Tillotson, R.P. (2006). Graduation Test Review: Student Choice as an Indicator of Self Efficacy. Instructional Technology Monographs 3 (2). Retrieved <insert date>, from http://itm.coe.uga.edu/archives/fall2006/rtillotson.htm.


Graduation Test Review: Student Choice as an Indicator of Self-Efficacy


Robin P. Tillotson
University of Georgia



The purpose of this study was to examine the effects of student efficacy, as demonstrated by student choice of topics, toward reviewing content in an effort to improve performance on the Georgia High School Graduation Test (GHSGT) in science. Two specific questions were (1) Can teachers provide the students with a review for the GHSGT that bolsters their self-efficacy towards the test? and (2) Will a more effective review result from the students choosing their topics and groups rather than being assigned topics and groups?

The intention was to examine the effects of self-efficacy on the GHSGT in science; this study was a pretest-posttest control group design. Participants were eleventh and twelfth grades in the College Preparatory physics classroom groups. Participants were randomly assigned to either the control group, who had their topics assigned to them or the experimental group, who were allowed to choose their topics. The treatment was administered over a two and one half week period and was measured by both the changes in scores from pre-test to post-test and by student perception surveys. There is no evidence indicating that the experimental group showed greater gains from pre-test to post-test than the students in the control group. Responses from the student perception surveys indicate that the students in the control group increased their feelings of confidence and preparedness for the GHSGT by 59.5 % and 44.9% respectively while 52.6% of the students in the experimental group reported greater preparation and 39% increased confidence from participating in the review activity.

This study did not find any evidence indicating differences existed between allowing students choice or not. Further research is needed to explore the effects of giving choice during classroom activities.


Literature Review Methods Results and Discussion Conclusions References



My heart sank when Lola came into my class on the afternoon of the science graduation test and exclaimed, “Mrs. T., I’m sorry, but I ‘Christmas tree-ed’ the science test. I just read the first ten questions and bubbled in the rest.” I knew that Lola had the knowledge and skills to be successful on the test. Why was she so proud of not trying on the test? I believe the theory of self-efficacy, the belief a student has in her own ability to complete the task, can at least partially explain her attitude and her effort on the test (Margolis & McCabe, 2006). How can I use my students’ individual self-efficacies toward science topics tested on the graduation test to provide them with an effective review for the science graduation test?

Over the last few decades, standardized testing has taken command of educational systems worldwide: driving curriculum changes, homogenizing teaching and learning, and quantifying the inadequacies of U. S. students. Because of this testing, the nation focuses its attention on the poor performance of American students on mathematics and science tests (Chen, 2004). At the Covenant College Educators’ Conference, Howard Gardner explained to the audience that he “… often had the opportunity to engage in discussions with the heads of education in many countries,” and that without exception, “educational leaders around the world state that their greatest mission for their country’s educational system is to have the highest scores on standardized tests” (public presentation on February 23, 2006). This statement gives a very clear indication of the importance of the performance of our students on high-stakes tests. Consequently, providing students with the necessary knowledge and skills to be successful on standardized tests is an important aspect of a teacher’s job.

While standardized test scores have been important to parents, teachers, and administrators for decades, these scores take on an even more significant role because of current educational legislation, specifically the Adequate Yearly Progress (AYP) requirement of the No Child Left Behind (NCLB) legislation (No Child Left Behind Act of 2001). According to the Georgia Department of Education website, AYP “is an annual measure of student participation and achievement on statewide assessments and academic indicators” (DOE, What is AYP, 2006). Either a school meets AYP, currently based on the performance of students on language arts and mathematics exit exams, or it does not meet AYP. When a school does not meet AYP, it joins other poor performing schools on a Needs Improvement list. If a school is on the Needs Improvement list for two consecutive years, it moves to the list of Poor Performing Schools. At the present time, AYP determination is based on student performance on the language arts and mathematics high school exit exams.

However, beginning in 2007, science exit exams will also factor into determining AYP. The current law only requires that every state must give a science test by 2007. The legislation does not set standards for the science test; therefore, no one knows exactly how the science test will affect the determination of AYP. Because students consistently perform poorly on the science portion of the Georgia High School Graduation Test (GHSGT) throughout the state, there is much concern over the science scores. Although the metro-area schools perform much better than the state average, our school, City High School (CHS), is on the lower end of the metro-area range, with only 79% of the students passing the test in 2005.

Apparently, the students are not retaining the content from year to year and that the eleventh grade teachers need to strengthen the review for the test. Although each year the students’ performances on these tests receive more and more emphasis, test scores on the science portion of the Georgia High School Graduation Tests (GHSGT) are not significantly improving. While we are beginning to see a small increase in the science scores, the passing rate for first–time test takers in science is still significantly lower than the passing rates for language arts and math. An instructional gap might exist when students perform lower on standardized tests than in their individual science classes. With a failure rate of 21%, it is logical to assume that instructional gaps in science do exist for the students.

Assuming that an instructional gap exists, science educators should ask two important questions: (1) Why are students consistently scoring poorly on the science GHSGT; and (2) What can teachers do to improve their students’ chances of success on the science GHSGT? Although interested in solving the issues inherent in both of the preceding problems, eleventh-grade science teachers should focus on solving the second question since their students are the students who will take the GHSGT during the spring of their junior year. While not responsible for teaching the students all the content tested, these teachers are their students’ last hope for an effective review that could improve their chances of passing the test. Since test scores have not significantly improved using any of the current methods of teaching the content and reviewing the material, high school science educators need a new approach.

As the science department chair, my objective was and continues to be to improve City High School students’ performances on the science portion of the Georgia High School Graduation Test. Because the GHSGT tests the students on biology, chemistry, and physics, we have the daunting task of re-teaching the previous material effectively, while the students are involved in learning physics for the first time. I wanted to explore whether allowing students to exercise their self-efficacies toward science content by choosing their topics is a more effective teaching strategy than assigning the topics to the students based on deficiencies identified by the diagnostic test.

Not only do educators want our students to have the content and skills necessary to succeed on this test, but we also want the test scores to improve in light of many external factors such as home environment, work schedules, substance abuse, and pressures to succeed. One such external factor is the AYP requirement of No Child Left Behind (No Child Left Behind Act of 2001). The hope is that by providing students with a more effective review for the science GHSGT, the instructional gap will shrink and the students will be better prepared to recall the information on the test, therefore improving their test scores.

I began to believe that a review unit incorporating Multiple Intelligences theory (MI) and Problem-Based Learning strategies (PBL) could provide students with a more effective review of the content tested on the GHSGT. Consequently, this more effective review would attribute to greater self-efficacy in my students, building more confidence for their success in taking the science portion of the Georgia High School Graduation Tests. Howard Gardner’s MI theory could provide a pathway for students to learn and retain more content. In this theory, Gardner (1983,1999) postulates that humans possess eight different intelligences , with different ones being more highly developed in different individuals (Gardner, 1983, 1999). Conceivably, a science review unit that incorporated MI theory would provide students with the opportunity to use their more highly developed intelligences while reviewing content for the science graduation test. Following the completion of this unit, the students should retain more content, build their self-efficacy toward science, and, consequently, improve their scores.

A second learning strategy that could increase students’ knowledge and skills is problem-based learning (PBL). In PBL, students must research information to solve a problem and produce a product. Since the person doing the work of learning is the person who retains the most content, the incorporation of PBL into a review of content for the science graduation test should improve students’ knowledge base and retention of information (Jensen, 2000). Again, greater self-efficacy should result, which in turn should result in improved scores on the GHSGT.

I first implemented this MI- and PBL-based Graduation Test Review Project as a pilot test during the fall term of 2005. Gains shown by comparing pre-test to post-test results of the forty-eight students indicated only a 0.9 average increase. As I pondered over why this increase was so slight, a couple of ideas came to mind. First, I administered the post-test on December 14, not an optimal time for adolescent concentration. I have little influence on the concentration abilities of teenagers the week before winter break. However, the second idea really caught my attention. As the final step of the project, the student-teams presented their websites to the class. As I observed my classes, it struck me that the students were paying as little attention to their peers as they do to their teachers when they lecture. Therefore, I decided to change the way in which the students share their websites. During spring term, 2006, the students shared their websites in a Rotating Review museum approach (Kagan, 1992, as cited in Baloche, p. 107). I asked each group to write a scavenger hunt and a quiz to accompany their website. The class then spent two days moving from website to website, solving the scavenger hunt, and completing the quiz. Following these two implementations of the project, the passing rate of the science GHSGT improved from 79% in 2005, to 83% in 2006.

Because “choice is a major motivator” (Pintrich and Schunk, 2002, as cited in Margolis & McCabe, 2006, p. 222), I allowed the students the freedom to choose their topics and their group members. In choosing their own topics, I believed that the students were choosing to work with the content they felt they knew and understood best. This would be a real-life example of self-efficacy. I am now wondering three things: (1) Could giving the students control over topics and group dynamics allow for greater motivation in completing the review project? (2) Could assigning groups and topics based on deficiencies shown on the diagnostic test be more effective? (3) Which of these approaches would influence Lola’s self-efficacy or confidence in her knowledge of the material enough for her to walk into the science GHSGT and know that she can pass it?

Therefore, the current research problem was to determine if a difference exists between allowing the students the power over the make-up of their groups and topics and assigning the groups and topics to the students based on the deficiencies shown on the diagnostic test. For the purposes of this study, students’ choices of topics represent their confidence in their content knowledge indicating the students’ efficacies toward the topics. The real question was which one of these teaching strategies would contribute more to my students’ increased self-efficacies toward the GHSGT in science? Consequently, as we taught the graduation test review unit in our eight College Prep Physics classes during fall term, the physics teachers attempted to quantify a difference between allowing the students to exercise control in choosing their topics and the assignment of topics to the students by the teachers. In half of the classes, the students chose their own groups and topics, while in the other four classes the teachers assigned the group members and topics based on the students’ performances on a diagnostic pre-test.

As a science teacher and as the science department chair, the pressure to raise the passing rate for the science GHSGT is definitely present. All science departments of which I have been a member placed great importance on preparing our students for the science test. We tried many different approaches and learning strategies. Yet each year following the test, I would listen to the testing proctors talk about how quickly the students completed the test, that they completed the test far more quickly than the math or language arts tests. The proctors stated their belief that the students finished the test so quickly because they felt defeated, gave up, and simply bubbled in random answers. This leads me to believe that, for the most part, the students did not believe that they could pass this test; they lacked self-efficacy for this test.

The purpose of this intervention study was to explore whether students’ self-efficacies toward their abilities to pass high-stakes can be influenced by allowing them more control over their learning activities. Since the classes were heterogeneous in make-up, the students’ demographic differences and ability levels for my eleventh grade College Prep Physics students at a large suburban high school in the Southeastern United States represented our school community at large. The independent variable was the teaching strategy we applied to each group: students were given the control to choose their own topics and groups or teachers assigned topics and groups according to previously identified content weaknesses. The belief is that choice of topic indicates the content about which the students feel more confident, i.e. the topics about which they are more self-efficacious. For the purposes of this study, self-efficacy was generally defined as the students’ beliefs in their knowledge of the topics tested on the science GHSGT (Bandura, 1982, Wang, 2001). The dependent variable was the change in achievement exemplified by comparison of the students’ scores on the pre-test and post-test.



Literature Review


The purpose of this study was to investigate allowing students to use their own self-efficacies toward science content in the context of a problem-based learning strategy incorporating multiple intelligences theory in the context a review of the high school science curriculum. The purpose of the review of literature was to place this research in the historical context of MI, PBL, and self-efficacy, to rationalize the scholarly importance of the research problem, to quantify the research variables and to support the methodological choice. Specifically, the review of the related literature is intended to answer these three questions:

  • What is the current research and practice of MI, PBL, and self-efficacy?
  • What are the effects of using MI as an instructional strategy?
  • Can self-efficacy, in the form of student choice, be incorporated into the classroom environment?


Multiple intelligences theory.

First published in Frames of Mind: The Theory of Multiple Intelligences ( Gardner, 1983), and refined in Intelligence Reframed: Multiple Intelligences for the 21 st Century, ( Gardner, 1999), Howard Gardner’s theory of multiple intelligences continues to be an innovative way to consider human intelligence. Rather than defining intelligence as a single score on a test written to perceive intelligence as thinking like a scientist, Gardner defines intelligence as “a biopsychological potential to process information that can be activated in a cultural setting to solve problems or create products that are of value in a culture” (Gardner, 1999, p. 33). To date, Gardner (1999) identifies and defines eight intelligences. The combination of these eight intelligences makes up each person’s overall intelligence. Individuals have some intelligences more highly developed than others. Brief explanations of each of these intelligences are

  • Linguistic intelligence relates to the spoken and written word. Most individuals whose careers involve the spoken and written word have highly developed linguistic intelligence.
  • Logical/Mathematical intelligence relates to a person’s ability to conduct problem-solving activities.
  • Spatial intelligence relates to an individual’s ability to work with patterns.
  • Musical intelligence relates to a person’s affinity for music.
  • Bodily-Kinesthetic intelligence relates to the use of the human body: either whole or in parts. These are our “hands-on” students.
  • Interpersonal intelligence relates to the ability of an individual to interact with others.
  • Intrapersonal intelligence relates to how well a person understands himself/herself.
  • Naturalist intelligence relates to an individual’s ability to see patterns in nature and distinguish between separate species. (Gardner, 1999; Lazear, 2000; Stanford, 2003; & Osciak and Milhem, 2001)

The first two intelligences listed above are the ones most highly valued in our traditional educational system. Therefore, persons exhibiting strong academic prowess are most often those with strong linguistic and logical intelligences. These are also the persons who will score the highest on both paper-and-pencil intelligence tests and standardized multiple-choice tests. The goal of incorporating MI theory into the graduation test review unit is to allow students whose stronger intelligences are not linguistic or logical to use their naturally stronger intelligences as the method to learn the material (Gardner, 1999; Lazear, 2000; Stanford, 2003; & Osciak and Milhem, 2001). This, in turn, should give them the knowledge base to score well on the standardized tests.

Educational trainer David Lazear has taken Gardner’s eight intelligences and provided the classroom teacher with a “toolbox” of activities they can incorporate into the classroom to allow students to use their more dominant intelligences in activities and assessments. These activities provide practical, easy-to-use resources to take the academically written definitions of the intelligences and place them into a user-friendly modality. With this toolbox in hand, an educator can easily design instruction that incorporates activities for all intelligences. Likewise, students may choose from the toolbox activities to complete the tasks given to them by their teachers (Lazear, 2000).

Goodnough (2001) reported on the successes of one science teacher in incorporating MI theory into his classroom using a case study method. The teacher, “Dave,” designed his lessons to include several intelligences at once. He taught the students about the theory and gave them a MI inventory to help them identify their own strongest intelligences. He occasionally gave the students the choice of which intelligence to use to solve a problem. He used learning activities, such as the writing of raps to teach vocabulary (activating the linguistic, logical, rhythmic, and kinesthetic intelligences) and creating skits to demonstrate knowledge (activating the interpersonal and kinesthetic intelligences), to teach a six-week unit on astronomy. During this unit, both Goodnough (2001) and “Dave” reported that the students were more engaged by these activities and seemed to enjoy the work more. They reported, however, that the teacher-made test results did not reflect a significant increase in student performance. On the unit taught using MI theory, the class average was 68% while the previous class average was 64% (Goodnough, 2001). However, the test for the MI unit contained more higher-level thinking skills than the previous test. It would appear that if there was an increase in the class average on a test that included a greater percentage of higher-level thinking skill questions, then a more significant increase exists than just the four percentage points reported. Multiple intelligences theory provides the framework for all students to use their innate strengths in the pursuit of knowledge. By allowing students to play to their strengths, educators are providing all children with the ability to attain the skills necessary to be successful both in the classroom and on standardized tests.

Problem-based learning.  

Problem-based learning (PBL) is an instructional strategy in which students actively resolve complex problems in realistic situations” (Glazer, 2001, paragraph 1). PBL is one of several instructional strategies that offer students the ability to assemble their own learning. These strategies are collectively known as constructivism: the philosophy that gives students the freedom to construct their own understanding (Haney, 2003). PBL provides students with authentic situations or problems to solve by working together in collaborative groups to formulate an answer. These resolutions are most often in the form of some type of physical product. Written as a series of problems that the students need to solve, an example of a textbook that incorporates problem-based learning is the American Chemical Society’s publication, Chemistry in the Community, or ChemCom (Heikkinen, 2002) . For example, the students have to solve the mystery of a large fish kill in a fictitious town, Riverwood. In order to discover the reason behind the fish kill, the students must learn the chemistry content necessary to identify the issues that could affect the fish in this manner (Heikkinen, 2002). Once learned, the students must place this content back into the scenario in order to propose a solution to the problem. This learning strategy is a powerful tool in the hands of an effective educator.

Problem-based instruction originated for the training of Canadian medical personnel and has since evolved to become applicable to a myriad of instructional settings (Gijbels, 2005). PBL shares many characteristics with project-based learning and learning by design (Glazer, 2001). All three of these instructional approaches provide the students with a learning-centered environment with which to solve authentic problems. The basis of these strategies is that students will recognize what information they need to solve the problem (Glazer, 2001). This leads them to assess what information they already know and what information they need to acquire via research. The incorporation of this learning strategy into a review for the science GHSGT will allow the students to focus on reviewing the content necessary for successful completion of the test. By recognizing and differentiating the content they already possess from the content they need to acquire via research, students will be able to review for the GHSGT in a more effective manner. Since each student’s strengths and weaknesses are different, the use of problem-based learning will allow for a more individualized instructional approach.

In problem-based learning, the classroom is no longer a teacher-centered one; the focus of the instruction and the classroom has shifted to become student-centric. The teacher serves as a guide to take the students through the learning cycle, also known as the PBL tutorial process (Hmelo-Silver, 2004). The steps of this process, as outlined by Hmelo-Silver, are

  1. Presentation of the problem.
  2. Analysis of the problem.
  3. Formulation of hypotheses. Identification of gaps in knowledge.
  4. Application of new knowledge. Once the students address the learning issues, conduct research to resolve those deficiencies, and gain new knowledge, the students are ready to apply their new knowledge in evaluating the strengths of each of their hypotheses. Here the students will design, organize, and implement their plans for the graduation test review.
  5. Evaluation of their original hypothesis in light of their new knowledge. (Hmelo-Silver, 2004)

The steps above show in detail the process of using PBL as an instructional strategy to produce a more effective review for the content tested on the science portion of the GHSGT. One can see that while the teacher serves as a guide during the instructional phase of this process, she must produce a problem that is both authentic and challenging--a problem that requires the students to access the appropriate content in order to find a solution. The teacher must also provide the scaffolding necessary to support the students in their research, leading them to the proper information to solve the problem at hand (Lipsomb & West, 2004). Working along with the teacher, the media specialist and technology specialist can also help provide the students with the tools they need to successfully research pertinent information and produce their websites and handouts.

Ideally, problem-based learning allows students to direct and control their own learning, as well as allowing educators to provide more individualized instruction, making the learning more engaging to the learner. This type of instruction provides for greater flexibility in the organization of the classroom. I hope that this will move our classrooms away from the herding mentality. Herding is defined by Prensky (2005) as “students’ involuntary assignment to specific classes or groups, not for their benefit but for ours” (p. 10). With PBL, students should choose their own group members. Not only could students be given the opportunity to chose group members within their own classroom, but also the opportunity to chose partners from other classrooms as well. This way, students from one teacher’s class may collaborate with students in a second teacher’s class in order to produce a better product. With the advent of the Internet and the communications revolution that has ensued, students could meet, communicate, and learn with other students from around the globe (Prensky, 2005).

Although preliminary research studies indicate that problem based-learning “supports students’ self-directed learning” (Moust, 2005, p. 63), at present, few research studies focus on how well heterogeneous groups of high school students are actually able to direct their own content learning. Most of the research studies conducted on problem-based learning activities involve medical students and gifted students (Hmelo-Silver, 2004, p. 249).While many 9-12 educators believe problem-based learning is a possible solution to the ever-lagging student engagement and achievement, most of the research so far does not involve high school (9-12) students. One limitation to our research base is the lack of solid evidence of the effectiveness of PBL in the K-12 setting (Hmelo-Silver, 2004).

Student Choice and Self-efficacy.

Self-efficacy is the belief someone has in her own ability to achieve a goal or to complete a task (Margolis & McCabe, 2006). The student mentioned above did not believe she possessed the knowledge and skills necessary to pass the GHSGT in science. By spending insufficient time answering the questions on the test, the student was exhibiting her feelings that she could not succeed on the test, an indicator of low self-efficacy (Bandura, Barbaranelli, Caprara, & Pastorelli, 1996). Therefore, a change in teaching strategies must occur in an effort to help the students not only gain content knowledge, but also gain self-confidence in their knowledge.

No personal belief is held more strongly “than people’s beliefs in their capabilities to exercise control over”themselves and their situations(Bandura, 1993, p.118). The students themselves will explain their behavior in rushing through the test and these explanations will help them maintain control over the situation. “Students with a high sense of efficacy for accomplishing an educational task will participate more readily work harder, and persist long when they encounter difficulties than those who doubt their capabilities” (Bandura, 1999, p. 204). Tollefson explains that individuals with high self-efficacy will persist at a difficult task longer then individuals with low self-efficacy will. Lola’s lack of persistence and willingness to give up when completing the graduation test is an indicator of her low self-efficacy. In all probability, it is not a lack of ability that causes these feelings in the students, but rather the beliefs students have in their abilities and the way those perceptions affect their emotions (Seifort, 2004).

Therefore, science educators must face the reality that it is not enough merely to present the material. Teaching content in a relevant and meaningful manner will enhance the students’ abilities to internalize the content. One way to achieve this is to allow the students more control in their own learning. According to Tollefson (2002), “Control of the difficulty of the task and the amount of effort needed for a successful achievement outcome is critical to developing outcome and efficacy beliefs that promote achievement,” (p, 69). By giving students a measure of control over their instructional tasks, successes on those tasks should indicate greater content attainment thereby boosting their beliefs in their abilities to complete the tasks. Once the students attain and internalize the content, their self-efficacy toward science will strengthen.

As stated previously, the impetus for this study was the poor performance of students on the GHSGT in science. In 2005, only 68 percent of the eleventh graders throughout the state of Georgia met or exceeded the passing score of five hundred (Report Card, 2005, p.3). With the continued pressure of the No Child Left Behind legislation (No Child Left Behind Act of 2001) and the looming promise of the addition of the science exit exams to the 2007Adequate Yearly Progress determination, the importance of increasing the passing rate of this test should not be underestimated. Not only will there be no escaping from the requirement of passing these tests, but also no escaping the ramifications of these poor performances on schools not qualifying for AYP.

Since the beginning of the GHSGT in science, science educators have tried many avenues to help our students review the content and prepare for the test. However, failure rates remain high. Hence, teachers need to employ methods that are more effective in order to help more students pass the test. The purpose of this study is to determine whether the implementation of a graduation test review that integrates the use of multiple intelligence theory and problem based learning strategies can provide students with the knowledge and skills necessary to be successful on the GHSGT. This research will explore the effects of using a modified instructional model that integrates multiple intelligence theory into problem-based learning unit in order to reach students’ individual intelligence strengths in a situated learning context. Therefore, the objective was to create more effective science review experiences to help students internalize the content at a deeper level, bolster their self-efficacies toward science content and the science GHSGT, and result in increased passing rates for the science graduation.

Summary of Literature Review

The continuing poor performance of our students on the GHSGT in science suggests that a change in teaching strategies is needed. Based on the research, a graduation test review project was written to incorporate the teaching strategies of Multiple Intelligence Theory and Problem-Based Learning. MI theory seeks to allow the students to use their strongest intelligences in both attaining new content and reviewing content previously learned. The integration of PBL strategies in the classroom gives the students an authentic problem, which they must solve. Research indicates that both of these strategies lead to greater gains in student content acquisition and student achievement. Research also shows that when students are given some control over their learning, their feeling of being adequate to complete the work, their self-efficacies, are increased. Therefore, the amalgamation of MI theory, PBL strategies, and increased student control should provide our students with the tools necessary to be successful on the GHSGT in science.





This research study is based on the belief that with the implementation of a graduation test review project that integrates MI theory with PBL, the students would find more relevance and meaning in the content. With this added relevance, the students would make greater connections between science content and real-life situations. These enhanced connections would result in greater acquisition of content. This, in turn, would augment their self-efficacies for both science content and the science graduation test. The evidence of this increased self-efficacy was the increase in time spent actually taking the test and improved passing rates for the test.


If students enjoy the control to choose their own topics during the completion of a problem-based learning review unit for the graduation test in science, their content knowledge is strengthened, their self-efficacies toward science content increases, and their scores on the graduation test improve. For the purposes of this study, the students demonstrated their self-efficacy toward science in the choices of group members and review topics, the students chose the content areas in which the felt the most academically prepared. The evidence of this improvement was the change in their scores from pre-test to post-test.

Site of Research

This study was conducted at a large, suburban high school, which serves 3000 students. For the purposes of this research, the high school was known as City High School. City High School is located in the suburbs of a major metropolitan area in the southeastern United States. City is on a 4 x 4-block schedule. On a 4 x 4-block schedule, students receive an entire year’s instruction during one semester. Each class period, or block, is 93 minutes, with students attending four classes per day. Tables 1 and 2 illustrate the 2005 demographics for this high school.

Table 4.1 2005 Demographics of City High School delineated by race


Percentage in City High School

Asian/Pacific Islander


Black/African American








Table 4.2 2005 Demographics of City High School based on socioeconomic groups

Demographic Group

Percentage in City High School

Free/Reduced Lunch


Limited English Proficiency


Special Education


The students have previously taken and passed the coursework necessary to be successful on the GHSGT. However, traditional methods of instruction--lectures, laboratory activities, and homework assignments--have not resulted in the students’ abilities to pass the GHSGT on their initial attempts. Traditional review strategies have resulted in few gains in students’ performances on the GHSGT. The Table 3 illustrates the results of the 2005 administration of the science GHSGT.

Table4. 3 Results of the 2005 Administration of the science GHSGT

Science Test Ability Level

2005 Results

Overall passing percentage


Fail percentage


Number of students tested


Number of students failing




This research study involved approximately 158 students in eight college prep physics classes with four different teachers. The study included all students taking College Prep Physics during the fall term of the 2006-2007 school year. All the students were on the College Prep Diploma track, meaning that they were planning to go straight from high school to some form of college. The College Prep Physics classes were heterogeneous in make-up both in demographic terms and in ability levels. The classes included both eleventh graders who have not previously attempted the GHSGT and twelfth graders who have attempted the test at least once.

The classroom groups of students were convenience groups because they were already intact groups, i.e. the physics classes (Creswell, 2003). The four classes that comprised the control group are Mrs. Alpha’s two classes, one of Dr. Omega’s classes and my own class. The other four classes made up the experimental group. These were Mr. Gamma’s two classes, Dr. Omega’s second classes, and Mrs. Delta’s one class. Four of the eight classes, and approximately half of the students, comprised the control group while the other half made up the experimental group. Since this study tested the effects of student choice of science content on student success on the science GHSGT, the teachers assigned the control group members to groups and assigned their topics. For the experimental group, the physics teachers allowed the students to choose both their groups members and their topics.

The tables below present the current demographics of City High School as determined on September 1, 2006. Table 4 below displays the demographic for race and ethnicity, while Table 5 exhibits other non-racial demographics.

Table 4.4 2006 Demographics of City High School delineated by race


Percentage in City

High School

Asian/Pacific Islander


Black/African American






Native American/Alaskan Native




 Table 4.5 Demographics of City High School based on socioeconomic groups


Percentage in City

High School

Free/Reduced Lunch


Limited English Proficiency


Special Education


Gifted Students



The independent variable for this study was student choice as an indicator of the students’ self-efficacies toward science content. The students fell into two groups. The delineation of these groups was based on students being allowed to choose their topics (experimental group) and students being assigned to certain topics based on deficiencies that appear on a diagnostic pre-test (control group). The dependent variable was the difference in the scores from pre-test to post-test. Again, the hypothesis was that if students choose their topics, rather than having the topics assigned to them, then the difference between their pre-test and post-test scores would be greater. An improvement from pre-test to post-test should indicate an increased potential for the students’ successes on the science GHSGT.  

Research Procedure

An Individual Review Board Authorization (IRB) form was completed and filed with the University, prior to the beginning of the study. It is the responsibility of the IRB to protect the rights and welfare of human subjects involved in research (Office of the Vice President for Research, UGA website, 2006). The local school system in which the study occurred also required the filing of an IRB prior to the commencement of the study. Once permission to research was granted, the teachers and students began to work on the project.

The teachers asked their students to complete a Graduation Test Review project, from which the final product was review websites. The teachers administered a diagnostic test to the students as a beginning point for this project. They also surveyed the students to determine the concepts tested on the GHSGT with which the students feel most comfortable and least comfortable. Approximately half of the students chose their group members and content area while the other half had their groups and topics assigned to them based on deficiencies shown on the diagnostic pre-test and the students’ own perceived comfort levels.

In an effort to minimize the bias, all the classes for each physics teacher, except Dr. Omega, were part of the same research group. For example, both of Mrs. Alpha’s classes were members of the control group while all of Mr. Gamma’s classes were members of the experimental group. Since each teacher only interacted with one type of group, the likelihood of bias should decrease. In a continuing effort to alleviate bias, every classroom followed the same procedure. The project procedure follows.

  • All students took the pre-test and completed the initial self-assessment survey.
  • All members of the experimental group chose their own group members and topics.
  • All members of the control group were assigned to their groups and topics.
  • All students participated in the project in the same basic manner.
    • pre-test and initial comfort survey
    • topic acquisition
    • research
    • website design and production
    • worksheet and quiz production
    • website sharing
    • post-test
  • All students took the post-test.

Once the data was collected, t-tests were used to determine if the results showed a statistical significance. “The t-test is used to determine whether significant differences exist between means (Cothran et al., 1998, p. 140).

I filled the role of a participant-observer and recorded the students’ reactions, facial expressions, comments, and apparent levels of motivation. Following the completion of the project, all the students in the eight classes of College Prep Physics completed a student perception survey. The other physics teachers also completed a survey concerning their observations and perceptions about the project and the students’ motivation during the implementation phase.

The study lasted six class days. During the first four days, students researched their content area, designed, and produced their websites, handouts, and quizzes. Because each student group produced a topic-specific website, the remaining two days of project time allowed for the sharing of the websites with the other members of the class. Once the web sites were completed, the teachers published them on the internet to have them available for the students to use later in the year, for students to show their parents, and as a community service to provide review materials for other students who were not involved in the project.

In this project, the student groups produced websites to help other students review for the science GHSGT. Because of a lack of website-producing software, the websites were simply converted PowerPoint presentations of the material. To make more effective review materials, the websites included at least three different multiple intelligences, contained a worksheet to accompany the site, and a practice quiz on the information. The project required the students to chose to incorporate any three of the eight multiple intelligences into their presentation. This could be as simple as including written information (verbal/linguistic), pictures and diagrams (visual, spatial), or patterns found in nature (naturalist). The following screen shots shows two example project pages. (See Figure 4.1 and Figure 4.2)

Example Web Page

Figure 4.1 Example web page

Example Web Page

Figure 4.2 Example web page

In order to accomplish this task, the students required ready access to the mobile laptop carts, the Internet, print texts, and printers. Once completed, the students shared their websites with their classmates via a museum/conference set up. Computers were set up around the room with the students’ websites and the printed handouts available at each station. The students moved throughout the room exploring each of the websites, completing the worksheets and taking the practice quizzes. An evaluation rubric and their scores on the practice quizzes comprised the final grades for the students.

At the beginning of the project, the physics teachers administered a diagnostic pre-test to their students. They also asked their students to rank the nine topics of the project in decreasing order of comfort with the material. Based on the pre-test and the self-assessment, Mrs. Alpha and Mrs. Beta assigned their students to groups and topics. Dr. Omega also assigned groups and topics in one of her two classes. These assignments were based on the topics for which the students showed a deficiency and/or with which they indicated a lack of confidence; this represented the control group. The other teachers allowed their students to decide the make up and topic of their groups based on student choice; this was the experimental group.

A combination of student surveys, teacher surveys, and personal discussions with other physics teachers, participant observation, and the change in student scores from pre-test to post-test, and a personal log was used to produce data. The teacher survey is located in Appendix B. Both surveys included six questions to which the participants responded. The teacher survey also included four open-ended, free-response questions to allow them to give additional information to the researcher. Students were observed in other physics classes as well as my own students. Using this combination of data-gathering techniques gave me a more complete view of both student motivation and how motivation affects the students’ products.

At the conclusion of the project, the teachers repeated the procedures from the start of the project. The teachers administered a post-test to their students. If a statistically significant difference between the pre-test and post-test was found to exist, then a difference between the achievements of the two groups must also exist.


Parker (1993) defines internal validity as “the extent to which extraneous variables are controlled” (p. 130). At least two forms of internal validity exist for this study: history, or outside events that affect the results of the study (Griffee, 2004; Parker, 1993) and testing (Parker, 1993). In this study, history might refer to any conversations between students working on the same review topics in different teachers’ classrooms. These conversations could result in the students progressing at a more rapid pace than could occur in the isolated classrooms alone; ideas from one group could lead to new discoveries in the other groups that would not have occurred otherwise. The second threat to internal validity in this study is the incorporation of the pre-and post-tests. Two different situations could occur from this method of testing. Either the students remember their answers to the pre-test and therefore score better because of their memories or they may become so tired of testing that they may totally disregard the post-test, actually scoring lower than before (Parker, 1993). Therefore, if the students’ post-test scores are lower than their pre-test scores, one could infer that they were rebelling against taking another test.

“External validity threats arise when experimenters draw incorrect inferences from the sample data to other persons, other settings, and past or future situations” (Creswell, 2003, p. 171). Since the sample size is approximately one hundred fifty students, any results from this research study may be generalizable to other eleventh grade populations.


As with any research study, limitations of the research existed. “Bias may be activated by personal infatuation with any current innovation and the strong belief that this teaching is effective” (Griffee, 2004, p. 2). My personal biases existed in the hope for a positive result and in the years of experience I have as a classroom teacher. I wanted the students to excel at this project just as I continue to want them to be successful on the GHSGT. Since this study did not produce evidence that allowing for student choice made a significant difference in the achievement of the students, more research is needed to explore this important topic.

Upon reflection on the procedure and outcome of this project, I wondered if the classroom teacher made a difference in the outcomes of the study. Could the different teachers had an unknowing impact of the students’ performances on the post-test in particular? This is also a direction for further investigation.



Results and Discussion

Three types of data in this study include (1) changes in scores from the pre-test to the post-test, (2) student and teacher surveys, and (3) researcher observations. Each of these modalities was designed in order to gather enough information to form a more complete picture of the results of this project.

Performance Tests

Prior to analyzing the data, Mr. Gamma discovered that his post-test data was tainted. He left the post-test as an assignment for the students while a substitute was in charge of the class. Upon his return, the students told him that they believed the post-test was a mistake since they had taken the pre-test. Therefore, the data from his two classes were excluded from the analysis. That lowered the number in the experimental group from 158 to 79 students.

To determine if the differences between the students’ pre-test scores and their post-test scores were the result of the experimental treatment, t-tests were conducted. Table 4.1 shows the means and standard deviation for each group. The mean pretest score for the experimental group was 27.3 answers correct while the mean pretest score the control group was 27.41 answers correct. The experimental group’s mean post-test was 26.1 as compared to the control group’s mean post-test score of 27.89. According to the gains between pre-test and post-test, the control group outperformed the experimental group 0 .42 and -1.26 respectively.

Table 4.6 Statistics on Means and Standard Deviations for both groups.



Standard Deviation










Experimental Group








Control Group
















A T-test was conducted to compare the mean difference between the experimental and control groups. The T-test on the group difference is t (116) = 1.98, p = 0.17, which is not significant at the 0.05 level. Therefore, there is no evidence to indicate that the two groups differ in their performance on the test.


All participants in this research were asked to complete a survey about their experiences. One hundred fifty-eight students returned their completed survey forms, which represents 100 % of the population of these eight classes. Because Mr. Gamma’s surveys were completed on a different date than the post-test, there was no evidence that the survey results were spoiled. Therefore, the survey results from all students were analyzed. The student survey can be found in Appendix C. The four of the five teachers, excluding myself, completed teacher surveys.

Student surveys.

Student surveys were examined in two ways. First each student’s responses to the survey questions were cataloged. Means, standard deviations, and t-tests were conducted on this data in an effort to quantify student perceptions. Table 4.3 and 4.4 below illustrate the results from the individual student data.

Table 4.2 displays the means and standard deviations for each of the six survey questions. The Qq value for the control group was 3.52 while the Qq value for the experimental group was 3.57. Based on this analysis, there is no evidence to support an appreciable difference between the control group and the experimental group.

Table 4.7 Statistics on Means and Standard Deviations for both groups on the Student Survey.


Student Responses to Survey Questions





St. Error Mean


Experimental Group






Control Group








The T-test on the group difference is t (154) = -.389, which is not significant at the 0.05 level. Therefore, there is no evidence to indicate that the two groups differ in their performance on the test

The student survey data were also analyzed based on the responses levels for each of the six questions. While the analysis of each individual students’ responses to the questions do not provide evidence of a significant difference between the two groups, an analysis of the total number of responses for each scale level for each question give a slightly different picture.

In Survey Question One, student responded to the question of their level of enjoyment in participating in the project. According to the data, the choice group enjoyed this project more than the assigned group did. Fifty-one of the students in the experimental group agreed or strongly agreed with this question while only thirty-seven of the control group students felt the same way. The most interesting statistic for this question is that thirty-one of the seventy-nine students surveyed in the control group had no opinion about their enjoyment of the project.

Survey Question Two queried the students concerning their feelings of being prepared to take the GHSGT. Both groups indicated feelings of preparedness for the GHSGT. Forty-seven of the control group students and 41 of the experimental group students answered with either agree or strongly agree. This data indicates that the students’ confidence levels for success on the GHSGT were high.

For Question Three the students were asked to indicate feelings of improvement in areas where they originally felt weak. More of the control group students, those students assigned to the topics by their teachers, felt that they were stronger in theses areas than the experimental group students. This data indicates that students assigned to their topics had greater feelings of improvement over students who chose their topics.

Student responses to Survey Questions Four indicated that the majority of the students in both the control and experimental groups felt that their website would be beneficial to themselves and to others as the time for the GHSGT approaches. Sixty-one students in the control group and 65 of the students in the experimental group answered with either agree or strongly agree response.

The students’ responses to Survey Question Five provided evidence that they were proud of the work they completed and of the product, their website, regardless of choosing their topic or being assigned to it.

In answering Survey Question Six the students exhibited how their confidence levels had increased from having completed the project. The data shows that the students in the control group felt more confident to succeed on the GHSGT than did the students in the experimental group.

According to the findings, four positive themes emerge.

  • The students enjoyed the project. Eighty-eight students out of the 158 students surveyed indicated this sentiment: 37 or 46.8% of the control group and 51 or 64.6 % of the experimental group.
  • The students feel prepared for the administration of the GHSGT. Forty-seven students from the control group and 41 students from the experimental totally 88 students indicated this feeling.
  • The students feel their websites are beneficial to themselves and to others. One hundred twenty-six students, 61 students from the control group and 65 students from the experimental group, responded with a positive reaction to this question.
  • The students are proud of their work and their product. Fifty-eight students from the control group and 61 students from the experimental group responded in this manner.

Teacher survey.

The responses from the teacher survey questionnaire indicated that the other teachers involved in this project felt positively about the project and their students’ successes. The open-ended responses from the teachers review four major issues:

  • Students saw value in the activity which resulting in the students being more serious and dedicated to completing the project.
  • The cooperative learning environment was a beneficial aspect of the project.
  • Technology was necessary to the project but also the most negative aspect of the project.
  • The review unit should come later in the term.


 As a participant-observer, I moved into and out of each of the eight CP Physics classes while the students were working on this project. I found that virtually all students were engaged in the task and interested in creating a good product. The students communicated well in each group whether they chose the group members and topics or their teachers chose the members and topics.

The students in the control group were disappointed to discover that their group members and topics were assigned to them. However, they quickly met with one another, used their self-efficacies to divide the work up among them, and went to work. It was clear during the observations that the students chose to work on tasks within the groups with which they were most comfortable. The students with the greater content knowledge and research skills became responsible for the content, the more artistic students took on the tasks of designing the website, and the technology-savvy students produced the PowerPoint presentations and converted them to WebPages. In one group, Mark became the webpage designer and began working on the page almost immediately. Sally began researching the topic on the internet, while Matthew discussed the content specific objectives for the project with his teacher. In another group, Kylie began researching material, while her partners looked on. At first glance, it seemed that the other two students were not on task. However, upon further inspection that they were researching together, and then Emma wrote the worksheet and Bob created the quiz.




This study did not find any evidence indicating that differences exist between the control group and the experimental group in terms of performance on the post-test. Based on the data, allowing the students to use their own beliefs in their competencies in a given content area, their self-efficacies, to choose their topics and group members did not result in greater gains from pre-test to post-test. In this instance, Parker’s (1993) thoughts about the students’ performance on pre-tests and post-tests may be correct; at least some of our students may have grown so weary of testing that they completely disregarded the post-test and actually scored lower than before. Why was the incidence of this greater in the experimental group than in the control group?

The student survey results give a different picture of the students’ feelings and attitudes. While the majority of students responded favorably to all the questions on the student survey, two questions provided responses most applicable to the concept of self-efficacy. The self-confidence levels of the experimental group compared to the control group led to very different responses to questions three and six. The students in the control group felt that they had improved in their weaker areas and that they were more confident going into the graduation test than the students who chose their own groups felt. These differences correlate to the differences between the control group and the experimental group in their performances on the post-test. The greater feelings of improvement and confidence resulted in the students in the control group being less likely to disregard the post-test than the students in the experimental group were. Not only could these increased feelings of self-confidence translate into greater self-efficacy in taking and succeeding on the post-test but also to taking and succeeding on the GHSGT. This increased self-efficacy toward the GHSGT could mean that the students will be more persistent in completing the test and less willing to surrender when they encounter difficult questions on the test, resulting in increased passing rates on the GHSGT. As this study provided no evidence that allowing students the freedom to choose their own cooperative groups made any significant difference in their achievement on a problem-based learning project, more research is needed.




Baloche, L. A. (1998). The cooperative classroom: Empowering learning. Upper Saddle River, NJ: Prentice-Hall, Inc.

Bandura, A. (Ed.). (1999). Self-Efficacy in changing societies. Cambridge, UK: Cambridge University Press.

Bandura, A. (1993). Perceived self-efficacy in cognitive development and functioning. Educational Psychologies, 29(2), 117-149.

Bandura, A. (1982). Self-Efficacy mechanism in human agency. American Psychologist. [Electronic version]. 37(2), 122-147.

Bandura, A., Barbaranelli, C., Caprara, G.V., & Pastorelli, C. (1996). Multifaceted impact of self-efficacy beliefs on academic functioning. Child Development, 67, 1206- 1222.

Brophy, J. (1999). Toward a model of the value aspects of motivation in education: Developing appreciation for particular learning domains and activities. Educational Psychologist, 34, 75.

Cothron, J.H., Giese, R.N., & Rezba, R.J. (2000). Students and research: Practical strategies for science classrooms and competitions (3 rd ed.). Dubuque: Kendall/Hunt Publishing Company.

Chen, L., Benton B., Cicatelli E., and Yee, L. (2004, May/Jun ). Designing and implementing technology collaboration projects: Lessons learned. [Electronic version]. TechTrends: Linking Research & Practice to Improve Learning, 48.

GADOE.org (n.d.). What is AYP? No Child Left Behind Adequate Yearly Progress (AYP)-The Basics. Retrieved: March 18, 2006, from Georgia Department of Education Web site: http://www.doe.k12.ga.us/support/plan/nclb /ayp_faq_q7.asp

Gardner, H. (1999). Intelligence reframed: multiple intelligences for the 21 st century. New York, NY: Basic Books.

Gardner, H. (2006, February 23). Changing minds: the art and science of changing our own and other people’s minds. Public presentation at the 2006 Covenant College Educator’s Conference. February 23, 2006. Lookout Mountain, GA.

Gijbels, D., van de Watering, G., & Dochy, F. (2005, February). Integrating assessment tasks in a problem-based learning environment. [Electronic version]. Assessment & Evaluation in Higher Education, 30.

Glazer, E. (2001). Instructional models for problem-based inquiry. In M. Orey (Ed.), Emerging perspectives on learning, teaching, and technology. Retrieved: April 29, 2006 from http://www.coe.uga.edu/epltt/ProblemBasedInstruct.htm

Goodnough, K. (2001, April). Multiple intelligences: a framework for personalizing science curricula. [Electronic version]. School Science & Mathematics. 101(4).

Governor’s Office of Student Achievement. (2005) 2004-2005 State of Georgia K-12 Report Card for State of Georgia in State of Georgia. Page 3. Retrieved April 29, 2006, from http://reportcard2005.gaosa.org/k12/PDF/OSA-K12_ALL-ALL.pdf

Griffee, D.T. (2004). Research tips: Validity and history. [Electronic version]. Journal of Developmental Education, 28(1), 38.

Haney, J. J., Czerniak, C. M., & A. T. (2003, December). Constructivist beliefs about the science classroom learning environment: perspectives from teachers, administrators, parents, community members, and students. [Electronic version]. School Science and Mathematics, 103.

Heikkinen, H. (Ed.). (2002). Chemistry in the community. 4 ed. New York: W. H. Freeman.

Hmelo-Silver, C. E. (2004, Sept). Problem-based learning: what and how do students learn? [Electronic version]. Educational Psychology Review.16(3).

Jensen, E. (2000) Brain-based learning. San Diego, CA: The Brain Store.

Lazear, D. (1999). Eight ways of knowing: teaching for multiple intelligences. 3 rd edition. Arlington Heights, IL: Skylight Training and Publishing.

Lipscomb, L., Swanson, J., West, A. (2004). Scaffolding. In M. Orey (Ed.), Emerging perspectives on learning, teaching, and technology. Retrieved February 18, 2006 from: http://www.coe.uga.edu/epltt/scaffolding.htm

Margolis, H. & McCabe, P.P. (2006) Improving self-efficacy and motivation: What to do, what to say. [Electronic version]. Intervention in School and Clinic. 41 (4), 218-227.

Moust, J., Roebertsen, H., Savelberg, H., & De Rijk, A. (2005, March) Revitalising PBL groups: Evaluating PBL with study teams. [Electronic version]. Education for Health.18(1).

No Child Left Behind Act of 2001, 20 U.S.C. § 6301 (2002).

Osciak, S. Y., & Milheim, W. D. (2001). Multiple intelligences and the design of web- based Instruction. [Electronic version]. International Journal of Instructional Media.28(4).

Parker, R.M. (1993) Threats to the validity of research. [Electronic version]. Rehabilitation Counseling Bulletin, 36(3), 130.

Placek, R. (2003). Stats without stress. (Class handout available from members of the honors classes at City High School, 1555 Old Peachtree Road, Suwanee, GA 30024)

Prensky, M. (2006). Listen to the natives. Educational Leadership, 63. Rumsey, D. (2003). [Electronic version]. Statistics for Dummies. Hoboken, NJ: Wiley Publishing, Inc.

Rumsey, D. (2003). Statistics for Dummies. Hoboken, NJ: Wiley Publishing, Inc.

Seifert, T. (2004) Understanding student motivation. [Electronic version]. Educational Research, 46(2), 137- 149.

Stanford, P. (2003, November).Multiple Intelligence for every classroom. [Electronic version]. Intervention in School and Clinic.39(2).

The Office of the Vice President for Research, (n.d.). The University of Georgia Website. Retrieved July 1:2006 from: http://www.ovpr.uga.edu/hso/guidelines/index.html

Tollefson, N. (2000). Classroom application of cognitive theories of motivation. [Electronic version]. Educational Psychology Review, 12.

Wang, S. (2001). Motivation: General overview of theories. In M. Orey (Ed.), Emerging perspectives on learning, teaching, and technology. Available Website: http://www.coe.uga.edu/epltt/Motivation.htm

What is ayp?. (n.d.). Retrieved March 18, 2006, from Georgia Department of Education Website: http://www.doe.k12.ga.us/support/plan/nclb/ayp_fq_q7.asp




Appendix A

Student Survey

Please complete the following survey concerning the Graduation Test Review Project you just completed









No opinion





1. I enjoyed this project.


2. I feel this project has helped to prepare me for the graduation test.


3. I feel that I am stronger in the areas of content in which I originally felt weak.


4. I feel our website will be beneficial to our groups and to others who wish to review for the graduation test.


5. I am proud of our work on the website.


6. My confidence has increased in regards to taking the graduation test.



Appendix B

Teacher Survey

Please complete the following survey concerning the Graduation Test Review Project your students just completed.









No opinion





1. My students focused on the task of research information to build an informative website.


2. My students appeared engaged in their work.


3. My students appeared to gain confidence in their knowledge of the topics.


4. My students appeared to be proud of their websites.



5. I feel that this time was beneficial in preparing the students for the graduation test.


6. My students appear to feel confident about taking and succeeding on the graduation test.


Free-Response Questions:

If I had to name the one most positive aspect of this project, it would be:

If I had to name the one most negative aspect of this project, it would be:

I would change this project by:

I feel my students have grown in the following ways because of this project.