Abstract
(type = abstract)
The quality of public education has become the focus of policymakers, educators and researchers since the launch of Sputnik by the Soviet Union (Bybee & Fuchs, 2006; Rutherford, 1997). In particular, science, technology, engineering and math (STEM) are under great scrutiny (Hill, Corbett, & St Rose, 2010). There are valid concerns that the growing generation may not do well in this globalized world due to lack of skills in STEM. Indeed, research showed that US students lag behind their international peers in science. In addition, most US students lack scientific literacy skills (Grigg et al., 2006), as was apparent in their PISA’s scores (Bybee & McCrae, 2006). There is much emphasis on mastering scientific content in order to increase scientific literacy and students’ international science test scores. This leaves less room for answering questions like “Why don’t students do well in science?” In order to improve science education, the root cause of this problem should be explored. There could be many valid reasons behind this issue. For example, the complexity of scientific phenomena and their intricately interrelated components makes students perceive science as a difficult subject (Aschbacher et al., 2010; Graesser, Singer & Trabasso, 1994). Similarly, many students found science irrelevant to their daily lives (Kadlec, Friedman, & Ott, 2007). Weak science identities and lack of motivation could be another reason. As research showed, students gradually lose their interest in science as they become adolescents (Osborne, Simon, & Collins, 2003). Likewise, students with strong science identities are more likely to participate and succeed in their science classes than students with weak identities (Gresalfi, 2009). Therefore, it is vital to explore students’ science identities and motivation in science in order to understand the root causes of science education problems. Since students’ science identities and motivation play such an important role in their science learning, the present study aimed at investigating high school students’ science identities, expectations of success in science, and values of science. Additionally, it looked into students’ environmental attitudes since environmental issues relate to science learning, and the planned studies of this research were related to the domain of environmental science learning. In order to achieve the study’s goals, a new survey instrument, called SIEVEA, was developed. This instrument was used to collect data so further analyses could be conducted. The data was collected in two various contexts: for a large group of students from multiple school districts and states, and for a smaller group of urban high school students participating in a collaborative online project. The research was made of three studies. Study 1 encompassed the design, development and validation of an instrument called Science Identities, Expectations of Success in Science, Values of Science and Environmental Attitudes (SIEVEA). The developed instrument is a convenient online survey that can be used to measure students’ science identities, expectations of success in science, values of science, and environmental attitudes. Study 1 was made of three sub-studies. In Study 1A, the SIEVEA was designed and used to collect data of 1,911 high school students (grades nine to twelve) from 11 school districts in New Jersey, Pennsylvania and Connecticut. The collected data was used to validate the survey as a measurement instrument. The data was analyzed using both descriptive statistics and exploratory factor analysis (EFA). The EFA results provided useful insights into the factor structure of the data and led to the formation of three candidate models: the two-factor, the three-factor and the four-factor models. All three models were evaluated based on their fit to data, their alignment with the research constructs, and their factor loadings. As a result of this evaluation, the three-factor model was selected as the final model. In Study 1B, the three-factor model was evaluated using partial-confirmatory factor analysis (PCFA) and confirmatory factor analysis (CFA). The PCFA used the original data, whereas the CFA used a newly collected data, which contained survey responses of 1,495 high school students from three schools (urban and suburban) in New Jersey and Connecticut. Additionally, Study 1B conducted reliability and validity tests on the model, including tests for convergent and discriminant validities. In addition, the instrument was tested for measurement invariance. As a result of these analyses and tests, the three-factor model’s selection as the best model was confirmed. Next study, Study 1C, conducted the Rasch analysis on the SIEVEA survey instrument in order to accomplish the following goals: explore the instrument’s psychometric properties and find areas of improvement, conduct additional validation tests on the instrument, and convert the ordinal scores of SIEVEA’s data to interval scale in preparation of conducting parametric statistical tests. Study 2 aimed at measuring and analyzing the three constructs related to student science learning as suggested by the three-factor model: students’ science identities and motivation in science, values of science and attitudes toward the environment. In order to do this study, two data sets collected by SIEVEA were combined into a single sample. This study showed that students’ science subject preferences influenced their science identities and motivation in science. In addition, significant gender-related differences were discovered in students’ science subject preferences, science identities and motivation, and values of science. Males had stronger science identities and motivation in science than females, whereas females ascribed higher value to science. Surprisingly, there were no statistically significant differences between males’ and females’ environmental attitudes. Lastly, statistically significant differences were found between urban and suburban students’ environmental attitudes and science subject preferences. Finally, Study 3 scrutinized how urban high school students’ science identities shifted during their participation in a project based on an online, collaborative learning environment called the River City. There were 8 student participants in this project, which took about two weeks. The project facilitated student learning of scientific inquiry and 21st century skills via a game-based, multi-user virtual environment (MUVE). The study’s results suggested that students’ science identities were not stagnant, but rather that they could change and evolve. However, since this was a short-duration project and due to measurement errors, these results were not conclusive. An additional, longitudinal research is recommended in order to confirm this study’s findings.