Mathers Schmidt- Effect Size Continued

Front 1

what does statistically significant mean

Back 1

the probability is less than .5 or .1 and null hypothesis is void

Front 2

which test is stronger- parametric or nonparametric test

Back 2

parametric tests are stronger

Front 3

define evidence based practice

Back 3

3 componenets 1- systematic research 2-clinical expertise 3-taking clients and client's family member perspectives and values into account

Front 4

treatment effectiveness

Back 4

results of procedure applied in everyday practice

Front 5

treatment efficiency

Back 5

extent to which one treatment provides relatively better outcomes than others

Front 6

sensitivity

Back 6

% of participants who have a disorder and are correctly identified as having a disorder

Front 7

specificity

Back 7

% of individuals who do not have a disorder and are correctly identified as such

Front 8

how do we improve the external validity of a study

Back 8

random selection from a large pool of people. increase sample size.

Front 9

direct replication is

Back 9

duplicating the original study in every possible way

Front 10

systematic replication is

Back 10

change one factor of the study in order to tease out nuances and a variety of results. (ex. change from clinical setting to naturalisitic setting)

Front 11

how do single subject (time-series) designs differ from one-shot case study design

Back 11

the flexibility of the time frame

Front 12

how is experiemental control established in single-subject vs. group design

Back 12

the # of individuals

Front 13

can you test for the null hypothesis in a single subjet design

Back 13

no

Front 14

define a class I study

Back 14

well-controlled studies, e.g. randomized treatment and control-group studies and time-series single-subject designs

Front 15

define a class II study

Back 15

nonrandom group assignment research, program evaluation, quality improvement studies

Front 16

define class III studies

Back 16

case studies and historical reports. case studies have no control, because they are just studying one person at a time.

Front 17

list the levels of evidence considerations

Back 17

relevance, quality, number and consistency of findings. All of these support a clear link between clinical procedure and clinical outcome.