Logistic Regression Analysis of Social Environmental Factors and Smoking Behavior of College Students

Abstract

Teenagers are the people who should be paid more attention to in smoking control work, among which college students are a special group. It is of great theoretical and practical significance to study the smoking behavior and its influence factors of college students. This study constructed an index system of social environmental factors from four dimensions of family environment, peer environment, campus environment and off-campus environment, and applied binary logistic regression to explore the main social environmental factors affecting college students’ smoking behavior in the index system. The study tested the fitting and optimization effect of the regression model, and drew several conclusions through data analysis: social environmental factors are closely related to college students’ smoking behavior; family environment, peer environment, campus environment and off-campus environment all contain factors that have weak significant impact on college students’ smoking behavior; six indexes of social environmental factors have a significant impact on college students’ smoking behavior, among which the no smoking sign in universities is the key index affecting college students’ smoking behavior; college students’ rebellious psychology should be fully considered in smoking control actions, and the indexes of social environmental factors should be reconsidered based on this. The relevant conclusions provide a theoretical basis for the formulation of smoking control policies in universities, and further improve the pertinence and effectiveness of campus smoking control measures, which has guiding significance for the practice of smoking control for college students.