Why Graph?
You need five primary things in life: food/water, clothing, shelter, friends and family, and graphing. Graphs make life fun and exciting, and if they were edible, they’d be the best food you would ever have, times infinity.
Graphs make data analysis easier. However, data can be graphed differently, producing different interpretations of data. Therefore, the chart you choose needs to be the graph that best shows the major trends in the data.
Huh?
Alright, here is an example. You want to know how extreme wall staring affects heart rate. You decide on a duration of 5 minutes of extreme wall staring with the subjects’ heart rate measured at the end of each minute. You first measure the subjects’ resting heart rate and record the measurements in your data table. The subjects begin the five minutes of extreme wall staring.
Wait. Is a data table a type of graph?
No. A data table is a place to input data during an experiment and must be created before experimentation. A graph is the analysis of data in the table after the experiment is complete.
Can you elaborate?
Yep. Let’s start with how to create a data table.
How to Make a Data Table
Before you begin an experiment, you need to create a data table. A data table should contain enough rows and columns for your data, and each row and column needs to be titled correctly and includes units. The data table needs a descriptive title that briefly explains the data.
Here is an example of a simple data table:
Tables that include very little data and that are made correctly are usually easy to read for analysis. (You still want to graph the data because a graph, when done correctly, makes data analysis easier and faster.)
Let’s make a data table for the extreme wall staring example. To start, figure out how many rows and columns you will need for data collection. For the extreme wall staring experiment, we will be collecting the subjects’ resting heart rate data, heart rate data each minute during extreme wall staring, and the average HR data for males and females.
This data hurts my head.
That is a normal response. As you can see, there is a lot of information in the data table above. Analyzing and interpreting the data using the table can be confusing. It is possible that instead of one large data table, two or three smaller tables can be made. However, no matter how many data tables you use, it is best used to collect data, not analyze it. You need to graph data to appreciate it fully and marvel at its awesomeness (or lack of).
Sparkline Graphs
Sparkline graphs are a great way of showing the primary trend in the data. Here is a spark graph of the Dow Jones Industrial Index.
What is the trend?
The Dow Jones increased.
Good.
By how much?
About 220 points.
Awesome.
How long did it take the Dow to rise by 220 points?
Umm, I don’t know.
Why?
There are no axis labels.
Exactly. What does the horizontal dotted line represent?
Well, I am not sure.
It represents the closing stock index from the previous day.
How am I supposed to know that?
You can’t know if it is not labeled. Sparklines are great at showing the primary trend, but they do not give enough information for in-depth data analysis.
So, what type of graphs should I make?
Well, that depends on the data you are graphing. However, line graphs and column graphs will be the most common graphs you make this school year.
Sparklines pros and cons:
- Pros – Quick look at the major trend
- Cons – No axis labels; not good for in-depth data analysis
When to Make a Line Graph
You want to make a line graph when comparing changes in a variable or trends over time (continuous). Line graphs are preferred over bar graphs when differences in data over time are small because it is much easier to see the trend(s).
A line graph should include labels and units on the x-axis and y-axis, the x-axis and y-axis are adjusted to the right scale, data points, a legend (key), and a title that briefly describes the graphed data.
Pros and cons of graphing the exercise heart rate per minute for all ten subjects:
- Pros – All the subjects’ data is displayed
- Cons – Too many lines can make analysis difficult
Pros and cons of graphing the average heart rate per minute for all ten subjects:
- Pros – Makes data analysis of the major trend(s) easier
- Cons – Does not show the individuals’ data
When to Make a Column (Bar) Graph
You want to make a column graph when comparing different categories or one point in time to another point (non-continuous). Bar graphs can plot changes over time, but only if the changes are not continuous.
A column graph should include labels and units on the x-axis and y-axis, the x-axis and y-axis are adjusted to the right scale, and a title that briefly describes the graphed data. A legend (key) may or may not be needed.
When to Make a Line Graph with TWO Y-Axes
You want to make a double y-axis line graph when comparing two variables that have different units or two things that have drastically different quantities over time. Double y-axis graphs allow you to see if there is a pattern between the two variables.
A double y-axis line graph should include labels and units on the x-axis and BOTH y-axes, the x-axis and y-axes are adjusted to the right scale, data points, a legend (key), and a title that briefly describes the graphed data.
When to Make a Histogram
You want to make a histogram when you are looking at quantitative data and the frequency in which that data occurs. Histograms look similar to column/bar graphs, but there are some notable differences. A bar graph can have either qualitative or quantitative data on the x-axis, and there is a separation between each bar/column. Histograms only have quantitative data on the x-axis, and there are no spaces between each column/bar. Also, histograms look at the frequency at which a particular quantity appears, which produces either a stabilizing, disruptive or directional curve.
A histogram should include labels and units on the x-axis and y-axis, the x-axis and y-axis are adjusted to the right scale, and a title that briefly describes the graphed data. A legend (key) may or may not be needed.
When to Make a Pie Chart
Pie charts are used when comparing parts of a whole.
Pie charts should include the percentage of each slice, a legend, a title that briefly describes the chart’s data, and each slice is of a different color.
Here is a video on how to make the pie chart pictured below.
(A hematocrit is a laboratory test that measures the ratio of erythrocytes (red blood cells) to the total volume of blood.)
When to Make a Scatter Plot
You want to make a scatter plot when comparing two different variables to see if there is a correlation between them. Adding a trend line and R2 value will help interpret the degree of correlation between the two variables.
A scatter plot should include labels and units on the x-axis and y-axis, the x-axis and y-axis are adjusted to the right scale, and a title that briefly describes the graphed data. A trend line and R2 valves need to be added, as well.
