Essay #1. The research questions for this experiment involved the relative effectiveness of a shot and a nasal spray in keeping people from getting the flu. The hypothesis appears to have been that there would not be a significant difference between the two, with an eye toward proving that the nasal spray is just as effective as the shot. A nasal spray would be easier to deliver to patients and therefore more salable. 16 percent of the people who took the shot ended up getting the flu, while 24 percent got the flu in spite of the shot. This is a significant level statistically, with a p value of 0.0008. The null hypothesis would be that there is a significant difference, and in this case, the researchers would fail to reject the null hypothesis.
While the sample size appears to have been appropriate for this study, there are some questions that the prompt does not provide the answer to. First, a breakdown of the samples by age and gender would have been helpful. The samples are large enough to overcome many of the limitations associated with those predictors, but knowing the numbers involved would have been helpful. After all, people are more likely to come down with the flu when they are very young and again when they are old, so two populations that skewed significantly differently in those directions could render the results less meaningful than the numbers indicate. Also, the existing health of the people would be worth knowing. Someone who catches the flu every year is more likely to end up falling sick again, even with a vaccination.
Practical and statistical significance are two completely different concepts. Statistical significance can lead to assumptions that turn out not to be true under practical considerations. For example, there could be a statistically significant correlation between ethnicity of players in a soccer tournament and the number of goals that each player produces. One might look at statistics from an international soccer tournament and determine that South American players are better than Asian players, because South American players scored many more goals than Asian players did in the tournament. If one is a casual fan of soccer, one might also know that soccer is more prominent in South American cultures than it is in Asian ones as a pastime, both for playing and spectating. However, it would be a mistake to go a step further and exclude Asian players in favor of South American ones in all cases, because there are some Asian players with a great degree of skill. In practical terms, it would be much more useful to conduct a player tryout instead of simply eliminating players with Asian surnames in favor of players with surnames that appear to be of Spanish or Portuguese (the two predominant languages in South America) descent. Making that sort of exclusion could lead to a team with much less quality, as a tryout would allow coaches to match players with available openings. If going with statistical trends were enough in sports, coaches would never need to hold in person tryouts. However, the fact that teams continue to go to the trouble of holding those tryouts means that practical significance has elements that statistics simply cannot provide.
Essay #2. My first response to the finding is that the researcher provides no information about the sample size of the test population. While a correlation of 0.75 or greater is generally considered a strong one, there are several questions that need to be answered as well. A correlation at this level would tend to support the connection between high IQ and high GPA, this is not the most meaningful set of results, at least not in isolation. The implication, of course, is that students with high IQs will have high GPAs, but there are several variables that might also in play. First, educators have identified the phenomenon of the “high achiever” when it comes to academic achievement. Having the status of a high achiever has little to do with high IQ, but it is closely connected to getting high grades. Also, there are many factors that can keep children with high IQ from receiving high grades. Many who fall into the category of “gifted and talented,” but the divergent thinking that goes along with that status often keeps students from getting higher grades, because they do not produce the answer that the teacher wants. Also, studies have shown that socioeconomic status has a solid connection with performance in the classroom. A child’s IQ is independent of socioeconomic status, but the sort of nurturing that turns a high IQ into high grades is often missing for students of lower socioeconomic status.
All of this is to say that that correlation and causation are a long way apart. Just because two factors show a correlation does not necessarily mean that one causes the other. For example, many people who are excellent distance runners are also short. However, there are no studies that show that quality in distance running keeps one short or that being short causes one to develop the traits of a good distance runner. The factor that one has less body mass and a shorter pair of legs to work into a running rhythm are positive factors when one is trying to start a distance running career, but the difference between correlation and causation becomes clear.
With that said, the fact that many students with high IQ also have high grades does not mean that one predicts the other. There are many redheads who sunburn easily, but again, having red hair is not a predictive factor. Instead, combining factors such as socioeconomic status and high IQ could also predict a high GPA. Those two factors combine to produce more high achievers than either factor alone. A two-factor test would go a long way toward identifying students that are more likely to end up with high grades on their report cards.
This research finding is an excellent example of the limits of statistical research. While statistics show a lot, they do not show everything, and there are times when an apparent correlation can be mistakenly extended to connote causation. The end result, though, can turn into mistaken assumptions. In the present case, this assumption could be used to assume that students with high IQ do not need as much intervention from their teachers, even when their grades begin to slip, because their high IQs mean that they are more likely to right the ship and begin achieving at a high level again. The fact that so many other factors play a part in academic performance is often frustrating for education policy planners, who are looking for a solution that will make producing high grades easier for teachers. However, such things as home life, motivation, psychological issues and other factors are beyond the purview of the social planner, but because they affect grades, they are a part of this equation. Expecting students with high IQ to automatically get high grades can be frustrating to them, as they feel like others are putting assumptions on them that are unfair. Resolving all of these is an important goal for the educational establishment, but easy correlations like the one in this hypothetical situation are not helpful toward that end.
Essay #3. After splitting the samples into lower and higher halves, the lower group has a smaller range, standard deviation, skew and kurtosis. However, doubling the groups after the splitting does not change any of the statistical values. Simply cloning the findings over and over again does not add anything meaningful to the information. However, this does not mean that sample size is insignificant.
First, sample size tends to eliminate many of the other limitations that take place within a statistical study. This is why the television ratings calculators do not need to place a box on every television throughout a country. Instead, it is just necessary to put enough in the number of homes that will deliver statistically significant outcomes. Factors like age, gender, ethnicity and other identifying factors tend to become less significant when the sample size becomes large enough. This is why the sample size is important. Whether one is performing a quantitative or qualitative analysis, having enough respondents is important to give the findings from a study validity.
Research Analysis.
The paper under study is “A Quantitative Review of Ethnic Group Differences in Experimental Pain Response: Do Biology, Psychology and Culture Matter?” by a team led by Bridgett Rahim-Williams. The method consisted of a systematic literature review and an analysis of experimental pain studies that used stimuli to measure pain sensitivity across multiple ethnic groups. Covering the time period between 1944 and 2011, the researchers used the PUBMED database of articles, which includes more than 17 million citations. Effect sizes, racial/ethnic group categories, pain measures and stimuli and findings about factors of psychology, sociology and culture that all contribute to group differences.
The researchers discovered 472 studies analyzing ethnic group difference and the presentation of pain. 26 of the studies lived up to the inclusion criteria of investigating the experimental nature of the pain. Most of the studies dealt with comparison between non Hispanic whites and African Americans. The results showed consistently moderate to big sizes of effect for tolerance of pain across several different stimulus modalities. In general, African Americans showed a lower tolerance for pain. With regard to pain threshold, findings generally went in one direction, but the effect sizes were varied across ethnic groups, and only limited information was available for pain ratings of a suprathreshold nature. The conclusion was that potentially significant ethnic and racial group differences exist in perception of experimental pain. Clarifying these ethnic group differences have translational value for providing clinical care that is culturally competent and for reducing disparities in pain treatment among groups that are ethnically or racially diverse.
Taking a look at this study from a sociological perspective, it is worth asking what the value of the findings would be. After all, knowing that members of certain racial and ethnic groups have responded in certain ways to experimental pain tests does not necessarily provide value for individual care plans. The fact that African Americans, in general, may or may not have responded to experimental pain studies in such a way as to show a larger sensitivity to pain does not mean that the African American in the care of a particular nurse on a particular day wil have that same degree of sensitivity. In fact, these sorts of findings could create a culture of expectation within the care facility that could end up being corrosive for both caregivers and patients. If caregivers have the sense that a patient is going to be sensitive to pain because of his or her ethnic background, that could negatively impact their treatment. On the one hand, the caregiver could end up evincing anger about the extra steps that appear to be necessary because of the sensitivity to pain. If patients became aware that there were different treatment protocols for patients from different ethnic backgrounds, that could create a culture of expectation among the patients, and the psychosomatic effects could end up leading to the increased perception of pain among other African Americans (or whichever group the caregivers are treating in a different way). Over time, the results could end up becoming even more statistically significant, because of the simple effect of awareness.
This all begs the question of the types of research that would actually benefit people who are receiving care in a clinical environment. Instead of trying to figure out which gender or which race is more likely to have a higher sensitivity to pain, it might be more effective to identify changes to clinical environments that work well for all patients, or at least for a majority of patients, irrespective of ethnic background. After all, there are other predictors that are more likely to provide a more helpful answer as to which factors produce a heightened sensitivity to pain. For example, socioeconomic factors could be one, as those from lower socioeconomic levels might be less likely to seek medical care because of a lack of access to medical insurance. As a result, they simply have to live with many conditions and develop a higher tolerance for pain over time. Also, longevity of ancestors can also be a predictor of sensitivity to pain. Those with long lives in their family trees are more likely to show a general sturdiness and positive response to pain, rather than changing their lifestyle significantly when pain shows up.
Research studies that focus on ethnic and race in their findings simply end up feeding the old stereotypes that, for too many centuries, informed the way that people interact with one another. Even if it turns out that one group reports a higher sensitivity to pain within a particular study, that does not necessarily have any clinical benefit for patients in a particular setting. The corrosive effect of stereotypes is more dangerous, over time, than any individual effect might be. The last three or four centuries of human history suffice to show the dangers of organizing any sort of study on the basis of race.
While the study in question appears to have followed accepted methods in analyzing a set of articles to find a quantitative pattern, the lack of apparent value from the findings means that the project is actually much less valuable than the protocols might suggest. Taking the time to test experimental pain responses is admirable, but it is important to remember that it has only been a few decades since American scientists were conducting experiments on prisoners in Philadelphia’s Holmesburg prison that would primarily benefit commercial enterprise that looked at race as a category. The findings of such a report would likely have little value, and it is the recommendation of this writer that such studies find new areas of emphasis or study, rather than trotting out the old language of ethnic hatred.
Works Cited
Rahim‐Williams, B., Riley, J. L., Williams, A. K., & Fillingim, R. B. (2012). A quantitative review of ethnic group differences in experimental pain response: do biology, psychology, and culture matter?. Pain Medicine,13(4), 522- 540.
