research - FINAL EXAM Flashcards
Reality is…
QUANTITATIVE: quantifiable, objective
QUALITATIVE: based on observer, subjective
Cause & effect is…
QUANTITATIVE: linear
QUALITATIVE: based on many factors, unique, based on situation
Research is…
QUANTITATIVE: reproducible, base on manipulation of variables
QUALITATIVE: individual, based on what happens naturally
The researcher is…
QUANTITATIVE: separate from study, as objective as possible
QUALITATIVE: intertwined in study, as objective as possible
Participant selection is…
QUANTITATIVE: random
QUALITATIVE: usually based on knowledge/ experience
Study design…
QUANTITATIVE: can be observational/ experimental, generalizable to population, involves quantifiable outcomes, decided beforehand
QUALITATIVE: use open ended interviews, open ended surveys, direct observations, involves descriptions, may change as study proceeds
data…
QUANTITATIVE: presented in numbers
QUALITATIVE: presented in words
QUALITATIVE
-no difference between observer & observation
-does not assume objective external reality
-takes all observations as relevant & valid
-does not assume reproducibility, all outcomes situationally dependent
-assumes cause & effect based on multiple variables
-open-ended interviews & surveys, or direct observations
*data expressed in words, not numbers!
QUANTITATIVE
-quantifiable reality
-linear cause & effect
-summarize & manage large amounts of data
-reproducibility
-can be experimental / observational
*data can be expressed as a number – can be quantified!
ANATOMY of an ARTICLE
-Abstract – summary
-Introduction – most references here
-Methods – how was study carried out?
-Results – what happened? qualitative/ quantitative, objective data only
-Discussion/Conclusion – authors interpretation of data
-References – how does this relate to past research?
LEVELS of EVIDENCE
- meta-analysis
- systemic reviews
- randomized controlled trials
- cohorts studies
- case control studies
- case reports, case series
- animal/ lab studies
RANDOMIZED CONTROLLED TRIALS
-establish cause & effect relationship
-reduces risk of bias
-weighted more heavily than non-randomized
CORRELATION
AKA population survey
-uses existing data
-looks for statistical association between exposure & outcome
*correlation DOES NOT MEAN causation!
LITERATURE REVIEW
*summarize existing evidence on a topic
-question is posed
-guidelines set; which types of research will be included
-research analyzed to see how strong it is
-important; if result is repeated over & over again, more likely to be true!
= ‘reproducibility’
literature review TYPES
Narrative Review
Meta Analysis
Systematic Review