Scribd
Upload
173 views4 pages

Regression Analysis in Minitab

The document discusses using regression analysis in Minitab to analyze the relationship between height and weight using data from 22 individuals. A scatterplot reveals a positive linear relationship between height and weight. Running a linear regression analysis produces the equation: weight = -315.3 + 6.902 height, indicating that for each 1 inch increase in height, weight increases by approximately 6.9 pounds on average. The regression has an R-squared value of 89.6%, meaning height well explains the variation in weight.

Uploaded by

Radu Vasile
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
173 views4 pages

Regression Analysis in Minitab

The document discusses using regression analysis in Minitab to analyze the relationship between height and weight using data from 22 individuals. A scatterplot reveals a positive linear relationship between height and weight. Running a linear regression analysis produces the equation: weight = -315.3 + 6.902 height, indicating that for each 1 inch increase in height, weight increases by approximately 6.9 pounds on average. The regression has an R-squared value of 89.6%, meaning height well explains the variation in weight.

Uploaded by

Radu Vasile
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Regression analysis in Minitab

The Minitab project file Heightweight.MPJ contains data on heights of 22 individuals (both men and women) and their desired weights. In these notes you will be shown step by step how to obtain the equation of the regression line that best fits the data. We use height as explanatory variable and weight the response. We first construct a scatterplot of the data.
Scatterplot of weight vs height
240 220 200 weight 180 160 140 120 100 60 62 64 66 68 70 height 72 74 76 78

The scatterplot reveals a fairly strong, positive linear association. Therefore, we can proceed to find the equation of the regression line. In order to do that we select: STAT->Regression->Fitted line plot (see below). In the window that opens select the Predictor (or explanatory variable) height, and the response weight and type of regression: linear, and press ok.

The window that opens will show the scatterplot together with the regression line and the equation of the regression line.

Fitted Line Plot


weight = - 315.3 + 6.902 height 240 220 200 weight 180 160 140 120 100 60 62 64 66 68 70 height 72 74 76 78
S R-Sq R-Sq(adj) 12.9925 89.6% 89.1%

The slope of the equation is positive (as we correctly interpreted the scatterplot) and indicates that for each increase in height of 1in, the average weight increases with approximately 6.9 pounds. Notice that the negative y-intercept means that for a person of zero inches in height (nonsense!) the average weight is -315.3 pounds! Of course this does not make sense, therefore we cannot apply this model far outside the range of our data. For example it would not make sense to make predictions for weights of toddlers based on his model! Take a toddler measuring say 36 inches, the model would predict a weight of -66.8 pounds! Notice also that we have a big R-square of 89.6%, meaning that 89.6% of the variation in weight is explained by the linear dependence on height. A plot of the residuals can b obtained if in the Fitted line plot window above you select Graphs -> Residuals versus variable and select weight in the variable box. Click then ok and again ok. You will get the following plot besides the one shown above:

Residuals Versus height


(response is weight) 30 20 10 Residual 0 -10 -20 -30 60 62 64 66 68 70 height 72 74 76 78

Since the points are scattered on both sides of the horizontal line in a chaotic way, we can say that the linear model is a good fit for the given data.

You might also like