Answer:
Step-by-step explanation:
First method information:
Sample size n₁ = 549
x₁ ( people ignoring on-line harassment ) = 412
Proportion of people ignoring on-line harassment p₁
p₁ = 412/549 p₁ = 0,75 q₁ = 1 - p₁ q₁ = 0,25
Second method ( people chosen to respond)
Sample size n₂ = 368
number of people responding x₂ = 305
Proportion of people responding
p₂ = 305/ 368 p₂ = 0,83 then q₂ = 0,17
a) Looking for differences in the approachs
Hypothesis test:
Null Hypothesis H₀ p₁ = p₂
Alternative Hypothesis Hₐ p₁ ≠ p₂
Alternative Hypothesis indicates that the test is a two-tail test
Choosing a confidence interval CI = 95 % then a significance level is α = 5% α = 0,05 and from z-table we get z(c) = 1.64
To calculate z(s)
z(s) = ( p₁ - p₂ ) / EED
EED = √ p₁*q₁)/n₁ + (p₂*q₂)n₂
EED = √ (0,75*0,25)/549 + 0,83*0,17)/368
EED = √ 3,41 *10⁻⁴ + 3,83*10⁻⁴
EED = 10⁻² *√ 3,41 + 3,83
EED = 10⁻² * 2,69
EED = 0,0269
z(s) = ( p₁ - p₂ )/0,0269
z(s) = - 0,08 / 0,0269
z(s) = - 2,97
p-value for z(s) is pvalue = 0,0015
Comparing z(s) and z(c) modules
z(s) > z(c) therefor z(s) is in the rejection region we have to reject H₀ and claim that there is a difference between the proportions
b) How large is the difference
