Cum hoc ergo propter hoc

The articles Correlation # Correlation and causal connection , spurious correlation and Cum hoc ergo propter hoc overlap thematically. Help me to better differentiate or merge the articles (→ instructions ) . To do this, take part in the relevant redundancy discussion . Please remove this module only after the redundancy has been completely processed and do not forget to include the relevant entry on the redundancy discussion page{{ Done | 1 = ~~~~}}to mark. Xiooix ( discussion ) 15:58, Feb. 21, 2015 (CET)

Cum hoc ergo propter hoc ( Latin for 'with this, consequently therefore') denotes the fallacy of pseudo causality , in which the common occurrence of events ( coincidence ) or the correlation between characteristics isunderstoodas a causal connection without closer examination. But implies a correlation is not causation ( English Correlation does not imply causation ), even if the relationship seems like a causal ( spurious correlation ). Without a causal connection, however, cause and effect are arbitrarily assigned without a well-founded justification.If one wants to express that there is a fallacy according to the pattern cum hoc ergo propter hoc , one says cum hoc non est propter hoc (Latin: 'With this is not because of it.').

description

A coincidence of two events observed several times generally gives rise to the assumption that there could be a connection between these events. However, the temporal and spatial coincidence of two events does not establish a causal connection. Whether such a connection exists, and if so, which of the two events is the cause and which is the effect, or whether both events are the result of a third event, cannot be deduced from this. In statistics, instead of cause and effect, one speaks of dependency or, in the case of random variables, of correlation - even with a clear covariance in the course of the values of the variables, it cannot be ruled out that there is a third common factor or that the direction of action between the recorded values is different proceeds as expected.

There are several possible cases:

{\ displaystyle A, B}

A and B appear
together

{\ displaystyle A \ Rightarrow B}

{\ displaystyle A}

caused

{\ displaystyle B}

{\ displaystyle A \ Leftarrow B}

{\ displaystyle B}

caused

{\ displaystyle A}

{\ displaystyle A \ nLeftarrow B, A \ nRightarrow B}

{\ displaystyle A}

and are out of context

{\ displaystyle B}

{\ displaystyle x \ Rightarrow A, B}

A common cause has immediate and consequence

{\ displaystyle x}

{\ displaystyle A}

{\ displaystyle B}

{\ displaystyle y \ Rightarrow x, x ',}

{\ displaystyle x \ Rightarrow ... \ Rightarrow A, x '\ Rightarrow ... \ Rightarrow B}

A common cause starts two causal chains, into which other conditions can flow that ultimately cause and

{\ displaystyle y}

{\ displaystyle A}

{\ displaystyle B}

An impartial observer should generally assume that any of these cases could exist. In particular, the case without interaction can be excluded by statistical testing of the null hypothesis for data from repeated observation . A sample bias should be excluded, since this can also lead to spurious correlations between independent features.

A similar fallacy is post hoc ergo propter hoc (Latin; 'afterwards, therefore'). In contrast to cum , post hoc emphasizes that the supposed effect occurs after the supposed cause. Therefore - unlike with " cum hoc" - assumed cause and effect cannot be mixed up. However, the chronological sequence is also necessary, but not sufficient for a causal relationship.

Examples

illustration

Event : "In adolescence, chocolate consumption increases." ${\ displaystyle A}$
Event : "Acne is more common in adolescence." ${\ displaystyle B}$

Conclusion: "Chocolate causes acne in adolescents."

This statement concludes from the common occurrence of a cause-effect relationship (causality) of two events. Event is arbitrarily assumed as cause, event as effect. From the multitude of conceivable relationships, the first that comes best is selected without being checked. However, different connections are logically conceivable here: ${\ displaystyle A}$ ${\ displaystyle B}$

1	Chocolate consumption leads to acne disease, for example because of the ingredients in chocolate such as fats or sugar.	${\ displaystyle A \ Rightarrow B}$
2	Severe acne leads to a craving for chocolate, for example because of increased sebum production or to increase mental well-being	${\ displaystyle A \ Leftarrow B}$
3	Frequent chocolate consumption and acne disease have no common cause. Both are not uncommon, the simultaneity observed several times is a coincidence .	${\ displaystyle A \ nLeftarrow B, A \ nRightarrow B}$
4th	Chocolate consumption and acne have a third common cause, such as increased hormone secretion	${\ displaystyle x \ Rightarrow A, B}$
5	During puberty, identity is called into question and the associated uncertainty leads to increased stress reactions which, mediated by regions in the hypothalamus , increase the neuronal response to sweet taste stimuli on the one hand, which leads to more frequent consumption of chocolate if it is readily available, and on the other hand hormonally increase lipogenesis in sebum glands, which more often leads to acne disease when the skin flora changes.	${\ displaystyle y \ Rightarrow x, x ',}$ ${\ displaystyle x \ Rightarrow ... \ Rightarrow A, x '\ Rightarrow ... \ Rightarrow B}$

This fictitious example shows that no decision can be made between these alternative interpretations on the path of logic; Additional knowledge from suitable sources is necessary to substantiate or refute the claim made at the beginning.

Further examples

A recurring example in the mass media is a correlation between aggressive behavior and the duration of daily occupation with computer games ; A ban on " first person shooters " is even regularly called for. Other, possibly both underlying social or personal factors are ignored.

An almost classic example is the correlation between the return of storks and the increase in the number of births in spring, or, in a more modern version, the decrease in stork nests in Europe and the decrease in the birth rate there . Per cum hoc one could conclude that the storks are causally involved in the birth or, for example, “bring the babies” . The common cause is the economic and social changes in Europe. More intensive agricultural techniques as well as increased settlement and road construction on formerly agricultural areas are affecting the storks' habitat. Although z. For example, in Germany you can find a connection between the number of births and the number of storks across the regions, but this is probably due to the degree of urbanization: In rural areas there are relatively many storks and more children are born in the city there are relatively few storks and fewer children are born.

Another example that is occasionally mentioned in textbooks is that the world population is apparently linked to the age of the British Queen: both sizes increase every year. A common cause occurs again: time.

According to the fictitious after the Bundestag deputies Jakob M. Mierscheid named, conceived in 1983 as a satire mierscheid law corresponds to the percentage of votes the SPD (in percent) in the year in a federal election the index of German crude steel production (in western Germany), measured in million tonnes . In the case of early federal elections, the raw steel values of the original and actual election year are to be averaged. Even if the parameter “crude steel production” was selected at the time the law was formulated on the basis of purely quantitative coincidence, the “law” also reflected the SPD election result with astonishing accuracy at the next election. Any relationships that can be traced back to a third variable, e.g. B. that in the case of an improved / worsened economy, both crude steel production and the SPD share of the vote may increase / decrease, are possible but not mandatory.

The Impact of Decreasing Pirate Numbers on Global Warming

A now famous, current example of cum hoc ergo propter hoc as an ironic and instructive stylistic device by critics is the statement of the physicist Bobby Henderson that the only cause of global warming , hurricanes and all other natural disasters is the falling number of pirates since the beginning of the 19th century . Century is responsible. The Gore effect is an ironic term for untimely snowy weather or cold spells in connection with events about and demonstrations against the dangers of global warming . According to CNN's Bob Marciano, this has already established itself as a running gag among experts and in the media environment . “Just bad timing. Whenever there is a corresponding climate conference, there is a cold outbreak. ”As with the Pauli effect, the exact background is unknown.

The Okun'sche law describes a correlation between economic growth and unemployment . From this correlation z. B. Statements regarding the employment threshold (necessary economic growth to prevent rising unemployment) are derived without showing a causal relationship, which is necessary for the formulation of meaningful statements.

Another example of cum hoc ergo propter hoc is the implicit assumption of a causal relationship between the property being investigated and a disease when a risk factor occurs in medicine. The risk factor is not proof of the existence of such a relationship, but merely indicates how much more often a disease occurs in a group with the corresponding characteristic compared to a control group.

literature

David G. Myers : Psychology . 2nd, expanded and updated edition. Springer, Heidelberg 2008, ISBN 978-3-540-79032-7 , pp. 30-35 ( Springer textbook ).

Web links

Tyler Vigen: Spurious Correlations. Retrieved on September 8, 2014 (English, collection of statistically determined correlations without causal relationships).

Individual evidence

↑ Eg violence increases aggression. on: sueddeutsche.de February 20, 2014.
↑ Playback of a radio recording of the American Morning Program on CNN, January 5, 2010

[1] Eg violence increases aggression. on: sueddeutsche.de February 20, 2014.

[2] Playback of a radio recording of the American Morning Program on CNN, January 5, 2010