A conceptual model is a visual or written description of predicted correlations between various variables in a certain set up. Conceptual models are usually characterized with causal relationships, where the person modeling tries to establish the connection between different variables as well as the causes of certain effects observed between the variables (Haanel, 2007). In this scenario: “the consumption of ice cream increases, so do the instances of drowning”. A causal relationship is being purported, in which it is suggested that the increase in ice cream consumption causes a direct increase in the cases of drowning. In this case the consumption of ice cream is taken to be the independent variable and the drowning incidences as the dependent variable. In that, the any change in drowning cases depends on ice cream consumption, suggesting that if the consumption rose the drowning cases also rise, however; if the consumption declines the drowning cases also decline. Out of the normal, human intuitive nature, one is likely to believe that there is a direct relationship between ice cream consumption and drowning cases. This is because of the normal human nature of trying to draw causes for every observed happening or event (effects). However, statistically we know that correlation does not necessarily mean causation (Anon. 2010). The occurrence of two distinct events with variables changing in a common direction does necessarily mean that they have a direct relationship. Therefore, this empirically observed relationship can be explained in three ways.
Firstly, this occurrence may be as initially assumed-that an increase in the consumption of ice cream indeed increases the cases of drowning. Thus its reduction would automatically reduce the drowning cases. In this case on variable is dependent on the other. On the other hand, in actual sense there might be no correlation between the two variables whatsoever. The observed or assumed correlation may not be real, but a rather a factor of a third variable unknown that may be affecting the two variables (ice cream consumption and drowning). Therefore, in the second case we may say that the two variables have either a direct or indirect relation to this third variable, but they actually tend to move in the same direction (the increase direction) (Haanel, 2007). Thirdly, the occurrence of the seemingly direct correlation may be as a matter coincidence or chance seen through an appeal of chronology. In this case the two variables may have happened to vary in the manner they did not because there is any relationship, but due to mere coincidence and a common occurrence in time (Haanel, 2007).
The probability of the first scenario being true depends on whether there is indeed a direct relationship. If correlation cannot be established the relevant coefficient then there is zero probability of scenario one being true. This can be tested by varying one variable in this case the ice cream consumption to find out whether there is a change in the second variable (drowning). The second scenario depends on whether a third factor can be found that will have a correlation and a coefficient that will relate to each variable independently of each other. This can be analyzed by establishing a third factor that may relate to both. If there is no found third factor the probability of scenario two is nil. Finally, the probability of the third scenario occurring entirely depends on the first two. If none of them can explain the relationship or occurrence then chances are high or the probability is high for scenario three being the explanation for the occurrence.
Anonymous, (2010), The Illusion of Cause-Vaccines and autism, retrieved on 20th November, 2010 from http://geraldguild.com/blog/tag/illusion-of-cause/
Haanel, F.C. (2007), Cause and Effect Kessinger Publishing, LLC.