Asia-Pacific Forum on Science Learning and Teaching, Volume 20, Issue 2, Article 2 (Jun., 2021)
María-Antonia MANASSERO-MAS1 & Ángel VÁZQUEZ-ALONSO2
Science teacher education on nature of science through explicit and reflective curriculum development

Previous Contents Next


Results

The results present the teachers' beliefs across the ten NOS issues and the qualitative productions drawn from their personal reflections throughout the TLS-based intervention. Further, the change of teachers’ NOS beliefs was elaborated through pre-test and post-test comparisons and the analysis of the teachers’ qualitative reflections.

Teachers’ prior beliefs about scientific investigations

The initial individual baseline profiles for the nine teachers of Cohort 1 across the ten assessment questions are represented by the pre-test question mean indexes (Figure 1), which showed some macroscopic trends. First, the question mean scores showed a broad dispersion across teachers, which means that teachers’ prior NOS beliefs were quite different. For instance, Question 70611 (Scientists’ Personality) presented the largest dispersion across the teachers, and Question 70621 (Brilliant scientists) displayed the lowest dispersion although its top-bottom difference extended over about 0.4 points.

Figure 1. Initial baseline profiles for the ten teachers (Cohort 1) across the ten assessment questions (pre-test mean question indexes)

Second, teachers’ prior beliefs depended on the issue; Questions 70221 (Controversy closure) and 90411 (change) displayed the highest and most positive mean scores, whereas the lowest and most negative score corresponded to Questions 70611 (Scientists’ Personality) and 90621 (Utility).

Third, a slight global positive trend was perceptible, as Figure 1 displayed more points above zero than below zero across the ten questions; in fact, the cohorts’ great mean was slightly positive (+.080).

The above qualitative features of teachers’ initial NOS beliefs hold broadly for the second cohort although the representative questions may vary somewhat. Thus, these features allowed concluding that the initial beliefs were very variable across teachers and questions. Moreover, they were not definitely positive, thus confirming previous research that claims that teachers’ NOS education is insufficient and inadequate to qualify them for quality NOS teaching.

Change across questions

The appraisal of teachers’ change of their NOS beliefs is based on the effect size of the differences between the final (post-test) and initial (pre-test) mean indexes for each of the ten questions (Table II). Overall, teachers improved in most questions, as Table II displays a majority of positive effect sizes (111/190, 58.4%) and fewer negative effect sizes (79/190, 41.6%), revealing an overall improvement in teachers’ NOS beliefs.

Table II. Teachers’ change profiles across the ten assessment questions between post-test and pre-test mean indexes, represented through the effect size of the differences for each question.

Teachers

Assessment questions

Teacher

 

10113

60211

70221

70611

70621

90111

90411

90611

90621

90631

Mean

T1

0.847

0.236

-0.495

1.025

0.548

0.975

-0.136

0.435

0.802

0.195

0,411

T2

-0.266

-0.428

-0.257

-1.047

0.230

1.109

0.355

1.202

-0.419

1.070

0,194

T3

-0.274

0.094

0.550

0.423

-0.061

-0.305

0.000

-0.139

-0.061

0.049

-0,011

T4

-0.139

0.407

0.108

1.030

-0.493

0.429

0.346

0.141

0.215

0.427

0,226

T5

0.082

0.742

0.336

0.111

-1.520

0.209

-0.965

-0.103

0.668

-0.084

0,016

T6

-0.307

-0.433

-0.567

0.308

-0.921

-1.813

-0.385

0.402

0.525

0.327

-0,209

T7

0.237

-0.551

0.000

0.000

-0.956

-0.500

0.167

0.643

0.394

0.230

0,095

T8

-0.553

-0.113

-0.233

-1.008

1.154

-0.196

-0.474

-0.529

0.000

0.086

-0,193

T9

0.075

0.735

0.259

0.063

0.473

-0.176

0.089

0.442

-0.217

-0.066

0,156

T10

0.506

0.166

-0.807

0.499

0.000

1.746

-0.578

0.859

1.298

-0.093

0,430

T11

1.640

0.000

0.490

0.684

0.757

0.236

0.000

0.331

-0.061

-0.521

0,224

T12

-0.169

0.651

-0.115

0.378

-0.633

-0.704

0.000

0.503

0.000

0.790

0,114

T13

0.554

-0.504

-0.428

0.252

0.132

-0.867

0.000

1.413

-0.316

-0.139

0,115

T14

0.530

-0.214

-0.642

0.000

-0.657

-0.947

-0.150

0.471

0.317

-0.162

-0,052

T15

0.413

0.725

-0.685

0.000

0.000

-0.585

0.521

0.491

-2.079

-0.085

-0,057

T16

-0.306

-0.515

-0.874

0.360

0.000

0.168

0.448

-0.100

-0.279

0.167

-0,085

T17

0.332

-0.855

-0.706

-1.071

-0.510

-0.179

0.317

-0.048

0.182

0.516

-0,141

T18

0.984

-0.577

1.016

-0.208

0.101

-0.817

-0.425

0.511

1.411

-0.139

0,133

T19

-0.244

0.328

-0.226

0.377

0.102

0.000

0.747

0.428

0.000

0.553

0,201

Question
Mean

0.208

-0.005

-0.172

0.115

-0.119

-0.117

-0.006

0.387

0.125

0.164

Five of the ten questions reached a positive mean effect size (d>.10), only three questions reached a negative, albeit nonsignificant, mean effect size (-.20<d<-.10), and the remaining questions hardly changed, with a close-to-zero effect size (.10<d<.10). Question 90611 (Scientific method) reached the highest positive effect size improvement, which is close to the significance threshold, and Question 70221 (Controversy closure) reached the lowest negative effect size improvement.

It is well known that the sample’s overall mean decreased the measurement variability, so that it underestimated the effect size. In fact, the effect size of the individual questions across teachers (inner cells, Table II) showed higher scores than its great mean (last row, Table II). For example, Question 90611 (Scientific method), which reached the largest average effect size, showed that 12 teachers exceeded the minimum threshold effect size (d>.40), with some teachers exhibiting high scores (Table II), although the question’s great mean did not reach the minimum threshold (d<.40).

Change across teachers

The analysis of the teachers’ changes across the final-initial differences (last column, Table II) showed the uniqueness of each teacher's evolution. Most teachers (10) improved their NOS beliefs (d>.10), and two teachers exceeded the minimum significant effect size (d>.40). These two teachers displayed six and five questions that reached a large significant effect size but the presence of one and two questions, respectively, whose effect size was significant and negative, is also noteworthy.

In contrast, only three teachers reached an overall negative effect size (d<-.10), although none of them even reached one half of the effect size threshold. Again, these three teachers with negative global change profiles presented one question that reached a significant and positive effect size. The remaining teachers (6) displayed an overall close-to-zero effect size, and their profiles across questions displayed a variable number of questions with significant effect size scores, either positive or negative.

The teachers’ change pattern was quite inhomogeneous and inconsistent across questions, as each teacher simultaneously exhibited significant positive and negative changes, that is, questions that worsened and questions that improved their profile. The teachers with the highest positive improvement profiles also displayed some questions with significant negative effect sizes, and conversely, the teachers with the lowest negative change also displayed some questions with significant positive improvement.

Qualitative Results

The elaboration and presentation of the TLS on investigations and the answers to two self-reflective questions about their test results are some products of the teachers’ reflections on NOS issues that could contribute to another qualitative study by themselves. However, the available space limits its presentation to some core results of the first cohort’s personal reflections on their question ratings. The instructor provided each teacher with their own ratings and requested two responses about each assessment question: "explain the reasons for your ratings" and "explain the reasons that justify your change/no change in each question when comparing your first responses to the second ones.”

Table III. Excerpts of teachers' reflections on the first reflective item (Explain the reasons that justify your ratings) across three significant teachers and three assessment questions

Teacher displaying the highest positive change

Teacher displaying the lowest negative change

Teacher displaying the largest absolute change

10113
I find acceptable those responses that describe the process of doing science as a set of activities-methods designed to explain our environment (nature-world-universe). The phrase D (use of technology to discover Nature) does not seem acceptable because it is limited to a single way of doing science, such as the use of technology.

10113
For me, the process of doing science is to discover the mysteries of the universe and the behavior of Nature, using new technologies and qualitative and quantitative methods.

10113
I agree to a greater or lesser extent with all the answers. Therefore, I have given them all a score above 6. The premises that I think best describe the process of doing science are everything we do, discovering order and proposing and validating explanations, so I gave them a score of 8 (high agreement).

90111
Observations are the results of the experiment directly (without treatment) or sanitized results.
Sometimes, treatment can produce incorrect data or erroneous observations.

90111
The observations are always the same but either they cannot explain all the knowledge or they will differ very little, depending on the theories.

90111
Observations must always be objective, but the interpretation of what we see will depend on the mindset and ideas of each person.

90621
I think the best scientists are those who use any method (including the scientific method) to obtain results.
The serendipity of many discoveries shows that it is really not necessary to always follow a particular method.

90621
Scientists using the proper scientific method in their field will achieve better results, but chance discovery also produces occasional successes.
The scientific method is subject to change but these changes must be based on scientific arguments, to avoid achieving false results.

90621
I agree with the idea that most scientists continue to use this method for their research because it assures valid, clear, logical and accurate results, although these results can also be negative.

Table III presents verbatim some exemplar excerpts of the responses to the first item by three prototypical teachers: the teacher who achieved the overall highest improvement, the teacher who achieved the lowest negative change, and the teacher who achieved the highest absolute changes, either positive or negative. Overall, the excerpts display another facet of the results and exemplify the teachers’ diverse evolution patterns across the informed and misinformed NOS conceptions. Further, they also display the subtlety of the nuances that differentiate the teachers’ positions from each other, and between informed and misinformed conceptions.

The highest-improvement teacher justifies his responses, considering the centrality of scientific explanations in the process of doing research, stating that observations are much more than simple data and any method is useful to obtain results. The lowest-improvement teacher sustains traditional positivist positions on science (discovering the mysteries of the universe and the behavior of Nature) and observations (observations are always identical) but he considers that theories can project some influence on the observations; he also views the scientific method as the way to achieve better results, while simultaneously acknowledging that “scientific method is subject to change.” The highest-change teacher expresses an informed view on observations, a holistic conception of science as a process, and also stresses the centrality of explanations; however, she joins the traditional positivist views in her conception of the scientific method.

The second item requests teachers to identify and explain the changes produced in their thinking when comparing their initial and final responses. Overall, teachers' responses to this requirement followed two different patterns: on the one hand, a concrete pattern, specifying and mentioning the question phrases where they perceived the presence or absence of change; on the other hand, a generic pattern, when the teacher acknowledged change/no change, without mentioning specific evidence.

For instance, Teacher #2 rightly argues for a change in question 70611 (Scientists' personality) through this specific reasoning:

My answers to this question have changed slightly. You can see a trend toward thinking that [a scientist’s] character influences the result. Certainly, this change is due to the fact that the debates in class have led me to see the diversity of opinions on the same theories.”

The comment specifies the nature of change and evaluates the amount of change (slight). However, the teacher underestimated the amount of change, as the initial and final scores of most sentences showed quite relevant positive changes, so that the teacher actually achieved a greater change than he perceived.

Globally, teachers’ explanations suggest some difficulty to accurately perceive and account for the changes between their initial and final scores. Thus, teachers' self-perceptions of their changes were elaborated by computing hits and misses in each question (Table IV). The procedure compares the teacher's decision expressed in their comments (either change or no change/same) with their actual quantitative ratings (change or no change/same) of the final-initial quantitative differences for each question. If a teacher's position on a question (change or no change) agreed with the actual quantitative difference between question ratings (both change or both no change), the comparison was assessed as a hit. On the contrary, if the two traits were different (change vs. same), a miss was assigned.

Table IV. Frequency of teachers' hits and misses drawn from the reflections elicited by the second item (Explain the reasons that justify your changes) when teachers compared their quantitative ratings on the assessment questions

 

Hits

Misses

 

Teacher/Ratings

Teacher/Ratings

Teacher/ Ratings

Teacher/Ratings

 

change/change

same/same

change/same

same/change

 

Cohort 1

Cohort 2

Cohort 1

Cohort 2

Cohort 1

Cohort 2

Cohort 1

Cohort 2

Teacher 1

1

5

4

1

2

2

3

2

Teacher 2

4

5

4

1

 

2

2

2

Teacher 3

3

3

4

4

3

 

 

3

Teacher 4

4

3

3

4

3

1

 

2

Teacher 5

2

5

2

1

6

1

 

3

Teacher 6

4

5

3

 

2

2

1

3

Teacher 7

6

2

1

4

2

4

1

 

Teacher 8

2

3

4

5

2

1

2

1

Teacher 9

4

4

1

4

1

1

4

1

Teacher 10

 

6

 

1

 

1

 

2

The most interesting results of Table IV are not the perception of hits (left side), as the teachers were expected to be able to correctly identify their changes, but the perception of errors (misses), because they were unexpected (right side), taking into account that the evidence (teachers’ ratings) was provided to the teachers. Overall, the teachers made about 34% of mistakes when they assessed their conception changes. Furthermore, there were still two types of misses: on the one hand, when the teacher perceived that their rating had changed and the evidence of ratings showed there was no change (change/same); on the other hand, when they did not perceive change, and the ratings showed that the change was relevant (same/change). Both errors were split nearly into halves, although the former (change/same) was more frequent.

The procedure allows the identification of the teachers according to their self-perceptions of changes; Teacher #5 displayed the highest number of errors (misses) about his perception of changes (6 change/same errors), and in contrast, Teachers #8 and #9 (cohort 2) displayed the lowest number of errors (2).

All in all, the teachers' qualitative reflections contributed to confirming many of the quantitative assessment results drawn from the test indexes. Further, teachers’ reflections also showed the many mistakes they make when trying to identify changes in their NOS learning and beliefs through self-assessment, thus supporting the need to deepen the reflections on NOS beliefs.

Relationship between qualitative and quantitative results

The main quantitative products are the indexes of the questionnaires, which represent the teachers' initial and final responses to the assessment instrument questions. The main qualitative products of this experience are the TLS and the reflections on two items (self-justification of scores and self-perception of the changes between the initial and final assessment). The qualitative products of the education course were graded by the researchers through rubrics, with consensus-building in case of disagreement on a rubric cell (prior to grading).

The relationships between the quantitative products, questionnaire indexes, and the qualitative products, quantified through the grades obtained by the teachers, were analyzed through Pearson correlations (Table V). For each teacher, the analysis considered the following variables: initial and final assessment average indexes, the average test change scores (quantitative variables), and the average of the three grades of the qualitative products (TLS, and the answers to the two items about the analysis and identification of changes).

Table V. Pearson correlation coefficients between the test indexes and the grade of the reflective identification of the improvement task for the two cohort teachers (n = 19)

 

Change identification (grade)

Pearson (r)

p-significance (two-tailed)

Pre-Test Scores

.071

.773

Post-Test Scores

.568

.011

Improvement

.586

.008

Most variables showed irrelevant correlation coefficients that lack statistical significance except for the qualitative variable change identification, which significantly correlated with the final evaluation indexes (post-test scores) and the NOS understanding improvement indexes (table V). These results mean that the higher the teachers’ effectiveness to identify changes in their NOS beliefs, the greater their index in the final evaluation (post-test), and the higher the improvement in their NOS understanding achieved through the education course. It should also be noted that the TLS score and the justification reflection score did not significantly correlate with the test indexes.

 


Copyright (C) 2021 EdUHK APFSLT. Volume 20, Issue 2, Article 2 (Jun., 2021). All Rights Reserved.