0% found this document useful (0 votes)
4 views15 pages

Data Visualization and Analysis Techniques

The lecture notes cover methods for displaying and describing the distribution of quantitative variables, including histograms, boxplots, and scatterplots. It discusses appropriate measures of center and spread for different types of distributions, emphasizing the use of median for skewed distributions and standard deviation for symmetric distributions. Additionally, it addresses the identification of outliers and the calculation of Z-scores to compare individual observations to a normal distribution.

Uploaded by

kaiwenhsu98
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
4 views15 pages

Data Visualization and Analysis Techniques

The lecture notes cover methods for displaying and describing the distribution of quantitative variables, including histograms, boxplots, and scatterplots. It discusses appropriate measures of center and spread for different types of distributions, emphasizing the use of median for skewed distributions and standard deviation for symmetric distributions. Additionally, it addresses the identification of outliers and the calculation of Z-scores to compare individual observations to a normal distribution.

Uploaded by

kaiwenhsu98
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Lecture Notes

Chapters 1-2

Part III

Turning Data Into Information

Stat 7 – Armstrong
Chapters 1-2 Part 3
1
Displaying the
distribution of a
quantitative
variable

Large Datasets Small Datasets

Histograms Boxplots Stemplots

X axis:
Variable of Interest 5 Number Similar to a
Summary histogram, but
instead of bars we
have the actual
values

Y axis: Minimum
Frequency or Q1
Relative Frequency Median
Q3
Maximum

When we consider two variables at once we will look at a SCATTERPLOT:

The X axis plots the explanatory variable (independent variable)


The Y axis plots the response variable (dependent variable)

Stat 7 – Armstrong
Chapters 1-2 Part 3
2
Describing the
Distribution of a
QUANTITATIVE
VARIABLE

SHAPE CENTER SPREAD

SHAPE

SYMMETRIC SKEWED

or has OUTLIERS

POSITIVELY NEGATIVELY
UNIFORM NORMAL SKEWED SKEWED

RIGHT SKEWED LEFT SKEWED

Other features to consider:

o Gaps: Spaces separating data

o Outliers: Data values that are set far apart from the rest of the body of the distribution

Stat 7 – Armstrong
Chapters 1-2 Part 3
3
APPROPRIATE
MEASURES OF
CENTER

SYMMETRIC DISTRIBUTION SKEWED DISTRIBUTION


OR
DISTN WITH OUTLIERS

MEAN MEDIAN

(average) (middle number)


Not Resistant Resistant

APPROPRIATE
MEASURES OF
SPREAD

SYMMETRIC DISTRIBUTION SKEWED DISTRIBUTION


OR
DISTN WITH OUTLIERS

Standard Deviation Inter Quartile Range

(typical deviation from the mean) (the middle 50% of your data)
Not Resistant Resistant

Stat 7 – Armstrong
Chapters 1-2 Part 3
4
MEDIAN

Appropriate measure
of CENTER for

SKEWED
DISTRIBUTIONS

and

DISTRIBUTIONS
WITH OUTLIERS

The median is
not influenced
Find the Find the by outliers.
MEDIAN MEDIAN
by hand by calculator

1st: 2nd: 1st: 2nd:


It is a
Put all the data Find the middle Put all the data Calculate RESISTANT
in order from position into a list in your measure of
smallest to calculator 1 VAR STATS
center.
largest n1
2

Note: For an odd number of values: Single middle value is the median

Ex: 10, 11, 15, 19, 20

Note: For an even number of values we take the average of the two middle values
Ex: 10, 11, 15, 19, 20, 80

Stat 7 – Armstrong
Chapters 1-2 Part 3
5
QUARTILES

The range is quick &


easy, but not very
informative.

Split up the IQR


distribution into
Inter Quartile Range
quarters

25th percentile 50th percentile 75th percentile Q3 – Q1

First Second Third Measures


Quartile Quartile Quartile Spread for
Skewed Dist

Q1 Q2 Q3

25 % of data 50 % of data 75 % of data


below this below this below this
observation observation observation

___% of data ___% of data ___% of data


above this above this above this
observation observation observation

Stat 7 – Armstrong
Chapters 1-2 Part 3
6
Ex:
Below are the numbers of deaths from tornadoes in the U.S. from 1990 through 2000.

53 39 39 33 69 30 25 67 130 94 40

a) Calculate the median.

b) Calculate the Interquartile range (IQR)

Stat 7 – Armstrong
Chapters 1-2 Part 3
7
Ex: French Exams
The Modern Language Association provides listening tests that measure understanding of spoken
French. The range of scores is 0 to 46. Here are the scores of 21 high school French teachers at
the beginning of an intensive summer course in French (in order).

9 15 20 20 22 23 23 30 30 31 31 32 34 34 35 39 40 42 42 45 46

Create a histogram:

Create a stemplot:

Create a boxplot:

What is the shape of the distribution?

What is the best measure of center for this distribution? WHY?

What is the best measure of spread for this distribution? WHY?

Stat 7 – Armstrong
Chapters 1-2 Part 3
8
Outliers:
Step 1: Determine Quartile 1 and Quartile 3 from the data set.

Step 2: Compute the Interquartile Range (IQR) IQR= Q3-Q1

Step 3: Determine the Fences. Fences serve as cutoff points for determining outliers

Mild Lower Fence = Q1- 1.5(IQR) Extreme Lower Fence = Q1- 2*1.5(IQR)
Mild Upper Fence= Q3 + 1.5(IQR) Extreme Upper Fence= Q3 + 2*1.5(IQR)

Step 4:

If observation < lower fence then it is an OUTLIER

If observation > upper fence then it is an OUTLIER

If lower fence < observation < upper fence it is NOT an OUTLIER

Outliers

What should be done with outliers? First try to understand them in the context of the data. A
histogram can show how the outlier fits with the rest of the data:

 Is there a large gap between the outlier and the rest of the data?

 Is the outlier a value at the end of a stretched out tail?

 Could the outlier be an error?

What you should NOT do:

 Leave an outlier in place without comment, and proceed as if nothing were unusual.

 Drop an outlier without comment just because it is unusual.

Stat 7 – Armstrong
Chapters 1-2 Part 3
9
Ex: The data set below is the case prices (in dollars) of wines produced by a vineyard in
Napa Valley.

150 135 90 122 128 67 142 140 128 132 127 140 129

Stat 7 – Armstrong
Chapters 1-2 Part 3
10
MEAN

Measures
CENTER
for
SYMMETRIC
Distributions

The mean is
Find the Find the
SAMPLE MEAN SAMPLE influenced
MEAN
X by outliers.
by hand X
by calculator

1st: 2nd: 1st: 2nd:


It is a
Add up all of the Divide Put all the data Calculate NON-
data values the sum by the
sample size
X 1  ...  X n into a list in your
calculator 1 VAR STATS
RESISTANT
measure of
n center.

 The true mean of a population uses the notation 

Stat 7 – Armstrong
Chapters 1-2 Part 3
11
Comparing the Mean and the Median:

When we look at distributions, where is the mean and where is the median?

NORMAL “SYMMETRIC” CURVE SKEWED DISTRIBUTIONS

 The mean will be pulled towards the tail of the skewed data.
 For Normal Bell Shaped Distributions, the mean is the more appropriate measure of the
center.
 For Skewed Distributions, the median is the more appropriate measure of the center.

Why use the mean then?

It is easy to do calculations on the mean.

The mean has nice properties that we will use later in this course.

Stat 7 – Armstrong
Chapters 1-2 Part 3
12
STANDARD
DEVIATION

Measures
SPREAD
for
SYMMETRIC
Distributions

Find the Find the The standard


SAMPLE POPULATION deviation is
STANDARD STANDARD influenced
DEVIATION DEVIATION by outliers.
s 

By hand By Calculator By hand By Calculator It is a


NON-
RESISTANT
measure of
spread.

Will not be 1st: Not possible. Not Possible.


asked to find Put all the If you need Do not let the
by hand. data into a , it should calculator
Formula in list in your be given to fool you.
text. calculator you.

2nd:
Calculate

1 VAR
STATS
Stat 7 – Armstrong
Chapters 1-2 Part 3
13
Variance is the standard deviation squared.

Standard Deviation is the positive square root of the variance

It is a distance measure so it has to be positive

It is the “typical” distance of the datapoints to the mean.

A standard deviation of zero means ________________________.

Values that are very close together have a small standard deviation.

Values that are very far apart have a large standard deviation.

Ex: 1, 2, 3, 4, 5 Calculate the standard deviation

Ex: 1, 1, 1, 1, 1 Calculate the standard deviation

NOTATION:

Sample Mean X
Population Mean 

Sample Standard deviation s


Sample Variance s2

Population Standard deviation 


Population Variance 

Stat 7 – Armstrong
Chapters 1-2 Part 3
14
Suppose we have that a single observation (X) comes from a normal distribution with 
(population mean) and  (population standard deviation) both given.

We write that the distribution of X is:

X ~ Normal (  ,  )
We say: “X is normally distributed with mean = and standard deviation = ”

X 
Z
Z-Score calculation: 
The Z SCORE is unitless.

The Z SCORE tells us how far our observation is from the mean in terms of standard deviations.

Ex: Two friends are training for the Boston marathon. James is training on a hilly jogging loop. For
the general population of runners, the time to complete this loop follows a normal distribution with a
mean of 167 minutes and standard deviation 25 minutes. Rob is training on a flat jogging route. The
time to complete this flat route follows a normal distribution with a mean of 143 minutes and
standard deviation 20 minutes. If it takes James 91 minutes to complete his loop, and it takes Rob 86
minutes to complete his loop, who is in better condition?

Draw a picture for each of the distributions.

What is the Z-Score for each runner?

Who is in better condition?

Stat 7 – Armstrong
Chapters 1-2 Part 3
15

Common questions

Powered by AI

When interpreting outliers, it's important to consider the context and potential reasons for their presence, such as errors in data collection or natural variability. Outliers should not be dismissed without investigation as they may highlight important insights or anomalies. Analysis should determine whether these values indicate errors or unique cases that need separate attention .

The median is a more suitable measure of center for distributions with outliers because it is resistant to the effects of extreme values. It represents the middle value, thereby remaining stable regardless of outliers, unlike the mean, which can be significantly skewed by outliers .

Scatterplots allow for the analysis of relationships between two variables by plotting the explanatory variable on the X-axis and the response variable on the Y-axis. This visual representation helps identify correlations, trends, patterns, and the strength and direction of the relationship between the variables .

Quartiles divide a dataset into four equal parts, providing insights into the spread and center of data by indicating where 25%, 50%, and 75% of observations lie. The interquartile range (IQR), calculated as Q3 minus Q1, measures the spread of the middle 50% of the data, offering a resistant measure against outliers and depicting variability within the central portion .

Z-scores normalize scores from different normal distributions by converting individual scores into standard deviations away from the mean. This allows for direct comparison across different datasets by accounting for the mean and standard deviation specific to each distribution, hence evaluating relative performance or condition between different distributions .

The mean is the appropriate measure of center for a normally distributed data set because it represents the central point of a symmetric distribution and is not influenced by the skewness or outliers typically present in non-normal distributions .

The mean is not considered a resistant measure of center because it is influenced by extreme values or outliers, which can significantly skew the mean away from its representative central position in a dataset .

Standard deviation is preferable for symmetric distributions as it accurately reflects the dispersion around the mean. However, its limitation is that it is a non-resistant measure, being sensitive to outliers and skewed data, which can misrepresent the data's actual variability .

Histograms display the distribution of a quantitative variable using bars to show frequency or relative frequency for different intervals, which is useful for identifying the shape of the distribution and patterns. Boxplots, on the other hand, visually represent the five-number summary, including quartiles and potential outliers, which highlight the data's spread and symmetry .

For skewed distributions with outliers, the interquartile range (IQR) should be used because it is a resistant measure of spread. Unlike standard deviation, IQR is not affected by extreme values and provides a reliable measure of variability for the middle 50% of the data .

You might also like