Charts & Graphs
Post Reply
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
18-01-2017, 08:48 PM
RE: Charts & Graphs
[Image: scatter-Artboard_1.png]

https://www.nytimes.com/interactive/2017...ecord.html

Of the 17 hottest years ever recorded, 16 have now occurred since 2000.

“I am quite sure now that often, very often, in matters concerning religion and politics a man’s reasoning powers are not above the monkey’s.”~Mark Twain
“Ocean: A body of water occupying about two-thirds of a world made for man - who has no gills.”~ Ambrose Bierce
Find all posts by this user
Like Post Quote this message in a reply
18-01-2017, 10:33 PM
RE: Charts & Graphs
Weathermaps are charts of sorts.

[Image: stroke.gif]

"If we are honest—and scientists have to be—we must admit that religion is a jumble of false assertions, with no basis in reality.
The very idea of God is a product of the human imagination."
- Paul Dirac
Find all posts by this user
Like Post Quote this message in a reply
[+] 4 users Like The Organic Chemist's post
19-01-2017, 10:18 PM
RE: Charts & Graphs
[Image: infographic-stop-home-hacked.png]

“I am quite sure now that often, very often, in matters concerning religion and politics a man’s reasoning powers are not above the monkey’s.”~Mark Twain
“Ocean: A body of water occupying about two-thirds of a world made for man - who has no gills.”~ Ambrose Bierce
Find all posts by this user
Like Post Quote this message in a reply
19-01-2017, 10:36 PM
RE: Charts & Graphs
[Image: Screen-Shot-2017-01-18-at-1.08.35-PM.png]

“I am quite sure now that often, very often, in matters concerning religion and politics a man’s reasoning powers are not above the monkey’s.”~Mark Twain
“Ocean: A body of water occupying about two-thirds of a world made for man - who has no gills.”~ Ambrose Bierce
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes Full Circle's post
05-02-2017, 07:20 AM
RE: Charts & Graphs

Is America enriching the world at its own expense? That’s globaloney.


The world is far less globalized than people tend to think.
By Pankaj Ghemawat and Steven A. Altman
https://www.washingtonpost.com/postevery...b5e8f26cdb

[Image: 2300-o-globalized1-02xxIMG.jpg&w=1484]

“I am quite sure now that often, very often, in matters concerning religion and politics a man’s reasoning powers are not above the monkey’s.”~Mark Twain
“Ocean: A body of water occupying about two-thirds of a world made for man - who has no gills.”~ Ambrose Bierce
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes Full Circle's post
09-02-2017, 11:13 AM (This post was last modified: 09-02-2017 11:27 AM by GirlyMan.)
Human Factor Chart Drawing Guidelines
So, as it turns out I am the PI for a human subject study called Understanding Data Graphs under the tutelage of the Banjo of experimental psychologist drummers. (He tells me the proper way to do things and then I go do them. I only agreed to it because the absurdity of a mediocre computer scientist as PI and the world class researcher dude as an associate investigator makes me chuckle.) He has devised a number of graph drawing guidelines based on human factors. I will share them with you here to help you get through those scary scary nights when you think all hope is lost.
.

There is only one really serious philosophical question, and that is suicide. -Camus
Find all posts by this user
Like Post Quote this message in a reply
09-02-2017, 11:21 AM
Guidelines in Gillan That Reference The Relations of Figures to Text
Guidelines in Gillan That Reference
The Relations of Figures to Text
The Graphing of Psychological Experiment Data
and
Graph Design for Papers with Multiple Graphs

Use simple phrases consistent with the text for labels.

Information in the graph should be congruent with the information in the text.

Label the y-axis with the name of the dependent variable used in the text.

Label the x-axis with the name of the relevant independent variable used in the text.

Labels for indicators should use the same names as those in the text for the levels of the relevant independent variable.

The relations shown between the independent and dependent variables should both relate to the hypotheses described in the paper’s introduction and reflect the analyses described in its results section.

The graphical representation of the data should help the reader to interpret the results that are described in the text.
Refer to the graph in the text early in the section in which its data are being described. Do not cite it only after describing the data.

To the extent possible, coordinate the text and graph so that the same printed page will contain the graph(s) and text that describe the same data.

An article that presents a series of related experiments using the same basic experimental design and analytical methods should maintain consistent graph structures and symbol use. This consistency will allow the reader to perceive the related meaning across experiments and directly compare the data across experiments.

Use the same type of graph across similar experiments. In other words, if a bar graph displays the data for Experiment 1, then a bar graph should display comparable data for Experiment 2.

Use the same layout of the data in the graph across similar experiments. In other words, if a graph for Experiment 1 has “Type of Interface” on the x-axis, then place “Type of Interface” on the x-axis of the comparable graph for Experiment 2.

Use the same symbols and patterns for coding the same independent variables (and levels thereof) across the graphs for different experiments.

Use the same numerical scales on the axes across graphs from different experiments that display the same dependent variable.

Make the critical feature that distinguishes each graph from its neighboring graphs visually prominent.

Search will be particularly challenging for readers looking across multiple graphs with many levels of an independent variable. To design multiple graphs with many levels of an independent variable, an author might (a) place all graphs in one Figure to facilitate search across graphs; (b) use spatial proximity for graphs that the reader might search in sequence (e.g., a series of response time graphs for a sequence of conditions might be aligned horizontally); © maintain visual consistency across graphs (e.g., use the same size axes, the same types of indicators, and the same coding for indicators for the same variables); (d) maintain semantic consistency across graphs (e.g., use the same scale on the y axis); (e) eliminate redundant labels (e.g., in a series of horizontally aligned graphs, the labels for the y axes should be placed only by the leftmost graph, and the verbal label naming the x axis should be centered under the series); and (f) avoid using a legend to label indicators (a legend requires multiple scans between the indicators and the legend, thereby disrupting visual search).

If a study measures different dependent variables (e.g., response time, accuracy, and usability) and the reader is likely to compare the results for the different dependent variables (either to observe the same pattern of results or to observe differences), make the graphs visually similar to facilitate comparisons. Specifically, use the same type of graph (line, bar, etc.), same layout of data, same symbols and patterns for coding the independent variables, and same length and labeling for the x axes. In addition, keep the length of the y-axis similar across the different graphs. (Because of differences in the values of the dependent variables, the graphs will most likely use different y-axis scales.).

If different graphs are intended to communicate unrelated information within an experiment or between experiments (e.g., graphs addressing different hypotheses), visual differences can be used to discourage readers from making comparisons that are not meaningful.

To create visual differences that match the differences in meaning between graphs, use different types of graphs across the different conditions or experiments.
Use different coding symbols or patterns for the independent variables across the different graphs.

If a graph contains two independent variables – one independent variable plotted on the x-axis and the second independent variable plotted as separate lines for each level of that variable – place the labels that specify the levels of the second independent variable close to the lines.

Sometimes, however, this can mean that indicators are too close to be labeled unambiguously. Use a legend only in such cases.

A legend should show symbols for all indicators clearly and, for each indicator, should show the entire symbol (e.g., both the redundantly coded plotting symbol and line for each line graph indicator) with the label in close proximity.

Separate the legend visually from the indicators in the graph by placing it in a box. Put the legend close to the indicators to reduce scanning distance but not so close as to interfere with reading the indicators.

The order of the symbols in the legend should match the order of the indicators in the graph.

Readers should be able to remember the data in relation to the hypothesis, theoretical point, or design implications of article.

Design the graphs so that the most important variables for the hypothesis and/or applications are available to global perception and are most salient.

Refer to the relevant graph or graphs in the textual discussions relevant to the hypotheses or applications.

Readers’ memories for graphical information may be subject to various memory biases. For instance, readers tend to remember the slope of a line in a graph as closer to 45° than it actually is. Likewise, readers tend to remember a curved line in a graph as more symmetrical than it actually is.

Textual descriptions of the critical characteristics of the data should be clear, accurate, and consistent with the visual representation in the graph.
The scaling of values (i.e., minimum and maximum values and spacing between them) on two adjacent graphs depicting the same variables should be the same.

There is only one really serious philosophical question, and that is suicide. -Camus
Find all posts by this user
Like Post Quote this message in a reply
09-02-2017, 11:23 AM
Claims in Gillan that Reference Notions of Salience
Claims in Gillan that Reference Notions of Salience

Readers of graphs of the size published in journals tend to examine the global features of the graph and/or indicator before examining the local features.

Readers will process the most salient features of the graph first.

Make the main point of the graph available to the reader at the global level. In other words, avoid making the reader search for the main point in the details of the graph.

Make the main point of the graph the most salient feature. In other words, attract the reader’s attention to the data most relevant to the message of the paper.

In bar graphs, use dark, heavy lines as the pattern code of a bar if the reader should pay particular attention to those data.

In line graphs, use dark, filled plotting symbols or dark, thick lines if the reader should pay particular attention to those data.

In scatter plots that show a best-fitting line that summarizes the data, make the best- fitting line thick and dark relative to the data points so that the reader will pay particular attention to the summary.

In line or bar graphs that use error bars to show variability, do not make the error bars thick and dark relative to the indicator. Longer error bars show the conditions with the greatest variability. Typically, readers want to attend to the conditions with the least variable, most reliable data points.

Unnecessary visual elements tend to slow search, so eliminate clutter by graphing only essential information. Make less essential information less salient by “ghosting” or “lowlighting” it (i.e., by representing it in a shade of gray that is visible but discriminably lighter than that used for more salient information).

Stevens’s law describes the general relation between physical amount and per ceived amount as a power function: perceived amount = a(physical amount)b (e.g., Stevens, 1975). When the exponent b = 1.0, the increase in perceived amount corresponds to the increase in the physical amount; when b < 1.0, the perceived amount increases more slow ly than the physical amount; and when b > 1.0, the perceived amount increases more rapidly than the physical amount. Because the physical dimensions that display simple linear extent in either vertical height or horizontal length pro duce Stevens’s law exponents of 1.0, a reader’s perception of extent with bar graphs and line graphs will accurately correspond to the physi cal distances shown in the graph. In contrast, the physical dimensions of the area of a rectangle or a circle and the volume of a cube typically produce exponents in Stevens’s law of less than 1.0, resulting in misestimation of the size of indicators. The physi cal dimension of the lightness of shades of
gray typically produces an exponent in Stevens’s law of greater than 1.0. When possible, use a physical dimension with a Stevens’s law exponent close to 1.0, such as linear extent, to indicate values in a graph. Psychophysical studies on segments of pie charts have been less consistent than those on linear extent. Some studies have shown a Stevens’s law exponent of 1.0 relating the perceived proportion indicated by a pie segment and the actual proportion of that segment, but other studies have shown a Stevens’s law exponent less than 1.0. Accordingly, use of a pie chart to display proportions may lead to misperception by the reader under some (as yet unknown) conditions. One solution would be to label the proportional value of each pie segment.

For pie charts, to draw the readers’ attention to one part from the whole, one approach is to separate that part from the whole (e.g., in an exploded pie chart). However, overuse of this technique by separating more than one part will lead to little increase in salience and will require the reader to perceptually reorganize the parts.

There is only one really serious philosophical question, and that is suicide. -Camus
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes GirlyMan's post
14-02-2017, 10:33 AM
RE: Charts & Graphs
[Image: median_income_2016.png]

Surprised that Mississippi is not on the bottom of that scale as well.

“I am quite sure now that often, very often, in matters concerning religion and politics a man’s reasoning powers are not above the monkey’s.”~Mark Twain
“Ocean: A body of water occupying about two-thirds of a world made for man - who has no gills.”~ Ambrose Bierce
Find all posts by this user
Like Post Quote this message in a reply
14-02-2017, 06:05 PM
RE: Charts & Graphs
(14-02-2017 10:33 AM)Full Circle Wrote:  [Image: median_income_2016.png]

Surprised that Mississippi is not on the bottom of that scale as well.

Looks like it is.

There is only one really serious philosophical question, and that is suicide. -Camus
Find all posts by this user
Like Post Quote this message in a reply
Post Reply
Forum Jump: