Midterm Flashcards
What is research?
systematic inquiry
what are the 2 types of research?
Basic and Applied (clinical)
Types of Applied Research
explanatory, exploratory, descriptive
What is qualitative research?
natural phenomena
What is quantitative research?
controlled, measured at beginning and end
Efficacy
ideal conditions
controlled
CaC
Effectiveness
real-world scenarios
What does each part of PICO stand for?
P- patient
I- intervention/exposure (IV)
C- comparison (DV)
O- outcome
what are concepts?
operationally defined, measured (weight)
What are constructs?
abstract concepts
what are propositions?
linkages
Difference between deductive and inductive reasoning
Deductive reasoning is going from broad to specific
Inductive reasoning is going from specific to broad
“Ways of Knowing” triangle from big to small
scientific evidence
ded. <–> ind. reasoning
experience
authority
tradition
3 pillars of Evidence Based Practice
best available research evidence
patient values and preferences
clinical experience
Hierarchy of Evidence order (best to worst)
SR/MA RCTs
RCTs
SR cohort
cohort
SR case-control
case-control
case study
clinical experience
The Process of EBP (5As)
Ask, Acquire, Appraise, Apply, Assess
Research differs from clinical practice in what 3 ways
intents, innovative, plan
What are the 3 guiding ethical principles of research
respect for persons
benefidence
justice
Who are the vulnerable people within research?
children, prisoners, pregnant women, disability
Types of “harm” in research
physical, economic, and social
What are the ‘rules’ for the IRB
at least 5 members
different genders
not from 1 group
1 nonscience member
1 “public” member
name 2 examples of nominal and why it is
blood type, color, gender, religion
no order or rank
name 2 examples of ordinal and why it is
degree earned, pain, MMT
has order/rank but no equal groups
name 2 examples of interval and why it is
calendar, time, SAT/ACT, temperature
has order/rank, equal groups, but no true zero
name 2 examples of ratio and why it is
income, cm, degrees, age, weight, height
has order/rank, equal groups, and true zero
what is a continuous variable
ex: age, ROM, weight
a value that has “no end”
decimals possible
what is a discrete variable
only in whole units
ex: children, cups of coffee
is dichotomous
what are the 2 types of independent variables
active: treatment, manipulation (changes)
attribute: can’t change (race, eye color, shoe size)
Independent groups
different groups of people for each level
repeated measures
same people measured at all “levels” of IV
What is it called when there is one independent variable?
single factor, one-way design
What is it called when there is 2+ independent variables?
multifactoral design, “two-way” design
dependent variables name when there is more than one
multivariate designs
Dependent variables name when there is only 1
univariate designs
reliability definition
consistency
validity definition
accuracy, correctness
types of measurement error
systematic error, random error
explain systematic error
consistently off
explain random error
unpredictable, due to chance
measurement error explained
3 trials needed at least
more measurements = the closer (regress) to the mean
sources of measurement error
rater, instrument, variability
relative reliability
ICC & kappa
NO UNITS
Absolute reliability
standard error of the measurement (SEM)
UNITS
Types of reliability
test-retest
intrarater
interrater
test-retest
2 separate occasions
interrater
“between” raters, 2+ raters
intrarater
“with-in” rater
reliability for questionnaires
internal consistency, split-half, alternate forms
internal consistency
very specific
consistently off
questionnaires
types of validity
content, construct, criterion-related
types of construct validity
convergent, divergent, known groups
types of criterion-related validity
predictive, concurrent
what is face validity?
it does what it looks like
it is not useful and not studied
Minimal Detectable change (MDC)
value at which observed change is NOT measurement error
above this value = real change
below this value = measurement error
Minimal Clinically Important Difference (MCID)
the ability of an instrument to detect minimally important change
above this value = patient notices a difference
content validity
questionnaires
not way to objectify
group of people determine (yes/no) so it is Nominal (dichotomous)
Construct validity: convergent
together (both measure shoulder ROM)
Construct validity: divergent
separated (one shoulder and one knee ROM)
Criterion-related validity
gold standard, concrete thing
concurrent (criterion-related)
@ same time
diagnostic test (yes/no)
predictive (criterion-related)
with some certainty what will happen in the future
-ACL tests that mean high/low risk
define population
persons, objects, or events that meet criteria
define target population
larger population, who results will be generalized around
define accessible population
actual population chosen for the study
define sample
subgroup of population of interest
two types of selection (eligibility) criteria
inclusion and exclusion
define inclusion criteria
make someone eligible to be a participant
define exclusion criteria
keep someone from being the subject
sampling bias
sample systematically misrepresents population
-conscious
-unconscious
sampling error
sample is randomly misrepresenting population
probability sampling types
simple random
systematic
stratified random
cluster
simple random sampling ex
dice, out of a hat
systematic sampling ex
every 10th person
stratified random sampling ex
specific number from group/category
cluster sampling ex
multilayer/stages
types of nonprobability sampling
convenience, quota, purposive
what is convenience sampling
when subjects are chosen based on their availability
-volunteers/flyers
**most common
what is quota sampling
like stratified but not random
-don’t randomly pick groups
what is purposive sampling
subjects handpicked by specific criteria
random sampling is
how we obtain our sample
random assignment is
what we do with the sample
Nuremberg Code
first formal guidelines
voluntary consent to participate
competence of the investigator
Declaration of Helsinki
independent review of protocols
National Research Act
clearly stated research design
informed consent
Institutional Review Board (IRB)
Belmont Report
The Common Rule
- Respect for persons
- Beneficence
- Justice
Levels of Measurement: Number of Chairs
Ratio (and discrete)
Levels of Measurement: Color of markers
nominal
Levels of Measurement: results of a coin flip
nominal
Levels of Measurement: elapsed time
ratio
Measurement Error equation
error = observed score - true score
observed score = true score + or - error
if I measure 95 degrees, and the SEM = 5 degrees…the TRUE value ROM is ________
90 - 100 degrees