Instructor Key

1st Hourly

Math 1107

Summer Term 2002

 

Case One

Random Variables

 

We have a pair of dice – a fair d3 {faces 0,2,5} and a d4 {faces 1,2,3,4} – note the probability models for the dice below. 

 

d4

 

 

d3

 

Face

Probability

 

Face

Probability

1

0.10

 

0

1/3

2

0.40

 

2

1/3

3

0.40

 

5

1/3

4

0.10

 

 

 

 

We assume that the dice operate separately and independently of each other. Suppose that our experiment consists of tossing the dice, and noting the resulting pair of faces.

 

1.a)       List the possible pairs, and compute a probability for each.

 

Denote pair as (d4 face,d3 face).

 

(1,0)     Pr{(1,0) shows} = Pr{1 from d4}*Pr{0 from d3} = (1/10)*(1/3) = 1/30 @ .03333

(2,0)     Pr{(2,0) shows} = Pr{2 from d4}*Pr{0 from d3} = (4/10)*(1/3) = 4/30 @ .13333

(3,0)     Pr{(3,0) shows} = Pr{3 from d4}*Pr{0 from d3} = (4/10)*(1/3) = 1/30 @ .13333

(4,0)     Pr{(4,0) shows} = Pr{4 from d4}*Pr{0 from d3} = (1/10)*(1/3) = 1/30 @ .03333

 

(1,2)     Pr{(1,2) shows} = Pr{1 from d4}*Pr{2 from d3} = (1/10)*(1/3) = 1/30 @ .03333

(2,2)     Pr{(2,2) shows} = Pr{2 from d4}*Pr{2 from d3} = (4/10)*(1/3) = 4/30 @ .13333

(3,2)     Pr{(3,2) shows} = Pr{3 from d4}*Pr{2 from d3} = (4/10)*(1/3) = 1/30 @ .13333

(4,2)     Pr{(4,2) shows} = Pr{4 from d4}*Pr{2 from d3} = (1/10)*(1/3) = 1/30 @ .03333

 

(1,5)     Pr{(1,5) shows} = Pr{1 from d4}*Pr{5 from d3} = (1/10)*(1/3) = 1/30 @ .03333

(2,5)     Pr{(2,5) shows} = Pr{2 from d4}*Pr{5 from d3} = (4/10)*(1/3) = 4/30 @ .13333

(3,5)     Pr{(3,5) shows} = Pr{3 from d4}*Pr{5 from d3} = (4/10)*(1/3) = 1/30 @ .13333

(4,5)     Pr{(4,5) shows} = Pr{4 from d4}*Pr{5 from d3} = (1/10)*(1/3) = 1/30 @ .03333

 

 

1.b)      Define LOWTIE as either the lesser of the two faces when unequal, or the common face value when equal. List the possible values for LOWTIE, and compute a probability for each.

 

(d4 face,d3 face)ÞLowTie.

 

(1,0) Þ 0 @ .03333

(2,0) Þ 0 @ .13333

(3,0) Þ 0 @ .13333

(4,0) Þ 0 @ .03333

 

Pr{ LT=0 } = Pr{ one of (1,0),  (2,0),  (3,0),  (4,0)  shows} = 10/30 @ .3333

 

(1,2) Þ 1 @ .03333

(1,5) Þ 1 @ .03333

 

Pr{ LT=1 } = Pr{ one of  (1,2) , (1,5)  shows} = 2/30 @ .06666

 

(2,2) Þ 2 @ .13333

(3,2) Þ 2 @ .13333

(4,2) Þ 2 @ .03333

(2,5) Þ 2 @ .13333

 

Pr{ LT=2 } = Pr{ one of  (2,2), (3,2), (4,2), (2,5)  shows} = 13/30 @ .4333

 

(3,5) Þ 3 @ .13333

 

Pr{ LT=3} = Pr{ (3,5)  shows} = 4/30 @ .13333

 

(4,5) Þ 4 @ .03333

 

Pr{ LT=4} = Pr{ (4,5)  shows} = 1/30 @ .03333

 

Case Two

Probability Computational Rules

 

In this experiment we have a weird pair of dice – they are telepathically linked so that they do not operate independently. In fact, the dice produce the following face-pairs with the following probabilities:

 

(d2face,d4face)

@

Pr{(d2face,d4face)}

1

2

1

(1,1) @ .10

(2,1) @ .10

2

(1,2) @ 0

(2,2) @ .15

3

(1,3) @ .10

(2,3) @ .10

4

(1,4) @ .20

(2,4) @ 0

5

(1,5) @ .10

(2,5) @ .15

 

In this experiment we toss this weird pair of dice and note the resulting pair of faces. In each of the following, show your intermediate steps and work. If a rule is specified, you must use that rule for your computation.

 

2.a)       Compute Pr{ exactly one face shows even } using the Additive Rule.

 

(d2face,d4face)

@

Pr{(d2face,d4face)}

1

2

1

(1,1) @ .10

(2,1) @ .10

2

(1,2) @ 0

(2,2) @ .15

3

(1,3) @ .10

(2,3) @ .10

4

(1,4) @ .20

(2,4) @ 0

5

(1,5) @ .10

(2,5) @ .15

 

 

Pr{ exactly one face shows even } = Pr{ one of  (1,2), (1,4), (2,1), (2,3), (2,5) shows } = 0+.2+.1+.1+.15 = .55

 

2.b)      Compute Pr{ product of the two faces > 2 } using the Complementary Rule.

 

(d2face,d4face)

@

Pr{(d2face,d4face)} \

Product of Faces

1

2

1

(1,1) @ .10 \

1

(2,1) @ .10 \

2

2

(1,2) @ 0 \

2

(2,2) @ .15 \

4

3

(1,3) @ .10 \

3

(2,3) @ .10 \

6

4

(1,4) @ .20 \

4

(2,4) @ 0 \

8

5

(1,5) @ .10 \

5

(2,5) @ .15 \

10

 

 

Complementary Event

 

Pr{ product of the two faces £ 2 } = Pr{one of (1,1), (1,2), (2,1) shows} = .1+0+.1 = .2

 

Pr{ product of the two faces > 2 } = 1 - Pr{ product of the two faces £ 2 } =1 - .2 = .8

 


Case Three

Perfect Samples and the Long Run Interpretation

Xyrkztin’s Syndrome

 

Xyrkztin's Syndrome (XS) is a rare, fictitious disease exhibiting the following symptoms:

 

Itchy Skin

Progressive Failure of Immune System Function

Progressive Failure of Skeletal System

Perceives Invisible, Talking Evil Frog

Involuntary Funny Gait/Walk

Death

 

Upon diagnosis of XS, treatment is initiated, and the patient is tracked over time. Suppose further that the patients'

disease status is classified as: Fatal, Severe, Moderate, Mild, Cure/Remission, and that the table below gives the

probabilities for the population of XS patients at 5 years past diagnosis:

 

Status 5 Years after Diagnosis

Probability

Fatal

.35

Severe

.10

Moderate

.15

Mild

.25

Remission/Cure

.15

Total

1.00

 

In our experiment, we draw individual patients (with replacement) from the XS patient population, noting the severity of the case.

3.a)       Interpret the probabilities in terms of repeated trials of draws with replacement from the XS patient (at 5 years past diagnosis) population.

           Repeated draws with replacement from the population of  XS patients, five years past diagnosis, will yield approximately 35% of sampled cases as fatal, 10% of sampled cases as severe, 15% of sampled cases as moderate, 25% of sampled cases as mild and 15% of sampled cases in remission.

3.b)      Describe the perfect sample for 75 draws with replacement from the XS patient (at 5 years past diagnosis) population. Briefly describe the relationship between this perfect sample and actual samples of 75 draws with replacement from the XS patient (at 5 years past diagnosis) population.

Status 5 Years after Diagnosis

Probability

Perfect Sample Count (n-=75)

Fatal

.35

.35*75 = 26.25

Severe

.10

.10*75 = 7.5

Moderate

.15

.15*75 = 11.25

Mild

.25

.25*75 = 18.75

Remission/Cure

.15

.15*75 = 11.25

Total

1.00

1.00*75 = 75

           Random samples of n=75 XS patients (five years past diagnosis), drawn with replacement, consist of approximately 26.25 deaths, 7.5 severe cases, 11.25 moderate cases, 18.75 mild cases and 11.25 cured cases/cases in remission. 


Case Four

Conditional Probability

Draws without Replacement: Color Bowl

Case Description: Compute conditional probabilities for pairs of draws (without replacement).

Here is our bowl, in tabular form:

Color

# in Bowl

Proportion of Bowl

Blue

7

7/15

Green

5

5/15

Red

3

3/15

Total

15

15/15

Suppose that on each trial of this experiment that we make three (3) draws without replacement from the bowl.

Compute Pr{ green shows 3rd | red shows 1st and green shows 2nd };

Color

# in Bowl

Proportion of Bowl

Blue

7

7/13

Green

4

4/13

Red

2

2/13

Total

13

13/13

Pr{ green shows 3rd | red shows 1st and green shows 2nd } = 4/13 @ .3077

Compute Pr{ red shows 3rd  | red shows 1st and red shows 2nd };

Color

# in Bowl

Proportion of Bowl

Blue

7

7/13

Green

5

5/13

Red

1

3/13

Total

13

15/15

Pr{ red shows 3rd  | red shows 1st and red shows 2nd } = 1/13  @ .0769

Compute Pr{ blue shows 2nd | blue shows 1st }.

Color

# in Bowl

Proportion of Bowl

Blue

6

6/14

Green

5

5/14

Red

3

3/14

Total

14

14/14

Pr{ blue shows 2nd | blue shows 1st } = 6/14 @ .4286

Hint: Work these out directly - do not use the usual formulas. Use the same approach as in Case Study 1.13.

Case Five

Clinical Trial

Donepezil for Dementia in Parkinson's Disease

 

Parkinson's disease (PD) is the second most common neurodegenerative disorder after Alzheimer's disease (AD).

 

Dementia is a common problem late in the course of the disease, and there is no effective therapy. Dementia severely

impairs patients' functional status and limits the treatment of the motor manifestations of PD. No effective therapy for

dementia in PD is available.

 

Dementia is a clinical state characterized by loss of function in multiple cognitive domains.

Diagnostic features of dementia include: memory impairment and at least one of the following:

 

Aphasia (Diminished ability to correctly use and comprehend language. Aphasia is a language disorder caused by damage to the temporal lobe or higher up in the frontal lobe. It causes problems with receptive and expressive functions. Aphasia is an impairment in understanding and/or formulating complex, meaningful elements of language. It causes problems with words and word order making difficulties in reading and writing.)

 

Apraxia (A motor disorder in which voluntary movement is impaired without muscle weakness. The ability to select and sequence movements is impaired. Oral apraxia affects one ability to move the muscles of the mouth for non-speech purposes. Someone with oral apraxia would have trouble coughing, swallowing, wiggling their tongue or blowing a kiss when asked to do so. Verbal apraxia, or apraxia of speech is an impairment in the sequencing of speech sounds.),

 

Agnosia (An inability to recognize and identify objects or persons despite having knowledge of the characteristics of the objects or persons. People with agnosia may have difficulty recognizing the geometric features of an object or face or may be able to perceive the geometric features but not know what the object is used for or whether a face is familiar or not. Agnosia can be limited to one sensory modality such as vision or hearing. For example, a person may have difficulty in recognizing an object as a cup or identifying a sound as a cough..),

 

In addition, the cognitive impairments must be severe enough to cause impairment in social and occupational functioning.

 

The primary outcome measure in this trial is the change in patients' dementia status. The secondary outcome measures will include other scales of cognitive function, activities of daily living, mood, and quality of life.

 

Sketch a basic clinical trial for Donepezil in the treatment of Dementia in patients with Parkinson's Disease.

 

Disease of Interest: Dementia in Parkinson’s disease (PD)

 

Subjects of Interest: Patients with PD also presenting dementia

 

Treatments of Interest: Donepezil, an anti-dementia drug, and Placebo (matched to Donepezil)

 

Subject candidates diagnosed with PD who have a diagnosis of dementia, who are briefed as to the risks and benefits

Of study participation, who are eligible for study participation and who give informed consent (or whose proxies give informed consent)

are randomly assigned to either Donepezil or to Placebo.

 

Neither study participants nor study workers are aware of actual subject treatment assignments (double blinding).

 

Study subjects are tracked for modification of symptoms of dementia: memory impairment, aphasia, agnosia, apraxia, social function, occupational function.

 

Study subjects are tracked for cognitive function, quality of life, mood.

 

Study subjects are tracked for drug tolerance/side effects,  as well as for toxic reactions to drug.