Correlation - cause and effect, Applied Statistics

Cause and Effect

Even a highly significant correlation does not necessarily mean that a cause and effect relationship exists between the two variables. Thus, correlation does not necessarily imply causation or functional relationship though the existence of causation always implies correlation or association between variables. It might sometimes happen that a fair degree of correlation is observed between the two variables but this is only due to chance as the theory might indicate non-existence of cause and effect relationship. Such a correlation is known as spurious correlation. While interpreting the correlation coefficient, it is necessary to see whether there is any cause and effect relationship between the variables under study. If there is no such relationship, the correlation observed is meaningless.

Figure 1

733_cause and effect.png

Ferrochrome is a material used in the manufacture of stainless steel. Suppose, in recent times there has been such a large increase in the supply of ferrochrome worldwide that its international price has dropped.

The simplified figure above reveals that as supply increased from S1 to S2, prices fell from P1 to P2. Hence, there is an inverse correlation between supply and price, that is, when supply increases, prices fall. Further we can say that the increase in supply was the CAUSE that led to the EFFECT of falling prices. So far so good.

Now consider the figure given below:

Figure 2

1313_cause and effect1.png

Here we have plotted supply against price. We may now be tempted to say that the fall in price from P1 to P2 was the CAUSE that led to the EFFECT of increase in supply from S1 to S2. This does not make economic sense because a fall in price does not lead to an increase in supply. The problem is that we have reversed the cause (increase in supply) and the effect (fall in price).

Continuing the example, suppose supply increases from S1 to S2, but demand increases by a disproportionately larger quantity. As demand increases more than supply the price will increase. In such a case if we plot price against supply the simplified graph would be as follows:

Figure 3

1900_cause and effect2.png

Hence, when supply increases from S1 to S2, price increases from P1 to P2. This is again poor economics because an increase in supply should reduce prices and not increase prices. The problem here is that the main CAUSE for the increase in price is the disproportionate increase in demand which is not shown in the graph. Hence, by ignoring the vital factor of demand and by comparing supply with price we come up with a spurious positive correlation between supply and price which defies good economics.

Posted Date: 9/15/2012 4:02:00 AM | Location : United States







Related Discussions:- Correlation - cause and effect, Assignment Help, Ask Question on Correlation - cause and effect, Get Answer, Expert's Help, Correlation - cause and effect Discussions

Write discussion on Correlation - cause and effect
Your posts are moderated
Related Questions
Problem : A company supplying electrical products, places a rush order for electric wires. Consignments of wires are to be sent immediately when they are available. Previous

Characteristics of Index Number  On the analysis of various definitions of index number the following may be its characteristics: 1.      Expressed in Number : Index number

As we stated above, we start factor analysis with principal component analysis, but we quickly diverge as we apply the a priori knowledge we brought to the problem. This knowled

In simple regression the dependent variable Y was assumed to be linearly related to a single variable X. In real life, however, we often find that a dependent variable may depend o

Charts when the Mean and the Standard Deviation are not known We consider the data corresponding to the example of Piston India Limited. Since we do not know population mean a

Measures of Dispersion Box 3: Food vs. Oil Below are the figures for foodgrain procurement   and cr

Using log(x1), log(x2) and log(x3) as the predictors, do pair wise scatterplots of all pairs of variables (including the response) and comment (use the pairs function). Do you thin

Statistical Process Control The variability present in manufacturing process can either be eliminated completely or minimized to the extent possible. Eliminating the variabilit


ogives graph