Author Summary
This survey was conducted to explore how visual methods impact public awareness of child labor in cobalt mines in the Democratic Republic of the Congo. Participants compared photos, graphics, charts, and their combinations to assess which format evoked stronger emotional responses and improved awareness. The data revealed that emotionally compelling visuals were more effective than text-based or graphic representations, and were most effective when presented with relevant text. While respondents supported awareness-raising campaigns, they showed less support for policy-related actions. These results demonstrate the potential of strategically designed visual campaigns to influence public perception for discussions on responsible sourcing and conflict minerals policies.
1. Introduction
1-1. Background
The current electric vehicle (EV) market aims to protect the environment and lead cutting-edge technological innovation. According to the International Energy Agency (IEA), “We currently expect to see 14 million in sales by the end of 2023, representing a 35% year-on-year increase” (International Energy Agency, 2023). As demand for EVs grows, the industry prioritizes battery technology innovation, such as Lithium-based technology, over traditional internal combustion engines (ICEs) (Koech et al., 2024).
Most lithium batteries in EVs contain cobalt, offering high performance and cost advantages (Koech et al., 2024). Cobalt is used in EV batteries and widely in 3C electronics, such as computers and mobile devices, due to its high energy density (Wu et al., 2022). According to the IEA(2023), rising EV battery demands have driven a 15% increase in cobalt use, reaching 150kt in 2023.
Korea plays a significant role in the global battery and electronics sectors; Korean companies such as LG Chem and Samsung SDI accounted for more than one-third of global lithium-ion battery production in 2020 (Jeong & Mah, 2022). Despite this industrial prominence, Korean professionals’ perspectives on the social and ethical implications of mineral sourcing are still underrepresented in international discourse.
1-2. Problems
However, behind this growing demand, using cobalt presents several challenges. Although efforts are underway to reduce reliance on cobalt, it remains essential in battery production due to its critical role in enhancing energy density and chemical stability (Ambrose & O’Dea, 2021).
Around 70% of the global supply of mined cobalt in 2019 originated from the Democratic Republic of Congo (DRC) (World Bank, 2022), but the benefits to the country remain limited. The DRC has not seen corresponding economic development despite its vast natural resources. China has played a central role in Central Africa’s cobalt industry, and the DRC also provided China, a growing force on the international stage, with direct access to its massive copper and cobalt resources (Byamungu, 2022, pp. 2, 4). However, the wealth generated has largely flowed out of the region. Consequently, the DRC is one of the poorest in the world, with an estimated 73.5% of Congolese people living on an income of less than $2.15 a day (World Bank, 2024).
Beyond economic concerns in the DRC, child labor is a more serious and often overlooked issue. According to the U.S. Department of Labor (2023), more than 25,000 children work in cobalt mines across the DRC (U.S. Department of Labor), which raises serious ethical concerns. Furthermore, artisanal and small-scale mining accounts for approximately 15-30% of the DRC’s cobalt production, where the risk of human rights, including child labor, is much higher (World Economic Forum, 2020, p. 4). Research conducted in Peru shows that children exposed to heavy metals are at significantly higher risk for health problems such as nosebleeds, chronic colic, dermatologic alterations, mood alterations, and other symptoms related to neurological harm (Fandiño Piñeiro et al., 2021). These realities highlight the urgent need to address both the labor conditions and the health problems for child workers in DRC cobalt mines.
International organizations such as the OECD, UN, and EU aim to stabilize the global mineral trade through measures like the Dodd-Frank Act (Section 1502), which includes provisions for mine safety that require companies to disclose whether they use conflict minerals and where they come from (Woody, 2019). Similarly, in 2021, the EU’s Conflict Minerals Regulation aimed to disconnect the mineral trade (3TG) from conflict-related activities and human rights abuses (Macchi, 2021). Cobalt has been associated with conflict financing in addition to 3TG, but it is not included in the regulation (IPIS, 2021). Consequently, child miners in the DRC remain in hazardous conditions, and the lack of regulation demonstrates the need for public awareness about ethical cobalt mining.
Therefore, this study investigates how different visual methods affect public awareness of child labor in cobalt mines, explicitly focusing on workers in Korea’s IT and electronics fields, as visual methods can offer a more direct understanding of the public perceptions than data collected solely by the researcher (Rich et al., 2000).
Furthermore, this study examines the issue from multiple perspectives to better capture the complexity of the social impact of visual methods by exploring, to a limited extent, whether emotionally driven shifts in awareness may influence policy-related attitudes.
1-3. Research Question and Hypothesis
The research question is, "What types of visual methods are most effective in increasing the awareness of child labor in cobalt mining among Korean IT and electronic industry professionals to advocate for a conflict mineral?
Depending on the type of visual method, there will be significant differences in the level of change in awareness of child labor among Korean IT and electronics industry workers. It is hypothesized that visual methods combining direct text and photos will be more effective than other visual methods in raising awareness and promoting support for designating cobalt as a conflict mineral.
2. Literature Review
2-1. Public Awareness Gap in Cobalt Mining
Current research on cobalt mining has focused heavily on supply chains, corporate responsibility, and economic impacts. However, the role of public awareness, particularly its emotional and social implications and how it may directly impact children, remains underexplored in the academic literature. The U.S. Department of Labor’s COTECCO (Combating Child Labor in the Democratic Republic of the Congo’s Cobalt Industry) project was developed to address child labor in mining areas, but it states that “the beneficiaries are primarily the stakeholders whose activities relate to the prevention, protection, and relief of children in mining sites” (U.S. Department of Labor, 2022, p. 11). These beneficiaries include government agencies, mining companies, and civil society organizations. (U.S. Department of Labor, 2022, p. 11). This suggests that while the project focuses on institutional stakeholders, it is barely delivering tangible improvements to child miners. Such limitations can undermine broader awareness efforts beyond mining regions.
As a result, awareness-raising efforts targeting the general public outside mining regions are still limited, and global perception of the ethical issues associated with cobalt production remains low. Existing studies on visual methods and public awareness suggest that emotional and cognitive engagement through visual methods can effectively close this gap.
2-2. Impact of Visual Methods on Awareness
Although research has explored the effectiveness of visual campaigns, few studies have focused on how the general public perceives such methods, highlighting the need for this study.
Pfau et al. (2006) investigated this relationship by testing how different print methods influenced what people thought. According to the study, “photographs do indeed have a tremendous impact” and serve as “a powerful tool in shaping public opinion” (Pfau et al., 2006).
In this research, respondents were given different types of content, such as images related to the Iraq War, descriptive captions, text without pictures, and a combination of images and text. Results showed that the pictures alone presented were more effective than text-only or image-with-text conditions (Pfau et al., 2006).
The Stanford University paper, which is “Then and Now: Evolving Representations of Children in UNICEF Photographs,” is another notable study that analyzes examples of using photographs to protect children’s rights. This paper examines the way UNICEF raises awareness of children’s rights internationally and emphasizes them when these rights are threatened in conflict situations. Photographs are powerful tools for education, awareness, fundraising, and national and international politics (Vasavada, 2016). They also support by allowing viewers to observe people and places without being physically present (Vasavada, 2016). Debrix (1996, as cited in Vasavada, 2016 ) argues that photographs are not only an emotional tool but also can inform the public about child labor realities (qtd. in Vasavada). Furthermore, the way international organizations like UNICEF portray children continues to influence global politics and cultural norms (Vasavada, 2016).
However, these portrayals tend to depict general hardship rather than specific cases, such as child labor in extractive industries. While these studies highlight the global effectiveness of visual methods, research on how Korean audiences perceive these issues remains limited. To address this gap, this study investigated how Korean IT and electronics industry workers interpret and respond to these issues. As Young at Northwestern University noted, “What is not measured remains invisible,” highlighting the need to initiate this research.
3. Methods
3-1. Participations
To explore their perceptions and attitudes toward child labor in the cobalt supply chain, this study surveyed Korean IT and electronics professionals. While Korean EVs and batteries continue their global expansion, it has been reported that three major firms accounted for 42.5% of battery production in the EU, with their products making up 63.5% of the EU battery market (Jo, 2024). In addition, major precursor and cathode producers in South Korea and Japan reported a 16% y/y increase in demand (Cobalt Institute, 2022). This industrial background supports the relevance of this study. A total of 76 respondents participated, of which 70 fully completed the survey and met the inclusion criteria. In other words, six responses were excluded because the participants were not employed in relevant job categories. All eligible participants were current employees in the IT or electronics industries, with no restrictions on age or gender, and they were informed that their responses would remain anonymous and were collected with their consent. Also, to ensure privacy, no participants’ information was collected, including names, email addresses, and other private details.
3-2. Data Collection
The survey was conducted in Korean via Google Forms between February 10 and 16 and distributed through email. Participants were first screened for job relevance; only those in the relevant category proceeded to the full survey.
The survey had two main parts.
First Section (a) examined emotional and cognitive reactions to images about child labor in cobalt mines. It was separated into (a-1), which focused on emotional responses to four types of images (e.g., child workers, mining workplaces, unrelated peaceful images), and (a-2), which compared pairs of visual formats (photos, graphic images, images + text, and text + charts) to evaluate which method more effectively raised awareness. From 5-10, it focuses on comparing the visuals to find the most effective visual for emotional reactions. In both subparts, participants responded using a 5-point Likert scale or selected the more effective image out of two. A Likert scale was used since Likert-type responses are psychometric item-scoring schemes for quantifying people’s opinions on the issues (Bishop & Herron, 2015).
Second Section (b) is an exploratory component of this study that seeks to examine, in a limited way, how emotional responses may be linked to changes in policy attitudes. It also explores whether the emotional impact of visual methods can lead to actual social engagement or policy support rather than merely shifting perception.
Section (b) focused on participants’ awareness of conflict minerals, their view on whether cobalt ought to be labeled a conflict mineral, and their opinions on the most effective ways to advocate for international action. This Section used three types of questions to gather a wide range of perspectives. Likert-scale items measured participants’ level of agreement or awareness.
3-3. Data Analysis
This study used the Friedman Test, the Chi-square Goodness-of-Fit Test, and Frequency Analysis to evaluate the statistical reliability of the survey responses. The main statistical analysis was conducted using Google Sheets, with reference to web-based resources such as Laerd Statistics. Visual materials, such as graphs and charts, were created using the Canva program.
4. Results and Discussion
4-1. Emotional Impact of Visual Materials (Section a-1)
The analysis of section (a-1) focused on evaluating two aspects. First, the study analyzed how respondents showed different emotional responses to the four images to evaluate the influence of the visual content of the pictures on emotional reactions. Second, the study attempted to evaluate whether greater emotional responses were associated with greater changes in awareness of child labor.
To this end, in the first analysis, respondents were presented a Group photo of miners (no specific focus on children) (1-1), A photo directly depicting child labor (2-1), Peaceful image of African children (not related to labor) (3-1), and Labor scene without faces (focused on hands only) (4-1), and analyzed statistical differences in emotional levels for each image using the Friedman test. In the second analysis, the study evaluated the relationship between emotional intensity and the level of cognitive change using Spearman correlation analysis.
In the first analysis, the Friedman test, which is suitable for repeated measures, was used to examine whether there were statistically significant differences in the emotional reactions of 70 respondents after viewing four different types of images. The Friedman test is a nonparametric method for comparing three or more repeated measures in a within-subjects design (Laerd Statistics, n.d.-a). It is a nonparametric statistical method suitable for cases where respondents rank the scores, and it is ideal for survey data using a Likert scale (1 to 5 points) because it does not assume normality.
The result showed a highly significant difference at the level of p < .001 through the Friedman test, χ² = 162.58, p = 5.09 × 10⁻³⁵. After viewing the four pictures, each participant’s emotions (discomfort levels) were statistically significantly different, showing that the emotional changes differed depending on the images.
The following analysis analyzed the correlation between the magnitude of the emotions respondents felt after viewing each image and the resulting change in perception. Spearman correlation analysis is an important tool used in the paper to evaluate the relationship between two variables. This analysis can be applied under the following three assumptions: First, the two variables must be ordinal, such as a Likert scale; second, the two variables must consist of paired observations; third, there must be a monotonic relationship between the two variables, where the values increase together or when one increases, the other decreases (Laerd Statistics, n.d.-b).
Based on the valid responses of 70 participants (70 out of 76), the relationship between emotional responses (1-1, 2-1, 3-1, 4-1) and changes in awareness (1-2, 2-2, 3-2, 4-2) was analyzed. As shown in Table 1, each pair showed a statistically significant positive correlation. These results support that the stronger the emotional discomfort, the more significant the change in perception of child labor.
This pattern was especially evident in Image 2, which generated the strongest reactions. After viewing Image 2-1, 90% of respondents reported feeling either “very” or “somewhat uncomfortable.” For 2-2, 85.7% said the image had a “somewhat” or “very strong” impact on raising their awareness of the issue. This result supports the hypothesis that emotionally powerful visual materials can effectively shift public awareness.
4-2. Comparing Images, Charts, and Text in Raising Awareness (Section a-2)
Survey questions 5 through 10 presented participants with pairs of visual methods, including photographs, graphics, charts, and text combinations, delivering similar messages about child labor in the DRC. Each question asked respondents to evaluate which of the two formats elicited a stronger emotional response (questions 5 to 7) or more effectively increased their awareness of the issue (questions 8 to 10).
The results indicated that visual images elicited a stronger emotional and cognitive response than textual or chart-based formats. In particular, real photographs were perceived as more effective than graphic representations.
To analyze section (a-2), this study divided the responses into two parts: Questions 5 to 7, which focused on emotional impact, and Questions 8 to 10, which examined informational effectiveness. For each group, the Chi-square Goodness-of-Fit Test was applied to determine whether the distribution of responses across the three categories (Image A, Image B, No Difference) deviated significantly from a uniform distribution.
The chi-square goodness-of-fit test is widely used to evaluate whether observed frequencies differ significantly from expected values. As Bolboacă et al. (2011) explain, it is “the most well-known statistic used to test the agreement between observed and theoretical distributions, independence, and homogeneity” (Bolboacă et al., 2011).
Questions 5 to 7 assessed which type of visual method triggered stronger emotional responses. Question 5 compared A: real photographs with B: graphic images, Question 6 compared A: text + image with B: text + chart, and Question 7 presented both images with a background photo of child labor, comparing A: direct text and B: indirect text.
In question 5, 59 out of 70 participants selected the real photo (A), while only 7 chose the graphic image (B), χ²=84.94, p < .001. Similar patterns were observed in Questions 6 and 7. The chi-square test results revealed that respondents had statistically significant preferences in all three questions, consistently favoring Image A (real photographs, text + image, or direct messages) over Image B (graphic images, text + charts, or indirect messages).
This result supports the hypothesis of this study. In particular, there was a clear tendency for visual methods that induce strong emotional responses, such as real photos or direct texts, to be more effective than graphic images or indirect expressions. This suggests that images or direct texts that cause immediate emotional reactions play an important role in raising awareness of child labor. However, it is necessary to examine whether methods that combine text and images are the most effective in changing awareness through the following analysis.
Next, Questions 8 through 10 compared the differences between visual methods types A and B in how they might effectively communicate information about child labor issues to the public. Each question presented participants with two types of visual methods and asked which type would more effectively communicate information about child labor issues in the DRC. The response categories were the same as the previous questions, consisting of three categories: ‘Image A,’ ‘Image B,’ and ‘No Difference,’ and the Chi-square Test was applied for analysis.
The combination of visual methods used for each question is as follows. In Question 8, the same background photo was used, but type A included a concise phrase linked to the battery drain image, while type B presented only the original photo without any text. In Question 9, type A used a graphic image that gave a strong visual impression, while type B presented a chart and a short explanatory text. Finally, Question 10 presented data in type A, which added simple text to actual photos, and type B, which combined a chart and text. The analysis results for the three questions are summarized in Figure 1.
The three questions compared which type of visual methods the participants felt were more effective in explaining the child labor issue, focusing on the differences in the information delivery power of the visual methods. The result of the chi-square test analysis showed statistically significant differences in all three questions (Q8, Q9: p < .05, Q10: p < .001), which shows that the participants did not simply choose randomly but perceived certain types of visual methods as more effective.
Question 8, 36 respondents chose Type A (photo with simple text and overlay), 21 chose Type B (the same image without text), and eight responded with “no difference.” The result, p = .003, shows statistically significant differences that suggest that many participants felt that visual methods with added text delivered information better. However, the fact that the difference was less strong than in other questions suggests that text or diagrams added to images may not necessarily have a significant effect.
Question 9 reveals a preference for chart-type data (B: 44.9%) that conveys information quickly over graphic images (A: 33.8%) that simply express the reality of child labor. This suggests that graphics may not offer the same level of communication or immersion as actual photos.
According to Question 10, 64.3% of participants selected Type A (real photo with short text) as the most apparent preference, while 34.8% selected Type B (chart with text), and 11.4% chose “no difference.” The result showed that the highly significant (p < .001), suggesting that real photographs with short text were perceived as more direct and emotionally engaging than abstract data visualizations.
Overall, the results from this section support that a combination of images and text is more effective than charts or text alone in increasing public awareness of child labor. This supports the idea that combining visual realism (e.g., real photographs) with brief, meaningful phrases can be a powerful strategy in advocacy communication.
4-3. Emotional Awareness vs. Policy Support Gap (Section b)
The final part of the survey analysis, Section b (Questions 11 to 13), focused on analyzing participants’ thoughts on effectively raising awareness of child labor in cobalt mines and participants’ understanding and support for designating cobalt as a ‘conflict mineral.’ The data showed that respondents had strong emotional responses to the visual images and were interested in raising awareness. However, they responded passively regarding practical policy support, such as designating cobalt as a conflict mineral. It shows a potential gap between emotional awareness and concrete advocacy.
However, after a brief explanation in the survey, only 62.9% supported designating cobalt as a conflict mineral (Q12-1), while 32% selected ‘no opinion’ and 7.1% opposed it. In Question 12-2, respondents who supported designating cobalt as a conflict mineral were asked to choose the most effective advocacy methods. The most selected approaches were using SNS and Social Media to raise awareness (56.8%) and official requests to international organizations like the UN or EU (54.5%), showing that community-driven initiatives and institutional pressure are important. On the other hand, the method of directly pressuring companies (6.8%), requesting the government to change the law, was chosen by 15.9% of respondents. This may be because people feel that such methods are hard to achieve and unlikely to bring real change. In particular, legal reforms can be slow and complicated, and there may be a lack of trust that requests will be handled transparently or effectively.
Question 11 asked how much respondents knew about ‘conflict minerals,’ and only 30.4% of respondents (n=69) answered ‘Very familiar’ or ‘Somewhat familiar.’ On the other hand, more than half of the respondents, 56.5%, answered ‘Heard of it, but sure what it is,’ and 13% answered ‘Not familiar at all,’ showing that the overall awareness of conflict minerals was not high.
Question 12-1 asked respondents whether they supported designating cobalt as a conflict mineral, along with a brief explanation. Of the 70 participants, 62,9% answered ‘yes,’ 7.1% ‘no,’ and 30% selected ‘no opinion.’ While earlier responses indicated strong emotional empathy (around 90%), participants expressed more caution when it came to policy support, which recorded the highest proportion of neutral responses in the survey. A chi-square test confirmed a statistically significant difference across responses (χ² = 32.94, p < .001), indicating the result was not due to chance.
The following question 12-2 asked respondents who supported the designation of cobalt as a conflict mineral how to most effectively inform the international community about this (select up to 2 out of 6 items, Figure 2). The most preferred methods among respondents were raising awareness through social media (56.8%) and appeals to international organizations (54.5%), highlighting the importance of community-based and institutional approaches. By contrast, boycotts (6.8%) and signature campaigns for legal reforms (15.9%) attracted relatively little support, as many perceived these methods to be less practical or impactful.
The Q13 asked respondents to choose the most effective method for publicizing child labor in the DRC and supporting cobalt’s designation as a conflict mineral.
The options included using emotional images of child labor, combining those images with strong messages, or presenting statistical data through charts and graphs.
The response rate was highest for ‘Photo and Text’ at 73.8%, followed by ‘Photo only’ at 23.2%, and ‘Chart or Graph’ at 13%. The Chi2 value was 29.83, showing a statistically significant difference (p < .001). The apparent gap in Q 13preference supports the study’s hypothesis that visual materials with strong messages are most effective in changing perception. It also suggests that combining direct text with imagery enhances the way people understand and respond to a problem.
4-4. Integrated Analysis and Interpretation
This survey showed that there is a more complex layer than a simple linear relationship between emotional responses, cognitive change, and policy attitude. In Section (a-1), images that directly showed child labor triggered the strongest emotional response, and this discomfort significantly correlated with the level of cognitive changes. In Section (a-2), the format that combined real photos and direct text was the most effective in information delivery and awareness-raising effects. Meanwhile, the auxiliary analysis, Section (b), showed that emotional empathy does not necessarily lead to policy support. In other words, it is not enough to make people feel something; visuals also need to help them understand and take action.
5. Limitations and Implications
While this research provides meaningful insights about child labor, there are some limitations. Since many people may have participated in the survey on their mobile phones, it may have been difficult to see the image details, such as screen size, phone brightness, and color contrast, which may have influenced the responses. Another limitation is that many questions in the survey asked about emotional reactions, which can vary greatly depending on each respondent’s personal experience, values, and background.
Despite these limitations, the results of this study provide stakeholders, such as policymakers, with a more effective way to use appropriate visual methods to support policies and raise public awareness to improve the lives of children suffering from child labor.
In addition, this study was intentionally designed to target workers in the IT and electronics industries in Korea rather than the general public. Although generalization is limited, the focused nature of this research sample allows for a more targeted analysis of industries closely associated with cobalt mining and child labor.
6. Conclusion
This study investigated how different types of visual methods influence public awareness of child labor in cobalt mining. The findings clearly demonstrate that visual methods combining direct text and photos would be more effective than other visual methods. Real photographs and direct textual messages are the most effective in eliciting emotional discomfort and driving perceptual change. In particular, respondents reacted more strongly to direct images of child labor, and this emotional response was positively correlated with shifts in awareness.
The survey showed that while almost 90% of participants felt emotional discomfort after seeing images of child labor, only about 63% supported officially designating cobalt as a conflict mineral. This highlights that emotions alone are insufficient to drive action, pointing to the importance of follow-up campaigns and clearer information.
In addition, the research results differ from those of Pfau et al., who concluded that using images alone was the most effective; the study found that combining strong, direct text with visual images significantly improved awareness. Awareness changes depend not only on the visual stimulus but also on the composition of the visual methods or the characteristics of the text. It means that further research is needed to elaborate on how factors such as how text and images are combined, the specificity of the message, and the respondents’ cultural background affect the change in awareness. Also, future studies can focus on finding a method to create synergy alongside the visual impacts.
Lastly, while the supplementary data showed that emotional responses did not always support political tendency, the core findings of this study point to one clear conclusion: Visual methods that combine real photos and direct text are the most effective in raising awareness of child labor in cobalt mining. This suggests that when designing a visual campaign, it is important to engage emotions and deliver clear, honest messages that help people understand the issue more deeply.
Conflict of Interest
The author declares no conflict of interest.