Research Design & Statistics: Foundation for Clinical Science Flashcards Preview

praxis > Research Design & Statistics: Foundation for Clinical Science > Flashcards

Flashcards in Research Design & Statistics: Foundation for Clinical Science Deck (21):
1

levels of measurement

nominal,
ordinal,
interval,
ratio scales

2

statistics defined

the field of study involved with data organization, summary, & analysis,
based on the theory of probability

3

external validity threats

limit the degree to which internally valid results can be generalized:
the hawthorne effect
(participants knowledge),
multiple-treatment interference,
reactive or interactive effects of pretesting

4

internal validity threats

result in failure to establish cause-effect relationship:
instrumentation,
history,
statistical regression,
maturation,
attrition,
testing,
subject selection biases,
interaction of factors

5

external validity

refers to generalizability,
to what settings, populations, treatment variables, & measurement variables the effect can be generalized

6

internal validity

degree to which data reflects a true cause-effect relationship

7

pretest-posttest design

prototype of experimental design groups,
similar experimental & control groups,
differ only after experimental group receives treatment,
allows evaluation of the effect of the independent variable on the behavior of the experimental group

8

ABAB reversal design

target behavior is base-rated (A),
taught to the subject (B),
reduced by teaching its counterpart or incompatible behavior (A),
then taught again (B)
to show the treatment was effective

9

confounding variable

the influence of extraneous variables
that may also affect the disorder or the skill

10

sample

in the exact middle of a distribution,
an equal # of scores fall above & below it

11

median score

small # of subjects drawn from the population,
should be randomly selected
& representative of the population

12

ABAB withdrawal design

target skills base-rated w/o treatment (A),
new treatment offered (B),
treatment withdrawn (A),
same treatment reinstated (B)

13

ordinal scale

represented by #s that can be arranged according to rank orders or levels,
intervals between #s are unknown & probably not equal

14

correlational relationships

0.00 indicates no relationship,
-1.0 to +1.0 range

15

ex post facto research

involves the effect of independent variables that have occurred in the past
& investigator is making retrospective search for causes of events

16

single-subject design

experimental design demonstrating cause-effect relationships,
involving anywhere from 1-6 subjects,
& recording individual performance,
ABA & ABAB designs are commonplace,
A phase = no treatment or baseline,
B phase = treatment,
multiple baseline designs also common

17

descriptive research

no demonstration of cause-effect relationships:
ex post facto
(case study or retrospective design),
survey,
comparative or standard-group comparison,
developmental or normative,
correlational,
ethnographic

18

independent variable

directly manipulated by the experimenter to cause changes in the dependent variable being studied

19

dependent variable

affected by the manipulation of the independent variable,
the effect

20

reliability

has scores that are consistent across repeated testing or measurement,
greater when closer to 1.00

21

concurrent validity

new test that correlates well with other established similar tests