Connecting Dots of Repeated Measures Data across Bar Plots

[Mik_code]

How can I connect with lines the dots over the first bar with the dots over the second bar and the dots of the second bar with the dots of the third bar.

Each dot represent a single subject in a repeated measure design experiment.

For the barplot and the scatterplot the data came from different dataset.

std.error <- function(x) sd(x)/sqrt(length(x))


  
  session_order <- c("Pre", "During", "Post_1")
  ggplot() +
    geom_bar(data = my_data_isolation_ordered[my_data_isolation_ordered$Manipulation=="Attended Visual Field" & my_data_isolation_ordered$Stimulation=="Sham",],
             aes(x = Session, y = Accuracy),
             color = "#20B2AA", fill = "#20B2AA", stat = "summary", width = 0.5) +
    geom_errorbar(data = my_data_isolation_ordered[my_data_isolation_ordered$Manipulation=="Attended Visual Field" & my_data_isolation_ordered$Stimulation=="Sham",],
                  aes(x = Session, y = Accuracy, ymin = mean(Accuracy) - std.error(Accuracy), ymax = mean(Accuracy) + std.error(Accuracy)),
                  color = "black", width = 0.2) +
    geom_point(data = my_data_isolation_ordered[my_data_isolation_ordered$Manipulation=="Attended Visual Field" & my_data_isolation_ordered$Stimulation=="Sham",],
               aes(x = Session, y = Accuracy),
               alpha = 1.5, shape = 21, size = 2.5, color = "black") +
    geom_bar(data = my_data_sham[my_data_sham$Session == "Pre" & my_data_sham$Manipulation == "Attended Visual Field",],
             aes(x = Session, y = Accuracy),
             color = "#20B2AA", fill = "#20B2AA", stat = "summary", width = 0.5) +
    geom_errorbar(data = my_data_sham[my_data_sham$Session == "Pre" & my_data_sham$Manipulation == "Attended Visual Field",],
                  aes(x = Session, y = Accuracy, ymin = mean(Accuracy) - std.error(Accuracy), ymax = mean(Accuracy) + std.error(Accuracy)),
                  color = "black", width = 0.2) +  
    geom_point(data = my_data_sham[my_data_sham$Session == "Pre" & my_data_sham$Manipulation == "Attended Visual Field",],
               aes(x = Session, y = Accuracy),
               alpha = 1.5, shape = 21, size = 2.5, color = "black") +
    geom_bar(data = my_data_sham[my_data_sham$Session == "Post_1" & my_data_sham$Manipulation == "Attended Visual Field",],
             aes(x = Session, y = Accuracy),
             color = "#20B2AA", fill = "#20B2AA", stat = "summary", width = 0.5) +
    geom_errorbar(data = my_data_sham[my_data_sham$Session == "Post_1" & my_data_sham$Manipulation == "Attended Visual Field",],
                  aes(x = Session, y = Accuracy, ymin = mean(Accuracy) - std.error(Accuracy), ymax = mean(Accuracy) + std.error(Accuracy)),
                  color = "black", width = 0.2) +
    geom_point(data = my_data_sham[my_data_sham$Session == "Post_1" & my_data_sham$Manipulation == "Attended Visual Field",],
               aes(x = Session, y = Accuracy),
               alpha = 1.5, shape = 21, size = 2.5, color = "black") +
    scale_x_discrete(limits = session_order)

Also, is there a best way to do it?

[jrkrideau]

I am not sure about what you want to do but if you want a parallel coordinates chart have a look at the ggally package.

As a point of genoral graphing, may I suggest dropping the barplot? It is compressing your data and nod adding any useful information.

[M_AcostaCH]

Maybe is better use geom_jitter(), for avoid overlap the points.

[jrkrideau]

Yes, I keep forgetting geom_jitter() .