170 likes | 294 Views
Predicting Heights. Sarah Beeson, Mike Levin, and Andy. Introduction. What we measured: Length of the hand Length of nose to the belly button Circumference of the ankle Measured the hand from the bottom of the hand to the top of the middle finger
E N D
Predicting Heights Sarah Beeson, Mike Levin, and Andy
Introduction • What we measured: • Length of the hand • Length of nose to the belly button • Circumference of the ankle • Measured the hand from the bottom of the hand to the top of the middle finger • Measured from the tip of the nose to the belly button • Measured the circumference of the ankle
Height vs. Circumference of Ankles • Linear • Positive slope • Moderate r= .52 r2= .27 27% of the change in the height is due to the change in the circumference of the person’s ankle.
Height vs. Circumference of Ankle Residual Plot • Scattered • Good linear model
Height vs. Circumference of Ankleby Gender • Female LSR Line is on the bottom; male on the top • Males were likely to have bigger ankles than females. • Females: • Scattered • Positive slope • Strong • Males • Scattered • Positive slope • Moderately weak
Height vs. Length of the Nose to Belly Button • Linear • Positive slope • Moderate r= .47 r2= .22 22% of the change in the height is due to the change in the length from the person’s belly button to their nose.
Height vs. Length of Nose to Belly Button Residual Plot • Scattered • Good linear model
Height vs. Length of the Nose to Belly Button by Gender • Female is on the bottom line; male on the top line • Males were likely to have a longer distance between their nose and belly button • The slopes for each are almost parallel • Female: • Linear • Positive slope • Weak • Male: • Linear • Positive slope • Moderate
Height vs. Length of the Hand • Linear • Positive slope • Moderately strong r= .74 r2= .55 55% of the change in the height is due to the change in the length of a person’s hand.
Height vs. Length of the HandResidual Plot • Scattered • Good linear model
Height vs. Length of the Handby Gender • Females LSR line on the bottom; males on top • Males tend to have longer hands than females • Again, the LSR lines appear parallel • Female: • Linear • Positive slope • Moderate • Male: • Linear • Positive slope • Moderately strong
Best Model • Our best model (LSLR) was the Height vs. Hand Length. This is the best because it has the highest correlation (.74) of our three graphs. It is also linear and has a positive slope.
Group Member’s HeightsHeight = 6.14Hand + 24.2 • Mike= 68.715 inches (Actual height= 71 inches) • Residual= 2.285 inches • Sarah= 67.18 inches (Actual height= 67) • Residual= -0.18 inches • Andy= 65.645 inches (Actual height= 71.5) • Residual= 5.855 inches
Teacher’s Predicted HeightsHeight = 6.14Hand + 24.2 • Mrs. McNelis= 70.25 inches • Mrs. G= 65.645 inches • Mrs. V= 67.18 inches • Mrs. Arden= 64.11 inches
Confidence in Predictions We are not too confident in our prediction of the teachers heights because the residuals for our own heights were pretty large. Even though our correlation for the height vs. hand length plot (.74) was our strongest correlating plot, our predictions were pretty off besides Sarah’s prediction that only had a residual of -.18 inches.
Bias and Error • Starting at the same spot on each hand • Whether the person is wearing socks or whether we measured around the sock • Where we started measuring on the nose
Conclusion We found that the length of your hand, the length of your belly button to your nose and the circumference of your ankle have no association with your height. Each variable did not have a strong correlation, therefore it was difficult to have accurate predictions.