Effectiveness of particulate nature of matter (PNM)-based intervention studies in improving academic performance: a meta-analysis study

Abstract

Through a meta-analysis, this study examines how effective particulate nature of matter (PNM)-based intervention studies are at improving academic performance. Well-known databases (e.g., ERIC, Springer Link, Taylor & Francis, and ScienceDirect) were used to look for the PNM-based intervention studies via specific keyword patterns. Also, a manual search of related journals and dissertations was conducted to find any missing papers. Subsequently, this meta-analysis included 66 papers (44 dissertations, 21 articles, and one proceeding) published from 1992 to 2022. All statistical data from the papers were initially inserted into an Excel sheet and then imported into comprehensive meta-analysis (CMA) statistics software to calculate Hedges’ g values. The findings indicated that the overall effect-size for the random-effects model was 0.90, which means that the PNM-based intervention studies have a large effect for academic performance. Furthermore, it was revealed that moderator variables, namely educational level and type of intervention, positively affected participants’ PNM-based academic performance (p < 0.05). In light of the findings, it can be concluded that the PNM-based intervention studies are effective at improving the participants’ academic performance. Moreover, given the findings regarding educational level, it can be deduced that K–8 students are able to learn the fundamental features or dimensions of the PNM. Since the meta-analysis includes few extreme values, further research should be undertaken to test the effectiveness of the intervention types on academic performance. Furthermore, the current study excluded a notable number of papers because they lacked sufficient data; therefore, science researchers should take care to include sufficient data or effect-size value for their papers to facilitate dissemination, generalization and comparison of their results.

---

Introduction

Because the particulate nature of matter (PNM) is associated mainly with such topics as solids, liquids, gasses, phase change, osmosis, diffusion, dissolution, heat and heat transfer, electric currents and chemical reactions (e.g., de Vos and Verdonk, 1996), it plays an important role at conceptualizing advanced concepts of science, chemistry, physics and biology (Novick and Nussbaum, 1978; Nussbaum, 1985; Ayas and Özmen, 2002; Harrison and Treagust, 2002; Ayas et al., 2010). Thus, it is one of the core concepts in science that enables students to visualize and meaningfully make sense of any related scientific phenomenon at the sub-microscopic level (Harrison and Treagust, 2002; Riaz, 2004; Duschl et al., 2007; Ayas et al., 2010; Beerenwinkel et al., 2011; Hadenfeldt et al., 2016). Hence, teachers need to initially clarify their content knowledge of the PNM to help their students scientifically comprehend it and deploy their pedagogical content knowledge (Riaz, 2004).

Unfortunately, the related literature is replete with dozens of PNM studies reporting that students, pre-service and in-service teachers have difficulties in understanding the PNM and visualizing some scientific phenomena at the sub-microscopic level (e.g., de Vos and Verdonk, 1996; Coll and Treagust, 2003; Riaz, 2004; Ayas et al., 2010). Also, previous studies have indicated several reasons for these difficulties; for example, the abstract nature of the PNM, lack of proper visualization of the PNM, two-dimensional presentation of the PNM in textbooks, and teachers’ inability to explain the PNM and develop proper instructional materials (Haidar and Abraham, 1991; Tsai, 1999; Riaz, 2004; Nakhleh et al., 2005; Beerenwinkel et al., 2011; Tang and Abraham, 2016). For instance, some science textbooks present a large gap between particles in liquids and do not mention any attractive force between them (Riaz, 2004). Thereby, such an illustration confuses students and undermines their conceptual understanding of the PNM. Likewise, students are unable to comprehend the arrangement of particles in three physical states due to the abstract nature of the PNM (Adadan and Ataman, 2021). Furthermore, when particles are lined up, side by side without any space, students are unable to visualize how they continuously move (Nakhleh et al., 2005; Ayas et al., 2010). Students tend to carelessly attribute such macroscopic properties as colour, taste, or softness to microscopic particles (de Vos and Verdonk, 1996). Also, even if they scientifically grasp the PNM, they have deficiencies at applying it to novel situations (Haidar and Abraham, 1991; Tsai, 1999; Ayas et al., 2010).

Moreover, when considering Piagetian operational stages, educational level may prevent students’ understanding of the PNM (e.g., Nakhleh and Samarapungavan, 1999; Cokelez and Dumon, 2005; Boz, 2006; Gökdere and Çalık, 2010). In consideration of students’ cognitive thinking levels, the Turkish science curriculum moved the PNM to grade 6 from grade 4; in other words, to students at the formal operational stage, who are able to think more abstractly (Ministry of National Education, 2018). This movement was a result of the PNM-based research in the related literature. Around the world, national educational systems have recognized the need for abstract thinking abilities to learn the PNM, for instance, grades 6–8 in the USA (Next Generation Science Standards, 2017), grade 8 in Australia (Australian Curriculum, 2022), the Philippines (Kelly et al., 2020a) and North Macedonia (Kelly et al., 2020c), and grades 7–9 in Malaysia (Kelly et al., 2020b) and the UK (called Key Stage 3) (Department for Education, 2015). Moreover, as seen from Table S1 (see ESI†), the countries under investigation generally have similar academic performance indicators of the PNM, which prioritize achievement and conceptual understanding. These are: students are able to (a) know/understand the fundamental features of the PNM, (b) associate the PNM with daily-life issues or other science topics, and (c) explain the advanced features of the PNM.

Given students’ difficulties with abstract science concepts (e.g., the PNM), science educators have made efforts to improve students’ academic performance by bridging pre-existing knowledge to scientific ones (Tang and Abraham, 2016). The Learning in Science Project (LISP) and Children's Learning in Science (CLIS) project are educational milestones that have stimulated discussions about the importance and place of students’ questions, ideas and interactive teaching in school science (Freyberg and Osborne 1982; Biddulph and Osborne, 1985; Scott et al., 1987; Bell, 2005). Hence, they challenged students’ difficulties by illustrating how to integrate them into school science and overcome them. Other educational researchers have attempted to facilitate students’ academic achievement by addressing their conceptual understanding (called academic performance in the current paper) via different pedagogical approaches or models or methods (e.g., brain-based learning, computer-assisted instruction, conceptual change, context-based learning, cooperative learning, multiple representation and project-based learning). Fortunately, there have been some researchers who have undertaken the task of reviewing and synthesizing these studies. The following section highlights these earlier efforts.

Previous systematic reviews of the PNM

The science education literature has included several systematic reviews directly focusing on the PNM (Sözbilir, 2002; Özmen, 2013; Altay and Balım, 2021) and handling the PNM within the topic of matter (Hadenfeldt et al., 2014) and the unit of “matter and change” (Küçük and Yıldırım, 2017). Özmen (2013), who reviewed 79 papers from 1981 to 2012, reported that traditional instruction was ineffective in helping students to develop a sound understanding of matter, phase changes and the PNM. He pointed out that most of the new methods positively affected students’ conceptual understanding and remedied their alternative conceptions of the PNM and related concepts. Likewise, Sözbilir (2002), who reviewed nine papers of the PNM (called selected research papers), denoted that students from primary school to university level had difficulties in understanding the PNM. Similarly, Altay and Balım (2021), who examined 17 Turkish studies on the PNM published between 2002 and 2020, found that most of the studies focused on determining the misconceptions, whilst few studies concentrated on overcoming them and developing diagnostic tests. Furthermore, Hadenfeldt et al. (2014), who synthesized 82 articles (from 2003 to 2012) published in leading journals (e.g., Journal of Research in Science Teaching, International Journal of Science Education, Science Education, Studies in Science Education, and Research in Science Education), found that the reviewed studies showed a focus shift (e.g., from categorizing students’ conceptions to analysing their progression) on understanding matter. Many of the studies identified typical pathways that may facilitate their progression in understanding matter through the four categories identified by Liu and Lesniak (2005). Moreover, Küçük and Yıldırım (2017) analysed 33 Turkish papers (from 2000 to 2016) concentrating on teaching the concepts of the “Matter and Change” unit in secondary school (grades 4–8) and meta-analytically calculated their effect-sizes. They reported that the findings of Kruskal–Wallis H test showed no significant difference between moderator variables (the used methods, publication year and grade) and achievement.

Of these studies, only one study (Küçük and Yıldırım, 2017) employed a meta-analysis that focused on Turkish studies involving secondary school students. However, this study did not mention heterogeneity, publication bias, model fit and so forth. Furthermore, it failed to include international studies with different educational levels. The rest of the previous systematic reviews descriptively synthesized the studies and did not calculate their effect-sizes to compare the effectiveness levels in improving students’ academic performance. Given these issues, the relevant literature calls for the current study going over the extent to which the PNM-based intervention studies affect academic performance.

Rationale and significance of the study

Science educators, who are mainly concerned with students learning scientific concepts (e.g., PNM and matter) and teachers facilitating student learning (Liu and Lesniak, 2005), have thoroughly investigated students’ and teachers’ conceptions of science concepts (e.g., PNM and matter) (e.g., Talanquer, 2009; Özmen, 2013; Hadenfeldt et al., 2014). Furthermore, the related literature has tested several instructional methods to improve students’ academic performance and overcome their alternative conceptions. Likewise, some researchers have suggested teaching pathways and models to facilitate the transition from naïve through novice to scientific views and to better comprehend possible reasons for students’ learning difficulties (e.g., Talanquer, 2009; Hadenfeldt et al., 2014). Given that there have been many intervention efforts to help students transition from concrete to abstract thinkers, especially regarding the PNM, it would be beneficial to have a comprehensive overview from these studies about what might work for science education.

As noted, even though there have been some efforts to review and synthesize the findings from the PNM-based studies, the related literature needs a comprehensive meta-analysis to portray their effectiveness levels in improving students’ academic performance. Hence, science educators, science teachers and curriculum developers may have insights about which intervention types are more effective in teaching the PNM and developing students’ academic performance (meaning achievement and conceptual understanding). Furthermore, they may learn how educational level, as a moderator variable, influences students’ academic performance of the PNM. Thus, this meta-analysis not only meets recency of the related literature but also provides a broader sense of the PNM-based intervention studies by incorporating all types of papers published between 1992 and 2022. Hence, it intends to fill an important gap in the related literature and challenges limitations of previous systematic reviews. For example, Hadenfeldt et al. (2014) only preferred major journals, whose impact factors were higher than 1.0, while Küçük and Yıldırım (2017) and Altay and Balım (2021) only included Turkish papers. Moreover, researchers may probe extreme values/points in the meta-analysis for their future research. Overall, having recruited a meta-analysis method may shed more light on the consistency and efficiency of the PNM-based intervention studies.

The purpose and research questions

The purpose of this study is to examine the effectiveness of the PNM-based intervention studies in improving academic performance through a meta-analysis. The following questions guide the present study.

• How do the PNM-based intervention studies affect their participants’ academic performance?

• Do moderator variables (namely, educational level and type of intervention) positively influence the participants’ academic performance?

Methodology

Through a meta-analysis design, this study investigated the effectiveness of the PNM-based intervention studies in improving academic performance. Hence, it gathered the findings of the PNM-based research papers and compared them via a statistical analysis (e.g., Üstün and Eryılmaz, 2014; Karadag, 2020). That is, calculating effect-sizes of the experimental papers, the current study attempted to examine the practical significance of the PNM-based interventions (e.g., Borenstein et al., 2009; Ellis, 2010; Üstün and Eryılmaz, 2014). Because meta-analysis gives an opportunity for researchers to find effects or relationships between dependent (e.g., academic performance) and independent (e.g., PNM-based instructional interventions) variables, the current study endeavoured to handle related papers in a very organized and systematic way (Lipsey and Wilson, 2001; Üstün and Eryılmaz, 2014; Sezen-Vekli and Çalik, 2023).

Sample and selection criteria

The research involved finding PNM-based intervention studies from a variety of well-known databases: Academic Search Complete, ERIC, Open Dissertations, Education Source, ULAKBIM, Springer Link, Taylor & Francis, Wiley Online Library Full Collection, Science Direct, ProQuest Dissertations and Theses Global, Google Scholar, Scopus, and Council of Higher Education Thesis Center in Türkiye. The authors employed some keywords (‘particulate nature of matter or particle theory or particulate theory or submicroscopic level of matter’ and ‘experimental or intervention or treatment’ and ‘chemistry education or science education’) via the keyword patterns from the abstracts (see Appendices). Also, the authors implemented a manual search of the related journals and dissertations by examining the references of the papers. Thus, they identified some missing papers (n = 6). Care was taken to avoid duplications, because some papers appeared in more than one database or some dissertations were published in journals as research papers. Therefore, the authors excluded 31 papers due to duplications in databases and research papers from the dissertations. Then, they read all papers to implement the inclusion criteria (PNM-based intervention, learning outcomes—conceptual understanding and achievement, quasi-experimental design and publication language—Turkish and English). Thereby, the authors eliminated 31 papers with pre-experimental design, different dependent variables (e.g., attitudes), and different languages (e.g., Arabic, Chinese, Greek and Korean). Later, they examined statistical data of each paper to calculate any effect-size(s). Hence, they excluded 34 papers that lacked sufficient data (e.g., mean, standard deviation, sample size, paired p-value, and paired t-value) for meta-analysis. For example, some of the papers only provided the findings of non-parametric analysis (e.g., mean rank, sum of ranks, and U-value), which are insufficient for running a meta-analysis. Similarly, some of them, which carried out the teaching intervention, only included qualitative data from pre- and post-interview or limited descriptive statistical results (e.g., frequency and percentage) from open-ended questions. Finally, the authors identified 66 papers (44 dissertations, 21 articles, and one proceeding) that focused on the effect of the PNM-based intervention studies on academic performance. Most of the papers were published from 2010 to 2019; one paper was published between 1992 and 1999 and five papers were from 2000 to 2009. Also, years 2020 and 2022 included 9 papers. Fig. 1 summarizes the selection process.

Fig. 1 Flow chart of the selection process.

Nine of the papers had several experimental groups (Williamson, 1992; Kaya, 2005; Doymuş et al., 2009; Çavdar, 2016; Çavdar et al., 2016; Okumuş, 2017; Okumuş et al., 2017; Nuic and Glažar, 2020; Ozyalcin, 2020), where the same teaching design was implemented at different cohorts or different schools at rural and downtown locations. In that case, the authors inserted their statistical values (e.g., mean, standard deviation, and sample size for the experimental and control groups) into comprehensive meta-analysis (CMA) statistics software and exploited the ‘subgroups within the study’ option to calculate combined effect-sizes (by using the study as the unit of analysis) given the type of intervention. For example, because Williamson (1992), Kaya (2005) and Nuic and Glažar (2020) followed the same intervention type (inquiry-based learning, computer-assisted instruction and computer-assisted instruction, respectively) for their experimental groups, the researchers produced only one effect-size for the type of intervention. However, the rest of them also incorporated different intervention types for their experimental groups. Thus, they were divided into individual studies given their intervention types. For example, the authors handled Çavdar (2016), Çavdar et al. (2016), Okumuş (2017), Okumuş et al. (2017), Doymuş et al. (2009), and Ozyalcin (2020) (including cooperative learning and computer-assisted instruction or enriched learning environment with different methods as type of intervention) as two individual studies, which compared each of the experimental groups with the control (comparison) one. In brief, it intended to minimize individual studies to overcome inflating impacts of these studies on the overall effect-size (see ESI,† for the list of papers in the meta-analysis).

Coding procedure

Since coding, as a data extracting process, needs clear and appropriate data (Karadag, 2020), the authors created and applied the following coding form to the papers: reference of the paper, sample size, grade, dependent and independent variables and quantitative values—mean scores, standard deviation, t, and p. The inter-rater consistency value was 0.82 after the papers were independently coded. The authors discussed any disagreement and resolved through negotiation.

Calculating effect-sizes

Effect-size, as a standard measurement value, determines the strength and direction of the relationship between two variables or degree of practical effect (Borenstein et al. 2009). Meta-analysis studies usually choose Cohen's d and Hedges’ g to calculate the effect-size. However, Hedges’ g is more accurate and less biased with relatively small sample sizes as compared to Cohen's d (Borenstein et al., 2009; Kansızoğlu, 2017; Güler et al., 2022). Therefore, the authors used Hedges’ g calculation for the meta-analysis and initially inserted all statistical data for the papers into an Excel sheet and then imported them into comprehensive meta-analysis (CMA) statistics software. Hence, the inclusion of the data received via email and the decision to report combined and two effect-sizes from the aforementioned nine papers meant that the corpus of the papers under investigation consisted of 72 studies from 66 unique sources. While calculating Hedges’ g value, the papers supplied different data:

• 62 individual studies (from 56 papers, some of which included several experimental groups and were imported as individual studies) included mean scores, standard deviations and sample sizes for the experimental and control groups.

• Six studies (six papers) supplied mean scores and sample sizes of the experimental and control groups and independent groups’ p-values.

• Three studies (three papers) reported mean scores and sample sizes for the experimental and control groups and independent groups’ t-values.

• One study (one paper) contained mean scores and sample sizes for the experimental and control groups, pre-post correlation and F for difference in change.

The effect-size values were interpreted using the following range: 0.14 and below (negligible); 0.15–0.39 (low); 0.40–0.74 (medium); 0.75–1.09 (large); 1.10–1.44 (very large); and 1.45 and above (perfectly huge) (Güler et al., 2022).

meta-Analysis model

To decide between two main meta-analysis models, the fixed-effects model and random-effects model, it is necessary to consider how the characteristics of the papers match with the meta-analysis model and its prerequisites (e.g., Borenstein et al. 2009; Karadag, 2020). The fixed-effects model assumes that all studies in the meta-analysis have only one true effect-size and all differences in the observed effects come from only sampling error (Üstün and Eryılmaz, 2014). According to Hedges and Vevea (1998), the fixed-effects model has the advantage of making inferences about the relevant parameters in the papers. Furthermore, considering the papers with the same functionality, this model estimates the effect-size for only one population (Karadag, 2020). On the other hand, the random-effects model alleges that true effect-size could be varied because of some moderator variables (e.g., age, grade, and sample-size) (e.g., Borenstein et al., 2009; Üstün and Eryılmaz, 2014). Namely, if the papers do not possess equal functionality, the random-effects model can be used to estimate and generalize the effect-size for greater populations (Karadag, 2020).

Given the foregoing explanations, the researchers looked for the presence of heterogeneity by calculating the Q-value and I² test. As seen from Table 1, the Q-value was found to be 911.966, which is higher than the critical value of 91.670 (with 71 degrees of freedom for 95% confidence interval). Furthermore, the p-value was statistically less than 0.05. This means that the current meta-analysis initially met the heterogeneity needed to employ the random-effects model. Moreover, the researchers examined the I² value before fully deciding on the use of the random-effects model (Borenstein et al., 2009). Given the heterogeneity criteria (low heterogeneity for nearly 25% of the I² value, medium heterogeneity for 50% of the I² value and high heterogeneity for 75% of the I² value) suggested by Higgins et al. (2003), the I² value was calculated to be 92.215. This means that the present meta-analysis met the heterogeneity criterion to run the random-effects model.

Table 1 The findings of the heterogeneity test

Model	Number of studies	Effect-size and 95% confidence interval					Test of null (2-tail)		Heterogeneity				Tau-squared
		Point estimate	Standard error	Variance	Lower limit	Upper limit	Z-Value	P-Value	Q-Value	df (Q)	P-Value	I-Squared	Tau-Squared	Standard error	Variance	Tau
Fixed	72	0.716	0.026	0.001	0.664	0.767	27.021	0.000	911.966	71	0.000	92.215	0.610	0.158	0.025	0.781
Random	72	0.900	0.098	0.010	0.709	1.092	9.210	0.000

---

Consequently, the researchers applied the random-effects model in the meta-analysis processes to calculate the effect-sizes via a comprehensive meta-analysis (CMA V2) software. Finally, they identified two moderator variables to respond to the second research question: educational level and type of intervention. The ‘educational level’ variable covers K–8 (kindergarten to grade 8), high school (grades 9–12) and university (undergraduate or bachelor). Also, the ‘type of intervention’ one includes multiple representation, inquiry-based learning, enriched learning environment with different methods, brain-based learning, conceptual change and so forth (see Appendices for descriptions of the ‘type of intervention’ moderator and Table S2 in the ESI†).

Publication bias

The authors acknowledged that some researchers criticize meta-analysis for having publication bias (e.g., Üstün and Eryılmaz, 2014). Therefore, to avoid this bias, this study employed the visual inspection of a funnel plot, the trim-and-fill method, the classic fail-safe N and Orwin's fail-safe N. Thereby, it objectively presented and interpreted the findings of publication bias (for instance, the number of missing studies needed to bring Hedges’ g under 0.1 or p value to > alpha).

Findings

The findings of publication bias

The results of the funnel plot can be found in Fig. 2; they indicate that the relationship between standard error and effect-size has an asymmetrical structure (e.g., Karadag, 2020). The fact that the papers with larger standard error (smaller sample size) have larger effects could be viewed as evidence of publication bias (Rahman and Lewis, 2020). With evidence of an asymmetric distribution in the funnel plot, Duval and Tweedie's trim and fill test was applied to further examine the publication bias for the random-effects model. As can be seen from Table 2, there was a difference between the observed and adjusted values. Fig. 3 (red circles in the funnel plot) presents counterpart data to offset this asymmetry or to overcome the impact of observed asymmetry on the overall results. But, the difference between the observed and adjusted estimates was 19.55%, which is less than the negligible cut-off value (20%) suggested by Kepes et al. (2012), Vevea et al. (2019), and Chang et al. (2022). This means that the publication bias can be neglected to run the current meta-analysis for the effectiveness of the PNM-based intervention studies. To further investigate any publication bias, the classic fail-safe N and Orwin's fail-safe N values were calculated. As seen from Table 3, the classic fail-safe N value was 4895. This means that 4823 additional papers with non-significant findings would be necessary to nullify the effect of the PNM-based intervention studies on the academic performance. That is, considering the formula [N/(5k + 10)] (where k means the total number of the studies in the meta-analysis), this ratio for the current meta-analysis was 13.035, which is higher than the cut-off point (1.00) offered by Mullen et al. (2001). So, the analysis ensured that the data selection process was robust and not affected by publication bias. A final confirmation about this issue can be seen in Orwin's fail-safe N value (see Table 4), which was 444. This value means that 372 additional papers with effect-size of 0.00 000 would be necessary to make the mean effect of this meta-analysis trivial (Üstün and Eryılmaz, 2014). In brief, the foregoing values mean that the current meta-analysis had no evidence of publication bias for the papers under investigation.

Fig. 2 Funnel plot of standard error by effect-size.

Table 2 The finding of Duval and Tweedie's trim and fill test for the random-effects model

	Studies trimmed	Point estimate	Confidence interval (CI)		Q value
			Lower limit	Upper limit
Observed values		0.90 029	0.70 871	1.09 188	911.96 552
Adjusted values	10	1.11 914	0.89 075	1.34 752	1706.22 236

---

Fig. 3 Funnel plot of standard error by effect-size after Duval and Tweedie's trim and fill test. (Note: red circles mean a counterpart data to offset the asymmetry or overcome the impact of observed asymmetry on the overall results).

Table 3 The findings of the classic fail-safe N test

Z-Value for observed studies	28.25 762
P-Value for observed studies	0.00 000
Alpha	0.05 000
Tails	2.00 000
Z for alpha	1.95 996
Number of observed studies	72
Number of missing studies that would bring p value to > alpha	4895

---

Table 4 Findings of Orwin's fail-safe N test

Hedges’ g in observed studies	0.71 552
Criterion for a ‘trivial’ Hedges’ g	0.10 000
Mean Hedges’ g in missing studies	0.00 000
Number missing studies needed to bring Hedges’ g under 0.1	444

---

The overall impact of PNM-based intervention studies on academic performance

Visual representations of the effect-sizes illustrate the findings of meta-analysis effectively (Lipsey and Wilson, 2001). Therefore, this research presents a stem and leaf plot (Table 5), and forest plot (Fig. 4) to summarize the effect-sizes. As seen from Table 5, 23 of the effect-size values were higher than 1.00. Also, 29 of the effect-sizes ranged from 0.5 to 1, whilst 20 of them were at 0.5 and below. As seen from Fig. 4, the meta-analysis calculated 72 effect-sizes from 66 papers. According to the forest plot, while 56 of the effect-sizes were significant (p < 0.05), 16 of them were non-significant (p > 0.05). Given effect-size classification offered by Güler et al. (2022), nine of the effect-sizes in the meta-analysis fell into the negligible effect. The effect-sizes from five papers were classified under the low effect, whilst 20 of them were labelled beneath the medium one. Finally, the frequencies of the effect-sizes categorized under the large, very large and perfectly huge effects were 16, 6 and 16, respectively. Also, the overall effect-size for the random-effects model was found to be 0.900, which means that the PNM-based intervention studies have a large effect for academic performance.

Table 5 Stem and leaf plot of the effect-sizes

Frequency	Stem	Leaf
1	−2	6
1	−0	7
4	−0	0133
14	0	00 011 233 444 444
29	0	55 556 666 666 677 777 777 888 888 899
8	1	01 234 444
3	1	557
7	2	0 011 234
2	2	67
3	3	022

---

Fig. 4 Forest plot of the papers included in the meta-analysis.

The findings of moderator variable (educational level)

The current meta-analysis viewed the ‘educational level’ as a moderator variable and covered K–8 (kindergarten to grade 8), high school (grades 9–12) and university (undergraduate or bachelor). As can be seen from Table 6, the effect-size difference between the educational levels was found to be statistically significant (Q-value = 6.122; df = 2; and p < 0.05). This means that educational level acted as a moderator variable for the positive effect of the PNM-based intervention studies on the participants’ academic performance. As a matter of fact, K–8 and high school as the educational levels were significant in academic performance (p < 0.05), while university as an educational level was non-significant in academic performance (p > 0.05). Moreover, Hedges’ g values indicated different classification levels based on the criteria suggested by Güler et al. (2022). That is, the mean effect-size of K–8 as an educational level fell into the large effect. Also, the mean effect-size of high school was categorized under the medium effect, whilst that for university fell into the low one.

Table 6 The findings of moderator analysis for educational level

Educational level	N	Point estimate	Standard error	Confidence interval (95%)		Z-Value	p-Value	Q-Value	df (Q)	p-Value
				Lower limit	Upper limit
Note: other was omitted since it only included one paper.
K–8	55	1.043	0.117	0.813	1.272	8.906	0.000
High school	8	0.603	0.305	0.004	1.202	1.975	0.048
University	8	0.323	0.300	−0.266	0.911	1.075	0.282
Total between								6.122	2	0.047

---

The findings of moderator variable (type of intervention)

As can be seen from Table 7, the effect-size difference between the intervention types was found to be statistically significant (Q-value= 20.448; df = 10; and p < 0.05). This means that the type of intervention acted as a moderator variable for the positive effect of the PNM-based intervention studies on the participants’ academic performance. However, ‘brain-based learning, conceptual change and context-based learning’ were non-significant in academic performance (p > 0.05). Moreover, the mean effect-size for multiple representation had the highest value (Hedges’ g = 2.193) for the intervention types. They also showed various classification levels in regard to criteria suggested by Güler et al. (2022). That is, Hedges’ g values for multiple intelligence, multiple representation and problem-based learning fell into the perfectly huge effect, while the values for computer-assisted instruction, constructivist learning environment, context-based learning, enriched learning environment with different methods, and inquiry-based learning were labelled under the large one. The values for brain-based learning, and cooperative learning were categorized under the medium effect, whilst that of conceptual change was classified under the low one.

Table 7 The findings of moderator analysis for type of intervention

Type of intervention	N	Point estimate	Standard error	Confidence interval (95%)		Z-Value	p-Value	Q-Value	df (Q)	p-Value
				Lower limit	Upper limit
Note: other (N = 4) was omitted since it only included one each paper from project-based learning, ARCS motivation model, family involvement and use of posters.
Brain-based learning	2	0.575	0.557	−0.517	1.666	1.032	0.302
Computer-assisted instruction	10	0.884	0.253	0.388	1.381	3.492	0.000
Conceptual change	4	0.174	0.405	−0.620	0.967	0.428	0.668
Constructivist learning environment	8	1.020	0.287	0.459	1.582	3.560	0.000
Context-based learning	4	0.757	0.407	−0.041	1.555	1.859	0.063
Cooperative learning	8	0.592	0.285	0.032	1.151	2.074	0.038
Enriched learning environment with different methods	14	0.746	0.217	0.322	1.170	3.445	0.001
Inquiry-based learning	9	0.955	0.273	0.421	1.489	3.505	0.000
Multiple intelligence	2	1.483	0.594	0.319	2.647	2.497	0.013
Multiple representation	5	2.193	0.381	1.446	2.940	5.752	0.000
Problem-based learning	2	1.831	0.597	0.661	3.001	3.068	0.002
Total between								20.448	10	0.025

---

Discussion

The overall effect-size value of the studies (which fell into the large effect) (see Fig. 4) means that the interventions are effective in improving the PNM-based academic performance. The intervention studies seem to have served their targeted aims. Indeed, an improvement in academic performance is an expected outcome after the teaching intervention or experimental design.

It is important to point out that some of the intervention studies were classified under the perfectly huge effect. This may result from the scope and content of the teaching intervention. For example, Çavdar (2016), who used several combinations of cooperative learning, modelling and seven principles for good practice in varied experimental groups, produced the perfectly huge effects. This means that embedding cooperative learning and modelling within seven principles for good practice has resulted in the highest effect-size value. In other words, enriching the learning environment with different methods seems to have increased the effectiveness of the teaching intervention vis-a-vis the use of separate methods. As a matter of fact, Ataman Mortaş (2011), Kaçar (2012), Öksüz (2019) and Adadan and Ataman (2021), who used various methods together (e.g., augmented reality, animations, simulations, conceptual change texts, problem-based learning, modelling, and multiple representations), provided good evidence for the role of the enriched learning environment. This means that accompanying different methods/strategies with each other helps students visualize, conceptualize and explain the PNM at sub-microscopic level.

The effect of educational level on academic performance

As seen from Table 6, the educational level, as a moderator variable, positively affected the PNM-based academic performance since there were statistically significant differences between the effect-size values of the educational levels (Q-value= 6.122; df = 2; and p < 0.05). This may be viewed as an indicator for the importance of educational level in teaching the PNM. Furthermore, this result is inconsistent with Küçük and Yıldırım's (2017)meta-analysis study that found no difference between grades 4–8, as educational level, to teach the PNM-related unit (matter and change).

It should be acknowledged that the majority of the intervention studies focused on K–8. This may stem from explicitly involving the PNM as a core concept in K–8 science curricula (Department for Education, 2015; Next Generation Science Standards, 2017; Ministry of National Education, 2018; Australian Curriculum, 2022). As compared with high school and university levels, the overall effect-size value of K–8 points out that this educational level is appropriate to handle and teach the fundamental dimensions and/or features of the PNM. On the other hand, this means that the more the dimensions or features of the PNM advance (e.g., high school and university cover advanced level of the PNM, which contains more detail and modelling at the later stages in education), the more the overall-effect-size value of educational level decreases. This may result from different instructional goals or representative order (from macroscopic level to sub-microscopic and then to symbolic) of the PNM (Singer et al., 2003). For example, Harrison and Treagust (2002) recommend that secondary school chemistry should emphasize only fundamental dimensions and/or features of the PNM, while its advanced dimensions and/or features should be taught in high school chemistry. Likewise, Liu and Lesniak (2005) suggest using specialized content and teaching methods for teaching the PNM to varied educational levels.

When the relative effect-size of each level is seen as an indicator for an improvement in students’ understanding of the PNM across educational level, the current meta-analysis refutes the established expectations that an increase in educational level should evolve and improve their understanding of the PNM, as an outcome of epistemological and ontological growth. This may come from many of the intervention studies in the meta-analysis that targeted K–8 students. Unfortunately, because of the low number of studies conducted for the “high school and university” categories, it is still unclear if an increase in educational level explicitly results in an improvement in students’ understanding of the PNM. Moreover, this may stem from the type of PNM-based assessment at each level. For example, the studies with undergraduate students (university level) generally preferred using an open-ended questionnaire (e.g., Williamson, 1992; Doymuş et al., 2009; Çavdar et al., 2016; Okumuş et al., 2017). Interestingly, the studies with K–8 and high school students mostly used multiple-choice questions to assess the students’ conceptual understanding and achievement levels of the PNM (e.g., Bunce and Gabel, 2002; Özmen, 2011; Barthlow and Watson, 2014; Nuic and Glažar, 2020; Eroğlu and Bektaş, 2022) (see Table S3 in the ESI†). That is, students, who respond to multiple-choice questions by marking one of the offered choices, may have easily reflected their PNM-based academic performance. On the contrary, open-ended questions, which require them to state their conceptual understanding and achievement levels of the PNM in words or drawings, may have acted as an influential role at probing PNM-based academic performance.

The effect of intervention type on academic performance

As can be seen from Table 7, type of intervention, as a moderator variable, positively influenced PNM-based academic performance (Q-value = 20.448; df = 10; and p < 0.05). The mean effect-size for multiple representations had the highest value (Hedges’ g = 2.193) amongst the intervention types; this may come from matching the nature of the PNM with common features of the multiple representations. In particular, as a part of a multiple representation intervention, students are asked to create their own representations visualizing, modelling and explaining the PNM concepts (Bunce and Gabel, 2002; Adadan et al., 2010). Because multiple representation requests students to transfer their gained knowledge or understanding of the PNM to new models and representations, it necessitates re-considering and synthesizing their learning. Therefore, this type of intervention may have facilitated participants’ academic performance (Wu and Puntambekar, 2012; Namdar and Shen, 2018) and resulted in the perfectly huge effect.

Even though the mean effect-sizes of brain-based learning, conceptual change and context-based learning fell into the perfectly huge effect, the results of the interventions were non-significant for the effect of the ‘type of intervention’ variable on academic performance (p > 0.05). This may result from limited intervention studies employing these methods and strategies. Similarly, it might be interpreted that the effect-sizes of the intervention studies using brain-based learning, conceptual change and context-based learning were at different-levels or included negative values. For example, two studies on brain-based learning had different effect-sizes: very large effect (1.424) for Aktas and Bilgin (2015) and negligible effect (−0.31) for Alaca (2014).

The mean effect-sizes of computer-assisted instruction (Hedges’ g = 0.884), constructivist learning environment (Hedges’ g = 1.020), context-based learning (Hedges’ g = 0.757), enriched learning environment with different methods (Hedges’ g = 0.746), and inquiry-based learning (Hedges’ g = 0.955) were classified under the large effect. These interventions, except for context-based learning, were significant in academic performance (p < 0.05) (see Table 7). This means that these intervention types seem to have enabled participants to improve PNM-based academic performance. This is not surprising as their theoretical frameworks prescribe engaging participants in building their own learning, challenging their alternative conceptions to replace them with scientific ones, handling their pre-existing knowledge in designing the content. Consequently, the scope of the intervention is designed to improve students’ knowledge and conceptual understanding.

The fact that all these different types of interventions fell into the same effect-level may be because they relate to the constructivist learning theory. This theory has influenced context-based learning, and enriched the learning environment with different methods and inquiry-based learning. Since most of the science curricula (e.g., the USA, England, and Türkiye) have been underpinned by constructivist learning theory, some of the intervention studies (e.g., Bektaş, 2003, 2011; Kenan, 2014) attempted to test it within the PNM or PNM-based unit.

Several of the intervention studies (e.g., Özmen, 2011; Çavdar, 2016; Okumuş, 2017) employed the enriched learning environment using some combinations of different methods (e.g., cooperative learning, modelling, multiple representations, augmented reality, argumentation, animation and conceptual change text) and attempted to enhance the effectiveness of their interventions. Given the overall effect-size, they seem to have accomplished this goal to some extent. Because inquiry-based learning involves participant engagement, stimulates their learning curiosity and encourages them to inquire about any phenomenon or concept or knowledge in the light of evidence, some studies (e.g., Barthlow and Watson, 2014; Çolak, 2014; Özkanbaş, 2018; Özkanbaş and Taştan Kırık, 2020) focused on the effect of inquiry-based learning on PNM-based academic performance. The findings of the meta-analysis revealed that inquiry-based learning, as the type of intervention, fostered the participants to think about the PNM via evidence-based learning and empowered their PNM-based academic performance.

The findings of the meta-analysis had mixed results for computer-assisted instruction. Even though this type of intervention provides instant feedback and gives students learning experiences to make unfamiliar and abstract concepts (such as PNM) more concrete and familiar (Yang et al., 2012; Gökçe and Saraçoğlu, 2018), the overall effect-size of the computer-assisted instruction was lower than those for some of the intervention types (e.g., constructivist learning environment, inquiry-based learning, multiple intelligence, and multiple representation). This means that the computer-based instruction is somewhat effective at developing PNM-based academic performance (Williamson, 1992; Tang and Abraham, 2016; Ateş, 2018). However, some of the intervention studies on computer-assisted instruction (Aydost, 2011; Öksüz, 2019) yielded the perfectly huge effect. This may result from the implementation or duration of the computer-assisted instruction.

Regarding context-based learning, even though Rusçuklu (2017) reported the very large effect for PNM-based academic performance, other intervention studies on context-based learning (e.g., Kirman Bilgin et al., 2017; Büyük Kuloğlu, 2019) seem to have had difficulties at finding proper context for the PNM. This supports the idea that daily life examples may not be enough to attain context-based learning (Sözbilir et al., 2007). Also, even though teachers and educators design a proper context to associate the PNM with daily life, students may not find it meaningful or may have a deficiency at linking this context with the PNM. Therefore, this may explain why the overall effect-size was the large effect, but non-significant for the effect of context-based learning on academic performance (p > 0.05).

The mean effect-sizes of brain-based learning (Hedges’ g = 0.575) and cooperative learning (Hedges’ g = 0.592) were classified under the medium effect, while that for conceptual change (Hedges’ g = 0.174) was labelled under the low one. Furthermore, only one of them (cooperative learning) was significant in academic performance in terms of the type of intervention (p < 0.05) (see Table 7). This means that brain-based learning and cooperative learning as the type of intervention have a reasonable effect on PNM-based academic performance. Interestingly, even though conceptual change attempts to overcome students’ alternative conceptions or misconceptions and replace them with scientific ones, the mean effect of conceptual change studies fell into the low effect. This may stem from the features of the control or comparison group, which generally follows the national science curriculum underpinned by a constructivist learning environment (e.g., 5E learning model). Also, this may come from the type of PNM-based assessment. That is, Eroğlu (2010), Beerenwinkel et al. (2011), Ceylan (2015), and Ezema et al. (2022) used multiple-choice questions but Ezema et al. (2022) also exploited two-tier questions (see Table S3 at ESI†). Thus, the type of PNM-based assessment and features of the intervention groups may have resulted in a negative value for the study by Ezema et al. (2022). Of course, such a case may have influenced the mean effect-size of conceptual change studies.

Despite the fact that brain-based learning asserts that each student uses different cerebral hemispheres while processing information in the brain (Davis, 2004; Aktas and Bilgin, 2015; Alaca, 2014), few studies have concentrated on it or its derived models to integrate individual differences (e.g., learning styles and processing functions of the brain) into PNM-based interventions. Even though brain-based learning gives an opportunity for students to understand, model, visualize and apply the PNM, the studies on brain-based learning possessed inconsistent effect-sizes (e.g., the very large effect for Aktas and Bilgin, 2015 and negligible one for Alaca, 2014). The finding may come from a mismatch or limited match between the nature of the PNM and intervention type. Unfortunately, because there are few studies that used these intervention types, the researchers were prevented from accurately interpreting how they serve as a moderator variable for effect-sizes.

Similar results were found for cooperative learning, which includes jigsaw puzzle, home group, working group, collaboration and competition between groups (e.g., Çavdar, 2016; Okumuş, 2017). Therefore, any intervention needs a match between the nature of the topic (e.g., PNM) and features of instructional methods, strategies and approaches (i.e., cooperative learning). For example, of these studies, Çavdar (2016) had the perfectly huge effect and Okumuş (2017) possessed the large effect. This also indicates that some studies seem to have intertwined the nature of the PNM with features of cooperative learning. However, the mean effect-size of cooperative learning points to the limited match for PNM-based academic performance.

Conclusions and implications

Given the mean effect-size of this meta-analysis, it can be concluded that the PNM-based intervention studies are effective at improving the participants’ academic performance. This difference is mostly dependent on matching the nature of the PNM with the features of the intervention type. For instance, the fact that the use of multiple representations fell into the perfectly huge effect supports this match. Furthermore, participant engagement and preparedness (e.g., formal operational term) that clearly influence the participants’ academic performance levels play a pivotal role at bridging the nature of the PNM to the intervention type.

This study confirmed that the fundamental features or dimensions of the PNM can be taught at K–8. Also, this result advocates the content and context of the PNM in science curricula in terms of educational level. Since the PNM is a core concept, waiting to introduce it until high school or university may allow too much time for students to form misunderstandings and incorrect knowledge of fundamental science concepts. For this reason, rather than continuing to debate when to introduce students to the PNM, discussions should focus on the type of intervention and its scope that is the best for K–8 students.

It is still important, however, to explore PNM-based interventions for high school, undergraduate students, and teachers. Since the meta-analysis includes few extreme values, further research should be undertaken to test the effectiveness of the intervention types on academic performance. For example, further studies may investigate in-service teachers’ pedagogical content knowledge and conceptual understanding of the PNM after the PNM-based interventions.

In addition to educational levels, further investigation of different intervention types is needed. For example, this meta-analysis included very few studies on project-based learning, family involvement activities, the use of posters and the ARCS motivation model. Furthermore, given the mean effect-size of the context-based learning, future research should go beyond daily-life experiences and provide better context(s) to teach the PNM. Hence, perceiving the relationship(s) between the PNM and context may result in better improvements in students’ academic performance. Similarly, taking the mean effect-size of cooperative learning into consideration, its core dynamics should be well designed and embedded within any course plan or teaching intervention to result in better science learning. To provide better insights into the effectiveness of these interventions, more researchers need to apply them and report their findings.

The current study excluded a notable number of papers that lacked sufficient data. This may be seen as the limitation of the paper. Even though the researchers sent several emails to the authors of these papers, they did not respond to this request or were unable to provide further information/data from their papers. For this reason, science educators ought to pay more attention to include sufficient data or effect-size value for their papers to facilitate dissemination, generalization and comparison of their results. Moreover, the fact that the current study only included Turkish and English papers may be viewed as another limitation of the study. Unfortunately, the authors proposed to contain the papers published in different languages (e.g., Arabic, Chinese, Greek, and Korean), but their English abstracts were not informative to get sufficient data. Because the authors are unfamiliar with these languages, they excluded these papers from the meta-analysis instead of spontaneously translating them into English or Turkish. For this reason, it is better to add an extensive English or easily translatable abstract to effectively disseminate the results. Since the papers with pre-experimental design were very small, the current study excluded them from the meta-analysis. Therefore, future research may handle types of experimental design to compare the effectiveness of pre-experimental and quasi-experimental designs with each other.

Conflicts of interest

There are no conflicts to declare.

Appendices

A. Keyword patterns used in the database search

Pattern number	Keywords
Pattern 1	Particulate nature of matter, experimental and chemistry education or science education
Pattern 2	Particulate nature of matter, treatment and chemistry education or science education
Pattern 3	Particulate nature of matter, intervention and chemistry education or science education
Pattern 4	Particle theory, experimental and chemistry education or science education
Pattern 5	Particle theory, treatment and chemistry education or science education
Pattern 6	Particle theory, intervention and chemistry education or science education
Pattern 7	Particulate theory, experimental and chemistry education or science education
Pattern 8	Particulate theory, treatment and chemistry education or science education
Pattern 9	Particulate theory, intervention and chemistry education or science education
Pattern 10	Submicroscopic level of matter, experimental and chemistry education or science education
Pattern 11	Submicroscopic level of matter, treatment and chemistry education or science education
Pattern 12	Submicroscopic level of matter, intervention and chemistry education or science education

B. Descriptions of the ‘type of intervention’ moderator

Type of intervention	Descriptions
Brain-based learning	Includes strategies and principles derived from an understanding of the brain and depicts how it handles any information to accomplish meaningful learning (Jensen, 1995). By incorporating hands-on learning strategies and active learning processes in a social environment, it purposes to increase the functions of the brain and neural connections that enable students to retain newly generated knowledge in long-term memory (Caine and Caine, 1991; Degen, 2014).
Computer-assisted instruction	Refers to the use of computers and computer-based applications (e.g., animation, simulation, educational games, web-based learning environment) in the learning–teaching process to facilitate student learning (Bayraktar, 2001; Tang and Abraham, 2016). Thus, it serves the goals of science education by modelling any scientific phenomenon at sub-microscopic level and enhancing students’ capacities of learning via enriched technological tools/components (e.g., audio, sound, visualization, interaction, feedback and hypertext) (Doymuş et al., 2009; Tang and Abraham, 2016; Nuic and Glažar, 2020).
Conceptual change	Intends to replace alternative conceptions or misconceptions with scientifically accepted ones and integrate new conceptions with pre-existing ones (Hewson and Hewson, 1983; Çalık et al., 2009). Thus, it supports meaningful learning and a sound understanding of science concepts that play a generative or organizing role in thought (Guzzetti et al., 1993; Beerenwinkel et al., 2011; Demircioğlu, 2017).
Constructivist learning environment	Claims that learning is an interaction between pre-existing knowledge and new knowledge. Furthermore, students, who actively participate in the learning process and take their own responsibility for learning, construct new knowledge in light of their prior conceptions (Guzzetti et al., 1993). To involve essential features of constructivism (e.g., eliciting prior knowledge, creating cognitive dissonance, application of new knowledge with feedback, and reflection on learning) in the learning process, such instructional models as learning cycle, generative learning model, 5E learning model are offered (Baviskar et al., 2009).
Context-based learning	Aims at making connections between real life and the scientific content of school science courses (Ültay and Çalık, 2012). Hence, it not only develops students’ curiosity about the natural world (Bennett et al. 2005; Bulte et al. 2006), but also makes students become more aware of the connection between chemistry/science and their daily lives. Also, it actively engages students in their own learning processes (Stolk et al. 2009) to promote their conceptual understanding and induce meanings using contexts. Thus, it creates a ‘need-to-know’ basis for learning scientific content (Bennett et al. 2005; Bulte et al. 2006; Ültay and Çalık, 2012).
Cooperative learning	Requires students to work in small cooperative groups on a common task, which is in harmony with the goals of the science curriculum. Thus, an effective cooperative learning includes such essential features as positive interdependence, accountability, promotive interactions, teaching interpersonal skills and group processing (Johnson et al., 1998; Rahman and Lewis, 2020). Phrased differently, it not only fosters students to have intellectual conversations about problem-solving skills and creative thinking, but also increases group interaction to enhance their own learning (e.g., achievement and conceptual understanding) (Johnson and Johnson, 1990).
Enriched learning environment with different methods	Refers to the use of some combinations of different methods (e.g., cooperative learning, modelling, multiple representations, augmented reality, argumentation, animation, conceptual change text, concept cartoon and stories) to increase the effectiveness of any intervention (Özmen, 2011; Kenan, 2014; Çavdar, 2016; Türkoguz and Ercan, 2022). Hence, it merges the advantages of different methods into an enriched teaching intervention to meet various learning styles and pose students’ capacities of learning. For this reason, it purposes to result in better learning outcomes, e.g., achievement and conceptual understanding (Bağ and Çalık, 2022; Er Nas et al., 2022; Türkoguz and Ercan, 2022).
Inquiry-based learning	Refers to a process of enquiry that asks students to answer a question or solve a problem by conducting experiments, collecting and analysing data, and drawing conclusions (Orosz et al., 2022). Thus, they are able to acquire new knowledge and inquiry skills, and develop attitudes towards science and understand the nature of science as the outcomes of a student-centred method (Schwartz et al., 2004; Kaya, 2005; Bell et al., 2010; Barthlow and Watson, 2014; Villagonzalo, 2014; Pedaste et al., 2015).
Multiple intelligence	Views intelligence as the ability to solve problems or create products that are valued within one or more cultural settings. It asserts that each person has different types of intelligence (visual/spatial, verbal/linguistic, musical/rhythmic, logical/mathematical, bodily/kinaesthetic, interpersonal, intrapersonal, and naturalistic) that can be cultivated for attaining better learning (Gardner, 1993). Even though all people have all eight intelligence areas, some of them are more dominant over others. Therefore, students’ intelligence types should be considered to develop multiple intelligence-based curricula, lesson plans, and assessments. Hence, teachers are able to meet their students’ individual needs based on their different capacities to learn (Gardner, 1993; Akyol, 2007; Ateş, 2007).
Multiple representation	Involves the use of models, pictures, and a combination of them or various visual tools that enable people to communicate ideas or concepts (Tsui and Treagust, 2004; Adadan et al., 2010). By capturing students’ attention to the concepts to be taught, it improves students’ conceptual understanding (Ainsworth, 1999; Adadan et al., 2010). Since it provides diverse opportunities to students in constructing the same knowledge from multiple perspectives, any weaknesses dealt with one particular representation might be replaced by another one (Adadan et al., 2010). Thus, it facilitates the interpretation of abstract or complicated concepts/knowledge and helps students achieve meaningful learning (Adadan et al., 2010; Adadan and Ataman, 2021).
Problem-based learning	Exploits complex and real-life problems to motivate students to identify and research the concepts and principles needed for solving the related problems (Duch et al., 2001; Rahman and Lewis, 2020). In other words, students work in their small collaborative groups on a contextually framed problem to identify necessary information and then communicate with each other and integrate information into feasible solutions (Duch et al., 2001; Rahman and Lewis, 2020). Thus, it not only facilitates student learning of concepts and issues, but also develops science process skills, scientific inquiry skills and social skills (Delisle, 1997; Torp and Sage, 1998; Diercks, 2002).
Other	Covers different intervention types (e.g., project-based learning, ARCS motivation model, family involvement and use of posters) reported by only one paper.

Acknowledgements

We would like to thank Associate Professor Jennie Farber Lane from Bilkent University, Türkiye for her kind help in language polishing.

References

1. Adadan, E. and Ataman, M. M., (2021). Promoting senior primary school students’ understanding of particulate nature of matter through inquiry instruction with multiple representations, Education, 49(3), 317–329 DOI:10.1080/03004279.2020.1854960.
2. Adadan, E., Trundle, K. C. and Irving, K. E., (2010), Exploring grade 11 students' conceptual pathways of the particulate nature of matter in the context of multirepresentational instruction, J. Res. Sci. Teach., 47(8), 1004–1035 DOI:10.1002/tea.20366.
3. Ainsworth, S. E., (1999), The functions of multiple representations, Comput. Educ., 33, 131–152.
4. Aktas, İ. and Bilgin, İ., (2015), The effect of the 4MAT learning model on the achievement and motivation of 7th grade students on the subject of particulate nature of matter and an examination of student opinions on the model, Res. Sci. Technol. Educ., 33(1), 1–21 DOI:10.1080/02635143.2014.968536.
5. Akyol, C., (2007), İlköğretim fen ve teknoloji dersinde uygulanan şiirle öğretim tekniğinin başarı ve tutuma etkisi [The effect of poetical teaching technique applied in science and technology lessons to the success and attitude of students in primary education schools], (unpublished Master's thesis), Niğde, Türkiye (In Turkish): Niğde University, Graduate School of Social Sciences.
6. Alaca, Ö., (2014), Kuantum öğrenme modeline dayalı fen bilimleri öğretiminin ortaokul öğrencilerinin akademik başarı, tutum ve öğrenmenin kalıcılığı üzerine etkisi [Quantum of science based learning model learning secondary school students ' academic achievement, attitudes and its effect on retention of learning], (unpublished Master's thesis), Çanakkale, Türkiye (In Turkish): Çanakkale Onsekiz Mart University, Graduate School of Educational Sciences.
7. Altay, E. and Balım, A. G., (2021), Maddenin tanecikli yapısı ile ilgili kavram yanılgılarının tespiti [Detecting misconceptions regarding the particulate structure of matter], Western Anatolia J. Educ. Sci., 12(2), 576–592 DOI:10.51460/baebd.956631 (In Turkish).
8. Ataman Mortaş, M. M., (2011), Promoting grade 6 students' scientific understanding of the particulate nature of matter and the role of models in science through multi-representational instruction, [Doctoral dissertation], İstanbul, Türkiye: Bogazici University, Graduate School of Natural and Applied Sciences.
9. Ateş, R. O., (2007), 6.sınıflarda maddenin tanecikli yapısı konusunun çoklu zeka kuramına dayalı öğretimi [Multiple intelligences theory based instruction of the particulate natüre of the matter at 6th grade level], (unpublished Master's thesis), Balıkesir, Türkiye (In Turkish): Balıkesir University, Graduate School of Natural and Applied Sciences.
10. Ateş, A., (2018), 7. sınıf fen ve teknoloji dersi “maddenin tanecikli yapısı ve saf maddeler” konusunda artırılmış gerçeklik teknolojileri kullanılarak oluşturulan öğrenme materyalinin akademik başarıya etkisi [7th grade science and technology course “structure of the substance and pure substances” on augmented reality created using technology learning material effect on academic achievement], [unpublished Master's thesis], Niğde, Türkiye (In Turkish): Niğde Ömer Halisdemir University, Graduate School of Educational Sciences.
11. Australian Curriculum (2022), F-10 curriculum: Science. https://www.australiancurriculum.edu.au/f-10-curriculum/science/.
12. Ayas, A. and Özmen, H., (2002), A study of students' level of understanding of the particulate nature of matter at secondary school level, Bogazici Univ. J. Educ., 19(2), 45–60.
13. Ayas, A., Özmen, H. and Çalik, M., (2010), Students’ conceptions of the particulate nature of matter at secondary and tertiary level, Int. J. Sci. Math. Educ., 8(1), 165–184 DOI:10.1007/s10763-009-9167-x.
14. Aydost, Y., (2011), The effect of computer supported education to the concept knowledge and attitude, [unpublished Master's thesis], Eskişehir Osmangazi University, Graduate School of Educational Sciences.
15. Bağ, H. and Çalık, M., (2022), Argumentation-driven educational digital game: a case of matter subject, J. Mater. Educ., 44, 67–76.
16. Barthlow, M. J. and Watson, S. B., (2014), The effectiveness of process-oriented guided inquiry learning to reduce alternative conceptions in secondary chemistry, School Sci. Math., 114(5), 246–255 DOI:10.1111/ssm.12076.
17. Baviskar, S. N., Hartle, R. T. and Whitney, T., (2009), Essential criteria to characterize constructivist teaching: derived from a review of the literature and applied to five constructivist-teaching method articles, Int. J. Sci. Educ., 31(4), 541–550 DOI:10.1080/09500690701731121.
18. Bayraktar, S., (2001), A meta-analysis of the effectiveness of computer-assisted instruction in science education, J. Res. Technol. Educ., 34(2), 173–188 DOI:10.1080/15391523.2001.10782344.
19. Beerenwinkel, A., Parchmann, I. and Gräsel, C., (2011), Conceptual change texts in chemistry teaching: a study on the particle model of matter, Int. J. Sci. Math. Educ., 9(5), 1235–1259 DOI:10.1007/s10763-010-9257-9.
20. Bektaş, O., (2003), Maddenin tanecikli yapısı ile ilgili lise 1. sınıf öğrencilerinin yanlış kavramaları, nedenleri ve giderilmesi [The misconceptions of high school first year students related to the particle structure of matter, the reasons behind them and their elimination], [unpublished Master's thesis], Ankara, Türkiye (In Turkish): Gazi University, Graduate School of Educational Sciences.
21. Bektaş, O., (2011), The effect of 5E learning cycle model on tenth grade students’ understanding in the particulate nature of matter, epistemological beliefs and views of nature of science, [Unpublished doctoral dissertation], Ankara, Türkiye: Middle East Technical University, Graduate School of Natural and Applied Sciences.
22. Bennett, J., Gräsel, C., Parchmann, I. and Waddington, D., (2005), Context-based and conventional approaches to teaching chemistry: comparing teachers’ views, Int. J. Sci. Educ., 27(13), 1521–1547 DOI:10.1080/09500690500153808.
23. Bell, B., (2005), Pedagogies developed in the learning in science projects and related theses, Int. J. Sci. Educ., 27(2), 159–182 DOI:10.1080/0950069042000276703.
24. Bell T., Urhahne D., Schanze S. and Ploetzner, R., (2010), Collaborative inquiry learning: models, tools, and challenges, Int. J. Sci. Educ., 32(3), 349–377 DOI:10.1080/09500690802582241.
25. Biddulph, F. and Osborne, R., (1985), Overviews: Selected working papers of the learning in science project (primary), Hamilton University of Waikato.
26. Borenstein, M., Hedges, L. V., Higgins, J. P. T. and Rothstein, H. R., (2009), Introduction to meta-analysis, John Wiley & Sons.
27. Boz, Y., (2006), Turkish pupils’ conceptions of the particulate nature of matter, J. Sci. Educ. Technol., 15(2), 203–213 DOI:10.1007/s10956-006-9003-9.
28. Bulte, A. M. W., Westbroek, H. B., de Jong, O. and Pilot, A., (2006), A research approach to designing chemistry education using authentic practices as contexts, Int. J. Sci. Educ., 28(9), 1063–1086 DOI:10.1080/09500690600702520.
29. Bunce, D. M. and Gabel, D., (2002), Differential effects on the achievement of males and females of teaching the particulate nature of chemistry, J. Res. Sci. Teach., 39(10), 911–927 DOI:10.1002/tea.10056.
30. Büyük Kuloğlu, Z., (2019), Yaşam temelli öğrenme yaklaşımının 6. sınıf öğrencilerinin maddenin tanecikli yapısı konusundaki başarılarına ve motivasyonlarına etkisi [The effect of context based learning approach on the success and motivation of 6th grade students’ about the particular structure of the matter], [unpublished Master's thesis], Düzce, Türkiye (In Turkish): Düzce University, Graduate School of Natural and Applied Sciences.
31. Caine, R. N. and Caine, G., (1991), Making connections: Teaching and the human brain, Alexandria, VA: Association for Supervision and Curriculum Development.
32. Çalık, M., Ayas, A. and Coll, R. K., (2009), Investigating the effectiveness of an analogy activity in improving students’ conceptual change for solution chemistry concepts, Int. J. Sci. Math. Educ., 7, 651–676 DOI:10.1007/s10763-008-9136-9.
33. Çavdar, O., (2016), Fen ve teknoloji dersinin öğretiminde iyi bir eğitim ortamı için yedi ilke ve modellerin işbirlikli öğrenme yöntemiyle uygulanması [The application of the seven principles for good practice and models with cooperative learning method in teaching of science and technology course], [Unpublished doctoral dissertation], Erzurum, Türkiye (In Turkish): Atatürk University, Graduate School of Educational Sciences.
34. Çavdar, O., Okumuş, S., Alyar, M. and Doymuş, K., (2016), Effecting of using different methods and models on understanding the particulate nature of matter, Erzincan Univ. J. Educ., 18(1), 555–592 DOI:10.17556/jef.34457.
35. Ceylan, Ö., (2015), Fen öğretiminde kavram karikatürü kullanımının 7.sınıf öğrencilerinin akademik başarılarına ve bilişsel yapılarına etkisinin incelenmesi [Analyzing the effects of concept cartoon usage on 7.grade students' science achievement and their cognitive structure of learning], (unpublished Master's thesis), Sakarya, Türkiye (In Turkish): Sakarya University, Graduate School of Educational Sciences.
36. Chang, H. Y., Binali, T., Liang, J. C., Chiou, G. L., Cheng, K. H., Lee, S. W. Y. and Tsai, C. C., (2022), Ten years of augmented reality in education: a meta-analysis of (quasi-) experimental studies to investigate the impact, Comput. Educ., 191, 104641 DOI:10.1016/j.compedu.2022.104641.
37. Cokelez, A. and Dumon, A., (2005), Atom and molecule: upper secondary school French students’ representations in long-term memory, Chem. Educ. Res. Pract., 6, 119–135 10.1039/B4RP90005G.
38. Çolak, O., (2014), Sorgulayıcı-araştırmaya dayalı fen öğretimi yönteminin fen okuryazarlığı ve bazı alt-boyutları üzerine etkisi [The effect of the inquiry-based science teaching method on scientific literacy and the certain sub-scale of scientific literacy], [unpublished Master's thesis], Edirne, Türkiye (In Turkish): Trakya University, Graduate School of Natural and Applied Sciences.
39. Coll, R. K. and Treagust, D. F., (2003), Investigation of secondary school, undergraduate, and graduate learners’ mental models of ionic bonding, J. Res. Sci. Teach., 40(5), 464–486 DOI:10.1002/tea.10085.
40. Davis, A., (2004), The credentials of brain-based learning, Journal of Philosophy of Education, 38(1), 21–36 DOI:10.1111/j.0309-8249.2004.00361.x.
41. Degen, R., (2014) Brain-based learning: The neurological findings about the human brain that every teacher should know to be effective, Amity Global Business Rev., 77(1), 915–23.
42. Delisle, R., (1997), How to use problem-based learning in the classroom, Alexandria, VA: Association for Supervision and Curriculum Development.
43. Demircioğlu, H., (2017), Effect of PDEODE teaching strategy on Turkish students' conceptual understanding: Particulate nature of matter, J. Educ. Train. Studies, 5(7), 78–90 DOI:10.11114/jets.v5i7.2389.
44. Department for Education (2015), National curriculum in England: Science programmes of study. https://www.gov.uk/government/publications/national-curriculum-in-england-science-programmes-of-study/national-curriculum-in-england-science-programmes-of-study#key-stage-3.
45. de Vos, W. and Verdonk, A. H., (1996), The particulate nature of matter in science education and in science, J. Res. Sci. Teach., 33(6), 657–664 DOI:10.1002/(SICI)1098-2736(199608)33:6%3c657::AID-TEA4%3e3.0.CO;2-N.
46. Diercks, R. W., (2002), Problem-based learning in elementary science methods: Exploring a format to prepare teachers for the 21st century, (unpublished Doctoral Thesis), Nebraska, USA: the University of Nebraska, Graduate College, UMI Number: 3059945.
47. Doymuş, K., Şimşek, Ü. and Karaçöp, A., (2009), The effects of computer animations and cooperative learning methods in micro, macro and symbolic level learning of states of matter, Eurasian J. Educ. Res., 36, 109–128.
48. Duch, B. J., Groh, S. E. and Allen, D. J. E., (2001), The power of problem-based learning: A practical “how to” for teaching undergraduate courses in any discipline, Sterling, VA: Stylus Publishing, LLC.
49. Duschl, R. A., Schweingruber, H. A. and Shouse, A. W., (2007), Taking science to school: Learning and teaching science in grades K-8, National Academies Press.
50. Ellis, P. D., (2010), The essential guide to effect sizes: Statistical power, meta-analysis, and the interpretation of research results, Cambridge University Press.
51. Er Nas, S., İpek Akbulut, H., Çalik, M. and Emir, M. İ., (2022), Facilitating conceptual growth of the mainstreamed students with learning disabilities via a science experimental guidebook: a case of physical events, Int. J. Sci. Math. Educ., 20, 45–67 DOI:10.1007/s10763-020-10140-3.
52. Eroğlu, N., (2010), 6. sınıf “maddenin tanecikli yapısı” ünitesindeki kavramların öğretiminde öğrenci ürünü karikatürlerin kullanımı [Using the student-made cartoons in teaching the concepts of “particulate structure of matter” unit in sixth grade], (unpublished Master's thesis), Sakarya, Türkiye (In Turkish): Sakarya University, Graduate School of Natural and Applied Sciences.
53. Eroğlu, S. and Bektaş, O., (2022), The effect of 5E-based STEM education on academic achievement, scientific creativity, and views on the nature of science, Learn. Individual Differences, 98, 102181 DOI:10.1016/j.lindif.2022.102181.
54. Ezema, M. J., Ugwuany, C. S., Okeke, C. I. and Orji, E. I., (2022), Influence of cognitive ability on students’ conceptual change in particulate nature of matter in physics, J. Turkish Sci. Educ., 19(1), 194–217 DOI:10.36681/tused.2022.118.
55. Freyberg, P. and Osborne, R., (1982), Final Report: Learning in Science Project (Form 1–4), Hamilton: University of Waikato.
56. Gardner, H., (1993), Multiple intelligences: The theory in practice, Basic books.
57. Gökçe, H. and Saraçoğlu, S., (2018), The effect of computer assisted instruction on eighth grade students’ academic achievement, logical thinking ability and attitude, related to the unit of acid and bases, Kastamonu Educ. J., 26(4), 1383–1394 DOI:10.24106/kefdergi.434178.
58. Gökdere, M. and Çalık, M., (2010), A cross-age study of Turkish students’ mental models: An “atom” concept, Didact. Slov.-Pedagos. Obzorja, 25(2), 185–199.
59. Güler, M., Bütüner, S. Ö., Danişman, Ş. and Gürsoy, K., (2022), A meta-analysis of the impact of mobile learning on mathematics achievement, Educ. Information Technol., 27, 1725–1745 DOI:10.1007/s10639-021-10640-x.
60. Guzzetti, B. J., Snyder, T. E., Glass, G. V. and Gamas, W. S., (1993), Promoting conceptual change in science: a comparative meta-analysis of instructional interventions from reading education and science education, Reading Res. Quarterly, 28, 116–159.
61. Hadenfeldt, J. C., Liu, X. and Neumann, K., (2014), Framing students’ progression in understanding matter: A review of previous research, Studies in Science Education, 50(2), 181–208 DOI:10.1080/03057267.2014.945829.
62. Hadenfeldt, J. C., Neumann, K., Bernholt, S., Liu, X. and Parchmann, I., (2016), Students’ progression in understanding the matter concept, J. Res. Sci. Teach., 53(5), 683–708 DOI:10.1002/tea.21312.
63. Haidar, A. H. and Abraham, M. R., (1991), A comparison of applied and theoretical knowledge of concepts based on the particulate nature of matter, J. Res. Sci. Teach., 28, 919–938 DOI:10.1002/tea.3660281004.
64. Harrison, A. G. and Treagust, D. F., (2002), The particulate nature of matter: Challenges in understanding the submicroscopic world, in J. K. Gilbert (ed.), Chemical Education: Towards Research-based Practice, Dordrecht: Springer, pp. 189–212.
65. Hedges, L. V. and Vevea, J. L., (1998), Fixed-and random-effects models in meta-analysis, Psychological Methods, 3, 486–504 DOI:10.1037/1082-989X.3.4.486.
66. Hewson, P. W. and Hewson, M. G., (1983), Effect of instruction using students’ prior knowledge and conceptual change strategies on science learning, J. Res. Sci. Teach., 20, 731–743 DOI:10.1002/tea.3660200804.
67. Higgins, J. P., Thompson, S. G., Deeks, J. J. and Altman, D. G., (2003), Measuring inconsistency in meta-analyses. BMJ, 327(7414), 557–560 DOI:10.1136/bmj.327.7414.557.
68. Jensen, E., (1995), Brain-based learning & teaching, Thousand Oaks, CA: Corwin Press.
69. Johnson, D. and Johnson, R., (1990), Cooperative learning and achievement, in S. Sharan (ed.), Cooperative learning: Theory and research, New York: Praeger, pp. 23–37.
70. Johnson, D. W., Johnson, R. T. and Smith, K. A., (1998), Cooperative learning returns to college. What evidence is there that it works? Change: The Magazine of Higher Learning, 30, 26–35 DOI:10.1080/00091389809602629.
71. Kaçar, S., (2012), Görsel sanatlarla bütünleştirilmiş probleme dayalı öğrenme yönteminin öğrencilerin fen akademik başarılarına, bilimsel yaratıcılıklarına ve sanat etkinlikleriyle fen öğrenme tutumlarına etkileri [The effects of problem based learning method integrated visual arts on students’ academic achievements, scientific creativities and attitudes towards science teaching with art activities], [unpublished Master's thesis], İzmir, Türkiye (In Turkish): Dokuz Eylül University, Graduate School of Educational Sciences.
72. Kansızoğlu, H. B., (2017), The effect of graphic organizers on language teaching and learning areas: a meta-analysis study, Educ. Sci., 42(191), 139–164 DOI:10.15390/EB.2017.6777.
73. Karadag, E., (2020), The effect of educational leadership on students’ achievement: a cross-cultural meta-analysis research on studies between 2008 and 2018, Asia Pacific Educ. Rev., 21, 49–64 DOI:10.1007/s12564-019-09612-1.
74. Kaya, O. N., (2005), Tartışma teorisine dayalı öğretim yaklaşımının öğrencilerin maddenin tanecikli yapısı konusundaki başarılarına ve bilimin doğası hakkındaki kavramalarına etkisi [The effect of teaching based on argumentation theory on the achievement of the students related to the particulate nature of matter and conception of the nature of science], (Unpublished doctoral dissertation), Ankara, Türkiye (In Turkish): Gazi University, Graduate School of Educational Sciences.
75. Kepes, S., Banks, G. C., McDaniel, M. and Whetzel, D. L., (2012), Publication bias in the organizational sciences, Organizational Res. Methods, 15(4), 624–662 DOI:10.1177/1094428112452760.
76. Kelly, D. L., Centurino, V. A. S., Martin, M. O. and Mullis, I. V. S., (2020a), TIMSS 2019 Encyclopedia: Education Policy and Curriculum in Mathematics and Science, Philippines. https://timssandpirls.bc.edu/timss2019/encyclopedia/philippines.html.
77. Kelly, D. L., Centurino, V. A. S., Martin, M. O. and Mullis, I. V. S., (2020b), TIMSS 2019 Encyclopedia: Education Policy and Curriculum in Mathematics and Science, Malaysia, https://timssandpirls.bc.edu/timss2019/encyclopedia/malaysia.html.
78. Kelly, D. L., Centurino, V. A. S., Martin, M. O. and Mullis, I. V. S., (2020c), TIMSS 2019 Encyclopedia: Education Policy and Curriculum in Mathematics and Science, North Macedonia, https://timssandpirls.bc.edu/timss2019/encyclopedia/north-macedonia.html.
79. Kenan, O., (2014), “Maddenin tanecikli yapısı” ünitesine yönelik zenginleştirilmiş bilgisayar destekli öğretim materyalinin geliştirilmesi ve etkililiğinin araştırılması [Developing and investigating the effectiveness of enriched computer based instruction material on “the particulate nature of matter” subject], [Unpublished doctoral dissertation], Trabzon, Türkiye (In Turkish): Karadeniz Technical University, Graduate School of Educational Sciences.
80. Kirman Bilgin, A., Demircioğlu Yürükel, F. N. and Yiğit, N., (2017), The effect of a developed REACT strategy on the conceptual understanding of students: “particulate nature of matter”, J. Turkish Sci. Educ., 14(2), 65–81 DOI:10.12973/tused.10199a.
81. Küçük, A. and Yıldırım, N., (2017), The effects of the methods used in teaching the concepts of the “substance and change” unit to academic successful: a meta analysis study, Route Educ. Soc. Sci. J., 4(5), 251–268.
82. Lipsey, M. W. and Wilson, D. B., (2001), Practical meta-analysis, Sage Publications.
83. Liu, X. and Lesniak, K. M., (2005), Students’ progression of understanding the matter concept from elementary to high school, Sci. Educ., 89, 433–450 DOI:10.1002/sce.20056.
84. Ministry of National Education, (2018), Science curriculum for primary and secondary schools (3rd-8th grades), Ankara.
85. Mullen, B., Muellerleile, P. and Bryant, B., (2001), Cumulative meta-analysis: a consideration of indicators of sufficiency and stability, Personality Social Psychology Bulletin, 27(11), 1450–1462 DOI:10.1177/01461672012711006.
86. Namdar, B. and Shen, J., (2018), Knowledge organization through multiple representations in a computer-supported collaborative learning environment, Interactive Learn. Environ., 26(5), 638–653 DOI:10.1080/10494820.2017.1376337.
87. Nakhleh, M. B. and Samarapungavan, A., (1999), Elementary school children's beliefs about matter, J. Res. Sci. Teach., 36(7), 777–805 DOI:10.1002/(SICI)1098-2736(199909)36:7[double bond splayed right]777::AID-TEA4[double bond splayed left]3.0.CO;2-Z.
88. Nakhleh, M. B., Samarapungavan, A. and Saglam, Y., (2005), Middle school students’ beliefs about matter, J. Res. Sci. Teach., 42(5), 581–612 DOI:10.1002/tea.20065.
89. NGSS, (2017), Next generation science standards, https://www.nextgenscience.org/sites/default/files/AllTopic.pdf.
90. Novick, S. and Nussbaum, J., (1978), Junior high school pupils’ understanding of the particulate nature of matter: an interview study, Sci. Educ., 62, 273–281.
91. Nuic, I. and Glažar, S. A., (2020), The effect of e-learning strategy at primary school level on understanding structure and states of matter, Eurasia J. Math. Sci. Technol. Educ., 16(2), em1823 DOI:10.29333/ejmste/114483.
92. Nussbaum, J., (1985), The particulate nature of matter in the gaseous phase, in Driver, E. Guesne and A. Tiberghien (ed.), Children's ideas in science, Open University Press, pp. 124–144.
93. Öksüz, M., (2019), “Maddenin tanecikli yapısı” ünitesine yönelik zenginleştirilmiş öğretim materyalinin etkililiğinin tespiti [Investigation the effectiveness of enriched teaching material on “the particulate nature of matter”], [unpublished Master's thesis], Trabzon, Türkiye (In Turkish): Trabzon University, Graduate School of Education.
94. Okumuş, S., (2017), “İyi bir eğitim ortamı için yedi ilke”nin işbirlikli öğrenme ve modellerle birlikte uygulanmasının fen bilimleri dersinin anlaşılmasına etkisi [The effect of implementing “the seven principles for good practice” using with cooperative learning and models on understanding of science], [Unpublished doctoral dissertation], Erzurum, Türkiye (In Turkish): Atatürk University, Graduate School of Educational Sciences.
95. Okumuş, S., Çavdar, O., Alyar, M. and Doymuş, K., (2017), The effect of different teaching methods to understanding of chemical equilibrium at micro level, Elementary Educ. Online, 16(2), 727–745 DOI:10.17051/ilkonline.2017.304730.
96. Orosz G., Németh V., Kovács L., Somogyi Z. and Korom E., (2022), Guided inquiry-based learning in secondary-school chemistry classes: a case study, Chem. Educ.: Res. Pract., 24, 50–70 10.1039/D2RP00110A.
97. Özkanbaş, M., (2018), 6. sınıf maddenin tanecikli yapısı ünitesinin öğretiminde süreç odaklı rehberli sorgulamayla öğrenme yönteminin öğrencilerin akademik başarıları, sorgulayıcı öğrenme becerisi algıları ve mantıksal düşünme becerileri üzerine etkisi [Effect of process oriented guided inquiry learning (POGIL) on 6th grade students’ achievement, logical thinking skills and inquiry learning skills perceptions in the context of particulate nature of matter], [unpublished Master's thesis], Adana, Türkiye (In Turkish): Çukurova University, Graduate School of Social Sciences.
98. Özkanbaş, M. and Taştan Kırık, Ö., (2020), Implementing collaborative inquiry in a middle school science course, Chem. Educ. Res. Pract., 21(4), 1199–1217 10.1039/C9RP00231F.
99. Özmen, H., (2011), Effect of animation enhanced conceptual change texts on 6th grade students’ understanding of the particulate nature of matter and transformation during phase changes, Comput. Educ., 57(1), 1114–1126 DOI:10.1016/j.compedu.2010.12.004.
100. Özmen, H., (2013), A cross-national review of the studies on the particulate nature of matter and related concepts, Eurasian J. Phys. Chem. Educ., 5(2), 81–110.
101. Ozyalcin, B., (2020), Artırılmış gerçeklikle zenginleştirilmiş jigsaw etkinliklerinin “maddenin tanecikli yapısı”na ilişkin başarıya ve teknolojik farkındalığa etkisi [The effect of jigsaw activities enriched with augmented reality on success and technological awareness relating to the “particulate structure of matter, [unpublished Master's thesis], İstanbul, Türkiye (In Turkish): İstanbul University, Graduate School of Educational Sciences.
102. Pedaste M., Mäeots M., Siiman L. A., de Jong T., van Riesen S. A. N., Kamp E. T., Manoli C. C., Zacharia Z. C. and Tsourlidaki E., (2015), Phases of inquiry-based learning: Definitions and the inquiry cycle, Educ. Res. Rev., 14(1), 47–61.
103. Rahman T. and Lewis S. E., (2020), Evaluating the evidence base for evidence-based instructional practices in chemistry through meta-analysis, J. Res. Sci. Teach., 57(5), 765–793 DOI:10.1002/tea.21610.
104. Riaz, M., (2004), Helping children to understand particulate nature of matter, Alberta Sci. Educ. J., 36(2), 56–59.
105. Rusçuklu, P., (2017), Yaşam temelli öğrenme yaklaşımının 6.sınıf öğrencilerinin “maddenin tanecikli yapısı” ünitesindeki akademik başarı ve kalıcılıklarına etkisi [The effects of context-based learning approach to academic achievements and retention of knowledge in the “particular structure of matter” unit of the sixth graders], [unpublished Master's thesis], Bursa, Türkiye (In Turkish): Uludağ University, Graduate School of Educational Sciences.
106. Schwartz, R., Lederman, N. and Crawford, B., (2004), Developing views of nature of science in an authentic context: An explicit approach to bridging the gap between nature of science and scientific inquiry, Sci. Educ., 88(4), 610–645.
107. Scott, P., Dyson, T. and Gater, S., (1987), A constructivist view of learning and teaching in science: Children's learning in science project, Leeds: Centre for Studies in Science and Mathematics Education.
108. Sezen-Vekli, G. and Çalik, M., (2023), The effect of web-based biology learning environment on academic performance: a meta-analysis study, J. Sci. Educ. Technol. DOI:10.1007/s10956-023-10033-4.
109. Singer, J. E., Tal, R. T. and Wu, H.-K., (2003), Students’ understanding of the particular nature of matter, School Sci. Math., 103(1), 28–44 DOI:10.1111/j.1949-8594.2003.tb18111.x.
110. Sözbilir, M., (2002), Students’ ideas and misunderstandings of the particulate nature of matter: a review of selected literature, Kastamonu Educ. J., 10(2), 345–350.
111. Sözbilir, M., Sadi, S., Kutu, H. and Yıldırım, A., (2007), Context-based teaching approach in chemistry education and its applications in the world, Presented at the First National Chemistry Education Congress, 20(22), 108.
112. Stolk M. J., Bulte A. M. W., de Jong O. and Pilot A., (2009), Towards a framework for a professional development programme: empowering teachers for context-based chemistry education, Chem. Educ. Res. Pract., 10, 164–175 10.1039/B908255G.
113. Tang, H. and Abraham, M. R., (2016), Effect of computer simulations at the particulate and macroscopic levels on students’ understanding of the particulate nature of matter, J. Chem. Educ., 93(1), 31–38 DOI:10.1021/acs.jchemed.5b00599.
114. Talanquer, V., (2009), On cognitive constraints and learning progressions: The case of “structure of matter, Int. J. Sci. Educ., 31(15), 2123–2136 DOI:10.1080/09500690802578025.
115. Torp, L. and Sage, S., (1998), Problems as possibilities: Problem-based learning for K-12 education, Alexandria, VA: Association for Supervision and Curriculum Development.
116. Tsai, C.-C., (1999), Overcoming junior high school students’ misconceptions about microscopic views of phase change: a study of an analogy activity, J. Sci. Educ. Technol., 8(1), 83–91 DOI:10.1023/A:1009485722628.
117. Tsui, C. Y. and Treagust, D. F., (2004), Learning genetics with multiple representations: cross-case analyses of student' conceptual status. Paper presented at the National Association for Research in Science Teaching, Vancouver, BC, Canada.
118. Türkoguz, S. and Ercan, İ., (2022), Effect of visual anthropomorphic stories on students' understanding of the particulate nature of matter and anthropomorphic discourse, Chem. Educ. Res. Pract., 23(1), 206–225 10.1039/D1RP00057H.
119. Ültay, N. and Çalık, M., (2012), A thematic review of studies into the effectiveness of context-based chemistry curricula, J. Sci. Educ. Technol., 21(6), 686–701 DOI:10.1007/s10956-011-9357-5.
120. Üstün, U. and Eryılmaz, A., (2014), A research methodology to conduct effective research syntheses: meta-analysis, Educ. Sci., 39(174), 1–32 DOI:10.15390/EB.2014.3379.
121. Vevea, J. L., Coburn, K. and Sutton, A., (2019), Publication bias, in Cooper H., Hedges L. V. and Valentine J. C. (ed.), The handbook of research synthesis and meta-analysis, 3rd edn, Russell Sage Foundation, pp. 383–430.
122. Villagonzalo, E. C., (2014), Process oriented guided inquiry learning: An effective approach in enhancing students’ academic performance. Paper presented at DLSU Research Congress, 2(1), 1–6.
123. Williamson, V. M., (1992), The effects of computer animation emphasizing the particulate nature of matter on the understandings and misconceptions of college chemistry students, [Unpublished doctoral dissertation], Oklahoma, USA: The University of Oklahoma, Graduate College.
124. Wu, H. K. and Puntambekar, S., (2012), Pedagogical affordances of multiple external representations in scientific processes, J. Sci. Educ. Technol., 21(6), 754–767 DOI:10.1007/s10956-011-9363-7.
125. Yang, J. C., Chien, K. H. and Liu, T. C., (2012), A digital game-based learning system for energy education: an energy conservation PET, Turkish Online J. Educ. Technol.-TOJET, 11(2), 27–37.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3rp00027c

---