Correlation is a statistical test that measures the strength of the relationship between two variables. Correlation coefficients have values ranging from -1 to 1. Correlation coefficient with a value greater than zero is said to be positively correlated while a value less than zero is a negative correlation. A correlative of +1 implies the variables are perfectly related and -1 implies the variables are perfectly unrelated ( Cohen et al., 2014) . A strong positive correlation between student SAT scores and their family’s income does not mean that a family’s income causes the SAT grades to rise. Correlation does not have a causation effect; a strong positive correlation between SAT scores does not imply an increase in family’s income causes an increase in SAT scores.
However, it implies that an increase is SAT scores is strongly associated with an increase in family's income. Studies have shown that student's SAT scores have been a strongly associated with the level of family's income ( Wang et al., 2015 ) . Students from rich families are able to get access to resource-rich schools with small class sizes which are associated with good performance in SATs. However high salaries cannot make student pass exams since intelligence is inborn that cannot be bought.
Delegate your assignment to our experts and they will do the rest.
Correlation between high-income families and high SAT will, however, will not be that strong. An increase in SAT scores does not necessarily imply an increase in future income. Similarly, any correlation here whatsoever will not imply that high SAT scores will cause high income in the future ( Cohen et al., 2014) . Generally, the main limitation of correlation is that it does not measure causation. A correlation between two variables will not imply that an increase or decrease in one variable will cause an increase or decrease in the other as the case of SAT scores and family's income. Furthermore, correlation only measures linear relations. And finally, in the test for correlation we are only restricted to the given data, we cannot go beyond it.
References
Cohen, P., West, S. G., & Aiken, L. S. (2014). Applied multiple regression/correlation analysis for the behavioral sciences.
Wang, W., Arora, R., Livescu, K., & Bilmes, J. A. (2015, April). Unsupervised learning of acoustic features via deep canonical correlation analysis. In Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on (pp. 4590-4594). IEEE.
