Data Visualization Step Two

Research Question 1

What is the relationship between teacher’s perspectives on the school’s attitude toward using ICT technology in the classroom?
(This question aims to explore institutional attitude toward information and communication technology (ICT) influence individual teacher behavior in adopting technology in the classroom).

Variable selection:
it2g06a: ICT use at school when teaching; it2g14a: Your School/Use of ICT in teaching at your school/ICT is considered a priority for use in teaching; cntry: Country(Denmark, Finland, and Uruguay); partt: Survey participation; idteach: Teacher ID;

Visualize with tables

Show code

datasummary_skim(rq1_data, 
                 type = "categorical")

		N	%
cntry	DNK	854	20.7
	FIN	2451	59.4
	URY	818	19.8
it2g06a	Never	76	1.8
	Less than once a month	194	4.7
	At least once a month but not every week	444	10.8
	At least once a week but not every day	946	22.9
	Every day	2463	59.7
it2g14a	Strongly Agree	1787	43.3
	Agree	1948	47.2
	Disagree	360	8.7
	Strongly Disagree	28	0.7

The summary table looks good so far, but we can make it better by integrating table-making packages. At here, I added output = “gt” to apply the gt() function to the rest of the table. However, the summary table looks still robust.

Show code

rq1_summary <- rq1_data %>% 
  datasummary_skim(type = "categorical", output = "gt") %>% 
  tab_header(
    title = "Summary of Teachers' ICT Perspectives (2018 & 2020)"
  ) %>% 
  cols_width(everything() ~ px(130)) %>% 
  opt_table_outline()

rq1_summary

Summary of Teachers' ICT Perspectives (2018 & 2020)
		N	%
cntry	DNK	854	20.7
	FIN	2451	59.4
	URY	818	19.8
it2g06a	Never	76	1.8
	Less than once a month	194	4.7
	At least once a month but not every week	444	10.8
	At least once a week but not every day	946	22.9
	Every day	2463	59.7
it2g14a	Strongly Agree	1787	43.3
	Agree	1948	47.2
	Disagree	360	8.7
	Strongly Disagree	28	0.7

Then, I decided to subgroup the dataset by country and year to show the counts and percentage for each category.

Show code

rq1_summary %>% 
  gt() %>% 
  tab_header(title = "Teachers' ICT perspectives by country and year (2018 & 2020)") %>% 
  cols_label(
    cntry = "Country",
    it2g06a = "ICT Use Frequency",
    it2g14a = "School ICT Priority",
    `2018_N` = "N (2018)",
    `2018_Percent` = "% (2018)",
    `2020_N` = "N (2020)",
    `2020_Percent` = "% (2020)"
  ) %>% 
  fmt_number(columns = c(`2018_Percent`, `2020_Percent`), decimals = 1) %>%
  tab_options(table.width = px(100),
              table.font.size = px(12)) %>% 
  tab_style(
    style = cell_text(whitespace = "nowrap"),
    locations = cells_body(columns = c(it2g06a,it2g14a))) %>% 
  cols_width(
    it2g06a ~ px(200),
    everything() ~ px(100)) %>% 
  tab_style(
    style = list(cell_fill(color = "#9EDBDB")),
    location = list(
      cells_body(columns = everything(), rows = it2g06a == "Every day"))
  ) %>% 
  tab_footnote(
    footnote = "Note: DNK refers to Denmark, FIN refers to Finland, and URY refers to Uruguay.")

Teachers' ICT perspectives by country and year (2018 & 2020)
Country	ICT Use Frequency	School ICT Priority	N (2018)	N (2020)	% (2018)	% (2020)
DNK	Less than once a month	Disagree	1	NA	0.0	NA
DNK	At least once a month but not every week	Strongly Agree	2	2	0.0	0.0
DNK	At least once a month but not every week	Agree	5	3	0.1	0.1
DNK	At least once a month but not every week	Disagree	1	NA	0.0	NA
DNK	At least once a week but not every day	Strongly Agree	51	21	1.2	0.5
DNK	At least once a week but not every day	Agree	46	22	1.1	0.5
DNK	At least once a week but not every day	Disagree	2	NA	0.0	NA
DNK	Every day	Strongly Agree	243	288	5.9	7.0
DNK	Every day	Agree	83	78	2.0	1.9
DNK	Every day	Disagree	5	1	0.1	0.0
FIN	Never	Agree	6	2	0.1	0.0
FIN	Never	Disagree	2	1	0.0	0.0
FIN	Less than once a month	Strongly Agree	12	10	0.3	0.2
FIN	Less than once a month	Agree	30	18	0.7	0.4
FIN	Less than once a month	Disagree	9	5	0.2	0.1
FIN	At least once a month but not every week	Strongly Agree	32	21	0.8	0.5
FIN	At least once a month but not every week	Agree	100	57	2.4	1.4
FIN	At least once a month but not every week	Disagree	19	8	0.5	0.2
FIN	At least once a month but not every week	Strongly Disagree	1	2	0.0	0.0
FIN	At least once a week but not every day	Strongly Agree	91	69	2.2	1.7
FIN	At least once a week but not every day	Agree	187	159	4.5	3.9
FIN	At least once a week but not every day	Disagree	37	15	0.9	0.4
FIN	At least once a week but not every day	Strongly Disagree	1	NA	0.0	NA
FIN	Every day	Strongly Agree	289	433	7.0	10.5
FIN	Every day	Agree	360	390	8.7	9.5
FIN	Every day	Disagree	45	31	1.1	0.8
FIN	Every day	Strongly Disagree	4	5	0.1	0.1
URY	Never	Strongly Agree	6	3	0.1	0.1
URY	Never	Agree	20	10	0.5	0.2
URY	Never	Disagree	16	5	0.4	0.1
URY	Never	Strongly Disagree	2	3	0.0	0.1
URY	Less than once a month	Strongly Agree	11	7	0.3	0.2
URY	Less than once a month	Agree	30	24	0.7	0.6
URY	Less than once a month	Disagree	20	15	0.5	0.4
URY	Less than once a month	Strongly Disagree	2	NA	0.0	NA
URY	At least once a month but not every week	Strongly Agree	16	15	0.4	0.4
URY	At least once a month but not every week	Agree	69	34	1.7	0.8
URY	At least once a month but not every week	Disagree	40	15	1.0	0.4
URY	At least once a month but not every week	Strongly Disagree	1	1	0.0	0.0
URY	At least once a week but not every day	Strongly Agree	29	36	0.7	0.9
URY	At least once a week but not every day	Agree	57	71	1.4	1.7
URY	At least once a week but not every day	Disagree	24	24	0.6	0.6
URY	At least once a week but not every day	Strongly Disagree	1	3	0.0	0.1
URY	Every day	Strongly Agree	43	57	1.0	1.4
URY	Every day	Agree	39	48	0.9	1.2
URY	Every day	Disagree	9	10	0.2	0.2
URY	Every day	Strongly Disagree	NA	2	NA	0.0
Note: DNK refers to Denmark, FIN refers to Finland, and URY refers to Uruguay.

Visualize with figures

The table summary can help me to get an idea of the structure of my dataset. However, it takes too much spaces. Therefore, I decided to visualize it with a plot.

My initial attempt was not that bad, but it looks too busy, and the color looks too bright. So I changed the position to dodge in geom_bar, and applied scale_fill_OkabeIto() to make it color-blind friendly.

Show code

rq1_p1 <- ggplot(rq1_data, aes(x = it2g14a, fill = it2g06a)) +
  geom_bar(position = "fill") +
  facet_wrap(year ~ cntry) +
  labs(
    title = "Teacher ICT Use in Teaching by School ICT Priority",
    x = "School ICT Priority",
    y = "Propportion",
    fill = "Teacher ICT Use Frequency in Teaching"
  ) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))
rq1_p1

Show code

rq1_p2 <- ggplot(rq1_data, aes(x = it2g06a, fill = it2g14a)) +
  geom_bar(position = "dodge") +
  facet_wrap(year ~ cntry) +
  labs(
    title = "Relationship between Teacher ICT Use and Schools' perspective on ICT",
    fill = "School ICT Priority",
    y = "Propportion",
    x = "Teacher ICT Use Frequency in Teaching"
  ) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))+
  scale_fill_OkabeIto()
rq1_p2

I am not satisfied with that because it is difficult to tell the difference. So I decided to factor year, and group the data by country, year, ICT use frequency, and school priority and got this.

Show code

rq1_data$year <- as.factor(rq1_data$year)

rq1_p3 <- rq1_data %>% 
  group_by(year, cntry, it2g06a,it2g14a) %>% 
  summarise(count = n(), .groups = "drop") %>% 
  ggplot(aes(x = year,
         y = count,
         group = interaction(cntry,it2g14a),
         color = it2g14a,
         linetype = cntry,
         shape = cntry))+
  geom_line() +
  geom_point(size = 2) +
  facet_grid(~ it2g06a) +
  labs(title = "Figure 1. ICT Use Across Countries (2018 & 2020)",
       x = "Year of Response",
       y = "Count",
       color = "School ICT Priority",
       linetype = "Country",
       shape = "Country"
  ) +
  scale_color_OkabeIto() +
  scale_shape_manual(values = c("DNK" = 0, "FIN" = 15, "URY" = 5))+
  theme(panel.spacing = unit(1, "lines"))

rq1_p3

Research Question 2

What is the relationship between teachers’ technology use in the classroom and their major teaching subjects?
(This question aims to explore whether teachers in different subject areas (e.g., mathematics, sciences, language arts) use technology more frequently or differently than their peers in other subjects).

Variable selection: it2g06a: ICT use at school when teaching
Main Subjects in School Year:
it2g03a: Language arts test language
it2g03b: Language arts foreign and other national languages
it2g03c: Mathematics
it2g03d: Sciences
it2g03e: Human sciences/Humanities
it2g03f: Creative arts
it2g03g: Information technology
it2g03h: Practical and vocational subjects
it2g03i: Other
Type of variable: Binary
IT2G14A: Your School/Use of ICT in teaching at your school/ICT is considered a priority for use in teaching – X-axis
cntry: Country(Denmark, Finland, and Uruguay)
partt: Survey participation
idteach: Teacher ID

Show code

rq2_data <- mergedtp %>% 
  clean_names() %>% 
  select(idteach, cntry, partt, year, it2g06a,it2g03a, it2g03b, it2g03c, it2g03d, it2g03e, it2g03f, it2g03g, it2g03h, it2g03i) %>% 
  filter(partt == 1)

rq2_data <- na.omit(rq2_data)

rq2_data <- rq2_data %>% 
  mutate(it2g06a = factor(it2g06a,
                          level = 1:5,
                          labels = c("Never", "Less than once a month", "At least once a month but not every week", "At least once a week but not every day", "Every day"), ordered = TRUE),
         across(starts_with("it2g03"), ~ factor(.x, levels = 1:2, labels = c("Checked", "Not checked"))))

rq2_aggregated <- rq2_data %>% 
  pivot_longer(cols = starts_with("it2g03"), names_to = "subject", values_to = "checked") %>% 
  filter(checked == "Checked") %>% 
  group_by(subject, it2g06a) %>% 
  summarise(count = n(), .groups = 'drop') %>%
  mutate(subject = factor(subject, 
                          levels = c("it2g03a", "it2g03b", "it2g03c", "it2g03d", "it2g03e", "it2g03f", "it2g03g", "it2g03h", "it2g03i"),
                          labels = c("Language arts test language", "Language arts foreign and other national languages", "Mathematics", "Sciences","Human sciences/Humanities", "Creative arts", "Information technology", "Practical and vocational subjects", "Other")))

Show code

p_subject <- rq2_aggregated %>% 
  ggplot(aes(x = it2g06a, y = subject, fill = count)) +
  geom_tile() +
  scale_fill_gradient(low = "white", high = "orange") +
  labs(title = "Relationship Between Teachers’ Technology Use in School and Teaching Subjects",
       x = "Technology Use",
       y = "Teaching Subject",
       fill = "Count") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

p_subject

Show code

p_subject_2 <- rq2_data %>% 
  pivot_longer(cols = starts_with("it2g03"), names_to = "subject", values_to = "checked") %>% 
  filter(checked == "Checked") %>% 
  mutate(subject = factor(subject, 
                          levels = c("it2g03a", "it2g03b", "it2g03c", "it2g03d", "it2g03e", "it2g03f", "it2g03g", "it2g03h", "it2g03i"),
                          labels = c("Language arts test language", "Language arts foreign and other national languages", "Mathematics", "Sciences","Human sciences/Humanities", "Creative arts", "Information technology", "Practical and vocational subjects", "Other"))) 

p_subject_p2 <- p_subject_2 %>% 
  ggplot(aes(x = subject, fill = it2g06a)) +
  geom_bar(position = "dodge") +
  facet_wrap(~year)+
  labs(title = "Frequncy of ICT Use by Subject",
       x = "Subject",
       y = "Proportion of Teachers",
       fill = "ICT Use Frequency") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))+
  scale_fill_OkabeIto()

p_subject_p2

Show code

p_subject_p3 <- p_subject_2 %>% 
  ggplot(aes(x = subject, y = as.numeric(it2g06a), fill = subject)) +
  geom_violin() +
  labs(title = "Density of Techology Use Frequency by Subject",
       x = "Subject",
       y = "ICT Use Frequency") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))+
  scale_fill_OkabeIto()

p_subject_p3

Show code

p_subject_p4 <- rq2_data %>%
  pivot_longer(cols = starts_with("it2g03"), names_to = "subject", values_to = "checked") %>% 
  filter(checked == "Checked") %>% 
  group_by(subject, it2g06a, year, cntry) %>% 
  summarise(count = n(), .groups = 'drop') %>%
  mutate(subject = factor(subject, 
                          levels = c("it2g03a", "it2g03b", "it2g03c", "it2g03d", "it2g03e", "it2g03f", "it2g03g", "it2g03h", "it2g03i"),
                          labels = c("Language arts test language", "Language arts foreign and other national languages", "Mathematics", "Sciences","Human sciences/Humanities", "Creative arts", "Information technology", "Practical and vocational subjects", "Other"))) %>% 
  ggplot(aes(x = subject, y = count, fill = factor(year))) +
  geom_bar(stat = "identity", position = position_dodge(width = 1)) +
  labs(title = "Relationship Between Teachers’ Technology Use in School and Teaching Subjects",
       x = "Teaching Subjects",
       y = "Number of teachers using ICT",
       fill = "Year")+
  facet_wrap(~ cntry)+
  theme(axis.text.x = element_text(angle = 45, hjust = 1))+
  scale_fill_OkabeIto()

p_subject_p4