Scatterplots
What is a scatterplot?
When investigating bivariate data making a scatterplot is the first step.
What types of scatterplot associations are possible?
Describe the Trend. (Linear or non-Linear)
Describe the Association. (Positive or Negative)
Describe the Strength. (Weak, Moderate, Strong)
Describe the Groups or clusters (possible reasons or causes)
Describe the Unusual Values (any worth noting)
Describe the Scatter (Even or changing)
If the scatterplot indicates a linear model is appropriate then we can proceed.
Discuss what you see in your scatterplots
"From the scatterplot it appears that as (blah) increases then (blah) tends to increase also."
Am I surprised by this? Why am I surprised or not by this (in context)
"i would expect that measurements of (blah) and (blah) are in proportion"
Investigate the effect of other variables or groups.
Examples of scatterplots
What do I look for in scatterplots?
Trend.
Association.
Strength.
Groups.
Unusual values.
Scatter.
Do you see a linear trend...
...or a non-linear trend?
Do you see a positive association...
...or a negative association?
As one variable gets bigger, so does the other.
Do you see little scatter (strong)...
As one variable gets bigger, the other gets smaller
...or lots of scatter (weak)?
Do you see any groupings?
Do you see any unusual values?
Do you see constant scatter...
Is the unusual value an error or just different?
Do NOT remove it unless it is an error
...or non-constant scatter?
Roughly the same amount of scatter as you look across the plot.
The scatter looks like a “fan” or “funnel”.