Exam three - Selecting the correct test

Question 1

What are the four questions to ask?

Answer 1

1. What data level is recorded
2. What TYPE OF COMPARISON/ASSESSMENT is desired?
3. How many groups are being compaired
   
   1. two or three different groups
4. is the data independent or related?
   
   1. o Data from the same (paired) or different groups (independent)
   2. “Pre- vs. Post-”, “Before vs. After”, “Baseline vs. End”

Question 2

Fillout the image

Answer 2

Question 3

Attach Sounds

Answer 3

Question 4

Fill out the diagram below

Answer 4

Question 5

Answer 5

Question 7

What kind of data does coincides with which level

Answer 7

• Nominal = mode
  
  • in terms of freq. and proportions
• Ordinal = median
  
  • in term of freq and proportions
• Interval = mean

Question 8

What does a chi test of independence test for and how does it relate to the cochran test?

Answer 8

• Test for the difference in 3 groups of data in an ordinal data. It does not idnetify where the difference occurs but the fact that there is one.
• Assumptions :
  
  • Chi square dsitribution for nominal data
  • No cell count 5< observation
• The cochran test applies to ordinal data but essential measures the same thing

Question 9

What are Bonferroni correction test used for ?

Answer 9

• it adusts the P value for for # comparisons being made.
• very conservative
  *

Question 10

What is the tukey scheffe tests used for?

Answer 10

• Compares all pairwise comparisons possible
• All groups must be equal in size
  
  • tukey test slightly more conservative than the stuN.K
  • Scheffe test less affected by violations in normality and homogenetity of variences - most effects

Question 11

What is the student newman Keul test?

Answer 11

• Compares all pairwise comparisons possible
• all groups must be equal in size

Question 12

what is the dunnett test?

Answer 12

• Compares pairwise comparisons against a single control
• all groups must be equal in size

Question 13

What is the dunn test?

Answer 13

Compares all pairwise comparisons possible

useful when all groups are not of equal size

Question 15

When and why should you used an post-hoc test?

Answer 15

for stastically significant findings (P< 0.05) in 3 or more comparisons.

TO determine where the differencies in stastically significant data (when their are 3 or more groups)

Question 16

Why are mutliple chi square tests never acceptable?

Answer 16

Risk of type 1 error increases with each additional test (almost guarnteed after 4 -5) test

Question 17

What is a kappa statistic?

Answer 17

A correlation test showing relationship or agreement between evaluators (consistency of “decisions”, “determinations” )

Question 18

How are kappa interpretations made?

Answer 18

+1 = the obserbers perfectly “classify” everyone exactly the same way

0 = there is no relationship at all between the observers “classifications”, above the agreement that would be expected by chance

-1 = the observers “classify” everyone exactly opposite of each other

Kappa (k) value can be + or - ; ( + ) is good agreement; ( - ) is poor agreement