How not to let the diagrams fool themselves
We continue to develop the theme of manipulation of diagrams in presentations (and not only in them) and today I will submit an interesting article. You will be able to experience all the techniques listed in it :)
By the way, we continue to record everyone in the free online video seminar " Manipulative Diagramology ", which will be held on March 5, 2009.
')
Modern technology allows you to quickly and visually represent a significant amount of information. Some of them — for example, a projector, a screen, and a plasma panel — provide data to a large number of people at the same time. Therein lies a great danger, because technology allows the use of special techniques that can mislead the audience. In particular, with the help of diagrams it is possible to present information that is unfavorable for this company in such a way that it will look as positive. Let's get acquainted with some tricks that will help uncover the deception and learn how to correctly interpret the data.
Especially actual timely understanding of the information that is presented in the process of public speaking, when the audience perceives only the visual picture. As a result, listeners do not have time to study the numbers in detail on each slide of the presentation, they often get the wrong impression. Positive, although in reality the speaker can not boast of significant achievements, or, conversely, negative - while the speaker is confident that the audience will appreciate the success of his company.
Why it happens? The answer lies in the essence of the diagrams themselves, because they are used to present data in such a way as to create persuasive presentations. They serve to influence people, sometimes even manipulate them. Therefore, we can say that the existence of truthful schedules is not just impossible, but often undesirable.
A chart is valid when the recipient of the message draws the same conclusion as the creator of the chart when analyzing information. False - when the creator intentionally hides adverse information for him. Let's do a little experiment. Suppose you are an investor who turns to a consultant in order to determine the choice of shares. The consultant opens the presentation and shows six slides: they present a comparative analysis of the activities of companies A and B. Look at Figures 1A, 2A, 3A, 4A, 5A, 6A and imagine that you are in the presentation. Do not stop looking at the "slides" for a long time and decide which company you preferred. If A, then bend the finger of the left hand, if B - then the right.
What company would you choose? Congratulations, if A - it is more successful, and much more successful, although at first glance it seems that B can boast the best performance. These diagrams are easier to read on paper - some of them show exact values. When the slides are shown using a projector and the audience is forced to perceive information from a long distance, the audience will simply not see the numbers and will make the wrong decisions based on visual perception.
Now let's take a look at how the comparison of companies A and B will look without “deception”.
One of the most popular data distortions is associated with the use of 3D graphics for all types of charts (pie, bar, bar). Wasn’t it so difficult to determine which of the sectors (company A or B) on comparison 1A and comparison 2A is more? But when converting to 2D (comparing 1B and comparing 2B) you will not have problems with perception: it is immediately clear that the company has a large market share.
Increasing / decreasing intervals on scales is another popular trick. For example, comparison of 3A shows that sales growth of company A is less stable than that of company B. This is due to the use of different ranges of scales: months on the left chart, quarters on the right chart. If you reorganize the charts correctly and present months everywhere (comparison 3B), it will become obvious: sales of company A are more stable than those of B.
Displaying diagrams in perspective is another sophisticated technique that allows you to present a downward trend in the form of a growth graph. If we remove the perspective, it becomes clear that the amount of profit will increase for company A, and not at all for B (comparison 4B).
Distortion of data is very veiled when comparing the dynamics of changes in several variables that are very different. For example, the value of shares of company A increased 43 times: from 1 to 43 cu, of shares of company B - by 2.7 times: from 300 to 800 cu However, in the 5A comparison, the data are presented in such a way that company B seems to be the clear leader.
On the other hand, the predominance of company A is clearly visible in comparison with 5B (with a logarithmic scale). Logarithmic scales are widely used in science, but, unfortunately, have not yet found wide application in business. There are several reasons: many experts do not know how effective the use of diagrams of this type can be, while others often do not know how to interpret them correctly.
Let's analyze the diagram presented in comparison 5B. How can it be obtained from the initial version presented in comparison 5A? Everything is very simple. If you build a chart using Excel in the "Axis Format ..." you can set the value of "logarithmic scale", and the chart will immediately be rebuilt. And if you create it yourself, all the values ​​of the Y axis are rebuilt using a simple formula: new value = lg (old value). From here, the Y scale changes: the values ​​are 10, 100, 1000, 10 000, etc., that is, degrees of 10 (100, 101, 102, 103, etc.).
A lot of distortion causes misuse of the scales. The extremely popular technique is the intersection of the Y axis with the X axis not in the zero value. For example, a comparison of 6A demonstrates that the risk of loss of investment in company A is 3 times higher compared to company B, but the risk is actually the most by 2%, and this is well presented visually in comparison 6B, where the Y axis intersects with the X axis in zero value.
There are many other tricks. Among them are the stretching / compression of the axes, the use of different scales, the seizure of data from the context due to the display of a part of the diagram, when a picture is favorable for the speaker, etc. These techniques allow you to visually increase or decrease the significance of a parameter.
Cheating can also be achieved with the help of ordinary design tricks. For example, to diminish the significance of an object, you can reduce the contrast and illumination of the desired column, ruler, or sector. And if the diagram is presented at the presentation, you can choose the color of the object in such a way that at a distance it will simply “accidentally” merge with the background color and will not be taken into account.
It is recommended to be critical about the diagrams that speakers demonstrate. Always question their visibility and pay attention not only to visual perception, but also to specific numerical values.
By the way, we continue to record everyone in the free online video seminar " Manipulative Diagramology ", which will be held on March 5, 2009.
')
Modern technology allows you to quickly and visually represent a significant amount of information. Some of them — for example, a projector, a screen, and a plasma panel — provide data to a large number of people at the same time. Therein lies a great danger, because technology allows the use of special techniques that can mislead the audience. In particular, with the help of diagrams it is possible to present information that is unfavorable for this company in such a way that it will look as positive. Let's get acquainted with some tricks that will help uncover the deception and learn how to correctly interpret the data.
Especially actual timely understanding of the information that is presented in the process of public speaking, when the audience perceives only the visual picture. As a result, listeners do not have time to study the numbers in detail on each slide of the presentation, they often get the wrong impression. Positive, although in reality the speaker can not boast of significant achievements, or, conversely, negative - while the speaker is confident that the audience will appreciate the success of his company.
Why it happens? The answer lies in the essence of the diagrams themselves, because they are used to present data in such a way as to create persuasive presentations. They serve to influence people, sometimes even manipulate them. Therefore, we can say that the existence of truthful schedules is not just impossible, but often undesirable.
A chart is valid when the recipient of the message draws the same conclusion as the creator of the chart when analyzing information. False - when the creator intentionally hides adverse information for him. Let's do a little experiment. Suppose you are an investor who turns to a consultant in order to determine the choice of shares. The consultant opens the presentation and shows six slides: they present a comparative analysis of the activities of companies A and B. Look at Figures 1A, 2A, 3A, 4A, 5A, 6A and imagine that you are in the presentation. Do not stop looking at the "slides" for a long time and decide which company you preferred. If A, then bend the finger of the left hand, if B - then the right.
What company would you choose? Congratulations, if A - it is more successful, and much more successful, although at first glance it seems that B can boast the best performance. These diagrams are easier to read on paper - some of them show exact values. When the slides are shown using a projector and the audience is forced to perceive information from a long distance, the audience will simply not see the numbers and will make the wrong decisions based on visual perception.
Now let's take a look at how the comparison of companies A and B will look without “deception”.
One of the most popular data distortions is associated with the use of 3D graphics for all types of charts (pie, bar, bar). Wasn’t it so difficult to determine which of the sectors (company A or B) on comparison 1A and comparison 2A is more? But when converting to 2D (comparing 1B and comparing 2B) you will not have problems with perception: it is immediately clear that the company has a large market share.
Increasing / decreasing intervals on scales is another popular trick. For example, comparison of 3A shows that sales growth of company A is less stable than that of company B. This is due to the use of different ranges of scales: months on the left chart, quarters on the right chart. If you reorganize the charts correctly and present months everywhere (comparison 3B), it will become obvious: sales of company A are more stable than those of B.
Displaying diagrams in perspective is another sophisticated technique that allows you to present a downward trend in the form of a growth graph. If we remove the perspective, it becomes clear that the amount of profit will increase for company A, and not at all for B (comparison 4B).
Distortion of data is very veiled when comparing the dynamics of changes in several variables that are very different. For example, the value of shares of company A increased 43 times: from 1 to 43 cu, of shares of company B - by 2.7 times: from 300 to 800 cu However, in the 5A comparison, the data are presented in such a way that company B seems to be the clear leader.
On the other hand, the predominance of company A is clearly visible in comparison with 5B (with a logarithmic scale). Logarithmic scales are widely used in science, but, unfortunately, have not yet found wide application in business. There are several reasons: many experts do not know how effective the use of diagrams of this type can be, while others often do not know how to interpret them correctly.
Let's analyze the diagram presented in comparison 5B. How can it be obtained from the initial version presented in comparison 5A? Everything is very simple. If you build a chart using Excel in the "Axis Format ..." you can set the value of "logarithmic scale", and the chart will immediately be rebuilt. And if you create it yourself, all the values ​​of the Y axis are rebuilt using a simple formula: new value = lg (old value). From here, the Y scale changes: the values ​​are 10, 100, 1000, 10 000, etc., that is, degrees of 10 (100, 101, 102, 103, etc.).
A lot of distortion causes misuse of the scales. The extremely popular technique is the intersection of the Y axis with the X axis not in the zero value. For example, a comparison of 6A demonstrates that the risk of loss of investment in company A is 3 times higher compared to company B, but the risk is actually the most by 2%, and this is well presented visually in comparison 6B, where the Y axis intersects with the X axis in zero value.
There are many other tricks. Among them are the stretching / compression of the axes, the use of different scales, the seizure of data from the context due to the display of a part of the diagram, when a picture is favorable for the speaker, etc. These techniques allow you to visually increase or decrease the significance of a parameter.
Cheating can also be achieved with the help of ordinary design tricks. For example, to diminish the significance of an object, you can reduce the contrast and illumination of the desired column, ruler, or sector. And if the diagram is presented at the presentation, you can choose the color of the object in such a way that at a distance it will simply “accidentally” merge with the background color and will not be taken into account.
It is recommended to be critical about the diagrams that speakers demonstrate. Always question their visibility and pay attention not only to visual perception, but also to specific numerical values.
Source: https://habr.com/ru/post/52417/
All Articles