Infrastructure and Gender Disparity in Information Communication Technology Literacy: A Cross-Country Comparative Study

1 Introduction

Information communication technology (ICT) skills are crucial for modern human life, especially in the digital era. For example, as shown by a report of United Nations (UN) Children’s Fund (UNICEF), children’s ICT literacy are important for them to develop essential technical and critical capabilities via education while reduce children’s exposure to the risks of the digital world (UNICEF 2019). Similarly, Food and Agriculture Organization of the UN (FAO) introduces that ICT literacy such as the ability to possess and properly use of mobile phones is a normal aspect of modern life, without which may lead to socioeconomic costs (FAO 2016). In addition, it is also noticed that ICT literacy can contribute to the increase of women’s confidence and ability to seek better socioeconomic status, which would be very helpful for the empowerment of women (e.g. FAO 2016; Kiondo 2007).

Since a major source of developing ICT literacy is school education especially science, technology, engineering, and math (STEM) education (e.g. Changpetch and Seechaliao 2020), ICT literacy has attracted attention from researchers and practitioners, particularly in the field of STEM education and sustainable development. One important aspect of ICT literacy which attracted academic and practical attention is gender disparity in ICT literacy. Gender disparity in ICT literacy has been included into the international framework of indicators for Sustainable Development Goals (SDGs) to measure the progress of gender equality and providing equitable quality education (UN Statistics Division 2022). More specifically, SDG Indicator 4.5.1 develops an index to measure the gender disparity in ICT literacy at country level, with a number of specific skills identified and assessed (e.g. UN Statistics Division 2022; UNESCO 2021). It is an indicator which directly reflects the progress of science education and associated gender disparity. Similarly, under the custodianship by International Telecommunications Union (ITU), SDG Indicator 4.4.1 monitors the overall proportion of population with ICT literacy at country level (ITU 2021a), which also shows a country’s progress in science education.

Infrastructure’s influence on ICT literacy is widely perceived and explored by researchers. For example, Samarakoon, Christiansen, and Munro (2017) demonstrate that lack of stable (supply and access of) electricity is a core barrier to improve the use of ICT in education in Africa. This could be a reason of relatively weaker ICT literacy in Africa. However, with the improvement of mobile phone technology including mobile networks, the challenge of insufficient electricity can be reduced due to their portability and capacity of battery storage (Samarakoon, Christiansen, and Munro 2017). In consideration of infrastructure’s important influence on education, ICT literacy, and also sustainable development in general, it is also included into the framework of SDG indicators. More specifically, SDG Indicator 9.c.1 measures the proportion of population covered by mobile networks (2G/3G/4G), which monitors the progress of building resilient infrastructure (ITU 2021b).

Although the importance of ICT literacy has been gradually realized, it is essential to explore more on the gender disparity in ICT literacy, as ICT literacy can play an important role in the empowerment of women, especially in the digital era, given women are widely perceived to have fewer access to ICT (e.g. Liu 2022). For example, women with improve ICT skills may better use ICT facilities in information seeking and career advancement. Therefore they can obtain better socioeconomic resources. Without adequate infrastructure, the empowerment of females via ICT facilities, and women may have fewer opportunities to obtain adequate ICT literacy in comparison with males due to existing gender gap in accessing relevant resources. Therefore, infrastructure’s influence on ICT literacy especially gender gap in ICT literacy is an important research question which may generate both scientific and practical implications.

Nevertheless, as will be examined in the next section, there is insufficient research on the relationship between infrastructure and gender disparity in ICT literacy. The majority of relevant existing literature tend to focus on gender disparity at micro levels such as schools, while studies at macro levels are relatively fewer. Therefore, there is insufficient evidence to support macro level policies and interventions which aim to reduce gender inequalities in ICT literacy. Such limitations create spaces for studies on the relationship between infrastructure and gender disparity in ICT literacy with more interdisciplinary perspectives including science education, which is an essential component of sustainable development.

To reduce the above-mentioned gaps, this article uses country-level data from the SDG Global Database to investigate the relationship between infrastructure and gender disparity in ICT literacy. This study aims to generate more insights into gender equality in science education including STEM, and also enrich the knowledge of science education for sustainable development. It is also expected to generate empirical evidence and implications for relevant policies and interventions. The next section of this article will review of existing studies on gender disparity in ICT literacy, which summarizes the contribution of previous literature and further identifies the knowledge gaps to be addressed.

2 Literature Review

There is an on-going discussion on the status quo of gender gaps in ICT literacy and the associating factors. For example, gender division in STEM education achievements and attitudes towards science are usually believed as reasons of gender disparity in ICT literacy (Tam, Chan, and Lai 2020). However, Ho et al. (2020) demonstrate that the relationship between gender and learners’ achievements in STEM education has not reached a global consensus. While it is often perceived that females are less representative in STEM related subjects and have lower self-perceived ability and performance in mathematics (Anaya, Stafford, and Zamarro 2022), Aini, Rachmatullah, and Ha (2019) find that in Indonesian universities, there are substantially more females than males enrolled in STEM disciplines. In addition, it is observed that in Indonesia ‘Female students showed a higher attitude toward science than male students in general’ (Aini, Rachmatullah, and Ha 2019, p. 654). Similarly, although females are still under-representative in math-intensive STEM fields such as computer science, gender gaps in math performance are relatively small in more developed countries and countries with lower general social inequalities (Breda and Napp 2019; Breda, Jouini, and Napp 2018). Therefore, it can be inferred that in these countries gender gaps in ICT literacy would be smaller. Such disparities across countries become a reason for this article to conduct a cross-country comparative research based on a global data source which covers both developed and developing countries.

The existing literature about infrastructure’s influence on gender disparity in ICT literacy is relatively few. For example, Kiondo (2007) argues that weak infrastructure and gender inequality in education in Africa are both reasons for excluding women from accessing and using ICT. This may lead to gender disparity in ICT literacy. Similarly, Williams, Istifanus, and Ajufo (2017) demonstrate that in Nigeria, inadequate school infrastructure is a barrier for girls to access education including STEM education. This is a possible reason for females’ relatively lower ICT literacy. Based on evidence from Ghana, Opoku and Kuranchie (2014) find that infrastructure development policies such as ‘One Laptop Per Child’ can support to improve students’ ICT literacy, and although female students have not received preferential treatment in ICT education, they show stronger positive attitude to computer education. Therefore, it can be inferred that female students may develop better ICT literacy if the essential infrastructure is available and they do not suffer disadvantages in accessing ICT resources. The scarcity of existing studies on the relationship between infrastructure and gender disparity in ICT literacy, as well as their strong focus on developing countries with relatively low incomes, call for more empirical evidence to explore whether and how infrastructure affects gender disparity in ICT literacy, including cross-country comparisons. It is a core reason for this article to use the SDG Global Database for examining infrastructure’s influence on gender gaps in ICT literacy.

The existing literature has made substantial contribution to the measurement of ICT literacy and the associated gender gaps. For example, Dimitrov (1999) argues that previous studies have emphasized the use of multi-choice questions to measure learners’ skills in science-related subjects (including those strongly related to ICT) while there is more space for using open-ended questions for such measurements. Based on a case study in Germany, Jansen, Schroeders, and Lüdtke (2014) find that the self-concept and achievements of science including ICT subjects are multi-dimensional. Ho et al. (2020) use the average scores in the school level standard tests to measure learners’ achievements in STEM, including essential ICT literacy. However, these advancements in measurements are mainly at micro level. As demonstrated by Hafkin and Huyer (2007), the paucity of reliable data and statistics to measure gender gaps in ICT literacy becomes a challenge of conducting further research and providing relevant policy recommendations in this field. The limitations in gender-disaggregated statistics and indicators is a key consideration for this article to use the SDG Global Database and investigate the relationship between infrastructure and gender disparity in ICT literacy with selected SDG indicators, which will be explained in more detail in the next section.

3 Research Methodology

4 Research Results

Pearson’s correlation is used to have an initial examination of the relationship between infrastructure development, gender gap in ICT literacy, and the overall ICT literacy of a country’s population. The coefficients and their statistical significances are shown in Table 2. The results suggest that the proportion of population covered by 4G mobile network, which reflects a country’s level infrastructure development, is significantly and positively correlated with a country’s gender parity index in ARSP skills. In addition, gender parity index in ARSP skills is positively and significantly correlated with a country’s overall ICT literacy, which is reflected by the percentage of population with ARSP skills. The significant correlation results between these variables provide spaces for further regression analysis with OLS. Also, since the percentage of 4G mobile network coverage and the proportion of population with ARSP skills are significantly correlated with each other, essential multi-collinearity checks will be performed in OLS regressions.

The results of OLS regressions (with different independent variables) are available in Table 3. Model 1 suggests that if no other factors considered, the coverage of 4G mobile network in a country has positive while modest impact on the country’s gender disparity in ARSP skills, and the impact is statistically significant (β = 0.004, p = 0.052). That means if a country has better infrastructure as reflected by its 4G mobile network coverage, it is likely to have higher female/male ratio in possessing ARSP skills. When the country’s national income level is included into consideration, such modest while statistically significant positive impact still remains (β = 0.004, p = 0.088). However, even not significant in term of statistics, it is still a bit surprising to notice that in Model 2, a country’s national income level is negatively associated with the gender parity index in ARSP skills (β = −0.020, p = 0.813).

When adding the proportion of population with ARSP skills (SDG Indicator 4.4.1) as a control variable, it is noticed that although the coverage of 4G mobile network still has modest positive impact on the gender parity index in ARSP skills, such impact becomes insignificant in statistics (β = 0.003, p = 0.213). Even when the country’s national income level is included as a control variable (Model 4), such impact still exists although statistically significant (β = 0.004, p = 0.134). The percentage of population with ARSP skills in a country is positively but insignificantly linked with the country’s gender parity index in ARSP skills. That means, in countries with higher percentage of population with ARSP skills, the female/male ratio of possessing ARSP skills is higher.

If interactive factors are included into consideration (Model 5), then some changes are observed. Firstly, the proportion of population with ARSP skills becomes a significant factor which has positive impact on the gender parity index in ARSP skills. National income level’s impact on gender parity index in ARSP skills turn into positive, although still not statistically significant (β = 1.769, p = 0.424). The interaction effects between a country’s national income level and the proportion of population with ARSP skills are negative and statistically significant (β = −0.016, p = 0.008). That suggests in high income countries, the impact of proportion of population with ARSP skills on gender parity index in ARSP skills is weaker than such impact in middle or low income countries. It is also noticed that R square values increased substantially (from lower than 0.2 to around 0.42) when interactive factors are added.

The insignificant and fluctuating impact of country’s national income level on gender parity index in ARSP skills calls for further cross-country comparative analysis. The explorative statistics divided by national income levels (grouped into high income and middle/low income countries) are presented in Table 4, including the result of Mann-Whitney U Test. It is noticed that there is no statistically significant difference between gender parity index in ARSP skills between high income countries (mean = 0.82, standard deviation = 0.11) and middle/low income countries (mean = 0.82, standard deviation = 0.24). This finding also corroborates the regression results in Table 3, where we can notice that country class (by income level) does not have statistically significant influence on gender parity index in ARSP skills.

Table 4 reveals a significant difference in the coverage of 4G mobile networks between middle/low income countries (mean = 79.01, standard deviation = 19.12) and high income countries (mean = 98.11, standard deviation = 2.06). The difference in the proportion of population with ARSP skills is also significant between high income countries and middle/low income countries. For example, on average near 43 % of population in high income countries have ARSP skills (standard deviation = 13.08), but this figure for middle/low income countries is only 17.53 (standard deviation = 10.04).

5 Discussion

The statistical analysis in the previous section suggests that a country’s level of infrastructure development, as reflected by its coverage of 4G mobile network, is associated with gender disparity in ICT literacy in the country, which is shown by the gender parity index in ARSP skills. A core reasons for this result is that infrastructure development especially mobile technology can support the integration of ICT into formal science education (Dibaba and Ramesh 2017; Paul and Mondal 2012), and formal science education at school is relatively gender-neutral (Miliszewska et al. 2006). Therefore, infrastructure development has a positive impact on reducing gender inequality in ICT literacy. Another reason for this result is that with infrastructure development especially the expansion of 4G mobile network, people (regardless of gender) will have more opportunities to access ICT and acquire some ICT literacy. In addition, the expansion of mobile network itself is a new channel which brings conveniences to people especially the more disadvantaged groups (such as females and ethnic minorities) to access and acquire ICT literacy, especially via more convenience access to essential STEM education. For example, it is able to simply use a mobile device at a convenient place to develop ICT literacy, so that the disadvantaged group such as females may not need to go to designated places for such skills (e.g. training centers which are sometimes male-dominant), which can reduce the gender disparity in ICT literacy.

The statistical results also demonstrate that the general level of ICT literacy, which reflects a country’s overall standard of science education, have positive impact on the female/male ratio with ICT literacy including ARSP skills. Since infrastructure development is positively associated with a country’s general level of ICT literacy as shown in Table 2, it may become an intermediate reason to explain the influence of infrastructure development on gender gaps in ICT literacy.

This finding corroborates some existing literature which argues that weak infrastructure is a factor causing gender disparity in ICT literacy (e.g. Kiondo 2007). It also provides evidence that infrastructure development can help reduce gender inequality in ICT literacy, and therefore policies and interventions in support of infrastructure development can also generate wider benefits in other fields such as science education and women’s empowerment (e.g. Opoku and Kuranchie 2014). In particular, since the development of mobile technology can improve STEM education and reduce the challenge in enhancing learners’ (including female learners’) ICT literacy caused by infrastructural factors such as unstable electricity supply (Samarakoon, Christiansen, and Munro 2017), it is recommended to further expand the coverage of mobile networks in order to reduce gender gaps in ICT literacy and better integration of ICT into STEM education. As the relationship between ICT and STEM education has been demonstrated by existing studies (Changpetch and Seechaliao 2020), it can also be inferred that infrastructure development can contribute to STEM education.

Countries’ income levels’ influence on gender gaps in ICT literacy is not statistically significant. In other words, in high income countries, gender gap in ICT literacy is not significantly larger or smaller than in middle/low income countries. This is particularly the situation as Table 3 shows that adding country class as a control variable does not lead to substantial changes in the R square values. This finding corroborates existing literature which argues that gender disparities in computer sciences and education (which are directly related to ICT literacy) remain substantial in developed countries, even the overall gender gap in STEM education in developed countries is smaller than in developing countries (Breda and Napp 2019). Also, this result supports the argument that the cross-country disparity in gender gaps in ICT literacy is not only caused by differences in national incomes, but also due to several other factors such as culture, level of STEM education, social division, and availability of teachers (Aini, Rachmatullah, and Ha 2019; Breda, Jouini, and Napp 2018; Dibaba and Ramesh 2017; Febro, Catindig, and Caparida 2020). Future research may conduct more cross-country comparisons in gender disparity in ICT literacy and further explore the associating factors.

However, in different countries, infrastructure’s influence on gender disparity in ICT literacy may vary. Similarly, the association between general level of population’s ICT literacy and gender disparity in ICT literacy may vary across countries. As revealed by Model 5 in Table 3, in high income countries, both general level of population’s ICT skills and infrastructure development’s impact on gender disparity in ICT literacy are weaker in high income countries than in middle/low income countries. This is possibly because of high income countries have better level of STEM education and infrastructure development, and may also have more resources to increase the gender responsiveness of relevant policies (Khalifa and Scarparo 2021), including in the fields such as science education and infrastructure.

The significant association between different SDG indicators suggests that there are interactions across SDGs, especially around equitable science education, empowerment of women, and infrastructure development (Le Blanc 2015; Liu 2020). For example, this article finds that a country’s infrastructure development is linked with its population’s ICT literacy and the associated gender disparity. Therefore, SDGs 4, 5, and 9 may have interactions or even trade-offs. Such interactions across different SDGs call for more holistic approaches to design and implement policies including educational reforms. For example, STEM educational reforms and interventions should be more gender sensitive (Febro, Catindig, and Caparida 2020), and ICT should be properly integrated into education (Dibaba and Ramesh 2017; Özkan and Tekeli 2021; Richardson 2008). More interdisciplinary studies in science education and ICT literacy with the angles of infrastructure and women’s empowerment could enrich the academic knowledge and generate relevant practice-oriented implications. Future research may also join the emerging trend to explore the interactions across SDGs, especially around science education, gender equality, and infrastructure.

6 Conclusions and Implications

This research examines the influence of infrastructure on gender disparity in ICT literacy in 30 countries based on national level data from the SDG Global Database. Relevant SDG indicators were selected to measure infrastructure development, gender disparity in ICT literacy, and the general level of population’s ICT literacy in different countries. The results show that infrastructure development can contribute to reduce the gender disparity in ICT literacy in both high income and middle/low income countries. Also, if a country has a higher proportion of population with ICT literacy, the gender disparity in ICT literacy is relatively lower. The cross-country comparison demonstrates that gender disparity in ICT literacy does not vary substantially between high income and middle/low income countries. However, in high income countries, infrastructure’s influence on gender disparity in ICT literacy is relatively weaker than in middle/low income countries. Possible reasons for the statistical outcomes are discussed in the article. For example, this article believes that a core reason for infrastructure development’s influence on reducing gender gaps in ICT literacy is that infrastructure development can contribute to STEM education at school, which is an important source to give male and female students with more equal opportunities to acquire ICT literacy.

This article enriches the knowledge in the field of STEM education because it uses a more macro perspective and global data source to identify the importance of infrastructure in STEM education and its contribution to reduce the gender inequality in ICT literacy at national level. It expands the current research spectrum in this field, which mainly analyze relevant data at micro level such as schools or communities. Also, the article uses the SDG Global Database and SDG indicators to operationalize the relevant variables especially gender disparity in ICT literacy, which is an advancement to overcome the challenges of lack of reliable indicators and data in this field (e.g. Hafkin and Huyer 2007). In particular, the SDG Global Database has information on the national gender parity index in ARSP skills (SDG Indicator 4.5.1), which is suitable to measure gender disparity in ICT literacy because ARSP skills can properly reflect the multi-dimensional nature of ICT literacy, including its strong link with STEM education. For example, ARSP skills can reveal learners’/users’ abilities in both mathematics and computer sciences, as well as their scientific reasoning ability especially the capability in controlling of variables.

In addition, this article joins the emerging research field of science education for sustainable development. Both science education and ICT literacy including the associated gender disparities are important aspects of sustainable development, and are included into the SDG indicator framework. The connections between ICT literacy, infrastructure, and empowerment of females demonstrate the possible interactions across SDGs, and also calls for future studies to explore the relationship between STEM education and sustainable development, which is currently under-explored and does not have sufficient existing literature.

Another contribution of this article is that it generates practical implications in relevant fields, especially at macro level. As the empirical evidence in the article shows the substantial connections between infrastructure development, national incomes, ICT literacy and associated gender disparities, more holistic approaches may be adopted in designing and implementing policies including reforms on science education. For example, ICT should be better integrated into STEM education with sufficient gender-sensitivity. Gender-responsive policies on STEM education aim to improve the general level of population’s ICT literacy may also reduce the gender gap in ICT skills. Therefore, policies of STEM education should be made and delivered with wide involvement of different ministries rather than only relying on educational authorities.

This article is not immune from limitations, which provides spaces for future endeavors in this research field. Firstly, the availability of data is still not perfect, although the SDG Global Database has significant strengths in comparison with other data sources. Both SDG Indicator 4.5.1 (gender parity index in relevant ICT literacy) and SDG Indicator 4.4.1 (the percentage of population with relevant ICT literacy) are ‘Tier 2 indicators’ as classified by the UN Statistics Division (2022). That means, data for these indicators are not regularly available although the concepts, methods, and standards for the indicators are clear. This is a reason that the number of countries included for analysis in this article is not large, which impairs the statistical power of analysis. Therefore, there are still spaces to improve the data availability for future research. For example, after the 55th session of the UN Statistical Commission in 2024, more data would be available in the SDG Global Database.

Secondly, this article uses relatively simple methods such as correlations and OLS regressions. Although simple methods have advantages such as easier to be understood by non experts, it is suggested that more sophisticated methods can also be used in future studies. For example, if more data become available in the future, more sophisticated research methods such as structural equation modeling can be used (Özkan and Tekeli 2021).

Thirdly, while this paper has shown that the gender parity index in ARSP skills is a suitable indicator for reflecting gender disparity in ICT literacy, it may be beneficial for future studies to explore additional indicators or develop compound indicators for a more comprehensive measurement of gender disparity in ICT literacy.

Last but not least, it is advised for future studies to conduct some qualitative analysis, in addition to the quantitative evidence provided in this article. The qualitative analysis would be helpful to better interpret statistical results, and may also tell interesting ‘stories behind numbers’. For example, interviews with officials at national authorities of education, infrastructure, and/or females’ rights may generate insights to explain why infrastructure in some countries have stronger influence on STEM education and gender disparity in ICT literacy than in other countries. In short, while quantitative evidence such as data from SDG Global Database are very useful, this article realizes that qualitative data could also generate valuable insights via future research in the relevant fields.