Correlations normally show the relationship between two variables. The three types of correlations are either positive, negative or no correlation at all.
Correlation implies that there is an existence of some definite relationship in two or more variables such that if for instance two quantities vary in a manner that movement in one is accompanied by movement in the other, these two quantities are said to be correlated. What correlation therefore does is to determine the degree of relationship that there is in two or more variables. It does not bring out the cause effect relationship between variables identified to have a correlation. The risk of causal conclusions is likely as follows:
- The correlation may be as a result of pure chance if small samples are used. One may fail to find any relationship in the universe for these same variables.
- The correlated variables may be under the influence of one or more other variables and therefore wrong conclusion arrived at.
- Both variables could be mutually influencing each other such that none can be designated as the cause and the other as the effects.
The behavior under study is divorce which has been linked to poverty, family background of the couple as to whether their parents were divorced or not and finally the racial aspect. Other environmental factors that can affect behavior include: The economic growth which in most cases depends on other factors such as balance of trade, government subsidies and inflation. If the there is low economic growth, majority of people could end up living below the poverty line. The other factor is lifestyle changes. People change their lifestyles for healthy and economic reasons. The other one is religion which shapes the way of life for believers and actions considered right or wrong.
In the first case, the researcher is linking divorce to poverty line, in the second case, divorce has been linked to social economic status. In the third case, the researcher is linking divorce to the marriage status of parents to those who are in the marriage under consideration. Finally, the last case, divorce is now being linked to death rate. In all these four cases, the researchers have ignored other causes of the outcome under study, only relating two variables at a time.
A variable is defined as anything that changes over a given timeframe or space for instance time, temperature and production. In the first case, where divorce has been linked to poverty, the researcher held constant the divorce variable. In this case, the researcher is implying if the economic status of a married couple changed and they begun living under poverty line, their chances of divorcing could me more. In the fourth case, where death has been linked to divorce, the researcher has held constant the divorce variable and thus arguing that those who divorce are likely to increase their chances of dying early.
The chance that an individual picked out of 1000 government employees could be a spy can be determined in two approaches. The first one is through probabilistic thinking in which case, the chances are 95 out of 100,000 chances which is 0.00095 chances.
The second approach is through human judgment which is based on the past experience. In this case, chances of the picked individual being a spy could be at 95%. This is because, according to the principal of humanity, everyone is basically the same and as such, the likely hood of an individual being a spy is at 95%.