Date post: | 17-Dec-2015 |
Category: |
Documents |
Upload: | jordan-julian-porter |
View: | 217 times |
Download: | 4 times |
INTRODUCTION
We chose to measure: Hip-to-knee length Head-nose circumference Top of palm
We measured in quarter inches
MEASUREMENTS: HEIGHT
We based our measurement of height based on a scale marked on the white board
Each subject was then measured off the scale with a book as a guide
MEASUREMENTS: HIP TO KNEE
We decided that a femur was a good measurement of height
The closest measurement possible is from the top of the hipbone to the middle of the knee when bent
We asked subjects to identify top of hip and then bend knee
MEASUREMENTS: HEAD-NOSE CIRCUMFERENCE
For our circumference, we measured around the head at the tip of the nose
Subjects held the tape at the nose while we measures around
MEASUREMENTS: TOP OF PALM
In order to measure the length of the palm, subjects bent wrist and knuckles and measurement was taken from wrist bone to middle finger knuckle
Measurements were taken in quarter inches
SCATTER PLOT OF HIP TO KNEE
56
58
60
62
64
66
68
70
72
74
76
HiptoKnee
16 18 20 22 24 26 28 30
Height = 1.463HiptoKnee + 35.4; r2 = 0.59
Collection 1 Scatter Plot
Collection 1
Height
HiptoKnee 0.767435
S1 = correlation
•The form of the scatter plot is roughly linear with a positive association. The strength is moderately strong with r = 0.767435.•The r2 of 0.59 shows that 59% of the variability in height is attributed to the variability in hip to knee length.
SCATTER PLOT OF HIP TO KNEE BY GENDER
Genderf m
56
58
60
62
64
66
68
70
72
74
76
HiptoKnee
16 18 20 22 24 26 28 30
Height = 1.304HiptoKnee + 37.3; r2 = 0.63
Height = 0.636HiptoKnee + 55.8; r2 = 0.28
Collection 1 Scatter Plot
• The two LSR lines vary between the genders •For females (the lower line), the correlation is stronger than the males (the upper line), with r’s of 0.79 and 0.53, respectively. •Both have a moderately strong association with linear qualities •The males show a steeper line than the females, and both show a positive association
Female: r = 0.793725Male: r= 0.529150
RESIDUAL PLOT OF HIP TO KNEE
56
58
60
62
64
66
68
70
72
74
76
HiptoKnee
16 18 20 22 24 26 28 30
Height = 1.463HiptoKnee + 35.4; r2 = 0.59
-4
0
4
8
16 18 20 22 24 26 28 30
HiptoKnee
Collection 1 Scatter Plot
• The residual plot (lower plot) for the hip to knee data shows no obvious patterns, showing that our LSRL fit our data
r = 0.768115r2 = 0.59
SCATTER PLOT OF NOSE-HEAD CIRCUMFERENCE
56
58
60
62
64
66
68
70
72
74
76
NoseHead
19 20 21 22 23 24 25
Height = 0.102NoseHead + 64.7; r2 = 0.0013
Collection 1 Scatter Plot
Collection 1
HiptoKnee
NoseHead -0.15286
S1 = correlation
• The form of nose-head circumference and height shows hardly any form or direction, with a very slight negative association. The form is scattered. The strength is weak, show by the correlation of -0.15286.•Our r2 of 0.0013 shows that .13% of the variability in height is attributed to the variability in nose-head circumference.
SCATTER PLOT OF NOSE-HEAD CIRCUMFERENCE BY GENDER
Genderf m
5658606264666870727476
NoseHead
19 20 21 22 23 24 25
Height = -0.0102NoseHead + 64.2; r2 = 0.000036
Height = 0.0821NoseHead + 68.6; r2 = 0.0019
Collection 1 Scatter Plot
• The scatter plot broken down by gender shows a difference in head circumference between males and females; yet, the form of both LSR lines were similar in their positive direction, weak scatter form, and weak strength. The males had the higher line while the females the lower.
Females: r = 0.006Males: r = 0.04359
RESIDUAL PLOT OF NOSE-HEADCIRCUMFERENCE
56
58
60
62
64
66
68
70
72
74
76
NoseHead
19 20 21 22 23 24 25
Height = 0.102NoseHead + 64.7; r2 = 0.0013
-8
-4
0
4
8
19 20 21 22 23 24 25
NoseHead
Collection 1 Scatter Plot
• The residual plot shows granularity which means that the model of the LSRL does not fit our data.
r2 = 0.0013r = 0.03606
SCATTER PLOT OF PALM LENGTH
56
58
60
62
64
66
68
70
72
74
76
Palm
3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2
Height = 6.47Palm + 41.7; r2 = 0.36
Collection 1 Scatter Plot
Collection 1
Height
Palm 0.600379
S1 = correlation
•The form is clustered at each data point with a slight linear association. The direction is positive while the strength is moderate. Our correlation is 0.600379•Our r2 of 0.36 shows that 36% of the variability in height is attributed to the variability in palm length.
SCATTER PLOT OF PALM LENGTH BY GENDER
Genderf m
56
58
60
62
64
66
68
70
72
74
76
Palm
3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2
Height = 2.3Palm + 55.5; r2 = 0.071
Height = 4.07Palm + 53.7; r2 = 0.38
Collection 1 Scatter Plot
• Our scatter plot of palm length broken down by gender has two LSR lines, with the males having the higher line and the females the lower.•The male’s data was linear with a positive association. The data was moderately strong with a correlation of 0.616.•The female’s data was slightly linear with a positive association. The data was moderately weak with a correlation of 0.266.•The male’s data LSRL was more reliable than the female’s.
Males: r = 0.616441Females: r = 0.26646
RESIDUAL PLOT OF PALM LENGTH
56
58
60
62
64
66
68
70
72
74
76
Palm
3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2
Height = 6.47Palm + 41.7; r2 = 0.36
-12
-8
-4
0
4
3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2
Palm
Collection 1 Scatter Plot
• The residual plot for palm length shows obvious granularity in the stair step appearance. This shows that our LSR line does not fit the data.
r2 = 0.36r = 0.6
BEST MODEL
The LSRL chosen by our group was the hip-to-knee because it had the highest correlation for both genders
We will base the female teachers off of the LSRL created by only the female data but approximate the male data off of the LSRL created by both the male and female data for better accuracy
Female: height = 1.304(hip-to-knee) + 37.3 Male: height = 1.463(hip-to-knee) + 35.4
PREDICTIONS OF GROUP MEMBERS
Kathy: Female model used with hip-to-knee measurement of 18” gave height of 60.772” which was 2.228” away from her actual height of 63”
Sarah: Female model used with 21” hip-to-knee length gave height as 64.684” which was 1.316” away from the actual height of 66”
Tyler: Total model used with hip-to-knee measurement of 23” gave height of 69.049” which was 3.451” away from actual height of 72.5”
Our models were consistently short
PREDICTIONS OF TEACHERS
Ms. Tannous: 1.304(21) + 37.3 = 64.684”
Ms. Arden: 1.304(22) + 37.3 = 65.988” Mrs. Robinson: 1.304(21) + 37.3 =
64.684” Mr. Lake: 1.463(23) + 35.4 = 69.049” Mr. Walsh: 1.463(21) + 35.4 = 66.123”
CONFIDENCE OF PREDICTIONS
For ourselves, our models were overall short, and therefore we can presume that our estimations for the teachers may also fall short from their real heights.
BIAS AND ERROR
Some subjects felt awkward with measuring the top of the hipbone
Our other two measurements showed too uniform of distributions to be useful
We did not specify shoes or no shoes for everyone
Some subjects may not have correctly identified the top of their hipbone
Cargo pants were difficult to find the top of knee
CONCLUSION
The measurement of hip-to-knee was the most reliable compared with the measurement of the palm and of the head over the nose.
Our predictions seemed to fall slightly short, suggesting that the measuring the femur through the hipbone is not as reliable.
Our female data for the hip-to-knee measurement had the highest correlation, followed by the LSRL for both genders that we used for males.