Problem No.
2(Two independent random sample test/ Fisher’s t-test/Population
variances are unknown but equal and samples are small)
Two levels of a certain fertilizer are tried to compare their effects on the increase
in yield of a local rice variety. The following are two random samples of
observations.
Per hectare yield in quintals
Level1 (X1) 34 32 30 35 34 34 32 30 31 35 35
Level1 (X1) 22 24 22 20 26 21 20 20 22 25 26 21 25
Test that there is no difference in the average effects of the two levels of the
fertilizer.
Solution to the Problem no.2
Here, 𝐻0 : 𝜇1 = 𝜇2
𝐻1 : 𝜇1 ≠ 𝜇2
We have, 𝑛1 = 11 𝑎𝑛𝑑 𝑛2 = 13
σ 𝑋𝑖 σ 𝑋𝑖
ത
We can also find 𝑋1 = ത
= 32.91 and 𝑋2 = = 22.62
𝑛1 𝑛2
σ(𝑋𝑖 −𝑋ത1 )2 σ(𝑋𝑖 −𝑋ത2 )2
𝑠12 = = 3.89 and 𝑠22 = = 5.26
(𝑛1 −1) (𝑛2 −1)
𝑛1 −1 𝑠12 + 𝑛2 −1 𝑠22 (10∗3.89+12∗5.26)
We know, s= = = 2.15
(𝑛1 +𝑛2 −2) 22
(𝑋ത1 −𝑋ത2 )
The test statistic t = 1 1
with (𝑛1 + 𝑛2 − 2) d.f.
𝑠 (𝑛 +𝑛 )
1 2
(𝑋ത1 −𝑋ത2 )
The test statistic t = 1 1
with (𝑛1 + 𝑛2 − 2) d.f.
𝑠 (𝑛 +𝑛 )
1 2
32.91−22.62
= 1 1
with 22 d.f.
2.15 (11+13)
= 11.96 with 22 d.f.
Comment: The critical value of t at 1% level of significance with 22d.f. is
2.82, the calculated value is greater than the critical value. Hence, we
reject the null hypothesis.
• To calculate sample variance we need the following table
𝑋1𝑖 2 𝑋2𝑗 2
𝑋1𝑖 𝑋2𝑗
34.00 1156 22.00 484.00
32.00 1024 24.00 576.00
30.00 900 22.00 484.00
35.00 1225 20.00 400.00
34.00 1156 26.00 676.00
34.00 1156 21.00 441.00
32.00 1024 20.00 400.00
30.00 900 20.00 400.00
31.00 961 22.00 484.00
35.00 1225 25.00 625.00
35.00 1225 26.00 676.00
21.00 441.00
25.00 625.00
σ 𝑋1𝑖 = 362 2
σ 𝑋1𝑖 σ 𝑋2𝑗 = 294 2
σ 𝑋1𝑖
= 11952 = 6712
• We know working formula for sample variance is
1 2 (σ 𝑋1𝑖 )2 1 (362)2
𝑠12 = σ 𝑋1𝑖 − = 11952 − = 3.89
(𝑛1 −1) 𝑛1 10 11
1 (σ 𝑋2𝑗 )2 1 (294)2
2
𝑠22 = σ 𝑋2𝑗 − = 6712 − = 5.26
(𝑛2 −1) 𝑛2 12 13