This video describes how to create a bar graph and then overlay the individual data points for each group to show the within-group variability. Creating publication quality graph in excel ( • Creating a publication... ).
Huge Thank you! This was extremely helpful. The reviewers on my paper asked for the individualized data in the graphs. I would not have figured it on my own. You got a new subscriber here!
Wow, this is so helpful- thanks a ton! I want to submit to a journal that requires individual data points, means and SD's. Very clearly explained, but it's too bad that the process is so convoluted in Excel.
Thanks for the tutorial. I've ran into some issues with this, however. I have 20 data points, yet only 3 are displayed in the plot when I follow your method. Any idea on what might bewrong? I've plotted the data points with respect to the secondary y-axis which covers the whole range of the values, I've also added a small jitter, so I know that the data points are not overlaying each other. The displayed datapoints also seem to be picked randomly.
Thx for the helpful video. My data points (y values) are quite close together and the dots overlap in the scatterplot. Should I change the x values so that the data points spread out horizontally or should I keep them all in the same vertical line?
I would say that this is a stylistic decision that should be left up to the author(s). You might consider changing the y-axis because a restricted y range could make it appear that the individual points are close together when they aren't as similar. Assuming that you have a representative y range, a tight cluster is a good feature of your data in so far as the members of the group (or observations) are very similar. I wouldn't alter the x-axis because the point is to show how the individuals do (or do not) represent the group/condition average.
Thank for nice explanation. If I have a number of values (i.e. 20) how can i see each value on the bar graph. I mean how can I make bar graph and scattered plot together?
Sorry, I don't know. I would have to figure it out. I would look at using box plots in R. They seem to be very easy to provide the kind of graph you are seeking. I hope this helps.
