Outcomes of Service-Learning in an Introduction to Sociology Course
number of Sociology courses (Ender, Kowalewski,
Cotter, Martin & DeFiore, 1996; HondagneuSotelo & Raskoff, 1994; Galura, 1993; Parker--
Gwin, 1996; Porter & Schwartz, 1993; Taub,
1991), there are no studies of outcomes in servicelearning courses in the discipline. The contribution
of my study is to expand our knowledge of outcomes to Sociology and examine the effects of service-learning on predominantly first-year students
in a relatively large course (at least by four-year
liberal arts college standards).
Methods
The design is that of a natural experiment which
compares two sections of the same course,
Introduction to Sociology, at the State University
of New York's College at Cortland. I taught both
sections back to back in the Fall semester, 1994.
The courses were as identical in content as I could
make them, except that in one class I required 20
hours of service experience and in the other I did
not. In order to make the workloads comparable,
students in the non-service-learning section were
assigned various articles from the weekday New
York Times.
My objectives for the service-learning and New
York Times reading requirements were the same -
to improve students' ability to identify and apply
relevant course concepts to everyday experience.
For example, when I introduced culture, I defined
the concept, described components of culture
(shared symbols, values, beliefs, norms), illustrated
each component with examples, and invited students to do the same. I used service-learning and
the Times reading to have students apply the concept to different contexts. Service-learning students were asked to analyze the cultures of their
agencies, while the Times readers were asked to
find examples of how culture is represented in the
news. In addition, a reading of the research literature led me to hypothesize that service-learning
would improve students' performance on knowledge content exams, knowledge application problems, and attitudes toward social responsibility and
personal efficacy.
The project is modeled after (although not identical to) a similar experiment conducted by
Markus, Howard, and King (1993) in a political
science course at the University of Michigan (hereinafter referred to as the Michigan study). Some
measures of learning outcomes are also similar to
those used by Giles and Eyler (1994) in their study
of service-learning programs at Vanderbilt
University (hereinafter referred to as the Vanderbilt
study). Like Markus, Howard, and King, and
Giles and Eyler, I examined the effect of service
learning on a variety of learning outcomes, including student performance in the course, attitudes
toward social responsibility and personal efficacy,
and student evaluations of the course and their
learning. Changes in social responsibility and personal efficacy were measured using the Social
Responsibility Inventory (administered as a preand post-test) developed by Jeffrey Howard and
Wilbert McKeachie. Student assessments of the
course were made using the University of
Michigan's Center for Research on Learning and
Teaching questionnaire. Tables 1, 2, and 3 show
the questions I used to assess students' attitudes
toward social responsibility, personal efficacy, and
course learning.
There are three performance measures used in
the class: percent of correct answers on four multiple choice examinations to measure understanding of core concepts, average of scores on two
essay examinations to measure ability to apply
course concepts, and a composite grade which
combines multiple choice and essay scores with a
participation grade based on attendance, completion of daily homework assignments, and extra
credit. Extra credit was awarded in the servicelearning course for participation above the 20-hour
minimum, and in the non-service-learning course
for summarizing and presenting relevant New York
Times articles which were not assigned in class.
I made grading in the two courses as comparable
as possible while differentiating between knowledge content and quality of thought in order to correct for two weaknesses of previous studies on service-learning. Miller (1994) pointed out that some
service-learning studies which compare students'
grades use two different methods to evaluate students - one method for the traditional classroom
course (such as in-class exams and term papers),
and another method for service-learning courses
(like journal writing and papers which apply course
concepts to the service experience).' Batchelder
and Root (1994) have called for separating knowledge content (ability to define concepts and recall
facts, for example) from quality of thought (like
the ability to reason and solve problems). The use
of multiple choice tests to measure knowledge content and essay questions to measure quality of
thought is an attempt to differentiate between the
two.2
Finally, I gathered background information on
characteristics such as sex, age, race, high school
GPA, SAT scores, class year in college, college
GPA, and prior and contemporaneous experience
as volunteers. These variables were selected to
control for effects on learning outcomes of extraneous factors.3
73
