In most circumstances that are practical in nature, outliers do reduce the value of a correlation coefficient and also enfeebles the regression connection. However, there is s possibility that in certain situations, an outlier may cause an increase in the correlation value and boost regression. For instance, influential outliers are those points in a set of data that have influence in the regression equation and at the same time improve the correlation (Wen et al., 2013).
Correlation in most instances does not imply that it also has an effect on causation. A seasonal example that supports the thesis that correlation does not indicate that one variable is responsible for the other is the effect of the period of winter in the United Kingdom. Just because folks in the U.K are inclined to spend more in the shops during the cold season and less during the summer does not imply that the cold causes an increase in the expenditure among the residents. The high spending could be because of the coincidence between Christmas and the cold season and the new year where people tend to spend more probably because of the festivities (Verhulst, Eaves " Hatemi, 2012).
Another supporting argument is the correlation between the sales of sunscreens and the sales of ice creams. Statistical data of the sales of the two commodities shows that their sales decreases and increases in a systematic manner in a year. However, the simultaneous changes in the annual volume of sales cannot be reason enough to conclude that there is a direct relationship between the sales of ice cream and sunscreen and that one causes the other. The connection could be due to the impacts of the season. For example, during the hotter season, there is an increase in the consumption of ice cream and also wearing sunscreen (Verhulst et al., 2012).
References
Verhulst, B., Eaves, L. J., " Hatemi, P. K. (2012). Correlation not causation: The relationship between personality traits and political ideologies. American journal of political science, 56(1), 34-51.
Wen, Y. W., Tsai, Y. W., Wu, D. B. C., " Chen, P. F. (2013). The impact of outliers on net- benefit regression model in cost-effectiveness analysis. PloS one, 8(6), e65930.
