Study Design
Calculations
Bias
Variables
Potpourri
100
This type of study is experimental and subjects undergo an intervention
Clinical trail
100
Calculate sensitivity and specificity
Sensitivity=A/(A+C)
Specificity=D/(D+B)
100
You are testing drugs for insomnia and more OB/gyn residents want to sign up than the general population
Selection bias
100
The _________ variable is sometimes called the “response,” the “symptoms”, or the “outcome”. These variables are not affected by other variables the study measures. In contrast, the _______ variable is often the focus of the research study.
independent and dependent variables
100
This type of variable influences both the dependent variable and independent variables, causing a spurious association
Confounder or confounding variable
200
A ___________ study looks back at past data to analyze an outcome, while a ___________ study follows a group of individuals forward in time to observe how an exposure might influence a future outcome
Retrospective, prospective
200
Describe the calculation of relative risk (RR)
RR = [A/(A+B)] / [C/(C+D)]
200
A patient reports that they got chicken pox shortly after receiving the HPV vaccine
Recall bias aka reporting bias
200
Mild, moderate, and severe pain
Categorical ordinal variables (have a specific order)
200
This type of statistical error occurs when you state there is a difference observed, when in reality there is not
Type I error (alpha) aka p-value
300
This type of observational study choses subjects based on the presence or absence of risk factors to see if they develop a disease/condition. Yields incidence and relative risk
Observational cohort study
300
Calculate odds ratio (OR)
OR = (A/B)/(C/D)=(A*D)/(B*C)
300
A surgeon who typically doesn't wash their hands, does so before surgery after the hospital started a SSI prospective observational study
Observation bias aka Hawthorne effect
300
Ca-125 values of 17.36, 364.4, and 3,489.3
Numerical continuous variables (cannot be counted)
300
This type of error occurs when you report there is no difference, when in actuality there is a difference between the groups
Type II error (beta)
Power=1-beta
400
In this type of observational study, subjects are chosen based on presence or absence of disease/condition. Yields odds ratios
Case control
400
Calculate attributable risk (AR)
AR=[A/(A+B)] - [C/(C+D)]
400
A participant doesn't report their full sexual history on a research survey
Response bias
400
Dysgerminoma, high-grade serous, and sertoli-leydig ovarian cancers
Categorical nominal variables - have no logical order
400
This measures of the ability of an experimental design to detect a particular effect based on a percent difference observed
Power
500
This type of randomized clinical trial compares a new treatment to an existing standard treatment to determine if the new treatment is not significantly less effective
Non-inferiority trial
500
This type of risk is calculated the same way as attributable risk (AR), but the term refers to the probability of a disease increasing due to an exposure or a risk factor
Absolute risk reduction (ARR)
AR=ARR=[A/(A+B)] - [C/(C+D)]
500
Doing these 4 things in your study can reduce bias
1. Blinding
2. A control group
3. Crossover (everyone is their own control)
4. Randomization
500
Para 1, para 2, para 3, para 4, para 5, para 6
Numeric discrete variables (can be counted)
500
A child is born to parents who are both carriers of achondroplasia. What are the odds (in a fraction or %) that the offspring is a carrier?
2/3 or 67%