Correlation and Regression The equation Sleep = 3500 – 0.15 totwrk – 11.20 educ + 2.29 age + ε based on Biddle and Hamermesh (1990) multiple regression model indicates that there is a negative correlation between sleep and work. This is because one is traded of for the other. When one increases the other one decreases, as a person increases the total work, measured in minutes per week sleep also measured in minutes per week decreases.The coefficient of total work being negative indicates that if the total work increases by 1 minute per week there will be a decrease in sleep by 0.15 minutes per week.Factors of total work, education and age are not the only factors that affect sleep. The way these factors work is determined by factors such as families. Family people will tend to have fixed family and leisure time. These times tend to be constant and are fixed because they do not vary with either work or education. This means there is a point beyond which sleep or work cannot go. The equation would then need to be modified to account for variation of sleep due to family and leisure.Other factors that should be taken into