Introduction

While the cognitive dimension remains central in science education, as evidenced by the implementation of curricula, selective teaching methods and reflections on students’ learning in terms of conceptual understanding, there is a growing emphasis in both research and educational systems on increasing the influence of affective factors in this process [1–4]. The specialised literature emphasises the importance of considering students’ attitudes as a key determinant of educational quality and effectiveness and advocates their cultivation in science education [5–9].

There is a broad consensus in the research community on the central role of positive attitudes towards science, which are considered to be a key determinant of variables such as academic achievement in the discipline [5, 10, 11], as well as long-term interest in science-related careers [12, 13]. Nevertheless, fostering positive attitudes in individuals through formal education is recognised as a formidable challenge [14, 15].

Furthermore, it has been observed that this challenge becomes more pronounced during the secondary education phase, where not only is there no increase in positive attitudes towards science compared to earlier stages of education, but a decline is even observed as students get older or progress through school. This trend has been highlighted in several research efforts [16–19]. For example, in a comprehensive literature review, Christidou [20] concluded that the transition from primary to secondary school often coincides with a declining interest in science, with students increasingly perceiving it as an untenable career path. Similarly, a literature review that included an analysis of 56 publications investigating the effects of interventions using different teaching approaches on attitudes towards science and mathematics found a notable negative correlation between school year and attitudes towards science [21].

While it is true that this trend of changing attitudes in secondary education is a major challenge for the entire education system, there is also a large body of scientific literature claiming that attitudes can be positively influenced by various factors [8, 18, 22–26]. Among these factors, educational interventions seem to play a key role in their development [5, 19, 27, 28].

Building on this idea, Helvaci & Yilmaz [29] suggested that pedagogical practices developed by teachers have an impact on students’ motivation and therefore promote the formation of positive attitudes towards science. In the same vein, Osborne et al. [5] highlight the specific characteristics of educational interventions as the main contributors to achieving positive attitudes towards science. Indeed, it has been found that negative attitudes towards science are mainly due to ineffective teaching methods, which are sometimes teacher-centred [28, 30], leading students to perceive science-related subjects as difficult or boring. With this in mind, it is important to identify the characteristics of educational interventions that achieve positive attitudes towards science among middle school students in order to increase the effectiveness of these interventions.

The purpose of the scoping review presented in this paper is to examine the scientific evidence available over the last 10 years on the characteristics of educational programmes that promote positive attitudes towards science in secondary school students.

These characteristics were analysed in terms of teaching approaches, duration, the person responsible for the intervention, the educational setting in which it was implemented, the monitoring of the development process, and the effects of the intervention on the attitude variable. As can be seen from the above statement, in contrast to previous literature reviews, which have mainly used quantitative methods to focus on and explain the results of educational interventions on attitudes towards science, the present study explored key features of educational interventions developed with the intention of understanding how they were designed and analysing gaps in the knowledge base.

In addition, the study focused on the construct of attitudes towards science as a whole, including data from disciplines such as physics, chemistry and biology, within the corresponding period of secondary education, a stage where a greater decline in this affective variable is observed and is considered fundamental for the development of scientific careers. In this context, the overarching question that guided this study was: What features are reported to be effective in educational programmes to promote positive science attitudes in middle school students?

Factors influencing teaching science to middle school students

Educational intervention or instruction is understood as the process by which knowledge, skills and attitudes are deliberately and intentionally fostered, including the entire instructional process, from planning and delivery to evaluation and feedback [31]. In science education in particular, the success of this process, that is, the ability of students to understand and engage in debates about science-related issues that shape our world, is crucial [12]. In this venue, it is important to consider a number of factors that support this educational action. Among these factors, instructional approach is particularly relevant in identifying effective science teaching practices. This is relevant because of the complexity of learning itself, as students require different types of explicit instructional support to understand and engage with the body of scientific knowledge [6].

There are many references in the literature to the effect that one or more teaching approaches can have on attitudes to science at different educational levels, among which those mediated by technological or virtual environments, based on inquiry strategies, out-of-classroom activities or cooperative learning, stand out [32–37]. For instance, a correlation between the achievement of positive attitudes towards physics and chemistry was found in a sample of Chinese students and cooperative learning strategies [38]. The impact of teaching approaches on students’ attitudes towards science was also observed in the 2015 Programme for International Student Assessment (PISA). This analysis showed that attitudes towards science, including interest in scientific topics, enjoying science, and participating in science-related activities, showed better results with an inquiry-based approach compared to other approaches such as adaptive teaching or direct teacher instruction, despite the latter being associated with better academic performance [39].

Although teaching approaches are diverse and each has specific characteristics, they can be grouped according to their nature and intervention purposes [19, 21, 38, 40, 41]. To provide a more flexible and comprehensive framework for interpretation, teaching approaches found by Savelsbergh et al. [21] were adopted in this study and further supplemented by additional categories from Schroeder et al. [40]. This integration allows for a broader, more nuanced analysis of teaching strategies, accommodating the complexity and diversity inherent in science education. The categories are outlined in Table 1, with approaches 1 to 5 from Savelsbergh et al. [21] and approaches 6 to 11 from Schroeder et al. [40].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Approaches to the teaching of science.

https://doi.org/10.1371/journal.pone.0315757.t001

Table 1 shows the pedagogical approaches used in science education. It should be noted that some of the categories identified by the authors overlap in terms of the objectives of the pedagogical approach, even if they are not labelled as such. Some examples illustrating this situation are approaches related to inquiry-based learning, technology-mediated approaches and collaborative approaches. It was also found that the category “enhanced context strategies” identified by Schroeder et al. [40] could be included in the category “extracurricular” proposed by Savelsbergh et al. [21]. In all cases, the labels coined by the latter author were adopted.

In addition, it should be noted that the category of manipulative strategies, which refers to the teaching approach where the teacher provides opportunities for students to work with physical objects, also includes work with intangible elements, which mainly refers to declarative knowledge that can be represented tangibly, as in the case of a concept map.

Method

This literature review adopted the scoping review methodology proposed by Peters et al. [48]. This methodological design was chosen due to the need to scan the existing literature on the topic and expand knowledge beyond simply addressing the effectiveness of the intervention itself [49, 50]. Previous educational research using this approach has confirmed that this type of study can provide relevant information for research and educational practice in the context of changing attitudes towards science [51, 52].

The following search parameters were used to conduct a thorough search of the academic databases Eric, SciELO, Google Scholar, Dialnet, EbscoHost, Scopus and Web of Science for literature on educational interventions aimed at fostering attitudes towards science in high school students: studies focused on assessing the construct of attitudes towards science, involving the implementation of a training programme or educational intervention for students, focused on students between 12 and 16 years of age, conducted in the last 10 years and published in English or Spanish. As established, the sources of information for the scoping review could include any existing literature [53]. Therefore, the inclusion of literature was not limited exclusively to peer-reviewed primary research studies, but also included meta-analyses, systematic reviews and other scoping reviews. Although other types of literature, such as handbooks or books, were considered for inclusion in the corpus of documents, they were not found using the search criteria established.

The search was carried out during two months of 2023, specifically September and October, with the most recent exploration of literature conducted on October 21 of the same year. This process followed the methodology proposed by Peters et al., which consists of three main phases [48]. Each phase is described below, together with its implications for the development of the study.

1. Initial search in databases (DB). Starting from an initial search equation, a first exploratory exercise was carried out with the intention of finding concepts that would allow the refinement of the search equations. This phase was carried out in two databases, one in Spanish (SciELO) and another that would allow the retrieval of documents in English (ERIC). In the SciELO database, the initial search was performed with titles that contained one or more of the following terms: “actitudes hacia la ciencia” OR “interés hacia a ciencia” + “estudiantes de secundaria” OR “adolescents” + “intervención” OR “programa” OR “instrucción”. In contrast, the ERIC database was searched using English-language equivalents of these terms: “attitudes toward science” OR “interest toward science” + “middle school students” OR “adolescents” + “intervention” OR “evidence-based program” OR “instruction”. The selection of these terms was aligned with the population, concept and context of the review, which are fundamental elements in a scoping review for establishing a priori inclusion criteria [48].
2. Subsequent DB search. This phase consisted of restructuring the initial search equations to search for primary documents for analysis in all the selected databases. Some keywords were included that refer to the breakdown of science into some of its disciplines (attitudes towards physics, attitudes towards chemistry, attitudes towards biology), as well as the keyword "secondary education" as a synonym for "middle school" in the English language. It was also emphasised that studies conducted with teachers or other populations that did not aim to develop and/or validate instruments to assess attitudes, or that only carried out an educational intervention, should not be included. In addition, after entering the search equations, some specifications were made in each database due to its characteristics. For example, in ERIC the temporality of the studies was specified, or in Scopus the search was restricted to social science studies. At the end of this phase, 133 documents were found.
3. Examination of the results. The third phase consisted of a manual selection based on the inclusion criteria specified in each of the DBs. This was done by reading the title, keywords and abstract of each document found. The full text was only consulted directly in those cases where the exposed elements could be clearly observed. After this step, 27 primary documents remained.
4. Searching reference lists. In order to ensure that material that was not included in the study and made an important contribution to the research was included, an additional step was taken to review the reference lists of review studies found in previous searches. These references also identified key journals that could contribute to the primary documents through manual review. A further 16 documents were located, but only 10 articles were included in the analysis corpus according to the inclusion criteria.

Where there was doubt about the inclusion or exclusion of retrieved texts, general agreement was reached within the research team on the basis of the inclusion criteria. For example, documents such as that by Houseal et al. reported the implementation of an intervention with students and teachers, so it was considered to include the study by reporting only the part related to the intervention with students [54]. In another case, that aimed to address attitudes towards other disciplines in addition to science was retained [21]. Furthermore, care was taken to ensure that only full texts were retained, excluding web pages, and that there were no duplicate articles.

The final outcome of the process was a total of 37 primary documents. It is worth noting that the elements of interest for the development of the study mentioned so far, such as objectives, research questions, inclusion criteria and methods of analysis, were specified and documented in advance in a research protocol (S1 Protocol). The literature search process is summarised in Fig 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Literature search process.

https://doi.org/10.1371/journal.pone.0315757.g001

Results

Some characteristics of the studies in terms of demographic and methodological information (country of origin, year of publication, research design and methodology, number of participants and factors influencing students’ attitudes towards science) and general components of the educational interventions found are presented before presenting the results in relation to the specific research questions posed. This information provides an overview of the information analysed in the current study.

The main characteristics of the educational interventions found are presented in Table 2. It should be noted that this table does not include the documents identified as literature reviews, as they do not allow the visualisation of elements of interest for this part of the research.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Approaches to interventions on middle school students’ attitudes towards science.

https://doi.org/10.1371/journal.pone.0315757.t002

However, these documents are included in other analyses of the study. Following Table 2, the results are presented in different sections in order to answer the specific research questions posed. According to the adopted methodological proposals (Peters et al., 2020), the aim is to provide sufficient details on the characteristics of the concept, population and context of interest.

This scoping review focused on studies published within the last decade, as mentioned above. Fig 2 shows the results of the analysis based on the year of publication, where the continued interest since 2018 in addressing students’ attitudes towards science through an educational programme is evident. Although no studies in this area were identified in the databases consulted in 2023, it is possible that publications occurred after the primary document search.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Annual scientific production.

The figure displays the number of publications from 2013 to 2023. A notable increase in recent years, particularly in 2022, suggesting growing academic interest in this topic.

https://doi.org/10.1371/journal.pone.0315757.g002

On the basis of the country in which the research was carried out or the data collected, the studies were categorised by continent of origin (Table 3). Due to the trend observed in the Americas, where the majority of publications were found in a minimum number of countries, it was decided to divide the continent into two parts: North and Central America on the one hand, and South America on the other. The research team felt that this decision would allow a better presentation of the results of the study in the first category of analysis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Number of publications per continent.

https://doi.org/10.1371/journal.pone.0315757.t003

Table 3 shows that the continent with the most contributions to this field of knowledge is Europe. Within this result, it is worth highlighting that the vast majority of studies were carried out in Turkey (14 studies). After Turkey, other countries that contributed significantly to this topic were, in decreasing order, the United States (6), Rwanda (3), Canada (2) and Colombia (2). Countries such as Ethiopia, Bhutan, Thailand, Oman, Nigeria, Qatar, Chile, Portugal, Lebanon and the Netherlands contributed only one document each to the corpus (Fig 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Distribution of the publication per country.

Countries are color-coded according to the number of publications, with darker shades representing a higher number of publications. The scale at the bottom ranges from 1 to 14 publications, with the highest concentration observed in Turkey. Following Turkey, the USA and Rwanda appear in lighter shades. We acknowledge the use of MapChart (https://www.mapchart.net) for creating Fig 3 in this article under a Creative Commons Attribution-ShareAlike 4.0 International License.

https://doi.org/10.1371/journal.pone.0315757.g003

In terms of research design, which is defined as the plan or proposal for guiding research that involves the intersection of philosophical underpinnings, research strategies, and specific methods [81], it was found, as shown in Fig 4, that studies approached from a quantitative perspective predominated, with 23 of the located documents falling within this subcategory. Eight studies located within the mixed methods subcategory and two within the qualitative subcategory completed the sample. The analysis shown in Fig 4 does not include documents that aim to conduct a literature review. It only includes documents that describe the implementation of an educational intervention and/or report on its results. Therefore, the total number of studies included in the analysis of both designs and research methods is 33.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Research designs.

The pie charts presents the distribution of research designs reported by the main authors of the studies. This distribution highlights the preference for quantitative approaches in the field.

https://doi.org/10.1371/journal.pone.0315757.g004

From the different research designs presented, the research strategies or methodologies found are reported, understood as “types of quantitative, qualitative and mixed methods research designs that provide specific guidance in a research design” [81]. The findings are summarised in Table 4.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Research methods.

https://doi.org/10.1371/journal.pone.0315757.t004

Studies using a quasi-experimental research methodology account for the vast majority of research within a quantitative design and total number of documents. They are followed by studies that used a pure experimental design. The document that reported the implementation of a survey [64] remains in this literature review as it addresses the characteristics of interest within an educational intervention.

In terms of mixed methods research design, the findings were categorised based on a proposed framework [82]. Three different research designs were identified. However, three documents [26, 55, 65] did not provide a categorisation in terms of the research design conducted.

On the other hand, there was a great diversity both in the age and school years of the students participating in the interventions found, as well as in the number of students involved in each of them. It is worth noting that the education systems from which the data in this study are derived have different organisation in terms of the grades considered for secondary education, so that in some cases they are not equivalent to each other. It is therefore difficult to synthesise and present the data found in relation to the school year.

There is also a significant difference in the number of participants in each intervention. In order to establish a point of analysis, a distinction can be made based on the research design. It was found that the minimum number of students involved in an intervention was 15 in a study that used a mixed research design [30]. Similarly, the study with the largest number of students involved reported a total of 442 participants [1]. Considering only the studies with a quantitative research design (quasi-experimental and true experimental methods), the average number of participants in the 22 studies included in this analysis is 100. However, it is important to accentuate that some of the documents analysed reported the inclusion of more than one experimental group due to the involvement of more than one independent variable [57, 59, 60, 74].

Finally, with the aim of exploring the characteristics of secondary school students, an attempt was made to identify factors that have been underscored by researchers as elements that may influence attitudes towards science within an educational intervention, without the need to intervene in them. In the case of the literature review studies included in this section, the general conclusions highlighted by the researchers in each of the identified variables are presented. Table 5 shows the results obtained based on the construct addressed by each of the researchers in their studies.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Factors influencing the attitudes of students in the context of an educational intervention.

https://doi.org/10.1371/journal.pone.0315757.t005

Six factors were identified which were measured in the attitudes of secondary school students and which were of interest in 8 of the 37 documents included in this study (21.6%). The variables that received the most attention were those related to sex / gender (6 studies) and those related to school year (3 studies).

Research question 1: How has the construct of attitudes towards science been treated in the literature in terms of approach, source references and scale dimensionality?

It has been noted that defining attitudes towards science encompasses a multitude of concepts that refer to activities carried out inside and outside the classroom, as well as different components that can be included in unidimensional or multidimensional constructs [8]. As a result, this phenomenon becomes a construct that encompasses a great deal of complexity. For this reason, an attempt was made to identify the general characteristics of the construct in the primary documents obtained, as well as the main theoretical references that support each of its definitions.

In a preliminary analysis of the construct of attitudes towards science, it was found that only seven of the total identified studies referred to a conceptual definition of the construct of interest (18.9%). Table 6 summarises the definitions provided by the authors of the primary interventions.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Definition of the construct under study.

https://doi.org/10.1371/journal.pone.0315757.t006

Table 6 shows the agreement between authors in identifying the construct under study as a phenomenon composed of different components, specifically affective, cognitive and instrumental. However, only in one case is there a reference to the dimension or dimensions that make up the construct [75]. In the vast majority of these cases, a general definition is used, even though the studies focus on specific disciplines. The remaining 30 studies (78.4%) simply emphasise the need to study this variable without directly defining what is meant by attitudes towards science.

As can be seen from the analyses presented, the review of the literature made it possible to identify the differences that existed in the corpus of documents with regard to the construct under study, which allowed a distinction to be made between each of them. As a result, the nature of the construct was investigated, specifically the object of interest toward which the attitudes are directed. Based on consensus in the scientific literature [5, 8], we identified articles addressing attitudes toward the following objects:

• A unitary concept of science (attitudes toward science in general);
• An integrated concept of scholar science course (attitudes toward a science course);
• A specific discipline such as biology, physics or chemistry (attitudes toward a science discipline);
• A particular topic within a discipline or scholar course, such as “evolution” or “cell functions” (attitudes toward a particular topic);
• Other.

Fig 5 shows the prevalence of studies that addressed the construct of attitudes towards science in a general way. It can also be seen that a significant proportion of these studies attempted to investigate attitudes towards some of the scientific disciplines covered by the study. In this category, the same number of studies was recorded for each of the disciplines covered (physics, chemistry and biology), with a total of 4 for each. One study investigated attitudes towards physics and chemistry [1], according to the authors, due to the characteristics of the science curriculum in the country where the research was conducted, hence there is no alignment between the count of the subcategory Attitudes towards a science discipline and the individual count of each discipline.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Focus on the construct or concept of interest.

The main constructs addressed in the studies include: attitudes toward science in general (ATSCG), attitudes toward a science course (ATSCC), attitudes toward a science discipline (ATSCD), attitudes toward a particular topic (ATPT), and constructs that do not fall into any of the mentioned categories (Other).

https://doi.org/10.1371/journal.pone.0315757.g005

Regarding the results in the "other" subcategory, studies were found that addressed attitudes towards the STEM (Science, Technology, Engineering, and Math) proposal, including Arts (STEAM), or studies that, based on the authors’ definition, addressed attitudes towards science learning.

Within the general structure of the construct, the aim was to determine whether the attitude variable consisted of one or more dimensions. In this regard, 21 studies (56.7%) used a multidimensional construct, 6 (16.2%) were based on a unidimensional construct, 8 studies (21.6%) did not provide details on the specific characteristics of the concept under investigation, and 2 other studies (5.4%) were not applicable for this analysis. For studies based on a multidimensional construct, it was examined whether the results were presented by dimension. It was found that 13 studies presented this type of data, representing 35% of the total sample.

For this study, it was of interest to identify the references of the attitude constructs mentioned in the documents. On this basis, three categories were established: 1) when the theoretical proposal/conceptual definition was taken from only one reference; 2) when the theoretical proposal/conceptual definition was based on an adaptation from different authors; and 3) studies that did not specify the origin of the term or that did not apply to such an analysis.

In Fig 6, it can be seen that the documents reporting a study that focused on the construct of attitudes towards science tended to adopt a conceptual definition from a single author, as 10 out of 15 studies (66.6%) that addressed this construct did so. For studies that investigated attitudes towards a science course, it was found that all of them used only one reference (seven documents). As far as the documents presenting research on attitudes towards a specific discipline are concerned, there is a diversification in the origin of the references to the phenomena of interest, since five studies (45.4%) were based on an adaptation by more than one author. This subcategory was the only one that presented this phenomenon. Finally, the studies dealing with attitudes towards other types of constructs showed a tendency to adopt the position of a single reference.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. The origin of the theoretical and conceptual foundations of the construct of interest.

Source of theoretical-conceptual support for the attitude constructs of interest. This figure illustrates the origins of the references used for each attitude construct mentioned in the reviewed studies. The constructs include attitudes toward science in general (ATSCG), attitudes toward a science course (ATSCC), attitudes toward a science discipline (ATSCD), and constructs that do not fall into any of the mentioned categories (Other). Total references per construct are also displayed.

https://doi.org/10.1371/journal.pone.0315757.g006

On the basis of the references found, it can be noted the coincidence of two theoretical-conceptual proposals in the documents that were part of the analysis: 1) that of Benli´s proposal, originally from 2010 and cited in his later work [27] on the construct of attitudes towards science, which is also referenced in another study [3]; and 2) that of Prokop et al.´s proposal [83], on attitudes towards biology, which is cited in three documents [57, 59, 76].

On the other hand, an examination was made of the information gathering instruments used to estimate the variable related to students’ attitudes. It was found that all studies, with the exception of those that aimed to conduct a literature review, used a questionnaire or an inventory with a Likert scale, usually of 5 points, as an information gathering instrument. Some studies complemented the information collection process with other instruments, such as two studies that reported the use of semi-structured interviews [26, 30] and one that indicated the use of an unstructured interview [62]. In addition, the use of an observation format as an information-gathering tool was documented, but the aim was not to assess students’ attitudes towards science but rather to examine the extent to which the designed intervention was characterised by the proposed methodological strategy.

Discussion

The review of 37 primary documents published in the last decade has allowed a deeper understanding of the key characteristics that define educational interventions, emphasizing the socio-demographic variables of importance, the theoretical-conceptual underpinnings and the facets of the proposed instructional designs. In addition, an attempt has been made to identify the main challenges and recommendations identified by authors in the course of intervention development, culminating in the identification of elements that guide best practice in the promotion of positive attitudes towards science.

The primary finding of this study highlights a prevalent trend in intervention programmes aimed at cultivating positive attitudes towards science: a lack of examination or delineation of specific components inherent in the instructional design, with a predominant emphasis on the outcomes achieved. This observation aligns with findings from a literature review indicating that the majority of the studies reviewed reported outcomes related to students’ attitudes, motivation and interest mainly through questionnaires or interviews, without delving into an evaluation of the educational interventions in terms of their constitutive elements [19]. The most common category in this analysis, referred to as ’portraits of students’ interest, motivation and attitudes’, focused on measuring the impact of interventions on these variables, relegating the analysis of the characteristics of educational interventions to a secondary role for primary researchers. While this review covered a broader scope in terms of the constructs examined, it serves as an indicative reflection of the prevailing trend observed to date regarding intervention strategies aimed at shaping students’ attitudes. The latter construct, along with interest and motivation, is recognised as an integral part of the affective dimension of learning [1, 21, 67].

It’s worth noting that more than half of the studies examined were carried out exclusively with students from just two countries: Turkey and the United States, together accounting for 51.3%. In addition, only a small number of studies were conducted in Latin America (3). These findings raise important questions about whether the concentration of research in specific countries has influenced the results in terms of attitudes towards science and, ultimately, the level of scientific literacy of the participants—a goal expected by the international education community and enshrined in various references that describe the competencies students should have in science by the age of 15 [12].

Furthermore, these findings prompt reflections on the establishment of a well-supported field of science education research (SER) on these countries, a trend previously identified in other literature reviews [19, 51]. The prominence of these two nations in the SER landscape has been consistently stressed, with both being ranked among the top 10 most productive countries in the field [84, 85]. Several factors may explain this trend, including the emergence of initiatives aimed at improving science education, the early development of science educations as an academic discipline, the presence of graduate programs specializing in this area, or financial support for researchers [85]. In the case of Turkey specifically, studies on attitudes and perceptions has been reported as one of the top three most frequently investigated topic by science education researchers. This area of research may also be influenced by the growing interest among Turkish scholars in publishing in international journals, driven by national promotion policies [86].

Most importantly, the results underscores the necessity of capturing the diversity of educational contexts in scientific literature, especially regarding the specific characteristics of science education interventions that other regions. We also acknowledge that this absence of information may be due to research from other areas being reported using different indicators. This issue will be revisited in the next section.

Quantitative research designs, including quasi-experimental and true-experimental methods, were also found to be predominant in the exploration of this research area. The prevalence of these approaches has several advantages, including the use of robust statistical techniques and the early detection of potential threats to the internal validity of experiments. However, unlike previous literature reviews that have predominantly examined quantitative research findings, the data presented in this study provide insights that are crucial for understanding the intricacies of interventions aimed at cultivating attitudes towards science, an area of research previously highlighted as lacking [5].

In terms of student demographics, an average enrolment of 78 students was found across the interventions, with instances of students being allocated to multiple experimental groups [57, 59, 60, 73, 74]. There were no discernible correlations between attitudes towards science and variables such as gender, socioeconomic status or ethnic background. However, a notable negative correlation was observed with school year, as discussed above. Conversely, positive effects were found for variables related to academic achievement and self-efficacy. While these findings are valuable in identifying gaps in the factors influencing students’ attitudes, it is advisable to treat such findings with caution. Firstly, the data includes a diverse age range within the student population [19], and secondly, some reports come from studies that go beyond attitudes towards science alone to include a wider range of variables or subjects [64]. Both of these factors may significantly influence the results obtained.

Among the studies analyzed, few specifically addressed the barriers faced by minorities in engaging with science, despite existing research showing disparities in science engagement among certain social groups, such as women, Black, and Latinx individuals [87]. While some studies did not report significant differences, this could be attributed to methodological limitations, such as the absence of intersectional and longitudinal approaches or insufficient sample sizes to analyze diverse identities [88]. These limitations may lead to an underestimation of the obstacles faced by disadvantaged groups and provide an incomplete understanding of the external factors that influence attitudes toward science [89]. In fact, some studies suggest that structural discrimination may contribute to negative attitudes toward education [87] and limit access to resources that could enhance scientific engagement. Additionally, the intersection of race, gender, and socioeconomic status often exacerbates these challenges, perpetuating a cycle of underrepresentation in science and technology fields [90].

Adopting an intersectionality approach that includes a broader range of factors in the analysis could be more effective in identifying how ethnicity, class, gender, and other social categories influence the relationship between certain communities and science [91]. For example, some studies have demonstrated differences in science engagement between Latinx boys and girls, as well as between youth in urban and suburban areas [90]. This suggests the need for an approach that incorporates multiple intersecting factors. Furthermore, studies conducted in diverse contexts and populations are essential for understanding the ways in which underrepresented students engage with science [91]. In summary, there is a pressing need for research that gathers robust validity evidence on the experiences of underrepresented groups in science education.

The study has revealed a divergence in the conceptualisation of attitudes, depending on the specific object towards which these attitudes are directed. Apart from authors who emphasise the construct within disciplines that include science (such as chemistry, physics and biology), notable differences emerge when attitudes towards science are assessed in a broader context. This includes students’ feelings and values about scientific endeavour, which are not limited to a specific academic subject, in contrast to attitudes towards science courses, which are more oriented towards activities or practical engagements within the school environment. This finding resonates with previous findings that underlined the potential disparity between attitudes towards science in a general sense and those experienced within the confines of formal education, highlighting the need to recognise this distinction [8]. Nevertheless, the current research calls attention to instances where the delineation of this boundary proves to be complicated, as certain studies examining attitudes towards science include factors related to enjoyment of science education [54, 66, 78].

The above observations may suggest, as mentioned earlier, a lack of consensus on the conceptualisation of the construct, pointing up a significant opportunity to adopt precise theoretical definitions in the study of attitudes. This ambiguity is reflected in certain findings. For example, some studies use the constructs of attitudes towards science and attitudes towards science education interchangeably, leading to potential confusion about the phenomenon under investigation. In addition, a notable feature of these studies is the wide range of theoretical frameworks employed by researchers; despite the relatively modest size of the corpus of primary documents in this review, congruence in theoretical-conceptual propositions was found in only two specific instances. A similar pattern emerges with regard to the tools used to collect data, each of which incorporates different factors or dimensions. Finally, the limited percentage of documents that provide a concrete definition of the attitudinal construct suggests that this variable may play a subordinate role compared to the priority given to other variables in these studies, such as academic achievement.

The top three teaching approaches identified, in descending order of prevalence, were: immersive learning environments rich in information technologies, research and collaborative elements. However, while it is evident that researchers can improve attitudes towards science through the use of electronic devices, software or computer-mediated interventions—due to their ability to facilitate the creation of mental models and enable direct observation of phenomena that are otherwise difficult to access in real, controlled and authentic contexts, alongside the use of various online scaffolding tools [26, 74] the optimal effectiveness of this teaching strategy has yet to be conclusively demonstrated.

In this context, various aspects integral to the design and implementation of each proposal were examined, including the duration of the interventions. Although this factor has been recognised as crucial for promoting positive changes in students’ attitudes towards science, the available data present a nuanced picture. For example, interventions lasting as little as two weeks have been associated with favourable outcomes [58, 65], in contrast to cases of significantly longer interventions that failed to produce notable improvements in students’ attitudes [1]. To reconcile these discrepancies, it would be prudent to assess the frequency of sessions delivered in terms of effective teaching hours, rather than simply considering the duration in weeks. However, the analysis of such data is hampered by the fact that many documents do not provide this information. Furthermore, the results of the present study show a consistent pattern among interventions that did not produce significant changes in students’ attitudes: they were predominantly delivered in conventional classroom settings, lacked an evaluation process for the intervention design, and did not specify whether an evaluation of the implementation process was carried out in line with the study’s objectives [1, 66–70] with only one exception.

Ultimately, the research questions served as a lens through which to identify key challenges and recommendations from researchers in their respective studies. While these data were not prevalent in the majority of documents, probably because the focus was on establishing the presence of an effect by presenting an independent variable rather than scrutinising the educational intervention itself, the pertinent elements that were identified are discussed below.

Researchers have identified a central challenge related to the difficult task of shaping students’ attitudes towards science [67–69, 79]. Some scholars highlight the complexity of changing affective dimensions such as attitudes, a claim supported by cases where interventions led to increases in other variables such as academic performance rather than the targeted attitudes [66, 67]. It is therefore clear that the allocation of time for this purpose is of paramount importance, requiring a clear delineation of its use within the educational intervention. Conversely, a critical gap that has been identified relates to the lack of adequate training in the proposed educational approach, whether for the educators tasked with leading the intervention or for the students to familiarise themselves with it [69, 75]. This shortcoming has been implicated in the failure to achieve the objectives related to attitudes towards science.

Coupled with the findings from the segment on the challenges of educational interventions, there is an imperative to assess the lasting impact of attitudes long after programme implementation [18, 65, 70] to tailor interventions specifically to minority cohorts [64], and to subject proposed instructions to pre-implementation scrutiny [61]. These instructions will act as a catalyst for further investigation, strengthening the aims of the trial and establishing central benchmarks for reporting standards for educational interventions. This aligns seamlessly with the assertion regarding the need for comprehensive delineation of interventions [19]. Ultimately, these recommendations stand as pillars to guide the development and evaluation of educational methodologies. Notably, our study reveals that only seven documents acknowledged the use of evaluation tools to assess programme implementation against objectives or intervention protocols [26, 54, 55, 59, 68, 75, 77], with only four considering tools or methods to assess programme design [56, 61, 65, 79].

Conclusions

This scoping review has outlined and described the impact of the educational interventions aimed at fostering positive attitudes towards science. The analysis of the corpus of documents revealed a discernible distinction in the concepts used to explore students’ attitudes, reflecting different objects of interest, such as attitudes towards science versus attitudes towards a science course. While these may represent distinct phenomena, there is a pervasive lack of clarity regarding their delineation, as evidenced by their interchangeable use within certain documents reviewed. There is therefore an urgent need for further research into their operational definition, facilitating the establishment of differentiating dimensions or a hierarchical organisation of the constructs.

In this review, we also identified salient features of educational interventions that demonstrated different pedagogical approaches to fostering attitudes towards science. However, none of these approaches showed clear effectiveness for their intended purpose. Consequently, there is an urgent need to assess the impact of other components within educational interventions, such as the competence of those overseeing implementation, evaluation of the developmental process, or the duration of effective instruction—factors that significantly influence the achievement of stated goals. This underlines the importance of comprehensive evaluation and refinement to optimise the effectiveness of interventions targeting attitudes towards science.

The identification of the main challenges and recommendations identified by the primary researchers in their studies also underlines the need to critically assess the timing of the development of interventions aimed at fostering positive attitudes towards science. It also underscores the need for a proactive approach by both teachers and students to the proposed programme. Finally, the recognition of intrinsic student variables influencing their attitudes, particularly the nuanced factors of racial/ethnic identity and self-efficacy, underscores the need for comprehensive exploration and elucidation. Remarkably, these variables have remained largely unaddressed in interventions, representing an untapped area for understanding the intricate interplay between these constructs and attitudes towards science. Thus, there is a compelling mandate to forge instructional designs that conscientiously incorporate and address these elements in order to realise their desired outcomes.

Supporting information