By: Kristin Hunter-Thomson
Help, my plots with time look really strange!
Have you ever had this experience? You plot your data and the x-axis comes up looking at bit off.
OK, not because it is set up backwards. We are talking about the depth in the water that we found these fish (Black Sea Bass). So we put the surface of the water (0m) at the top and deeper into the water down at the bottom (60m).
At first glance the graph looks good, right? There are data points across time in a line. It seems the depth that we have found Black Sea Bass over time from 1967 to 2016 has overall remained similar. But then what about when you look at the x-axis to get a sense of time these data were collected? Wait, it doesn’t seem right…
The space on the graph between 1967 and 1972 is the same as the space between 1972 and 1973. Even though the gaps differ from 5 years and 1 year. How can that be? Why are they graphing the same when there is obviously a difference?
What I want is a graph that actually looks like this…
In this version, the space along the x-axis in this graph accurately and equally represents all years. Meaning the dots are spaced according to their year rather than being just the next row in the data table.
So, how do I make the second rather than first graph?
If your data look like this as the data are not collected at equal intervals (meaning the amount of time between each new data collected is not the same) then you need to make your graph in Excel or Google Sheets a bit differently than you would expect.
While you are making a line graph, to see a change over time, the trouble comes in executing that graph in common graphing software/platforms.
In Excel and other graphing packages, their backend coding expects the data being plotted in a “Line Chart” to be categorical, rather than continuous. Meaning that these packages read the “Year” column as words “1967” “1972” “1973” as if they were “Apples” “Bananas” “Kiwis”. Therefore, the graphing package does not plot the data as numbers with a numerical value.
This is why the interval, space, between each data point (aka “1967” “1972” “1973”) are treated as equal steps rather than as numbers. The package doesn’t know that it should be treated differently.
The way around this is to choose a “Scatterplot” (for which you connect the data points) rather than a “Line Chart.” These graphing packages expect that the data in a Scatterplot are numerical and thus the packages read the “Year” column “1967” “1972” “1973” as numbers as if they were “1,967” “1,972” “1,973”. Therefore, the data points are plotted based on their numerical value rather than an arbitrary difference.
So, rather than using “Marked Lines/Line with Markers” as your go to graph choice for line graphs, consider teaching your students to use “Connected Scatterplot/Scatter with Straight Lines and Markers.”
Have fun with line graphs!