Read the summary and the most important questions on Management research methods 2

• 1 ANOVA

  This is a preview. There are 11 more flashcards available for chapter 1
  Show more cards here

• When can one use ANOVA?

  - The OV is quantitative
  - The PV is categorical, so has two or more categories (0,2 or 4 beers predict if something will happen)
  - participants differ from each other (different test groups do not have the same participants)
  - population is normally distributed
  - populations must have the same variance

• Variance testing can distinguish two tests, which two?

  Variance testing within groups
  Variance testing between groups (this is the one that says the most)

• What does the F-test tell (ANOVA)?

  Variance between sample means / variance within the samples
  
  
  Variance between sample means = mean sum of squares of the model / Degrees of Freedom (number of groups - 1)
  
  Variance within the samples = Mean square residuals / Degrees of Freedom (number of participants - number of groups)

• When is there effect from a PV on an OV?

  When the explained variance (between groups) is large relative to the unexplained variance (within groups) --> High F score

• What is a Brown-Forsyth or Welch test?

  It is a criterion that tells if ANOVA can be done P > 0.05

• What happens with R2 when there is only one PV?

  Than R2 becomes eta2

• A T-test cannot be done to test for statistical evidence because there are used more than 2 groups to test (3x40N). Why can one not do multiple t-tests on these groups to find evidence?

  Because of alpha inflation (type 1 error)
  
  To test with 3 groups, three t-tests have to be done.
  
  1-0.95 = 0.05
  1-0.95^3 = 0.14

• What are the two criteria for planned contrast tets?

  1. Sum of the weights must be 0
  2. If multiple contrasts are done (so multiple hypotheses are tested) they must be orthogonal. This means that the multiplied comparison of the tetst must be 0. 
  
  
  Example: H1 = group 2 > group 3 --> 0, 1, -1
                  H2 = group 1 > rest --> 1, -0.5, -0,5
  
  multiplied comparison: 
  0*1 = 0
  1*-0.5 = -0,5
  -1*-0.5 = 0.5
  
  These added up is 0, so they are orthogonal!

• What does the Levene probability tell?

  If the significance is >0.05, than equal variances can be assumed.

• What is a post hoc test?

  Statistical test afterwards to check the reliability of the earlier tetst done.