4wednesday

4:00-4:50 p.m. SCIENCE EDUCATION (Discussion Session) Salon A

Presider: Bonita White, The University of Memphis

EIGHTH-GRADE SCIENCE TEACHERS USE OF INSTRUCTIONAL TIME: COMPARING QUESTIONS FROM THE THIRD INTERNATIONAL MATHEMATICS AND SCIENCE STUDY (TIMSS) AND NATIONAL SCIENCE FOUNDATION QUESTIONNAIRES

Anne B Davidson. George Mason University

             Did the Third International Mathematics and Science Study (TIMSS) ask eighth- grade science teachers the right questions about their use of instructional time? TIMSS asked teachers to recall a lesson that they had taught, and, then, to group activities into 11 categories. This study examined the TIMSS question "How did the lesson proceed?" by videotaping six classes of eighth-grade science in Alabama and Virginia and comparing observer coding of the video to the teachers' recalled descriptions of the same class.
             The manner in which the TIMSS data were collected and manner in which data were collected from teachers in this video tape study suggested the use of a repeated measures analysis of variance (ANOVA) model. Using a repeated measures analysis allowed a look at the interactions between teachers and observer and the 11 TIMSS activities, between teachers and observer and the 26 NSF student activities, and between teachers and observer and the 11 NSF teacher activities.
             The difference between observer and teacher responses using TIMSS categories was not significant; however, 43% of the total variance was explained by whether the teacher or the observer reported the times for the instructional activities. The teachers also responded to questions from the NSF Local Systemic Change Through Teacher Enhancement K-8 Teacher Questionnaire to describe the same class. The difference found between the teacher and the observer coding was not significant, but the amount of variance explained by the data source (observer or teacher) dropped to 33% when using NSF student activity categories and to 26% when using NSF teacher activity categories.
             The conclusion of this study was that questionnaires to survey science teachers about their instructional activities should include operational definitions, methods of classifying single activities into multiple instructional categories, and questions that are more accurate in describing quality science instructional activities.

USING COVER, COPY, COMPARE TO TEACH STUDENTS WITH DISABILITIES PARTS OF THE HEART

Tawnya J. Smith and Karen I. Dittmer, Mississippi State University, and Christopher H. Skinner, The University of Tennessee-Knoxville

             Students with learning disabilities often have difficulty across content areas. Various procedures have been developed for use with these students. Cover, copy, compare is a self-management procedure that allows students to acquire and generalize strategies allowing them to guide and regulate learning and performance. The purpose of this paper was to extend the research on cover, copy, compare procedure across students with learning disabilities in the content area of science.
             Three high school students with specific learning disabilities in a science special education class participated in the study. Students were instructed the techniques of CCC to learn 15 parts of the heart. During each session, students were given a copy of a heart diagram with all of the parts labeled on the outside of a manilla file folder. The inside contained three unlabeled diagrams with three liens for each part of the heart. Students were instructed to look at the labeled diagram and then open the folder completely and write the names of the first five parts of the heart on the unlabeled sheet. The students were to then check their answers. If the student did not label the part correctly, the student was to write the correct answer two more times on the blank lines adjacent to the first response. After each session, assessments were given to determine learning.
             The results of the study suggested that students with learning disabilities were able to learn the parts of the heart using a CCC procedure. Learning occurred over 11 sessions, and an average of 80% accuracy across all students was obtained for each set of the five heart parts.

COGNITIVE ILLUSIONS AS HINDRANCES TO LEARNING COMPLEX ENVIRONMENTAL ISSUES

Fred H. Groves and Ava F. Pugh, University of Louisiana at Monroe

The purpose of this study was to determine the effectiveness of an environmental science course for teaching the ozone depletion problem. Students were pre- and posttested in the spring 2000 semester with a 38-item questionnaire consisting of three sets of Likert-style statements, plus multiple choice questions. The first group of statements centered on results of ozone depletion, the second set on causes of ozone depletion, and the third set targeted ways to lessen ozone depletion. The seven multiple choice questions targeted factual information. The Likert- style rating ranged from "I am sure this is correct" to "I am sure this is wrong." The course, Life in the Environment, focused on current environmental issues, including global warming and ozone depletion. The instructional format was lecture/discussion supplemented by Power Point presentations.
Overall, posttest scores were lower than pretest scores, and students showed a strong tendency to conflate cause/effect relationships between several different environmental issues that are not actually related. This matches with previous research and indicates that as students become more generally aware of environmental problems, they tend to assume relationships that are not correct. This is especially pronounced for global warming and ozone depletion, with most students incorrectly believing that ozone depletion leads to global warming when the reverse is true, that global warming may increase ozone depletion. These results revealed the difficulty in guiding students to construct proper concepts for complex subject matter by means of traditional, direct instruction methods, even with the use of presentation devices such as Power Point. Students must be challenged to see that their prior assumptions do not properly accord with current explanations for complex problems, and they must be guided to develop sound concepts.

4:00-4:50 p.m. STUDENT ACHIEVEMENT (Discussion Session) Salon B

Presider: Cathy Stockton, Louisiana Tech University

UNDER FIRE: OUTCOMES OF A HIGH SCHOOL'S TUTORING EFFORTS IN THE WAKE OF A STATE TAKEOVER

Dennis C. Zuelke and J. Gordon Nelson, Jacksonville State University

             The problem was the process and outcomes involved in a small city high school's tutoring program for at-risk students (school enrollment under 300 students grades nine-twelve). The problem grew out of a state takeover of the high school in 1999-2000 because of low standardized achievement test scores. Though a nonprofit community agency had abandoned its after-school tutoring program in the city's schools, the high school's staff decided to continue tutoring on its own.
             Two professors from a local university pursued an assessment plan that included a template of variables, variable definitions, and data matrix. At least 25 volunteer tutees were subjected to the assessment. Twelve independent variables, plus Piers-Harris Self Concept Scale scores, were used with two dependent variables, reading and math end-of-year grade point averages (spring 2000 standardized test scores for the tutees also were used). Among the independent variables were hours tutored, type of tutoring (including kinds of tutors used), tutor pay, and tutee absenteeism.
             When tutoring ended in mid May 2000, the high school's staff collected the data and entered them on a computer spreadsheet. The staff prepared a computer data diskette that was then given to the professors. The professors used SPSS software to analyze the data from the diskette with descriptive statistics (including single order correlations).
             The tutoring literature clearly indicated that personalized tutoring with qualified tutors, including peer tutors, had a positive influence on student achievement. Contrary to the literature, prior years' assessments of the nonprofit community agency's after-school tutoring program in the city schools showed that achievement among tutees in reading and math did not improve. The professors determined the findings from the data analyses during the summer 2000. The findings may be instrumental in the state's decision about what to do with this high school in 2000-2001.

STUDENT MOTIVATION IN LARGE SCALE ASSESSMENT

Beverly M. Klecker and Robert Wetter, Kentucky Department of Education

             Student motivation in large scale assessment has been the focus of few research studies (O'Neil, Sugrue, & Baker, 1996; DeMars, 1999). The school is the unit of analysis for Kentucky's accountability assessment. Aggregate scores are accurate only if students respond with good faith effort. This study was designed to explore: (1) the status of the problem and (2) possible solutions.
             Data for this study were responses to student questionnaires by fourth- (N=45,751), seventh- (N=47,772), and eleventh- (N=19,546) grade students taking the Reading assessment and fifth (N=46,532), eighth (N=47,324), and 11th (N=19,308) grade students taking the Mathematics assessment (questionnaire matrixed for the 11th grade) Questions were "How hard did you try on this reading (mathematics) test?" and "How many of the reading (mathematics) questions tested things you learned in school?"
             Percentages of student responses were compared by grade level, gender, and race. The percentage of students reporting that they either "did not try" or "tried a little" increased as grade level increased for both content areas. Percentages of students who reported that they were taught either "none" or "some" of the material assessed also increased by grade level. A surprising 38.2% of students taking the eleventh-grade Reading assessment and 43.2% of eleventh-grade students taking the Mathematics assessment responded in this manner. Percentages of males responding "did not try" or "tried a little" was higher than the percentage of females at every grade level in both subjects. Most striking was the difference in eleventh-grade Reading (Males=18.17%, Females=8.41%). The differences in percentages between African American and white students decreased as grade level increased in both subjects.
             The confounding of student motivation and student knowledge and possible strategies for increasing student motivation were discussed.

PRE- AND POST-KERA STUDENTS' WRITING SKILLS - A COMPARATIVE STUDY

Qaisar Sultana and Lisa Kay, Eastern Kentucky University

             In 1989, Eastern Kentucky University (EKU) implemented the University Writing Requirement (UWR) as one of EKU's standards for graduation. The UWR consists of a prompt that is occasionally repeated. Students write a comprehensive essay on the prompts, which are scored by trained readers according to a rubric.
             The Kentucky Educational Reform Act (KERA) was passed in 1990. The center piece of KERA is accountability measured through the Commonwealth Accountability Testing System (CATS). One component of CATS is the writing portfolio. Since 1991 all fourth, eighth, and twelfth graders do the writing portfolio.
             This study was conducted to determine whether the writing skills, as measured by the UWR, of Kentucky high school graduates who underwent the CATS writing portfolios, had improved as compared to their pre-KERA peers. Specifically, the objectives were to: (1) compare the UWR scores of pre- and post-KERA students and (2) determine the significance of the difference.
             The researchers randomly selected 50 each UWR essays of Kentucky's high school graduates on a prompt, Spectators Sport, given in fall 1989 (Pre-KERA) and summer 1997 (post- KERA). The pre- and post-KERA groups were recorded by the last four digits of the social security number. Each essay was xeroxed to mask the effects of aging. Four UWR trained and experienced readers were asked to score the essays. They were not given any information pertaining to the purpose or methodology of the research.
             The scored essays were separated into pre-KERA and post-KERA groups using the four digits of the social security numbers. Means of the two groups were computed and t-tests were performed to determine the significance of the difference. No significant differences were found at any level between the means of the two groups. Results have implications both for UWR and CATS.

4:00-4:50 p.m. TEACHER EMPLOYMENT (Discussion Session) Meeting Room 1

Presider: Barbara Kacer, Western Kentucky University

TEACHER INDUCTION: IS THIS THE ANSWER TO TEACHER RETENTION?

Cheryl J. Cummins and Camille B. Branton, Delta State University

One issue facing many schools is the recruitment and retention of quality teachers. Teachers are becoming disillusioned and leave the profession within the first two to five years. Others fail to even teach at all. One of the major reasons novice teachers give for abandoning the professions is the lack of support and assistance in the beginning years.
What is the answer to teacher retention? Schools are now establishing teacher- induction programs to aid in the recruitment and retention of quality teachers. This paper addressed the role teacher induction plays in helping transform preservice teachers into competent seasoned professionals. Included was a discussion of the most common components of teacher induction programs such as mentoring, peer coaching, orientation inservice, and teacher-guided observations. A discussion of the importance of classroom management skills, student discipline techniches, and professional obligations was held as well.

THE ATTITUDES, CONCEPTUALLY CONSISTENT WITH OCCUPATIONAL BURNOUT, HELD BY UNITED STATES TEACHERS

Paige L Tompkins, Western Kentucky University

Based on (1) the importance of the retention of a quality teacher workforce as an education policy concern, (2) recommendations from researchers to give greater consideration to the role that teachers' attitudes and perceptions play in teacher burnout and subsequent attrition from the teaching profession, and (3) past research implicating burnout as the impetus for some of the most qualified teachers to leave teaching, the purpose of this study was to identify what attitudes and perceptions (conceptually consistent with occupational burnout) have been held by United States teachers. In order to determine, on a national level, what attitudes and perceptions were held, data from the latest round (1993-94) of the Schools and Staffing Survey (SASS) were used. This survey was conducted by the National Center for Education Statistics (NCES) of the United States Department of Education. These data were weighted by NCES to provide national estimates, thereby enabling conclusions generalizable to the nation's population of teachers.
To determine the attitudes and perceptions held by United States teachers, more than 40 survey items assessed by the 1993-94 SASS were used. The latent dimensions measured by the SASS items were obtained through an exploratory factor analysis. For reasons of practicality and ease of interpretation, the factors, representing composites of the SASS items, were then used as the indicators of teacher attitudes and perceptions for all subsequent analyses.
Implications for educational policy decisions regarding the retention of a quality United States teacher workforce included the amount of attention paid or not paid to the severity of attitudes consistent with occupational burnout, and the role these attitudes play in teacher attrition.

ESTABLISHING CUT SCORES FOR INITIAL TEACHER TESTING

Donna E. Pascoe and Glennelle Halpin, Auburn University

The school reform movement experienced during the previous decade turned attention on teachers and their qualifications to teach. With findings that indicate that the quality of teachers is related to the improvements in student performance, the push for better qualified teachers has intensified (Darling-Hammond, 2000). By increasing educational standards to improve educational outcomes, teachers are required to show that they have acquired the basic educational skills necessary to teach. These basic competencies are most often measured by standardized criterion testing. The practice of teacher testing for initial licensure has been instituted in 35 states with more to follow (NASDTEC, 1999).
This review of literature investigated past and current trends in standard setting for initial teacher licensing and an account of the methodology used to determine the cut scores for each state. A state-by-state compilation of how standards are set in the licensure process is included. The standard-setting process discussed includes four sections: (1) planning, (2) the passing score study, (3) setting the passing score, and (4) evaluating the effects of using the score. The findings of this review of literature should assist those state departments of education needing procedural directives for determining a cut score.

4:00-4:50 p.m. STATISTICS (Discussion Session) Meeting Room 2

Presider: Larry G. Daniel, University of North Florida

CREATIVE USES OF FACTOR ANALYSIS IN PSYCHOTHERAPY RESEARCH: PAST EXAMPLES AND FUTURE POSSIBILITIES

James M. Adams, The University of Southern Mississippi

             Of general linear model analyses, factor analysis has been dubbed as the "reigning queen of the correlational methods" (Cattell, 1978, p. 4). Indeed, factor analysis is an extremely flexible data analytic method that allows researchers to examine multiplicities in simpler terms. Most commonly, factor analysis is invoked to reduce multiple variables (typically items on attitudinal or achievement tests) into a smaller set of factors.
             However, the logic of factor analysis can be applied to many research situations in which variables are not the units being factored. For example, in so-called higher order factor analysis, factors themselves are factored in an attempt to determine whether the factors can be further summarized by one or more higher order factors. A commonly recognized example of this application is the currently popular Big Five factor model of personality, in which many variables have been used to identify personality characteristics, and these characteristics have, in turn, been summarized by five major aspects of personality.
             Furthermore, factor analysis can be applied to factor other units such as people and occasions. These two-mode analyses are generally limited by the creativity of the researcher. Essentially, whenever the researcher has reason to believe that units of analysis (whatever they may be) will cluster into meaningful subgroups, some form of factor analysis can be invoked to examine this possibility.
             The purpose of this paper was to provide an introductory framework for the many applications possible for factor analysis beyond the traditional factoring of variables. Both higher order and two-mode factoring were discussed regarding their theoretical foundations and potential applications. Furthermore, uses of these methods in psychotherapy research were reviewed, including discussion of examples from published literature. The role of factor analysis in examining future trends in psychotherapy research was also examined.

TOWARD REFLECTIVE JUDGMENT IN EXPLORATORY FACTOR ANALYSIS DECISIONS: DETERMINING THE EXTRACTION METHOD AND NUMBER OF FACTORS TO RETAIN

Jennifer L. Knight, The University of Southern Mississippi

             Factor analysis is frequently used to reduce variable complexities into more parsimonious models. While the logic of factor analysis allows for factoring people, occasions, and other units in two-mode strategies, factor analysis is most frequently applied to variables in both measurement and substantive studies.
             In exploratory factor analysis (EFA), many decisions must be made about how to proceed with the analysis. The impact of these decisions on the results obtained can be substantial. As such, each decision should be reflective of thoughtful researcher judgment in the context of the individual study. Unfortunately, empirical studies indicate that many researchers over-utilize default selections in statistical software packages. While these defaults are appropriate for some contexts, superior methods are often available.
             The paper addressed three EFA decisions in an attempt to present options available to the researcher. The primary purpose was to aid the researcher in making informed decisions during the factor analysis instead of relying on defaults in statistical programs or traditions of previous researchers. First, the importance of determining the number of participants and variables was discussed. Second, the similarities and differences in the two most common factor extraction methods, principal components (PCA) and principle factor analysis (PFA), were addressed. Finally, the paper compared accuracy of different methods for retaining factors (e.g., K1 rule, scree plot).
             Although PCA and PFA results converge with increasing numbers of variables and communality estimates, PFA may be the better choice for attempting to determine underlying factors. Furthermore, parallel analysis and minimum average partial methods show great promise to facilitate precision in factor retention. However, researchers should take into consideration a combination of methods in order to be as accurate as possible in developing the factor model. Empirical studies indicate that superior methods (e.g., parallel analysis) are often grossly underutilized.

AN EXAMINATION OF STATISTICAL TESTS USED TO DETERMINE THE NUMBER OF COMPONENT SUPBPOPULATIONS IN A FINITE MIXTURE MODEL

Frank R. Lawrence and James E. McLean, The University of Alabama at Birmingham

             The most fundamental parameter in the definition of a finite mixture distribution is k, the number of component subpopulations. Its importance derives from technical considerations and from its role in subsequent analyses.
             Techniques exist for estimating k. Formal techniques for estimating k involve statistical tests. The Akaike Information Criterion (AIC) has been used in a mixture modeling context to determine the correct value for k. The AIC is used for testing non-nested models with different values of k ( AIC = -2*logL+2*r where r is the number of free model parameters).
              A second test for k, uses the Bayesian Information Criterion (BIC). The BIC is computed from the log likelihood (BIC = -2*log L + r*ln n) where r is as above and n is the sample size. The BIC values are plotted against k for a series of theoretically plausible models. The value of k corresponding to the break in the BIC curve is taken to be the candidate model.
             Finally, the Bayes factor (Bk) has been used to determine the number of subpopulations in the data. The Bayes factor represents the posterior odds that the alternative hypothesis (H1: k > 1) is correct versus a null hypothesis (H0: k = 1) given the data. The larger the value of Bk the more likely the alternative hypothesis is true.

4:00-4:50 p.m. PROFESSIONAL DEVELOPMENT (Training Session) Meeting Room 5

LIFE AFTER THE DISSERTATION: PREPARATION FOR THE ACADEMY AND FIELD

Carol A. Mullen, Richard C. Kunkel, Rhonda Porter, and James E. Witte, Auburn University, and Carolyn Reeves-Kazelskis, University of Southern Mississippi

             A panel of deans, professors, editors, guest editors, published authors, chairs of search committees, and successful job-seekers provided snapshots of issues related to preparation for the academy and field. This training session offered a broad set of interrelated topics centered on the professional/leadership development of graduate students nearing graduation, job-seekers, and beginning faculty. Novice writers learned tips for conducting well-designed research in the context of work assignments and for publishing.
             The subjects covered included the nature of the job market (university), requirements and credentials for academic positions, and networking and other forms of mentoring that promote development, such as organizational involvement through team projects. Publishing issues (i.e., tips for getting started and continuing to publish, strategies for dissemination, venues that include journals, and calls for papers on thematic issues) were shared.
             The topics presented featured input from the different panelists. Participants' questions were encouraged, and papers were made available. Participants were asked to imagine that they are learning how to ride a bicycle for the first time in their life: "Jot down words capturing how you would feel riding a bicycle with training wheels. Now record how you would feel riding a bike for the first time without training wheels." Participants shared their lists. The trainers built on the words shared by the audience (e.g., "scary" and "risky" for being without training wheels, and "reassured" and "supported" for being with training wheels). The message was that novice academics must find "training wheels" to support their professional development. Participants returned to their "bicycle" lists to share what they learned (e.g., new understanding and mentoring tips) that provided them with "training wheels" for the academy and field.