Você está na página 1de 6

Biology 300, Biometrics Name: KEY

Key to Exam #1a, Winter Quarter 2010

PART I
A. Multiple Choice (30 points). Circle the best answer. Only one choice is “best”.

1. In general, which of the following statements is true?


(a) The arithmetic mean is greater than or equal to the geometric mean.
(b) The harmonic mean is less than or equal to the geometric mean.
(c) The median is a measure of central tendency.
(d) The median is the same as the 50’th percentile.
E (e) All of the above are true.

2. A standard normal probability distribution


A (a) has a standard deviation equal to one.
(b) has a coefficient of variation equal to one.
(c) is a uniform probability distribution.
(d) is symmetric around one.
(e) All of the above are true.

3. Joan has a high-density lipoprotein (HDL) level that is at the 45th percentile and Isabel’s HDL
level is at the 25th percentile for women of their age. This means that
(a) Joan has a lower HDL level than Isabel.
(b) There is a 0.75 probability of a woman their age having a lower HDL level than Isabel.
C (c) 20% of women in their age group have a HDL level between Isabel and Joan.
(d) All of the above are true.
(e) None of the above are true.

4. Which of the following is an example of ordinal data?


(a) birth weight of an infant
(b) number of cousins for an infant
(c) sex of an infant
(d) color of an infant’s hair
E (e) none of the above

5. A bird lays a clutch of seven eggs. If each egg has a 80% chance of hatching, then the number
of chicks produced will follow a
(a) Z-distribution.
B (b) binomial distribution.
(c) normal distribution.
(d) Poisson distribution.
(e) uniform distribution.
Biometrics — Winter 2010 Key to Exam #1a Page 2

6. Which of the following statistics is not a measure of dispersion?


(a) the range
(b) the mean deviation
(c) the standard deviation
(d) the coefficient of variation
E (e) All of the above are measures of dispersion.

7. A couple are both carriers for the recessive allele for phenylketonuria (PKU). Each of their
children will have 1/4 probability of inheriting PKU. If the couple plans on having two
children, the which of the following statements is true?
A (a) From the multiplication rule there is a (3/4)(3/4) = 9/16 both of their children will be
normal.
(b) From the addition rule there is a (1/4)+(1/4) = 1/2 probability that both of their children
will inherit PKU.
(c) From the multiplication rule there is a (1/4)(3/4) = 3/16 probability that one of their two
children will inherit PKU and the other child will be normal.
(d) All of the above are true.
(e) None of the above are true.

8. The standard error of the mean


(a) decreases as the sample size increases.
(b) is a measure of the error associated with a sample estimate of the population mean.
(c) is smaller than the standard deviation.
D (d) all of the above
(e) none of the above

9. The central limit theorem states that as the sample size increases
(a) the standard deviation decreases.
(b) the sample median tends towards the sample mean.
(c) an estimated probability tends towards a binomial distribution.
D (d) the distribution of the sample mean tends towards a normal distribution.
(e) all of the above

10. The State of Florida has records on the number of shark attacks per year from 1948-2009. If we
assume shark attacks are random events, the number of attacks per year should follow a
(a) binomial distribution.
(b) uniform distribution.
C (c) Poisson distribution.
(d) normal distribution.
(e) Z-distribution.
Biology 300, Biometrics Name: KEY
Key to Exam #1a, Winter Quarter 2010

PART II
B. Answer each question and show all intermediate calculations. Be Neat! You may use one sheet
of paper containing statistical formulas and statistical table A. Use the back sides of these sheets
if you need more space.

1. The following data represent the dry weight (in g) of leaves:

0.2, 0.6, 1.1, 0.5, 0.7

(a) Compute the arithmetic mean, the geometric mean, the harmonic mean, and the median to
three decimal places. (8 points)

Rank X LN(X) 1/X Sample size: n = 5


1 0.2 -1.6094 5.000
3 0.6 -0.3567 1.667
Arithmetic mean: x =  X)/n = 3.1/5 = 0.620 g
5 1.1 0.0953 0.909
2 0.5 -0.6931 2.000
4 0.7 -0.5108 1.429 Geometric mean: G = exp[( Ln X)/n] = exp[–3.0748/5] = exp[–0.61496]
SUM: 3.1 -3.0748 11.004
= 0.541 g

Harmonic mean: H = n/ (1/X)] = 5/[11.004] = 0.454 g

Median: n=5 is odd, so median is the X-value with rank


(n+1)/2 = (5+1)/2 = 6/2 = 3.
median = X3rd = 0.600 g

(b) Compute the standard deviation, the mean deviation, the range, and the 20th percentile to
three decimal places. (8 points)

Rank X X*X |X-x| Standard deviation: s = Sqrt([( X2) – ( X)2/n] / [n-1])


1 0.2 0.04 0.42 = Sqrt([(2.35) – (3.1)2/5] / [5-1])
3 0.6 0.36 0.02 = Sqrt([2.35 – 9.61/5] / 4)
5 1.1 1.21 0.48 = Sqrt([2.35 – 1.922] / 4)
2 0.5 0.25 0.12 = Sqrt(0.428 / 4) = Sqrt(0.107) = 0.327 g
4 0.7 0.49 0.08
SUM: 3.1 2.35 1.12
Mean deviation =  |X – x|)/n = 1.12 / 5 = 0.224 g

Range = Max(X) – Min(X) = X5th – X1st = 1.1 – 0.2 = 0.900 g

20th percentile: P=20, n=5, (nP)/100 = (5*20/100) = 100/100 = 1 which


is an integer, so the 20th percentile is the average of the X-values with
ranks 1 and 2.
20th percentile = (X1st + X2nd)/2 = (0.2 + 0.5)/2 = 0.7/2 = 0.350 g
Biometrics — Winter 2010 Key to Exam #1a Page 4

2. About 20% of the world’s population has blood type B+. If three people are randomly selected,
what is the probability distribution for the number of people with blood type B+. Show your
work and then write your answers in the table below to three decimal places. (12 points)

Number of blood type B+ Probability

0 0.512

1 0.384

2 0.096

3 0.008

Each person represents an independent trial and each trial has two possible outcomes: Blood type B+ or not blood type B+.
Therefore, the probabilities can be determined using the binomial distribution:

N! k N–k
Prob  x=k  = ------------------------- p q
k!  N – k !

where A = blood type B+,


B = not Blood type B+,
p = Prob(A) = probability that a person is blood type B+,
q = Prob(B) = 1–p,
x = the number of people selected with blood type B+,
N = total number of people selected.
For this problem, p = 0.2, q = 0.8, and N = 3. The probabilities are

3! 0 3–0 6
Prob  x=0  = ------------------------  0.2   0.8  = -----------  1   0.512  = 0.512
0!  3 – 0 ! 16

3! 1 3–1 6
Prob  x=1  = ------------------------  0.2   0.8  = -----------  0.2   0.64  = 0.384
1!  3 – 1 ! 12

3! 2 3–2 6
Prob  x=2  = ------------------------  0.2   0.8  = -----------  0.04   0.8  = 0.096
2!  3 – 2 ! 21

3! 3 3–3 6
Prob  x=3  = ------------------------  0.2   0.8  = -----------  0.008   1  = 0.008
3!  3 – 3 ! 61
Biometrics — Winter 2010 Key to Exam #1a Page 5

3. The following table represents the number of fruit flies per trap for sticky traps that were placed
in fruit trees:

Number of fruit
Number of traps Probability Expected values
flies per trap

0 18 0.2466 19.73

1 22 0.3452 27.62

2 30 0.2417 19.33

3 10 0.1665 13.32

(a) Compute the mean, standard deviation, and median for the number of fruit flies per trap to
two decimal places. (10 points)
X f f*X f*X^2 Ranks Sample size: n =  f = 80
0 18 0 0 1-18
1 22 22 22 19-40 Arithmetic mean: x = ( f*X)/n = 112/80 = 1.40 fruit flies/trap
2 30 60 120 41-70
3 10 30 90 71-80
SUM: 80 112 232 Std. Dev. = s = Sqrt([ ( f*X2) – ( f*X)2/n ]/[n–1])
= Sqrt([ (232) – (112)2/80 ]/[80–1])
= Sqrt([ (232) – (12544)/80 ]/[79])
= Sqrt([ 232 – 156.8 ]/[79])
= Sqrt(75.2/79) = Sqrt(0.951899) = 0.98 fruit flies/trap

Median: n=80 is even, so median is the average of the X-values with


ranks (n/2)=40 and 1+(n/2)=41.
Median = (1+2)/2 = 1.5 fruit flies/trap.

(b) Fit a Poisson distribution to these data using the sample mean computed in part (a). Enter
the probabilities (to four decimal places) and expected values (to two decimal places) for
the number of fruit flies per trap into the table above. (16 points)
The probabilities can be determined using the Poisson distribution: Prob(X=k) = (e–)k )/(k!), where the sample mean
x= 1.40 is used in place of .

Prob(X=0) = exp(–1.40) (1.40)0/(0!) = (0.246597) (1)/(1) = 0.2466 Expt = n Prob(X=0) = 80 (0.2466) = 19.73
1
Prob(X=1) = exp(–1.40) (1.40) /(1!) = (0.246597) (1.40)/(1) = 0.3452 Expt = n Prob(X=1) = 80 (0.3452) = 27.62
2
Prob(X=2) = exp(–1.40) (1.40) /(2!) = (0.246597) (1.96)/(2) = 0.2417 Expt = n Prob(X=2) = 80 (0.2417) = 19.33

Prob(X3) = 1 – Prob(X2) = 1 – [ Prob(X=0) + Prob(X=1) + Prob(X=2) ]


= 1 – [ 0.2466 + 0.3452 + 0.2417 ] = 1 – 0.8335 = 0.1665
Expt = n Prob(X3) = 80 (0.1665) = 13.32
Biometrics — Winter 2010 Key to Exam #1a Page 6

4. Soil extracts from a landfill had a mean dioxin concentration of  = 532 g/ml and a standard
deviation of  = 50 g/ml.
(a) What is the percentile (to two decimal places) for an dioxin concentration of 500 g/ml?
(5 points)
For this problem, = 532 g/ml,  = 50 g/ml, and X1 = 500 g/ml. The percentile is given by

percentile = Prob(X < X1) (100%)


= Prob(Z < Z1 = (X1 – )/) (100%)
= Prob(Z < (500 – 532)/50) (100%)
= Prob(Z < –32 / 50) (100%)
= Prob(Z < –0.64) (100%)
= [0.5 – Prob(0 < Z < 0.64)] (100%)
= [0.5 – A] (100%)
= [0.5 – 0.2389] (100%) = (0.2611) (100%) = 26.11% or 26th percentile.

(b) What dioxin concentration (to zero decimal places) corresponds to the 80th percentile? (5
points)
For this problem, = 532 g/ml,  = 50 g/ml, and the percentile = 80%. The corresponding value of X1 is the solution to
Prob(X < X1) = 0.80. From the normal probability table, (0.5+A) = 0.80, so A = 0.80–0.5 = 0.30. This gives a value of Z1 =
0.84 from Table A.

0.80 = Prob(Z < Z1 = (X1 – )/)


= Prob(X < X1 =  + Z1)
= Prob(X < X1 = 532 + (50) (0.84))
= Prob(X < X1 = 532 + 42) = Prob(X < X1 = 574)

Thus, X1 = 574 g/ml is the 80th percentile.

(c) In what proportion (to four decimal places) of soil samples can one expect dioxin
concentrations that are between 500 g/ml and 600 g/ml? (6 points)
For this problem, = 532 g/ml,  = 50 g/ml, X1 = 500 g/ml and X2 = 600 g/ml.

Prob(X1 < X < X2) = Prob( (X1 – )/ < Z < (X2 – )/)
= Prob( (500 – 532)/50 < Z < (600 – 532)/50) )
= Prob( –32/50 < Z < 68/50 )
= Prob( –0.64 < Z < 1.36 )
= Prob( –0.64 < Z < 0 ) + Prob( 0 < Z < 1.36 )
= A1 + A2 = 0.2389 + 0.4131 = 0.6520.

Thus, 65.20% of the soil samples are predicted to have dioxin concentrations between 500 g/ml and 600 g/ml.

Você também pode gostar