Biostats Notes on Two Sample Tests of Hypotheses

Flashcards for key concepts on two sample tests of hypotheses in biostatistics, covering independent and dependent samples, hypothesis tests, p-values, and confidence intervals.

What are independent samples in biostatistics?

Independent samples are two samples where one sample is not related to the other.

What are dependent samples also known as?

Dependent samples are also called paired samples or matched samples.

How do you determine if samples are dependent or independent with a follow-up questionnaire?

If the same subjects answer both the initial and follow-up questionnaires, the samples are dependent.

What is the null hypothesis form for paired samples?

Ho: There is no mean difference (Pa = do, for equal population means, do = 0).

What is the significance level assumed in hypothesis testing if not given?

A significance level of α is typically assumed to be 0.05.

What statistic is computed in hypothesis testing for paired samples?

Test statistic TS = (d - do) / (Sd / √n) where d is the mean of differences.

What does the p-value represent in hypothesis testing?

The p-value represents the probability of obtaining the observed results if the null hypothesis is true.

How is the confidence interval for paired samples calculated?

The confidence interval is calculated as P.E. ± M.O.E., where P.E. is the point estimate and M.O.E. is the margin of error.

What is the margin of error for a confidence interval?

The margin of error is calculated using the t-distribution and depends on the standard deviation and sample size.

What does it mean if p-value < α?

If p-value < α, then we reject the null hypothesis.

In a confidence interval, what does a result like (0.833, 8.167) mean?

We are 95% confident that the mean difference in scores lies between 0.833 and 8.167.

What is the alternative hypothesis for testing if the mean score after viewing is higher?

H₁: The mean score after viewing is greater than the mean score before viewing.

What does it indicate if Ho is rejected?

If Ho is rejected, there is sufficient evidence to support the research hypothesis.

What are the degrees of freedom for the paired t-test?

Degrees of freedom for the paired t-test is calculated as df = n - 1.

What is Bartlett's F-Test used for?

Bartlett's F-Test is used to test the equality of variances from two populations.

What skewness is associated with the F-distribution?

The F-distribution is right skewed.

What is the conclusion for a F-test if p-value < α?

Reject the null hypothesis, indicating that the variances are statistically different.

What should you do if significant deviations from normality occur in Bartlett's test?

Consider using Levene's Homogeneity of Variances Test instead.

How do you interpret a confidence interval that includes zero, such as (-1.10264, 4.60264)?

We cannot conclude that there is a significant effect; the mean difference could be zero.