TLX Analysis
Anthony Conrardy
2024-07-10
Read in Excel File with Task Load Index (TLX) Data
Read in the Excel file and change column names for easier analysis. Create a Team Size measure.
tlx_df <- read_excel("D:/Documents/R_Working_Directory/TLX Data Analysis/TLI Data(working).xlsx")
colnames(tlx_df)[colnames(tlx_df)=="Number"] <- "Team"
tlx_df <- tlx_df |> group_by(Team, Scenario) |>
mutate(team_size = n_distinct(LN)) |>
ungroup()
tlx_df$Scenario <- factor(tlx_df$Scenario)Correlation Matrix with the variables for individual participants
Create an individual correlation matrix displaying values for each Task Load Index factor.
tlx_df_corr <- tlx_df |> select(Performance, Effort, Frustration, MD, TD, PD)
cor_matrix <- cor(tlx_df_corr, use = "complete.obs")
new_labels <- c("Performance", "Effort", "Frustration", "Mental Demand", "Temporal Demand", "Physical Demand")
colnames(cor_matrix) <- new_labels
rownames(cor_matrix) <- new_labels
corrplot(cor_matrix, method = "color", type = "upper",
addCoef.col = "black", # Add correlation coefficients to plot
tl.col = "black", tl.srt = 45, # Rotate x-axis labels for readability
title = "Correlation Matrix of Individual Measures",
mar = c(0,0,2,0)) 
Create Team Performance, Frustration, Mental Demand, Effort, Physical Demand, and Temporal Demand Measures
These factors are weighted by team size.
tlx_df <- tlx_df %>%
group_by(Team, Scenario) %>%
mutate(team_Performance = round(mean(Performance),2),
team_Frustration = round(mean(Frustration),2),
team_MD = round(mean(MD),2),
team_Effort = round(mean(Effort),2),
team_PD = round(mean(PD),2),
team_TD = round(mean(TD),2))|>
ungroup()Correlation Matrix with the variables for aggregated teams measures
Create an correlation matrix displaying values for each Task Load Index factor.
tlx_df_corr_team <- tlx_df |> select(team_Performance, team_Effort, team_Frustration, team_MD, team_TD, team_PD)
cor_matrix_team <- cor(tlx_df_corr_team, use = "complete.obs")
new_labels <- c("Performance", "Effort", "Frustration", "Mental Demand", "Temporal Demand", "Physical Demand")
colnames(cor_matrix_team) <- new_labels
rownames(cor_matrix_team) <- new_labels
corrplot(cor_matrix_team, method = "color", type = "upper",
addCoef.col = "black", # Add correlation coefficients to plot
tl.col = "black", tl.srt = 45,# Rotate x-axis labels for readability
title = "Correlation Matrix of Team Measures",
mar = c(0,0,2,0)) 
## Plot Performance by Frustration for Individual Values by Scenario
suppressMessages(ggplot(tlx_df, aes(x = Performance, y = Frustration))) +
geom_point() +
geom_smooth(method = "lm", se=TRUE, linewidth = 1.0, color="red") +
facet_wrap(~Scenario) +
labs(title = "Performance by Frustration for Individual Values",
x = "Performance",
y = "Frustration")## `geom_smooth()` using formula = 'y ~ x'
#theme(legend.position = "none") +
#scale_color_gradient(low="lightseagreen", high="red") +
#guides(color=guide_legend(reverse = FALSE))Plot Overall Performance by Team and Scenario
suppressMessages(ggplot(tlx_df, aes(x = Scenario, y = team_Performance, group = 1))) +
geom_line(aes(color=Team), linewidth=1.0) +
geom_point(aes(color=Team)) +
geom_smooth(method = "lm", se=FALSE, linewidth = 1.0, color="black") +
facet_wrap(~Team) +
labs(title = "Team Performance by Scenario",
x = "Scenario",
y = "Performance") +
theme(legend.position = "none") +
scale_color_gradient(low="lightseagreen", high="red") +
guides(color=guide_legend(reverse = FALSE))## `geom_smooth()` using formula = 'y ~ x'
ggplot(tlx_df, aes(x = as.numeric(Scenario), y = team_Performance)) +
geom_point(color = "darkblue") + # Add points to the plot
geom_smooth(method = "lm", se=TRUE, size =1.0, color="red") +
xlab("Scenario") + # Label x-axis
ylab("Performance Mean") # Label y-axis## Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
## ℹ Please use `linewidth` instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.## `geom_smooth()` using formula = 'y ~ x'
ggplot(tlx_df, aes(x = as.factor(Scenario), y = team_Performance)) +
geom_boxplot(fill = "darkblue", color = "limegreen") + # Add boxplot layer
xlab("Scenario") + # Label x-axis
ylab("Performance Mean") # Label y-axis 
Plot Frustration by Team and Scenario
ggplot(tlx_df, aes(x=Scenario, y=team_Frustration, group = 1)) +
geom_line(aes(color=Team), size=1.0) +
geom_point(aes(color=Team)) +
geom_smooth(method = "lm", se=FALSE, linewidth = 1.0, color="black") +
facet_wrap(~Team) +
labs(title = "Team Frustration by Scenario",
x = "Scenario",
y = "Frustration")+
theme(legend.position = "none") +
scale_color_gradient(low="lightseagreen", high="red") +
guides(color=guide_legend(reverse = FALSE))## `geom_smooth()` using formula = 'y ~ x'
ggplot(tlx_df, aes(x = as.numeric(Scenario), y = team_Frustration)) +
geom_point(color = "darkblue") + # Add points to the plot
geom_smooth(method = "lm", se=TRUE, size =1.0, color="red") +
xlab("Scenario") + # Label x-axis
ylab("Frustration Mean") # Label y-axis## `geom_smooth()` using formula = 'y ~ x'
Plot Mental Demand by Team and Scenario
ggplot(tlx_df, aes(x=Scenario, y=team_MD, group = 1)) +
geom_line(aes(color=Team), size=1.0) +
geom_point(aes(color=Team)) +
geom_smooth(method = "lm", se=FALSE, linewidth = 1.0, color="black") +
facet_wrap(~Team) +
labs(title = "Team Mental Demand by Scenario",
x = "Scenario",
y = "Mental Demand") +
theme(legend.position = "none") +
scale_color_gradient(low="lightseagreen", high="red") +
guides(color=guide_legend(reverse = FALSE))## `geom_smooth()` using formula = 'y ~ x'
ggplot(tlx_df, aes(x = as.numeric(Scenario), y = team_MD)) +
geom_point(color = "darkblue") + # Add points to the plot
geom_smooth(method = "lm", se=TRUE, size =1.0, color="red") +
xlab("Scenario") + # Label x-axis
ylab(" Mental Demand Mean") # Label y-axis## `geom_smooth()` using formula = 'y ~ x'
Regression Analysis Performance vs. Frustration
subset_team_2 <- tlx_df |> select(Team, Scenario, team_Performance, team_MD, team_Frustration) |>
distinct() |>
ungroup()
#write_xlsx(subset_team_2, "team_test.xlsx")reg_perf <- lm(team_Performance ~ team_Frustration + team_MD, data = subset_team_2)
summary(reg_perf)##
## Call:
## lm(formula = team_Performance ~ team_Frustration + team_MD, data = subset_team_2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -4.0916 -1.1020 0.0985 1.2518 3.8540
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 16.31263 1.01791 16.026 < 2e-16 ***
## team_Frustration -0.53056 0.08438 -6.288 2.53e-08 ***
## team_MD 0.06244 0.11231 0.556 0.58
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.683 on 69 degrees of freedom
## Multiple R-squared: 0.4564, Adjusted R-squared: 0.4406
## F-statistic: 28.97 on 2 and 69 DF, p-value: 7.363e-10residuals <- resid(reg_perf)
plot(fitted(reg_perf), residuals,
xlab = "Fitted Values", ylab = "Residuals",
main = "Residuals vs. Fitted Values")
abline(h = 0, col = "red", lty = 2) 
reg_perf_int <- lm(team_Performance ~ team_Frustration + team_MD + (team_Frustration * team_MD), data = subset_team_2)
summary(reg_perf_int)##
## Call:
## lm(formula = team_Performance ~ team_Frustration + team_MD +
## (team_Frustration * team_MD), data = subset_team_2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.201 -1.170 0.093 1.090 3.664
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 10.69517 3.15191 3.393 0.00116 **
## team_Frustration 0.03290 0.31107 0.106 0.91609
## team_MD 0.58150 0.29741 1.955 0.05467 .
## team_Frustration:team_MD -0.05019 0.02671 -1.879 0.06449 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.653 on 68 degrees of freedom
## Multiple R-squared: 0.4832, Adjusted R-squared: 0.4604
## F-statistic: 21.2 on 3 and 68 DF, p-value: 8.381e-10residuals <- resid(reg_perf_int)
plot(fitted(reg_perf_int), residuals,
xlab = "Fitted Values", ylab = "Residuals",
main = "Residuals vs. Fitted Values")
abline(h = 0, col = "red", lty = 2)
reg_perf_s <- lm(team_Performance ~ team_Frustration + team_MD + Scenario, data = subset_team_2)
summary(reg_perf_s)##
## Call:
## lm(formula = team_Performance ~ team_Frustration + team_MD +
## Scenario, data = subset_team_2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.7575 -1.0023 -0.1337 1.2126 3.3518
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 14.74230 1.21311 12.152 < 2e-16 ***
## team_Frustration -0.52873 0.08371 -6.316 2.9e-08 ***
## team_MD 0.11375 0.11462 0.992 0.3248
## Scenario2 0.89322 0.68077 1.312 0.1942
## Scenario3 0.86404 0.68697 1.258 0.2130
## Scenario4 1.04142 0.69060 1.508 0.1365
## Scenario5 1.43708 0.68988 2.083 0.0412 *
## Scenario6 1.58491 0.69807 2.270 0.0266 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.667 on 64 degrees of freedom
## Multiple R-squared: 0.5056, Adjusted R-squared: 0.4515
## F-statistic: 9.35 on 7 and 64 DF, p-value: 6.33e-08residuals <- resid(reg_perf_s)
plot(fitted(reg_perf_s), residuals,
xlab = "Fitted Values", ylab = "Residuals",
main = "Residuals vs. Fitted Values")
abline(h = 0, col = "red", lty = 2) 
reg_model_sug <- lm(team_Performance ~ team_Frustration + team_MD + (team_Frustration * team_MD) + Scenario, data = subset_team_2)
summary(reg_model_sug)##
## Call:
## lm(formula = team_Performance ~ team_Frustration + team_MD +
## (team_Frustration * team_MD) + Scenario, data = subset_team_2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.2996 -1.2304 -0.0847 1.2778 3.1302
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 8.33009 3.29698 2.527 0.0140 *
## team_Frustration 0.09980 0.31251 0.319 0.7505
## team_MD 0.69692 0.30138 2.312 0.0240 *
## Scenario2 0.99349 0.66541 1.493 0.1404
## Scenario3 1.15540 0.68415 1.689 0.0962 .
## Scenario4 1.11136 0.67408 1.649 0.1042
## Scenario5 1.55168 0.67480 2.299 0.0248 *
## Scenario6 1.65204 0.68130 2.425 0.0182 *
## team_Frustration:team_MD -0.05606 0.02691 -2.084 0.0413 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.625 on 63 degrees of freedom
## Multiple R-squared: 0.5375, Adjusted R-squared: 0.4787
## F-statistic: 9.151 on 8 and 63 DF, p-value: 2.982e-08residuals <- resid(reg_model_sug)
plot(fitted(reg_model_sug), residuals,
xlab = "Fitted Values", ylab = "Residuals",
main = "Residuals vs. Fitted Values")
abline(h = 0, col = "red", lty = 2) 
reg_model_square <- lm(team_Performance ~ team_Frustration + I(team_Frustration^2) + team_MD + I(team_MD^2) + (team_Frustration * team_MD) + Scenario, data = subset_team_2)
summary(reg_model_square)##
## Call:
## lm(formula = team_Performance ~ team_Frustration + I(team_Frustration^2) +
## team_MD + I(team_MD^2) + (team_Frustration * team_MD) + Scenario,
## data = subset_team_2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.2100 -1.0849 -0.0694 1.0702 3.3492
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5.29716 4.21014 1.258 0.2131
## team_Frustration 0.17407 0.54072 0.322 0.7486
## I(team_Frustration^2) -0.04652 0.02880 -1.615 0.1114
## team_MD 1.23991 0.87299 1.420 0.1606
## I(team_MD^2) -0.05840 0.05015 -1.164 0.2488
## Scenario2 0.78417 0.66754 1.175 0.2447
## Scenario3 1.11165 0.67709 1.642 0.1058
## Scenario4 1.06260 0.67187 1.582 0.1189
## Scenario5 1.50117 0.67127 2.236 0.0290 *
## Scenario6 1.69710 0.67375 2.519 0.0144 *
## team_Frustration:team_MD 0.02099 0.05028 0.418 0.6778
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.605 on 61 degrees of freedom
## Multiple R-squared: 0.5631, Adjusted R-squared: 0.4915
## F-statistic: 7.862 on 10 and 61 DF, p-value: 5.892e-08residuals <- resid(reg_model_square)
plot(fitted(reg_model_sug), residuals,
xlab = "Fitted Values", ylab = "Residuals",
main = "Residuals vs. Fitted Values")
abline(h = 0, col = "red", lty = 2) 
reg_model_square_f <- lm(team_Performance ~ team_Frustration + I(team_Frustration^2) + team_MD + I(team_MD^2) + (team_Frustration * team_MD), data = subset_team_2)
summary(reg_model_square_f)##
## Call:
## lm(formula = team_Performance ~ team_Frustration + I(team_Frustration^2) +
## team_MD + I(team_MD^2) + (team_Frustration * team_MD), data = subset_team_2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.2205 -1.1224 -0.0258 1.2126 3.7219
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 7.270470 4.112268 1.768 0.0817 .
## team_Frustration 0.001063 0.541590 0.002 0.9984
## I(team_Frustration^2) -0.040665 0.028441 -1.430 0.1575
## team_MD 1.279746 0.866583 1.477 0.1445
## I(team_MD^2) -0.065029 0.050186 -1.296 0.1996
## team_Frustration:team_MD 0.025924 0.050257 0.516 0.6077
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.637 on 66 degrees of freedom
## Multiple R-squared: 0.5079, Adjusted R-squared: 0.4706
## F-statistic: 13.62 on 5 and 66 DF, p-value: 3.929e-09residuals <- resid(reg_model_square_f)
plot(fitted(reg_model_square_f), residuals,
xlab = "Fitted Values", ylab = "Residuals",
main = "Residuals vs. Fitted Values")
abline(h = 0, col = "red", lty = 2)
min(tlx_df$team_Frustration)## [1] 4max(tlx_df$team_Frustration)## [1] 16.2min(tlx_df$team_MD)## [1] 6.17max(tlx_df$team_MD)## [1] 16.5reg_model_square_f <- lm(team_Performance ~ team_Frustration + I(team_Frustration^2)
+ team_MD + I(team_MD^2)
+ (team_MD * team_Frustration), data = subset_team_2)
summary(reg_model_square_f)##
## Call:
## lm(formula = team_Performance ~ team_Frustration + I(team_Frustration^2) +
## team_MD + I(team_MD^2) + (team_MD * team_Frustration), data = subset_team_2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.2205 -1.1224 -0.0258 1.2126 3.7219
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 7.270470 4.112268 1.768 0.0817 .
## team_Frustration 0.001063 0.541590 0.002 0.9984
## I(team_Frustration^2) -0.040665 0.028441 -1.430 0.1575
## team_MD 1.279746 0.866583 1.477 0.1445
## I(team_MD^2) -0.065029 0.050186 -1.296 0.1996
## team_Frustration:team_MD 0.025924 0.050257 0.516 0.6077
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.637 on 66 degrees of freedom
## Multiple R-squared: 0.5079, Adjusted R-squared: 0.4706
## F-statistic: 13.62 on 5 and 66 DF, p-value: 3.929e-09# Assuming reg_model_square_f is your fitted model object
# Extract residuals
model_residuals <- residuals(reg_model_square_f)
# Create the residual plot
plot(fitted(reg_model_square_f), model_residuals,
xlab = "Fitted Values", ylab = "Residuals",
main = "Residuals vs. Fitted Values")
# Add the horizontal line
abline(h = 0, col = "red", lty = 2)
min(tlx_df$team_Frustration)## [1] 4max(tlx_df$team_Frustration)## [1] 16.2min(tlx_df$team_MD)## [1] 6.17max(tlx_df$team_MD)## [1] 16.5Scatter Plots
ggplot(subset_team_2, aes(x= team_MD, y = team_Performance)) +
geom_point() +
geom_smooth(method = "lm") +
labs(x = "Mental Demand", y = "Performance")## `geom_smooth()` using formula = 'y ~ x'
ggplot(subset_team_2, aes(x= team_Frustration, y = team_Performance)) +
geom_point() +
geom_smooth(method = "lm") +
labs(x = "Frustration", y = "Performance")## `geom_smooth()` using formula = 'y ~ x'
ggplot(subset_team_2, aes(x= team_Frustration, y = team_MD)) +
geom_point() +
geom_smooth(method = "lm") +
labs(x = "Frustration", y = "Mental Demand")## `geom_smooth()` using formula = 'y ~ x'