Introduction

Nowadays technology plays an essential role in our society, and its importance is expected to increase in the future. For this reason, it is essential that students are interested in following technological careers (Zapata-Ros, 2015). Several factors influence scientific and technological vocations, and students' attitudes towards technology are considered to play a significant role (Ardies et al., 2014; Regan & DeWitt, 2015). It is important to define at this point the term “attitude” as a multi-dimensional construct (Ankiewicz, 2016) which comprises a broad variety of aspects such as enthusiasm, enjoyment/boredom, interest, career aspirations or perceived difficulty (Ardies et al., 2014).

Due to the fact that students’ attitudes towards Science, Technology, Engineering and Mathematics (STEM) education in advanced societies are declining (Institution of Engineering and Technology, 2019; Tytler et al., 2008), there is an ongoing international concern about this matter (DeWitt et al., 2013). Given that early exposure for elementary students to STEM initiatives motivates them to enrol in science and math in later years (DeJarnette, 2012), and that future interests in science and technology are generally formed before the age of 14 (Ardies et al., 2014), it seems valuable to address this issue earlier in an individual's schooling experience.

School-aged children’s attitudes towards technology are a relevant factor when choosing technological careers, and, for this reason, are in the focus of research. A number of studies have addressed these attitudes in secondary education students. For example, Autio et al. (2015) studied the differences in 11 to 13 year-old students' attitudes towards craft and technology education in Finland, Estonia and Iceland; Becker and Maunsaiyat (2002) assessed attitudes and understanding of technology among students before a technology curriculum reform; Rupnik and Avsec (2019) related attitudes towards technology and technological literacy in students between 11 and 14 years old; and Van Rensburg et al. (1999) assessed learners' attitudes towards technology in grades 9 and 10 in South Africa. Nonetheless, there is a scarcity of research at primary education levels, as Ankiewicz (2016) shows in an extensive review.

Regarding gender and attitudes towards technology use, the differences between boys and girls have been revealed through research from different perspectives. For instance, Virtanen et al. (2015) attested differences in terms of the motivation towards technology learning in primary school students (grades five to six). From another perspective, Padilla-Meléndez et al. (2013) carried out a study with a sample of 484 undergraduate students and identified significant differences between males and females regarding playfulness, attitude towards technology and intention to use it in a blended learning scenario. In particular, although the females showed higher scores with regard to playfulness and attitude, the males’ rating of intention to use indicated that boys were more prone to use technology than girls.

According to a recent meta-analysis concerning gender and attitudes toward technology, there has been a minimal reduction in the gender attitudinal gap in the last 20 years, with males still having a more favourable attitude toward technology use than females (Cai et al., 2017). To date, there are studies analysing gender differences in attitudes towards technology in higher education (Al-Emran et al., 2016; Huffman et al., 2013; Sáinz & López-Sáez, 2010), but again, few studies have addressed this issue at elementary levels, especially in the early years of this stage.

Even though children start engaging with their future careers around age 10, they start choosing gender-adequate activities and interests at age 3 (Buccheri et al., 2011). Therefore, this is another strong reason to analyse their interest towards technology in the early stages of Primary Education.

Computational Thinking (CT) is one of the skills related to STEM education (Weintrop et al., 2016), and consists of “solving problems, designing systems, and understanding human behaviour, by drawing on the concepts fundamental to computer science” (Wing, 2006, p. 33). The importance of CT in the future was predicted by Wing (2006) in her influential article where she stated that “this kind of thinking will be part of the skill set of not only other scientists but of everyone else” (p. 34). As many have pointed out, there are countless opportunities to integrate CT in primary, middle and high-school (Lu & Fletcher, 2009), so students must begin to work with it in K-12 (Barr & Stephenson, 2011), or even from the first year in primary school (Djurdjevic-Pahl et al., 2017).

Regarding its teaching in schools, two main approaches are being used. The plugged approach, which involves computer programming exercises, and the unplugged approach, which involves activities that do not require the use of digital devices or any specific hardware (Brackmann et al., 2017).

Several studies have been conducted to analyse the development of young students’ CT skills when subjected to specific CT or robotics instruction. Brackmann et al. (2017) proved with 5th and 6th grade students that an unplugged approach may be effective for the development of this ability. Likewise, Chalmers (2018) and Chiazzese et al. (2019) demonstrated how these skills also increased after integrating robotics and coding into 1st to 6th and 3rd to 4th grade classrooms, respectively. The interventions were also successful in del Olmo-Muñoz et al. (2020), where the CT skills and motivation of 7 and 8 year olds towards plugged and unplugged approaches were assessed, revealing that the inclusion of unplugged activities in CT instruction at these ages is beneficial to both variables or aspects. Other research has explored the relationship between computational work through visual programming with other cognitive processes. For example, Falloon (2016) analysed the relationship of basic coding activities with general and higher order thinking skills of 5 and 6 year old students.

As evidenced by a recent meta-analysis (Higgins et al., 2019), many other studies have been based on analysing or promoting attitudes towards STEM through the use of technology. For instance, Eyyam and Yaratan (2014) designed mathematics lessons for the seventh grade of secondary education (13 years old), with and without using technological tools, and then measured academic achievement and attitudes towards technology use in class, finding that students had a positive attitude towards technology use, and that those who used technology obtained better results in mathematics. Witherspoon et al. (2018) examined learning and motivation in middle school (6th–8th grade) through educational robotics using a visual programming language and obtained positive results in relation to skill acquisition and attitudes towards programming. Nugent et al. (2010) tested the impact of two interventions, also concerning robotics and geospatial technology, on children’s (mean age of 12.28 years) learning of, and attitudes towards, STEM. Their results showed greater learning and an increase in attitudes and motivation in the group that carried out the instruction. Similarly, the study by Leonard et al. (2016) analysed the CT strategies and attitudes towards STEM of fifth to eighth grade students after engaging in robotics and game design. Their results attested that self-efficacy on videogaming increased more in a combined robotics/gaming environment, compared to a gaming-only context, and that student attitudes towards STEM did not change significantly. Also relating robotics and coding games to children's attitudes towards coding activities, Sharma et al. (2019) noted that highly engaging and collaborative coding activities significantly impacted children's (aged between 8 and 17 years old) attitudes. In a similar vein, Merino-Armero et al. (2018) showed how robotics-based CT instruction positively influenced the motivation of third-grade students.

Some studies have focused on girls’ attitudes, such as the research study by Mammes (2004), conducted during the 3rd year of elementary education, in which students were exposed to technology education and then the change in their interest in technology was compared, showing that both boys' and girls' interests were fostered, and that gender differences were significantly reduced. Master et al. (2017) tested an intervention with 6-year-old children to develop girls’ STEM motivation by programming robots, reporting, for these girls, higher technology interest and self-efficacy (confidence in one’s ability to succeed in a specific task), and not reporting a significant gender gap with the boys. Again with a robotics experience, this time in high school (grades 5–7), Kucuk and Sisman (2020) analysed students' attitudes towards robotics and STEM, concluding that gender has no effect on STEM attitudes, but it does have an effect on an individual’s confidence and desire to learn robotics. Likewise, Screpanti et al. (2018) presented a project to raise and study interest in STEM education and careers among K-12 students (and girls in particular) from 11 to 13 years old through robotics. The experience was positive according to the authors, since the girls improved in terms of achievement and interest. Expanding the age range, Marth and Bogner (2019) monitored the interests and social aspects of technology in different age groups, including adults, indicating a gender gap across all age-groups, showing a significantly higher interest in males.

The research presented here is part of a larger project that includes multiple objectives. While other project works have analysed the effectiveness of instruction in the development of CT skills or motivation del Olmo-Muñoz et al. (2020), this one is cognisant of the aforementioned studies and the necessity for further research in terms of interest and attitudes (Boston & Cimpian, 2018; Cai et al., 2017; Regan & DeWitt, 2015) by proposing CT instruction with unplugged and plugged activities, and focusing the analysis on extrapolating results to signal how the instruction affects students' attitudes towards technology.

To this end, the following research questions are posed:

RQ01.:

Does CT instruction improve pupils' attitudes towards technology?

RQ02.:

What approach, unplugged-plugged activities or plugged activities only, is more effective in improving pupils’ attitudes towards technology when introducing CT in the early years of Primary Education?

RQ03.:

Is there a gender gap in attitudes towards technology in the early years of Primary Education?

RQ04.:

Could CT instruction be beneficial to reduce an eventual gender gap?

RQ05.:

What approach, unplugged-plugged activities or plugged activities only, is more effective in reducing an eventual gender gap when introducing CT in the early years of Primary Education?

Method

Design

In this section we describe the research design. The study was developed with natural groups of students so, given the impossibility of randomly assigning participants to the conditions, a quasi-experimental study was conducted. The development of the experiment consisted of five 45-min sessions (one per week) of instruction which in turn were divided into two phases, as well as a pre-test session and a post-test session. Figure 1 summarizes the different phases of the experiment.

Fig. 1
figure 1

Experiment design

Full size image

During the first session, the students took the pre-test for the evaluation of their prior attitudes towards technology. After this, the instruction began. It was divided into two different phases: the first one, in which both groups participated in three sessions, consisting of different activities (an unplugged instruction and a plugged instruction), and the second one, in which both groups participated in two sessions, consisting of same set of plugged activities. Once the second phase of instruction finished, the participants from both groups were invited to do the post-test following the same method used for the pre-test.

Participants

The sample of our study was composed of 84 students enrolled in the 2nd grade of Primary Education (7–8 years old), from four groups of three different schools located in Albacete (Spain). The unplugged group (unplugged-plugged activities) and the plugged group (plugged activities) were made up of 42 students each. With regard to the gender of the participants, its homogeneity can be verified in Table 1, which presents the demographic data of the sample.

Table 1 Demographic data of the sample.
Full size table

Instrument

This section details the instrument used in the pre- and post-tests to assess the participants' attitudes towards technology. Attitude is a multi-dimensional concept with many factors of influence (Osborne et al., 2003), and this has to been taken into account when selecting the instrument for its measurement. Students' attitudes towards technology have been investigated for just over three decades (Ankiewicz, 2019), with the Pupils’ Attitudes Toward Technology (PATT) survey, initially developed by Raat and de Vries (1985), being the most notable used instrument. However, its length (80 items) and its non-validation for use with young children (Frantom et al., 2002) have led it to be reconstructed and revalidated in several forms (Ankiewicz, 2016), with the validated reduction by Ardies et al. (2013) being the instrument selected for this study.

Thus, the instrument is a questionnaire consisting of 24 items and six factors or dimensions: Technological career aspirations (A), Interest in technology (I), Tediousness towards technology (T), Consequences of technology (C), Technology is difficult (D) and Technology is for both boys and girls (G).

The choice of answers is guided by a Likert scale, which are ordinal scales used to determine the levels of agreement or disagreement of the students. For each item there is a 'five-point choice' consisting of 'strongly disagree/disagree/not sure/agree/strongly agree', supported by emoticons for pupils to freely indicate their feelings about each statement.

Something that is often done, because it is inherent to this type of testing, is to fit the instrument to the regional context (Ardies et al., 2013). In our case, the language of the test was changed to Spanish, and the context-specific items had to be adapted (e.g., we talk about 'extracurricular technology activities instead of 'school technology club). Table 2 shows the original instrument items and dimensions in detail. To facilitate its tracking, the original item number of the PATT-USA questionnaire (Bame et al., 1993) is included before each question, as it also appears in Ardies et al. (2013).

Table 2 Instrument items and dimensions.
Full size table

In the statistical analysis, the scales were inverted for the negative items, so they were oriented in the same direction as the positive items. As a consequence, higher scores in any dimension are associated with more positive outcomes.

Procedure

The procedure design of the instruction sessions is based on different courses from the Code.org website, which are a good option to help students lay the foundations of computer science, thanks to their characteristics (Kalelioğlu, 2015).

The source courses correspond to the 2nd level of primary education, these being: Course 1, for early-readers who have little or no previous computer science experience; Course 2, for readers who have little or no previous computer science experience; and Course B developed with first-graders in mind, and tailored to a novice reading level.

The selection of activities for each session, and its order, is based on working on similar computational concepts with both groups, following, as faithfully as possible, the original sequencing proposed in the guidelines of the creators (Code.org, 2018a, 2018b), so that the concepts (directions, sequences and loops) are worked on in increasing difficulty through the lessons. These equivalences are explained below for the first three instructional sessions.

In Session 1, the activities for both groups served as an introduction and were taken from the same course (Course 2),Footnote 1 where they appear almost consecutively; first the unplugged and then the plugged activity.

For Session 2, even though the activities do not belong to the same original course, both groups work on the computational concept of sequences and on the computational practice of debugging. The unplugged activity corresponds to Lesson 6 in Course B,Footnote 2 which is immediately followed in that course by equivalent activities to the selected plugged ones from Course 1Footnote 3 (Lessons 5, 7 and 8).

The same thing occurs in Session 3, where loops are worked on. The unplugged activity corresponds to Lesson 9 in Course B, with Lessons 10 and 11 in that course being equivalent to the selected plugged activities from Course 1 (Lessons 13 and 14 in that course, respectively).

For further appreciation of the mentioned equivalences, some of the activities are presented in Table 3, and a detailed explanation of the sessions can be found at del Olmo-Muñoz et al. (2020).

Table 3 Examples of activities.
Full size table

Results

Considering the research questions posed for this study, this section presents the results obtained related to each research question. As the data gathered with the PATT survey are ordinal, non-parametric tests have been employed in the analysis. In particular, in the comparison between different groups the Mann -Whitney test is employed, while the Wilcoxon signed-rank test is used when comparing related conditions. Statistical significance is considered at a confidence level of 95%. Additionally, we reported r as a measure of the effect size for all the comparisons.

RQ01.: Does CT instruction improve pupils' attitudes towards technology?

In order to answer RQ01, Table 4 shows the results in the pre-test and the post-test. The results suggest that the pupils’ attitudes slightly improved in the post-test compared to the pre-test. However, as the table includes the results split by dimensions, a differential impact of CT instruction across dimensions should be noted; thus, in the case of ‘Technology is difficult’ a statistically significant improvement took place. After the intervention, the students perceived technology as more accessible. Indeed, the gain in this dimension could be considered as medium-sized according to Cohen (1992). Furthermore, even though the participants had very high scores in the dimension ‘Technology is for both boys and girls’, the results were even more positive after the intervention.

Table 4 Pre- and post-test results.
Full size table

RQ02.: What approach, unplugged-plugged activities or plugged activities only, is more effective in improving pupils’ attitudes towards technology when introducing CT in the early years of primary education?

Table 5 includes the results of both the unplugged and plugged groups in the pre-test and the post-test. The differences between both test scores reveal that whereas the unplugged group slightly improved their interest in technology, their perceptions about the consequences of technology worsened, the plugged group significantly improved in the ‘Technology is difficult’ and in the ‘Technology is for both boys and girls’ dimensions. Specifically, with regard to the latter dimension, this improvement can be considered as medium-large sized.

Table 5 Pre- and post-test results based on group.
Full size table

RQ03.: Is there a gender gap in attitudes towards technology in the early years of primary education?

Table 6 presents the results of each of the groups split by dimension and gender in the pre-test, in order to answer RQ03. These results attest that there were initially significant differences based on gender in several dimensions. In this respect, boys had more technological career aspirations and a higher interest in technology, while girls tended to consider that technology is difficult to a lesser extent. Overall, the results indicate a medium-sized effect size.

Table 6 Pre-test results based on gender.
Full size table

RQ04.: Could CT instruction be beneficial to reduce an eventual gender gap?

Table 7 reveals the results separated by dimension and gender in the post-test, which shows a reduction in the gap after the instruction. Despite the fact that after the intervention there was still a significant difference in favour of males concerning their technological career aspirations, it should be noted that both boys and girls improved their post-test scores in comparison to the pre-test.

Table 7 Post-test results based on gender.
Full size table

Regarding the gender divides identified in the pre-test in the dimensions ‘Interest in technology’ and ‘Technology is difficult’, both gaps narrowed after the instruction. In this regard, it should be said that the reduction in the interest gap is simultaneously caused by a small decrease in the boys’ interest, and by an improvement in the females’ interest in technology.

RQ05.: What approach, unplugged-plugged activities or plugged activities only, is more effective in reducing an eventual gender gap when introducing CT in the early years of primary education?

Table 8 summarizes the pre- and post-test results broken down by group and gender, facilitating analysis of RQ05. Thus, it can be verified that the gap is reduced in most of the dimensions, as seen below.

Table 8 Pre and post-test results based on group and gender.
Full size table

Technological career aspirations (A): The gender gap is reduced in the unplugged group, where it decreased from a large effect size before the intervention to a medium-sized effect after the instruction. In this case, although the boys also improved their scores in this dimension, the reduction is due to a greater improvement among the females. However, the gap widens in the plugged group (from r = 0.29 to r = 0.38), due to two factors: aspirations improved among the boys and decreased among the girls.

Interest in technology (I): The gap narrows in the plugged group, changing from a medium to a small effect. This reduction is largely based on the fact that the girls' interest increased after the intervention. It is noteworthy that the gap disappears in the unplugged group, since it was medium-sized in the pre-test.

Tediousness towards technology (T): Scores in this dimension reveal no significant differences before and after the intervention for the unplugged group. However, in the plugged group a reduction in the differences between males and females took place due to an improvement in the girls’ attitude after the instruction. In this case, the divide decreased from a medium to a small effect.

Technology is for both boys and girls (G): This dimension started from optimal levels in the pre-test for both the boys and girls in the unplugged group; however, it worsened a little in the post-test in the case of the boys; there was an increase in the gap from a small to a moderate effect. The opposite occurred in the plugged group, since although the girls remain at maximum levels after the instruction, the boys, who in this case did not obtain such high levels in the pre-test, did better in the post-test, completely reducing the gap (from small to a non-significant effect).

Consequences of technology (C): Regarding the perception of the effects of technology, the initial picture of the results showed that these were higher for the girls in the unplugged group (medium-sized effect), and higher for the boys in the plugged group (small-sized effect). On the other hand, the post-test scores show us that only the girls in the plugged group improved in this dimension. Consequently, although to a lesser extent, the small initial gap remains in the unplugged group (where girls continue to have higher levels), while the gap completely disappears in the plugged group.

Technology is difficult (D): This dimension is the one with the lowest levels in all senses, for the boys and girls, in both the plugged and unplugged groups, and in both the pre- and post-test. Regarding the gap between the boys and the girls, the girls in both groups obtained better results in the pre-test (medium-sized effect in both cases), but the same thing did not occur in the post-test in the case of the unplugged group, where the girls worsened while everyone else improved. Thus, it revealed that the gap remains in both groups, as effect sizes for both plugged and unplugged groups are still medium-sized in the post-test.

Discussion

The results of this study reveal important findings and provide answers to the five research questions, whose order, for the sake of clarity, will be followed again in the discussion section.

As regards RQ01, "Does CT instruction improve pupils' attitudes towards technology?", observing the results of similar studies, we find that Nugent et al. (2010) and Eyyam and Yaratan (2014) reported significant improvements in attitudes towards technology or STEM, in general, after their proposed instruction. In our case, the results would be more in line with those of Leonard et al. (2016), where there was no statistically significant change in the participants’ attitudes towards STEM or STEM careers. Despite the fact that in other studies a short intervention improved the students’ attitudes as "relatively superficial nature of the short-term activities led to perceived goal achievement with little effort" (Nugent et al., 2010, p. 404); this could be because the sole objective was to develop attitudes, and in that case small workshops focused on it can be very effective. However, the main objective of the present project was the development of CT, and the modest improvement appreciated in the pupils’ attitudes in this case could be due to the short duration of the proposed instruction, since “a larger amount of time spent on learning about technology correlates with a higher interest” (Ardies et al., 2014, p. 60). Besides, the improvement should be valued insofar as the participants already started in the pre-test from considerably high levels in most of the dimensions. Taking all of the above into consideration, the answer to RQ01 would be affirmative, but it does not occur to the same extent in all dimensions. Thus, improvements in the belief that technology is for boys and girls, and in aspirations, stand out, and even more significant is the improvement in the perception of difficulty.

Regarding the comparison between groups questioned in RQ02, "What approach, unplugged-plugged activities or plugged activities only, is more effective in improving pupils’ attitudes towards technology when introducing CT in the early years of Primary Education?", although the improvement in attitudes was slightly greater in the plugged group, the results do not offer a resounding answer in one direction, since there were only small differences in some dimensions. Therefore, it cannot be conclusively determined that one instruction reported more benefits than the other in terms of attitudes in general, but the analysis of the dimensions separately can be taken into account when advocating different educational approaches in the future. On the one hand, the unplugged group generally showed moderate improvements across dimensions. On the other hand, the fact that the plugged group came out better in their perception of the difficulties may be due to their greater exposure, or direct contact with coding. This was discussed by Sáez-López et al. (2016) when highlighting the benefits of visual programming in primary schools to alleviate the perception of the difficulty of computer science by students, as well as improving its perceived usefulness. The improvement in this group is also greater in the belief that technology is for both boys and girls, something which could be explained by observing that the unplugged group started at higher levels.

In relation to RQ03, "Is there a gender gap in attitudes towards technology in the early years of Primary Education?", Ankiewicz (2019) reviewed numerous studies from the last 30 years on pupils’ perceptions and attitudes towards technology. Most of them dealt with pupils from ten years onwards, and they generally indicated that boys had more positive attitudes toward technology than girls. In that review it can also be seen that gender differences have been found to already exist at the age of 10. Conversely, Ardies et al. (2014) collected several studies in which, at the age of 10, interest towards STEM did not differ between boys and girls, and it was quite high. Related to this, our study provides new insights into this topic, since results from younger pupils (7–8 years) are presented. Although, on the one hand, the higher results of the boys in some dimensions ('Technological career aspirations' or 'Interest in technology') are aligned with those studies regarding the greater attitudes of boys, in our study girls excelled in the 'Technology is difficult' dimension, differing with respect to previous studies such as that of Jones et al. (2000), in which girls tended more to regard science as a difficult subject, or that of Ardies et al. (2014), in which all the students perceived the difficulty of technology equally. This being said, the remarkable differences in favour of males at this age regarding their interest in technology, or their interest in making a career in technological domains, are a major finding, as this underlines that important gender gaps are formed at early ages. These results verify the conformation of attitudinal gender gaps at early ages, as Eccles et al. (1993) demonstrated in other domains.

As for RQ04, "Could CT instruction be beneficial to reduce an eventual gender gap?", a narrowing of the gap occurred. This is mainly because the boys 'and girls' interest in technology was almost completely balanced, and they also saw the difficulty of technology in a similar way. There were positive test scenarios for the dimensions 'Technology is for both boys and girls' and 'Tediousness towards technology', where the boys’ improvement in the first case, and the girls’ in the second, resulted in a quite high, similar score for everyone in these dimensions. Finally, special mention should be made in the case of aspirations, where there was a significant improvement with a medium-strong effect size in both the boys and girls, the importance of which other authors have highlighted (DeWitt et al., 2013; Tytler & Osborne, 2012).

The short answer to RQ05, "What approach, unplugged-plugged activities or plugged activities only, is more effective reducing an eventual gender gap when introducing CT in the early years of Primary Education?", is that there was a greater narrowing of the gap in the unplugged group, although this could be because the gap was greater in that group at the starting point. The extended answer is that the gap narrowed, or at most remained the same in all dimensions for both groups, except in two cases. The first, regarding technological career aspirations, where the gap widened in the plugged group because the girls got slightly worse and the boys got better. This can be explained by the fact that boys perform better with coding than girls from an early age (Sullivan & Bers, 2016), taking into account the relationship between attitudes and learning outcomes (Metsärinne & Kallio, 2016). The second, for the dimension 'Technology is for both boys and girls', where the gap widened this time in the unplugged group because the boys got worse.

To recap, in addition to answering the research questions raised in this study, it was also the objective of the work to propose CT instruction for early ages which would lead to the promotion of children’s attitudes towards technology, and to the reduction of an eventual gender gap in that matter. Related to this, DeJarnette (2012) stated that “early exposure to STEM initiatives and activities positively impacts elementary students' perceptions and dispositions” (p. 77), and Virtanen et al. (2015) stated that girls' lower self-efficacy, and lower intellectual and practical interest in STEM field subjects, lead to a loss of potential in the fields of science and engineering, and how improving girls' hitherto negative attitudes towards STEM would avoid that loss. In this sense, it is in view of the results of this work that this objective has been also successfully achieved.

Another notable finding of the study is related to the perception of the difficulty of technology, in which the lowest levels were obtained in all cases: before and after the instruction, regardless of the approach and gender of the students. This attests that children perceive technology, or technology-related jobs, as difficult or unattainable, and given that this perceived difficulty is a predictor of decreased perceived competence, which is in turn associated with an increase in disengagement (Patall et al., 2018), future efforts should be made to improve this dimension.

As for limitations, one of them affects the instrument, in which the gender dimension might be gender-biased, since “the statements in this category seem to trigger gendered reactions to the stated gender differences” (Svenningsson et al., 2018, p. 81). This could have been reflected in the reported results, so future studies should measure attitudes towards technology with complementary instruments. Regarding the duration of the instruction, although its brevity was not an obstacle to improve CT skills del Olmo-Muñoz et al. (2020), attitudes may require more long-term work, something that could be implemented in other studies. Another limitation of the study was the impossibility of randomizing the sample, leading to the initial scores not being homogeneous in all dimensions, thus making it difficult to interpret the results. Again, future works could try to randomize the sample and, at the same time, make it more numerous in order to favour the generalization of the results.

Conclusions

In conclusion, the demands of future society point to the promotion of primary students' attitudes to STEM as an important issue to address in the current educational landscape. The results obtained here provide empirical evidence concerning the effectiveness of CT instruction in enhancing children's attitudes towards STEM in general, and technology in particular. Furthermore, this evidence responds to these demands in two ways: incorporating CT into education, and avoiding wasting any talent students may have in relation to STEM, especially in the case of females. Future possible research could delve into how the incorporation of new STEM-related approaches to education, such as the one presented here, would affect children's learning and attitudinal aspects, taking into account their different characteristics.