Print - Back

C. Muse---- Project 3 REVISED
Generated Apr 30, 2018 by cmuse28

   

Christine Muse

The data I have chosen is the Alcohol usage in adults found on Stat Crunch website.  Data gathered asked individuals questions regarding age, sex, the individual cause to drink, the number of drinks the person drink a week and education.  The two quantitative variables I focused on is, the dependent (y) number of drinks consumed and independent (x) the age in a week using a sample size of 100 people.

The formula equation for linear regression is y = b+ b1x.  In the equation y is the dependent  or response is the number of drinks consumed and x is the independent or predictor is the age of individuals.  My linear regression equation is 11.4457 - 0.0987 Age=Average.  Since the b1 is negative the regression line will go in a downward angle.  The linear correlation coefficient, r, is a descriptive measure of the strength and also tell the direction of the linear relationship between two variables.  My correlation coefficient is -0.1409, which means my regression line is a weak and negative.  The coefficient of determination or R-squared is 0.0199 and 1.99% the change in y is explained by the regression equation and changes in the independent variable.  Although a regression can be done, I can conclude that the regression line should not be done; it is not a good predictor because r-squared is close to 0.

In conclusion, my data focus on the quantitative data with number of drinks consumed by an individual per week (Dependent variable) and the age of said person (independent variable). The linear regression equation is 11.4457 - 0.0987 Age.  The scatterplot shows the data is clustered around the regression line with a few outliers on the chart.  My correlation coefficient is -0.1409, which means my regression line is a weak negative. The coefficient of determination or R-squared is 0.0199 and 1.99% the change in y is explained by the regression equation and changes in the independent variable.  Although, the linear regression data can be use it shouldn’t be because the coefficient of determination is close to 0.   

Result 1: Simple Linear Regression   [Info]
Right click to copy



Result 2: Consumption of drinks in Adults by Age   [Info]
Simple linear regression results:
Dependent Variable: No of drinks
Independent Variable: Age
No of drinks = 11.445749 - 0.098717052 Age
Sample size: 100
R (correlation coefficient) = -0.14091474
R-sq = 0.019856963
Estimate of error standard deviation: 9.5784805

Parameter estimates:
ParameterEstimateStd. Err.AlternativeDFT-StatP-value
Intercept11.4457492.7523425 ≠ 0984.1585482<0.0001
Slope-0.0987170520.070059568 ≠ 098-1.40904450.162

Analysis of variance table for regression model:
SourceDFSSMSF-statP-value
Model1182.15567182.155671.98540650.162
Error988991.234391.747289
Total999173.39