Making-visualizations

From Earlham CS Department
Revision as of 01:48, 27 August 2012 by Elena (talk | contribs) (→‎Reading Summaries)
Jump to navigation Jump to search

List each item you identify using the following format. The easiest way is to copy and paste the template. For now don't bother grouping them, we'll collect a bunch first and then see what the appropriate categories are based on what we find.

Google Doc instead? Yes

Pattern

  1. Another pithy idea. Why it's important. How to accomplish it. [Where It's From, page number/URL. curator initials]

Example

  1. Choose color combinations with good contrast. This makes it easier for people to separate the principle components. Identify a set to use and then ask your colleagues for feedback, use a web-based color choosing tool. [Charlie's Book of Viz, page 33. cfp]

Entries


Start-Up

  • "Excellence in statistical graphics consists of complex ideas communicated with clarity, precision, and efficiency. Graphical displays should:
    • show the data
    • induce the viewer to think about the substance rather than about methodology, graphic design, the technology of graphic production, or something else
    • avoid distorting what the data have to say
    • present many numbers in a small space
    • make large data sets coherent
    • encourage the eye to compare different p[pieces of data
    • reveal the data at several levels of detail, from a broad overview to the fine structure
    • serve a reasonably clear purpose: description, exploration, tabulation, or decoration
    • be closely integrated with the statistical and verbal descriptions of a data set"
    • - "The Visual Display of Quantitative Information", pg. 13
  • "Graphics should be reserved for the richer, more complex, more difficult statistical material." - "The Visual Display of Quantitative Information", pg. 30
  • Data Density = # of entries in data matrix/area of data display
    • Graphics should be based on large data matrices and have high data density ("VDoQI", pg. 161, RJL)
    • Tables or charts better for smaller data matrices to maximize data density ("VDoQI", pg. 161, RJL)
  • Integrate words/numbers/pictures in one space (“The Visual Display of Quantitative Information”, Chapter 9, p.180-181, ES):
    • Integrate supportive text to the plotting filed to make the information perception easier for the viewer (180 ES)
    • The size of type could be quite small(180 ES)
    • Keep table, graph, words packaged in one place of the page – they all speak the same information(180 ES)
    • Words on and around the graphics are highly effective in telling viewers how to allocate their attention to the various parts of data display(180 ES)
    • For graphics words should tell the viewer how to read the design and not what to read in terms of content(180 ES)
    • Make the reader choose how to perceive an information, do whatever it takes to understand the material (from presentation)(180 ES)
  • Graphical practices
    • If the nature of the data suggests the shape of graphic, follow the suggestion. Otherwise, move towards horizontal graphics about 50 percent wider than tall ((“The Visual Display of Quantitative Information”, Chapter 9,p. 186-190, ES)
      • Horizontally stretched time-series are more accessible to the eye (186 ES)
      • Shaded, calm, high contrast display might be better (187 ES)
      • Ease of labeling: easier to read from left to right on a horizontally stretched plotting-field (187 ES)
      • Emphasis on casual influence (187 ES)
      • Wider-than-tall shapes make it easier to follow from left to right (188 ES)
    • Mapped Pictures to portray large quantities of data at high densities(“Beautiful Evidence”, p.12-46, ES)
      • Include labels directly on the picture (43 ES)
      • Important comparisons among images should be pointed out to viewers by means of annotation, arrows, labels, highlighting (45 ES)
      • Useful to show both unmapped and mapped images (45 ES)
      • Every image should always reside on the universal measurement grid of 3-space and time, should accompany rescaling and zooming in and out (45 ES)
    • Micro/Macro Readings (“Envisioning Information”, p.37-53, ES)
      • Reports immense detail, organizing complexity through multiple and hierarchical layers of contextual reading (38 ES)
      • Often simplification leads to ambiguity - instead cluster & confusion is a failure of design, not of too much data ("EI" 51 RJL)
    • Multiples
      • Can be used to reveal changes, patterns, and surprises through comparison ("VE" pg 105, RJL)
      • Can show passage of time, space, etc. ("VE" pg 108 RJL)
      • "Shrink principle" - graphics (especially multiples) can often be shrunken significantly without any loss of readability ("VDoQI", p.167, RJL)
      • Also great for large quantities of multivariate data ("VDoQI", p. 169, RJL)
    • Choose friendly data graphic. (“The Visual Display of Quantitative Information”, Chapter 9, p.183, ES)
      • Words are spelled out, elaborate encoding is avoided (183 ES)
      • Words run from left to right(183 ES)
      • Little messages help explain data(183 ES)
      • Elaborately encoded shadings, cross-hatching, colors are avoided. Instead labels are placed on the graphic, no legend required(183 ES)
      • Colors if used are chosen so that color-deficient/ color-blind can make sense of it (choose blue) (183 ES)
      • type is clear, precise, modest (183 ES)
      • Type is upper and lower case, with serifs (183 ES)
      • the more the letters are different from each other, the easier is the reading (183 ES)
      • all upper case letter is the hardest read (183 ES)
    • graphical elements look better when their relative proportions are in balance. (“The Visual Display of Quantitative Information”, Chapter 9, p.184-186, ES)
      • lines in data graphics should be thin (185 ES)
      • choose intersection of lines of different weights – contrast in line weight represents contrast in meaning (185 ES)
      • Choosing perpendicular intersections of lines of different weight, the heavier line should be a data measure (186 ES)
  • Manners of presenting (from the lecture, ES)
    • Research problem, examples of problem solutions
    • Content: what the problem is, relevance, solution
    • Dedicate 8 min of you presentation for audience to pre-read of the material
    • Present the material as it was given for the pre-read. Nothing like repetition improves the understanding
    • Credibility on speech: give the reasons to believe - documentation
    • Conclude with question to personalize/particularize
    • Inefficiency of PowerPoint – it is stuck in time

Web Design

  • "Information becomes the interface" - Basically don't clutter your interface except with the information the audience wants - Visual Explanations, pg. 146
  • "Measures of the informational performance of a screen include:
    • the proportion of space on the screen devoted to content
    • to computer administration, and to nothing at all;
    • character counts and measures of typographic density (making comparisons with printed material as well as computer interfaces);
    • The number of computer commands immediately available (more are better, if clearly but minimally displayed)" - Visual Explanations, pg 150
  • "Computer selects, organizes, customizes data; paper makes visible the high-resolution information in portable and permanent form." - Visual Explanations. pg 149

Financial Data - Elena

  1. Avoid visual distortion in data graphics. This will allow the viewer to perceive reality more accurately. Table - best way to show numbers (20 numbers or less>prefer table to a graph). Representation of numbers should be directly proportional to the numerical quantities represented. Clear, detailed labeling to defeat graphical distortion and ambiguity. Show data variation, not design variation. (The Visual Display of Quantitative Information, Chapter 2, p.56, p.61, ES)
  2. Don't tell lies in your graph by planning graphical representation accurately. That avoid inaccurate reflection of reality. If plotting government spending and dept over the years, take population and inflation into account. In time-series displays of money, deflated and standardized units of monetary measurement are nearly always better than nominal units(The Visual Display of Quantitative Information, Chapter 2, p.68 ES)
  3. A use of 2 or 3 varying dimensions to show one-dimensional data is a weak and inefficient technique. The number of information should not exceed the number of dimensions in the data (The Visual Display of Quantitative Information, Chapter 2, p.71 ES)
  4. Context is essential for graphical integrity. Graphics must not quote data out of context. In the quantitative thinking data graphics should answer “Compared to What?” question. (The Visual Display of Quantitative Information, Chapter 2, p.74 ES)
  5. Blending quantitative multiples, narrative text and images is useful for monitoring data-rich processes. Multiples help to monitor and analyse typical to finance multi-variable processes, combining overview with detail (Visual Explanations, p.110-111 ES)
  6. Consider sparklines for data-rich representation: datawords – data-intense, data-simple, word-sized graphics. Tracks and compares changes over time, by showing overall trend along with local detail; shows recent change in relation to many pat changes; sparklines gives a context for nuanced analysis and better decisions; shows intensity/frequency of occurrence. Should often be embedded in text and tables : possibility of writing with data graphics; Variation in slopes are best detected when slopes are 45 average; moderately greater in length than in height; changing the relative weight of the data-lines and also muting the contrast between the data and background to reduce optical noise; avoiding strong frames around that create unintentional optical clutter; printing and viewing data density of 500 spraklines on A3 size paper (Beautiful Evidence, p.46-63 ES)
  7. Choose text-table representation for the display of quantitative material. Best way to show exact numerical values (for many small data sets); Easier localized data comparison. Choose to order by content or data values (not alphabetically); Don’t choose pie charts: Low data density, failure to order numbers along a visual dimensions, Hard comparison of data located in spatial disarray within the pies. (“The Visual Display of Quantitative Information”, Chapter 9, p.178-179 ES)
  8. Choose wordy data graphic for sets of highly labeled numbers (“The Visual Display of Quantitative Information”, Chapter 9, p.180 ES)
  9. Graphical Practices (“The Visual Display of Quantitative Information”, Chapter 1)
    • Data maps is the most powerful way of displaying statistical information (16 ES)
    • Time-series plot gives the design strength and efficiency of interpretation (28 ES)
      • Shape of the data in a comparative perspective (30 ES)
      • Casual explanation: adding additional variables into the graphic design (38 ES)
      • Vivid design : before-after (39 ES)
      • Explanatory power of time-series displays is to add spatial dimensions to the design of the graphic (40 ES)
    • Keep the principle of graphical excellence – the well-designed presentation of interesting data, a matter of substance, statistics and of design (51 ES)
      • Complex ideas communicated with clarity, precision and efficiency (51 ES)
      • Give the greatest number of ideas in the shortest time with the least ink in the smallest space


Reading Summaries

Retrieved from "https://wiki.cs.earlham.edu/index.php?title=Making-visualizations&oldid=13342"