Researchers hypothesize that patients who are married or have a partner are more likely to have improved health overall than those who are not married and not living with a partner. They believe that being married or having a partner leads to decreased stress levels and increased security, which in turn leads to improved health.

Specifically, researchers want to know the probability that patients who are married or living with a partner have a greater mean PROMIS score than other types of patients. They want to know this probability for comparing married/living with partner vs. each of the following groups of patients: 1) separated/divorced 2) widowed 3) single/never married 4) separated/divorced, widowed, or single/never married (all combined) Use a single linear regression model that can be interpreted appropriately for this question. Consider the goals of the analysis to determine whether other variables should be included in the model, and provide a brief justification of your decisions.