My fourth research question involved associations between students’ perceptions of six aspects of classroom environment (Student Cohesiveness, Teacher Support, Involvement, Personal Relevance, Critical Voice, Student Negotiation) and two attitude scales (Attitude to Scientific Inquiry, Enjoyment of Science Lessons). Data from my sample of 1,396 students were used. This section reports results for associations between students’ perceptions of their learning environment and attitudes.
To investigate the relationships between students’ perceptions of their learning environment and student attitudes, simple correlation and multiple regression analyses were conducted. Simple correlations (r) described the bivariate relationship between each student attitude scale and each of the six learning environment scales from IBLES. Multiple regression analysis was conducted to determine the joint influence of the set of correlated learning environment scales on each attitude scale. The multiple correlation (R) was used to describe the multivariate association between an attitude scale and the set of all learning environment scales. Standardized regression coefficients (β) were used to provide information about which environment scales contributed significantly to the variance in students’ attitudes when all other environment scales were mutually controlled.
Table 4.10 provides the simple correlation between each learning environment scale and each student attitude scale. All of the six IBLES scales showed a statistically significant correlation with Attitude to Scientific Inquiry. For Attitudes to Scientific Inquiry, correlations ranged from 0.08 (Student Cohesiveness) to 0.13 (Student Negotiation). For Enjoyment of Science, correlations were statistically significant for every learning environment scale except Critical Voice. Significant correlations ranged from 0.11 (Student Cohesiveness) to 0.29 (Teacher Support). Higher learning environment scores were linked to higher Inquiry and Enjoyment scores.
For each attitude scale, the multiple correlation with the set of six environment scales was statistically significant. The multiple correlation for the sets of IBLES scales and Attitudes to Science Inquiry was 0.16 and for Enjoyment of Science was 0.34. Inspection of the regression coefficients revealed that:
• Student Negotiation was a positive independent predictor of Attitude to Scientific Inquiry.
• Each of the six WIHIC and CLES scales was a positive independent predictor of Enjoyment of Science.
Interestingly, all statistically significant univariate and multivariate associations in Table 4.10 are positive, suggesting that there was a positive relationship between environment scales and students’ attitudes. These results replicate considerable past research that provides convincing evidence that classroom environment is a strong determinant of students’ attitudes towards science (Fraser, 2012).
Table 4.10 Simple Correlation and Multiple Regression Analyses for Associations between Learning Environment and Attitude Scales
Scale Attitude–Environment Association
Attitude to Inquiry Enjoyment of Science
r β r β IBLES Student Cohesiveness 0.08** 0.00 0.11** 0.07* Teacher Support 0.10** 0.04 0.29** 0.20** Involvement 0.10** 0.01 0.25** 0.12** Personal Relevance 0.11** 0.04 0.19** 0.08** Critical Voice 0.09** 0.04 0.04 0.09** Student Negotiation 0.13** 0.07* 0.22** 0.12** Multiple Correlation R 0.16** 0.34** *p<0.05, ** p<0.01
Figure 4.3 Classroom Environment Scales that Contributed Uniquely and
Significantly to Variance in Two Attitude Scales Using Standardized Regression Coefficients (β)
Figure 4.3 is a visual representation of the of environment scales that contributed significantly and uniquely to variation in students’ attitudes. Beta weights indicated that all environment scales contributed significantly and independently to the variance in Enjoyment of Science Lessons, with beta weights ranging from 0.07 for Student Cohesiveness to 0.20 for Teacher Support. In contrast, Student Negotiation (β=0.07) was the only environment scale that contributed significantly and independently to Attitude to Scientific Inquiry. This results support the view that students’ attitudes to science learning are closely associated with their learning environment perceptions.
Overall, the results of data analysis reported in this section suggest that there is a statistically significant association between students’ perceptions of the classroom learning environment and their attitudes towards science learning. My findings are consistent with those obtained in past studies which showed positive and statistically significant relationships between students’ attitudes and their classroom environment perceptions (Allen & Fraser, 2007; Fraser, Aldridge & Adolphe, 2010; Kim, Fisher
Attitude to Science Inquiry
Student Cohesiv eness Teacher Support Invo lvem ent Persona l Relevan ce Studen ts Negoti ation
Enjoyment of Science Learning
β= 0.07 β= 0.12 β= 0.0 9 β= 0.08 β= 0.12 β= 0.20 β= 0.07 Uncerta inty of Science
& Fraser, 2000; Martin-Dunlop & Fraser, 2008; Ogbuehi & Fraser, 2007; Quek, Wong & Fraser, 2005a; Wolf & Fraser, 2008).
4.5 Chapter Summary
This chapter focused on providing answers to my four research questions which were discussed in Chapters 1 and 3. It described the analyses of IBLES and TOSRA data from 1,396 high school students in 35 biology and earth science classrooms. The research questions are: firstly, is it possible to develop valid and reliable measures of science students’ perceptions of the classroom learning environment and their attitudes towards science? Secondly, is inquiry-based instruction effective in terms of students’ perceptions of learning environment and attitudes towards science? Thirdly, is inquiry-based instruction differentially effective for male and female students in terms of perceptions of learning environment and attitudes towards science? Fourthly, are there associations between students’ perceptions of inquiry- based learning environments and their attitudes towards science?
This chapter began with a report of the analyses for research question 1: the validity and reliability of IBLES and TOSRA scales. The IBLES consists of six scales (three scales selected from WIHIC and three scales from CLES).Three TOSRA scales were used in the initial version of the attitude questionnaire. To examine the validity and reliability of the learning environment and attitude scales, principal axis factoring followed by varimax rotation was used to determine the factor structure. Also the internal consistency reliability was calculated for each scale with individual student as the unit of analysis. Factor analysis supported the three-scale structure of the WIHIC and the three-factor structure of CLES, but a two-scale structure for the TOSRA.
The total amount of variance accounted for was 58.12% for the WIHIC, 55.1% for the CLES, and 44.06% for TOSRA. Eigenvalues for the WIHIC scales ranged from 1.74 (Student Cohesiveness) to 9.37 (Teacher Support), and for the CLES scales ranged from 1.71 (Personal Relevance) to 5.77 (Student Negotiation). The eigenvalue for TOSRA was 2.15 for Science Inquiry and 4.89 for Enjoyment of Science.
The internal consistency reliability was determined using the Cronbach alpha reliability coefficient with the individual student as unit of analysis. The alpha coefficient for the 6 scales of IBLES ranged from 0.78 (Personal Relevance) to 0.92 (Teacher Support) and for the 2 scales of TOSRA was 0.67 (Scientific Inquiry) to 0.87 (Enjoyment of Science). The highest alpha reliability was obtained for the Teacher Support and the lowest for the scale Scientific Inquiry. These results compare favorably with those obtained from previous studies reported in Table 4.5 for the WIHIC and in Table 4.6 for the CLES scales, as well as in previous studies with TOSRA (Fraser et al., 2010; Wong & Fraser, 1996).
Next, the chapter reported the effectiveness of inquiry-based instruction in terms of students’ perceptions and attitudes. Data from 1,396 students were analyzed using a two-way MANOVA with instructional method (inquiry-based classrooms vs. non- inquiry-based classrooms) and student sex as the independent variables. Using Wilks’ lambda criterion, MANOVA revealed significant results for instructional method and sex for the set of dependent variables as a whole, and therefore the two- way ANOVA results were interpreted separately for each learning environment and attitude scale. Statistically significant results emerged for: method of instruction for every learning environment scale and both attitude scales; and for sex for three learning environment scales but not for attitudes. Although instructional-method differences were statistically significant for all scales, effect sizes were moderate or large only for Involvement (0.63 SDs) and Teacher Support (0.93 SDs) according to Cohen’s (1988) criteria.
Sex differences were statistically significant for three environment scales (Student Cohesiveness, Teacher Support and Critical Voice) for which effect sizes were small and ranged from 0.15 to 0.31 standard deviations. Interestingly, the results show that females held somewhat more favorable perceptions than males for these three scales.
Two-way MANOVA was used to determine if there were differences in the effectiveness of using inquiry-based learning for males and females. The results showed that the instruction−by−sex interaction was statistically nonsignificant for every learning environment and attitude scale and, therefore, the use of inquiry-based learning was not differentially effective for males and females for any scale.
Next, associations between students’ perceptions of their classroom learning environment and their attitudes toward science were reported. Simple correlation and multiple regression analyses were used to explore associations between the learning environment scales and each student attitude scales. Simple correlation analysis revealed that all learning environment scales significantly correlated with Science Inquiry and that, for the Enjoyment of Science scales, all the environment scales except Critical Voice showed a positive and statistically significant correlation. Standardized regression coefficients (β) were used to provide information about which environment scales contributed significantly to the variance in students’ attitudes when all other environment scales were mutually controlled. Beta weights showed that only Student Negotiation scale was statistically, significantly, and independently related to Attitude to Science Inquiry. For the Enjoyment of Science scale, all of the six environment scales were statistically significantly and independently related to Enjoyment of Science Lesson. These results suggested that improved student attitudes are associated with more emphasis on the aspects of learning environment assessed in this study.
Chapter 5 summarizes the thesis and draws on previous chapters to present an overview, as well as a detailed discussion, of the major findings obtained. It describes the implications of the research and the effectiveness of inquiry-based learning in Californian public schools education system. The chapter concludes with consideration of the limitations of the present study and provides recommendations for subsequent studies on inquiry-based practices in science classrooms.
Chapter 5
DISCUSSION AND CONCLUSION
5.1 Introduction
This chapter concludes this thesis which reports an evaluation of inquiry-based learning among 1,396 high-school biology and earth science students from 35 classes in Los Angeles County public schools. My study focused mainly on how students’ attitudes and perceptions of their classroom environments differ according to instructional method (inquiry and non-inquiry-based learning) and student sex.
The State of California and Local Education Agencies (LEA) have adopted new science standards designed on the premise of inquiry and constructivist practices. Therefore, this study used data obtained from quantitative sources using IBLES (based on scales from WIHIC and CLES) and TOSRA. The study provided insight into how students’ perceptions and attitudes differed under alternative instructional methods. Also associations between various aspects of classroom environment and students’ attitudes were reported. The conclusions presented in this chapter are organized under the following headings:
Section 5.2 Overview of the Thesis Section 5.3 Major Findings of the Study
Section 5.4 Significance and Implication of the Findings for Educational Practice Section 5.5 Constraints and Limitations of the Study
Section 5.6 Suggestions and Recommendations for Further Research Section 5.7 Concluding Remarks.
5.2 Overview of the Thesis
Chapter 1 introduced the background, conceptual and theoretical framework, significance and aims of the study. It also provided insight into the development of the New Generation Science Standards (NGSS) and the recent adoption of federal and state educational reforms in the USA. The four research questions which were
the focus for this study were also delineated in this chapter. The four research questions are:
a. Is it possible to develop valid and reliable measures of science students’ perceptions of inquiry-based classroom learning environments and their attitudes towards science?
b. Is inquiry-based instruction effective in terms of students’ perceptions of learning environment and attitudes towards science?
c. Is inquiry-based instruction differentially effective for male and female students in terms of perceptions of learning environment and attitudes towards science?
d. Are there associations between students’ perceptions of inquiry-based learning environments and their attitudes towards science?
Although extensive research has been conducted in the field of learning environment, Chapter 2 mainly reviewed literature pertinent to my study and presented some historical background of classroom environment research. I comprehensively reviewed literature related to the origin, development, and uses of classroom environment and attitude instruments. Determinants of students’ perceptions of their classroom environments were reviewed in Chapter 2. Also, literature on students’ attitudes to science, as well as past studies of associations between students’ perceptions and learning outcomes, were also reviewed. The chapter concluded by presenting detailed characteristics of inquiry-based learning in science classrooms, as well as its implications and prospects in modern science classrooms in the Californian public school system.
Chapter 3 described the methods, research design, and techniques used for data collection. It also described the development and selection of environment scales for inclusion in the Inquiry-Based Learning Environment Survey (IBLES) and the attitude scales from the Test Of Science-Related Attitudes (TOSRA) used in this study. The 42-item 6-scale IBLES and the 30-item 3-scale TOSRA were used with a sample of 1,396 students in biology and earth science classrooms in Los Angeles County, California. A total of 886 students participated in classrooms where inquiry- based practices were implemented, while 510 students participated in non-inquiry- based classrooms. The sample comprised 735 male students and 661 female students.
The IBLES combines scales from the WIHIC and the CLES that were carefully selected to address specific psychosocial characteristics relevant to inquiry-based science classroom environments. Three scales from the WIHIC (Student Cohesiveness, Teacher Support and Involvement) and three scales from the CLES (Personal Relevance, Critical Voice and Student Negotiation) were selected for use in this study. From the TOSRA, the three scales of Attitude to Scientific Inquiry, Enjoyment of Science Lessons, and Social Implications of Science were selected for use in this study.
Chapter 3 also described the statistical analyses used to provide answers to my four research questions. To answer research question #1 concerning the validity and internal consistency reliability of my questionnaires, principal axis factor analysis followed by varimax rotation and Kaiser normalization was performed for my sample of 1,396 students for the 42-item 6-scale IBLES and 30-item 3-scale TOSRA to check the a priori factor structure. The Cronbach alpha coefficient was used with the individual student as the unit of analysis to determine the internal consistency reliability of each IBLES and TOSRA scale.
Answers to research questions #2 and #3 were obtained from a two-way multivariate analysis of variance (MANOVA) performed with the six scales of IBLES and two scales of TOSRA as dependent variables and with instructional method and sex as the independent variables. The results from the two-way MANOVA, together with effect sizes, were also used to determine the overall effectiveness of inquiry-based instruction, as well as whether inquiry-based learning was differentially effective for males and females.
Research question #4 was answered by performing simple correlation and multiple regression analyses with the individual as unit of analysis to determine associations between students’ perceptions of their classroom learning environment and their attitudes towards science. Standardized regression coefficients were used to identify which environment scales contributed uniquely to variance in student attitudes.