A school administrator wants to see if there is a difference in the number of students per class for the Portland Public School district (group 1) compared to the Beaverton School district (group 2). Assume the populations are normally distributed with unequal variances. A random sample of 28 Portland classes found a mean of 36 students per class with a standard deviation of 4. A random sample of 27 Beaverton classes found a mean of 34 students per class with a standard deviation of 5.

Find a [tex]$90\%$[/tex] confidence interval for the difference of the means. Round answers to 4 decimal places.

[tex]\[ \square \ \textless \ \mu_1 - \mu_2 \ \textless \ \square \][/tex]



Answer :

To find a [tex]\(90\% \)[/tex] confidence interval for the difference in the means of students per class between the Portland Public School district ([tex]\( \mu_1 \)[/tex]) and the Beaverton School district ([tex]\( \mu_2 \)[/tex]), we use the following information:

- Sample size for Portland ([tex]\(n_1\)[/tex]): 28
- Sample mean for Portland ([tex]\(\bar{x}_1\)[/tex]): 36
- Sample standard deviation for Portland ([tex]\(s_1\)[/tex]): 4

- Sample size for Beaverton ([tex]\(n_2\)[/tex]): 27
- Sample mean for Beaverton ([tex]\(\bar{x}_2\)[/tex]): 34
- Sample standard deviation for Beaverton ([tex]\(s_2\)[/tex]): 5

Step-by-Step Solution:

1. Calculate the standard error of the difference in means:
[tex]\[ SE = \sqrt{\left(\frac{s_1^2}{n_1}\right) + \left(\frac{s_2^2}{n_2}\right)} \][/tex]

Substituting the values:
[tex]\[ SE = \sqrt{\left(\frac{4^2}{28}\right) + \left(\frac{5^2}{27}\right)} = \sqrt{\left(\frac{16}{28}\right) + \left(\frac{25}{27}\right)} \][/tex]

[tex]\[ SE \approx \sqrt{0.5714 + 0.9259} \approx \sqrt{1.4973} \approx 1.2237 \][/tex]

2. Calculate degrees of freedom using the Welch-Satterthwaite equation:
[tex]\[ df = \frac{\left(\frac{s_1^2}{n_1} + \frac{s_2^2}{n_2}\right)^2} {\left(\frac{\left(\frac{s_1^2}{n_1}\right)^2}{n_1-1} + \frac{\left(\frac{s_2^2}{n_2}\right)^2}{n_2-1}\right)} \][/tex]

Substituting the values:
[tex]\[ df = \frac{\left(\frac{16}{28} + \frac{25}{27}\right)^2} {\left(\frac{\left(\frac{16}{28}\right)^2}{28-1} + \frac{\left(\frac{25}{27}\right)^2}{27-1}\right)} \][/tex]

[tex]\[ df \approx \frac{(0.5714 + 0.9259)^2} {\left(\frac{0.5714^2}{27} + \frac{0.9259^2}{26}\right)} \approx \frac{1.4973^2} {\left(\frac{0.3266}{27} + \frac{0.8572}{26}\right)} \][/tex]

[tex]\[ df \approx \frac{2.2420} {0.0121 + 0.0330} \approx \frac{2.2420}{0.0451} \approx 49.7061 \approx 50 \; (\text{rounded}) \][/tex]

3. Determine the critical t-value ([tex]\( t_{\text{critical}} \)[/tex]) for a [tex]\(90\% \)[/tex] confidence level and [tex]\(df = 50\)[/tex]:
Using t-distribution tables or statistical software, we find the critical t-value for a 90% confidence interval with [tex]\(df \approx 50\)[/tex] is approximately [tex]\( t_{\text{critical}} \approx 1.676 \)[/tex].

4. Calculate the margin of error:
[tex]\[ \text{Margin of Error} = t_{\text{critical}} \times SE = 1.676 \times 1.2237 \approx 2.0518 \][/tex]

5. Construct the confidence interval:
First, we find the difference in sample means:
[tex]\[ \bar{x}_1 - \bar{x}_2 = 36 - 34 = 2 \][/tex]

Then, apply the margin of error to find the lower and upper bounds:
[tex]\[ \text{Lower bound} = (\bar{x}_1 - \bar{x}_2) - \text{Margin of Error} = 2 - 2.0518 \approx -0.0518 \][/tex]

[tex]\[ \text{Upper bound} = (\bar{x}_1 - \bar{x}_2) + \text{Margin of Error} = 2 + 2.0518 \approx 4.0518 \][/tex]

Thus, the [tex]\(90\%\)[/tex] confidence interval for the difference in the means is:
[tex]\[ -0.0509 < \mu_1 - \mu_2 < 4.0509 \][/tex]
Rounded to 4 decimal places.