Justifying Claims: Difference of Two Means - AP Statistics

$(\bar{x}_1 - \bar{x}_2) ,\pm, z^* \sqrt{\frac{s_1^2}{n_1} + \frac{s_2^2}{n_2}}$. Uses difference of sample means plus/minus critical value times SE.

$H_0: \mu_1 - \mu_2 = 0$. States no difference exists between population means.

$H_a: \mu_1 - \mu_2 \neq 0$. States a difference exists between population means.

$\sqrt{\frac{s_1^2}{n_1} + \frac{s_2^2}{n_2}}$. Measures variability in the sampling distribution of differences.

$z^* \times \text{standard error}$. Critical value multiplied by standard error gives margin of error.

Approximately $1.96$. Standard normal value that captures 95% of distribution.

Check if $0$ falls within the interval range. Zero in interval suggests no significant difference between means.

It indicates no difference between means if in interval. Zero represents no difference between the two population means.

Sample size large or population normal. CLT applies with large samples; normality assumed otherwise.

When population standard deviations are unknown. t-distribution accounts for additional uncertainty from unknown σ.

More precision in the estimate of the mean difference. Smaller interval range indicates more precise estimation.

Mean of first group is larger than second group. All values positive means $\mu_1 > \mu_2$ consistently.

Mean of first group is smaller than second group. All values negative means $\mu_1 < \mu_2$ consistently.

Increases width of confidence interval. Higher confidence requires wider interval to capture parameter.

Wider confidence interval. Greater spread in data increases uncertainty in estimates.

Estimate range where a population parameter lies. Provides plausible range for unknown population parameter.

Independence between samples. Samples must be independent for valid statistical inference.

Observed effect unlikely due to chance alone. Result is unlikely to occur by random variation alone.

Higher variability leads to wider intervals. More variable data creates less precise interval estimates.

Single value estimate of a population parameter. Single best guess for the unknown parameter value.

No significant difference between means. Zero suggests means could be equal; no significant difference.

Sample means of two different groups. Sample means from groups 1 and 2 respectively.

Larger sample size reduces margin of error. Larger n decreases standard error, reducing margin of error.

Higher confidence level means less precision. Trade-off between confidence and interval width exists.

Lower bound is $5$. Lower bound is the smaller endpoint value.

Upper bound is $15$. Upper bound is the larger endpoint value.

Indicates a significant difference exists. Zero excluded means difference is statistically detectable.

CI includes $0$; no significant difference. $0$ in CI means no significant difference detected.

Wider confidence interval. Smaller n increases standard error, widening the interval.