Statistical Methods for Testing Relationships Among Multiple Variables: Multiple Linear and Logistic Regression Analysis
Overview
This paper examines statistical approaches for identifying relationships among multiple variables, specifically focusing on multiple linear regression and logistic regression. The analysis demonstrates practical applications of these methods by predicting baseline glucose levels from BMI, age, and gender, and evaluating the significance of various biomedical variables in predictive modeling.
What this coursework covers
- Introduction to regression analysis methods comparing multiple linear and logistic regression approaches
- Question 1 analysis using multiple linear regression to predict glucose from BMI, age, and gender with ANOVA results and coefficient tables
- Question 2 beginning analysis of relationships between glucose, cholesterol, systolic blood pressure, and BMI (incomplete)
Preview
Testing to Find Relationships Among Many Variables
Helena Lynch
Capella University
MPH 5509 - Principles of Biostatistics
Dr. Haley Cash
December 13, 2020
Testing to Find Relationships Among Many Variables
Introduction
There are several approaches that can be used to understand the relationship between variables. Some of the most common methods is multiple regression and logistic regression. In regression analysis, the goal is to determine the best fit between two or more variables. Usually, the independent variable is a continuous variable, and hence it allows you to determine how one of more of the independent variables predict the values of a dependent variable. In multiple regression, it is the basis for establishing how multiple independent variables predict the value of a dependent variable. On the other hand, a logistic regression is a statistical model that utilizes a logistic function to model a binary dependent variable, although there are other complex extensions that exist. Logistic regression is used to estimate the parameters of a logistic model.
Question 1: Demonstrate how baseline BMI, age, and gender (variables: bmi1, age1, sex1) can be used to predict baseline glucose (variable: glucose1).
The most appropriate approach that can be used to predict the variables glucose is multiple linear regression. This is because multiple linear regression is a continuous variable and the predictor variable BMI, age and gender are a mix of both categorical and continuous variables. The general equation for a multiple regression that can be used to predict glucose is as shown.