quiz 2 Flashcards
Types of validity : face
makes sense according to a lay person
Types of validity : content
makes sense according to an expert
Types of validity : predictive
properly measure an outcome
Types of validity : concurrent
results are the same as other similar outcomes
Types of validity : discriminant
scores are high and low as expected
Types of validity : construct
measures what it is supposed to
Types of research validity : external
can the findings be generalized to a larger group than the one in the study
Types of research validity : internal
the extent to which the observed results represent the truth in the population we are studying
Types of research validity : statistical
is the statistical analysis properly chosen and used
anatomy of an article: 6 sections
aimrdr
- abstract
- introduction
- methods
- results
- discussion/conclusion
- references
what is selection bias
which articles you select to include in your review
which subjects are selected to participate in study
what is publication bias
studies w positive results are more likely to get published
hawthorne effect
acting differently bc you know u are being watched
recall bias
issue w retrospective studies: people may not recall things accurately
MEAN = what
what we know as the average
add them all and divide by the number of items
mode = what
the number that appears the most
more useful for categorical data
what is a correlational study
population survey
uses existing data
looks for statistical association between an exposure and an outcome
what is study heterogeneity
studies in a lit review are too different to be compared
what is a null hypothesis
the idea that your hypothesis will not work
suggests there is no difference between study groups
what are some descriptive study designs
CASE STUDY - simplest form of descriptive study - one client only
CASE SERIES - group of similar clients - reports trends
CORRELATIONAL STUDY - population survey - uses existing data
QUALITATIVE STUDY - does not include quantitative statistics or analysis
what are the 2 explanatory studies
experimental
observational / analytical
they establish the relationship between two variables
experimental
evaluates efficacy
researchers has an active role in the intervention
relationship between exposure and outcome
ethical considerations as researcher is controlling the environments
observational
seeks to establish cause, factors and predictors
the investigator observes what is happening
relationship between exposure and outcome
less ethical consideration as exposures and outcomes happen naturally
experimental explanatory studies
before and after tx design
(a practical way to conduct research in your own clinic)
involves pre determined hypothesis
eligibility criteria
methods
baseline data
treatment
outcome
more rigorous form of case series
often no control group
experimental explanatory studies
clinical trial
intervention study
randomized trial
randomized controlled trial
best design for establishing a cause and effect relationship
randomization and inclusion of a control reduces the rick of bias
explanatory: observational studies
CROSS SECTIONAL
data is simultaneously collected across a population at a particular point in time
provides a snapshot of a situation
different groups of subjects are compared at the same time
explanatory: observational studies
CASE CONTROL
“retrospective study”
exposure and outcome have already occurred
people w the outcome are identified and then compared w people who did not have the outcome
explanatory: observational studies
COHORT
longitudinal or prospective studies
outcome has not yet occurred
people are grouped bases on whether they are exposed or not, and then followed to see if they experience the outcome
literature review study design
summarize the existing evidence on a topic
a question is posed
guidelines are set for which types of research will be included in the review
the research is analyzed to see how strong it is
this is important because if a result is repeated over and over, it is more likely to be true - this is called REPRODUCIBILITY
types of literature reviews (3)
- narrative review - examines a group of studies
- meta analysis - group of studies that fit into a set of pre selected criteria
- systematic review
what do lit reviews summarize
EXISTING EVIDENCE
what is a more in depth search for articles considered
systematic review
more refined version of a meta analysis
what does p score tell you
the odds of getting a particular result
based on # of participants in a study
if p score is small enough then we conclude that the outcome is not random
less than 0.05 is key
what is a false positive
when the null hypothesis is falsely rejected (alpha)
what is a false negative
when the null hypothesis is falsely accepted (beta)
means missing out on a real effect or difference b/c your test results mistakenly suggest there isn’t one
“concluding there is no effect when there actually is one”
levels of evidence
systematic review
critically appraised topics (evidence synthesis and guidelines)
critically appraised individual articles (synopsis)
randomized controlled trials
cohort studies
case controlled studies - case series/reports
background information/expert opinion
what is external validity
can the findings be generalized to a larger group than the one in the study
what is internal validity
the extent to which the observed results represent the truth in the population we are studying
the sum of all probabilities always adds up to
1 or 100%
what is complement in probability
the chances a thing will NOT happen
p-value table P>0.10 means what (TOP OF CHART)
no evidence against the null hypothesis
data appears consistent w null hypothesis
p-value table 0.05 < P <0.10 =
weak evidence
0.01<P<0.05 =
moderate evidence
0.001<P<0.01 =
strong evidence
P<0.001 =
very strong evidence
variable types: nominal
category
can be dichotomous - yes or no
or can have more categories
variable types: ordinal
ranked order
but distance is not equal between variables
(frequency)
variable types: interval
numbers, ranked w equal intervals
rating scales
variable types: ratio
numbers are ranked, equal intervals, zero point is meaningful
descriptive statistics
summarize large group of numbers
measures of central tendency
mean, median, mode
descriptive statistics - normal distribution
bell curve
the mean is 0
standard deviation is 1
scores occur 50/50 on each side of mean
mean, median and mode are =
what are inferential statistics
used to make an inference about a population by measuring a part of the population
used for estimates and hypothesis testing
what is confidence interval
using the margin of error
what is a researchers job
to disprove a null hypothesis
NOT PROVE
correlation does not =
causation
when the null hypothesis is rejected then we know that
H0 - H1
what does the effect size tell us
tells us HOW wrong the null hypothesis is
really important in massage therapy
formula for effect size
mean treatment -mean control
OVER
standard deviation
what is odds ratio
one way of calculating an effect size
what are the odds that the outcome will happen if there is an exposure
you need
1. two dichotomous variables
2. assumption of a casual relationship
what is statistical power
likelihood that a significance test will find
related to sample size
larger effect = needs smaller sample size
what is the threshold for power calculation
80%
what does power =
1-beta
when are parametric tests used
when you have interval or ratio variables (numbers)
when are non parametric tests used
for nominal or ordinal variables categories)
what are 4 things to look for in clinical research
number needed to treat
sensitivity
specificity
minimal clinically important difference
what is validity
the extent to which the results are accurately measured
threats to validity - external
lack of randomization, lack of control
threats to validity - internal
confounds, bias, history, maturation, testing, instrumentation, regression, selection, mortality
threats to validity - statistical
sample size (power)
lack of standardization
types of reliability
inter-rater
do different raters get the same results
types of reliability
intra-rater
does the same rater consistently get the same result
types of reliability
test-retest
is the same result repeated, time and time again
what is reliability
the extent to which the outcomes are consistent when the experiment is repeated more than oncewh
what is validity
the extent to which the instruments that are used in the experiment measure exactly what you want them to measure
reliability vs validity
reliability= outcomes are consistent when repeated
validity = measure exactly what you want them to measure
