We will check if there is a statistically significant difference between the two groups in terms of the proportion of workers who failed the course.
Let p_A be the proportion of workers who failed in group A and p_B be the proportion of workers who failed in group B.
The null hypothesis is that there is no difference in the proportions, which can be stated as:
H_0: p_A = p_B
The alternative hypothesis is that group B takes better advantage of the instruction:
H_1: p_A > p_B
The significance level given is 3%, which means \alpha = 0.03 .
Now, we will calculate the z-score and compare it to the critical z-value for rejection.
First, calculate the standard error of the difference between two sample proportions:
SE = \sqrt{p_{pool} \times (1 - p_{pool}) \times \left(\frac{1}{n_A} + \frac{1}{n_B}\right)}
where p_{pool} is the pooled sample proportion:
p_{pool} = \frac{X_A + X_B}{n_A + n_B}
X_A and X_B are the number of failures in groups A and B, respectively.
Then, calculate the z-score:
z = \frac{(p_A - p_B)}{SE}
Next, find the critical z-value at a significance level of 3%. Since it's a one-tailed test (we're checking if group B takes better advantage), the critical value is obtained by finding the z-value with a cumulative probability of 97%:
z_{\alpha} = 1.88
If the calculated z-score is greater than 1.88, we reject the null hypothesis.
Given:
- Group A: 25 failed out of 180 workers (n_A = 180, X_A = 25)
- Group B: 21 failed out of 170 workers (n_B = 170, X_B = 21)
Calculations:
p_{pool} = \frac{25 + 21}{180 + 170} = \frac{46}{350} \approx 0.1314
SE = \sqrt{0.1314 \times (1 - 0.1314) \times \left(\frac{1}{180} + \frac{1}{170}\right)} \approx 0.0335
p_A = \frac{25}{180} \approx 0.1389
p_B = \frac{21}{170} \approx 0.1235
z = \frac{0.1389 - 0.1235}{0.0335} \approx 0.459
We compare z = 0.459 to the critical z-value of 1.88. Since 0.459 < 1.88, we fail to reject the null hypothesis.
Therefore, we do not have enough evidence to conclude that group B takes better advantage of the instruction at a 3% significance level.