Misleading Statistics: Vitamin E - Research Proposal Example

Misleading Statistics: Vitamin E 1. Vitamin E is a proven antioxidant and may help in fighting cancer and heart disease.” Is there anything ambiguous about this claim? It is undisputable that many antioxidants are applicable in fighting cancer and heart diseases. Fighting cancer and heart diseases have is a serious challenge to the contemporary world. Consequently, due to desperation there have been misleading statements in regards to treatment of cancer and heart diseases. It is true that Vitamin E is a proven antioxidant. However, it is wrong and ambiguous to state that, since Vitamin E is a proven antioxidant it can be used to fight cancer and heart diseases given that antioxidants are used to fight cancer and heart diseases. Therefore, the above statement is ambiguous hence it is an example of the misleading statistics. Ambiguity of the statement arises in connecting antioxidant with fighting cancer and heart diseases. No doubt Vitamin E is a proven antioxidant. What’s more, antioxidants can fight cancer and heart diseases. Nonetheless, there are no statistics indicating effectiveness and efficiency of Vitamin E in fighting cancer and heart diseases despite the fact that it is an antioxidant. In this regards, the main misrepresentation or misuse demonstrated is detached statistic. Detached statistics is a misrepresentation where a statement is made without any supporting data (Hughes, Lavery, and Doran 2010). In most cases, quantitative data is usually employed in order to give supportive evidence on a given claim or preposition. Hence, there is need to conduct a research and obtain enough data for purposes of answering research question. The above research should aim at establishing the correlation between Vitamin E and fighting of cancer and heart diseases. Once enough quantitative data has been collected, there is need to carry out a multi-regression analysis, which helps in identifying the correlations between Vitamins E and fighting of cancer and heart diseases. According to Gelman and Nolan (2002), multi-regression analysis is an inferential statistical tool that attempts to prove the relationship existing between more than two variables. In this case, Vitamin E will be independent whereas cancer and heart diseases will be dependent variables (Best 2004). Multi-regression analysis in this case will involve either the use of SPSS software of determination of the sum of squares between XX as well as between YY. Given the independent and two dependent variables, the following equation for multi-regression analysis is applicable X = ß1Y1 + ß2Y2 Where; X is Vitamin E Y1 is the cancer Y2 is the heart ß1 and ß2; are correlation coefficients indicating the relationship independent and dependent variables. The two coefficients of correlations, ß1 and ß2 can be found using the following formulas: 1 In conclusion, many people make ambiguous statements that are detached from their statistical proofs. It is usually important for statements especially in medical fields to be accompanied by relevant data in which the argument is based. For instance, in this scenario though some antioxidants have been used in fighting cancer and heart diseases, it does not follow automatically that Vitamin E being an antioxidant will be effective in fighting the same health conditions. In future, any claim especially in the medical field should be accompanied with relevant data. 2. In a recent article, the author states that 71% of adults do not use sunscreen. Although 71% is a large percentage, explain why it could be misleading. Many researchers have a weakness in making their conclusions following a study on a particular research. Previous studies have indicated higher number of adults not using sunscreen. However, this does not mean that all adults do not use sunscreen. In any case, most of these studies have based their conclusions on given samples, which may not be adequate enough to represent the entire population. Consequently, there is serious need for researchers to engage in finding appropriate confidence and significance levels that make their results authentic hence comparable to the whole population. Lower confidence and significance levels usually result into increased biasness of data as provided for by researchers. Most articles are usually developed after conducting a series of research or study that employ sample data. Sample data is used in representing a given population (Best 2004). The mistakes most researchers make is assuming that since the sample employed posses a given feature then it becomes obvious that the population will also have the same feature without providing a room for any critique. In this case, probably the 71% was the percentage of adults who were not using sunscreen in the sample. It will be misleading to conclude that the entire adults do not use sunscreen without giving proper significance level in which the study was conducted. What’s more, the researcher ought to have given a statement relating the conclusion with the sample (Gelman and Nolan 2002). For instance, the statement could be, “following the results of the study in which 71% of the respondents do not use sunscreen, it can be conclude d that majority of adults in the population do not use sunscreen”. In the above scenario, the misrepresentation of misuse of data arises in the suspect samples. It is obvious that the researcher employed a given sample in coming up with the data. Nonetheless, the authenticity of the sample should be known in order for the researcher to vehemently conclude that the exact percentage in the sample conforms to the exact percentage of the population. Best (2004) clearly explains that wherever a sample data is applied in making statistical inferences, there is need to ensure higher confidence and significance levels capable of making the conclusions or recommendations authentic. Confidence and significance levels are usually based on the following formulas: And; In conclusion, making statistical inferences through sampling requires that higher levels of significance and confidence are attained. This is the only sure way of ascertaining that the obtained data does not only represent the population but also applicable in making inferences and conclusions regarding subject of the research. In future, once results are obtained from research studies, there is serious need for such researchers to link the information to both samples and population. Using the results to refer entirely to population without any link to applied sample is a misleading concept that should not be encouraged within statistics. In any case, most recommendations and prepositions about given population are usually obtained through statistical analysis of data obtained from sampling process. References Best, J. (2004). More damned lies and statistics: how numbers confuse public issues. Los Angeles, CA: University of California Press, 2004. Print. Gelman, A. & Nolan, D. (2002). Teaching statistics: A bag of tricks. Oxford, NY: Oxford University Press. Hughes, W., Lavery, J., and Doran, K. (2010). Critical Thinking: An Introduction to the Basic Skills. Buffalo, NY: Broadview Press. Read More