Charts and graphs

1
Charts and graphs
2
Agenda

• value and limits of graphical analysis
• how to create and read and interpret graphs
• basic types of graphs and conditions under which
  should be used

3
Basic principle

• Charts and graphs, like all numerical
  information, have one function
• To communicate information
• to an audience
• in support of a thesis or claim.

4
Main benefit of graphingEases interpretation of
data.

• Visual representation draws attention to relative
  amounts.
• Easier to see all data simultaneously.

5
Main drawbacks to graphs

• Lose sight of actual amounts.
• Easier to use for deception.

6
Interpretation

• Text explains meaning and significance of results
• Two elements Chart reading and chart
  interpretation

7
Chart reading

• Verbal expression
• of main features of the pattern
• What do the data show?
• Forest, not trees

8
Chart interpretation

• Commentary or analysis
• of appropriate
• conclusions or questions
• What do the data mean?

9
Qualities of good graph

• Appropriate to the data
• Self-explanatory effective title and axis labels
• 3. Simple and uncluttered
• 4. Not misleading

10
Appropriate to data
11
Pie charts

• Slices represent shares of a whole.
• The categories that comprise the variable must be
  qualitatively different or crude rankings.
• The slices/categories must exhaust the
  possibilities.
• The number of slices must be small. Avoid having
  many narrow slices.
• Must include a legend.

12
Slices shares of wholes
13

Slices need not be percentages
14
Total never a slice
15
Pie chart Appropriate data

• Categories represent different qualities
• Sex
• Religion
• Race/ethnicity
• Categories represent crude quantitative
  differences
• Large, small
• High, medium, low
• Strongly agree, agree, disagree, strongly
  disagree
• Freshman, sophomore, junior, senior

16
Pie chart Inappropriate data

• Precise quantities
• Age
• GPA
• of credits

17
Exhaustive categories
18
Bar and column graphs

• Interchangeable.
• Can be used with any kind of frequency count
  data.
• Can be used to show multiple series of data.
• Legend required if using more than one series.

19
Line graphsDont create them using the line
graph option in Excel. . . .The results are
funky. (A technical term)Use xy scatter with
connection of points instead.
20
xy scatter graphs

• Can only be used when both variables are
  quantitative.
• Show the relationship between two variables.
• The causal variable goes on the x (horizontal)
  axis.

21
Self-explanatory

• Key is title and labels
• Title should be clear, concise, complete.
• All variables must be named.
• Measurement units must be properly stated.
• Time period must be stated.
• Spatial or geographic domain must be specified.

22
Simple and uncluttered

• Limited number of variables
• No unnecessary legends.
• No distracting colors and shading.
• Consistent colors and shading.
• Use colors or patterns that will be distinct if
  printed in black white.

23
Misleading graphs and charts.

• Truncated y-axis
• Unlabeled axes
• Arbitrary axis dimensions
• 3-dimensional pictograms
• 3-dimensional pie charts
• Line graphs with data gaps

24
Truncated vertical y axis
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
Arbitrary dimensions
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
3-dimensional pictograms
33
Rising costs of prescriptions
1994
2004
34
Deceptive pictogram

• Problem is that changing one dimension while
  keeping proportions changes both dimension,
  leading to misleading greater change in volume.

35
Pictograms
36
3-dimensional pie charts
37
Three-dimensional pie graphs
38
Line graphs with data gaps