140.621.01 STATISTICAL METHODS IN PUBLIC HEALTH I

Introduces the basic concepts and methods of statistics as applied to diverse problems in public health and medicine. Demonstrates methods of exploring, organizing, and presenting data, and introduces fundamentals of probability, including probability distributions and conditional probability, with applications to 2x2 tables. Presents the foundations of statistical inference, including concepts of population, sample parameter, and estimate; and approaches to inferences using the likelihood function, confidence intervals, and hypothesis tests. Introduces and employs the statistical computing package, STATA, to manipulate data and prepare students for remaining course work in this sequence.

Students who successfully master this course will be able to: 1) Use statistical reasoning to formulate public health questions in quantitative terms [1.1 Understand the role of statistical reasoning within the scientific method; 1.2 Understand and apply the counterfactual definition of causal effects in public health research; 1.3 Distinguish and critique the relative merits of continuous, categorical, binary and time-to-event data; 1.4 Understand that evidence for establishing an association between a risk factor and health outcome is generated by comparing the distribution of an outcome in otherwise similar populations that have different levels of the risk factor; 1.5 Use stratification in design and analysis to minimize confounding and identify risk modification]; 2) Design and interpret graphical and tabular displays of statistical information [ 2.1 Create by hand and interpret stem and leaf plots, box plots, Q-Q plots and frequency tables; 2.2 Use the statistical analysis package Stata to make basic statistical computations and graphical displays; 2.3 Characterize the distribution of a variable – using the concepts of typical value, variability, and shape; 2.4 Use a variable transformation, such as the logarithm, to study a right skewed distribution such as hospital costs; 2.5 Analyze the sources of bias and variance in study measurements; 2.6 Graphically compare the distributions of two groups of observations of otherwise similar units (e.g. people or treatments) and interpret the display; 2.7 Explore study results for associations among multiple variables and interpret the findings]; 3) Use probability models to describe trends and random variation in public health data [3.1 Use the statistical analysis package Stata to make basic statistical computations in combination with graphical displays; 3.2 Use the concepts of probability to describe the effect of a treatment on a health outcome in a randomized trial; 3.3 Use the binomial distribution and the Poisson approximation to the binomial to calculate probabilities of events; 3.4 Use the Gaussian or normal probability model to approximate the distribution of a continuous public health measure and assess the quality of this approximation; 3.5 Generate and interpret a quantile-quantile (Q-Q) plot to compare two distributions]; 4) Use statistical methods for inference, including tests and confidence intervals, to draw public health inferences from data [ 4.1 Generate random numbers and appreciate the sources of variation among multiple observations of a random process; 4.2 Explain the implications of the Central Limit Theorem in determining the sampling distribution of the mean of n observations; 4.3 Use bootstrapping to determine confidence intervals and interpret them in a scientific context; 4.4 Use sampling distribution theory for the mean and for differences between two means to create confidence intervals and hypothesis tests; 4.5 Use stratification to eliminate the influence of a possible confounding variable in a study of the association of a risk factor and outcome; 4.6 Construct and interpret the appropriate two-sample confidence interval and t-test to assess whether average outcome is different between two groups; 4.7 Use the paired-sample t-test and confidence intervals to assess if the average change is different from zero; 4.8 Examine the consequence of using an inappropriate unpaired (two-sample) analysis when a paired analysis is appropriate; 4.9 Define and apply the term “effect modification” or equivalently “interaction” in the analysis and interpretation of data from a randomized trial].