Hypothesis Test for mean differences of the populations (µ)
H 0 : μ 1 = μ2 H 0 : μ 1 = μ2 H 0 : μ 1 = μ2
H1 : μ1 > μ2 H1 : μ1 < μ2 H 1 : μ 1 ‡ μ2
z-test > z-critic
z-test > z-critic z-test < - z-critic
Reject H0 z-test < z-critic
Reject H0
Reject H0
when population variance is known Use Z test. when population variance is unknown and n1 > 30, n2 > 30
The average durability duration of A brand and B brand floor The information about two samples taken from two different
polishes is to be compared. 20 samples of A brand floor polish populations, which are assumed to be normally distributed, is
are taken and it is observed that they have an average durability given in the table below. It is assumed that the populations from
of 3 months. 15 samples of B brand floor polish are taken and it which these samples are taken have the same means. Test whether
is observed that they have an average durability of 2.9 months. there is a difference in the difference of the means at a
The population variances of the durability duration of A and B significance level of α=0.05.
brand floor polishes are 0.1089 and 0.1296 respectively. Test Sample Number of data Mean St.deviation
whether the A brand floor polish is more effective than the B A 40 5 1
brand floor polish at a significance level of α=0.05. B 50 4.7 1.2
Given information: Given information:
Sample size for A (nA) = 20 Sample size for A (nA) = 40
Sample mean for A (x̄A) = 3 Sample mean for A (x̄A) = 5
Sample size for B (nB) = 15 Sample size for B (nB) = 50
Sample mean for B (x̄B) = 2.9 Sample mean for B (x̄B) = 4.7
Variance for A (s2A) = 0.1089 Standard deviation for A (sA) = 1
Variance for B (s2B) = 0.1296 Standard deviation for B (sB) = 1.2
Significance level (α) = 0.05 Significance level (α) = 0.05
H0: μA = μB H0: μA = μB
H1: μA > μB H1: μA ‡ μB
Calculate the z-test. Calculate the t-test.
z = [(3 - 2.9) – 0]/ √(0.1089/20 + 0.1296/15)
t = (5 - 4.7) / √((12/40) + (1.22/50))
z = 0.1 / √(0.00545 + 0.00864)
z = 0.1 / √0.01409
t = 0.3 / √(0.025 + 0.0288)
z = 0.1 / 0.1187 t = 0.3 / √0.0538
z = 0.8426 ≅ 0.843 t = 0.3 / 0.2319
t = 1.2923 ≅ 1.293
Determine the critical z-value from the z-table. For a two-tailed test with α = 0.05 and (nA + nB - 2) = 40 + 50 - 2
For a one-tailed test with α = 0.05, the critical z-value is 1.645. = 88 degrees of freedom, the critical value is ±1.987.
Since the calculated z-value (0.8426) is less than the critical value Since the calculated t-value (1.2923) is less than the critical value
(1.645), we fail to reject the null hypothesis. z-test < z-critic (1.987), we fail to reject the null hypothesis. t-value < t-critical value
At a significance level of 0.05, there is not enough evidence to At a significance level of 0.05, there is not enough evidence to
conclude that the A brand floor polish is more effective than the conclude that the means of the two populations are different.
B brand floor polish in terms of durability.
2nd WAY: 2nd WAY:
when population variance is unknown and equal to each other when population variance is unknown and equal to each other
(s 2 1 = s 2 2 ) (s 2 1 ≠ s 2 2 )
Two machines are used to fill plastic containers with dishwashing
detergent. From the first machine, n1 = 10 plastic containers are
selected, and from the second machine, n2 = 12 plastic containers
are selected. Upon examination, the first machine has filled the
containers with an average of 30.87 units of liquid, while the
second machine has filled the containers with an average of 30.68
units of liquid. The variances are found to be 0.0225 and 0.0324,
respectively.
Given this information, can it be said that the first machine fills
more liquid, assuming equal population variances, at a 5% Can it be said that the first machine fills more liquid, assuming
significance level? unequal population variances, at a 5% significance level?
Given information:
Sample size for first one (n1) = 10
Sample size for second one (n2) = 12
Sample mean for first one (x̄1) = 30.87
Sample mean for second one (x̄2) = 30.68
Variance for first one (s21) = 0.0225
Variance for second one (s22) = 0.0324
Significance level (α) = 0.05
H0: μ1 = μ2 (The means of the two populations are equal) H0: μ1 = μ2
H1: μ1 > μ2 (The first machine fills more liquid) H1: μ1 > μ2
Use t test. df = (0.0225/10 + 0.0324/12)2 / [(0.0225/10)2/ 9 +
(0.0324/12)2/11] = 19.99 ≅ 20
t test= (30.87 - 30.68)-0 / √ [0.0225/10 + 0.0324/12] ≅ 2.7
Pooled sample variance:
s2 = [(10-1)(0.0225) + (12-1)(0.0324)] / (10+12-2)= 0.027945
t = (30.87 - 30.68)-0 / (√0.02745(√(1/10 + 1/12)) ≅ 2.654
For a one-tailed test with α = 0.05 and (n1 + n2 - 2) = 10 t-critic value= t0.05,20 = 1.725
+ 12 - 2 = 20 degrees of freedom, the critical value is
1.725.
The calculated t-value (2.654) is greater than the critical The calculated t-value (2.7) is greater than the critical
value (1.725), so reject the null hypothesis. value (1.725), so reject the null hypothesis.
At a 5% significance level, there is enough evidence to At a 5% significance level, there is enough evidence to
conclude that the first machine fills more liquid than the conclude that the first machine fills more liquid than
second machine. the second machine.
Hypothesis Test for matched paired differences Hypothesis Test for differences of
population proportions
A paired t-test is used to compare two population means where you have
two samples in which observations in one sample can be paired with
observations in the other sample.
In a region, the prices of 8 randomly selected apartments for January and It is claimed that the old building rates in two
June are given below. Is there a statistically significant difference neighborhoods are different. For this purpose, 30 out
between the prices? (α=0.05) of 40 randomly selected buildings from the 1st
neighborhood and 35 out of 50 buildings from the
January 100.000 105.000 120.000 130.000 135.000 140.000 155.000 160.000 2nd neighborhood are old. Accordingly, test the
June 104.000 105.000 125.000 138.000 145.000 150.000 170.000 180.000
difference 4.000 0.000 5.000 8.000 10.000 10.000 15.000 20.000
accuracy of the claim at a 5% significance level.
H0: μd = 0 H0: p1 = p2
H1: μd ≠ 0 H1: p1 ≠ p2
sd2 = Σ(d - d̄)22 / (n-1)
Mean of the differences: xd = (4.000 + 0.000 + 5.000 + 8.000 + 10.000 + the sample proportions:
10.000 + 15.000 + 20.000) / 8 = 9.000 (p"1) = 30/40 = 0.75
(p"2) = 35/50 = 0.70
Variance of the differences: sd2 = Σ(d - d̄)22 / (n-1)
z = (0.75 - 0.70) / √(0.75).(0.25)/40 + (0.70).(0.30)/50
2 2 2 2
sd = [(4.000 - 9.000) + (0.000 - 9.000) + (5.000 - 9.000) + (8.000 - z ≅ 0.53
9.000)2 + (10.000 - 9.000)2 + (10.000 - 9.000)2 + (15.000 - 9.000)2 +
(20.000 - 9.000)2] / (8-1) = 40.2857
Standard deviation of the differences: sd = √40.2857 = 6.3577
t = (9.000 - 0) / (6.3577 / √8) = 4.0000
Determine the critical value: At a 5% significance level α=0.05, the critical value
For a two-tailed test with α = 0.05 and (n-1) = 8-1 = 7 degrees of for a two-tailed z-test is ±1.96.
freedom, the critical value is t0.025,7= ±2.365.
The calculated t-value (4.0000) is greater than the critical value of The calculated z-value (0.53) is within the critical
2.365, so we reject the null hypothesis. region (-1.96 to 1.96), so we fail to reject the null
hypothesis.
At the 5% significance level, there is a statistically significant At the 5% significance level, there is not enough
difference between the January and June apartment prices. evidence to conclude that the proportion of old
buildings in the 1st neighborhood is different from
the proportion of old buildings in the 2nd
neighborhood.
