Introduction
When examining large amounts of data, easy correlations can show a great deal of details. However, it is a lot more significant to assess relations within the data, particularly in health sciences. Via relationship calculations and hypothesis analysis, such correlations can be examined in depth, restricted only by the information accessible to the researchers (Frank, 2000). This paper explores the effects of confounders on variables.
Basically, confounding variables refer to additional variables whose presence influences the variables being investigated so that any outcome obtained does not replicate the real relations between such variables (Sullivan, 2012). In any experiment the major query the researchers ought to ask themselves is whether X influences Y. Where X is the potential cause and Y is the effect. Treatment in X is believed to cause a change in the final outcome (McNamee, 2005).
Statistical description
The statistical test the researcher would use to test the null hypothesis would be:
- State hypothesis
Null hypothesis = presumed value of parameter (H0) – statistical hypothesis that we test (strength of our sample data vs. the null hypothesis).
H0: (µ = 500)
>>There is no relationship between fast-food hamburger consumption and BMI
Alternative hypothesis = contingent value if null rejected (H1) – statement that specifies what we think is true (usually the opposite of the null hypothesis).
H0: µ ≠ 500 OR H0: µ > 500 OR H0: µ < 500. >> There is a relationship between fast-food hamburger consumption and BMI
Obesity instead of BMI
Now if the dependent variable was yes/no obesity instead of BMI the statistical test the researcher would use to test the null hypothesis would be:
- State hypothesis
Null hypothesis = presumed value of parameter (H0) – statistical hypothesis that we test (strength of our sample data vs. the null hypothesis).
H0: µ = 500
>> There is no relationship between fast-food hamburger consumption and Obesity.
Alternative hypothesis = contingent value if null rejected (H1) – statement that specifies what we think is true (usually the opposite of the null hypothesis).
H0: µ ≠ 100 OR H0: µ > 100 OR H0: µ < 100. >> There is a relationship between fast-food hamburger consumption and Obesity.
Confounders and Their Importance
The cofounders that should be included in this analysis are the age set, the hamburger eating habit, and the nutrition awareness (nutrition education) of the population being studied. The two fits in the key question the experimenter is expected to ask him/herself. A good example would be to study whether nutrition education (Cause) is related to an increase of BMI in the population being studied (Effect). Of course, as expected, it is logical to think that the BMI would increase as a result of eating fast food hamburger since the level of cholesterol and fat content is high in this particular food (Walpole et al., 1989). But if you conducted the study, you would realize that those eating hamburgers in the population are less likely to increase their BMI. It would seem that eating fast-food hamburger resulted in less/ no BMI increase (Gareen & Gatsonis, 2003). But the fact is that you missed a significant important variable – each specific person’s nutrition education. If you put each specific person’s nutrition education into account you would then realize that those educated have a nutrition menu that they strictly adhere to. The hamburger could have been slotted in their meals menu as the source of protein and carbohydrates.
Conclusion
Individual awareness is associated with not only the cause but also the effect. The confounder that must be included in the analysis is the age. The other cofounders such as the hamburger eating habit and the nutrition knowledge also have their effect on the overall result of the variable in this experiment.
References
Frank, K. (2000). Impact of a confounding variable on a regression coefficient. Sociological Methods & Research, 29(2), 147-194.
Gareen, F. & Gatsonis, C. (2003). Primer on multiple regression models for diagnostic imaging research. Radiology. 229(2), 305-310. Web.
McNamee, R. (2005). Regression modeling and other methods to control confounding. Occup Environ Med. 62(7), 500-506. Web.
Sullivan, M. (2012). Essentials of Biostatistics in Public Health (2nd ed.). Boston, Massachusetts: Jones & Bartlett Learning.
Walpole, I., Zubrick, S., & Pontré, J. (1989). Confounding variables in studying the effects of maternal alcohol consumption before and during pregnancy. Epidemiol Community Health. 43(2), 153-161. Web.
