Biostats

Pearson Correlation

Evaluates the linear relationship between two continuous interval variables without establishing causation

Chi-Square Test (X^2)

Evaluates whether two categorical nominal variables are significantly related based on frequency counts

Independent Two-Sample t-Test

Compares the mean of one continuous interval variable across two separate, independent nominal groups;

Example: Comparing systolic blood pressure between men and women

Paired t-Test (Matched Pairs t-Test)

Compares the mean of one continuous interval variable across two linked or non-independent groups

Example: Measuring blood pressure before and after taking the USMLE Step 1 exam, or comparing blood pressure differences between identical twins given Drug X versus placebo

One-Way ANOVA

Compares the means of one continuous interval variable across three or more independent nominal groups;

Example: Evaluating pain score reduction among three separate groups receiving Drug A, Drug B, or placebo

Two-Way ANOVA

Compares the means of one continuous interval variable across two independent nominal variables simultaneously;

Example: Evaluating blood pressure differences based on both biological sex (men vs. women) and nationality (US vs. Canada)

Repeated Measures ANOVA

Compares the means of one continuous interval variable across three or more linked or repeated measurements over time within the same subjects

Example: Tracking changes in pain improvement scores measured across multiple time points in the same individuals taking three different drugs

Case Report

An observational, descriptive report detailing the clinical presentation, diagnosis, and outcome of an individual patient

Example: Documenting a single patient presenting with an unusual adverse drug reaction or rare disease manifestation

Case Series

An observational, descriptive study tracking clinical outcomes in a small group of individuals with the same rare condition without a control group

Example: Gottlieb's 1981 report of Pneumocystis carinii pneumonia and mucosal candidiasis in previously healthy homosexual men that helped first identify AIDS

Ecological Study (Correlational Study)

An observational study analyzing the relationship between exposure and outcome using aggregate population-level data

Example: Comparing citywide HIV prevalence rates between communities with needle exchange programs and communities without needle exchange programs

Cross-Sectional Study (Prevalence Survey)

An observational snapshot study measuring exposure and outcome status simultaneously at a single point in time to determine prevalence

Example: Surveying crane operators and desk workers during annual physicals on a single day to compare the current prevalence of lower back pain

Case-Control Study

A retrospective observational study identifying individuals with disease (cases) and without disease (controls) to compare past exposures, reported via odds ratios;

Example: Comparing first-trimester maternal SSRI antidepressant use between infants born with cardiovascular birth defects and infants born without birth defects

Cohort Study (Longitudinal / Prospective Study)

An observational watch-and-wait study tracking exposed and unexposed disease-free individuals forward over time to calculate disease incidence and relative risk;

Example: The Framingham Heart Study following healthy residents over decades to observe who develops cardiovascular disease

Retrospective Cohort Study

An observational study using historical records to reconstruct past exposure groups and following their outcomes forward through time up to the present

Example: Reviewing military service employment records from 1990 to track the development of respiratory illness over the subsequent 20 years

Randomized Controlled Trial (RCT)

An experimental study where investigators randomly allocate participants to intervention or control groups to establish cause-and-effect

Example: Randomly assigning hypertensive patients to either WallerDope or a matching placebo to test blood pressure lowering efficacy

Crossover Study

An experimental trial where each subject receives an experimental treatment, completes a washout period, and then crosses over to receive the alternative control; Example: Administering a short-acting migraine drug to a patient for 2 weeks, completing a washout period, and then administering a placebo for 2 weeks so each patient serves as their own control

Systematic Review

A rigorous synthesis of multiple primary studies selected through standardized inclusion criteria to answer an answerable clinical question;

Example: A Cochrane review screening all published and unpublished randomized trials evaluating zinc supplementation for the common cold

Meta-Analysis

A quantitative synthesis pooling numerical data from multiple independent studies to calculate a single summary effect estimate at the top of the evidence pyramid; Example: Combining statistical results from 10 distinct statin trials into a forest plot to estimate the overall relative risk reduction in myocardial infarction

Screening Test

A test with high sensitivity designed to capture all possible cases and rule out disease in asymptomatic populations; Example: Blood bank HIV enzyme immunoassay (EIA) or fecal occult blood testing

Confirmatory Test

A test with high specificity used after a positive screen to eliminate false positives and rule in the disease;

Example: HIV Nucleic Acid Amplification Testing (NAAT) or Western blot performed after a reactive screening immunoassay.

Pearson Correlation Example

Example: Measuring the association between height and weight, or daily caloric intake and fasting blood glucose

Chi-squared example

Example: Comparing marital categories (single, married, partnered, divorced) between two groups, or assessing smoking status (yes/no) versus low birth weight (yes/no)

Quasi-Experimental Method

A study comparing groups exposed to an intervention or condition lacking true random assignment because the independent variable is an inherent attribute;

Example: Comparing standardized math test scores between randomly selected men and women where biological sex cannot be manipulated

Receiver Operating Characteristic (ROC) Curve

A graphical plot of Sensitivity versus 1 - Specificity across all potential test cutoffs, where Area Under the Curve (AUC) reflects overall diagnostic accuracy;

Example: Plotting test cutoffs for prostate-specific antigen (PSA) or cardiac troponin to select the threshold that balances true positives against false positives

Correlational Method

An observational design measuring two or more variables to identify associations without experimental manipulation or causal claims

Example: Observing cat ownership and recording the number of recorded falls in 500 older nursing home residents over one calendar year