8 % male) was consistent for previous years of general chemistry courses at Michigan
State University.

Table 2.

Ratio of Male to Female Students enrolled in general chemistry at Michigan State
University

 

2001 2002 2003 2004

 

Male 49.9 47.2 44.8 47.5

Female 50.1 52.8 55.2 52.5

 

The sample for this study is also consistent with national data for the ratio of
fewer male students compared to female students attending college (N CES, 2003).
Table 3.

Ratio of Male to Female Students Attending College

 

2001 2002* 2003* 2004*

 

Male 43.9 43.5 43.3 43.2

Female 56.1 56.5 56.7 56.7

 

*proj ected

76

Ethnicity and final grade in general chemistry. Student ethnicity was downloaded
from the SIS at MSU. Ethnicity was categorized into nine groups as follows: Caucasian,
Black, Chicano, Hispanic, American Indian, Asian/Pacific Islander, other, not reported,
not requested. The “not requested” category was usually comprised of international
students. Ethnicity on the SIS is taken from the self-reported information on applications
for admission to MSU and the international students are not required to report ethnicity.

Independent sample t-tests compared the mean final grades of Caucasian students
and all other students in the sample. The mean grades of Caucasian students (M=2.639)
were greater than non-Caucasian students (M=2.330) (tr—5.658, p<.001).

As stated previously, the sample for this study was 74.8% Caucasian (n=1685).
All other students totaled 25.2% (n=569). There were 144 Black students (6.7 %) and 140
Asian/Pacific Islander students (6.2 %) in the sample. There were missing data on 8.2%
of the sample (n=l 82).

A one way ANOVA was performed on the nine groups of ethnicity. This analysis
indicated a significant difference across all seven ethnic groups (F=13.92, df=7/2057,
p<.001).

A post-hoe test (Tukey’s HSB) of the mean final grade in general chemistry and
the nine ethnic groups revealed that Caucasian students (n=1682, M=2.64) did
significantly better than Black students (n=l43, M=l .82) and Chicano students (n=15,
M=1.40), but not Hispanic (n=29, M=2.38), American Indian (n=12, M=2.13) or
Asian/Pacific Islander (n=140, M=2.58).

The mean final grade of the group of students labeled “not reported” (n=15,

M=3.0) or “not requested” (n=29, M=2.97) was significantly higher than the Caucasian

77

students. These groups were primarily international students who do not normally include
their ethnicity on their application for admission to the University.

Response bias between participants and non-participants. The study participants
were students enrolled in CEM 141 and CEM 142 during the fall semester of 2003. For
analysis purposes to investigate for any response bias in the sample, the data were re-
coded into four groups: CEM 141 participants (n=1645), CEM 142 participants (n=246),
CEM 141 non-participants (N=321) and CEM 142 non- participants (N=40).

There were no survey data collected on non-participant groups. However, the
same archival data were collected on the participant and non-participant groups. These
data included the final grade in the course and all other information available from the
SIS.

The results from CEM 141 and CEM 142 were analyzed separately to determine
differences in final grades between participants and non—participants. Independent sample
t-tests indicted that the CEM 141 participant group earned a significantly higher mean
final grade (M=2.7l) versus the non-participant group (M=1.69) (t=15.8, p<001). The
same was true with the mean final grades for CEM 142 participants (M=2.84) and non-
participants (M=l.54) (tr-7.1, p< .001).

CEM 141 and CEM 142 were analyzed separately to determine differences in
final grades of the participant groups. Independent sample t-tests were used to compare
the mean final grades of students enrolled in the CEM 141 and CEM 142. The mean final
grade for the CEM 141 was 2.71 (n=1643) and the CEM 142 was 2.84 (n=244). The

significance level for the testing the means was set at .05. The difference between the

78

participants in CEM 141 and CEM 142, was almost, but not quite, significant (t= -l .89,
p=.059).

F irst-generation college students and final grade in general chemistry.
Independent sample t-tests showed that there was a significant difference in the mean
final grade for the students who reported they were the first in their family to attend
college (M=2.597, n=440) when compared to non-first-generation college students (M=
2.778, n=1430 (t= -3.333, p=.001).

This finding is consistent with the literature on first-generation college students
(Reihl, 1994; Ting 1998).

Physics in high school and final grade. The 1,284 students who reported taking
physics in high school were more successful in college general chemistry (M=2.84) than
the 590 students who did not take it (M=2.51) (t=6.56, p<.001). This finding is
consistent with Fletcher (1978), who found that a strong background in high school math
and science, especially physics, translated into academic success in college chemistry.
However, Yager (1989) determined that for high-achieving students, high school physics
had little effect on college chemistry.

Class standing (first-year students) and final grade. It is common for first-year
students to take general chemistry because it is a prerequisite for upper-level chemistry
and biology courses. This sample included approximately 66% first-year students.
Independent sample t-tests were used to compare first-year students with sophomores,
juniors, and seniors.

The mean final grade for first-year students was 2.62. The grades for first-year

students were significantly better than non-first-year students (M=2.43) (tr-3.64, p<.001).

79

F irst-year student final grades (M=2.62) were significantly better than
sophomores (M= 2.43) (t=3.295, p=.001) and juniors (M=2.42 ) (t=24, p<.05). There
was no significant difference between first-year students (M=2.62) and seniors (M=2.6)
(t=.433, p=.665). There were seven graduate or lifelong education students in the course
but their grades were not included in this analysis.

Honors college status and final grade in general chemistry. There were 118
students in the sample who were in the Honors College. The mean final grade of Honor’s
College students was 3.45. This was significantly higher than the 2.51 for all of the other
students in the sample (t=8.94, p<.001).

Major area of stuay and final grade in general chemistry. There were 112
majors listed on the SIS for the 2,252 students in the sample. The data were re-coded
into the following 13 groups: human biology (n=92), education/kinesiology (n=90),
Lyman Briggs College (n=l 75), mathematics (n=55), engineering (n=142), agriculture
and natural resources (n=75), chemistry/biochemistry (n=35), no preference (n=196), pre-
professional (n=413), psychology (n=63), natural science-no preference (n=22), pre-
nursing (n=l48) and other (n=746). A one way ANOVA indicated that there was a
significant difference between major and final grade (F =1 .835, df=12/2234, p<.05).

A post-hoe test (Tukey’s HSB) of the mean final grade in general chemistry and
the 13 majors revealed only one significant difference: the College of Engineering
students had the highest mean final grade (M=2.82) and the students in the College of
Agriculture and Natural Resources had the lowest (M=2.18).

This finding partially conflicts with Paul (197 8) who found that the highest grades

in chemistry courses were those students majoring in premedical technology. However,

80

in his study, the final grades of engineering students were greater in comparison to
students majoring in Agriculture and Forestry.

County of residence (if in Michigan) and final grade. Michigan State University
is a land grant university and the majority of students in the sample are from Michigan
(91.2%). The students in the sample represented 74 of the 83 counties in Michigan. The
majority of the students (54.6%) were from six counties. Thirty-nine percent of the total,
however, were from the tri-county area of southeast Michigan (Wayne, Oakland and
Macomb counties).

The counties of residence data were re-coded into a single variable with seven
categories that reflected the population of Oakland (n= 432), Wayne (n= 268) Macomb
(178), Ingham (n= 142), Kent (n= 117), Genesee (n= 93) and Other (n= 1022).

The students from Macomb County earned the highest mean final grade in the
course (M=2.73). The next highest grades were achieved by students from Oakland
county (M=2.65). Students from Wayne county earned the lowest mean final grades in
the course (M=2.39). A one way ANOVA indicated significant differences in final grades
across the categories of (F = 2.291, df = 6/2240, p<.05).

A post-hoe test (Tukey’s HSB) indicated a significant difference in the mean
final grades for students from Macomb county and Wayne county (Mean difference
=.338, p<.05) and from Oakland county and Wayne county (Mean difference = 2.26,
p<.05). However, there was no difference in the mean final grades of students from

Michigan versus non-Michigan (tr—.244, p=.807).

81

Age of student at time of study and final grade in general chemistry. A one-way
ANOVA indicated significant differences across the age groups of the students
participating in the study (F=3.640, dfi2/2226, p<.001).

The range of students in the sample was 18 to 47 years old. Ninety-eight and one
half percent of the students were 18 to 23 years old. The data were re-coded into seven
age groups: 18,19,20,21,22,23, and “all other.” The 18-year-old group’s (n=971) final
grades (M=2.67) were significantly better than the 20-year-olds (n=293, M=2.38), 21-
year-olds (n= 137, M=2.26), 22-year-olds (n=40, M=2.11) and the “all other” (n=50,
M=2.17) age groups.

Again, this finding conflicts with that of Boughan’s (1996), who found that age
was a strong predictor for academic success in community college chemistry, with older
students earning higher final grades. This may be because students in community college
differ substantially from sample used for this study and therefore may be an irrelevant
comparison. Boughan, however, was one of the few studies reviewed that included age
as an independent variable.

Repeat status and final grade. Independent sample t-tests indicated that there was
a significant difference in the mean final grades of student repeating the course and those
taking it for the first time (t= -7.795, p< 001). The 112 students repeating the course
earned a lower final grade (M=1.862) than those taking it for the first time (n=2071,
M=2.669).

This finding is consistent with Burdge and Daubenmire (1991), who found that

repeating a course was detrimental to academic success. They found that students, who

82

repeated the course more than one time, earned lower grades than previous attempts and
no student who repeated more than three times passed the course.

Financial aid and final grade in general chemistry. There were 1205 students in
the sample who reported receiving some type of financial aid. Six hundred and fifty-nine
students indicated that they did not receive any financial aid. Independent sample t-test
confirmed that there was no difference in the final grades of students who received

financial aid (M=2.75) and those who did not (M=2.67) (t=1.1, p=.269)

Research Question 3. What is the relationship between the blocks of variables related to
pre-college characteristics, in-class experiences, out-of-class experiences, instructional
resources, student-focused variables, instructor-focused variables, and the final grade in
general chemistry? (r-square change by block). ?

The data for analysis were tabulated from student surveys and archival
information from students enrolled in general chemistry. The interval data fi'om the
survey were Likert-type responses to 78 questions related to student experiences in and
out of the classroom. In addition to the 78 objective questions, there were 13 additional
questions on the survey associated with parental background and socio-economic status
for a total of 91 questions on the survey. The archival data were primarily demographic
information, high school data and standardized test scores gathered from the SIS at
Michigan State University.

The 91 survey items corresponded to the variables included in the six blocks in

the conceptual model. Factor analysis was used to reduce and replace the larger set of

83

variables by a smaller set which best summarized the data set and detected structure in
the relationships between variables.

The data from the responses for the General Chemistry survey and archival data
from the SIS were reduced to 30 summated scales corresponding to six blocks of data and
using principal component analysis with oblique rotation and minimum Eigenvalue of
one (Eigenvalue is the measure of variance explained by a factor).

The six blocks included: pre-college characteristics, in-class experiences, out-of-
class experiences, instructional resources, student-focused variables and instructor-
focused variables (Appendix A). The 30 scales contained in the six blocks were
transformed in SPSS into new variables using the “recode” function. The 30 new scales
were the arithmetic mean of the groups of variables derived from the factor analysis. A
description of each factor and the variables included within each factor is explained in
Appendix A. The variables that did not load and were omitted are listed below.

During the preliminary regression analysis, the coefficients for many of the
variables were negative. This result was because of the way the survey instrument was
designed and scored. The response choices to the questions on the survey instrument
were “Strongly agree,” “Agree,” “Neutral,” “Disagree,” and “Strongly Disagree.” They
were scored as: “Strongly agree” =1, “Agree”=2, “Neutral”=3, “Disagree”=4, and
Strongly disagree”=5. The negative regression coefficients were confusing when
looking at the model. To put all scales in the same direction, all the variables were
multiplied by a negative one. The recoded variables in the data set were all negative.

In-class experiences. The variables in the block under “in-class experiences”

loaded onto five factors. Three variables were discarded. The variable, “The lecture room

84

 

is comfortable,” did not load on any of the other factors. It was not correlated positively
or negatively to any of the other variables and was discarded. Two other variables
related to lecture size were not used. The responses to, “I like the large lecture format for
chemistry” and, “I prefer a small classroom environment for chemistry,” were highly
negatively correlated (-.751). They essentially measured the same response, did not load
onto any of the other factors, and were discarded.

Pre-college characteristics. The variables in the pre—college characteristics block
loaded onto three factors. There were no variables omitted.

Out-of-class experiences. The variables related to out-of-class experiences loaded
onto six factors. The variable, “I get along with my roommate,” did not load onto any of
the variables related to out-of-classroom experiences and were discarded.

Instructional resources. Four factors emerged from the variables related to
instructional resources. One variable, “The book of old exams is helpful for studying for
exams” did not load onto any of the factors and was omitted as a variable in for the study.

Student-focused variables. The student-focused variables loaded onto three
factors. One variable “I try to explain material to another student or fiiend” did not load
onto any of the other variables. It was not highly correlated positively or negatively with
any of the other variables and was discarded.

Another variable related to use of the ALLMSU web site for LON-CAPA
homework, was also discarded. The ALLMSU web site is a non-MSU sponsored web
site that student use to get help with their LON-CAPA homework. It is essentially a chat
room that students use to, among other things, rate professors, buy and sell athletic

tickets, and discuss LON-CAPA for chemistry and physics. Although LON-CAPA

85

problems are randomized, students can post the algorithm to solve the problem on the
ALLMSU web site. The students simply “plug and chug” the numbers assigned for their
problems and get the answer

The other variable discarded was “I do CAPA alone (not with a group or with
other students)” This variable did not load onto any of the other factors related to study
habits and was not correlated, either positively or negatively, with the other variables.

Instructor-focused variables. Five factors emerged from the variables related to
instructor focus. There were no variables omitted.
Model Tests

Standard multiple regression analysis was performed to determine the relative
influence of the six blocks of independent variables and the dependent variable. The six
blocks were pre-college characteristics, in-class experiences, out-of-class experiences,
instructional resources, student-focused variables, and instructor-focused variables. The
dependant variable was the final grade in general chemistry. Analysis was performed
using SPSS regression (Appendix A)

The first block of variables was pre-college characteristics and experiences.
These variables included; grades in high school science and high school math, parents
education and income, years of math and chemistry in high school, ACT English,
reading, science and comprehensive scores, ACT math score and math scores from tests
at summer academic orientation (AOP). The variables in block one contributed to 19.8

percent of the total variance in the model (R2 =.198, p<.001).

86

Controlling for all other variables in the block there were two significant
variables. One was students’ scores on ACT math and AOP math tests (t=7.3 80, p<.001)
and the other was grades in high school science (t=7.048, P<.001).

The variables parents education and income (t=1 .370, p=.171), total years of math and
chemistry in high school (t=.177, p=.860) and ACT English, reading science and
comprehensive sub scores (t=1.910, p=.056) were not significant in the pre-college
characteristics and experiences block.

The second block of variables was in-class experiences. This block included
student perceptions of course organization, classroom conditions, course enjoyment and
course satisfaction. This block accounted for largest portion (20 percent) of total
variance in the model (R2 = .200, p<.001).

There were two significant variables in this block, course enjoyment (t=-2.341, p<.01) and
satisfaction with the course (t=19.45, p<.001).

The third block of variables entered was out-of-class experiences. These
variables included issues such as campus life adjustment, making new friends, whether
they knew or were prior friends with their roommate, socializing, campus life connection,
whether they lived on a “party” floor or not. This block accounted for only .2 percent of
the total variance in the model (R2 = .002, p>.01). Controlling for all other variables in
the block all were significant except the one that asked if they knew their roommate or
they were fiiends with their roommate prior (t=1.345, p=.179). The other variables in the
block, campus life adjustment (t=-5. 138, p<.001), friendships (t=4.067, p<.001),
socializing (t= -4.109, p<.001), campus life connection (t= -3.521, p<.001) and live on a

quiet floor or party house (t= 3.111, p< .01) were all significant.

87

The fourth block of data were variables related to instructional resources. These
variables included; usefulness of instructional materials, use of the help room, sources of
help and instructor contact. Instructional resources contributed to 1.2% of the total
variance in the model (R2 = .012, p=.054). All four of the variables in this block were
significant. Instructional resources (t=—7.82, p<.001), use for the help room (tr-2.581,
p=.01), sources of help (t= -9.572, p<.001) and instructor contact (t= -2.436, p<.05).

The fifth block consisted of student-focused variables. These variables were
students’ answers to questions related to course commitment, study habits, time
commitments and total credits enrolled for the semester. Student focus variables
contributed to 6.9 percent of the total variance in the model (R2 = .069, P<.001). All of
the variables in this block are significant. These include, course commitment (t=8.491,
p<.001), study habits (t=2.346, p<.01), time commitments (t=4.168, p<.001) and total
credits enrolled for the semester (t= -5.586, p<.001).

The sixth block of variables was instructor focus variables. These variables
included students perceptions of instructor basic skills, instructor more advanced skills,
instructor attitude, instructor availability, and teaching assistants (TA). Instructor focus
variable contributed 1.3 percent of the total variance in the model (R2 = .013, p<.05).
Only instructor attitude (t=2.014, P<.05) and teaching assistant (TA) performance (t=
3.540, p<.001) were significantly related to final course grade.

As a predictor of final grade in general chemistry all six of the blocks of
variables the regression model were significant except out-of-class experiences. The full

model accounted for 49.4 percent of the total variance. Two blocks of variables, however,

88

in-class experiences and pre-college characteristics, accounted for the largest portion of

the shared variance (39.8 percent).

Research Question 4: What is the relationship between individual indicators within pre-
college characteristics, in-class experiences, out-of-class experiences, instructional
resources, student-focused variables, instructor-focused variables, and the final grade in
general chemistry? (The significance of individual predictors by block)

Standard multiple regression analysis was performed between the dependent
variable (final grade in general chemistry) and the 28 independent variables. Analysis
was performed using SPSS regression.

Regression analysis indicated that the model significantly predicted the frnal
grade in general chemistry (F=15.626, df=28/448, p<.001). The R2 for the model was
.494 and the adjusted R2 was .462. Table 4.4 depicts the Y-intercept, unstandardized
regression coefficients (B) and standardized regression coefficients (beta) for each
variable.

Nine individual variables in the model significantly predicted the final grade in
general chemistry. The strongest predictor was course satisfaction ([3: .481, t=11.172,

p<.001). The next most significant variable was course commitment (B: .225, t=6.625,
p<.001). The third variable was ACT math sub scores and AOP math scores (13: .202,
t=5.081, p<.001). The fourth was students’ grades in high school science courses (0:
.169, t=4.57, p<.001). The fifth was instructor contact ([3= -.126, t=-3.44, p=.001). The

instructor contact coefficient was negative. The sixth significant variable was the variable

composed of ACT sub scores in English, reading, science and their comprehensive score

89

([3: .108, t=2.86, p<.01). The seventh significant variable was the total number of credits

enrolled for the semester ([3: -.085, t2 -2.46, p=.014). The regression coefficient for total

credits enrolled was also negative. The eighth significant variable was if the students used

the help room for help with their LON-CAPA homework or other problems ([3: .085,

t=2.36, p=.019). The last significant variable was study habits ([3: .077, t=2.11, p<.05).
Summary

This chapter looked at the analysis of the data collected from student responses on
the general chemistry survey and archival data collected from the SIS. All data were
collected from students enrolled in CEM 141 and CEM 142 at Michigan State University
during fall semester 2003.

There were some initial problems with the survey that were detected during the
pilot study and corrected for the final data collection. There were no problems with the
final administration of the survey and the response rate was 83.9%.

The first analysis was a description of the sample and independent variables in the
conceptual model. The ratio of men to women students in the sample reflected prior
years in general chemistry courses at Michigan State University, as well as national data.
The descriptive statistics of the independent variables did not show any outliers.

There were eleven key demographic and control variables in the dataset. There
was no difference in the final grades between male and female students. White students
earned higher mean final grades than did non-white students.

There was a response bias in the sample. The survey participants earned
significantly higher final grades in general chemistry than non-participants. However,

the non-response rate was so low (16.1%) that it was not a concern.

90

Less than a third of the students in the sample were first-generation college
students, but they did not do as well in the course as non first-generation college students.
The mean final grades in general chemistry for first-generation college students were
significantly lower than non first-generation students.

Taking a physics course in high school positively influenced students’ final grade
in college general chemistry. From the data analyzed in this study, students who took a
physics course in high school earned significantly higher final grades in general
chemistry than those who did not.

There was a wide range of student ages in the sample. Moreover, first-year
students through seniors were represented. Age and class were important factors for
academic success in college general cherrristry. F irst-year students earned higher final
course grades than sophomores and juniors. However, there was no difference in first-
year students and seniors. Age was also an important factor. The 18-year-old students
earned higher final grades than any other group.

There were students in the sample representing 74 of the 83 counties in Michigan.
The greatest number of students was from Wayne, Oakland and Macomb counties. Of
these three counties, students earned the highest final grades from Macomb County and
the lowest from Wayne County.

Other important control variables were that students who repeated the course
earned significantly lower final grades than those who took it for the first time. Honors
college students earned higher grades than non-honors college students. Students in the
College of Engineering earned the highest final grades of all majors. Students in the

College of Agriculture earned the lowest final grades.

91

The conceptual model for this study was a variation of the College Impact Model
(Pascarella & Terenzini, 1991, 2005). The model included six blocks of independent
variables, which were: pre-college characteristics, in-class experiences, out-of-class
experiences, instructional resources, student-focused variables and instructor-focused
variables. The dependent variable was the final grade in college general chemistry.

Standard multiple regression analysis was used to analyze the relationship of the
six blocks of independent variables to the final grade in college general chemistry. Two
blocks of independent variables, pre-college characteristics (R2 = .198) and in-class
experiences (R2 = .200) explained a large portion of the variance. Together, those two
variables contributed to 39.8% in shared variability with the final grade in general
chemistry. The remainder of the change in R2 was attributed to out-of-class experiences
(R2 = .002), instructional resources (R2 = .012), student-focused variables (R2 = .069) and
instructor-focused variables (R2 = .013).

There were 28 independent variables in the model. A standard multiple
regression analysis was performed with the dependent variable (final grade in college
general chemistry) and 28 independent variables. There were nine significant predictors
of final grade in college general chemistry. The strongest predictor was satisfaction with
the course. Students who were satisfied with the course earned higher final grades.
Other significant predictors were ACT math sub scores, AOP math score, ACT English,
reading, science and comprehensive scores.

Another significant variable was instructor contact. However, the regression

coefficient was negative so the students with high instructor contact earned lower final

92

grades in college general chemistry. Components of the instructor contact scale asked if

students attended the instructor office hours and if the instructor knew their name.

93

CHAPTER 5

DISCUSSION AND CONCLUSION

The origin of this investigation of the factors that influence academic success in
college general chemistry occurred when the researcher represented the Chemistry
Department at summer academic orientation. The primary purpose during orientation
was to meet with prospective chemistry majors or students who required general
chemistry for their major.

The majority of the students enrolling in general chemistry did not want to
become scientists. They were required to take general chemistry as a prerequisite for
admission to medical school or another professional school, such as teacher education,
forensic science, or chemical engineering. Thus, the majority of the students worked with
would go on to graduate or professional schools, or apply for admission to the College of
Education or College of Engineering in their sophomore year. Fewer were in the College
of Agriculture, College of Nursing, and College of Natural Science. Although their final
academic goals were different, they all had been placed into the minimum of College
Algebra or higher on the mathematics placement test giving at AOP and all had a strong
desire to succeed.

The researcher of this study was the Academic Program Coordinator for General
Chemistry, and was able to follow many of the students from summer academic
orientation through their first-year of General Chemistry. Some of the students did very
well and some struggled in the course. Developing an understanding of why some

succeeded and others did not became the motivation for pursuing this research.

94

A review of the literature revealed that college students bring disparate
experiences and backgrounds with them when they matriculate. Once at the institution,
they spend their hours both in and out of the classroom. The foundation of this study was
Pascarella’s (1996) assertion that the impact of college on students was the result of
cumulative, interrelated experience, rather than any single event. There were a number of
models created to assess and predict the impact of college on students (Astin, 1991; Pace,
1979; Tinto, 1993; Weideman, 1989). These models were adapted for this study to
investigate the factors that influence academic success in college general chemistry.

The conceptual model for this study was a college impact model. This model was
modified primarily from Pascarella’s (1996) General Causal Model for Assessing the
Effects of Differential Environments on Student Learning and Cognitive Development
(2005) but also from Astin’s I-E-O model (1991), Pace’s Student Development and
College Impress Model (1979), Tinto’s Theory of student departure (1993) and
Weidman’s Model of Undergraduate Socialization (1989).

All of the models were similar. Astin’s Input-Environment-Out (I-E-O) was
perhaps the simplest but the basic tenets of all the models were similar. Students come to
college with varied backgrounds and experiences and the academic and non-academic
environments of college influence a variety of outcomes for students during their time at
the institution. The various models assessed outcomes such as learning, retention, and
socialization after college. The student-learning model created for this study focused on
one simple outcome, which was academic success in college general chemistry as
measured by the final grade in CEM 141 and CEM 142 at Michigan State University

during fall semester 2003. This outcome variable was examined in relation to the six

95

constructs of the student learning model modified for this study. The six constructs were
student pre-college characteristics, in-class experience, out-of-class experiences,
instructor-focused variables, student-focused variables and instructional resources.

To operationalize the student-leaming model, a 91-question General Chemistry
Survey (GCS) was developed. The questions on the GCS were related to the six
constructs that made up the student learning model. Some of the questions on the GCS
were unique to general chemistry at MSU, but most were borrowed from the College
Student Experiences Questionnaire (CSEQ) (Pace & Kuh, 1998) and The Inventory of
College Activities (ICA) (Astin, 1968).

An analysis of the data collected provided the following conclusions:

1. For the key demographic variables, Caucasian students earned higher final
grades than non-Caucasian. First-generation college students earned lower
grades than non first-generation students did. Students who had physics in
high school earned higher final grades in college general chemistry. First-
year students earned higher final grades than sophomores and juniors, but
there was no significant difference between first-year students and seniors.
Students in the Honors College earned higher grades than all other students
in the study. Engineering students earned higher grades than other majors.
Students from suburban counties earned higher grades than urban counties.
Students who repeated the course earned lower final grades than students

taking the course for the first time.

96

When looking at the r-square change by the six blocks of variables used in
this study, the largest percentage of the variance contributing to the final
grade in general chemistry was student experiences in the classroom and
students pre-college characteristics. Instructor-focused variables, student-
focused variables, and instructional recourses were also significant, but to
a lesser extent. The out-of-class experiences block was not significant.
When analyzing the significance of individual predictors within each
block, the significant variables were: grades in high school science, total
years of math and chemistry, ACT English, reading, science and
comprehensive sub scores, scores on ACT math and AOP math, course
satisfaction, instructor contact, course commitment, using the help room,

study habits and total credits enrolled during the semester.

Conclusion I

The key demographic variables in this study were analyzed primarily by comparing

one group to another in relation to the final grade in general chemistry. Depending on the

type of variable, independent sample t-tests or one-way ANOVA were used to assess any

group differences. This analysis provided little insight into the factors that influence

academic success in college general chemistry. The key demographic variables were

important, however, for looking at the sample and characteristics of the students

participating in the study.

In the study design and conceptual model, the key demographic variables were

collected and analyzed as controls but were not included as independent variables in the

97

model. It turned out that not including the key demographic variables in the conceptual
model was a flaw in the study design. Upon final analysis, there were some differences in
the demographics of the sample, but they were not included as mediating variables and
there was no analysis of their interaction with the dependent variable. The difference in
the key demographic characteristics of the sample was interesting but did not serve to
fully explain their influence on academic success in college general chemistry in relation
to student experiences in and out of the classroom.

For example, Caucasian students earned overall higher final grades in general
chemistry than all other groups of students. This finding is consistent with the literature
on academic success in college (Hagedorn, et a1, 1999; Mason & Mittag, 2001; Flowers
& Pascarella, 2003). However, the model and analysis for this study did not control for
any other variables such as ACT scores, parental education/ income, or county of
residence to determine if the students were from urban, suburban, or rural high schools,
or to see if simply being Caucasian influenced the final course grade.

The first-generation college students did not achieve as well as non first-
generation students in relation to the final grade in general chemistry. This finding was
also consistent with the literature. Ting (1998) found that first-generation students and
those from disadvantaged backgrounds tended to be academically at risk in college.
Hagedom et a1 (1999) found that students with parents that completed some higher
education had higher levels of college math achievement. Others (Naumann, Bandalos &
Gutkin, 2003) found that ACT scores were a good predictor of academic achievement for

first-generation college students.

98

This study only looked at the group differences between first-generation college
students and non first-generation college students in relation to the final grade in general
chemistry. It would be interesting to look at first-generation college students controlling
for pre-college characteristics, such as ACT scores and parental income to see whether or
not first-generation students with high standardized test scores and advantaged
backgrounds do better than students with lower scores and lower parental
income/education.

The analysis of the key demographic data collected for this study would suggest
that the most successful students in general chemistry would be Caucasian and a first-
year or senior student. She or he (gender did not make a difference) would be an
engineering major in the Honors College taking the course for the first time, and coming
from a suburban high school.

Other than Honors College status, the above description would include a large
number of students in the course so there must be other factors that influence academic

success in college general chemistry.

Conclusion 2
This study used the block method in the regression analysis. There were six
blocks of variables that formed the college impact model used for this study. These six
blocks included pre-college factors, in-class experiences, out-of-class experiences,
student-focused variables, instructor-focused variables and instructional resources
variables. The questions on the GCS corresponded to each of the constructs. The

variables were entered into a factor analysis by construct. For example, there were 17

99

questions on the GCS related to the course instructor. The factor analysis identified a
pattern of relationships within this block and five variables were created. These variables
made up the block entitled “Instructor-focused variables.” The five new variables were
the arithmetic average of each cluster.

The data analysis for this study was two-fold. First, the blocks of variables in the
model were entered into the regression equation to determine the contribution of each in
relation to the final grade in general chemistry. Second, the individual variables were
entered into a regression equation to determine their influence on the outcome, which was
the final grade in general chemistry.

For discussion purposes, conclusion one and conclusion two will be discussed
together. The discussion too, of the blocks of variables and the individual variables
within each block will provide a clearer picture of their influence on academic success in
general chemistry. This format is evident when looking at the in-class experiences block,
which was the most significant block in the model. Upon closer examination, there is one
highly significant variable, course satisfaction, whereas the rest of the variables in the
block were not significant.

Looking at the blocks of data and the individual variables within each block helps
make better sense of the data. It also reduces the need to keep referring back to the
model to determine which variables were included in each block. Moreover, it illustrates
the weakness of the block method of analysis (which will be discussed later in this
chapter). Therefore, the discussion will include both the contribution of the blocks of
variables and the individual variables within each block in relation to the influence on the

dependent variable, which was the final grade in general chemistry.

100

In-C lass Experience Block and Variables

The in-class experience block and variables had the greatest significance of the
six constructs in the conceptual model used for this study. In-class experiences
contributed to 20% (r-squared = .200) of the variance predicting the final grade in general
chemistry. The in-class experience variables included: course organization, classroom
conditions, course enjoyment and course satisfaction. The only significant variable
within the in-class experiences construct was course satisfaction and it was quite large
(t=1 1.172). The course satisfaction variable was made up of three questions on the
GCS. These included whether or not students were satisfied with their performance in
the course, whether or not their grades on exams were about what they expected, and
their overall satisfaction of the course.

In this study, course satisfaction was an extremely strong predicator of academic
success in general chemistry. This finding suggests that students who were satisfied with
the course earned a higher final grade than those who were less satisfied. Unfortunately,
from these data, it is unknown whether or not students were satisfied with the course
because they earned a good grade, or whether or not they earned a good grade because
they were satisfied with the course. The influence of course satisfaction on the final
course grade was an important predictor of the final course grade in this study and

requires further investigation.

Pre-College Characteristics and Experiences - Block and Variables

The next most significant block was pro-college characteristics and experiences.

This block included standardized test scores, parental educational and income, and high

101

school coursework and performance. This block of variables contributed 19.8% (r-
squared = .198) of the variance to the final grade in general chemistry.

It can be hypothesized from this study that an individual’s innate intelligence and
high school performance will translate into academic success in college general
chemistry. Students who earned good grades in high school math and science will be
more successful in college general chemistry.

In this study, student grades in high school science influenced their final grade in
college general chemistry. The variable of the grade in high school science was
composed of self-reported grades in algebra II, chemistry and physics. This is consistent
with the literature (Astin, 199; Deal, 1984; Ozsogomonyan & Loftus, 1979) that high
school performance and standardized test scores were predictors of academic success in
college.

Fletcher (1977) reported that students with four years of math and a full science
sequence were more successful academically in college chemistry. This particular
research found that the number of years of high school math and chemistry was not a
significant predictor of academic success in college general chemistry. However, earning
good grades in high school math and chemistry was a significant predictor. Therefore,
earning good grades in high school math and science was more important than the
number of years of math and science for academic success in college general chemistry.
In addition, students who had a physics course in high school also were more successful
in college general chemistry.

The literature on the relationship between high school chemistry and college

chemistry was mixed. Some found a relationship between high school chemistry and

102

college chemistry (Fletcher, 1977; Willingham, 1985) and some did not (Ogden, 1976).
In this study there was a significant relationship between both earned grades and number
of years of high school chemistry and academic success in college chemistry. Students
who earned higher grades in high school chemistry and had more than one year of high
school chemistry earned higher grades in college general chemistry

Higher ACT English, reading, science and comprehensive sub-scores translated
into higher final grades in college general chemistry. This finding was also true for
higher scores in ACT math and AOP math. In general, high achieving students as
measured by standardized test scores, were more successful in college general chemistry.
This result was interesting because it was not sub-score specific. That is, higher sub-
scores in English and reading were as important as the sub-scores in science, math and
their comprehensive score. Perhaps this relationship was a function of the type of
questions on the general chemistry exams. Although all the exams were multiple choice,
all the questions were story problems and required reading. Students must be able to
read and understand the questions on the exam before they could set up the problem and
do the necessary calculations. Understanding the math required to complete the problems
was not enough to do well on the exams. Students were required to read quickly and
understand the information presented in order to pick out the information necessary to
answer the question correctly.

Interestingly, in this study parental education and income were not a significant
predictor of academic success in college general chemistry. This finding was somewhat
surprising because first-generation college students did not do as well as non-first-

generation college students.

103

There may be one or more reasons for this discrepancy. One is that the variable
that measured parental education/ income was a combination of the highest degree
attained by each parent and the annual combined parental gross income. A high score on
this variable could be from a high income, from highly educated parents or both. Parents
that earned high incomes may not have had college degrees. An additional reason for the
discrepancy may be a result of the data analysis because the regression equation did not
control for first-generation college students.

The strong influence of pre-college characteristic has both positive and negative
implications for educational practitioners. One negative implication is that all the pre-
college characteristics and experiences occur prior to matriculation so colleges and
universities have little influence on these factors once the student arrives at the campus
gates. Still, understanding pre-college characteristics and experiences is valuable for
higher education professionals. Deal (1984) concluded that predicting course grades was
a usefirl tool for counselors and admissions professionals. Students at risk can be
identified early for remediation. Moreover, understanding the relationship between pre-
college characteristics may motivate students to work harder to achieve the grade they
desire if they understand that their preparation for college is not as strong as many of

their classmates.

Student-Focused Block and Student—Focused Variables
The third significant block in the conceptual model was related to student-focused
variables. This block of variables contributed to 6.9 % of the variance to final grade in

the course (r-square = .069). These variables included course commitment, study habits,

104

time commitments and the total number of credits enrolled for the semester. The
student-focused block was certainly not as robust as was expected although many of the
individual variables included in the block were significant.

One of the significant variables within the student-focused block was course
commitment. The course commitment variable was a strong predictor of the final grade
in the course. Course commitment included attendance at lecture, attendance at recitation
and completing the weekly LON-CAPA assignments. The students in this study reported
attending lecture regularly and doing their LON-CAPA, however, they reported neutral
rather than agree or disagree on the GCS when asked if they attended recitation.

Attendance at lecture and recitation is not required in CEM 141 and CEM 142.
There were no points or any other credit given for attendance at lecture or recitation.
Students must have perceived attendance at lecture valuable, however, because most
attended. They responded neutral on the GCS on questions related to attendance at
recitation although they reported that they understood their TA and their TA did a good
job in recitation. This was somewhat interesting because the majority of the TA’S
teaching recitation were international students. In the past, students have commented that
they had trouble understanding their TA but that was not the case in this study.

Most students reported completing their weekly LON-CAPA assignments.
Completing LON-CAPA was not required but contributed to 24 % of the total points
toward their final grade. Student had ten chances to answer each LON-CAPA question.
With a little effort, these were essentially “free” points so there was a strong incentive to

earn these points toward their final grade.

105

Another significant variable within the student focus block was study habits. The
study habits variable included doing practice problems, studying on weekends and total
number of hours per week the students studied chemistry. This variable was not as
strong a predictor as course commitment. This finding was somewhat surprising because
it would seem that the more practice problems students completed and more total hours
studying would translate into higher course grades.

The students did not spend much time studying chemistry. The majority of the
students in this study spent between two to four hours per week studying. This is less
than half the two hours outside of class recommended for each course credit hour (CEM
141 is a four credit course).

The total number of credit hours enrolled during the semester was also a
significant variable in student-focused variables block. The GCS asked students for the
total number of credits enrolled and the total number of physical science credits enrolled
during the semester. The students enrolled in more credits earned lower final grades than
those with fewer credits. This finding is somewhat inconsistent with the literature.
Szafran (2001) determined that students who enrolled in more credits per semester earned
higher grades than those who enrolled in fewer credits, however, he concurred that those
who enrolled in more difficult courses tended to earn lower grades.

The results of this study conflict with Szafran (2001) partly because of the data
collected on the GCS. In this study, the total number of credits enrolled and the total
number of physical science credits enrolled were asked as two separate questions on the
GCS but were combined into one variable (“Total credits enrolled”) for the data analysis.

The decision to combine both questions into one variable fit the characteristics of the

106

students in the sample. If a student enrolled in the general chemistry course (CEM 141 or
CEM 142), the general chemistry laboratory course and a college algebra and
trigonometry, he or she had 10 credits out of a normal 15 credit hour load for the
semester. This was also true for the students in this sample. The analysis of the data
indicated that the mean number of physical science credits enrolled was greater than the
mean number of non-physical science credits enrolled for the semester. This may
partially explain why students enrolled in more credits during the semester did not do as
well as students enrolled in fewer credits. When the data were analyzed in relation to
physical versus non-physical science courses the results of this were consistent with
Szafi'an (2001) regarding the relationship between grades, number of credits enrolled and
difficulty of courses.

The data also revealed that the total number of credits enrolled per semester
influenced academic success. Students in a large number of physical science courses
tended to be less successful academically. This outcome poses a difficult problem.
Students majoring in the physical sciences are usually required to enroll in traditionally
difficult courses. Furthermore, most upper level math and science courses require
sufficient prerequisites that may add to the total course load per semester.

Advisors must be aware that enrolling in a large number of difficult courses has a
negative effect on academic success. Advisors and academic program directors must
look at the sequencing of courses in the total academic program. How many courses are
too many? That is probably up to the individual student but they should be made aware
that in case of difficult courses, less is probably better and they must be prepared to

budget their time accordingly.

107

Instructor—Focused Block and Instructor-Focused Variables

The instructor-focused block contributed to 1.3% of the variance toward the final
grade in the course (r-square = .013). None of the variables in the block were significant.
This was very surprising because 19 questions on the GCS were related to student
perceptions of their course instructor with the anticipation that the course instructor
would be a significant factor that influenced academic success in the course. In the case
of this study and these data, the instructor did not appear to be a significant factor.
Again, more research is required into the relationship between course instruction and

other instructor-focused variables, and course satisfaction.

Instructional Resources Block and Instructional Resources Variables

The instructional resources block accounted for 1.2% of the total variance in the
final course grade (r-square = .012). The use of the help room to LON-CAPA and other
chemistry assistance had a positive influence on final course grade.

The only other significant variable in the instructional resources block was
instructor contact. Instructor contact had a negative association with final grade. The
instructor contact variable asked students if they attended the instructor’s office hours and
if the instructor knew their name. Very few of the students indicated they visited their
instructors during office hours and fewer reported that the instructor knew their names. It
appears that the students who visited their instructor for help were in dire straits and

needed more help than the instructor could provide to assist them in the course.

108

Out-of-Class Experiences Block and Variables

The out-of-classroom experiences block was not significant and no individual
variables within the block were significant. These findings are particularly interesting. A
student’s time outside of the classroom is significant. When Terenzini noted that over
85% of a student waking hours are spent out of the classroom, there was some
expectation that that out-of-class experiences would be a significant factor for academic
success in general chemistry. Perhaps out-of—class experiences are contributing factors
for overall student learning and retention but not in specific courses, which was the focus
of this study.

In summary, when looking at the six blocks of variables in the student learning
model used for this study, approximately 40% of the variance contributing to the final
grade in college general chemistry could be attributed to student pre-college
characteristics and their in-class experiences. Out-of-class experiences, instructional
resources, student-focused variables and instructor-focused variables contributed to only
9.4%.

The results of this study suggest that efforts to increase academic success in
college general chemistry must be concentrated on two specific areas: students’ pre-
college characteristics and their experiences in the classroom. The bad news, as
previously discussed, is that influencing students’ pre-college characteristics and
experiences are beyond the control of higher education professionals. In addition, the
relationship between course grade and the most significant individual variable, course

satisfaction, may be specious. It cannot be determined whether or not students earned a

109

good grade in the course because they were satisfied or whether or not they reported
Strong course satisfaction because they were earning good grades in the course.

The good news is that there were a number of variables that can be influenced that
will help students succeed academically in college general chemistry. One is instructor
contact. Instructors often lament that the students they see during their office hours are
the ones that need it the least. From the data in this study, students who attend their
instructor’s office hours needed help badly. Therefore, if it appears to an instructor that a
student is simply not “getting it,” they are probably correct. Instead of simply going over
a few problems, answering a few questions, or reiterating concepts covered in lecture, the
instructor should ask some questions to assess the knowledge level of the student. If it is
not possible or practical to help the student during limited office hours, then the student
should be referred to an alternate source of assistance such as their teaching assistant, the
help room, or private tutor.

Another way to make students more successful is to increase their commitment to
the course. Students who reported stronger course commitment earned a higher grade
than those who did not. Course commitment included attending lectures and recitations,
and completing their LON-CAPA. Instructors must make course lectures worth
attending. Required attendance may be one way of making lecture attendance more
attractive. Attendance was not mandatory for the sample used for this study. It would be
interesting to conduct an additional study with randomly assigned students to lectures
with required attendance to see if attendance at lecture had any effect on the final course

grade or if students who did not want to be there would not do well anyway.

110

Doing homework was also essential for academic success in general chemistry.
In this study, students were assigned LON-CAPA that served as computerized
homework. Students need to practice chemistry problems through guided homework.
Whether or not giving credit for homework problems is an effective method to induce
students to practice chemistry problems is another good question for firrther research.

From the results of this study, students who used the help room were more
successful in general chemistry. Again, more research is required to determine why
students who visit the help room were more successful in the course. Perhaps it is
because they are able to get questions answered on an individual basis in a less
threatening environment. Students may be intimidated by asking a “dumb” question in
lecture or recitation, but feel more comfortable working one-on-one with a teaching
assistant. There is also some informal peer tutoring and collaborative learning in the
help room. Students may speak with each other while they are waiting to meet with the
next available help room tutor and learn from the interaction. In this study student
success was negatively influenced by attending instructor office hours but positively
influenced by visiting the help room. An explanation of why one source of remediation
appears to be more efficacious for academic success in college general chemistry than the

other will require further research.

Suggestions for Model Revision
The framing of the study to look at academic success from a cognitive and non-
cognitive perspective was sound. Collecting data on student pre-college characteristics

and their experiences in the classroom and experiences out-of-the classroom captured

111

most of the desired variables. The block analysis is useful in identifying the overall
importance of distinct student characteristics and behaviors.

Future studies should use more complex models to study the interchange between
the blocks in the model. For example, the in-class experience block had the highest r-
square in the model, but it was most likely a fimction of one variable, course satisfaction.
When all the variables were entered into a regression equation, course satisfaction was an
extremely important factor while the rest of the in- class experiences variables were not
significant. When each block was added to the regression equation, in-class experiences
became significant when in reality the only significant variable was course satisfaction.

Further research might consider course satisfaction as an additional outcome
variable, one that precedes course grade. It is difficult to tell whether students were
satisfied with the course because they earned a good grade, or they earned a good grade
because they were satisfied with the course. Moreover, because course satisfaction
emerged as such a strong influence, it would be interesting to investigate the factors that
influence course satisfaction.

The third consideration is to look at some of the bi-directional influences of the
variables in the study. In the model used for this study the direction of the analysis was
all in one direction. Each block of variables was analyzed in relation to the final grade in
the course. A revised model would include a look at the relationship between pre-college
characteristics and final grade without controlling for all other variables. In addition, the
relationship between the individual blocks of variables would be investigated. For
example, student study habits and the student focus block may influence use of the help

room or other instructional resources.

112

Further Research

The data collected for this study examined factors that influence academic success
in college general chemistry. The GCS collected data on students enrolled in general
chemistry at Michigan State University during fall semester 2003. The strongest
predictor of academic success in general chemistry was course satisfaction (t =1 1.172, p
< .001). The more satisfied students earned higher final grades. This finding suggests
that further research on the relationship between satisfaction and performance is
warranted.

In this study, the course satisfaction variable was constructed fi'om three questions
on the GCS that asked the students whether or not they were satisfied with their
performance in the course, if their grades on exams were about what they expected, and if
they were satisfied with the course overall. To gain a greater understanding of the
relationship between course satisfaction and academic success, developing and testing a
new two-stage least squares model with course satisfaction, first as a dependent variable
and then as a predictor of academic success, seems worthwhile. Using course satisfaction
as a dependant variable requires eliminating from the satisfaction scale the item on exam
grades and satisfaction with performance.

Another consideration for further research would be to collect data on instructor
perceptions of the course. The instructor focus block was significantly related to
academic success but less so than the student pre-college characteristics and in-class
experience blocks. Two of the variables in the instructor focus block - the instructor
attitude variable and the instructor availability variable - were most significantly related

to academic success and justify further investigation. This investigation is especially

113

important because of the negative relationship between instructor contact and final course
grade. The data from this study indicate that students who attended their instructor office
hours earned a lower final grade in the course. It would be interesting to investigate this
finding from the instructor’s perspective and compare it to the perception of the students
in relation to instructor attitude and instructor contact.

In this study, the student pre-college characteristics and experiences block of
variables was a strong predictor of academic success. Student pre-college characteristics
and experiences contributed 19.8% in shared variability with the final course grade. It
would be beneficial to do further research incorporating student demographic data in
addition to other pre-college variables in relation to course final grade. Such data would
include creating a dummy variable for gender and ethnicity in the regression equation.

In addition, it would be interesting to include whether or not a student was from
an urban, rural, or suburban high school. The students in this sample represented almost
every county in Michigan in addition to non-resident and international students. The data
set included county of residence (if from Michigan) but it was not possible to determine
the type of high school attended (such as urban, rural, or suburban) from county of
residence data alone. For example, approximately 12% of the students in the sample
were from Wayne County, Michigan. Wayne County includes the Detroit Public Schools
(an urban school) and the Grosse Pointe Public Schools (a suburban school), representing
two very different demographics. One way to collect data on the type of high school
attended would be to include a question on the GCA asking students to classify their high

school. Another would be to contact the Admissions Office to see whether or not they

114

code the type of high school attended when they enter the student application information
on the admissions database.

As mentioned earlier, general chemistry is a service course for many other majors.
Therefore it is important for students to earn good grades in the course to be admitted to
the nursing, engineering, medical, and veterinary schools, or other professional programs.
It would be interesting to follow up with the students in this sample to see where they are
now. Did a poor grade in course impede them from achieving their vocational goal? Did
a good grade facilitate achieving their aspirations? Or did their grade make no difference
in the pursuit of their educational goals? There is much written about the “gatekeeper“
effect of courses such as general chemistry. Failing the course or dropping out of college
would certainly interfere with a student’s entry into medical school or engineering but
does earning a barely passing grade have the same restrictive effect? It would be
beneficial to investigate the impact of the “gatekeeper” effect of general chemistry on the

future academic, career, and economic lives of students.

Scope and Limitations
“One of the more seductive dangers in the development of a body of knowledge is
the temptation to seek premature closure; to think that we have learned enough.”
(Pascarella, 1985, p. 53). This study investigated the factors that influence academic
success in college general chemistry. The data and conclusions contained herein pertain
only to the students enrolled in CEM 141 and CEM 142 at Michigan State University
during the fall semester of 2003. Year after year, however, the population of students

enrolled in CEM 141 and CEM 142 is very consistent. Whether these factors generalize

115

to other chemistry courses at Michigan State University or general chemistry courses at

other institutions is beyond the scope of this study.

116

APPENDICES

117

APPENDIX A
MSU General Chemistry Survey Data

2003

118

Regression coefficients for blocks of variables In conceptual model

 

Total R2 Block A R2

 

.198 .198”
Block 1: Pre-college characteristics
Grades in high school science
Parents education/income
Total years of math and chemistry
ACT English, reading, science, comprehensive.
ACT math/AOP math

Block 2: In-Class Experiences .398 .200“
Course organization
Classroom conditions
Course enjoyment
Satisfied w/course

.399 .002

Block 3: Out of class experiences

Campus life adjustment

New or existing friends

Knew roommate or were friends

Socializing

Campus life connection

Live on quiet floor or party house

.412 .012“
Block 4: lnstmctional resources
Usefulness of instructional resources
Use help room for CAPA and other
Help from friends or other sources
Instructor contact

.481 .069"
Block 5: Student focus
Course commitment
Study habits
Time commitments
Total credits enrolled

.494 .013“
Block 6: Instructor focus
Instructor basic skills
Instructor more advanced skills
Instructor attitude
instructor avail/office hours adequate
TA does good job/understand TA

 

"p<.001
* p<.05

119

Regression Coefficients for the Variables In the Conceptual Model

 

 

8 Std. Beta 1 Sig.
Error
(Constant) 1 .199 .805 1 .490 .137
Pre-college characteristics
Grades in high school science .308 .067 .169 4.573 .000"
Parents education/income .022 .042 .019 .535 .593
Total years of math and chemistry .007 .053 .005 .126 .900
ACT-English, reading, science, comprehensive .016 .006 .108 2.856 .004”
ACT math, AOP math .053 .010 .202 5.081 .000“
In-class experiences
Course organization -.038 .092 -.020 -.415 .678
Classroom conditions .010 .064 .006 .156 .876
Course enjoyment -.049 .036 -.056 -1.373 .171
Satisfied wlcourse .451 .040 .481 1 1.172 .000“
Out-of-class experiences
Campus life adjustment -.018 .044 -.016 -.416 .678
New or existing friends .021 .035 .024 .602 .547
Knew roommate or were friends -.006 .020 -.011 -.292 .771
Socializing .034 .099 .012 .338 .736
Campus life connection .261 .221 .041 1.180 .238
Live on quiet floor or party house -.028 .037 -.026 -.755 .451
Instructional resources
Usefulness of instructional resources -.027 .052 -.020 -.518 .605
Use help room for CAPA and other .135 .057 .085 2.362 .019‘
Help from friends or other sources -.021 .036 -.021 -.567 .571
Instructor contact -.169 .049 -.126 -3.438 .001"
Student focus
Course commitment .309 .047 .255 6.625 .000”
Study habits .125 .059 .077 2.1 14 .035*
Time commitments .015 .051 .010 .287 .774
Total credits enrolled -.170 .069 -.085 -2.461 .014‘
Instructor focus
Instructor basic skills .022 .060 .017 .367 .714
Instructor more advanced skills -.063 .066 -.046 -.960 .338
Instructor attitude -.123 .066 -.085 -1 .867 .063
Instructor available/office hours adequate .093 .050 .069 1.875 .061
TA does mod job/understand TA -.047 .030 -.059 -1.588 .113

 

*“p < .001, "p < .01, *p < .05

120

Descriptive Statistics for General Chemistry Survey Items

 

N Min Max M SD

 

Course organization

 

 

 

Course goals clear 1884 1 5 1.71 0.67
Syllabus accurate 1887 1 5 1.45 0.58
Understand grade calculation 1881 1 5 1.67 0.83
Course calendar followed 1873 1 5 1.68 0.74
Course organized 1882 1 5 1.83 0.84
Classroom Conditions
Can hear instructor 1879 1 5 1.5 0.59
Can see instructor 1884 1 5 1.6 0.67
See board or overhead 1877 1 5 1.69 0.76
Course eniovment
Enjoy chemistry 1878 1 5 3.08 1.16
Take CEM if not required 1878 1 5 3.66 1.25
Course satisfaction
Satisfied w/perforrnance 1876 1 5 3.12 1.18
Grade about what expected 1877 1 5 3.53 1.18
Satisfied with course 1864 1 5 2.76 1.12
SLatinq preference (not used)
Prefer sit in front 1884 1 5 2.53 1.22
Prefer sit in back 1884 1 5 3.53 1.21
No seating preference 1879 1 5 3.35 1.22
Campus life adjustment
Get homesick 1865 1 5 3.15 1 .24
Consider leaving 1866 1 5 3.76 1.27
Go home weekends oft 1864 1 5 3.56 1.29
Attend closer to home 1864 1 5 4.11 0.98
Mistake to come 1873 1 5 4.46 0.81
New or existing friends
Made few friends 1874 1 5 3.84 1.18
Knew friends before 1870 1 5 3.62 1.23
Knew roommate or were frienjds
Knew roommate before 1840 1 5 3.36 1.64
Roommate friends before 1838 1 5 3.46 1.58
Go on dates (1=yes, 2=no) 1869 1 2 1.38 0.48
Go to parties (1 =yes, 2=no) 1869 1 2 1.23 0.42
Drink Alcohol (1 =yes, 2=no) 1852 1 2 1.36 0.48
Campus life connection
Commuter Student (1 =yes, 2=no) 1878 1 2 1.91 0.28
Live off campus (1=yes, 2=no) 1878 1 2 1.79 0.41
Live at home (1 =yes, 2=no) 1875 1 2 1.96 0.19

 

121

 

 

 

 

 

N Min Max M SD

Organizational affiliation

Belong to club/org (1=yes, 2=no) 1873 1 2 1.59 0.49

Attend religious services (1=yes, 2=no) 1874 1 2 1.73 0.45
Usefulness of instructional materials

Textbook helpful 1874 1 5 3.02 1.04

Study guide helpful 1867 1 5 2.57 1.00
Use hel room for CAPA or other

Help room for CAPA 1882 1 5 3.5 1.26

Visit help room 1885 1 5 2.35 1.23
Help from friends or other sources

Help from other sources 1888 1 5 3.98 1.15

Help from friends 1890 1 5 2.47 1.19
Instructor contact

Attend Instructor office hrs 1880 1 5 4.06 0.85

Instructor knows name 1883 1 5 4.46 0.80
Coprse commitment

Attend every lecture 1884 1 5 2.17 1.13

Attend every recitation 1880 1 5 3.34 1.38

Do all CAPA 1883 1 5 1.36 0.64
Time commitments

Ave hrs work for pay/week (1=1-5, 2=6-10, 3:11-15, 1670 1 5 1.85 1.23
4:16-20, 5> 20

Ave hrs non-paid work/week (1=1-2, 2=3—4. 3=5-6, 4=7-8, 1639 1 5 1.26 0.68
5> 9)
Total credits enrolled

Total physical sci cred enrolled (1 =4-5, 2=6e7, 3=8-9, 1858 1 5 2.65 1.24
4:10-11, 5> 12

Total non-physical sci credits enrolled (1 =1-2, 2=3-4, 3=5- 1839 1 5 3.1 1 1 .23
6, 4=7-8, 5> 9
Study habits

Do practice problems-not CAPA 1873 1 5 2.83 1.19

Study weekends 1866 1 5 2.52 1.08

Hrs/week study chemistry 1=0-1, 2=2-4, 3=5—7, 4=8-10, 1872 1 5 2.23 0.78
5>10
Instructor basic skills

Instructor prepared 1883 1 5 1.58 0.75

I like Instructor 1881 1 5 1.97 0.92

Instructor speaks intelligibly 1885 1 5 1.71 0.79

Instructor makes few errors 1883 1 5 2.01 1.00

 

122

 

 

 

 

 

 

N Min Max M SD

Instructor more advanced skills

Instructor uses examples in lecture 1888 1 5 1.79 0.78

Instructor asks questions in lecture 1881 1 5 2.24 0.95

Instructor answers questions lecture 1883 1 5 2.15 0.90

Instructor reviews/summarizes 1879 1 5 2.43 0.99
Instructor attitude

Instructor sounds excited 1885 1 5 2.26 0.99

Instructor likes teaching 1882 1 5 2 0.85

Instructor likes chemistry 1878 1 5 1.6 0.67

Instructor uses humor 1881 1 5 2.32 1.03

Instructor cares if learn 1874 1 5 2.54 0.94
Instructor available/office hours adeguate

Instructor avail office hrs 1858 1 5 2.39 0.76

Instructor office hrs adequate 1859 1 5 2.41 0.79
TA doeflood iob/underst_and TA

Understand TA 1881 1 5 2.55 1.22

TA does good job 1871 1 5 2.57 1.19
W

Grade in HS Algebra II 1857 0 4 3.6 0.65

Grade in HS Chemistry 1782 0 4 3.41 0.71

Grade in HS Physics 1371 0 4 3.38 0.80
Grades in high school science

Years of HS math (1=0, 2=1, 3=2, 4=3, 5=4) 1876 1 5 4.84 0.44

Years of HS chemistry (1 =0, 2=1, 3=2, 4=3, 5=4) 1871 1 5 2.17 0.59
W

Highest degree mom 1< HS, 2=HS grad, 3=BA/BS, 1843 1 5 2.79 0.82
4=MA/MS, 5=PhD

Highest degree dad 1<HS, 2=HS grad, 3=BA/BS, 1816 1 5 3.03 0.98
4=MA/MS, 5=PhD

Parents Income 1=0—39K, 2=40K-79K, 3=80K-124K, 1712 1 5 3 1 .17
4=125K-150K, 5>150K
ACT-Engflsh/rea_di_n_g_flscience/comprehensive

ACT English 1904 9 36 23.64 4.10

ACT Read 1904 10 36 24.57 4.77

ACT Science 1904 12 39 24.37 3.73

ACT Comprehensive 1904 12 35 24.51 3.38
ACT mflh/AOP ma_th

ACT Math 1904 13 36 24.86 4.02

AOP math unproctored 1510 2 28 14.85 5.13
Like large lecture 1881 1 5 3.14 1.22
Prefer small classroom 1882 1 5 2.29 1.14

 

123

 

 

N Min Max M SD
Lecture hall comfortable 1875 1 5 2.89 1.02
ALLMSU for CAPA 1874 1 5 2.07 1.06
Do CAPA alone 1864 1 5 2.73 1.31
Seldom use helproom 1885 1 5 2.35 1.23
Old exams helpful 1877 1 5 1.69 0.93
Explains material others 1880 1 5 2.32 0.98
Get along with roommate 1850 1 5 2.01 1.01
Live on designated quiet floor 1858 1 5 3.82 1.39
Live party floor or house 1859 1 5 3.23 1.26
Chose residence college (1=yes, 2=no) 1871 1 2 1.84 0.37
Rushed fratemity/sorority (1=yes, 2=no) 1876 1 2 1.91 0.28
First-generation college student (1=yes, 2=no) 1874 1 2 1.77 0.42
Receive financial aid (1 =yes, 2=no) 1868 1 2 1.35 0.48
Physics in HS (1=yes, 2=no) 1878 1 2 1.32 0.47
Final Grade in general chemistry 2247 0 4 2.56 1.13
Honors Student (1 =yes, 2=no) 2252 1 2 1.95 0.22
Level (1=undergrad, 2=other) 2252 1 2 1 0.06
Class (1 =first year, 2=sophmore, 3=junior, 4=senior) 2245 1 4 1 .48 0.77
Sex 1=MaIe. 2=Female 2252 1 2 1.55 0.50
Ethnicity (1 =cauc, 2=black, 3=chicano, 4=hisp, 5=native 2070 1 9 1 .65 1.69
American, 6=Asiaanl
7=other, 8=not reported, 9=not requested
HS grade point average 2039 0 4.8 3.60 0.33
SAT Verb 433 28 76 547.39 83.6
SAT Math 433 36 80 580.6 83.9
Group (1=141 fall part, 2=142 non part, 3=141fal| non-part, 2252 1 4 1.45 0.80
4:142 fall non-part)
Total credit this semester 2252 0 20 14.2 1.53
Age at time of survey 2252 18 47 19.1 1.82

 

124

Factor Loadings

 

 

Standardized
rotated factor Alpha
loadings"

Course Organization

Course goals clear .690

Syllabus accurate .805

Understand grade calculation .614

Course calendar followed .722

Course organized .707 .738
Classroom conditions

Can hear instructor .849

Can see instructor .916

See board or overhead .863 .843
Course Enjoyment

Enjoy chemistry .845

Would take CEM if not required .789 .762
Satisfied w/course

Satisfied w/performance in course .899

Grades on exams about what expected .843

Satisfied with course .848 .829
Seating preference

Prefer to sit in front - .923

Prefer to sit in back .926

No preference .626 not used
Campus IIfe adjustment

Get homesick at times .594

Consider leaving MSU at times .807

Go home weekends oft .629

Attend college closer to home .791

Mistake to come to MSU .746 .745
New or existing friends

Made few friends .886

Knew friends before .846
Knew roommate or were friends

Knew roommate before .987

Roommate was friend before .987 .979
Socializing

Go on dates .601

Go to parties .840

Drink alcohol .799 .600
Campus life connection

Commuter student .860

Live off campus .670

Live with parents .711 .559

 

125

 

Standardized

 

rotated factor Alpha
loadings‘

Organizational affiliation

Belong to club or org .750

Attend religious services/meetings .795 .306

Usefulness of instructional resources

Textbook helpful .824

Study guide helpful .804 .490
Use help room for CAPA and other

Help room for CAPA .774

Use help room .900 .708
Help from friends or other sources

Help from other sources .818

Help from friends .775 .431
Instructor contact

Attend instructor office hours .843

Instructor knows student name .881 .665
Course commitment

Attend every lecture .800

Attend every recitation .706

Do all CAPA .591 .470

Study Habits

Do practice problems-not CAPA .822

Study weekends .817

Hours per week study chemistry -.739 .526
Time commitments

Average hours work for pay/week .749

Average hours non-paid work/week .702 .084
Total credits enrolled

Total physical science credits enrolled - .912

Total non-physical science credits enrolled .915 -4.02
instructor basic skills

Instructor prepared .856

I like instructor .634

Instructor speaks intelligibly .863

Instructor makes few errors .862 .8391
Instructor more advanced skills

Instructor uses examples in lecture .777

Instructor asks questions in lecture .795

Instructor answers questions in lecture .764

Instructor reviews/summarizes .698 .749

 

126

 

Standardized

 

rotated factor Alpha
. loadings'

Instructor attitude

Instructor sound excited in lecture .863

Instructor likes teaching .874

Instructor likes Chemistry .768

Instructor uses humor in lecture .718

Instructor cares if learn .572 .827

instructor available/oflice hrs adequate

Instructor available during office hrs .915

Instructor office hours adequate .912 .803
TA does good job/understand

Understand TA .962

TA does good job in recitation .962 .919
Grades in high school science

Grade HS algebra II .744

Grade HS chemistry .769

Grade in HS physics .720 .589
Parents education/income

Highest degree mom .755

Highest degree dad .816

Parents combined yearly income .776 .674
Total years math and chemistry

Years of HS math .731

Years of HS chemistry .731 .124

A CT-Engiish/reading/ science/
comprehensive

ACT English .825

ACT Reading .869

ACT Science .787

ACT Comprehensive .985 .896
ACT math/A OP math

ACT math .798

AOP math(un-proctored) .908 .638

 

'Factor loadings- Principal component analysis and oblique rotation

127

BIBLIOGRAPHY

Anaya, G. (1999). College Effects on Student Learning: Comparing the use of self-
reported gains, standardized test scores and college grades. Research in Higher
Education, 40(5), 499-526.

Angelo, T. E. (1993). A teacher's dozen: Fourteen general, research-based principles for
improving higher learning in our classrooms. American Association for Higher
Education Bulletin, 45, 3-7, 13.

Astin, A. W. (1968). The College Environment. Washington: The American Council on
Education.

Astin, A. W. (1971). Predicting academic performance in college; selectivity data for
2,300 American colleges. New York: Free Press.

Astin, A. W. (1984). Student involvement: A developmental theory for higher education.
Journal of College Student Personnel, 25, 297-308.

Astin, A. W. (1991). Assessment for excellence: the philosophy and practice of
assessment and evaluation in higher education. New York: Macmillan.

Astin, A. W. (1993). What matters in college? Four critical years revisited. San
Francisco: Jossey-Bass.

Axtell, J. (2003). What’s wrong? and right? with American higher education? The
Virginia Quarterly Review, 79 (2), 189-208.

Bauer, K. W. & Liang, Q. (2003). The effect of personality and pre-college
characteristics on first-year academic activities and academic performance.
Journal of College Student Development, 44(3), 277-290.

Berger, J. B, & Milem, J. F. (1999). The role of student involvement and perceptions of
integration in a causal model of student persistence. Research in Higher
Education, 40(6), 641-664.

Boughan, K. (1996). Correlates of chemistry 101 course performance. ERIC Document
Reproduction Service No. ED 397919.

Bonwell, C .C. & J.A. Eison. 1991. Active learning: Creating excitement in the
classroom. ASHE-ERIC Higher Education Report No. 1. Washington, DC:
George Washington University

128

BouJaoude, S. B. & F. J. Giuliano (1994). Relationships between Achievement and
Selective Variables in a Chemistry Course for Nonmajors. School Science and
Mathematics. 94(6): 296-302.

Brown, T. L., LeMay, H. E., Jr., & Bursten, B. E. (2003). Chemistry: The central science.
(9th Edition). New JerseyzL Prentice Hall

Burdge, J. R. and S. W. Daubenmire (2001). "Raising Success Rates in Freshman
Chemistry. Journal of College Science Teaching. 30(5): 295-99.

Chickering, A. W. & Reeser, L., (1993). Education and identity. San Francisco: Jossey
Bass.

Cote’, J. E., & Levine, C. G. (2000). Attitude versus aptitude: Is intelligence or

motivation more important for positive higher-educational outcomes? Journal of
Adolescent Research, 15(1), 58-80.

Cove, P. G., & Love, A. G., (1996). Enhancing student learning: intellectual, social and
emotional integration. ERIC Digest No. ED 400741.

Ditzer, M.A. & Ricci, R.W. (1995). Discovery chemistry: A laboratory approach to
general chemistry. In New Directions for Teaching and Learning (Robert S.
Menges & Marilla D. Svinicki, eds) (6) Spring 1995. p.5-43. San Francisco:
Jossey-Bass.

Deal, W. J. (1984). Predictions of Course Grades: Uses and Uncertainties. Journal of
College Science T caching, 13(3): 154-56.

Feldman, K. A. & Newcomb, T. M. (1969). The impact of college on students. Vol.1. An
analysis of four decades of research. San Francisco: Jossey Bass.

Fletcher, R. K. (1977). A summary of the research literature predicting success in college
chemistry. ERIC Document Reproduction Service No. ED155054.

Fletcher, R. K. (1978). Probability of success in college chemistry. ERIC Document
Reproduction Service No. ED168884.

Flowers, L. & Pascarella, E. T.(2003). Cognitive effects of college: Differences between
African American and Caucasian students. Research in Higher Education. 44(1):
21-49.

Gainen, J. (1995). Barriers to success in quantitative gatekeeper courses. In Robert S.
Menges & Marilla D. Svinicki (Series Ed.) & Gainen, J. & Willemsen, E. W.
(Vol. Ed), New Directions for Teaching and Learning: Number 61. Fostering
student success in quantitative gateway courses. (Spring 1995. pp. 5-43). San
Francisco: Jossey-Bass.

129

Gagne’, R. M. & Medsker, K. L. (1996). The conditions of learning: training
applications. Orlando: Harcourt.

Kotz, J .C. & Treichel, P.,Jr. (1996). Chemistry & chemical reactivity (3rd ed.). New York:
Saunders College Publishing.

Kuh, G. D., Schuh, J. H., Whitt, E. J., and Associates. (1991). Involving colleges:
Successful approaches to fostering student learning and development. San
Francisco: Jossey Bass

Kuh, G. D., Pace, R. C. & Vesper, N. (1997). The development of process indicators to
estimate student gains associated with good practices in undergraduate education.
Research in Higher Education, 28, 435-454.

Hagedorn, L. S., Siadat, V. M., F ogel, S. F., Nora, A.& Parcarella, E. T. (1999). Success
in college mathematics: Comparisons between remedial and nonremedial first-
year college students. Research in College Mathematics, 40, 261-284.

Harwood, W. S. (1996). The one-minute paper: A communication tool for large
lecture classes. Journal of Chemical Education, 73, 229-233.

Hunter, P. W. (2003). Study guide to accompany Kotz and Treichel 's chemistry and
chemical reactivity fifth edition. Brooks/Cole.

Madhavi, J. & Nelson, J. S. (2002). Savvy decision making: An administrator’s guide to
using focus groups in schools. Thousand Oaks: Corwin Press.

Mason, D. & Mittag, K. C. (2001). Evaluating the success of Hispanic-sumame studne in
first-semester general chemistry. Journal of Chemical Education, 78 (2), 256-
259.

Milem, J. F ., & Berger, J. B. (1997). A modified model of college student persistence:
Exploring the relationship between Astin’s theory of involvement and Tinto’s
theory of student departure. Journal of College Student Development. 38(4),
387-399.

NCES (2003). Projections for education statistics to 2013.
http://nces.ed.gov/prpgrams/projections/

Naumann, W.C., Bandealos, D. & Gutkin, T. B. (2003). Identifying variables that
predict college success for first-generation college students. Journal of College
Admission. 181, 5-9.

Odgen, W. R. (1976). The affect of high school chemistry upon achievement in college
chemistry: A summary. School Science and Mathematics 76(2), 122-126.

130

Oliver-Hoyo, M. T. & Allen, D. (2005). Attitudinal effects of a student—centered active
learning environment. Journal of Chemical Educaion, 82(6), 944-949.

Ostrom, C.W. & Rogers, W. (1996). The undergraduate experience at Michigan State
University 1996. Unpublished report conducted by the Survey Research
Division, Institute for Public Policy and Social Research. Prepared for the Office
of the Vice President for Student Affairs and the Undergraduate University
Division. East Lansing, MI: Michigan State University.

Ozsogomonyan, A., & Loftus, D. (1979). Predictors of General Chemistry Grades.
Journal of Chemical Education; 56(3), 173-75.

Pace, R. C. (1979). Measuring outcomes of college: Fifty years of finding and
recommendations for the future. San Francisco: Jossey Bass.

Pace, R. C. (1984). Measuring the quality of college student experiences. Los Angeles:
University of California, The Center for the Evaluation, Graduate School of
Education.

Pace, R. C. & Kuh, G. D. (1998). College student experiences questionnaire, Fourth
Edition. Bloomington: Indiana University Center for Postsecondary Research
and Planning

Palomba, C. A. & Banta, T. W. (1999). Assessment essentials: Planning, implementing,
and improving assessment in higher education. San Francisco: Jossey Bass.

Paul, A. D. (1978). Factors affecting student performance in general chemistry. Journal
of College Science Teaching; 7(5), 301-304.

Paulson, D. R. (1999). Active learning and cooperative learning in the organic chemistry
lecture class. Journal of Chemical Education, 76(8), 1136-1140.

Pascarella, E. T. (1985). College environmental influences on learning and cognitive
development. In J. C. Smart (Ed.), Higher Education: Handbook of theory and
research: Vol. I. (pp. 1-61. New York: Agathon.

Pascarella, E. T., Edison, M., Hagedorn, L. S., Nora, A. & Terenzini, P. T. (1996).
Influences on students’ internal locus of attribution for academic success in the
first-year of college. Research in Higher Education, 37(6), 731-756.

Pascarella, E. T., Edison, M., Nora, A., Hagedorn, L. S., & Braxton, J. (1996). Effects of

teacher organization/preparation and teacher skill/clarity on general cognitive
skills in college. Journal of College Student Development, 37(1), 7-19.

131

Pascarella, E. T., & Terenzini, P. T. (2005). How college affects students: a third decade
of research. San Francisco: Jossey-Bass.

Pascarella, E. T., & Terenzini, P. T. (1991). How college affects students: Findings and
insights from twenty years of research. San Francisco: Jossey-Bass.

Perry, W. G. (1981). Cognitive and ethical growth: The making of meaning. In A.W.
Chickering & Associates (Eds.), The modern American college. P. 76-116. San
Francisco: Jossey Bass.

Pratt, D. D. (1998). Five perspectives on teaching in adult & higher education. Malabar,
FL: Krieger Publishing Co.

Richardson, S. M., & Sullivan, M. M. (1994). Identifying non-cognitive factors that
influence success in academically underprepared freshmen. Journal of the
Freshman Year Experience, 6(2), 89-100.

Reihl, R. J. (1994). The academic preparation, aspirations and first-year performance of
first-generation students. College and University, 70(1), 14-19.

Rosenberg, K.M. & Daly, H.B. (1993). Foundations of behavioral research: A basic
question approach. Fort Worth: Harcourt.

Royal Swedish Academy of Sciences (October 5, 2005). The Nobel Prize in Chemistry,
2005. http://www.kva.se/KVA Root/engj press/detail.asp?Newsld=697

Sanchez, K.& Betkouski, M (1986). A study of factors affecting student performance in

community college general chemistry courses. (ERIC Document Reproduction
Service No ED275549).

Szafran, R. F. (2001). The effect of academic load on success for new college students: Is
lighter better? Research in Higher Education, 42, 27-50.

Terenzini, RT. (1991). Out-of-class experiences research program. The Transition to
college project. Final report. Pennsylvania. (ERIC Document Reproduction
Service No ED357710).

Terenzini, P. T., & Pascarella, E. T. (1994). Living with myths: Undergraduate education
in America. Change, 28-32.

Ting, S. R. (1997). Estimating academic success in the first-year of college for specially
admitted white students: A model combining cognitive and psychosocial
predictors. Journal of College Student Development, 38(4), 401-409.

Ting, S. R. (1998). First-year grades and academic progress of college students of first-
generation and low-income families. The Journal of College Admission, 15-23.

132

 

Tinto, V. (1986). Theories of student departure revisited. In J. C. Smart (Ed.), Higher
Education: Handbook of theory and research: Vol. II. (pp. 359-384). New York:
Agathon.

Wagner, E. P., H. Sasser, et al. (2002). Predicting Students at Risk in General Chemistry

Using Pre-semester Assessments and Demographic Information. Journal of
Chemical Education. 79(6), 749-55.

Weidman, J .C. (1989). Undergraduate socialization: A conceptual approach. In J. C.
Smart (Ed.), Higher education: Handbook of theory and research: Vol. V. (pp.
289-321). New York: Agathon.

Whitt, E. J ., Marth, B. M, Pascarella, E. T., Pierson, C. T.& Neshein, BS. (2003).
Differences between women and men in objectively measured outcomes, and the
factors that influence those outcomes, in the first three years of college. Journal
of College Student Development, 44(5), 587-610.

Willingham, W. W. (1985). Success in college: The role of personal qualities and
academic ability. New York: College Entrance Examination Board.

Yager, R. E. (1989). Success in college physics and chemistry for high ability students

with corresponding high school courses. School Science and Mathematics,
89(8), 676-687.

133

 

 

IIIIIIIIIIIIIIIIIIIIIIIIIIIIII

IIlIIIIIIIIIIIIIIIIIIIIIIllIIIIIIIIIIIIIIIIIIIII