An advanced practice nurse (APN) manages an urban primary care clinic for the uninsured. She observes that patients with poorly controlled diabetes, defined as glycosylated hemoglobin levels (HbA1c) greater than 7%, often do not adhere to diabetes management and seem to suffer from certain mental health symptoms. In her literature review, she found an association between post-traumatic stress disorder (PTSD) symptoms and non-adherence to medical treatment in patients with heart disease, organ transplants, and HIV. However, she could find no published studies addressing the relationship in patients with diabetes.

The APN conducts a study at her clinic to investigate the association between HbA1c levels and PTSD symptoms. After getting informed consent, she has patients complete a survey questionnaire for demographic information and information about diabetes history and management. Height and body weight are measured and recorded. HbA1c levels are also measured. Patients are also assessed for depression and PTSD symptoms. A total of 50 patients are included in her study.

The APN analyzed her data to determine whether there was a difference in mean body weight for study participants with well-controlled diabetes (HbA1c < 7%) compared to those with poorly controlled diabetes (HbA1c > 7%). What would be an appropriate statistical test for this comparison and why

1. Analysis of variance (ANOVA)
2. Chi-square test
3. Independent samples t-test
4. Paired t-test

Which test would I use for the comparison? I believe its Independent samples t-test but Im not sure. Can you please explain this to me?