Discussion Posting

Provide a response to each posting listed below. Be sure each response is at least 200 words. Go in detail and provide more insight to the discussion.

1st Post: Parametric methods require specific rules to be observed, such as normal distribution of values, random sampling, a sample size of 30 and above (with some exceptions), and populations with similar variances (Pett, 2016; Schober & Vetter, 2020).  

Nonparametric statistics is used when rules for parametric statistics are violated because nonparametric methods do not require a strict assumption about the population from which samples are drawn (Schober & Vetter, 2020). Nonparametric tests also have some assumptions, such as independence of observations (except paired data); however, they have lesser requirements for population distribution. Nonparametric tests focus on rank ordering of ranks, medians, or frequencies instead of population means. Nonparametric testing also has less strict requirements regarding sample sizes. 

Despite having high potential, Pett (2016) states that nonparametric tests are underutilized in healthcare research. Nonparametric statistics should be considered in healthcare because it could be used with small sample sizes, categorical or ordinal-level data, and the acceptance of unusual and irregular sampling. Pett suggested the use of nonparametric statistics in the following situations: 

· When independent and dependent variables are nominal

· When data are ordered

· Rank-ordered data

· Unequal or small sample sizes

· Nonnormal distribution of dependent variable

· Unequal variances across groups

· Notable outliers in the data

From a clinical perspective, nonparametric statistics suit researchers with limited budgets, small samples or convenience, and a primary focus on improving patient outcomes in local settings. 

Here is an example of using Pearson chi-square in nursing research (Eion, 2019). The OR nurses wanted to know if a relationship between cold exposure and fever exists. The expected outcome is that most patients (at least 90) will develop fever from cold exposure. Out of 100 patients, 75 had a fever after cold exposure, and 25 had a fever without cold exposure. Using chi-square, OR nurses determined that her null hypothesis was correct, and exposure to cold could cause fever in 47% of patients. The practical application would be protecting patients from hypothermia during surgical procedures to avoid postoperative fever. 

The second example is using a chi-square test in a randomized trial to compare the incidence of bradycardia after phenylephrine versus norepinephrine in women during c-section (Schober & Vetter, 2019). After comparing two groups of women, authors determined a lower incidence of bradycardia in women receiving norepinephrine. That finding would help lower the incidence of hypotension in women who receive spinal analgesia during c-sections.

2nd Post: Statistics is a challenging field of study. It requires specific baseline skills in mathematics, logic, and critical thinking. I would not be able to complete assignments if I did not have a strong foundation from courses, I took in high school and college. The book written by Ravid (2020) was well-organized, and the chapters supplied needed information. However, in the last two weeks, the author dived too deep into calculations and equations. The supplemental readings and videos were beneficial, especially the video in week 7. I loved the website called Statistics by Jim because the author clearly explains complex concepts and illustrates explanations with graphs and tables. 

I liked that the course focused on concepts and statistics frequently seen in research articles. After taking this course, I feel confident about interpreting findings in educational and medical research articles. I appreciate that discussion questions and assignments focused on practical application of learned material which could help me solve real-life problems. Overall, I feel more competent in statistics after taking this course.