📉Graphical Representation: Cluster (Just the Quotes)

"To the untrained eye, randomness appears as regularity or tendency to cluster." (William Feller, "An Introduction to Probability Theory and its Applications", 1950) 

"Sometimes clusters of variables tend to vary together in the normal course of events, thereby rendering it difficult to discover the magnitude of the independent effects of the different variables in the cluster. And yet it may be most desirable, from a practical as well as scientific point of view, to disentangle correlated describing variables in order to discover more effective policies to improve conditions. Many economic indicators tend to move together in response to underlying economic and political events." (Edward R Tufte, "Data Analysis for Politics and Policy", 1974)

"The logarithmic transformation serves several purposes: (1) The resulting regression coefficients sometimes have a more useful theoretical interpretation compared to a regression based on unlogged variables. (2) Badly skewed distributions - in which many of the observations are clustered together combined with a few outlying values on the scale of measurement - are transformed by taking the logarithm of the measurements so that the clustered values are spread out and the large values pulled in more toward the middle of the distribution. (3) Some of the assumptions underlying the regression model and the associated significance tests are better met when the logarithm of the measured variables is taken." (Edward R Tufte, "Data Analysis for Politics and Policy", 1974)

"The scatterplot is a useful exploratory method for providing a first look at bivariate data to see how they are distributed throughout the plane, for example, to see clusters of points, outliers, and so forth." (William S Cleveland, "Visualizing Data", 1993)

"Multivariate techniques often summarize or classify many variables to only a few groups or factors (e.g., cluster analysis or multi-dimensional scaling). Parallel coordinate plots can help to investigate the influence of a single variable or a group of variables on the result of a multivariate procedure. Plotting the input variables in a parallel coordinate plot and selecting the features of interest of the multivariate procedure will show the influence of different input variables." (Martin Theus & Simon Urbanek, "Interactive Graphics for Data Analysis: Principles and Examples", 2009)

"Parallel coordinate plots are often overrated concerning their ability to depict multivariate features. Scatterplots are clearly superior in investigating the relationship between two continuous variables and multivariate outliers do not necessarily stick out in a parallel coordinate plot. Nonetheless, parallel coordinate plots can help to find and understand features such as groups/clusters, outliers and multivariate structures in their multivariate context. The key feature is the ability to select and highlight individual cases or groups in the data, and compare them to other groups or the rest of the data." (Martin Theus & Simon Urbanek, "Interactive Graphics for Data Analysis: Principles and Examples", 2009) 

"Be careful not to confuse clustering and stratification. Even though both of these sampling strategies involve dividing the population into subgroups, both the way in which the subgroups are sampled and the optimal strategy for creating the subgroups are different. In stratified sampling, we sample from every stratum, whereas in cluster sampling, we include only selected whole clusters in the sample. Because of this difference, to increase the chance of obtaining a sample that is representative of the population, we want to create homogeneous groups for strata and heterogeneous (reflecting the variability in the population) groups for clusters." (Roxy Peck et al, "Introduction to Statistics and Data Analysis" 4th Ed., 2012)

"Linking is a powerful dynamic interactive graphics technique that can help us better understand high-dimensional data. This technique works in the following way: When several plots are linked, selecting an observation's point in a plot will do more than highlight the observation in the plot we are interacting with - it will also highlight points in other plots with which it is linked, giving us a more complete idea of its value across all the variables. Selecting is done interactively with a pointing device. The point selected, and corresponding points in the other linked plots, are highlighted simultaneously. Thus, we can select a cluster of points in one plot and see if it corresponds to a cluster in any other plot, enabling us to investigate the high-dimensional shape and density of the cluster of points, and permitting us to investigate the structure of the disease space." (Forrest W Young et al, "Visual Statistics: Seeing data with dynamic interactive graphics", 2016)

"Dimensionality reduction is a way of reducing a large number of different measures into a smaller set of metrics. The intent is that the reduced metrics are a simpler description of the complex space that retains most of the meaning. […] Clustering techniques are similarly useful for reducing a large number of items into a smaller set of groups. A clustering technique finds groups of items that are logically near each other and gathers them together." (Danyel Fisher & Miriah Meyer, "Making Data Visual", 2018)

"Important features to look for in a scatter plot are whether there is one cloud of dots or several clusters, whether there is an upward or downward slope to the cloud of dots, and whether there is any curvature to the slope." (Robert Grant, "Data Visualization: Charts, Maps and Interactive Graphics", 2019)

"[...] scatterplots had advantages over earlier graphic forms: the ability to see clusters, patterns, trends, and relations in a cloud of points. Perhaps most importantly, it allowed the addition of visual annotations (point symbols, lines, curves, enclosing contours, etc.) to make those relationships more coherent and tell more nuanced stories." (Michael Friendly & Howard Wainer, "A History of Data Visualization and Graphic Communication", 2021)

📉Graphical Representation: Similarity (Just the Quotes)

"Two important characteristics of maps should be noticed. A map is not the territory it represents, but, if correct, it has a similar structure to the territory, which accounts for its usefulness. If the map could be ideally correct, it would include, in a reduced scale, the map of the map; the map of the map, of the map [...]" (Alfred Korzybski, "Science and Sanity: An Introduction to Non-Aristotelian Systems and General Semantics", 1933)

"Some believe that the vertical bar should be used when comparing similar items for different time periods and the horizontal bar for comparing different items for the same time period. However, most people find the vertical-bar format easier to prepare and read. and a more effective way to show most types of comparisons." (Robert Lefferts, "Elements of Graphics: How to prepare charts and graphs for effective reports", 1981)

"In order to be easily understood, a display of information must have a logical structure which is appropriate for the user's knowledge and needs, and this structure must be clearly represented visually. In order to indicate structure, it is necessary to be able to eemphasiz, divide and relate items of information. Visual emphasis can be used to indicate a hierarchical relationship between items of information, as in the case of systems of headings and subheadings for example. Visual separation of items can be used to indicate that they are different in kind or are unrelated functionally, and similarly a visual relationship between items will imply that they are of a similar kind or bear some functional relation to one another. This kind of visual 'coding' helps the reader to appreciate the extent and nature of the relationship between items of information, and to adopt an appropriate scanning strategy." (Linda Reynolds & Doig Simmonds, "Presentation of Data in Science" 4th Ed, 1984)

"Scatter charts show the relationships between information, plotted as points on a grid. These groupings can portray general features of the source data, and are useful for showing where correlationships occur frequently. Some scatter charts connect points of equal value to produce areas within the grid which consist of similar features." (Bruce Robertson, "How to Draw Charts & Diagrams", 1988)

"Two types of graphic organizers are commonly used for comparison: the Venn diagram and the comparison matrix [...] the Venn diagram provides students with a visual display of the similarities and differences between two items. The similarities between elements are listed in the intersection between the two circles. The differences are listed in the parts of each circle that do not intersect. Ideally, a new Venn diagram should be completed for each characteristic so that students can easily see how similar and different the elements are for each characteristic used in the comparison." (Robert J. Marzano et al, "Classroom Instruction that Works: Research-based strategies for increasing student achievement, 2001)

"Arbitrary category sequence and misplaced pie chart emphasis lead to general confusion and weaken messages. Although this can be used for quite deliberate and targeted deceit, manipulation of the category axis only really comes into its own with techniques that bend the relationship between the data and the optics in a more calculated way. Many of these techniques are just twins of similar ruses on the value axis. but are none the less powerful for that." (Nicholas Strange, "Smoke and Mirrors: How to bend facts and figures to your advantage", 2007)

"We tend automatically to think of all the categories represented on the horizontal axis of a column Chart as being equally important. They vary of course on the value axis. Otherwise, there would be little point in the chart, but there is somehow this feeling that they are in other respects similar members of a group. This convention can be put to good use to manipulate the message of the most boring bar or column chart." (Nicholas Strange, "Smoke and Mirrors: How to bend facts and figures to your advantage", 2007)

"For a given dataset there is not a great deal of advice which can be given on content and context. Those who know their own data should know best for their specific purposes. It is advisable to think hard about what should be shown and to check with others if the graphic makes the desired impression. Design should be let to designers, though some basic guidelines should be followed: consistency is important (sets of graphics should be in similar style and use equivalent scaling); proximity is helpful (place graphics on the same page, or on the facing page, of any text that refers to them); and layout should be checked" (graphics should be neither too small nor too large and be attractively positioned relative to the whole page or display)." (Antony Unwin, "Good Graphics?" [in "Handbook of Data Visualization"], 2008)

"A histogram for discrete numerical data is a graph of the frequency or relative frequency distribution, and it is similar to the bar chart for categorical data. Each frequency or relative frequency is represented by a rectangle centered over the corresponding value" (or range of values) and the area of the rectangle is proportional to the corresponding frequency or relative frequency." (Roxy Peck et al, "Introduction to Statistics and Data Analysis" 4th Ed., 2012)

"With further similarities to small multiples, heatmaps enable us to perform rapid pattern matching to detect the order and hierarchy of different quantitative values across a matrix of categorical combinations. The use of a color scheme with decreasing saturation or increasing lightness helps create the sense of data magnitude ranking." (Andy Kirk, "Data Visualization: A successful design process", 2012)

"Tree maps are similar to pie charts in that they show parts of a whole but, unlike pie charts, they can incorporate more individual pieces without cluttering the graphic. Tree maps are particularly good at presenting information like budgets, which often include more elements than can be effectively communicated through a pie chart." (Christopher Lysy, "Developments in Quantitative Data Display and Their Implications for Evaluation", 2013)

"Upon discovering a visual image, the brain analyzes it in terms of primitive shapes and colors. Next, unity contours and connections are formed. As well, distinct variations are segmented. Finally, the mind attracts active attention to the significant things it found. That process is permanently running to react to similarities and dissimilarities in shapes, positions, rhythms, colors, and behavior. It can reveal patterns and pattern-violations among the hundreds of data values. That natural ability is the most important thing used in diagramming." (Vasily Pantyukhin, "Principles of Design Diagramming", 2015)

"A histogram represents the frequency distribution of the data. Histograms are similar to bar charts but group numbers into ranges. Also, a histogram lets you show the frequency distribution of continuous data. This helps in analyzing the distribution" (for example, normal or Gaussian), any outliers present in the data, and skewness." (Umesh R Hodeghatta & Umesha Nayak, "Business Analytics Using R: A Practical Approach", 2017)

"A taxonomy is a classification scheme that organizes categories in a broader-narrower hierarchy. Items that share similar qualities are grouped into the same category, and the taxonomy provides a global organization by relating categories to one another." (Jesús Barrasa et al, "Knowledge Graphs: Data in Context for Responsive Businesses", 2021)

"A well-designed dashboard needs to provide a similar experience; information cannot be placed just anywhere on the dashboard. Charts that relate to one another are usually positioned close to one another. Important charts often appear larger and more visually prominent than less important ones. In other words, there are natural sizes for how a dashboard comprises charts based on the task and context." (Vidya Setlur & Bridget Cogley, "Functional Aesthetics for data visualization", 2022)

"Unlike text, visual communication is governed less by an agreed-upon convention between 'writer' and 'reader' than by how our visual systems react to stimuli, often before we’re aware of it. And just as composers use music theory to create music that produces certain predictable effects on an audience, chart makers can use visual perception theory to make more-effective visualizations with similarly predictable effects." (Scott Berinato, "Good Charts : the HBR guide to making smarter, more persuasive data visualizations", 2023)

See also the quotes on Similarity in Systems Engineering

📉Graphical Representation: Executives (Just the Quotes)

"An exact knowledge of conditions, and consequent timely application of praise or of constructive criticism, is one of the chief forces of the executive in securing satisfactory results. Undeserved criticism is unjust and destroys-initiative, while unmerited praise tends to render the executive ridiculous in the eyes of his subordinates; both are detrimental to discipline and weaken the power of the executive." (Allan C Haskell, "How to Make and Use Graphic Charts", 1919)

"The problem of the executive, then - once his organization is perfected - is to secure live data covering the exact conditions of the business at all times. These data should be arranged so as to give him all the facts, subordinated according to their relative bearing upon net earnings, and do so with the least demand upon his time. Furthermore, these facts must be so exhibited that the general laws underlying the business may be easily and accurately deduced and standards of accomplishment set which will be a continual incentive to greater accomplishment." (Allan C Haskell, "How to Make and Use Graphic Charts", 1919)

"Wherever unusual peaks or valleys occur on a curve it is a good plan to mark these points with a small figure inside a circle. This figure should refer to a note on the back of the chart explaining the reason for the unusual condition. It is not always sufficient to show that a certain item is unusually high or low; the executive will want to know why it is that way." (Allan C Haskell, "How to Make and Use Graphic Charts", 1919

"An economic justification for computer graphics is that the organization spends an enormous amount of money on data processing, often providing managers with too many reports, too many data, and an overload of information. The report output has to be condensed into a more usable form. The computer graph essentially is the data represented in a structured pictorial form. The role of the graph is to provide meaningful reports. To the extent that it does. it can be justified." (Anker V Andersen, "Graphing Financial Information: How accountants can use graphs to communicate", 1983)

"Without meaningful data there can be no meaningful analysis. The interpretation of any data set must be based upon the context of those data. Unfortunately, much of the data reported to executives today are aggregated and summed over so many different operating units and processes that they cannot be said to have any context except a historical one - they were all collected during the same time period. While this may be rational with monetary figures, it can be devastating to other types of data." (Donald J Wheeler, "Understanding Variation: The Key to Managing Chaos" 2nd Ed., 2000)

"Organizations face challenges of all kinds after activating their new systems. To be sure, these challenges are typically not as significant as those associated with going live. Still, executives and end users should never assume that system activation means that everyone is home free. Systems are hardly self-sufficient, and issues always appear." (Phil Simon, "Why New Systems Fail: An Insider’s Guide to Successful IT Projects", 2010)

Most discussions of decision making assume that only senior executives make decisions or that only senior executives' decisions matter. This is a dangerous mistake. Decisions are made at every level of the organization, beginning with individual professional contributors and frontline supervisors. These apparently low-level decisions are extremely important in a knowledge-based organization." (Zach Gemignani et al, "Data Fluency", 2014)

Along with the important information that executives need to be data literate, there is one other key role they play: executives drive data literacy learning and initiatives at the organization." (Jordan Morrow, "Be Data Literate: The data literacy skills everyone needs to succeed", 2021)

📉Graphical Representation: Beauty (Just the Quotes)

"Nearly all those who produce graphics for mass publication are trained exclusively in the fine arts and have had little experience with the analysis of data. Such experiences are essential for achieving precision and grace in the presence of statistics. [...] Those who get ahead are those who beautified data, never mind statistical integrity." (Edward R Tufte, "The Visual Display of Quantitative Information", 1983)

"Maximizing data ink (within reason) is but a single dimension of a complex and multivariate design task. The principle helps conduct experiments in graphical design. Some of those experiments will succeed. There remain, however, many other considerations in the design of statistical graphics - not only of efficiency, but also of complexity, structure, density, and even beauty." (Edward R Tufte, "Data-Ink Maximization and Graphical Design", Oikos Vol. 58 (2), 1990)

"The main goal of data visualization is its ability to visualize data, communicating information clearly and effectively. It doesn’t mean that data visualization needs to look boring to be functional or extremely sophisticated to look beautiful. To convey ideas effectively, both aesthetic form and functionality need to go hand in hand, providing insights into a rather sparse and complex dataset by communicating its key aspects in a more intuitive way. Yet designers often tend to discard the balance between design and function, creating gorgeous data visualizations which fail to serve its main purpose - communicate information." (Vitaly Friedman, "Data Visualization and Infographics", Smashing Magazine, 2008)

"A beautiful visualization has a clear goal, a message, or a particular perspective on the information that it is designed to convey. Access to this information should be as straightforward as possible, without sacrificing any necessary, relevant complexity. [...] Most importantly, beautiful visualizations reflect the qualities of the data that they represent, explicitly revealing properties and relationships inherent and implicit in the source data. As these properties and relationships become available to the reader, they bring new knowledge, insight, and enjoyment."  (Noah Iliinsky, "On Beauty", [in "Beautiful Visualization"] 2010)

"For a visual to qualify as beautiful, it must be aesthetically pleasing, yes, but it must also be novel, informative, and efficient. [...] For a visual to truly be beautiful, it must go beyond merely being a conduit for information and offer some novelty: a fresh look at the data or a format that gives readers a spark of excitement and results in a new level of understanding. Well-understood formats" (e.g., scatterplots) may be accessible and effective, but for the most part they no longer have the ability to surprise or delight us. Most often, designs that delight us do so not because they were designed to be novel, but because they were designed to be effective; their novelty is a byproduct of effectively revealing some new insight about the world." (Noah Iliinsky, "On Beauty", [in "Beautiful Visualization"] 2010)

"The first requirement of a beautiful visualization is that it is novel, fresh, or unique. It is difficult (though not impossible) to achieve the necessary novelty using default formats. In most situations, well-defined formats have well-defined, rational conventions of use: line graphs for continuous data, bar graphs for discrete data, pie graphs for when you are more interested in a pretty picture than conveying knowledge." (Noah Iliinsky, "On Beauty", [in "Beautiful Visualization"] 2010)

"The key to the success of any visual, beautiful or not, is providing access to information so that the user may gain knowledge. A visual that does not achieve this goal has failed. Because it is the most important factor in determining overall success, the ability to convey information must be the primary driver of the design of a visual." (Noah Iliinsky, "On Beauty", [in "Beautiful Visualization"] 2010)

"[...] visual art, primarily serves the relationship between the designer and the data. [...] it often entails unidirectional encoding of information, meaning that the reader may not be able to decode the visual presentation to understand the underlying information. [...] visual art merely translates the data into a visual form. The designer may intend only to condense it, translate it into a new medium, or make it beautiful; she may not intend for the reader to be able to extract anything from it other than enjoyment." (Noah Iliinsky & Julie Steel, "Designing Data Visualizations", 2011)

"Graphics, charts, and maps aren’t just tools to be seen, but to be read and scrutinized. The first goal of an infographic is not to be beautiful just for the sake of eye appeal, but, above all, to be understandable first, and beautiful after that; or to be beautiful thanks to its exquisite functionality." (Alberto Cairo, "The Functional Art", 2011)

"The biggest thing to know is that data visualization is hard. Really difficult to pull off well. It requires harmonization of several skills sets and ways of thinking: conceptual, analytic, statistical, graphic design, programmatic, interface-design, story-telling, journalism - plus a bit of ‘gut feel.’ The end result is often simple and beautiful, but the process itself is usually challenging and messy." (David McCandless, 2013)

"Aesthetics plays an important part in the process of visualization. One dimension of aesthetics is in terms of beauty, elegance and desirability. These are important attributes for the rhetorical function of visualization." (Peter A Hall & Patricio Dávila, "Critical Visualization: Rethinking the Representation of Data", 2022)

"A perfectly relevant visualization that breaks a few presentation rules is far more valuable - it’s better - than a perfectly executed, beautiful chart that contains the wrong data, communicates the wrong message, or fails to engage its audience." (Scott Berinato, "Good Charts : the HBR guide to making smarter, more persuasive data visualizations", 2023)

"As beautiful as data can be, it’s not an al fresco painting that should be open to interpretation from anyone who walks by its section of the museum. Make bold, smart color choices that leave no doubt what the purpose of the data is." (Kate Strachnyi, "ColorWise: A Data Storyteller’s Guide to the Intentional Use of Color", 2023)

"When deeply complex charts work, we find them effective and beautiful, just as we find a symphony beautiful, which is another marvelously complex arrangement of millions of data points that we experience as a coherent whole." (Scott Berinato, "Good Charts : the HBR guide to making smarter, more persuasive data visualizations", 2023)

📉Graphical Representation: Good Design (Just the Quotes)

"Good design looks right. It is simple (clear and uncomplicated). Good design is also elegant, and does not look contrived. A map should be aesthetically pleasing, thought provoking, and communicative."  (Arthur H Robinson, "Elements of Cartography", 1953)

"Without adequate planning. it is seldom possible to achieve either proper emphasis of each component element within the chart or a presentation that is pleasing in its entirely. Too often charts are developed around a single detail without sufficient regard for the work as a whole. Good chart design requires consideration of these four major factors: (1) size, (2) proportion, (3) position and margins, and (4) composition." (Anna C Rogers, "Graphic Charts Handbook", 1961)

"Because ease of use is the purpose, this ratio of function to conceptual complexity is the ultimate test of system design. Neither function alone nor simplicity alone defines a good design. [...] Function, and not simplicity, has always been the measure of excellence for its designers." (Fred P Brooks, "The Mythical Man-Month: Essays", 1975)

"Good design protects you from the need for too many highly accurate components in the system. But such design principles are still, to this date, ill-understood and need to be researched extensively. Not that good designers do not understand this intuitively, merely it is not easily incorporated into the design methods you were taught in school. Good minds are still needed in spite of all the computing tools we have developed." (Richard Hamming, "The Art of Doing Science and Engineering: Learning to Learn", 1997)

"Good design, however, can dispose of clutter and show all the data points and their names. [...] Clutter calls for a design solution, not a content reduction." (Edward R Tufte, "Beautiful Evidence", 2006)

"Good design is an important part of any visualization, while decoration (or chart-junk) is best omitted. Statisticians should also be careful about comparing themselves to artists and designers; our goals are so different that we will fare poorly in comparison." (Hadley Wickham, "Graphical Criticism: Some Historical Notes", Journal of Computational and Graphical Statistics Vol. 22(1), 2013) 

"In the field of design, experts speak of objects having 'affordances'. These are aspects inherent to the design that make it obvious how the product is to be used. For example, a knob affords turning, a button affords pushing, and a cord affords pulling. These characteristics suggest how the object is to be interacted with or operated. When sufficient affordances are present, good design fades into the background and you don’t even notice it." (Cole N Knaflic, "Storytelling with Data: A Data Visualization Guide for Business Professionals", 2015)

"For every rule in data visualization, there is a scenario where that rule should be broken. This means that choosing the best chart or the best design is always a trade-off between several conflicting goals. Our imperfect perception means that data visualization has a larger subjective dimension than a data table. Sometimes we only need this subjective, impressionist dimension and other times we need to translate it into hard figures. Striving for accuracy is important, but it’s more important to provide those insights that only a visual display can reveal." (Jorge Camões, "Data at Work: Best practices for creating effective charts and information graphics in Microsoft Excel", 2016)

"A chart that knows its context well will naturally end up looking better because it’s showing what it needs to show and nothing else. Good context begets good design. Good charts are only the means to a more profound end: presenting your ideas effectively. Good charts are not the product you’re after. They’re the way to deliver your product - insight." (Scott Berinato, "Good Charts : the HBR guide to making smarter, more persuasive data visualizations", 2023)

"Good design isn’t just choosing colors and fonts or coming up with an aesthetic for charts. That’s styling - part of design, but by no means the most important part. Rather, people with design talent develop and execute systems for effective visual communication. They understand how to create and edit visuals to focus an audience and distill ideas." (Scott Berinato, "Good Charts : the HBR guide to making smarter, more persuasive data visualizations", 2023)

"Good design serves a more important function than simply pleasing you: It helps you access ideas. It improves your comprehension and makes the ideas more persuasive. Good design makes lesser charts good and good charts transcendent." (Scott Berinato, "Good Charts : the HBR guide to making smarter, more persuasive data visualizations", 2023)

"Graphic design is not just about making things look good. It is a powerful combination of form and function that uses visual elements to communicate a message. Form refers to the physical appearance of a design, such as its shape, color, and typography. Function refers to the purpose of a design, such as what it is trying to communicate or achieve. A good graphic design is both visually appealing and functional. It uses the right combination of form and function to communicate its message effectively. Graphic design is also a strategic and thoughtful craft. It requires careful planning and execution to create a design that is both effective and aesthetically pleasing." (Faith Aderemi, "The Essential Graphic Design Handbook", 2024)

📉Graphical Representation: Graphs (Just the Quotes)

"Graphs are all inclusive. No fact is too slight or too great to plot to a scale suited to the eye. Graphs may record the path of an ion or the orbit of the sun, the rise of a civilization, or the acceleration of a bullet, the climate of a century or the varying pressure of a heart beat, the growth of a business, or the nerve reactions of a child." (Henry D Hubbard [foreword to Willard C Brinton, "Graphic Presentation", 1939)])

"Graphs carry the message home. A universal language, graphs convey information directly to the mind. Without complexity there is imaged to the eye a magnitude to be remembered. Words have wings, but graphs interpret. Graphs are pure quantity, stripped of verbal sham, reduced to dimension, vivid, unescapable." (Henry D Hubbard [foreword to Willard C Brinton, "Graphic Presentation", 1939]) 

"The graphic language is modern. We are learning its alphabet. That it will develop a lexicon and a literature marvelous for its vividness and the variety of application is inevitable. Graphs are dynamic, dramatic. They may epitomize an epoch, each dot a fact, each slope an event, each curve a history. Wherever there are data to record, inferences to draw, or facts to tell, graphs furnish the unrivalled means whose power we are just beginning to realize and to apply."  (Henry D Hubbard [foreword to Willard C Brinton, "Graphic Presentation", 1939)])

"If one technique of data analysis were to be exalted above all others for its ability to be revealing to the mind in connection with each of many different models, there is little doubt which one would be chosen. The simple graph has brought more information to the data analyst’s mind than any other device. It specializes in providing indications of unexpected phenomena." (John W Tukey, "The Future of Data Analysis", Annals of Mathematical Statistics Vol. 33 (1), 1962)

"There is no more reason to expect one graph to ‘tell all’ than to expect one number to do the same." (John W Tukey, "Exploratory Data Analysis", 1977)

"[...] exploratory data analysis is an attitude, a state of flexibility, a willingness to look for those things that we believe are not there, as well as for those we believe might be there. Except for its emphasis on graphs, its tools are secondary to its purpose." (John W Tukey, [comment] 1979)

"We would wish ‘numerate’ to imply the possession of two attributes. The first of these is an ‘at-homeness’ with numbers and an ability to make use of mathematical skills which enable an individual to cope with the practical mathematical demands of his everyday life. The second is ability to have some appreciation and understanding of information which is presented in mathematical terms, for instance in graphs, charts or tables or by reference to percentage increase or decrease." (Cockcroft Committee, "Mathematics Counts: A Report into the Teaching of Mathematics in Schools", 1982)

"We would wish ‘numerate’ to imply the possession of two attributes. The first of these is an ‘at-homeness’ with numbers and an ability to make use of mathematical skills which enable an individual to cope with the practical mathematical demands of his everyday life. The second is ability to have some appreciation and understanding of information which is presented in mathematical terms, for instance in graphs, charts or tables or by reference to percentage increase or decrease." (Cockcroft Committee, "Mathematics Counts: A Report into the Teaching of Mathematics in Schools", 1982)

"Iteration and experimentation are important for all of data analysis, including graphical data display. In many cases when we make a graph it is immediately clear that some aspect is inadequate and we regraph the data. In many other cases we make a graph, and all is well, but we get an idea for studying the data in a different way with a different graph; one successful graph often suggests another." (William S Cleveland, "The Elements of Graphing Data", 1985)

"There are some who argue that a graph is a success only if the important information in the data can be seen within a few seconds. While there is a place for rapidly-understood graphs, it is too limiting to make speed a requirement in science and technology, where the use of graphs ranges from, detailed, in-depth data analysis to quick presentation." (William S Cleveland, "The Elements of Graphing Data", 1985)

"A first analysis of experimental results should, I believe, invariably be conducted using flexible data analytical techniques – looking at graphs and simple statistics – that so far as possible allow the data to ‘speak for themselves’. The unexpected phenomena that such a approach often uncovers can be of the greatest importance in shaping and sometimes redirecting the course of an ongoing investigation." (George Box, "Signal to Noise Ratios, Performance Criteria, and Transformations", Technometrics 30, 1988) 

"We are not saying that the primary purpose of a graph is to convey numbers with as many decimal places as possible. We agree with Ehrenberg (1975) that if this were the only goal, tables would be better. The power of a graph is its ability to enable one to take in the quantitative information, organize it, and see patterns and structure not readily revealed by other means of studying the data." (William Cleveland & Robert McGill, "Graphical Perception: Theory, Experimentation, and Application to the Development of Graphical Models", Journal of the American Statistical Association 79, 1984)

"It’s not easy to select more than a few clearly distinct colors. Also, 'distinct' is context-dependent, because: What will be the spatial relationships of the different colors in your output? You can successfully have fairly similar colors adjacent to each other, since the contrast is more obvious when they’re adjacent. However, if you want to use colors to track identity and difference across scattered points or patches, then you need bigger separations between colors, since you want to be able to see easily that patch 'A' here is of the same kind as patch 'A' there and different from patch 'B' somewhere else, when mingled with patches of other kinds. And size matters. Big patches of similar color (as on a map) can look quite distinct, while the same colors used to plot filled circular blobs on a graph might be barely distinguishable, and totally indistinguishable if used to plot colored '.'s or '+'s. [...] It’s all very psycho-visual and success usually requires experimentation!" (Ted Harding, R-help mailing list, 2004)

📉Graphical Representation: Slopes (Just the Quotes)

"If a chart contains a number of series which vary widely in individual magnitude, optical distortion may result from the necessarily sharp changes in the angle of the curves. The space between steeply rising or falling curves always appears narrower than the vertical distance between the plotting points." (Rufus R Lutz, "Graphic Presentation Simplified", 1949)

"Besides being easier to construct than a bar chart, the line chart possesses other advantages. It is easier to read, for while the bars stand out more prominently than the line, they tend to become confusing if numerous, and especially so when they record alternate increase and decrease. It is easier for the eye to follow a line across the face of the chart than to jump from bar top to bar top, and the slope of the line connecting two points is a great aid in detecting minor changes. The line is also more suggestive of movement than arc bars, and movement is the very essence of a time series. Again, a line chart permits showing two or more related variables on the same chart, or the same variable over two or more corresponding periods." (Walter E Weld, "How to Chart; Facts from Figures with Graphs", 1959)

"When approximations are all that are needed, stacked area graphs are usually adequate. When accuracy is desired, this type of graph is generally not used, particularly when the values fluctuate significantly and/or the slopes of the curves are steep." (Robert L Harris, "Information Graphics: A Comprehensive Illustrated Reference", 1996)

"In general. statistical graphics should be moderately greater in length than in height. And, as William Cleveland discovered, for judging slopes and velocities up and down the hills in time-series, best is an aspect ratio that yields hill - slopes averaging 45°, over every cycle in the time-series. Variations in slopes are best detected when the slopes are around 45°, uphill or downhill." (Edward R Tufte, "Beautiful Evidence", 2006)

"For linear dependences the main information usually lies in the slope. It is obvious that those points that lie far apart have the strongest influence on the slope if all points have the same uncertainty. In this context we speak of the strong leverage of distant points; when determining the parameter 'slope' these distant points carry more effective weight. Naturally, this weight is distinct from the 'statistical' weight usually used in regression analysis." (Manfred Drosg, "Dealing with Uncertainties: A Guide to Error Analysis", 2007)

"Diagrams furnish only approximate information. They do not add anything to the meaning of the data and, therefore, are not of much use to a statistician or research worker for further mathematical treatment or statistical analysis. On the other hand, graphs are more obvious, precise and accurate than the diagrams and are quite helpful to the statistician for the study of slopes, rates of change and estimation, (interpolation and extrapolation), wherever possible." (S C Gupta & Indra Gupta, "Business Statistics", 2013)

"The term shrinkage is used in regression modeling to denote two ideas. The first meaning relates to the slope of a calibration plot, which is a plot of observed responses against predicted responses. When a dataset is used to fit the model parameters as well as to obtain the calibration plot, the usual estimation process will force the slope of observed versus predicted values to be one. When, however, parameter estimates are derived from one dataset and then applied to predict outcomes on an independent dataset, overfitting will cause the slope of the calibration plot (i.e., the shrinkage factor ) to be less than one, a result of regression to the mean. Typically, low predictions will be too low and high predictions too high. Predictions near the mean predicted value will usually be quite accurate. The second meaning of shrinkage is a statistical estimation method that preshrinks regression coefficients towards zero so that the calibration plot for new data will not need shrinkage as its calibration slope will be one." (Frank E. Harrell Jr., "Regression Modeling Strategies: With Applications to Linear Models, Logistic and Ordinal Regression, and Survival Analysis" 2nd Ed, 2015)

"The idiom of scatterplots encodes two quantitative value variables using both the vertical and horizontal spatial position channels, and the mark type is necessarily a point. Scatterplots are effective for the abstract tasks of providing overviews and characterizing distributions, and specifically for finding outliers and extreme values. Scatterplots are also highly effective for the abstract task of judging the correlation between two attributes. With this visual encoding, that task corresponds the easy perceptual judgement of noticing whether the points form a line along the diagonal. The stronger the correlation, the closer the points fall along a perfect diagonal line; positive correlation is an upward slope, and negative is downward." (Tamara Munzner, "Visualization Analysis and Design", 2014)

"Important features to look for in a scatter plot are whether there is one cloud of dots or several clusters, whether there is an upward or downward slope to the cloud of dots, and whether there is any curvature to the slope." (Robert Grant, "Data Visualization: Charts, Maps and Interactive Graphics", 2019)

"In the case of the straight line, we can talk about the slope of the line, which defines how much the outcome changes if there is a change in the predictor. A steeper slope means a stronger 'effect'. We can simply show the slope, which is quantified by a regression coefficient, as a marker with error bars for its confidence interval. A positive slope indicates that observations with higher values of the predictor usually also have higher values of the outcome too. A negative slope means the opposite." (Robert Grant, "Data Visualization: Charts, Maps and Interactive Graphics", 2019)

"Researchers have studied how accurately people can read information displayed in different types of plots. They have found the following ordering, from most to leasta ccurately judged (•) Positions along a common scale, like in a rug plot, strip plot, or dot plot (•) Positions on identical, nonaligned scales, like in a bar plot (•) Length, like in a stacked bar plot (•) Angle and slope, like in a pie chart (•) Area, like in a stacked line plot or bubble chart (•) Volume and density, like in a three-dimensional bar plot (•) Color saturation and hue, like when overplotting with semitransparent points."  (Sam Lau et al, "Learning Data Science: Data Wrangling, Exploration, Visualization, and Modeling with Python", 2023)

📉Graphical Representation: Assumptions (Just the Quotes)

"Logging size transforms the original skewed distribution into a more symmetrical one by pulling in the long right tail of the distribution toward the mean. The short left tail is, in addition, stretched. The shift toward symmetrical distribution produced by the log transform is not, of course, merely for convenience. Symmetrical distributions, especially those that resemble the normal distribution, fulfill statistical assumptions that form the basis of statistical significance testing in the regression model." (Edward R Tufte, "Data Analysis for Politics and Policy", 1974)

"The quantile plot is a good general display since it is fairly easy to construct and does a good job of portraying many aspects of a distribution. Three convenient features of the plot are the following: First, in constructing it, we do not make any arbitrary choices of parameter values or cell boundaries [...] and no models for the data are fitted or assumed. Second, like a table, it is not a summary but a display of all the data. Third, on the quantile plot every point is plotted at a distinct location, even if there are duplicates in the data. The number of points that can be portrayed without overlap is limited only by the resolution of the plotting device. For a high resolution device several hundred points distinguished." (John M Chambers et al, "Graphical Methods for Data Analysis", 1983)

"One graph is more effective than another if its quantitative information can be decoded more quickly or more easily by most observers. […] This definition of effectiveness assumes that the reason we draw graphs is to communicate information - but there are actually many other reasons to draw graphs." (Naomi B Robbins, "Creating More effective Graphs", 2005)

"Exploratory Data Analysis is more than just a collection of data-analysis techniques; it provides a philosophy of how to dissect a data set. It stresses the power of visualisation and aspects such as what to look for, how to look for it and how to interpret the information it contains. Most EDA techniques are graphical in nature, because the main aim of EDA is to explore data in an open-minded way. Using graphics, rather than calculations, keeps open possibilities of spotting interesting patterns or anomalies that would not be apparent with a calculation (where assumptions and decisions about the nature of the data tend to be made in advance)." (Alan Graham, "Developing Thinking in Statistics", 2006)

"Where correlation exists, it is tempting to assume that one of the factors has caused the changes in the other (that is, that there is a cause-and-effect relationship between them). Although this may be true, often it is not. When an unwarranted or incorrect assumption is made about cause and effect, this is referred to as spurious correlation […]" (Alan Graham, "Developing Thinking in Statistics", 2006)

"Too often there is a disconnect between the people who run a study and those who do the data analysis. This is as predictable as it is unfortunate. If data are gathered with particular hypotheses in mind, too often they (the data) are passed on to someone who is tasked with testing those hypotheses and who has only marginal knowledge of the subject matter. Graphical displays, if prepared at all, are just summaries or tests of the assumptions underlying the tests being done. Broader displays, that have the potential of showing us things that we had not expected, are either not done at all, or their message is not able to be fully appreciated by the data analyst." (Howard Wainer, Comment, Journal of Computational and Graphical Statistics Vol. 20(1), 2011)

"Data visualization is a means to an end, not an end in itself. It's merely a bridge connecting the messenger to the receiver and its limitations are framed by our own inherent irrationalities, prejudices, assumptions, and irrational tastes. All these factors can undermine the consistency and reliability of any predicted reaction to a given visualization, but that is something we can't realistically influence." (Andy Kirk, "Data Visualization: A successful design process", 2012)

"Data visualization is a means to an end, not an end in itself. It's merely a bridge connecting the messenger to the receiver and its limitations are framed by our own inherent irrationalities, prejudices, assumptions, and irrational tastes. All these factors can undermine the consistency and reliability of any predicted reaction to a given visualization, but that is something we can't realistically influence." (Andy Kirk, "Data Visualization: A successful design process", 2012)

"With time series though, there is absolutely no substitute for plotting. The pertinent pattern might end up being a sharp spike followed by a gentle taper down. Or, maybe there are weird plateaus. There could be noisy spikes that have to be filtered out. A good way to look at it is this: means and standard deviations are based on the naïve assumption that data follows pretty bell curves, but there is no corresponding 'default' assumption for time series data (at least, not one that works well with any frequency), so you always have to look at the data to get a sense of what’s normal. [...] Along the lines of figuring out what patterns to expect, when you are exploring time series data, it is immensely useful to be able to zoom in and out." (Field Cady, "The Data Science Handbook", 2017)

"Some scientists (e.g., econometricians) like to work with mathematical equations; others (e.g., hard-core statisticians) prefer a list of assumptions that ostensibly summarizes the structure of the diagram. Regardless of language, the model should depict, however qualitatively, the process that generates the data - in other words, the cause-effect forces that operate in the environment and shape the data generated." (Judea Pearl & Dana Mackenzie, "The Book of Why: The new science of cause and effect", 2018)

📉Graphical Representation: Dimensions (Just the Quotes)

"Two dimensional charts for the representation of mathematical equations or experimental data are in very common use nowadays and are everywhere recognized as valuable devices for giving a clear conception of the manner in which the variables are related. Their application is generally restricted, however, to cases where there is but one variable and its function, if the variation to be shown is continuous. Nevertheless cases often arise in which there are two variables and a function to be represented and where it is desirable to show a continuousvariation for all three." (John B Peddle, "The Construction of Graphical Charts", 1910)

"Graphic comparisons, wherever possible, should be made in one dimension only." (Willard C Brinton, "Graphic Methods for Presenting Facts", 1919)

"In general, the comparison of two circles of different size should be strictly avoided. Many excellent works on statistics approve the comparison of circles of different size, and state that the circles should always be drawn to represent the facts on an area basis rather than on a diameter basis. The rule, however, is not always followed and the reader has no way of telling whether the circles compared have been drawn on a diameter basis or on an area basis, unless the actual figures for the data are given so that the dimensions may be verified." (Willard C Brinton, "Graphic Methods for Presenting Facts", 1919)

"Readers of statistical diagrams should not be required to compare magnitudes in more than one dimension. Visual comparisons of areas are particularly inaccurate and should not be necessary in reading any statistical graphical diagram." (William C Marshall, "Graphical methods for schools, colleges, statisticians, engineers and executives", 1921)

"In short, the rule that no more dimensions or axes should be used in the chart than the data calls for, is fundamental. Violate this rule and you bring down upon your head a host of penalties. In the first place, you complicate your computing processes, or else achieve a grossly deceptive chart. If your chart becomes deceptive, it has defeated its purpose, which was to represent accurately. Unless, of course, you intended to deceive, in which case we are through with you and leave you to Mark Twain’s mercies. If you make your chart accurate, at the cost of considerable square or cube root calculating, you still have no hope, for the chart is not clear; your reader is more than likely to misunderstand it. Confusion, inaccuracy and deception always lie in wait for you down the path departing from the principle we have discussed - and one of them is sure to catch you." (Karl G Karsten, "Charts and Graphs", 1925)

"The bar chart is one of the most useful, simple, adaptable, and popular techniques in graphic presentation. The simple bar chart, with its many variations, is particularly appropriate for comparing the magnitude, or size, of coordinate items or of parts of a total. The basis of comparison in the bar chart is linear or one-dimensional. The length of each bar or of its components is proportional to the quantity or amount of each category' represented. " (Calvin F Schmid, "Handbook of Graphic Presentation", 1954)

"The common bar chart is particularly appropriate for comparing magnitude or size of coordinate items or parts of a total. It is one of the most useful, simple, and adaptable techniques in graphic presentation. The basis of comparison in the bar chart is linear or one-dimensional. The length of each bar or of its components is proportional to the quantity or amount of each category represented." (Anna C Rogers, "Graphic Charts Handbook", 1961)

"An especially effective device for enhancing the explanatory power of time-series displays is to add spatial dimensions to the design of the graphic, so that the data are moving over space (in two or three dimensions) as well as over time. […] Occasionally graphics are belligerently multivariate, advertising the technique rather than the data." (Edward R Tufte, "The Visual Display of Quantitative Information", 1983)

"Graphical integrity is more likely to result if these six principles are followed:

The representation of numbers, as physically measured on the surface of the graphic itself, should be directly proportional to the numerical quantities represented.

Clear, detailed, and thorough labeling should be used to defeat graphical distortion and ambiguity. Write out explanations of the data on the graphic itself. Label important events in the data.

Show data variations, not design variations. 

In time-series displays of money, deflated and standardized units of monetary measurements are nearly always better than nominal units.

The number of information-carrying (variable) dimensions depicted should not exceed the number of dimensions in the data.

Graphics must not quote data out of context." (Edward R Tufte, "The Visual Display of Quantitative Information", 1983)

"The time-series plot is the most frequently used form of graphic design. With one dimension marching along to the regular rhythm of seconds, minutes, hours, days, weeks, months, years, centuries, or millennia, the natural ordering of the time scale gives this design a strength and efficiency of interpretation found in no other graphic arrangement." (Edward R Tufte, "The Visual Display of Quantitative Information", 1983)

"Maximizing data ink (within reason) is but a single dimension of a complex and multivariate design task. The principle helps conduct experiments in graphical design. Some of those experiments will succeed. There remain, however, many other considerations in the design of statistical graphics - not only of efficiency, but also of complexity, structure, density, and even beauty." (Edward R Tufte, "Data-Ink Maximization and Graphical Design", Oikos Vol. 58 (2), 1990)

"The ducks of information design are false escapes from flatland, adding pretend dimensions to impoverished data sets, merely fooling around with information." (Edward R Tufte, "Envisioning Information", 1990)

"We envision information in order to reason about, communicate, document, and preserve that knowledge - activities nearly always carried out on two-dimensional paper and computer screen. Escaping this flatland and enriching the density of data displays are the essential tasks of information design." (Edward R Tufte, "Envisioning Information", 1990)

"Binning has two basic limitations. First, binning sacrifices resolution. Sometimes plots of the raw data will reveal interesting fine structure that is hidden by binning. However, advantages from binning often outweigh the disadvantage from lost resolution. [...] Second, binning does not extend well to high dimensions. With reasonable univariate resolution, say 50 regions each covering 2% of the range of the variable, the number of cells for a mere 10 variables is exceedingly large. For uniformly distributed data, it would take a huge sample size to fill a respectable fraction of the cells. The message is not so much that binning is bad but that high dimensional space is big. The complement to the curse of dimensionality is the blessing of large samples. Even in two and three dimensions having lots of data can bc very helpful when the observations are noisy and the structure non-trivial." (Daniel B Carr, "Looking at Large Data Sets Using Binned Data Plots", [in "Computing and Graphics in Statistics"] 1991)

"Fitting is essential to visualizing hypervariate data. The structure of data in many dimensions can be exceedingly complex. The visualization of a fit to hypervariate data, by reducing the amount of noise, can often lead to more insight. The fit is a hypervariate surface, a function of three or more variables. As with bivariate and trivariate data, our fitting tools are loess and parametric fitting by least-squares. And each tool can employ bisquare iterations to produce robust estimates when outliers or other forms of leptokurtosis are present." (William S Cleveland, "Visualizing Data", 1993)

"The visual representation of a scale - an axis with ticks - looks like a ladder. Scales are the types of functions we use to map varsets to dimensions. At first glance, it would seem that constructing a scale is simply a matter of selecting a range for our numbers and intervals to mark ticks. There is more involved, however. Scales measure the contents of a frame. They determine how we perceive the size, shape, and location of graphics. Choosing a scale (even a default decimal interval scale) requires us to think about what we are measuring and the meaning of our measurements. Ultimately, that choice determines how we interpret a graphic." (Leland Wilkinson, "The Grammar of Graphics" 2nd Ed., 2005)

"It is tempting to make charts more engaging by introducing fancy graphics or three dimensions so they leap off the page, but doing so obscures the real data and misleads people, intentionally or not." (Brian Suda, "A Practical Guide to Designing with Data", 2010)

"One way a chart can lie is through overemphasis of the size and scale of items, particularly when the dimension of depth isnʼt considered." (Brian Suda, "A Practical Guide to Designing with Data", 2010)

"Using colour, itʼs possible to increase the density of information even further. A single colour can be used to represent two variables simultaneously. The difficulty, however, is that there is a limited amount of information that can be packed into colour without confusion." (Brian Suda, "A Practical Guide to Designing with Data", 2010)

"Bear in mind is that the use of color doesn’t always help. Use it sparingly and with a specific purpose in mind. Remember that the reader’s brain is looking for patterns, and will expect both recurrence itself and the absence of expected recurrence to carry meaning. If you’re using color to differentiate categorical data, then you need to let the reader know what the categories are. If the dimension of data you’re encoding isn’t significant enough to your message to be labeled or explained in some way - or if there is no dimension to the data underlying your use of difference colors - then you should limit your use so as not to confuse the reader." (Noah Iliinsky & Julie Steel, "Designing Data Visualizations", 2011)

"[...] the human brain is not good at calculating surface sizes. It is much better at comparing a single dimension such as length or height. [...] the brain is also a hopelessly lazy machine." (Alberto Cairo, "The Functional Art", 2011)

"Explanatory data visualization is about conveying information to a reader in a way that is based around a specific and focused narrative. It requires a designer-driven, editorial approach to synthesize the requirements of your target audience with the key insights and most important analytical dimensions you are wishing to convey." (Andy Kirk, "Data Visualization: A successful design process", 2012)

"A signal is a useful message that resides in data. Data that isn’t useful is noise. […] When data is expressed visually, noise can exist not only as data that doesn’t inform but also as meaningless non-data elements of the display (e.g. irrelevant attributes, such as a third dimension of depth in bars, color variation that has no significance, and artificial light and shadow effects)." (Stephen Few, "Signal: Understanding What Matters in a World of Noise", 2015)

"When we use the number of dimensions as the classification criterion of visual displays, we get four distinct groups: charts, networks, and maps, along with figurative visualizations as a special group." (Jorge Camões, "Data at Work: Best practices for creating effective charts and information graphics in Microsoft Excel", 2016)

"A time series is a sequence of values, usually taken in equally spaced intervals. […] Essentially, anything with a time dimension, measured in regular intervals, can be used for time series analysis." (Andy Kriebel & Eva Murray, "#MakeoverMonday: Improving How We Visualize and Analyze Data, One Chart at a Time", 2018)

"Color is difficult to use effectively. A small number of well-chosen colors can be highly distinguishable, particularly for categorical data, but it can be difficult for users to distinguish between more than a handful of colors in a visualization. Nonetheless, color is an invaluable tool in the visualization toolbox because it is a channel that can carry a great deal of meaning and be overlaid on other dimensions. […] There are a variety of perceptual effects, such as simultaneous contrast and color deficiencies, that make precise numerical judgments about a color scale difficult, if not impossible." (Danyel Fisher & Miriah Meyer, "Making Data Visual", 2018)

"Maps also have the disadvantage that they consume the most powerful encoding channels in the visualization toolbox - position and size - on an aspect that is held constant. This leaves less effective encoding channels like color for showing the dimension of interest." (Danyel Fisher & Miriah Meyer, "Making Data Visual", 2018)

📉Graphical Representation: Extrapolation (Just the Quotes)

"In working through graphics one has, however, to be exceedingly cautious in certain particulars, for instance, when a set of figures, dynamical or financial, are available they are, so long as they are tabulated, instinctively taken merely at their face value. When plotted, however, there is a temptation to extrapolation which is well nigh irresistible to the untrained mind. Sometimes the process can be safely employed, but it requires a rather comprehensive knowledge of the facts that lie back of the data to tell when to go ahead and when to stop." (Allan C Haskell, "How to Make and Use Graphic Charts", 1919)

"A piece of self-deception - often dear to the heart of apprentice scientists - is the drawing of a 'smooth curve' (how attractive it sounds!) through a set of points which have about as much trend as the currants in plum duff. Once this is done, the mind, looking for order amidst chaos, follows the Jack-o'-lantern line with scant attention to the protesting shouts of the actual points. Nor, let it be whispered, is it unknown for people who should know better to rub off the offending points and publish the trend line which their foolish imagination has introduced on the flimsiest of evidence. Allied to this sin is that of overconfident extrapolation, i.e. extending the graph by guesswork beyond the range of factual information. Whenever extrapolation is attempted it should be carefully distinguished from the rest of the graph, e.g. by showing the extrapolation as a dotted line in contrast to the full line of the rest of the graph. [...] Extrapolation always calls for justification, sooner or later. Until this justification is forthcoming, it remains a provisional estimate, based on guesswork." (Michael J Moroney, "Facts from Figures", 1951)

"Extrapolations are useful, particularly in the form of soothsaying called forecasting trends. But in looking at the figures or the charts made from them, it is necessary to remember one thing constantly: The trend to now may be a fact, but the future trend represents no more than an educated guess. Implicit in it is 'everything else being equal' and 'present trends continuing'. And somehow everything else refuses to remain equal." (Darell Huff, "How to Lie with Statistics", 1954)

"Almost all efforts at data analysis seek, at some point, to generalize the results and extend the reach of the conclusions beyond a particular set of data. The inferential leap may be from past experiences to future ones, from a sample of a population to the whole population, or from a narrow range of a variable to a wider range. The real difficulty is in deciding when the extrapolation beyond the range of the variables is warranted and when it is merely naive. As usual, it is largely a matter of substantive judgment - or, as it is sometimes more delicately put, a matter of 'a priori nonstatistical considerations'." (Edward R Tufte, "Data Analysis for Politics and Policy", 1974)

"Each part of a graphic generates visual expectations about its other parts and, in the economy of graphical perception, these expectations often determine what the eye sees. Deception results from the incorrect extrapolation of visual expectations generated at one place on the graphic to other places." (Edward R Tufte, "The Visual Display of Quantitative Information", 1983)

"Time-series forecasting is essentially a form of extrapolation in that it involves fitting a model to a set of data and then using that model outside the range of data to which it has been fitted. Extrapolation is rightly regarded with disfavour in other statistical areas, such as regression analysis. However, when forecasting the future of a time series, extrapolation is unavoidable." (Chris Chatfield, "Time-Series Forecasting" 2nd Ed, 2000)

"The first myth is that prediction is always based on time-series extrapolation into the future (also known as forecasting). This is not the case: predictive analytics can be applied to generate any type of unknown data, including past and present. In addition, prediction can be applied to non-temporal (time-based) use cases such as disease progression modeling, human relationship modeling, and sentiment analysis for medication adherence, etc. The second myth is that predictive analytics is a guarantor of what will happen in the future. This also is not the case: predictive analytics, due to the nature of the insights they create, are probabilistic and not deterministic. As a result, predictive analytics will not be able to ensure certainty of outcomes." (Prashant Natarajan et al, "Demystifying Big Data and Machine Learning for Healthcare", 2017)

"If you study one group and assume that your results apply to other groups, this is extrapolation. If you think you are studying one group, but do not manage to obtain a representative sample of that group, this is a different problem. It is a problem so important in statistics that it has a special name: selection bias. Selection bias arises when the individuals that you sample for your study differ systematically from the population of individuals eligible for your study." (Carl T Bergstrom & Jevin D West, "Calling Bullshit: The Art of Skepticism in a Data-Driven World", 2020)

📉Graphical Representation: Variation (Just the Quotes)

"To a very striking degree our culture has become a Statistical culture. Even a person who may never have heard of an index number is affected [...] by [...] of those index numbers which describe the cost of living. It is impossible to understand Psychology, Sociology, Economics, Finance or a Physical Science without some general idea of the meaning of an average, of variation, of concomitance, of sampling, of how to interpret charts and tables." (Carrol D Wright, 1887)

"Two dimensional charts for the representation of mathematical equations or experimental data are in very common use nowadays and are everywhere recognized as valuable devices for giving a clear conception of the manner in which the variables are related. Their application is generally restricted, however, to cases where there is but one variable and its function, if the variation to be shown is continuous. Nevertheless cases often arise in which there are two variables and a function to be represented and where it is desirable to show a continuousvariation for all three." (John B Peddle, "The Construction of Graphical Charts", 1910)

"Graphical integrity is more likely to result if these six principles are followed:

The representation of numbers, as physically measured on the surface of the graphic itself, should be directly proportional to the numerical quantities represented.

Clear, detailed, and thorough labeling should be used to defeat graphical distortion and ambiguity. Write out explanations of the data on the graphic itself. Label important events in the data.

Show data variations, not design variations. 

In time-series displays of money, deflated and standardized units of monetary measurements are nearly always better than nominal units.

The number of information-carrying (variable) dimensions depicted should not exceed the number of dimensions in the data.

Graphics must not quote data out of context." (Edward R Tufte, "The Visual Display of Quantitative Information", 1983)

"One can hide data in a variety of ways. One method that occurs with some regularity is hiding the data in the grid. The grid is useful for plotting the points, but only rarely afterwards. Thus to display data badly, use a fine grid and plot the points dimly [...] A second way to hide the data is in the scale. This corresponds to blowing up the scale (i.e., looking at the data from far away) so that any variation in the data is obscured by the magnitude of the scale. One can justify this practice by appealing to 'honesty requires that we start the scale at zero', or other sorts of sophistry." (Howard Wainer, "How to Display Data Badly", The American Statistician Vol. 38(2), 1984)

"A good description of the data summarizes the systematic variation and leaves residuals that look structureless. That is, the residuals exhibit no patterns and have no exceptionally large values, or outliers. Any structure present in the residuals indicates an inadequate fit. Looking at the residuals laid out in an overlay helps to spot patterns and outliers and to associate them with their source in the data." (Christopher H Schrnid, "Value Splitting: Taking the Data Apart", 1991)

"When visualization tools act as a catalyst to early visual thinking about a relatively unexplored problem, neither the semantics nor the pragmatics of map signs is a dominant factor. On the other hand, syntactics (or how the sign-vehicles, through variation in the visual variables used to construct them, relate logically to one another) are of critical importance." (Alan M MacEachren, "How Maps Work: Representation, Visualization, and Design", 1995)

"Comparing series visually can be misleading […]. Local variation is hidden when scaling the trends. We first need to make the series stationary (removing trend and/or seasonal components and/or differences in variability) and then compare changes over time. To do this, we log the series (to equalize variability) and difference each of them by subtracting last year’s value from this year’s value." (Leland Wilkinson, "The Grammar of Graphics" 2nd Ed., 2005)

"In general. statistical graphics should be moderately greater in length than in height. And, as William Cleveland discovered, for judging slopes and velocities up and down the hills in time-series, best is an aspect ratio that yields hill - slopes averaging 45°, over every cycle in the time-series. Variations in slopes are best detected when the slopes are around 45°, uphill or downhill." (Edward R Tufte, "Beautiful Evidence", 2006)

"Graphical displays are often constructed to place principal focus on the individual observations in a dataset, and this is particularly helpful in identifying both the typical positions of datapoints and unusual or influential cases. However, in many investigations, principal interest lies in identifying the nature of underlying trends and relationships between variables, and so it is often helpful to enhance graphical displays in ways which give deeper insight into these features. This can be very beneficial both for small datasets, where variation can obscure underlying patterns, and large datasets, where the volume of data is so large that effective representation inevitably involves suitable summaries." (Adrian W Bowman, "Smoothing Techniques for Visualisation" [in "Handbook of Data Visualization"], 2008)

"Given the important role that correlation plays in structural equation modeling, we need to understand the factors that affect establishing relationships among multivariable data points. The key factors are the level of measurement, restriction of range in data values (variability, skewness, kurtosis), missing data, nonlinearity, outliers, correction for attenuation, and issues related to sampling variation, confidence intervals, effect size, significance, sample size, and power." (Randall E Schumacker & Richard G Lomax, "A Beginner’s Guide to Structural Equation Modeling" 3rd Ed., 2010)

"A signal is a useful message that resides in data. Data that isn’t useful is noise. […] When data is expressed visually, noise can exist not only as data that doesn’t inform but also as meaningless non-data elements of the display (e.g. irrelevant attributes, such as a third dimension of depth in bars, color variation that has no significance, and artificial light and shadow effects)." (Stephen Few, "Signal: Understanding What Matters in a World of Noise", 2015)

"Upon discovering a visual image, the brain analyzes it in terms of primitive shapes and colors. Next, unity contours and connections are formed. As well, distinct variations are segmented. Finally, the mind attracts active attention to the significant things it found. That process is permanently running to react to similarities and dissimilarities in shapes, positions, rhythms, colors, and behavior. It can reveal patterns and pattern-violations among the hundreds of data values. That natural ability is the most important thing used in diagramming." (Vasily Pantyukhin, "Principles of Design Diagramming", 2015)

"Another word of caution for dot plots that show changes over time. The dot plot is, by definition, a summary chart. It does not show all of the data in the intervening years. If the data between the two dots generally move in the same direction, a dot plot is sufficient. But if the data contain sharp variations year by year, a dot plot will obscure that pattern (as it also does for bar charts)." (Jonathan Schwabish, "Better Data Visualizations: A guide for scholars, researchers, and wonks", 2021)

"The urn model is a simple abstraction that can be helpful for understanding variation.This model sets up a container (an urn, which is like a vase or a bucket) full of identical marbles that have been labeled, and we use the simple action of drawing marbles from the urn to reason about sampling schemes, randomized controlled experiments, and measurement error. For each of these types of variation, the urn model helps us estimate the size of the variation using either probability or simulation." (Sam Lau et al, "Learning Data Science: Data Wrangling, Exploration, Visualization, and Modeling with Python", 2023)

"Index number shows by its variations the changes in a magnitude which is not susceptible either of accurate measurement in itself or of direct valuation in practice." (Francis Y Edgeworth)

📉Graphical Representation: Smoothing (Just the Quotes)

 "The chief problems in the technique of historigram [aka histogram] plotting are those of base line scales, types of lines to use for the graphs and methods of and purposes of smoothing these curves. The size of page, ability of grasp by the eye, subsequent treatment of the illustration, etc., are determining factors. The variable factor is usually plotted from a base line along the ordinate axis. Spacing and rules for scales apply as in frequency diagrams." (William C Marshall, "Graphical methods for schools, colleges, statisticians, engineers and executives", 1921)

"A connected graph is appropriate when the time series is smooth, so that perceiving individual values is not important. A vertical line graph is appropriate when it is important to see individual values, when we need to see short-term fluctuations, and when the time series has a large number of values; the use of vertical lines allows us to pack the series tightly along the horizontal axis. The vertical line graph, however, usually works best when the vertical lines emanate from a horizontal line through the center of the data and when there are no long-term trends in the data." (William S Cleveland, "The Elements of Graphing Data", 1985)

"If the underlying pattern of the data has gentle curvature with no local maxima and minima, then locally linear fitting is usually sufficient. But if there are local maxima or minima, then locally quadratic fitting typically does a better job of following the pattern of the data and maintaining local smoothness." (William S Cleveland, "Visualizing Data", 1993)

"As a general rule, the fewer the time intervals used in the averaging process, the more closely the moving average curve resembles the curve of the actual data. Conversely, the greater the number of intervals, the smoother the moving average curve. […] Moving average curves tend to have a delayed reaction to changes." (Robert L Harris, "Information Graphics: A Comprehensive Illustrated Reference", 1996)

"The plot tells us the data are granular in the data source, something we could not ascertain with the histogram. There is an important lesson here. Statistics texts and statistical packages that recommend the histogram as the graphical starting point for a data analysis are giving bad advice. The same goes for kernel density estimates. These are appropriate second stages for graphical data analysis. The best starting point for getting a sense of the distribution of a variable is a tally, stem-and-leaf, or a dot plot. A dot plot is a special case of a tally (perhaps best thought of as a delta-neighborhood tally). Once we see that the data are not granular, we may move on to a histogram or kernel density, which smooths the data more than a dot plot." (Leland Wilkinson, "The Grammar of Graphics" 2nd Ed., 2005)

"Another method used to simplify the appearance of a graphic is smoothing. A regression line overlaid on a scatterplot is a smooth representation of the relationship between the two graph variables. For time series data, a moving average of the data over time is often used to smooth out the variation over small time steps in order to illustrate the overall trend." (Daniel B Carr & Linda W Pickle, "Visualizing Data Patterns with Micromaps", 2010)

"Scatterplots are the preferred medium for adding smooth curves to show a causal functional relationship or an association […] However, despite the advantage of the scatterplot for seeing some types of patterns, the linked micromap design adds geographic location to the information displayed and so enables searches for geographic patterns that the scatterplot omits." (Daniel B Carr & Linda W Pickle, "Visualizing Data Patterns with Micromaps", 2010)

"Smoothing is a technique that can be used to remove some of the variation in short-term data in favor of emphasizing long-term trends." (Andy Kriebel & Eva Murray, "#MakeoverMonday: Improving How We Visualize and Analyze Data, One Chart at a Time", 2018)

"Smoothing and aggregating can help us see important features and relationships, but when we have only a handful of observations, smoothing techniques can be misleading. With just a few observations, we prefer rug plots over histograms, box plots, and density curves, and we use scatterplots rather than smooth curves and density contours. This may seem obvious, but when we have a large amount of data, the amount of data in a subgroup can quickly dwindle. This phenomenon is an example of the curse of dimensionality." (Sam Lau et al, "Learning Data Science: Data Wrangling, Exploration, Visualization, and Modeling with Python", 2023)

📉Graphical Representation: Angles (Just the Quotes)

"In the labelling of the pie-chart, you will furthermore encounter typographical difficulties. It is not ordinarily a good thing to make a reader crane his neck at various angles to read writing along every point of the compass, so you should not, as so many do, write on radii from the center of the circle. On the other hand, unless the chart and its segments are very large as compared with the size of the printing, you will introduce tricky optical illusions if you write all labels in the same directions inside the segments." (Karl G Karsten, "Charts and Graphs", 1925)

"The disadvantages of the pie-chart are many. It is worthless for study and research purposes. In the first place, the human eye cannot easily compare as to length the various arcs about the circle, lying as they do in different directions. In the second place, the human eye is not naturally skilled at comparing angles - those angles at the center of the circle, formed by the various rays or radii and subtending the various arcs. In the third place, the human eye is not an expert judge of comparative sizes of areas, especially those as irregular as the segments of parts of the circle. There is no way by which the parts of this round unit can be compared so accurately and quickly as the parts of a straight line or bar. Moreover, when, as frequently happens, several pie-charts are shown together, the various slices in one chart cannot be so easily compared with the corresponding slices in the next, as can the various parts of one 100% bar with corresponding parts of another bar." (Karl G Karsten, "Charts and Graphs", 1925)

"First, it is generally inadvisable to attempt to portray a series of more than four or five categories by means of pie charts. If, for example, there are six, eight, or more categories, it may be very confusing to differentiate the relative values portrayed, especially if several small sectors are of approximately the same size. Second, the pie chart may lose its effectiveness if an attempt is made to compare the component values of several circles, as might be found in a temporal or geographical series. In such case the one-hundred percent bar or column chart is more appropriate. Third, although the proportionate values portrayed in a pie chart are measured as distances along arcs about the circle, actually there is a tendency to estimate values in terms of areas of sectors or by the size of subtended angles at the center of the circle." (Calvin F Schmid, "Handbook of Graphic Presentation", 1954)

"Circles of different size, however cannot properly be used to compare the size of different totals. This is because the reader does not know whether to compare the diameters or the areas" (which vary as the squares of the diameters), and is likely to misjudge the comparison in either ease. Usually the circles are drawn so that their diameters are in correct proportion to each other; but then the area comparison is exaggerated. Component bars should be used to show totals of different size since their one dimension lengths can be easily judged not only for the totals themselves but for the component parts as well. Circles, therefore, can show proportions properly by variations in angles of sectors but not by variations in diameters. " (Anna C Rogers, "Graphic Charts Handbook", 1961)

"Pie charts have weaknesses and dangers inherent in their design and application. First, it is generally inadvisable to attempt to portray more than four or five categories in a circle chart, especially if several small sectors are of approximately the same size. It may be very confusing to differentiate the relative values. Secondly, the pie chart loses effectiveness if an effort is made to compare the component values of several circles, as might occur in a temporal or geographical series. [...] Thirdly, although values are measured by distances along the arc of the circle, there is a tendency to estimate values in terms of areas by size of angle. The 100-percent bar chart is often preferable to the circle chart's angle and area comparison as it is easier to divide into parts, more convenient to use, has sections that may be shaded for contrast with grouping possible by bracketing, and has an easily readable percentage scale outside the bars." (Anna C Rogers, "Graphic Charts Handbook", 1961)

"The circle graph, or pie chart, appears to simple and 'nonstatistical', so it is a popular form of presentation for general readers. However, since the eye can compare linear distances more easily and accurately than angles or areas, the component parts of a total usually can be shown more effectively in a chart using linear measurement." (Peter H Selby, "Interpreting Graphs and Tables", 1976)

"The bar or column chart is the easiest type of graphic to prepare and use in reports. It employs a simple form: four straight lines that are joined to construct a rectangle or oblong box. When the box is shown horizontally it is called a bar; when it is shown vertically it is called a column. [...] The bar chart is an effective way to show comparisons between or among two or more items. It has the added advantage of being easily understood by readers who have little or no background in statistics and who are not accustomed to reading complex tables or charts." (Robert Lefferts, "Elements of Graphics: How to prepare charts and graphs for effective reports", 1981)

"We make angle judgments when we read a pie chart, but we don't judge angles very well. These judgments are biased; we underestimate acute angles (angles less than 90°) and overestimate obtuse angles (angles greater than 90°). Also, angles with horizontal bisectors" (when the line dividing the angle in two is horizontal) appear larger than angles with vertical bisectors." (Naomi B Robbins, "Creating More effective Graphs", 2005)

"The donut, its spelling betrays its origins, is nearly always more deceit friendly than the pie, despite being modelled on a life-saving ring. This is because the hole destroys the second most important value- defining element, by hiding the slice angles in the middle." (Nicholas Strange, "Smoke and Mirrors: How to bend facts and figures to your advantage", 2007)

"Communication is the primary goal of data visualization. Any element that hinders - rather than helps - the reader, then, needs to be changed or removed: labels and tags that are in the way, colors that confuse or simply add no value, uncomfortable scales or angles. Each element needs to serve a particular purpose toward the goal of communicating and explaining information. Efficiency matters, because if you’re wasting a viewer’s time or energy, they’re going to move on without receiving your message." (Noah Iliinsky & Julie Steel, "Designing Data Visualizations", 2011)

"A histogram for discrete numerical data is a graph of the frequency or relative frequency distribution, and it is similar to the bar chart for categorical data. Each frequency or relative frequency is represented by a rectangle centered over the corresponding value" (or range of values) and the area of the rectangle is proportional to the corresponding frequency or relative frequency." (Roxy Peck et al, "Introduction to Statistics and Data Analysis" 4th Ed., 2012)

"The use of the density scale to construct the histogram ensures that the area of each rectangle in the histogram will be proportional to the corresponding relative frequency. The formula for density can also be used when class widths are equal. However, when the intervals are of equal width, the extra arithmetic required to obtain the densities is unnecessary." (Roxy Peck et al, "Introduction to Statistics and Data Analysis" 4th Ed., 2012)

"Graphs can help us interpret data and draw inferences. They can help us see tendencies, patterns, trends, and relationships. A picture can be worth not only a thousand words, but a thousand numbers. However, a graph is essentially descriptive - a picture meant to tell a story. As with any story, bumblers may mangle the punch line and the dishonest may lie." (Gary Smith, "Standard Deviations", 2014)

"A scatterplot reveals the strength and shape of the relationship between a pair of variables. A scatterplot represents the two variables by axes drawn at right angles to each other, showing the observations as a cloud of points, each point located according to its values on the two variables. Various lines can be added to the plot to help guide our search for understanding." (Forrest W Young et al, "Visual Statistics: Seeing data with dynamic interactive graphics", 2016)

"Researchers have studied how accurately people can read information displayed in different types of plots. They have found the following ordering, from most to leasta ccurately judged (•) Positions along a common scale, like in a rug plot, strip plot, or dot plot (•) Positions on identical, nonaligned scales, like in a bar plot (•) Length, like in a stacked bar plot (•) Angle and slope, like in a pie chart (•) Area, like in a stacked line plot or bubble chart (•) Volume and density, like in a three-dimensional bar plot (•) Color saturation and hue, like when overplotting with semitransparent points."  (Sam Lau et al, "Learning Data Science: Data Wrangling, Exploration, Visualization, and Modeling with Python", 2023)

📉Graphical Representation: Size (Just the Quotes)

"Comparison between circles of different size should be absolutely avoided. It is inexcusable when we have available simple methods of charting so good and so convenient from every point of view as the horizontal bar." (Willard C Brinton, "Graphic Methods for Presenting Facts", 1919)

"Sometimes the scales of these accompanying charts are so large that the reader is puzzled to get clearly in his mind what the whole chart is driving at. There is a possibility of making a simple chart on such a large scale that the mere size of the chart adds to its complexity by causing the reader to glance from one side of the chart to the other in trying to get a condensed visualization of the chart." (Willard C Brinton, "Graphic Methods for Presenting Facts", 1919) 

"To the question "how many rulings is the 'right' number?" there is unfortunately no easy answer. Charts designed to perform the work of a large amount of tabular data, being primarily tabular in purpose, obviously require closer rulings than charts designed primarily to present a picture. But even within these two groups the decision may be influenced by the precise purpose of the chart, its size and shape, the nature of the data, the degree of reading accuracy needed, and to some extent, by the style of the medium in which the chart appears." (Kenneth W Haemer, "Hold That Line. A Plea for the Preservation of Chart Scale Ruling", The American Statistician Vol. 1 (1) 1947)

"Recognize effective results. Does the type of chart selected give a comprehensive picture of the situation? Does the size of chart and visual aid used satisfy all audience requirements? Do materials meet all reproduction problems? Is the layout well balanced and style of lettering uniform? Does the chart as a whole accurately present the facts? Is the projected idea an effective visual tool?" (Mary E Spear, "Charting Statistics", 1952)

"The number of grid lines should be kept to a minimum. This means that there should be just enough coordinate lines in the field so that the eye can readily interpret the values at any point on the curve. No definite rule can be specified as to the optimum number of lines in a grid. This must be left to the discretion of the chart-maker and can come only from experience. The size of the chart, the type and range of the data, the number of curves, the length and detail of the period covered, as well as other factors, will help to determine the number of grid lines." (Calvin F Schmid, "Handbook of Graphic Presentation", 1954)

"The bar chart is one of the most useful, simple, adaptable, and popular techniques in graphic presentation. The simple bar chart. with its many variations, is particularly appropriate for comparing the magnitude, or size, of coordinate items or of parts of a total. The basis of comparison in the bar chart is linear or one-dimensional. The length of each bar or of its components is proportional to the quantity or amount of each category' represented. " (Calvin F Schmid, "Handbook of Graphic Presentation", 1954)

"The number of grid lines should be kept to a minimum. This means that there should be just enough coordinate lines in the field so that the eye can readily interpret the values at any point on the curve. No definite rule can be specified as to the optimum number of lines in a grid. This must be left to the discretion of the chart-maker and can come only from experience. The size of the chart, the type and range of the data, the number of curves, the length and detail of the period covered, as well as other factors, will help to determine the number of grid lines." (Calvin F Schmid, "Handbook of Graphic Presentation", 1954)

"Simplicity, accuracy. appropriate size, proper proportion, correct emphasis, and skilled execution - these are the factors that produce the effective chart. To achieve simplicity your chart must be designed with a definite audience in mind, show only essential information. Technical terms should be absent as far as possible. And in case of doubt it is wiser to oversimplify than to make matters unduly complex. Be careful to avoid distortion or misrepresentation. Accuracy in graphics is more a matter of portraying a clear reliable picture than reiterating exact values. Selecting the right scales and employing authoritative titles and legends are as important as precision plotting. The right size of a chart depends on its probable use, its importance, and the amount of detail involved." (Anna C Rogers, "Graphic Charts Handbook", 1961)

"Without adequate planning, it is seldom possible to achieve either proper emphasis of each component element within the chart or a presentation that is pleasing in its entirely. Too often charts are developed around a single detail without sufficient regard for the work as a whole. Good chart design requires consideration of these four major factors: (1) size, (2) proportion, (3) position and margins, and (4) composition." (Anna C Rogers, "Graphic Charts Handbook", 1961)

"The bar of a bar chart has two aspects that can be used to visually decode quantitative information-size (length and area) and the relative position of the end of the bar along the common scale. The changing sizes of the bars is an important and imposing visual factor; thus it is important that size encode something meaningful. The sizes of bars encode the magnitudes of deviations from the baseline. If the deviations have no important interpretation, the changing sizes are wasted energy and even have the potential to mislead." (William S. Cleveland, "Graphical Methods for Data Presentation: Full Scale Breaks, Dot Charts, and Multibased Logging", The American Statistician Vol. 38 (4) 1984) 

"When a graph is constructed, quantitative and categorical information is encoded, chiefly through position, size, symbols, and color. When a person looks at a graph, the information is visually decoded by the person's visual system. A graphical method is successful only if the decoding process is effective. No matter how clever and how technologically impressive the encoding, it is a failure if the decoding process is a failure. Informed decisions about how to encode data can be achieved only through an understanding of the visual decoding process, which is called graphical perception." (William S Cleveland, "The Elements of Graphing Data", 1985)

"Good graphics can be spoiled by bad annotation. Labels must always be subservient to the information to be conveyed, and legibility should never be sacrificed for style. All the information on the sheet should be easy to read, and more important, easy to interpret. The priorities of the information should be clearly expressed by the use of differing sizes, weights and character of letters." (Bruce Robertson, "How to Draw Charts & Diagrams", 1988)

"The visual representation of a scale - an axis with ticks - looks like a ladder. Scales are the types of functions we use to map varsets to dimensions. At first glance, it would seem that constructing a scale is simply a matter of selecting a range for our numbers and intervals to mark ticks. There is more involved, however. Scales measure the contents of a frame. They determine how we perceive the size, shape, and location of graphics. Choosing a scale (even a default decimal interval scale) requires us to think about what we are measuring and the meaning of our measurements. Ultimately, that choice determines how we interpret a graphic." (Leland Wilkinson, "The Grammar of Graphics" 2nd Ed., 2005)

"What distinguishes data tables from graphics is explicit comparison and the data selection that this requires. While a data table obviously also selects information, this selection is less focused than a chart's on a particular comparison. To the extent that some figures in a table are visually emphasised. say in colour or size and style of print. the table is well on its way to becoming a chart. If you're making no comparisons - because you have no particular message and so need no selection (in other words, if you are simply providing a database, number quarry or recycling facility) - tables are easier to use than charts." (Nicholas Strange, "Smoke and Mirrors: How to bend facts and figures to your advantage", 2007)

"Designers are responsible for the project’s fit and finish, that is, specifying the geometry and sizes of components so they properly mate with each other and are ergonomically and aesthetically acceptable within the operating environment." (Dennis K Lieu & Sheryl Sorby, "Visualization, Modeling, and Graphics for Engineering Design", 2009)

"One way a chart can lie is through overemphasis of the size and scale of items, particularly when the dimension of depth isnʼt considered." (Brian Suda, "A Practical Guide to Designing with Data", 2010)

"Maps also have the disadvantage that they consume the most powerful encoding channels in the visualization toolbox - position and size - on an aspect that is held constant. This leaves less effective encoding channels like color for showing the dimension of interest." (Danyel Fisher & Miriah Meyer, "Making Data Visual", 2018)

"When it comes to presenting categorical data, pie charts allow an impression of the size of each category relative to the whole pie, but are often visually confusing, especially if they attempt to show too many categories in the same chart, or use a three-dimensional representation that distorts areas. [...] Multiple pie charts are generally not a good idea, as comparisons are hampered by the difficulty in assessing the relative sizes of areas of different shapes. Comparisons are better based on height or length alone in a bar chart." (David Spiegelhalter, "The Art of Statistics: Learning from Data", 2019)

"The sizes of charts in space reflect how we convey information to a reader. In a dashboard context, the content, size, and space that the various charts occupy should reflect the form and function of the main message. As you saw with the bento box metaphor from the introduction, there needs to be deliberate thought put into the placement and size of each individual chart so that they all work together in harmony." (Vidya Setlur & Bridget Cogley, "Functional Aesthetics for data visualization", 2022)