Answer:
We accept H₀ we don´t have enough evidence to support that more than 3 % of the packages are undeweight
Step-by-step explanation:
Sample size n = 1000
x ( defective packages found) = 40
p = 40 /1000 p = 0,04 p = 4 % then q = 0,96
np = 0,04 * 1000 and nq / 0.96 * 1000 both greater than 5
Then we can use the approximation of binomial distribution to normal distribution.
α is significance level α = 0,05
From z- table we find z(c) for α = 0,05 z(c) = 1,64
Test Hypothesis
Null Hypothesis H₀ p = 0,03
Alternative Hypothesis Hₐ p > 0,03
We can see from alternative Hypothesis that the test is a one tail test to the right of the bell shape curve of the normal distribution
To calculate z(c) = ( 0,04 - 0,03 ) / √ ( p*q)/n
z(c) = 0,01 * 31,62 / √ 0,04*0,96
z(c) = 0,3162 / 0,1959
z(c) = 1,6140 z(c) ≈ 1,61
Comparing z(s) and z(c)
z(c) > z(s) 1.64 > 1.61
Then z(s) is in the acceptance region for H₀ we accept H₀