Causal Fallacies

Causal fallacies occur due to ignorance of the scientific method. The most common error is known as the 'correlation/causation error' -

Correlation Does Not Equal Causation.
This error is based on the assumption that two correlated phenomena have a causal relationship. This fallacy occurs when we assume that because two things have either a positive relationship (the more it rains, the more your knee itches) or a negative relationship (The more you watch tv, the less you exercise) that this means that one thing is the cause of the other. This is not necessarily true, for while correlation is a necessary condition for causality, it is not a sufficient reason for a causality. There are are two types of correlational/causational errors. They follow below:

Directionality Error
The direction between cause and effect is reversed. Saying that carrying your umbrella with you makes it more likely to rain is an example.

Third Variable Error
This occurs when one takes correlational data and assumes a causal relationship. One thing is held to to cause another, when in fact they are both the joint effects of an underlying cause.

See also the section on "necessary and sufficient" causes.

False Cause
A false cause fallacy is any fallacy wherein one fails to demonstrate a causal link.

Post Hoc Ergo Propter Hoc
http://i602.photobucket.com/albums/tt102/hanniballecturer/candleinthedark/posthocergo.jpg

Also referred to as Non Cause Pro Causa, this is the fallacy found in ritualistic thinking. This fallacy occurs when events are seen to be causally connected simply due to the fact that they follow in temporal succession. In scientific research, this is referred to as the dreaded correlation/causality error


 * Example: "I asked Father Pio to pray for a cancer patient. 11 days later, he had a spontaneous recovering. Therefore Father Pio worked a miracle!" - Pope John Paul II.

Warning catchphrase to look for: In general, consider the scientific education of the arguer. Do they fail to present empirical evidence? Not refer to charts/data? Do they follow up naked assertions with statements like "therefore" as if the assertion was the proof? (I borrowed this from Chris Hitchens)  If so, the speaker may not understand the limits of correlational evidence.

Cum hoc ergo prompter hoc
http://i602.photobucket.com/albums/tt102/hanniballecturer/candleinthedark/cumhocergopropterhoc.jpg

Also known as concomitant variation, this causal fallacy is a special case of Pot hoc, in that the two events are said to occur simultaneously.

For example, it can be shown that global warming is negatively correlated with to the overall decrease in the number of pirates:

http://www.logicalfallacies.info/piratesglobalwarming.jpg

But the covariance does not imply causality.

Complex Cause
This fallacy is similar to post hoc ergo propter hoc, except that the putative causal entity given in the argument may in fact play a role. However, it is also likely that other causes exist, or that several factors must interact in order to bring about the effect. Example:


 * "Literacy rates have steadily declined since the advent of television. Clearly television viewing impedes learning."

This may be a valid point. However, the fallacy is to assume that television is the sole cause, or even the most important cause. The correlation/causality error may take place here, leading go the possibility of the 'directionality' error - i.e. that some other factor reduce literacy rates, which led people to watch more tv! Or it may be that there is the third variable problem at work - some other factor may have led to a reduction in literacy and an increase in television viewing. See also the "Texas Sharpshooter Fallacy"

Insignificance
Related to complex cause, here one entity is held to cause another, and it does play a part in the causality, but it is insignificant compared to other causes of the effect. You'll often see this fallacious claim in conjunction with therapies and medicines - hell, with just about any consumer product. Recall the old commercial phrase "part of a complete breakfast!" - (sure, so is the table cloth) that captures the concept quite well.

Shared Weakness
This is a specific type of fallacy of insignificance - it is also an example of "weak induction". The fallacy is commonly employed in politics, wherein a politician claims connection to a well respected politician's weakness, indirectly implying that they shared strengths. Dan Quayle's infamous claim of having the same experience and age as John Kennedy is an excellent example.

Fallacy of Regression
The regressive fallacy occurs when one fails to take into account that any measured phenomena will inevitably fluctuate for various reasons, and instead claims that some change over time MUST be be due to the cause of some purported independent variable. The tendency for any measurable variable to move toward the average expected measure for that variable and away from extremes was called "regression" by Sir Francis Galton. Today we call this phenomenon the regression towards the mean.

Nearly anything you measure will fluctuate over time - your mood, your physical well being, the stock market, sports teams performances, etc. Even if the measured phenomenon remained precisely the same, measurement error alone would indicate fluctuation! The regressive fallacy is one key argument against the legitimacy of the purported placebo effect. Simply put, there may be no real placebo effect at all... what is really occurring in medical research is that positive responses after placebo may be measuring regressive factors - such as natural healing over time, individual differences and measurement error.

Necessary vs. Sufficient conditions
A necessary condition is a condition that must be present to bring about some effect. A sufficient condition possesses, as one of its attributes, the necessary condition, along with other attributes that are not necessary to bring about the effect. The fallacy occurs when we mistakenly assume that the the secondary attributes of the sufficient condition are somehow necessary conditions.

Example: let's imagine that the desired effect we want is to shatter a window. We realize that force that exceeds the tensile strength of the window is a necessary condition for breaking the window. A hammer thrown at 30 mph. at the window is a sufficient condition for breaking a window. The fallacy occurs when we attempt to argue that is it necessary to possess something metal to break a window, or something of a specific size, etc.

Transductive Reasoning
The causal fallacies above are related in that they are both examples of transductive reasoning. Transductive reasoning is how pre-schoolers' reason. Because preoperational children (children typically under the age of 8 who have not yet reached Piaget's concrete operations stage, see my entry on Piaget) are not yet logical thinkers, their explanations are often based on collections of disconnected facts and contradictions. This means that young children tend not to reason from general laws to specific instances (deduction) or from evidence to general conclusions (induction) but from a specific events to other specific event (Transductive reasoning). Piaget called this transductive reasoning. Because their worldviews are not yet internally consistent, pre-schoolers tend to see causal connections between events simply because of their connections in contiguity. (Source: Development Through the Lifespan, Laura Berk, 1999)

What this means is that they form opinions on causal connections non logically. Here is a good example from the above cited text:

Mother: Why does it get dark at night? Child: Because that's when we go to bed!

Compare this answer from a four year old to the pope's statements from the false cause entry above:

Reporter: Why did the cancer disappear? Pope: Because father Pio prayed!

The transductive reasoning of a four year old is at work in both statements. In the child, it's cute. In the adult figurehead representing spiritual and intellectual leadership for a billion people, it's nearly beyond frightening.

Slippery Slope
http://i602.photobucket.com/albums/tt102/hanniballecturer/candleinthedark/slipperyslope.jpg

Another popularly known fallacy, this is actually an offshoot of the false cause fallacy. It occurs when an arguer claims one event must lead to a successive chain of less desirable consequences -without offering any other proof. There error comes from assuming that a current rate or progresion will continue on at the present rate indefintely.


 * Example: "If we vote for Clinton, a known pot smoker, soon the whole Whitehouse will be filled with drug addicts."

Fallacy of the Pre-Determined Outcome
This is a current favorite of mine, and I learned it from Michael Kay, a talented and intelligent sports announcer for the New York Yankees. Go Yankees!

Kay often runs into fans who say something akin to: "If only player x had gotten on base, then player y's home run right after x's at bat would have won the game."

Kay counters that we cannot assume player y would have hit the homerun in this hypothetical situation, because we have changed what led up to y's homerun so that all future outcomes are now in doubt - i.e. the playing field is literally changed. For specific examples, Kay would point out that after walking player x, the opposing pitcher might have been substituted, or that player y may have become more anxious knowing that the game was now on the line and struck out. In fact, aliens from outer space, who had placed wagers on the Yankee's opponents, may have chosen that instant to attack and annihilate Yankee stadium. There is simply no way to know for certain what would happen next. Whatever the specifics, we can never assume that an outcome would be exactly the same as it is now, if we were to hypothetically go back and change variables that led up to the outcome. The flaw in the assumption here is that the arguer is assuming that he can keep whatever outcomes that please him, while denying those that do not, when in fact the variables related to an outcome are probably interconnected.

Ecological Fallacy
Offering correlation evidence at at higher level of analysis for phenomena manifest at a lower level of analysis. For example research findings showing a positive correlation between the ideological liberalism of state's elected officials and the per capita conumsption of alcohol in a state cannot support the claim that "liberals drink more than conservatives."