Michelle Ji, Sam Shober, April Zhang
1. Shoulder to Floor
5. Group Members
2. Head Circumference
6. Predictions
3. Right Foot Length
7. Confidence
4. Best Model
8. Bias and Error
9. Conclusion




Ticks pre-marked on wall
Participants take of both shoes and
stand with feet as close to wall as
possible
Observer approximates which tick the
participants’ shoulder reached
Measured in inches
Scatterplot/LSR Line
Residual Plot
Male/Female
Difference
SCATTERPLOT AND LSR LINE



Positive
Moderately strong
Linear
RESIDUAL PLOT

Scattered
 LSR Line a good fit



r = 0.943398
r2 = 0.89
89% of the variation in
height is explained by
the variation in
shoulder to floor length
MALE
FEMALE


Positive
Moderately
Strong



Linear
Smaller Slope


Smaller
correlation:
0.8888





0.674
Generally
smaller values
Positive
Strong
Linear
Larger Slope


Larger
correlation:
0.9644
0.948
Generally
larger values




Participants lifted hair about head (for
long hair)
Tape measurer placed as tightly as
possible around head above ears
Measurement read as point where tick
and metal tip met
Measured in Inches
Scatterplot/LSR Line
Residual Plot
Male/Female
Difference
SCATTERPLOT AND LSR LINE



Linear
Positive
Moderately weak
RESIDUAL PLOT

Slight Horn Shape
 LSR Line not best fit




Outlier near 26
r = 0.42426
r2 = 0.18
18% of the variation in
height is explained by
the variation in head
circumference
FEMALE


Positive
Weak
MALE


 Smaller
correlation
: 0.02


Linear
Smaller
Slope
 0.0615
Positive
Weak



Linear
Larger slope


Larger
correlation:
0.305
0.71
Outlier: near
26




Participants made to take off their
right shoe
They were to line the heel of their foot
to the end of the ruler
Observer approximated the tick on
the ruler that the participants foot
touched (looked at the longest toe)
Measured in inches
Scatterplot/LSR Line
Residual Plot
Male/Female
Difference
SCATTERPLOT/ LSR LINE



Linear
Positive
Moderate
RESIDUAL PLOT

Scattered
 LSR Line is a good fit

Two possible outliers
 Near 11.5 and 12



r = 0.76811
r2 = 0.59
59% of the variation in
height is explained by
the variation right foot
length
FEMALE


Positive
Weak
 Smaller
MALE


 Larger
correlation
: 0.2966


Linear
Smaller
slope
 1.15
Positive
Moderate
correlation:
0.6557


Linear
Larger slope
 1.9

Shoulder to Floor
Length
 Strongest
correlation: r =
0.9434
 Female: r = 0.8888
 Male: r = 0.9644
 r2 = 0.89
 Female: r2 = 0.79
 Male: r2 = 0.93
MICHELLE




Shoulder to Floor: 50
inches
Height=.674(50) +28.6
= 59.3 inches
Actual Height= 63
inches
Residual =63-59.3= 3.7
inches
SAM
Shoulder to Floor: 57
inches
 Height=.674(57)
+28.6 = 67.018
inches
 Actual Height= 67
inches
 Residual =67-67.018=
-.018 inches

APRIL




Shoulder to Floor: 53 inches
Height=.674(53) +28.6 = 64.322
inches
Actual Height= 64 inches
Residual =64-64.322= -.322
inches
MR. LAKE


Shoulder to Floor: 59
inches
Height= .948(59) + 15.4
= 71.332 inches
MS. GEMGNANI


Shoulder to Floor: 55
inches
Height=.674(55) +28.6 =
65.67 inches
MR. WALSH


Shoulder to Floor: 56
inches
Height= .948(56) + 15.4
= 68.488 inches
MISS. TANNOUS


Shoulder to Floor: 56.5
inches
Height=.674(56.5) +28.6
= 66.681 inches
MRS. ROBINSON


Shoulder to Floor: 58
inches
Height=.674(58) +28.6 =
67.692 inches
MS. ARDEN


Shoulder to Floor: 53.5
inches
Height= .674(53.5) +
28.6 = 64.659
We are confident in our predictions because
our data has a moderately strong linear
shape and our LSR line has a strong
correlation, especially for the males. By using
different models for females and males, we
eliminate a possible lurking variable, making
us even more confident in our predictions. In
addition, our model accurately predicted our
own heights. Sam and April’s residuals were
very small, but Michelle’s was a little larger,
but not large enough to make us less
confident in our models.

Measurements taken by
different observers

Sam on foot measurements
 Variation in tightness of tape
between April and Michelle

 Michelle more exact than



Tightness of tape when
measuring head
circumference
Amount of hair in tape
measurer when measuring
head circumference
Exact location of
measurement for head
circumference
 Tried to place it in the same
place, can’t be exact

Participants may have
placed foot more forward
or back than others on foot
length measurement
Potential slouching during
shoulder to floor
measurement
Human error during
measurements
 Hard to approximate

Shoulder to floor length was best predictor
 Greatest correlation, strongest, most linear, lowest residuals
out of all three
Females have lower correlation for all three types of
measurements
 Females had smaller measurements than males

 With the exception of head circumference


Head circumference had little correlation to height
Future:
 Measure adults
 Make sure all participants have good posture
 Use more advanced equipment
▪ Height and foot measurer
 Measure height to nearest mm
 Be more accurate on foot length