Ordinal Data

• When data is ordinal, we can calculate a....

  Median and Mode
• t-tests compare the means of experimental groups, so can be conducted on what type of data?

  Interval and Ratio
• Mann-Whitney U test

  (1) Data is grouped together and ranked

  (2) Data is put back into original groups

  (3) Calculate sum of the ranks for each group separately
• As the U-value gets smaller, the P​-value gets…

  Smaller
• The smaller the U-value, the…

  Smaller the probability of observing our results, if the null hypothesis is correct and the bigger the probability of observing our results, if the null hypothesis is incorrect.
• The smaller the U-value, the less likely it is that the null hypothesis is correct, and the more likely we are to reject our null hypothesis.
• A smaller t-value gives us a…

  Bigger p-value
• t-value = less likely to reject null hypothesis

  U-value = more likely to reject null hypothesis
• The smaller the U-value…

  The smaller the p​-value.

  The more likely we are to accept the alternative hypothesis.
• The bigger the difference between the sum of the ranks of the two groups…

  The smaller the U-value

  The smaller the p-value
• In order to reject the null hypothesis our observed U-value must be…

  Less than or equal to the critical U-value.
• Why do we see rankings, like 7.5​, that are not integers?

  More than one person was ranked in the same place.
• If the researchers are using a matched pairs design, they should use…

  the same test as they would in a repeated measures design. which in this case is the Wilcoxon test.
• When researchers have ordinal data, and their experiment uses an independent groups design, they use a Mann-Whitney U Test.

  When researchers have ordinal data, and their experiment uses a repeated measures design, or a matched pairs design, they use a Wilcoxon Test.
• In the Mann-Whitney test, as the difference between before-and-after ratings increases, the U-value decreases.
• Wilcoxon Test

  The value we calculated tells us how far apart the before and after ratings are.

  The larger the difference between before and after ratings, the smaller the value.

  We reject the null hypothesis if our T​-value is less than or equal to the critical T​-value
• In the table of critical values for the Wilcoxon test, N is...

  The sample size subtract the people whose scores didn’t change from before to after.