Healthcare and Statistics: Examples - Term Paper Example

1. What is an example of “statistical abuse” in the media during the past week? Statistical abuse refers to the irresponsible use of statistics to represent significant numbers which are correlated with each other. According to Tarko (2006), people involved in statistics must use specific processes to show the relationship between factors and to explain the reason why such relationships exist. The media is one of the aspects of the community that sometimes become more prone to unreliably present relationships only to support a point. Although they must not be fully blamed on this, there are times that they should take the responsibility of the blame. This is very true when it comes to presenting statistically-defective health information as it can induce confusion and even hostility from the public that receives them. One of the latest issues when it comes to healthcare that is more or less statistically abusive is hailed in a report that discusses about the three prominent “healthcare leaders” talking about a study that allegedly found 969 cases of “preventable infections, bone breaks, bed sores and other harm” caused by the admission of these people to some of Nevada’s hospitals (Allen, 2010). There is, indeed, urgency in this information as it gives an idea about the true situation of the standard of healthcare of the specified state. Yet, it is deemed as a statistical abuse since the reporters did not indicate the two or more sides of the story but one. They too have solely depended on hospital records and did not even specify how they have accessed those supposedly-confidential documents. What makes it more complex is the fact that the primary goal of the conducted study was to expose the hypothetically-based Nevada’s “healthcare can hurt you” theory (Allen & Richards, 2010). Lastly, those who have conducted this research did not first consult the appointed statistical agencies to put the study under review, clarification and analysis for the purpose of validity. Instead, a newspaper article was eagerly made for the consumption of the many. 2. When would random sampling not be the best approach to sample selection? Give an example. Random sampling is a randomized technique of selecting samples from a certain population to represent that smaller population in the face of a larger one (Dattalo, 2010, p. 4). Dattalo also emphasized that many researchers prefer this as a method of representation as it increases the external validity of the data sample. With this technique, moreover, “each individual in the population has an equal chance of being selected” (Gravetter & Wallnau, 2009, p. 167). Hence, bias is prevented; and thus, generalization can be deemed statistically safe and the result of the study will be highly regarded to be the sure representative of the “probable” truth. On the other hand, the use of this sampling method also has its own disadvantages. Firstly, some researchers may not choose the populations through a thorough randomized process (Landoll, 2006, p. 122). Some simply have personal biases that can alter the selection of the population which in turn may pose relevant threat on the result of the study conducted. Other hindrances can include the variability of the traits that are put under study. In research surveys, the samples must possess uniformity over the scheduled time frame. If there is significant variance in the samples, then the samples will be held incompetent to correspond to the outcome. One example of the random sampling’s insufficiency as a method can be observed in a research which includes assessment of the quality of healthcare provided in one’s own workplace. Of course, if a researcher belongs to the same institution he or she is ought to measure the quality of, there would be a high possibility that bias in the side of the researcher may exist in choosing who to include in the survey as he or she already personally knows the subjects. Moreover, the consciousness of these people to the conducted research may blur the results by providing a proficient care during the time the study is going on which then would signify that the “probable truth” is not at all a “probable truth.” 3. What factors do you think would contribute to a distribution being skewed? A skewed distribution, according to Dever (1997), is a kind of asymmetrical distribution in which the majority of the collected data concentrate on a single side of the distribution line (p. 169). The author added that this is possible and is dependent on the existence of the practical and specification limits. Practical limits include “time measurements and counts of things” (Dever, 1997, p. 171). Time would, for example, pertain to the length of time a task in the research is needed to be. Limits to this can include rendering more time to observe a certain group of healthcare providers to measure the quality of the work done as compared to the prescheduled average time to do so. Counts of things could mean the amount of the things that are relevant in the fulfillment of the research such as the number of samples identified versus their availability to be part of it. On the contrary, specification limits are dictated mainly by the researchers before the study is started but are sometimes determined by the factors related to the subjects that the researchers may identify during the pilot study. These are set to render consistency and standardized quality of the data that must be collected from the subjects in relation to the specified requirements of data that the study must gather (Carpenter & Oglio, 1994, p. 3). If a researcher only opted to make use of the ward nurses of the morning shifts as subjects of his or her research, then it must be followed. 4. What are the definitions of means, medians and modes? Give an example of each that is reported in a health care organization. The mean in statistics is the “mathematical middle of the distribution” (Urdan, 2008, p. 18). It is the average of the total of the numbers in a certain distribution. A document in the World Health Organization site gave an example of the mean as it took note of the different hemoglobin (Hgb) levels observed in several individuals from both genders. They added all the Hgb levels of the males and the females separately and indicated that 13.2 was the average level of Hgb in males whereas it is 11.7 in females (“Describing and Analysing,” n.d.). Separately, the median is also a middle item of a distribution (DeFusco, McLeavey, Pinto, & Runkle, 2007, p. 81). However, choosing the median requires the arrangement of the numbers into an ascending or descending order. In a book, the Public Health Service of the United States (1992) exemplified this by stating that in an ascending pattern of age in years, the median age of the Americans would have to be more than 36 years old by the year 2000 (p. 2). The mode, on the other hand, is the item or number that recurrently appears in a set of numbers. In a study sponsored by the National Health Institute, it was reported that zero was the mode of the one-surface restoration variable of the active duty armies’ dental care needs (Parker, Schopper, Mangelsdorff, & Cheatham, 1979, p. 183). 5. What are the major differences between practical and statistical significance? Statistical significance is the consideration of the difference in the relationship between two or more variables using the numbers that represent each of them after the p-value was derived. Practical significance, distinctively, is the use of common sense in determining the significance of the difference in the variables (Schlotzhauer, 2007, p. 169). 6. Give one example of when practical significance would outweigh statistical significance? Explain your rationale. In an example given by Schlotzhauer (2007) in which a study aims to measure the difference in the quality of care provided by the hospital trainees and its experienced employees by initially conducting a written test with a possible score of 0 to 100 to identify the need of changing the training program of the institution (p. 169). The scores turn out to be at the average of 83.5 for the trainees and 85.2 for the experienced employees which would seem to have a statistically significant difference; but practically, the difference in the two average scores would deem insignificant to execute changes in the current training program (Schlotzhauer, 2007, p. 169) since it would be unrealistic to spend a considerable amount of time and money for the difference found. 7. Why is it important to learn what sample size was used in a study in which no statistical difference was found? According to StatSoft Electronic Statistics Textbook, statistical significance can only be identified if it “represents the probability that a similar outcome would be obtained” after an entire population is put to test (as cited in “Elementary Concepts,” n.d.). If the results show no statistical difference, it is important that the sample size be known as the probability of having the same result would only be significant if the sample size is large and would be otherwise if the sample size is small. A large sample size, even to the point of having the same size as the total population under study, is the only way to prove the absence of statistical difference; a smaller size would not efficiently mean that there is a lack of statistical difference as it does not represent the actual total population. 8. What is an example of two variables that are likely to be correlated because they are both changing over time? It is possible to find the correlation between the variables climate change and its effect on human beings’ health. It is apparent that these two can change over time. By examining this, healthcare providers will be able to identify what the health policies must be applied on specific times as the climate helps dictate the health of the population (World Health Organization, 2003); thus, help influence the healthcare quality and are correlated. 9. A study of employees at a large company found a negative correlation between weight and distance walked on an average day. In other words, people who walked more weighed less. Would you conclude that walking causes lower weight? Can you think of an alternate explanation for the findings? Although it is indicated in the findings that the employees who have spent more time in walking weighed lesser than those who do not walk often, the act of walking itself cannot be concluded to have caused the lower weight due to a number of factors. Primarily, walking must be paired with proper diet so that excess weight of those who are overweight will be lost (Schofield, 2006, p. 9). Additionally, walking is mainly used to burn the excess fat and then tone up the muscles (Bach & Schleck, 2003, p. 25). Hence, for those who do not have extra fat and cellulites stuck underneath their muscles, walking would only serve as a toner and not cause weight loss; instead, it only maintains it. Other factors may include initial weight prior to the study, individual metabolism rate, and dedication to the walking program specified by the researchers. Individual capability to perform the same level walking exercise is also a factor for this. An alternative explanation for the findings would be: The weight of a person is inversely proportional to the distance he or she travels by walking. References Allen, M. (2010, July 4). Health care leaders discuss Sun report’s findings. Las Vegas Sun. Retrieved from http://www.lasvegassun.com/news/2010/jul/04/willingness- work-improvement/ Allen, M., & Richards, A. (2010, June 27). Do no harm: hospital care in Las Vegas: health care can hurt you: Sun’s investigation of Nevada hospital data shows 969 incidents of inpatient injuries -- some that can be deadly. Las Vegas Sun. Retrieved from http://www.lasvegassun.com/news/2010/jun/27/health-care-can- harm-you/ Bach, M. L., & Schleck, L. (2003). Shape walking: six easy steps to your best body (2nd Ed.). Alameda, CA: Hunter House, Inc. Carpenter, C. A., & Oglio, E. R. (1964, July). Establishing specification limits by statistical analysis. Statistical methods for quality control of road and paving materials American Society for Testing and Materials, (Ed.). Baltimore, MD: American Society for Testing and Materials. Dattalo, P. (2010). Strategies to approximate random sampling and assignment. New York, NY: Oxford UP. DeFusco, R. A., McLeavey, D. W., Pinto, J. E., & Runkle, D. E. (2007). Quantitative investment analysis (2nd Ed.). Hoboken, NJ: John Wiley and Sons, Inc. Describing and analyzing research results. (n.d.). Retrieved from http://whqlibdoc.who.int/emro/2004/9290213639_chap8.pdf Dever, G. E. A. (1997). Public health practice: strategy and methods. Gaithersburg, Maryland: Aspen Publishers, Inc. Elementary concepts in statistics. (n.d.). Retrieved from http://www.statsoft.com/textbook/elementary-concepts-in-statistics/#Why significance of a relation between variables depends on the size of the sample Gravetter, F. J., & Wallnau, L. B. (2009). Statistics for the behavioral sciences (8th Ed.). Belmont, CA: Wadsworth. Landoll, D. J. (2006). The security risk assessment handbook: a complete guide for performing security assessments. Boca Raton, FL: Auerbach Publications. Parker, W. A., Schopper, A. W., Mangelsdorff, A. D., & Cheatham, J. L. (1979, March- April). Determination of the distribution of dental care needs of the active duty army. Public Health Reports, 94 (2), p. 183. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1431799/pdf/ pubhealthrep00137-0080.pdf Schlotzhauer, S. D. (2007). Elementary statistics using JMP. Cary, NC: SAS Institute Inc. Schofield, R. (2006). Weight-walking: a new path to health and fitness. USA: BookSurge, LLC. Tarko, V. (2006). The use and abuse of statistics: how to make a good psychological test. Retrieved from http://news.softpedia.com/news/The-Use-and-Abuse-of- Statistics-20979.shtml United States Public Health Service. (1992). Healthy people 2000: national health promotion and disease prevention objectives: summary. Sudbury, MA: Jones and Bartlett Publishers. Urdan, T. C. (2005). Statistics in plain English (2nd Ed.). Mahwah, NJ: Lawrence Erlbaum Associatess, Inc. World Health Organization. (2003). Climate change and human health -- risks and responses: summary. Retrieved from http://www.who.int/globalchange/summary/en/index9.html Read More