2. Materials and Methods
The research design was quasi-experimental and of a quantitative nature, with a pre-test and two post-tests, a control group, and an experimental group. Both descriptive and inferential data analyses were carried out. Likewise, didactic materials were developed to help prospective teachers in their professional future to design teaching/learning sequences with scientific–didactic rigor.
2.1. Objectives
The main objective of the study was to compare the influence of two teaching methodologies on the learning and teaching self-efficacy of the trainee teachers for physics content related to light and color. To achieve this objective, didactic interventions based on STEM practices and hyper-realistic simulations were designed, implemented, and compared with an expository academic teaching of the same concepts.
The general objective was broken down into the following specific objectives:
Specific Objective 1 (SO1): To validate from a didactic point of view the usefulness of the didactic tools developed for the learning of the selected optics contents.
Specific Objective 2 (SO2): To verify whether the concepts of light and color learned by the trainee teachers through different teaching methodologies last or are forgotten with the passage of time.
Specific Objective 3 (SO3): To test whether levels of teacher self-efficacy improve as a function of the teaching methodologies applied with trainee teachers.
2.2. Hypotheses
Based on the proposed objectives, the following study hypotheses were proposed:
Hypothesis 1 (H1). Students who use hyper-realistic simulations and STEM experiences to learn basic optics concepts related to light and color have similar average initial scores to students who follow an academic-expositional teaching intervention.
Hypothesis 2 (H2). There are no statistically significant differences in the short-term knowledge level variable of students who follow a didactic intervention based on the use of hyper-realistic simulations and STEM experiences compared to students who follow an academic-expositional didactic intervention.
Hypothesis 3 (H3). There are statistically significant differences in the long-term learning variable between students in the experimental group using STEM simulations and experiences and students in the control group following an academic-expositional teaching intervention.
Hypothesis 4 (H4). Hyper-realistic simulations and STEM experiences on light and color facilitate meaningful, long-term learning for trainee primary school teachers.
Hypothesis 5 (H5). The development and implementation of didactic interventions on basic concepts of optics related to light and color produce a positive evolution in the variable level of teaching self-efficacy in trainee teachers.
2.3. Sample
The sample, chosen by non-probabilistic convenience sampling due to the ease of access (as they were the students we teach) consisted of 173 primary school teachers-in-training. The participating subjects were studying experimental science didactics to become future teachers. In this subject, they were taught the scientific and didactic content for teaching/learning the concepts of the subject of natural sciences at the primary school stage so that they could explain them to their future pupils in a meaningful way. The students were divided into two homogeneous and equivalent groups in terms of initial knowledge level, a control group, and an experimental group to test the research hypotheses.
Table 1 shows the descriptive analysis of the sample according to group and gender.
As shown in
Table 1, the first group, called the Control Group (CG), consisted of 86 subjects. The second group, called the Experimental Group (EG), consisted of 87 students. These groups were made up of students from the degree of primary education, future primary education teachers. The absolute frequency and percentage of the variable gender are indicated. In the control group, 72.1% were female and in the experimental group approximately 67.8% were female. The age range of the participants was 20 to 26 years. Both the control group and the experimental group used the same amount of time to teach the contents. The didactic methodology of the control group was based on a more traditional teaching, using different presentations and theoretical explanations of the selected contents as learning resources. However, with the experimental group, practical sessions were carried out based on STEM experiences [
80] and the use of hyper-realistic simulations of our own elaboration [
74,
76,
77]. For the development and implementation of the simulations, we used software specifically designed for rendering photorealistic graphical environments, namely POV-Ray, an open-source ray tracer [
81]. The choice of this program was motivated by the need for a technique capable of faithfully imitating the optical system in a way that was consistent with the theoretical models involved. POV-Ray uses a ray tracing technique based on geometrical optics that simulates images with great realism [
82,
83,
84]. The software models the path of light following rays as they interact with optical surfaces, resulting in accurate simulations of optical phenomena. These simulations arise as a natural result of the combined use of the ray-tracing algorithm and a specific Monte Carlo algorithm for the synthesis of three-dimensional images with perspective. These hyper-realistic simulations show environments of optical phenomena that reproduce the behavior of real systems with a higher level of reality than traditional computer simulations. The STEM experiences are practical projects carried out by the students using easily available materials. In the experience, the didactics of the selected science, technology, engineering, and mathematics contents are explained to the students in an interrelated way. An observation sheet is also included, containing questions that focus the students’ attention on the contents being worked on. In addition, a guide is added where the didactic and methodological components of each experience are specified so that they can be easily carried out in any school. By way of example,
Figure 1 shows some of the simulations used with the experimental group. The simulation of a flat diopter, a coin immersed in water, a convergent lens, chromatic aberration, a prism, a concave mirror, and the same horse under two different illuminants is presented.
2.4. Measuring Instrument
Two measurement instruments were designed based on the variables under study: one to measure the variable level of knowledge and the other to measure the variable level of teaching self-efficacy. Each of them is described below, followed by the validation of each of these instruments.
First, a test was designed to detect misconceptions about basic concepts of optics related to light and color. The test consisted of 35 closed multiple choice questions with a single answer, based on previous studies [
79,
85] and designed considering the distractor theory and the scientific literature on misconceptions in optics. These questions could be grouped into several categories according to the specific concepts worked on. Specifically, they were grouped into four categories:
Category I aimed to analyze whether the learner remembered basic concepts about what light is and its nature, behavior, and characteristics and was called Category I— Light. Nature and Propagation.
Category II aimed to identify whether students could distinguish between light primary colors, ink primary colors, additive mixing, subtractive mixing, and the perception of the color of objects as a function of the illuminant used. This category was labelled Category II—Color.
The purpose of category III was to analyze whether students had preconceptions about the laws of reflection and refraction of light; the formation of images in a mirror; and the behavior of lenses, prisms, or filters, among other optical systems. This category was called Category III—Simple Optical Systems (laws of geometrical optics).
Category IV aimed to find out if students knew how rainbows are formed and their nature and was named Category IV—Rainbows.
As examples, some of the questions included in the questionnaires designed are shown in
Table 2.
The misconception detection test was used with the students at three different times, specifically once as a pre-test and twice as a post-test. The pre-test was given at the beginning of the didactic sessions before the teaching of the contents under study in the two groups began. The aim of post-test I was to check the effectiveness of the didactic methodology used in each group, as well as to check the persistence of the misconceptions after carrying out different teaching–learning sequences. Specifically, the students were tested after their respective teaching sessions (control and experimental) to find out the degree of acquisition of the contents explained according to the different didactic resources used in the sessions. The purpose of post-test II was to verify whether significant learning had taken place in the students and whether they remembered the content explained after the passage of time. The students in both groups were tested months after post-test I.
In addition, the aim was to analyze the level of self-efficacy of prospective primary school teachers in teaching content about light and color. For this purpose, a questionnaire was designed based on previous research [
15,
86,
87]. The questionnaire, on teaching self-efficacy on light and color, consisted of 28 items that were formulated based on the optical contents and activities to be developed in the classroom of the fifth and sixth grades of primary education. The trainee teachers had to rate on a 4-point Likert scale (0: Not at all competent, 1: Not very competent, 2: Fairly competent, 3: Fully competent) their level of teaching self-efficacy for the teaching of the selected light and color contents. This questionnaire was implemented before and after the didactic intervention to assess the evolution of the variable teaching self-efficacy. By way of example,
Table 3 shows the statements proposed.
2.5. Validation of the Evaluation Instrument: Calibration Indices of the Misconceptions Test
Firstly, the results referring to the validity and reliability of the misconceptions test used in the research are presented based on the analyses recommended by other studies [
88,
89,
90]. Based on the results obtained, we can affirm that the test for the detection of misconceptions designed presents an adequate degree of reliability and validity, constituting a reliable assessment instrument with an adequate level of difficulty and discriminatory power, as we will see below.
Specifically, the validity and reliability of this measurement instrument was determined through the consensus of opinions of a group of experts. Following the guidelines set by some previous studies [
88], a concordance test was carried out among experts, who were provided with eight assessment criteria on which they had to mark their degree of agreement (scored as 1) or disagreement (scored as 0). The degree of agreement is calculated as the result of the number of total agreements divided by the sum of the number of total agreements plus the total disagreements. The value obtained in this evaluation was 0.91, which indicates a degree of agreement classifiable as very good according to the literature [
88].
In addition, several psychometric tests were carried out following the methodology recommended by other authors [
89,
90,
91,
92]. Statistical tests focused on the assessment of the test items, such as the difficulty index, discrimination indices, point biserial coefficient, Ferguson’s Delta, and Kuder-Richardson’s 20 coefficient, were performed using the methodologies specified in previous studies. As shown in
Table 4, all values are within the recommended range.
The mean difficulty index (P) of the test indicates the degree of difficulty of the test so that the higher the index, the easier the question asked. The calculation of this index was carried out for all the questions that made up the test, obtaining similar difficulty values for all of them, which were within the established ranges.
Table 4 shows that an average value of
p = 0.49 was obtained, so in general, the degree of conceptual difficulty of the instrument is adequate for the research.
Subsequently, discrimination indices were calculated. Discrimination index 1 (D1) measures the discriminatory power of each item in a test, i.e., it allows us to conclude whether the test can distinguish those subjects with stronger knowledge who answer correctly from those subjects whose understanding is weaker. Discrimination index 1 (D1) was calculated for all questions included in the instrument to check whether there were questions that were excessively easy or excessively difficult that did not discriminate and therefore did not contribute to the reliability of the instrument. The average value obtained in the test, as specified in
Table 4, is D1 = 0.36, indicating an adequate discrimination index. Discrimination index 2 (D2) indicates the extent to which a question helps to distinguish between those who know the most and those who know the least, regardless of the easiness of the question, and can be considered satisfactory if it is at least higher than 0.50, that is, if more than half of the respondents belong to the group that knows the most. In our case, this fact was fulfilled in all the questions, with an average value of D2 = 0.72. This value is considered adequate by the literature.
The point biserial coefficient (r
pb) reflects the correlation between subjects’ scores on one item with their scores on the whole test. As shown in
Table 4, the average point biserial coefficient of the test is r
pb = 0.32, so it also meets the recommended criterion.
For more evidence, we obtained Ferguson’s Delta (δ). The literature recommends following the criterion that a test that offers good discriminatory power obtains a value greater than 0.90. The tests in the study have an index of approximately δ = 0.91, so in general terms, the instruments offer good discriminatory power.
Finally, the Kuder-Richardson 20 coefficient (KR-20), which is a measure of internal consistency reliability for measures with dichotomous choices, was calculated. A KR-20 value of 0.72 was obtained for the misconceptions test, indicating high reliability.
2.6. Validation of the Self-Efficacy Instrument: Calibration Indices
To validate the self-efficacy test designed, the reliability coefficient was calculated using Cronbach’s alpha to estimate the reliability of the instrument for measuring teacher self-efficacy. As pointed out by some authors [
93], the measurement of reliability for items formulated in Likert-type scales assumes that the items measure the same construct and that they are correlated. Thus, if the alpha value obtained is close to 1, the consistency of the items is excellent. On the other hand, the literature recommends obtaining the reliability of the scale with data from each sample. The result obtained was α = 0.967 for the 28 items that made up the questionnaire. This result allows us to conclude that the questionnaire for measuring the level of teacher self-efficacy has excellent reliability [
94].
4. Discussion and Conclusions
First, a summary of the hypotheses formulated and their implications for the research is shown in
Table 16.
Regarding the achievement of Specific Objectives 1 and 2, related to the cognitive domain, the descriptive and inferential statistical analysis of the data obtained has revealed an improvement in the cognitive domain of the trainee teachers in the process of teaching the scientific concepts under study. However, depending on the methodology used in the different didactic interventions, we can affirm that the students in the experimental group, who worked with hyper-realistic simulations and STEM experiences, assimilated the knowledge in a more satisfactory way, achieving significant learning. Specifically, their learning has lasted over time, and the misconceptions found in the pre-test have been combated with the intervention developed. However, the students in the control group, who used a more traditional methodology in their teaching process, forgot over time the contents learned, and some of the misconceptions found at the beginning of the research have resurfaced in them. This implies that the learning of the selected scientific concepts of light and color by means of a traditional methodology was short-term and probably more rote than that of the pupils in the experimental group. We consider these results to be proof of the didactic validity of the resources used in the experimental group when carrying out teaching interventions on optics concepts.
In addition, regarding the achievement of Specific Objective 3, related to the teaching self-efficacy variable, a positive evolution has been observed in the participants’ levels of teaching self-efficacy. As some studies [
52] point out, levels of teacher self-efficacy in science are raised when positive experiences with science teaching occur. Specifically, student participants have been found to improve their levels of teacher self-efficacy following the development and implementation of teaching interventions. These results seem to indicate that future primary school teachers feel more competent to teach optics concepts to their future students, which may improve the future teaching of these concepts from the early stages of school. These results are in line with the statements of other authors [
104,
105] who indicated that trainee teachers’ beliefs are often projected onto their future teaching in the primary classroom. If a teacher feels unprepared to teach science, this will contribute to fostering negative attitudes towards science learning in future students [
86]. Additionally, based on the results obtained in the research with respect to the didactic interventions developed, we can conclude that the use of hyper-realistic simulations and the STEM experiences designed promote the acquisition of scientific competence in the trainee teacher combats the misconceptions found in them and significantly increases the learning of optics with respect to more traditional teaching. These results are in line with those obtained in other studies [
74], where it was found that students who used hyper-realistic simulations to study optics concepts learned more than students who used schematic simulations and students who used the traditional laboratory. Similarly, other authors [
110] report that students using simulations obtained better results than those using real equipment. We therefore consider it relevant to carry out these type of activities and workshops with trainee teachers so that they can increase their levels of teaching self-efficacy, thus favoring their future professional development.
The results obtained show that the practical activities prove to be an effective tool for the promotion of the teachers’ scientific-experimental didactics [
57] and the learning levels of this group. Initially, insufficient content management was observed, but the use of STEM didactic tools with the experimental participant sample led to a considerable increase in the level of learning about optics and, consequently, an improvement in the scientific literacy of this group, which is necessary to better teach science. Primary school teachers must have a strong mastery of basic physics and physical science because of their influence as future teachers on many students [
111]. In this sense, we believe that teacher preparation based on science-based teaching is essential to ensure student learning. Teachers need to know the content to reconstruct, adapt, restructure, and simplify it to make it comprehensible to students [
112]. Therefore, we agree with other studies [
113] on the importance of the practical experimental approach in the physics teaching–learning process to improve the scientific preparation of teachers in these areas and to show them ways to transform scientific–technological content into didactic representations and use them in practice.
On the other hand, we consider that, although subject knowledge is indispensable, it does not in itself generate ideas of how to present particular content to students. A didactic knowledge of the content is required for a good teacher [
114,
115]. Three fundamental aspects in the development of didactic content knowledge [
116] are content knowledge, teaching practice, and emotional attributes. The affective domain is closely linked to teaching competence [
117,
118] since, as many authors [
119] maintain, trainee teachers who show high self-efficacy in teaching physics and chemistry have higher positive emotions and lower negative emotions towards teaching these subjects than those with low self-efficacy. In this sense, the results obtained in this study on the self-efficacy variable suggest that the use of teaching strategies that favor positive emotional states leads to an increase in feelings of competence and personal efficacy, thus coinciding with the contributions of [
120]. Teachers’ beliefs and attitudes regarding classroom management, discipline, behavioral control, and an effective learning environment can and should be modified, as these changes play a determining role in effective teaching [
121]. Accordingly, teaching teachers to become more self-effective should be a prerequisite in teacher training courses. High levels of teacher competence contribute to improving teachers’ abilities to manage personal and contextual resources linked to quality learning. Therefore, the teachers also improve students’ motivation and academic performance [
122,
123].
Additionally, teachers with good self-efficacy tend to be more confident in applying active teaching methodologies that focus on students and their learning. That is, self-effective teachers are open to innovation and are also eager to use new methods and strategies in teaching [
124]. To be successful in the study of science, and physics in particular, it is very important to recreate interesting, challenging, and fun learning situations that allow students’ active participation in the activities, facilitate interaction with invisible and multidimensional objects, reinforce theoretical concepts, and provide experiences related to real applications [
125]. In this sense, we agree with [
126] that illustrating the phenomena of optics by means of experiments made with homemade or low-cost materials allows each student to construct his or her own learning. Therefore, it is assumed that this research presents several effective didactic possibilities for learning optics in active learning contexts. However, implementing an experimental didactic proposal requires, in addition to the design of the prototypes and experiments, an instructional guide to support the teacher and the student to ensure the cognitive link of the programmed learning with the live experience of the phenomenon [
57].
Teaching physics involves building bridges between scientific knowledge and students’ previously constructed ideas. To do this, the teacher must rework scientific knowledge so that students can use it to interpret and transform their environment and encourage them to construct congruent models of scientific learning [
127]. Therefore, for future teachers to be able to explain and make predictions about a variety of physical phenomena, they should be trained through different methodologies that foster students’ interest and motivation [
127]. More effective science teaching requires teachers to be comfortable with the discourse, to believe in their ability to teach it, and to want to do so [
128]. Therefore, it is necessary to emphasize the basic competences that a teacher must have to conduct quality teaching–learning processes [
129]. The process of becoming a teacher is a long succession of stages in which the future professional is trained to teach as effectively as possible and to enter a profession that has always been classified as vocational [
130]. However, today’s society demands new challenges regarding the initial training of future teachers so that, focusing on students, they are trained using active methodologies [
131,
132]. Along these lines, including the interdisciplinary teaching of STEM areas in curricula will help educators to understand scientific–technological disciplines as entities interconnected with life [
133]. Moreover, with this educational model, teachers are not only experts in a single subject but also have the additional responsibility of guiding their students in all STEM subjects [
71]. Consequently, this will entail the reorganization of teacher education programs in universities [
101], as STEM education requires teachers to excel in the appropriate use of knowledge, skills, and attitudes towards scientific–technological disciplines [
133].