Professional Documents
Culture Documents
Statistics Assignment 3
Statistics Assignment 3
A. Given n=30,
Question: Which are more important to you in your vote for Congress
this November: domestic issues such as the economy, health care, and
immigration; or international issues such as Afghanistan, Iran, and
terrorism?
A)
given, confidence interval = 95% and n=1000. Hence α = 0.05
Zα/2 = Z0.025 = 1.96 (using z distribution table)
Population proportion for 95% confidence interval is given by,
p̄ ± 1.96(√ (p̄ (1- p̄)/n)
Gulf coast oil spill handled by Obama in August 2010, approved by 50%
people,
p̄ = 0.50 , n=1000
= 0.50 ± 1.96(√(0.50(1-0.50)/1000)
=(0.4695, 0.5305)
Gulf coast oil spill handled by Bush in March 2006, approved by 36%
people,
p̄ = 0.36, n=1000
= 0.36 ± 1.96(√ (0.36(1-0.36)/1000)
= (0.3302, 0.3897)
Hence, we can be 95% confident that the approval rates for handling
Hurricane by Obama lies in the range (0.47,0.53) and by Bush lies in the
range (0.33,0.39). That is approval rates is more for Obama.
B)
Importance of domestic issues in August 2010, voted by 73% people,
p̄ = 0.73, n=1000
= 0.73 ± 1.96(√ (0.73(1-0.73)/1000)
= (0.7024, 0.7575)
Importance of domestic issues in September 2006, voted by 43%
people,
p̄ = 0.43, n=1000
= 0.43 ± 1.96(√ (0.43(1-0.43)/1000)
= (0.3993, 0.4606)
Hence, we can be 95% confident that the importance of domestic issues
in August 2010 lies in the range (0.7,0.75) and in September 2006 lies in
the range (0.4,0.46). That is the importance given increased from 2006
to 2010.
C)
Using Left-tail Hypotheses test for score 1 & 2,
Hₒ: pₒ≥0.54 vs HA: pₒ<0.54 and p̄ = 0.5
Z = (p̄-pₒ)/(√(pₒ(1-pₒ))/n) (using central limit theorem for sample
proportion)
= (0.5-0.54)/ (√ (0.54(1-0.54)/122))
= -0.886
p-value = P(Z<-0.886)
= 0.1867
Here, p-value > α. therefore, by p-value approach, we accept the null
hypotheses
i.e.; cannot reject Hₒ.
Hence, at 5% significance level, we conclude that more than 54%
students score 1 or 2.
D)
From above, I can summarize that monetary incentives have neither
resulted in decreasing the proportion of failing scores of 1 & 2 nor
resulted in increasing the higher proportion of score 5. Hence, we can
say that the monetary incentives in improving student achievement have
no or very less effectiveness.