Descriptive Statistics

Box plots

When comparing two or more groups, it is generally a good idea to start by creating some form of side-by-side box plots or quantile plots. These plots reveal important information about the location, spread, and shape of the distribution of scores in each group.

The plot shows pain ratings of subjects treated with active magnets (Treatment) and those treated with inactive placebo devices (Control).


What is the mean for the group treated with active magnets?
9
Just over 8
Just over 4
4
Blank


The lowest pain rating of the control group appears to be equal to:
The mean of the treatment group
The median of the treatment group
The highest score in the treatment group
Blank
Overall, the group in the most pain is the
control group
treatment group
Blank
The middle 50% of the scores for the control group fall between:
2-6
5-10
8-10
0-10
Blank
The circle represents:
An outlier
A score not included in the calculations
A score which did not fit in the treatment group
Blank




The histograms shown below provide more details about the shapes of the distributions. The Y axis labeled density is equal to the frequency divided by the sample size. For the treatment group, 4 out of 29 subjects rated their level of pain as 0. Therefore the density for the first bar is 4/29 = 0.14.

The red curve in the distribution shows what a normal distribution with the same mean and variance as the data would look like.





The post-treatment scores for the placebo (control) group:
Have a positive skew
Are leptokurtic
Are platykurtic
Have a negative skew
Blank

 

Basic descriptive statistics are shown below.

 

Condition
N
Mean
Median
Trimean
Minimum
Maximum
25th percentile
75th percentile
s
se of mean
Skew
Kurtosis

Treatment
29
4.38
4.00
4.00
0.00
10.00
2.00
6.00
3.14
0.59
0.50
-0.74

Control
21
8.43
9.00
9.00
4.00
10.00
8.00
10.00
1.86
0.41
-0.96
-0.27

 


What was the range for the group that received the active magnets?

8.43
6
4.38
10
Blank


The pain ratings are more variable in
The treatment condition
The placebo (control) condition
Blank