Pharmacy exam 2018-01-19
Degrees: Pharmacy, Biotechnology
Date: January 19, 2018
Question 1
A study done on a group of senior people to determine the relation between age
Do the following:
- Estimate the number of times a 70-year-old patient will go to the doctor, according to a linear regression model.
- What will be the estimate equal to if you consider an exponential model instead of the linear one?
- Which of the two estimates is more reliable?
- A potential model has equation of the type
, where and are constants to be determined; what transformation should you apply to the variables and to change a potential model into a linear one?
Use the following sums for the computations:
-
Linear model of Visits on Age:
years, years² . visits, visits². years⋅visits. Regression line of Visits on Age: . visits. -
log(visits), log(visits)². years⋅log(visits). Exponential model of Visits on Age: . visits. -
Linear coefficient of determination of Visits on Age
. Exponential coefficient of determination of Visits on Age . Thus, the exponential model explains a little bit better the number of visits to the doctor with respect to the age. -
We must apply the logarithm to both Visits and Age:
.
Question 2
The grass pollen concentration in the center of a city in grains/m
- Health authorities have determined that the level of pollen did not pose a risk for 75% of the days in the year; what is the minimum level of pollen that is consider a health hazard?
- On days with pollen level between 575 and 860 health authorities issue a warning to citizens; on how many days of the last year there were warnings issued?
- Are there outliers in the above sample?
- Platanaceae has a pollen cycle similar to grass: if
are the pollen levels of grass, and are the levels of the platanaceae, it is known that . What will be the average pollen level for platanaceae? Which of the two averages is more representative? - Can one say that the level of grass pollen comes from a population that is normally distributed?
Use the following sums for the computations:
grains/m³. and , so the frequency of days with a warning is that correspond to days. grains/m³, grains/m³ and grains/m³. Fences: grains/m³ and grains/m³. Since all the values fall into the fences there are no outliers. grains/m³, (grains/m³)², grains/m³ and grains/m³, grains/m³ and . The mean of is more representative than the mean of as . and . As both of them are between -2 and 2, we can assume that the pollen concentrations are normally distributed.
Question 3
Polen level in Madrid in the year 2017 is normally distributed with mean equal to 90 particles per cubic meter. In 42 days of 2017, the level was above 120 particles per cubic meter. Do the following:
- Compute the standard deviation of the polen level in the year 2017.
- On how many days the polen level did not go over 50 particles per cubic meter of air?
- On 20% of the days the level of polen was high enough to pose a health risk for allergic people; what is the level of polen that triggers this high risk situation?
Let
grains/m³. that correspond to days. grains/m³.
Question 4
A study on two drugs to reduces the cholesterol levels in blood shows that drug
- Compute the percentage of the population for which only drug
works. - Assume that drug
works on a person; what is the probability hat drug will also work in that person? - On the other hand, if drug
has not worked for a person, what is the probability that drug will actually work? - Are the effects of the two drugs independent events?
, that is, a . . . , thus the events are dependent.
Question 5
The weekly average births on a hospital is equal to 14.
- Compute the probability that on a given day more than 2 births take place.
- Compute the probability that during a week there are more than one day without births taken place.
- Let
be the number of births in a day. . - Let
be the number of days without births in a week. . .
Question 6
A trial to develop a diagnosis test for a desease is tested on 250 people, of which 50 suffer the desease and 200 are healthy.
The medical team in charge of the trial wants for the test to have a positive predictive value of
- In order to get the values given above, how many of the healthy people should get a positive outcome in the test?
- And how many of the sick people should get a negative outcome in the test?
- What is the probability that a person with two positive outcomes in the test has the desase?
Let
persons. persons. .