This is a reading note I made after reading the book.
Now You See It: Simple Visualization Techniques for Quantitative Analysis teaches simple, practical means to explore and analyze quantitative data--techniques that rely primarily on using your eyes. This book features graphical techniques that can be applied to a broad range of software tools, including Microsoft Excel, because so many people have nothing else, but also more powerful visual analysis tools that can dramatically extend your analytical reach. You'll learn to make sense of quantitative data by discerning the meaningful patterns, trends, relationships, and exceptions that measure your organization's performance, identify potential problems and opportunities, and reveal what will likely happen in the future. Now You See It is not just for those with "analyst" in their titles, but for everyone who's interested in discovering the stories in their data that reveal their organization's performance and how it can be improved.
5. Definition of Data Visualization
• Computer-supported
• Interactive
• Visual Representations
• Abstract Data
• Amplify Cognition
The purpose of information visualization is not to make
pictures, but to help us to think
8. Visual Perception
• We do not attend
to everything that
we see.
• Visual perception is
selective, at it must
be for awareness
of everything
would overwhelm
us.
• Our attention is
often drawn to
contrasts to the
norm.
• Memory plays
an important
role in human
cognition, but
working
memory is
extremely
limited.
• Our eyes are
drawn to
familiar
patterns.
• We see what
we know and
expect.
19. Effectiveness of Visualization
• Ability to clearly and accurately represent information
• Ability to interact with visualization to figure out what
the information is
25. Comparing
• Provide a selection of graphs that support the full
spectrum of commonly needed comparisons
• Provide graphs that are designed for easy comparison
of those values and relevant patterns without
distraction
• Provide the means to place a great deal of
information that we wish to compare on the screen at
the same time, thereby avoiding the need to scroll or
move from screen to screen
27. Sorting
• Provide the means to sort items in a graph based on
various values, especially the values that are featured
in the graph
• Provide extremely quick and easy means to re-sort
data in different ways, ideally with a single click of the
mouse
• Provide the means to link multiple graphs and easily
sort the data in each graph in the same way, assuming
that the graphs share a common categorical variable.
30. Adding Variables
• Provide convenient access to every available variable
that might be needed for analysis
• Provide easy means to add a variable to or remove
one from the display, such as by directly grabbing the
variable and placing it or removing
31. Filtering
• Easy filtering based on any information in the connected data
source not just based o information that is currently being
displayed
• Allow date to be filtered rapidly using simple controls, the lag time
between issuing the filter command and seeing the result should
be almost unnoticeable.
• Provide means to directly select items in a graph and then
remove them from display by single/two click
• Visible feedback on filter
• Complex filter with multiple conditions
• Filter multiple graph that linked together
34. Highlighting
• Provide the means to highlight a subset of data by selecting
from lists of categorical items.
• Provide the means to highlight a subset of data by directly
selecting it in a graph (mouse click, brush)
• Highlight selected information so that it can be seen
independently from the rest while still allowing viewers to se
the entire set of data
• Provide the means to highlights a set of items I none graph
and have those same items automatically highlighted in
other graphs that share the same dataset (link)
35. Aggregating
• Provide the means to easily aggregate the quantitative
data to the level of items In a categorical variable
• Provide the means to easily aggregate data in a number of
useful way, especially summing, averaging and counting
• Provide the means to easily aggregate data based on
equal intervals of a quantitative variable.
• Process the transition from one level of aggregation to
another without noticeable delay (Drill down/up)
• Ad Hoc Grouping
36. Drill
• Define hierarchical relationship among categorical
variables
• Drill down/up through hierarchy with no more than
one/two click
• Can skip levels
• Support nature hierarchies such as time
39. Re-Expressing
• Switch current unit of measure to percentage
• Re-express values in terms of how they compare to a
reference value or as a rolling value
40. Re-visualizing
• Easily and Rapidly switch from one type to another
• List the available graph types that are appropriate for
current data
• Prevent or make more difficult the selection of the
graph that would display the data inappropriately
42. Zooming and Pan
• Directly select an area of a graph and then zoom into
it with a single click
• Zoom back
• Pan when some portion of the graph is out of the view
44. Re-scaling
• Change the quantitative scale from linear to
logarithmic
• Set log scale’s base
• Set starting and ending value for the scale
• Prevent or make inconvenient the use of log scale for
bar and box plot
45. Accessing Details on Demand
• View details related to
an item in a visualization
when needed, in form of
text
• Make details disappear
when it is no longer
required
46. Annotating
• Add notes to a visualization
so that they are associated
with the visualization as a
whole, a particular region,
or one or more particular
value
• The note should reposition
to the associated data
value
47. Bookmarking
• Save the state of an analysis for later access without
interrupting the flow of analysis
• Maintain a history of the steps and states during the
analytical process
50. Techniques and practices
• Optimal quantitative scales
• Reference lines and regions
• Trellises and crosstabs
• Multiple concurrent views and brushing
• Focus and context together
• Details on demand
• Over-plotting reduction
51. Optimal Quantitative Scales
• When using a bar graph, begin
the scale at zero and end at
the scale a little above the
highest value
• With every type of graph other
than a bar graph, begin the
scale a little below the lowest
value and end it a little above
the highest value
• Begin the end the scale at
round numbers, and make the
intervals round number as well.
53. Reference Line and Region
• Add reference line based on a specific value and ad hoc
calculation or statistical calculation
• Automated calculations for : mean, median, standard
deviation, specific percentiles, minimum and maximum
• Reference line based on the values that appear in the
graph only or on a larger set of value
• Label the reference lines to clearly indicate what the lines
represent
• Format the reference line as needed (hue, color intensity,
line weight, line styles etc)
64. Details on Demand
• Control in
tooltips
• Information for
multiple
selected data
points
65. Over-plotting Reduction
• Reduce the size of data
objects
• Remove fill color from data
objects
• Changing the shape of data
objects
• Jittering data objects
• Making data objects
transparent
• Encoding the density of values
• Reducing the number of values