Scientists can now see the causality function that contributing factors play in almost any system from a more systematic perspective thanks to a brand-new method for determining causation.
Cause and effect. This idea is something that we are already aware of. Tug on a draw doll’s series, and the toys follows. Obviously, things get much more complicated as a program grows, as the number of factors increases, and as sounds enters the picture. Sometimes it becomes nearly impossible to determine whether a factor is actually having an effect or just a function of correlation or association.
Consider an example from weather research. Researchers who study big atmospheric flow designs and their effects on global climate would like to know how these techniques might alter as the weather warms. Here, several variables come into play: lake and weather temperatures and pressures, ocean tides and depths, and even information of the moon’s movement over time. But which factors have the most measured results?
Data theory serves as the foundation for causation in this context. A technique that can be used to establish causation in such intricate systems has been developed by Associate Professor of Aerospace at Caltech, Adrien Lozano-Durán, and people of his class at MIT and Caltech.
The new scientific tool can identify the contributions made by each system’s individual variables to a recorded effect, both independently and, more important, in combination. The group describes its new technique, called synergistic-unique-redundant breakdown of causation ( SURD), in a report published immediately, November 1, in the journal Nature Communications.
Any circumstance where scientists are trying to identify the true reason or reasons of a recorded effect can be adapted using the new unit. That could include factors like what caused the stock market to decline in 2008, the role of several risk factors in brain malfunction, how coastal factors affect the people of some fish species, and what mechanical factors play a role in a material’s failure.
” Causal inference is very multidisciplinary and has the potential to drive progress across many fields”, says Álvaro Martínez-Sánchez, a graduate student at MIT in Lozano-Durán’s group, who is lead author of the new paper.
For Lozano-Durán’s group, SURD will be most useful in designing aerospace systems. For instance, by identifying which variable is increasing an aircraft’s drag, the method could help engineers optimize the vehicle’s design.
” Previous methods will only tell you how much causality comes from one variable or another”, explains Lozano-Durán. Our method’s ability to capture every detail of what is causing an effect is what makes it special.
Additionally, the new approach avoids identifying causalities incorrectly. This is largely due to the fact that it goes beyond simply determining the independent effects produced by each variable. In addition to what the authors refer to as “unique causality”, the method incorporates two new categories of causality, namely redundant and synergistic causality.
When more than one variable results in a measured effect, redundancy in causality occurs, but not every variable is required to produce the same result. For instance, a student can receive a high grade in class because of her diligence or her intelligence. Both could result in the good grade, but only one is necessary. The two variables are redundant.
Synergistic causality, on the other hand, involves multiple variables that must work together to produce an effect. Each individual factor wo n’t produce the same result on its own. For instance, a patient takes medication A, but he does not recuperate from his illness. Similarly, when he takes medication B, he sees no improvement. But when he takes both medications, he fully recovers. Medications A and B are synergistic.
SURD mathematically breaks down the contributions of each variable in a system to its unique, redundant, and synergistic components of causality. A conservation-of-information equation, which can then be used to determine the existence of hidden causality, i .e. variables that could not be measured or that were perceived to be unimportant, must be satisfied once the sum of all these contributions is reached. The researchers are aware that the variables they included in their analysis need to be reconsidered if the hidden causality turns out to be too large.
The team used SURD to analyze 16 validation cases, scenarios with known solutions that would ordinarily present significant difficulties for researchers trying to determine causality, to test the new method.
Gonzalo Arranz, a postdoctoral researcher at the Graduate Aerospace Laboratories at Caltech, who is also the paper’s author, says,” Our method consistently gives you a meaningful answer in all these cases.” Other methods combine causalities that are improper, and they occasionally get confused. They get a false positive identifying a causality that does n’t exist, for example”.
The team used SURD to investigate the creation of turbulence as air circling a wall in the paper. In this case, air flows more slowly at lower altitudes, close to the wall, and more quickly at higher altitudes. Some theories of what is happening in this scenario have previously suggested that the higher-altitude flow influences what is happening close to the wall rather than the opposite. The opposite is believed to be true, according to other theories, which include the air flow close to the wall and what is happening at higher altitudes.
” We analyzed the two signals with SURD to understand in which way the interactions were happening”, says Lozano-Durán. ” Causality, as it turns out, comes from the far-off velocity. In addition, there is some synergy where the signals interact to create another type of causality. This decomposition, or breaking into pieces of causality, is what is unique for our method”.
