Deck 19: Basic Principles of Analysis

Question 1

If there is some doubt about the lasting quality of an outcome, it is wise to :

A) Do a statistical test
B) Examine variability in the data
C) Continue monitoring
D) Terminate intervention

Answer 1

C

Question 2

One way of evaluating goal achievement is to use statistical significance as the criterion.

Answer 2

True

Question 3

If the change in problematic behavior is statistically significant, you know your intervention produced the change.

Answer 3

False

Question 4

If you see a discontinuity between the baseline and intervention data, this means your data are free of autocorrelation.

Question 5

None of the statistics available for analysis of time-series data should be used if the data are autocorrelated.

Question 6

First differences transformation is best used when:

A) There appears to be a linear direction (trend) in the data
B) The data are wildly fluctuating
C) The data are not autocorrelated
D) None of the above

Question 7

When comparing actual client changes with conceptually-based, predicted changes, you are examining

A) Practical significance
B) Clinical significance
C) Both a and b
D) Neither a nor b

Question 8

Effectiveness refers to the work that goes into a service program.

Question 9

When you attain statistical significance without clear practical and theoretical significance, it would be wise to:

A) Terminate
B) Maintain the present intervention
C) Do a withdrawal
D) None of the above

Question 10

An autocorrelation of .2 will inflate the t-test by 122%.

Question 11

If you measure how many hours you worked with a client, in the context of the large caseload you had to maintain at the same time, this would be an example of efficiency.

Question 12

The statement, "The change was statistically significant at the .05 level" means that there are only 5 out of 100 chances that the treatment did not cause the change.

Question 13

The term "analysis" is used in the text to refer only to the statistical significance of the data.

Question 14

You can always tell if your data are autocorrelated by visual inspection.

Question 15

If your data are auto-correlated, you could:

A) Not use statistics
B) Use statistics, but be aware of possible problems due to autocorrelation.
C) Transform the data
D) All of the above

Question 16

The first differences transformation plots the means between two data points over all baseline data.

Question 17

If you measured how many hours you worked with a client in relationship to what effects this had on changes in the client's behavior, this would be an example of effectiveness.

Question 18

If you see a notation that some outcome had a

Question 19

Yes or No? If you were able to show statistical significance and theoretical significance, but no practical significance in a given evaluated case situation, should you consider terminating the intervention for this target?