Non experimental research designs Flashcards
1
Q
quasi-experimental designs
A
- structure similar to experimental designs but lack either
- random assignment
- comparison groups
- or both
2
Q
quasi-experimental designs: one group pretest-posttest design
A
- quasi-experimental designs
- one set of repeated measures before and after treatment from one group of subjects
- effect of treatment–difference between pretest and posttest scores
- one IV (time) – 2 levels (pretest and posttest)
3
Q
one group pretest-posttest design- validity
A
- draw conclusion with caution; no comparison groups, can demonstrate change in DV, events other than treatment could have caused change
- can use when: previous research has documented behavior of a control group in similar circumstances
- time interval between measurements is short
4
Q
Exploratory designs –observation studies
A
- analyze how existing factors or individual characterisitics are associated with health outcomes
- intention: frequently to generate hypotheses about factors that may give rise to differences in population groups
- speculative in nature
5
Q
Exploratory designs –observation design: estimating risks
A
- a common use of observation design is examining effects of hypothesized risk factor
- compare outcomes of groups of subjects exposed to risk factor with group of subjects not exposed
- researches do not alter exposure
- results can suggest possible preventative intervention but do not provide conclusive evidence of causal link
- explanatory trials provide compelling evidence of causation
6
Q
longitudinal studies
A
- researchers follow subjects through time
- take measurements at prescribed intervals
- can be prospective (forward in time) or retrospective (backward in time)
7
Q
Prospective longitudinal studies
A
- information about potential risk factors recorded at start of study
- exposed and unexposed subject identified
- both groups followed forward in time and monitor outcomes
- investigates, participant unaware of eventual outcomes which minimizes threat of bias
- impractical for rare or slowly developing conditions
- potential threats to validity: testing effect (pretest acts as intervention) or attrition (not wanting to stay in a study for a long time
8
Q
retrospective longitudinal studies-
A
- EXAMINE PREVIOUSLY COLLECTED DATA
- medical records, databases, surveys
- efficient way to study conditions with long latency
- validity threat; outcome status often known at beginning of study
- need two protect against biased assessment
9
Q
Cross-sectional studies
A
- researches study groups of subjects at one point in time
- insight into current health status of population
- cannot determine if risk factor preceded outcome
- not conclusive evidence for causation
10
Q
cohort studies
A
- also known as follow up studies
- uses a group of people and determines exposure to a particular factor
- prospective: subjects do not have the outcome of interest
- retrospective: some subjects have outcome of interest
11
Q
Case-control studies
A
- one group of subjects with condition
- other group of subjects without the condition(control)
- difference in exposure history and presence of risk factor
12
Q
potential bias in Case-control studies
A
- selection bias
- observation bias
- interviewer bias
- recall bias
13
Q
selection bias
A
- relates to criteria used for selection of participants
- criteria must be unrelated to exposure of interest
- volunteers – self-selection bias
14
Q
observation bias
A
- outcomes known
- susceptible to biased researcher measurement or participant reporting
15
Q
interview bias
A
- change in the way they interview groups
- assessment different for control and cases