1950-1974: Re-birth of data visualization

Still under the influence of the formal and numerical zeitgeist from the mid-1930s on, data visualization began to rise from dormancy in the mid 1960s, spurred largely by three significant developments:

  • In the USA, John W. Tukey, in a landmark paper, "The Future of Data Analysis" (Tukey:1962), issued a call for the recognition of data analysis as a legitimate branch of statistics distinct from mathematical statistics; shortly, he began the invention of a wide variety of new, simple, and effective graphic displays, under the rubric of "Exploratory Data Analysis" (EDA). Tukey's stature as a statistician and the scope of his informal, robust, and graphical approach to data analysis were as influential as his graphical innovations. Although not published until 1977, chapters from Tukey's EDA book (Tukey:1977) were widely circulated as they began to appear in 1970-1972, and began to make graphical data analysis both interesting and respectable again.
  • In France, Jacques Bertin published the monumental Semiologie Graphique (Bertin:1967). To some, this appeared to do for graphics what Mendeleev had done for the organization of the chemical elements, that is, to organize the visual and perceptual elements of graphics according to the features and relations in data.
  • But the skills of hand-drawn maps and graphics had withered during the dormant "modern dark ages" of graphics (though every figure in Tukey's EDA (Tukey:1977) was, by intention, hand-drawn). Computer processing of data had begun, and offered the possibility to construct old and new graphic forms by computer programs. True high-resolution graphics were developed, but would take a while to enter common use.

By the end of this period significant intersections and collaborations would begin: (a) computer science research (software tools, C language, UNIX, etc.) at Bell Laboratories (Becker:1994) and elsewhere would combine forces with (b) developments in data analysis (EDA, psychometrics, etc.) and (c) display and input technology (pen plotters, graphic terminals, digitizer tablets, the mouse, etc.). These developments would provide new paradigms, languages and software packages for expressing and implementing statistical and data graphics. In turn, they would lead to an explosive growth in new visualization methods and techniques.

Other themes begin to emerge, mostly as initial suggestions: (a) various visual representations of multivariate data; (b) animations of a statistical process (c) perceptually-based theory (or just informed ideas) related to how graphic attributes and relations might be rendered to better convey the data to the eyes.

1957
Circular glyphs
Added: 2008-07-17

Use of metroglyphs in a graphUse of metroglyphs in a graph

Diagramming variables in more than 3 dimensionsDiagramming variables in more than 3 dimensions

Circular glyphs, with rays to represent multivariate data


References:
Anderson:1957
1957
Fortran
Added: 2007-02-01

Creation of Fortran, the Formula Translation language for the IBM 704 computer. This was the first high-level language for computing.


References:
none
1958
Phillips curve
Added: 2006-03-04

The Phillips CurveThe Phillips Curve

The Phillips CurveThe Phillips Curve

The "Phillips Curve,'' a scatterplot of inflation vs. unemployment over time shows a strong inverse relation, leading to important developments in macroeconomic theory


References:
Phillips:1958
Added: 2007-02-01

Early SYMAP image of ConnecticutEarly SYMAP image of Connecticut

Initial development of geographic information systems, combining spatially-referenced data, spatial models and map-based visualization. Example: Harvard Laboratory for Computer Graphics (and Spatial Analysis) develops SYMAP, producing isoline, choropleth and proximal maps on a line printer


References:
Abbott:1884 Chrisman:1988
1962
Modern statistical graphics
Added: 2007-02-01

Kruskal portraitKruskal portrait

Beginnings of modern dynamic statistical graphics (a 1 minute movie of the iterative process of finding a multidimensional scaling solution)


References:
none
1965
Exploratory data analysis
Added: 2008-07-17

Hanging rootogram for the fit of a Poisson distributionHanging rootogram for the fit of a Poisson distribution

Tukey portraitTukey portrait

Beginnings of EDA: improvements on histogram in analysis of counts, tail values (hanging rootogram)


References:
Tukey:1965
1966
Triangular glyphs
Added: 2008-07-17

Triangular glyphs to represent simultaneously four variables, using sides and orientation


References:
PickettWhite:1966
1967
Theory of graphic symbols
Added: 2008-07-17

Bertin's seven visual variablesBertin's seven visual variables

The reorderable matrixThe reorderable matrix

Bertin portraitBertin portrait

Bertin color portraitBertin color portrait

Comprehensive theory of graphical symbols and modes of graphics representation


Among other things, Bertin introduced the idea of reordering qualitative variables in graphical displaysto make relations more apparent--- the reorderable matrix.
References:
Bertin:1967 Bertin:1983
1968
Graphical rational patterns
Added: 2008-07-17

Bachi number patternsBachi number patterns

Systematic "graphical rational patterns'' for statistical presentation


References:
Bachi:1968
1969
Exploratory data analysis
Added: 2008-07-17

Graphical innovations for exploratory data analysis (stem-and-leaf, graphical lists, box-and-whisker plots, two-way and extended-fit plots, hanging and suspended rootograms)


References:
Tukey:1972
1969
Cathode ray tube
Added: 2008-07-17

Suggestion for displaying five variables by means of movements on a CRT


References:
Barnard:1969
1969
1st interactive system
Added: 2008-07-17

The first well-known direct manipulation interactive system in statistics: allowed users to interactively control a power transformation in realtime for probability plotting


References:
Fowlkes:1969
Added: 2008-07-17

Star plot of crime rates in US citiesStar plot of crime rates in US cities

Irregular polygon ("star plot'') to represent multivariate data (with vertices at equally spaced intervals, distance from center proportional to the value of a variable) [but see Georg von Mayr in 1877 cite[S. 78]{vonMayr:1877} for first use]


References:
Siegel-etal:1971
1971
Social indicator reporting
Added: 2008-07-17

Proposal to use statistical graphics in social indicator reporting, particularly on television


References:
Biderman:1971
1971
Biplot
Added: 2007-02-01

Biplot representation of blood chemistry dataBiplot representation of blood chemistry data

Biplot representation of ratings of automobilesBiplot representation of ratings of automobiles

Development of the biplot, a method for visualizing both the observations and variables in a multivariate data set in a single display. Observations are typically represented by points, variables by vectors, such that the position of a point along a vector represents the data value


References:
Gabriel:1971
1972
Fourier series
Added: 2007-02-01

Fourier function plot imageFourier function plot image

Form of Fourier series to generate plots of multivariate data


References:
Andrews:1972
1973
Chernoff faces
Added: 2008-07-17

Faces plot of automobile data, by originFaces plot of automobile data, by origin

Chernoff portraitChernoff portrait

Cartoons of human face to represent multivariate data


References:
Chernoff:1973
1973
Social indicator chartbook
Added: 2008-07-17

USA Government chartbook devoted exclusively to reporting social indicator statistics


References:
PresidentBudget:1973
1973 to 1976
Innovation
Added: 2008-07-17

Revival of statistical graphics innovation, use by U.S. Bureau of the Census


References:
none
1974
Bivariate matrix
Added: 2008-07-17

CDC map of incidence of stomach cancerCDC map of incidence of stomach cancer

Color-coded bivariate matrix to represent two intervally measured variables in a single map (Urban Atlas series)[but see Georg von Mayr in 1874 cite[Fig. XIX]{vonMayr:1874} for first use]


References:
USCensus:1974
1974
Comparative graphics
Added: 2008-07-17

Wainer portraitWainer portrait

Comparative experimental test of histogram, hanging histogram and hanging rootogram


References:
Wainer:1974
Added: 2008-07-17

Friedman portraitFriedman portrait

Start of true interactive graphics in statistics; PRIM-9, the first system in statistics with 3-D data rotations provided dynamic tools for projecting, rotating, isolating and masking multidimensional data in up to nine dimensions


References:
Fishkeller-etal:1974 Fishkeller-etal:1974b