library(tidyverse)
library(plotly)
library(readr)
survey_ai <- read_csv(
"E:/Cover Letter/RMIT/Classes/Sem 3/Data Vis/Assignment 3/Survey_AI.csv",
show_col_types = FALSE
)
theme_conversation <- function() {
theme_minimal(base_size = 14) +
theme(
plot.title = element_text(size = 16, face = "bold"),
plot.subtitle = element_text(size = 11, colour = "#383838"),
axis.title = element_text(face = "bold"),
panel.grid.minor = element_blank(),
legend.position = "bottom"
)
}The incorporation of artificial intelligence tools is revolutionizing today’s educational sector. Students utilize AI tools for completing their assignments, gaining information, generating ideas, and increasing efficiency. Nevertheless, even though there is a fast pace of adaptation, the universities and authorities lack knowledge of students’ reactions towards artificial intelligence.
This paper examines the perceptions students have regarding AI, what resources they use to learn about AI, their level of usefulness when using AI academically, and whether more AI exposure decreases their fear towards the technology. Interactive multivariate graphs designed in R are used in this paper to provide data-based stories, just like on The Conversation site.
The rate at which AI systems are being integrated into student learning is happening quicker than universities can formally integrate them. The usage of AI in learning is more common among students, but universities do not fully comprehend how students’ academic and emotional experiences with these systems are. This knowledge is important because AI systems are increasingly becoming part of learning experiences at all educational levels.
Many students already feel moderately familiar to highly familiar with artificial intelligence technology, thus proving that these technologies are already embedded into student’s lives.
p1 <- survey_ai %>%
count(Q1.AI_knowledge) %>%
ggplot(aes(
x = as.factor(Q1.AI_knowledge),
y = n,
text = paste(
"AI knowledge score:", Q1.AI_knowledge,
"<br>Students:", n
)
)) +
geom_col(fill = "#D8352A") +
labs(
title = "Most students already understand AI",
subtitle = "Self-rated AI knowledge is concentrated around mid-to-high scores.",
x = "AI knowledge score",
y = "Number of students"
) +
theme_conversation()
ggplotly(p1, tooltip = "text")Most students rated their AI familiarity between 5 and 8, indicating that AI technologies are already well integrated into student experiences.
In order to gain knowledge about Artificial Intelligence technologies, students depend on online resources much more than they use institutional education.
source_data <- survey_ai %>%
summarise(
Internet = sum(`Q2#1.Internet`, na.rm = TRUE),
Books_Papers = sum(`Q2#2.Books/Papers`, na.rm = TRUE),
Social_Media = sum(`Q2#3.Social_media`, na.rm = TRUE),
Discussions = sum(`Q2#4.Discussions`, na.rm = TRUE),
Not_Informed = sum(`Q2#5.NotInformed`, na.rm = TRUE)
) %>%
pivot_longer(
cols = everything(),
names_to = "source",
values_to = "count"
) %>%
mutate(
source = recode(
source,
Books_Papers = "Books / papers",
Social_Media = "Social media",
Not_Informed = "Not informed"
)
)
p2 <- source_data %>%
ggplot(aes(
x = reorder(source, count),
y = count,
text = paste(
"Source:", source,
"<br>Students:", count
)
)) +
geom_col(fill = "#29339B") +
coord_flip() +
labs(
title = "Students mostly learn about AI outside formal education",
subtitle = "Internet and social media dominate as AI information sources.",
x = "Information source",
y = "Number of students"
) +
theme_conversation()
ggplotly(p2, tooltip = "text")The findings imply that students acquire knowledge on AI mainly on their own within online ecosystems but not via classroom instruction in universities.
Students who perceive AI tools as academically useful often report stronger GPA outcomes.
p3 <- survey_ai %>%
ggplot(aes(
x = Q7.Utility_grade,
y = Q16.GPA,
text = paste(
"AI usefulness score:", Q7.Utility_grade,
"<br>GPA:", Q16.GPA
)
)) +
geom_jitter(
width = 0.15,
height = 0.05,
alpha = 0.65,
color = "#505AAF",
size = 3
) +
geom_smooth(method = "lm", se = FALSE, colour = "#383838") +
labs(
title = "AI usefulness is linked with higher GPA",
subtitle = "Each point represents one student response.",
x = "AI usefulness score",
y = "GPA"
) +
theme_conversation()
ggplotly(p3, tooltip = "text")## `geom_smooth()` using formula = 'y ~ x'The students who viewed AI as helpful in their learning had good academic results, indicating that AI is becoming a productive learning tool.
Students from different years of study and genders report varying levels of AI usefulness.
p4 <- survey_ai %>%
mutate(
Year = factor(Q13.Year_of_study),
Gender = factor(Q12.Gender,
labels = c("Male", "Female"))
) %>%
ggplot(aes(
x = Year,
y = Q7.Utility_grade,
fill = Gender
)) +
geom_boxplot(alpha = 0.8) +
labs(
title = "AI usefulness differs across student groups",
subtitle = "Usefulness scores vary by year of study and gender.",
x = "Year of study",
y = "AI usefulness score",
fill = "Gender"
) +
theme_conversation()
ggplotly(p4)Different student groups do not experience AI in the same way, suggesting that educational background, experience, and learning context may shape how AI technologies are perceived.
Students with higher AI knowledge do not necessarily feel more positively about artificial intelligence technologies.
p5 <- survey_ai %>%
ggplot(aes(
x = Q1.AI_knowledge,
y = Q5.Feelings,
colour = Q16.GPA,
text = paste(
"AI knowledge score:", Q1.AI_knowledge,
"<br>Feeling score:", Q5.Feelings,
"<br>GPA:", Q16.GPA
)
)) +
geom_jitter(
width = 0.2,
height = 0.15,
alpha = 0.75,
size = 3
) +
labs(
title = "AI knowledge does not always reduce concern",
subtitle = "This chart compares AI knowledge, feelings toward AI, and GPA.",
x = "AI knowledge score",
y = "Feeling score",
colour = "GPA"
) +
theme_conversation()
ggplotly(p5, tooltip = "text")Greater AI familiarity does not necessarily eliminate uncertainty or concern, highlighting the complex emotional relationship students have with rapidly emerging technologies.
Results indicate that AI has already penetrated into student learning ecologies. Students tend to acquire skills related to AI informally online, as opposed to institutional education, whereas students believing that AI tools can be helpful show better grades.
Nevertheless, more knowledge on the part of students does not mean less anxiety and worries regarding technology. This is indicative of the increasing gap between fast-paced implementation of technology and inadequate emotional adjustments in universities. With AI technologies getting integrated more widely into the learning environment, it might be necessary for educational institutions to consider changing their approaches to training students in AI.
Generative AI tools including ChatGPT were used to assist with debugging R code.
[1] Mendeley Data. (2025). Global Student ChatGPT Survey Dataset. Available at: https://data.mendeley.com/
[2] UNESCO. (2025). AI guidance in higher education institutions. Available at: https://www.unesco.org/
[3] Kaggle. (2025). Survey on Students’ Perceptions of AI in Education. Available at: https://www.kaggle.com/
[4] Wickham, H. et al. (2025). ggplot2 and tidyverse documentation. Available at: https://www.tidyverse.org/
[5] Sievert, C. (2025). Interactive Web-Based Data Visualization with Plotly and R. Available at: https://plotly.com/r/