Physics and Math Flashcards
Scientific Method
- generate a testable question
- Gather data and resources
- Form hypothesis
- Collect new data (exp. or obs.)
- Analyze data
- Interpret data/existing hypothesis
- Publish (and peer review)
- Verify results
FINER method; whats it for?
for evaluating a research question to determine whether it will add to the body of scientific knowledge
- Feasible?
- Interesting?
- Novel?
- Ethical?
- Relevant?
positive vs. negative control in an experiment
positive ensures a change in dependent variable when expected, negative is to ensure no change in dependent variable when no change is expected
what is accuracy/validity? how does that compare to precision/reliability?
accuracy/validity is ability of an instrument to measure true value, precision/reliability is ability of instrument to read consistently
give example of binary/continuous/categorical variables
- binary - yes vs no; better vs. worse
- continuous - amount of weight lost; % cardiac improvement
- categorical variables - state of residence; socioeconomic status
cohort study
subjects are sorted into groups based on different risk factors (exposures), then assessed at various intervals to determine how many subjects in each group had a certain outcome
cross-sectional study
attempts to categorize patients into different groups at a single point in time
- ex. study to determine prevalence of lung cancer in smokers/nonsmokers at given point in time
case-control study
identifies number of subjects w/ or w/o a particular outcome, then looks backwards to assess how many subjects in each group had exposure to a particular risk factor
- ex. study where patients w/ and w/o lung cancer are assessed for smoking history
Hill’s criteria
- Temporality - exposure (I.V.) must occur before outcome (D.V.)
- Strength - as more variability in the outcome variable is explained by variability in study variable, relationship is more likely to be cause
- Dose-response relationship - as study or I.V. increases, there is proportional increase in response. More consistent relationship = more likely to be causal
- Consistency - relationship is found to be similar in multiple settings
- Plausibility - reasonable mechanism for I.V. to impact D.V. supported by existing literature
- Consideration of alternative explanations - if all other plausible explanations have been eliminated, remaining explanation is more likely
- Experiment - if an experiment can be performed, causal relationship can be determined conclusively
- Specificity - change in outcome variable is only produced by associated change in I.V.
- Coherence - new data and hypothesis are consistent w/ current state of scientific knowledge
bias vs. confounding
Bias is result of flaws in data collection phase of experimental/observational study, confounding is error during analysis
detection bias
when educated professionals use their knowledge in an inconsistent way
hawthorne effect (observational bias)
behavior of study participants is altered simply because they recognize that they are being studied
beneficence vs nonmaleficence
beneficence is the obligation to act in patient’s best interest, nonmaleficence is obligation to avoid treatments in which potential for harm outweighs potential for benefit
autonomy vs justice
autonomy is responsibility to respect patients’ decisions/choices about their own healthcare, justice is responsibility to treat similar patients w/ similar care and to distribute resources fairly
morally relevant differences
differences b/w individuals that are considered an appropriate reason to treat them differently
- ex. age (a transplant would be given to child vs old person)
equipoise
in a study comparing two treatments, you cannot approach research that one treatment is superior
internal vs external validity (generalizability)
internal related to how well a study is conducted (structure), external is how applicable the findings are to the real world
statistical vs clinical significance
statistical significance can be significant mathematically (like drop in BP by 1mmHg), but that would not be a notable change to a patient’s health
measures of central tendency
describe the middle of a sample
how to determine if outlier by IQR
- Q1=n/4, then average that&next highest
- Q3=3n/4, then average that and next highest
- IQR = Q3-Q1
- LF=Q1-1.5IQR and UF=Q3+1.5IQR
standard deviation formula
sqroot((x-u)^2/n-1)
independent vs dependent events
independent events have no effect on each other (e.g. rolling dice), dependent events affect each other (e.g. picking balls out of a hat)
Probability equation of 2 or more events occurring at the same time:
P(A and B) = P(A) + P(B)
Probability equation of at least one of two events occurring:
P(A U B) = P(A) + P(B) - P(A and B)
null hypothesis
always a hypothesis of equivalence
- says 2 populations are equal or that a single population can be described as equal to a given value
alternate hypothesis 2 options are?
can be directional (populations are not equal) or directional (mean of population A > mean of B)
decisions based on p-value
- p>a (0.05) - fail to reject null hyp. and not statistically significant
- p<a></a>
Type I error
likelihood we report a different b/w two populations when one doesn’t actually exist
Type II error; what is Beta
when we incorrectly fail to reject null hyp.
- Beta = props. of TypeII error
confidence vs. power (statistics)
power is prob. of correctly rejecting a false null hypothesis, 1-Beta
- confidence is prob. of correctly failing to reject a true null hypothesis
what are pie and circle charts for?
used to represent relative amount of entities and are especially popular in demographics
