Assessing The Methodological Quality Of Research Papers Flashcards
Dunning-Kruger effect
Describes the relationship between confidence and level of competence one has in a field
- confidence is at its highest w/ no experience and drops overtime until you reach expert level
“ slope smiley face shaped”
Internal validity
“How well is a study designed”
Often leads to believing a study is flawless because of some internal reason
- NO research study is perfect
Results of study can be attributed to the following
- truth
- bias
- random chance
External validity
How useful is the research outside of the setting of the research itself
1st step to understanding a paper/study
What kind of research is it? What is its design
- is either primary or secondary
Primary research designs include what?
Randomized controlled trials
- best used for therapy and prevention questions
(is this intervention more effective than another and does this intervention reduce the risk of the disease?)
Lab/bench research
Blind studies
- best for diagnostician test questions (how accurate is this test?)
Case-control
Cross sectional
Cohort
- best used in prognosis and etiology questions (what’s the causes and what is the likely course over time for the patient?)
Qualitative
- best for quality of life questions
( how do patients live with this disease )
Secondary research designs include what?
Systematic review and meta analyses
Non-systematic reviews
Guidelines
taking primary research and comparing it to others
Meta-analyses and systematic reviews
SR:
A summary of research that addresses a clinical question
MA:
A statistical pooling or aggregation of results from different studies and determining the results/conclusion
- essentially SR w/ math
- are both secondary studies *
Clinical practice guidelines
Statements that include recommendations via systematic review and meta analyses results
- created by groups of experts for health care workers
Groups induce
- professional societies
- government groups
Randomized controlled trials
A type of study in which participants are randomized into intervention groups
considered the gold standard type of study for clinical research and are designed to reduce bias
Observational studies
Include cohort studies and can be used to find interesting/important associations
However, very susceptible to bias since correlations made via observational studies are often assumed to be causation
- It can NEVER be
Types of observational studies
Case control study
- look at the past and make correlations
Cross-sectional study
- look in the present and determine correlations
Cohort
- future based study shower you take a population in the present and just follow them dont intervening.
Case series and case reports
Descriptions of a series of patients and correlations associated with them
- generally weak studies and are used primarily in rare/new presentations only.
Ex: corona virus
Introduction methods results and discussion sections of a paper
Introduction: why the authors decided to do the research
Methods: how
Results: what they found
Discussion: what does it mean
Why is table 1 in most research papers the most important in randomized trials
It compares the groups with criteria that the researchers believe is important
- a lot of the information one wants to see is in this table
Questions to ask when looking at a randomized trial
1) was the assignment of patients to treatments randomized?
2) were groups similar at the start of the trial
3) aside from allocated treatment, were groups treated equally
4) were all patients who entered the trial accounted for?
5) were the patients any lived in the groups which they were originally randomized into
6) were measures objective or were the patients and doctors kept blind throughout the trial
