A statistician measures the difference in songbird warble ...

[Mollykgkg]

A statistician measures the difference in songbird warble variation in twenty-five mockingbirds, both before and after nutritional supplementation with Dr. Petit Oiseau's Musical Bird Seed.
 
  Warble variation is determined by a cranky librarian, who characterizes the birds, before and after treatment, as one of the following Lots of Notes There Fair Variety, and Quite The Redundant Little Song. What considerations determine choice of a test for this research? (Select all that apply.)
  a. The sample size is small to moderate.
  b. There is one group of birds and all receive the treatment.
  c. A before-after type of design is used.
  d. The independent variable is nominal level.
  e. Birds are divided into two groups.
  f. The dependent variable is ordinal level.
  g. The dependent variable is ratio level.

Question 2

Below the third table in a research report, concerning switching infants to cow's milk at six months of age instead of 12 months, is a notation: t(22) = 0.76, p = .41 What does this mean? (Select all that apply.)
 
  a. A t-test was performed.
  b. Apparently 24 subjects participated in the study.
  c. There was a difference between groups.
  d. The difference between groups was not statistically significant.
  e. If there was an intervention, the intervention was more effective than the control condition.
  f. The results are generalizable.
  g. The researcher failed to reject the null hypothesis.
  h. A type I error occurred.

[blfontai]

Answer to Question 1

ANS: B, C, D, F
The parametric statistics used to determine differences that are covered in this chapter are the independent samples t-test, paired or dependent samples t-test, and analysis of variance (ANOVA). If the assumptions for parametric analyses are not achieved or if study data are at the ordinal level, then the nonparametric analyses of Mann-Whitney U, Wilcoxon signed-ranks test, and Kruskal-Wallis H are appropriate techniques to use to test the researcher's hypotheses. In this study, one group is compared before and after treatment, and the dependent variable is at the ordinal level of measurement. In such an instance, a Wilcoxon signed-ranks test is appropriate. This test is useful when a paired t-test would otherwise be performed, if only the dependent variable weren't ordinal.

Answer to Question 2

ANS: A, B, C, D, G
One of the most common parametric analyses used to test for significant differences between group means of two samples is the t-test. It is conducted to discover whether a difference exists between two groups, or between a previous condition and the current condition in paired subjects. Degrees of freedom are equal to n  2. Not enough is known about the research to discern in which group the intervention was more effective. In this example, there was a difference between groups, but it was apparently not statistically significant, since p = .41, according to the notation below the graph. Only p-values of .05 and less would be statistically significant. If the researcher failed to reject the null hypothesis, no type I error could have occurred.