Print Page   |   Contact Us   |   Report Abuse   |   Sign In   |   Register
Biostatistics Questions

Sample Biostatistics Questions




1.  A one-tailed, one-mean Z test is conducted with the result that the null

     hypothesis is rejected.  What would be the outcome of a two-tailed test with

     the same null hypothesis, sample mean, alpha and standard deviation of the



       (a) The null hypothesis will be rejected.

       (b) The null hypothesis will be accepted.

       (c) We will fail to reject the null hypothesis.

       (d) Cannot tell the outcome, could be either reject or fail to reject

            depending on the data.


2. As a increases (e.g., from .01 to .05), the power of a one mean Z test


      (a) remains unchanged

      (b) increases

      (c) decreases

      (d) may increase or decrease, but the direction is not knowable.


3. Confidence intervals of the form are designed to estimate the

      (a) sample mean.

      (b) sample standard deviation.

      (c) population mean.

      (d) population standard deviation.


4.  What is the upper limit of the two-sided 90% confidence interval for the

       following sample:  X=14, 19, 15, 20, 26, 21?


       (a) 21.791

       (b) 23.090

       (c) 22.040

       (d) 22.749


5.   To carry out hypothesis tests or to compute confidence intervals for risk and

      odds ratios, we use the natural log of the sample risk or odds ratio in place of

      their actual values.  We do this because:


      a.  The sampling distributions of risk and odds ratios are approximately

           normally distributed.

      b.  The sampling distributions of  the natural log of risk and odds ratios

            are approximately normally distributed.

      c.  It increases the power to find significant risk and odds ratios.

      d.  It reduces the size of confidence intervals for any given level of



6.   What consequences does adjustment for multiple comparisons using

      Bonferroni’s procedure or a similar technique have?


      (a) It has almost no effect.

      (b) It increases the chance of making a Type I error.

      (c) It increases the power to find differences between pairs of group means.      

      (d) It reduces the power to find differences between pairs of group



7. A Chi Square test tells us


            (a) whether two continuous variables are correlated with each other.

           (b) whether two discrete variables are independent of each other.

           (c) the amount of variation in one variable explained by the other.

           (d) which categories of one variable are associated which categories of

                the other variable.





USF College of Public Health Alumni Society
12901 Bruce B. Downs Blvd, MDC 70, Tampa, FL 33612
Phone: 813-396-9665 I Fax: 813-974-2936

Membership Management Software Powered by YourMembership  ::  Legal