This reading focuses on presenting graphs in a way that accurately portrays the given information, rather than creating a misleading and/or overly complicated presentation. The article discusses six fundamental stages: “1. Determine your message and identify the necessary to communicate it. 2. Determine if a table, graph or combination of both is needed to communicate your message.” (3) If one or more graphs are required, then, “3. Determine the best means to encode the values. 4. Determine where to display each variable. 5. Determine the best design for the remaining objects. 6. Determine if particular data should be featured above the rest, and if so, how.” (3)
Concerning general concepts and practices, the article differentiates between tables versus graphs, quantitative versus categorical data, and common relationships in quantitative business data. “Tables work best when the display will be used to look up individual values or the quantitative values must be precise. Graphs work best when the message you wish to communicate resides in the shape of the data (in patterns, trends, and exceptions).” (4) Tables express data in text, while graphs express data graphically. Quantitative data consists of the numbers, while categorical data tells us what the quantitative data measures. Categorical scales have three fundamental types. “Nominal scales consist of discrete items that belong to a common category but…they differ in name only…Ordinal scales have intrinsic order… the items… do not represent quantitative values… Interval scales also consist of items that have an intrinsic order, but…they represent quantitative values as well… [It] is converted into a categorical scale by subdividing the full range of values into a sequential series of smaller ranges of equal size.” (5) Next, different relationships in quantitative business data was discussed. Time-Series relationships are, “when quantitative values are expressed as a series of measures taken across equal intervals of time.” (6). Ranking relationships are when quantitative values are organized by size. Part-to-Whole relationships are when quantitative values are presented to show the position that each value represents in relationship to some whole. Deviation relationships are, “when quantitative values are displayed to feature how one or more sets of values differ from some reference set of values.” (7) Distribution relationships show quantitative values pan out across the set’s entire range. Correlation relationships are, “when pairs of quantitative values , each measuring something different about an entity, are displayed to reveal if there is significant relationship between them.” (8) Lastly, nominal comparison relationships compares different values that may not have a particular relationship between them. I think that this section was helpful, because when making graphs I usually do not pay attention to what type of graph I use.
For encoding quantitative values in graphs, the best ways to display information are: points, lines, bars, and boxes. Points can, “be used to encode quantitative values along two quantitative scales simultaneously… and they can be used in place of bars when the quantitative scale does not begin at zero.” (11) Lines are useful when used with an interval scale. Lines connect values of a series and emphasize the shape of the data from value to value. Bars are useful when encoding data that needs to emphasize individual values by, “the 2-D position of the bar’s endpoint in relation to the quantitative scale, and the length of the bar.” (12) Boxes are similar to bars, but both ends encode quantitative values. Range bars are used to encode a range of values instead of a single value. The last part of this section emphasized removing distractions such as grid lines and some visual content. Misusing color can also be problematic by leading our brain to make connections that may not be there. This article recommends using soft color, low saturation, or colors found in nature are recommended. One should use bright or saturated colors when something needs to stand out.
First, it is important to determine the message of data one is trying to get across. Next, deciding whether a table, graph, or both would be the best way to communicate that message. After deciding the best way to encode the data, one must consider where to display each variable. Other design decisions also have to be made, which are concerned with visual appearance and placement. This would include ranges, legends, tick marks, location of scale, grid lines, and what descriptive text is needed to explain the message of the graph. Highlighting information with bight or dark colors or borders (for bars) are useful when necessary.