Agric45 Presents'''

1
Agric45 Presents...

• Presenting Data...
• Which Chart?

2
How Should Data be Presented?

• 1. Is a chart the best way to present my data or
  just text?
• 2. Which chart best fits my data and what I'm
  trying to show?
• 3. How do I make the chart clear, concise and
  interesting?

3
Two Rules of charting

• What is the message? If no message then no graph!
• Graphical excellence-gives viewer greatest number
  of ideas in the shortest time with the least
  amount of ink.
• The visual Display of Quantitative Information -
• Edward Tufte (Graphic Arts 1983)

4
Column Chart

• Compares one or more discrete variables over a
  period of time.
• Use a column chart to compare sales figures
  over several years.
• Errors??
• No x axis title
• No chart title

Value (Can x1000)
5
Stacked Column Chart

• Compare several items by plotting the values of
  each on the same chart either by stacking or
  by placing them side by side.
• Errors??

100
80
Electrical
60
Heat
40
Phone
20
0
Jan
Feb
Mar
6
Bar Charts

• Bar charts are column charts set horizontally.
  Bar charts often compare several items to a
  single variable at one time (or shows values
  distributed over a period of time )
• Errors on graph?
• No axis title
• Numbers too large
• Never end data at edge

7
Line

• Depicts trends over time, or even the
  distribution of one variable over another.
• Represents continuous rather than discrete
  series.
• Errors??

120
Supply
100
Consumption
80
60
Water Consumption (millions of L)
40
20
0
1991
1993
1995
1997
1999
2001
8
Combination
14000
Actual Sales
Overlays two chart types to compare two data
series For example Actual vs Predicted
sales Errors on graph?
12000
Predicted
10000
8000
6000
4000
2000
0
1991
1992
1993
1994
9
Double Y

• Compares combined trends in two sets of data
  measured in different units. Double-y graphs
  compare two trends that share one variable but
  not a second.
• Use a double-y graph to compare an engine's
  torque output and horsepower output over the
  same range of engine speeds.
• Errors?

Torque(ftlb)
Horsepower
RPM
10
Area

• Variation of a line chart
• Emphasizes quantity or volume.
• Adding multiple sets of data to the graph
  enables you to compare different sets of data
  over time or among categories.

11
Pie

• Compares relative parts that make up a whole.
• The pie represents the whole, and the size of
  each wedge shows that component's share.
• If you really want to make
• a pie and do not have 100
• of the data then use other
• or miscellaneous categories

12
Scatter

• An analytical tool to show the relationship of
  one variable to other variables. Chart plots
• x1,y1, x2,y2, x3,y3
• e.g. Display the relationship between yield and
  growing season precipitation.
• Note X data is variable
• not category

13
Polar

• Plot data points as described by their distance
  from the center of a circle (radius) and their
  relative angle in degrees from a reference
  point.
• e.g. Earthquake damage

14
PERT(Program Evaluation Review Technique)

• Used for Project Management
• Graphically represents the tasks, milestones,
  and dependencies involved in completing a
  project.

Widgets Inc.
15
Tree

• Illustrates hierarchical relationships.
• e.g. Corporate organizational chart, matches in
  a sports championship or genealogical histories.

16
Hi-Low

• Chart for reports on the stock market progress.
  
• Compare the high and low selling points of any
  given stock.
• Line charts are commonly used for plotting
  trends.

17
Error Bars

• Make chart more credible
• Beware - don't overuse and clutter!

18
Pictographs

• Add visual clues to nature of data

19
10 Worst Features inGraphic Presentations

• 10 Using Small Fat Serif Fonts for Labeling
• 9 Overdoing Fancy TypeFaces or Colors
• 8 Too many data points or items on the graph -
  clarify
• relationships, not obscure them!
• 7 Poor choice of axes and tick distribution
• 6 Over or under spaced

20
The 10 Worst Features...

• 5 Poor Titles and Labels
• 4 Using 3D graphs for 2D data
• 3 Choosing the wrong type of graph or chart
• 2 Displaying a chart when table or text would
  communicate more effectively
• 1 No Message!

21
All good charts have

• A title (single sentence, hanging indent)
  this is added in WORD if using DeltaGraph as
  charting program.
• Axes labeled clearly and easy to read (both or
  all with appropriate units)
• Axes values ( single numerals are easier to
  understand than hundreds, thousands etc.)
  change unit if necessary

22
All good charts have

• Legends are basic extra, redundant or
  extraneous information such as fancy graphics,
  text, clipart is eliminated.
• Axes ticks appropriate intervals.2s, 5s,
  10s Percent axes end at 100.
• Lines of line charts clearly defined few or no
  crossovers

23
All good charts have

• Right amount of data (avoid too few or too many
  columns, bars or lines)
• Helvetica font for labels and numbers (14 and
  bold for axes titles only)
• No solid black for fills use shades of gray and
  patterns
• Boxes, arrows, text labels used sparingly

24
Remember that less is moreavoid 3D technotype
graphing if data is not enhanced but rather
hidden by special effects.
25
Chart Analysis

• Can you identify the chart types?
• Are there errors in the charts and their titles?
• High impact vs textbook or print ready
• Good vs excellent charting (ie Does chart convey
  message immediately upon viewing?)

26
Value of various widget stocks in 1999 for Widget
Canada Inc.
Figure 1.
27
Figure 2.
Urban vs rural population in Saskatchewan.
28
The change in urban population and percent of
rural population in Saskatchewan over a 20 year
period.
Figure 2.
29
A comparison of the amount of land used for
agriculture and non-agriculture purposes in 1999.
Figure 3.
30
(No Transcript)
31
Remember.

• Check data table or figure?
• Choose correct chart type to convey your message.
• Print ready or presentation type?
• Extraneous information eliminated
• TITLE - key words - should stand alone