The flashcards below were created by user
epm49
on FreezingBlue Flashcards.
-
What does this describe:
What would have happened to the same people exposed to a causal factor, if they were simultaneously not exposed to the causal factor
The term "counterfactual"?
-
Name 3 criteria for establishing causality in a study?
- 1. temporal: cause must precede an effect in time.
- 2. empirical relationship btw the presumed cause and presumed effect
- 3. 3rd variable: relationship cannot be presumed to be due to a 3rd variable
-
What research design is considered "the gold standard" for establishment of causal relationships in variables
RCT's
-
What are the 3 key characteristics of an RCT?
- 1. manipulation
- 2. control:
- 3. randomization
-
In RCT's, the design feature of manipulation means that researchers manipulate the ________ by administering a __________ to some subjects and withholding it from others
- independent variable
- intervention/treatment
-
The independent variable is always considered the presumed ____________ of the phenomenon
cause
-
The dependent variable is considered the presumed ______________
outcome
-
In this example: what is the IV and the DV?
Gentle massage is an effective pain relief strategy for nursing home residents
- IV: massage
- DV: pain relief
-
What is a PCI?
Patient controlled intervention
-
What is meant by an "attention control group"
an added study group that hopes to control for the effect of "attention" that the participants receive from the intervention
-
Define "random" as in rand assignment or random allocation
Each subject has an equal chance of being assigned to any group
-
What is the "Table of Random numbers?"
table where any digit from 0-9 is equally likely to follow any other digit; going in any direction from any point in the table produces a random sequence
-
What is allocation concealment?
random assignment should involve this method which prevents those who enroll participants from knowing upcoming assignments; intended to prevent bias that could stem from knowledge of allocations before assignments actually occur
-
What is cluster randomization?
assigning groups or "clusters" of people to different treatment groups
-
________, _________and _______, and should all be done prior to random assignment?
- screening for eligibility
- obtaining informed consent
- collection of baseline data
-
What is baseline data?
Pre-intervention data which are collected to measure key outcomes
-
What is randomized consent AKA a "Zelen design?"
a randomization variant where subjects are randomized first AND then informed consent is obtained, thus eliminating possibility that the consent process will generate preferences and drop outs
-
Blinding, or masking prevents biases stemming from __________
awareness
-
________ involves concealing information from participants, data collectors, care givers, intervention agents or data analysts in order to enhance __________ and minimize_______
- Blinding or masking
- objectivity
- expectation bias
-
________bias refers to systematic differences in care provided to members of different groups of participants, apart from an intervention that is the focus of the inquiry
Performance
-
________ bias concerns systematic differences between groups in how outcome variables are measured, verified, or recorded, is addressed by blinding those who collect the outcome data, or in some cases those who analyze it
Detection or ascertainment
-
An ______study is one in which blinding is not used. A ________study results from masking.
-
When blinding is only used for one group of people (e.g. study participants) is is called___________
single-blind
-
_______ involves masking those who receive and those who deliver the treatment
double-blind
-
A _________ design involves data collection on the dependent variable only once, after randomization and completion of the intervention.
- post-test only/after-only design
- Basic Experimental Design
-
A ___________ design involves testing subjects at least twice (before and after the intervention.) In this design, researchers cn measure _________.
- Pre-test-post-test design/before-after design
- change
-
A _________ design includes data collection at multiple post-intervention points
repeated measures
-
Designs that involve collected data multiple times from two groups can be described as __________. Analyses can examine both differences ________ and changes _______ over time
- Mixed designs
- between groups
- within groups
-
An experimental design in which 2 or more independent variables are simultaneously manipulated, permitting a separate analysis of the main effects of the IVs and their interaction.
Factorial
-
Effects from experimentally manipulated variables
Main effects
-
Effects from combining treatments
interaction effects
-
An experimental design in which one group of subjects is exposed to more than one condition or treatment, preferably in random order.
Cross-over design
-
The process of systematically varying the order of presentation of stimuli or treatments to control for ordering effects, especially in a cross-over design.
counterbalancing
-
The influence that 1 treatment can have on subsequent treatments, notably in a cross-over design.
carry-over effects
-
A period of no treatment exposure to reduce the potential concern of carry-over effects
washout period
-
The effect on the dependent variable resulting from subjects' awareness that they are participants under study
Hawthorne Effect
-
Another term for Quasi-Experimental designs which involve an intervention in the absence of randomization
Controlled trials without randomization
-
Design includes 2 groups of participants, from whom outcome data are collected before and after implementing an intervention.
Non-equivalent control group pretest-posttest design
-
The term used in Quasi-experiments in lieu of control group?
comparison group
-
a sophisticated method of matching subjects on more than a couple variables
propensity matching
-
a metric that captures the conditional probability of exposure to a treatment given various pre-intervention characteristics; can be used to match comparison groups or as a statistical control
propensity score
-
In lieu of using a contemporaneous non-randomized comparison group, a _______ is often used
Historical comparison group
-
A quasi-experimental design involving collection of data over an extended period of time, with multiple data collection points both prior to and after the intervention
Time series design
-
An intervention study that tests the effectiveness of of an intervention with a single subject, typically using a time series design; sometimes called________
single subject experiment or "N of 1 experiment"
-
In quasi-experiemental designs, this is an analysis to assess whether larger doses of an intervention are associated w/ greater benefits
Dose-response analysis
-
A threat to internal validity in quasi-experimental designs; an alternative explanation, competing w/ the researcher's hypothesis, for interpreting the results of the study
rival hypotheses
-
studies in which the researcher collects data without introducing an intervention; also called observational research
non experimental research
-
These study designs are used in non experimental research where researchers study the effect of a potential cause that they cannot manipulate
correlational studies
-
a relationship or association between 2 variables, or a tendency for variation in one variable to be related to variation in the other variable
correlation
-
A study design that begins with the manifestation of the dependent variable (lung cancer) in the present, followed by a search for the presumed cause occurring in the past (smoking)
retrospective study design
-
a non-experimental design involving the comparison of a "case" and a matched control
case control design
-
a study design that begins with an examination of presumed causes and then goes forward in time to observe presumed effects; also called a _______ design
- Prospective study
- cohort design
-
a non-experimental study that takes advantage of a naturally occurring event that is explored for its effect on people's behavior or condition, typically by comparing people exposed to the event with those who were not
natural experiment
-
A regression-based procedure for testing causal models, typically using correlational data
path analysis
-
a statistical procedure for predicting values of a dependent variable based on one or more independent variables
regression analysis
-
a type of non-experimental, correlational research which seeks to describe relationships among variables rather than support inferences of causality
descriptive correlation research
-
a study that gathers information on the occurance, frequency of occurance, or average value of the variables of interest, one variable at a time, without focusing on interrelationships among variables
Univariant descriptive studies
-
Two types of descriptive studies from field of epidemiology
- 1. prevalence studies
- 2. incidence studies
-
__________studies are done to estimate the _________ rate of some condition at a particular point in time; rely on __________ designs in which data are obtained from the population at risk of the condition.
- prevalence
- prevalace
- cross-sectional
-
What is the formula for a Prevalence Rate?
- # of cases with the condition/disease
- at a given point in time (subjects in study with the condition)
- ___________________________________
- # in the population at risk of being a case
- (size of the sample)
-
___________ studies estimate the frequency of developing new cases
Incidence
-
__________ designs are need to estimate incidence b/c the researcher must first establish who is at risk of becoming a new case; that is who is free of the condition at the outset
Longitudinal
-
an estimate of risk of "caseness" in one group as compared to another, computed by dividing the absolute risk for one group by the absolute risk for another; also called the _______
relative risk or risk ratio
-
Limitations of correlational research include a weakness in their ability to _____________
support causal influences
-
In correlational studies, researchers work with pre-existing groups that were not formed at random, but rather via _____________, which increases _________bias.
-
_______________ may be a plausible alternative explanation for any group differences on the outcome variable in correlational studies.
Preexisting differences among groups
-
Correlational research is very efficient in ______________about a problem.
collecting a large amount of data
-
3 broad design categories
- 1. non experimental; observational
- 2. quasi-experimental
- 3. experimental
|
|