Page 121 - Data Science class 11
P. 121
However, in correlation between variables, change in one variable does not automatically mean that it causes change
in the values of the other variable. Causation on the other hand shows that one event is the outcome of the occurrence
of the other event; i.e. there is a causal relationship between the two events.
Two Types of Causation
There are two types of causation: cause-in-fact, and proximate cause.
• Cause-in-fact: It is also known as actual cause. For example, when a bus strikes a car, the bus driver’s actions are the
actual cause of the accident.
• Proximate cause: It also means 'legal cause'. It is therefore recognised by the law as the primary cause of the injury.
It may not be the first event that set in motion a sequence of events that caused the injury. It may also not be the
very last event before the injury occurred. It is rather an action that produced anticipated repercussions without
interference from anyone else. In this case, the plaintiff will have to show that the injuries were the direct repercussion
of the proximate cause, without which the injuries would not have occurred.
There are three conditions for causality: covariation, temporal precedence, and control for 'third variable'. The latter
comprise alternative clarification for the observed causal relationship.
Causation can occur without correlation when a lack of change in the variables is present. Lack of change in variables
occurs most frequently with insufficient samples. Here is the most fundamental example: if we have a sample of 1,
we have no correlation, because there is no other data point to compare against. Therefore, there is no correlation.
How is Causation Calculated?
It is only from a suitably designed experiment that the causation can be determined. In such experiments, similar
groups receive different treatments, and the outcomes of each group are analysed. We can only conclude that a
treatment causes an effect if the groups have remarkably non-identical outcomes.
A causal argument is one that focuses specifically on how something has caused, or led to, some particular problem. A
causal argument is a necessary argument type, as people many a times find reasons as to why things have happened,
but may not be confident, or have all of the required information.
Causation Theory
Correlation or statistical dependence under plausible assumptions, can determine some causal relationships when
three or more variables are considered, even though it cannot decide the causal relationship between two variables.
Three Rules of Causation
Causal statements must follow three rules:
• Visibly show the cause and effect relationship.
• Use specific and correct descriptions of what occurred instead of negative and unclear words.
• Identify the preceding system cause of the error rather than the human error.
Agent causation is the concept that only 'things' have the power to change the world. Agent causation requires the
specification of a 'thing' that caused the effect. For example, the brick broke the window. Modern versions of agent
causation are largely restricted to purposeful agents that are self-directed (teleology). The brick did not break the
window; the person throwing it did. Agent causation has been largely superseded by event causation.
Cause and Effect Analysis
There is a cause for every effect, and therefore, things happen for a reason. This can be explained through cause in
science. The effect describes what happened. This is known as cause and effect analysis.
Cause and Effect Examples in Sentences
• A tornado blew the roof off the house, and as a result, the family had to find another place to live.
• Because the alarm was not set, we were late for work.
• Since school was cancelled, we went to the mall.
Assessing Data 119

