1. Asuming that the following sample of 10 numbers arise from a Normal distribution with mean µ and standard deviation s = 6.0, obtain, from first principles, 95% and 99% confidence intervals for µ.
39.4 22.1 37.6 26.4 26.5 29.6 25.4 30.9 28.3 22.4
Evaluate a test statistic for the null hypothesis H0: µ = 30 against the two-sided alternative hypothesis H1 : µ = 30, and by plotting this on its reference distribution, identify and calculate the corresponding p-value associated with H0. What are your conclusions about H0?
One-Sample Statistics
N
Mean
Std. Deviation
Std. Error Mean
VAR00001
10
28.8600
5.80731
1.83643
One-Sample Test
Test Value = 30
t
df
p-value
Mean Difference
95% Confidence Interval of the Difference
Lower
Upper
VAR00001
-.621
9
.550
-1.14000
-5.2943
3.0143
Comments:
Since calculated value is greater than the tabulated value therefore we reject Ho.
2. Repeat all calculations for the confidence intervals for µ and the p-value for H0 for the data in Q1 in the absence of any assumption on value of the standard deviation s. Do your conclusions about H0 change?
One-Sample Statistics
N
Mean
Std. Deviation
Std. Error Mean
VAR00001
9
29.5778
5.66961
1.88987
One-Sample Test
Test Value = 30
t
df
Sig. (2-tailed)
Mean Difference
95% Confidence Interval of the Difference
Lower
Upper
VAR00001
-.223
8
.829
-.42222
-4.7803
3.9358
Comments:
From above calculation we noticed that value of Ho changes as per change in data of assumption.
3. Tables show that the 95% critical values for some t distributions are (to 2 dp) as follows:
Df69 12 20 30
critical values ±2.45 ±2.26 ±2.18 ±2.09 ±2.04
Use statistical tables or the TINV function in EXCEL to find the critical values for the t distribution with the following degrees of freedom
5 8 15 18 24 60 120
Combine the two tables, and hence argue that these critical values approach ±1.96 as the number of degrees of freedom tends to infinity.
Briefly comment on why we would should expect ±1.96 to be the limit here.
Investigate the behavior of the 99% critical values for these degrees of freedom, and comment on the limits in this case.