In: Statistics and Probability
In studies for a medication, 7 percent of patients gained weight as a side effect. Suppose 403 patients are randomly selected. Use the normal approximation to the binomial to approximate the probability that (a) exactly 29 patients will gain weight as a side effect. (b) no more than 29 patients will gain weight as a side effect. (c) at least 37 patients will gain weight as a side effect. What does this result suggest?
Let X be the random variable that number of patients will gain weight as a side effect.
X ~ Binomial (n = 403, p = 7% = 0.07)
Now here we can see that,
np = 403 * 0.07 = 28.21 > 5
n(1-p) = 403 * (1-0.07) = 374.79
Here n is too large and p is too small so we will use here normal approximation.
mu = np = 403 * 0.07 = 28.21
X ~ N(mu = 28.21, sd = 5.12)
Use the normal approximation to the binomial to approximate the probability that :
(a) exactly 29 patients will gain weight as a side effect.
To find P(X = 29)
Using continuity correction we have to find P(29-0.5 < X < 29+0.5).
i.e. we have to find P(28.5 < X < 29.5)
Convert Xbar = 28.5 and 29.5 into z-score.
z-score is defined as,
Now we have to find P(0.06 < Z < 0.25)
P(0.06 < Z < 0.25) = P(Z < 0.25) - P(Z < 0.06)
We can find this probability in excel.
syntax :
=NORMSDIST(z)
where z is z-score
P(0.06 < Z < 0.25) = 0.5994 - 0.5226 = 0.0768
(b) no more than 29 patients will gain weight as a side effect. (c) at least 37 patients will gain weight as a side effect. What does this result suggest?
To find P(X < 29)
By using continuity correction we have to find P(X < 29-0.5).
To find P(X < 28.5)
z-score for Xbar = 28.5 is 0.06.
Now we have to find P(Z < 0.06).
P(Z< 0.06) = 0.5226
(c) at least 37 patients will gain weight as a side effect. What does this result suggest?
To find P(X 37)
By using continuity correction to find P(X > 37-0.5)
To find P(X > 36.5)
z-score for x = 36.5 is,
z = (36.5 - 28.21) / 5.12 = 1.62
Now we have to find P(Z > 1.62)
P(Z > 1.62) = 1 - P(Z < 1.62) = 1 - 0.9472 = 0.0528