Research Principles Flashcards
What types of studies are subject to recall bias? (2)
Cross-sectional and case control
what type of study samples subjects at one point in time for presence or absence of exposure and disease?
Cross sectional
what parametric test compares means between two groups?
student t test
what type of error occurs when the null hypothesis is falsely rejected?
alpha error (type 1 error)
define relative risk (risk ratio)
risk of disease/death among people who are exposed divided by risk of disease/death among people who are not exposed
define prevalence
a measure of disease present at one point in time
Formula for sensitivity
TP / (TP+FN)
X is assumed to cause Y, when in truth Z is common also and causes X and Y. What is Z?
confounding variable
define beneficence
maximizing the benefits for the research project while minimizing the risks to the subjects
when is it appropriate to use risk ratio/relative risk?
with a cohort study. cannot use with case-control bc researcher controls how many cases and controls (i.e. the risk ratio)
when can you use an odds ratio?
with case control study or a cohort study
Prevalence affects which of the following?
- sensitivity
-specificity
- PPV
- NPV
NPV and PPV
Formula for specificity
TN / (TN + FP)
define power
probability of avoiding type II error/ ability of a statistical test to detect a true difference in the sample of data
define incidence
measure of individuals who develop a disease during a specified time period
for acute conditions, which is generally greater: incidence or prevalence?
incidence
name parametric and nonparametric tests that compare means between two groups
Parametric - student t test
Non parametric - Wilcoxon rank-sum test
name parametric and nonparametric tests that compare means between three ore more groups
parametric - ANOVA
non parametric - Kruskal Wallis test
name parametric and nonparametric tests that compare two quantitative measurements from one source
Parametric - paired t test
non parametric - Wilcoxon signed-rank test
name parametric and nonparametric tests that determine degree of association between two quantitative variables
parametric - Pearson coefficient of correlation
non parametric - Spearman rank-order correlation
what is statistical power?
probability of avoiding type II error (beta error) or ability of a statistical test to detect a true difference in the sample data
- type II error = null hypothesis is falsely accepted
what is odds ratio?
magnitude of association between an exposure and an outcome
what is relative risk ratio?
ratio of incidence among exposed vs unexposed
relative risk is used in what kind of study?
cohort studies
- cannot be used in case-control study bc researchers control the risk
odds ratio is used in what kind of study?
case control study
likelihood ratio formula
SN/1-SP
- 1-SP = prob of false positive
formula for NNT (2)
NNT = 1/ARR
- 1/ (%placebo - %tx)
formula for ARR
ARR = EER - CER
(diff in event rate between experimental and control group)
formula for RR
a/(a+b) divided by c/(c+d)
formula for OR (with a,b,c,d)
(a x d) / (b x c)
What outlines the ethical principles and guidelines for human subject research?
The Belmont Report
