Good day,
I am a medical student currently working on a thesis focused on metabolic syndrome. My sample size is 45 individuals. I have encountered multiple challenges in assessing the normality of my data.
One of the variables appears to exhibit symmetry when visualized on a histogram, and the data closely align with the reference line on the quantile-quantile (QQ) plot. Additionally, the box plot displays symmetrical characteristics. However, when subjected to the Shapiro-Wilk normality test, the p-value obtained is 0.005.
Another variable displays a noticeable positive skew on the box plot, but the corresponding Shapiro-Wilk test yields a p-value of 0.227.
A third variable exhibits a positive skew on the QQ plot, and the histogram does not resemble a typical bell-shaped distribution. Surprisingly, the Shapiro-Wilk test for this variable produces a p-value of 0.415.
These are the details of my thesis: I have at my disposal a dataset containing information about individuals who have experienced acute coronary syndrome, and among them, some have been diagnosed with metabolic syndrome while the rest do not exhibit this condition. Within this dataset, I am working with seven continuous variables, which encompass age, waist circumference, triglycerides, HDL (High-Density Lipoprotein), fasting blood glucose, systolic blood pressure, and diastolic blood pressure. Additionally, there are nine categorical variables, including gender, marital status, education level, employment status, whether individuals are receiving treatment for elevated serum triglycerides, treatment for low serum high-density lipoprotein cholesterol, treatment for hypertension, treatment for diabetes, type of acute coronary syndrome, and metabolic syndrome status. As for the comparison between the two groups, one consisting of individuals with metabolic syndrome and the other without, using categorical variables, I plan to apply the chi-square test of independence. However, when it comes to comparing these two groups concerning continuous variables, my selection between the Mann-Whitney U test and the independent samples t-test will be contingent upon the results of the normality test conducted on these variables. I would appreciate your assessment of the validity of this approach.
I would greatly appreciate your help in analyzing and understanding these results, if you don't mind.
