Case Study of Teacher Beliefs in Contemporary Science Education Goals and Classroom Practices

¶ … teacher beliefs in contemporary science education goals and classroom practices. Fischer-Mueller, J. & Zeidler, D.L.

The participants in the study included nine science teachers at the Souhegan High School (SHS) in Amherst, New Hampshire. However, three "typical case" teachers chosen for the qualitative data analysis as they were willing to allow an observer in their classroom and participate in multiple interviews.

There are three major research questions being addressed: (1) What is the extent that the science teachers a SHS support the contemporary goals of science education? (2) What is a SHS science teacher's degree of conviction concerning their beliefs about particular goals? (3) How is a teacher's belief in contemporary goals translated or expressed by their actual classroom teaching?

The methodology included surveys and interviews (Contemporary Goals of Science Education Survey; Zeidler & Duffy, 1994) but for question three it is predominately qualitative in that it uses observation, coding of observations, and interview data of the three teachers in conjunction with their responses on the survey to address its main questions. The interviews were recorded on an audiotape and were semi-structured using the Newmann et al. (1995) (Standards and Scoring Criteria for Classroom Instruction and Assessment Tasks) as a guideline, but were obviously not fully standardized.

The data collection procedure is quite complex for a qualitative study and will be briefly addressed here. The researchers support the data collection method by adhering to Lincoln and Guba's (1985) criteria for trustworthiness with adherence to credibility, transferability, dependability, and confirmability of the data. To make sure the data was credible the researchers employed a number of techniques including prolonged engagement, persistent observation, triangulation, and member checking. Interviews with the teachers were used as a means of checking the coding and observations. A reflexive journal was used to attend to dependability and confirmability to rule out the effects of personal bias and experimenter effects of the investigator.

While the data analysis for three teachers, their classroom observations, and interviews is qualitative in nature all nine teachers completed the Contemporary Goals Survey. Teachers generally demonstrated a conviction to contemporary goals over past goals (research question one). However, there was a mild discrepancy in the conviction of contemporary over past goals (the research question two). The results indicated that in one area the difference in the strength of the conviction between a corresponding past and contemporary goal was inconsequential. Weighted means for the conviction to the past and contemporary goals were calculated from the survey and the difference between the weighted means determined. Here the authors are a bit vague in their explanation, but as it turns out if the difference is less the .15 the strength of the conviction between past and contemporary goals is inconsequential (Zeider & Duffy, 1994). For the pair in question: "Science education should focus on knowledge acquisition and process skill unrelated to the interactions of science, technology and society" (past goal) and "Science education should stress the interactions among science, technology, and society" (contemporary goal) the difference was -.17, indicating overall no difference in the strength of conviction. The authors spend considerable time analyzing the responses of all nine teachers, and surprisingly conclude that it is hard to believe that those that expressed a conviction in the past goal did so with the same strength as they did for the contemporary goal-despite their earlier findings based on accepted standards. I think this is an interesting conclusion and it reflects the very bias that the investigators purport to be trying to control with their reflexive journal during the qualitative analysis. The fact that the group demonstrated no overall difference in the strength of the past and contemporary goal conviction says just that, there is no overall difference. Instead of trying to fit the result in with the hypothesis, a thorough follow-up analysis, re-questioning, and discussion with the participants is in order so that the researchers can fully understand this result in much the same manner that the researchers address potential problems with their qualitative analysis of the classroom behavior of the three teachers. When a researcher begins to try to explain a quantitative result away we can sense their own bias and also can become suspicious of other findings. Furthermore, since there is an established cutoff for this type of comparison there must be past research with these pairs of questions and the current investigators could have taken a look at past findings to determine how the trends in this data typically look. Perhaps this was not an unusual finding, or perhaps it has been explained before. This could have provided an explanation for the current findings in conjunction with a follow-up of the participants. There was one other pair that was close to the cutoff of .15 and the researchers explain this result straightforward fashion. All other 12 pair comparisons indicated a solid conviction of contemporary goals compared to past goals by the teachers in those areas.

For the final research question the data analysis consisted of focused observations, interviews, and taxonomic analysis of field notes, reflexive journals, and video tapes with participants. This process produced 81 codes for classroom practice, a very large number of variables. The researchers paired this down by pairing a code with a contemporary goal in the Survey of Contemporary Goals (Why not just do this in the first place?). This resulted in 30 codes and the researchers report on consistencies and inconsistencies in convictions to eight of these. The final conclusion, not surprisingly, was that in general degree of evidence that the teachers' beliefs in the contemporary goals of science were embedded in their classroom practice despite the great variance in the teachers' actual classroom behaviors.

The data analysis, results, and conclusions of any research paper are all interconnected. The type of data analysis used in a research design determines and automatically places limitations on the results of that design. For example, most quantitative data analysis is not very helpful in determining processes, whereas most qualitative data analysis is not very good at ruling out chance or determining statistical and clinical significance of findings. The findings in the results section is of course crucial in determining any conclusions that can be made. Therefore, the data analysis should be specifically determined by the hypothesis of the research. But there are a number of other biases that can undermine any research design despite the analyses, results, and conclusions.

In the current study the researchers attempted to answer three specific questions about a group of high school science teachers. Therefore, the generalziability of any findings in this study is very limited. Moreover, the head researcher is closely affiliated with the teachers and the program by her own admission and even though she takes steps to make sure that her findings are not biased, there is a real question of bias here. Do I believe that she would have published negative findings about the teachers or about the program? I am skeptical. When there was a minor discrepant finding such as with the attitude data, the authors went out of their way to sugarcoat it. A third-party investigator would have been more credible.