PHP2515: Fundamentals of Probability and Statistical Inference

Fall 2023

Instructor:
Arman Oganisian, PhD
Assistant Professor of Biostatistics
Email: aoganisi@[Link]

Teaching Assistant:
Zihan Zhou
Email: zihan zhou1@[Link]

Lectures: Mondays and Wednesdays, 9am-10:20am.

Location: Room 408, 121 South Main Street, Providence RI.
Office Hours: Tuesdays 4-5pm (ET) via Zoom:
Zoom Meeting ID: 941 4563 9183
[Link]

Overview

This course will provide an introduction to probability theory, mathematical statistics, and
their application to biostatistics. The emphasis of the course will be on mathematical and
probabilistic concepts that form the foundation for statistical inference. After developing
probability theory from the ground up, we will turn to inference: using data to learn about
the unknown probability models that govern random biomedical phenomena. The course
will cover fundamental ideas of probability, some foundational probability models (normal,
binomial, exponential and Poisson), sample and population moments, finite and approximate
sampling distributions, point and interval estimation, and hypothesis testing. Examples of
their use in modeling will also be discussed. R programming will be used throughout.

Prerequisites This course is designed for first-year masters students in biostatistics. Treat-
ment of probability and statistics will be calculus-based with the goal of providing a rigorous
theoretical foundation for upper-level graduate courses.
– Mathematics up to and including multivariate calculus (e.g. convergence of infinite se-
ries and sequences; limits; partial derivatives; double integrals; optimization). Usually
this is obtained through three semesters of undergraduate calculus.

– Working knowledge of R will be useful. Exams may have programming components.

The final project will involve R programming.
Course Materials
Texts:
- (required - main text) Casella, G. and Berger, R.L. (2002). Statistical inference. Vol.
2. Duxbury. Estimated cost: ≈ 80 US$.

- (optional, less technical version of Casella) Rice, J. A. (2006). Mathematical Statistics

and Data Analysis. Duxbury Press. Estimated cost: ≈ 150 US$.
Software:
- R programming language: [Link] Estimated cost: free.

- RStudio IDE: [Link] Estimated

cost: free.

- (optional) Use of LATEX editor for homework assignments: [Link]

org/. Estimated cost: free.

Course Expectations
Students in this course will be expected to do the following:
– Keep up with lectures.

– Read all assigned material prior to coming to class and actively participate in class
discussions.

– Review lectures/notes regularly - recommend doing this every weekend while transcrib-
ing notes.

– Complete and turn in all assignments on time. Solutions to homework must be clearly
written with all steps detailed. Students are encouraged to use LATEX for writing up
homeworks, but it is optional.

– Demonstrate an understanding of course material on examinations.

– Participation in group projects.

– Respect each other, each others questions, and each others discussion.

Attendance, Lateness, Missed/Late Assignments

Students are expected to keep up with lectures and submit all assignments by the due date.

Expected Time Commitment

Over the course of the semester students will spend at least the amounts of time shown below
on specific tasks related to this course:
 Task Hours Spent on Task
Class Time 40
Reading 35
Homework 100
Exams/Prep 25
Total 200

The total amount of time expected is 200 hours. Of course, these amounts are averages and
will vary by student.

Grading

– Homework - 25%

– Midterm Exam - 25%

– Final Exam - 35%

– Final Project - 10%

– Participation - 5%

Topics Covered

– Part I - Probability Theory and Modeling:

– Set theory, probability measures, counting techniques.

– Conditional probability
– Random variables, probability mass/density functions, distribution functions.
– Expectation, variance, moments, and moment-generating Functions.
– Common discrete and continuous distributions.
– Location-scale families and exponential families.
– Multiple random variables: joint, marginal and conditional distributions.
– Consequences of independence. Mutual and pairwise independence.
– Covariance, correlation, and hierarchical/mixture models.
– Multivariate distributions with close look at bivariate normal and multinomial
distributions.

– Part II - Statistical Inference:

– Statistics and sampling distributions.

– The Nonparametric Bootstrap.
– Convergence concepts: almost-sure convergence, convergence in probability, con-
vergence in distribution.
– Large sample theorems: central limit theorem, weak and strong laws of large
numbers, delta-method, Slutsky’s theorem, continuous mapping.
– Methods of point estimation: method of moments, maximum likelihood (score
functions, fisher information, CR-lower bound, invariance property, and asymp-
totic normality/consistency),
– Bayesian inference (prior choices and posterior updates in conjugate families).
– Methods of evaluating point estimators: bias, variance, and bias-variance trade-
off.
– Hypothesis testing: null/alternative hypothesis, critical values, rejection regions,
test statistics, p-values.
– Methods of testing: Likelihood ratio tests, asymptotic tests, score tests.
– Methods of evaluating tests: type I and II error rates, and power. Sample size
calculations.
– Multiple testing and controlling family-wise error rates.
– Methods of Interval estimation: test inversion and Bootstrap intervals.
– Methods of evaluating intervals: size and coverage.
Brown ScM/AM Biostatistics Core Competencies and Assessment

1. Use probabilistic and statistical concepts and methods to describe and draw inferences
from biomedical data

Assessment: Each homework assignment will include questions on distributions and

assumptions on binary, count, and continuous variables. Most exercises will report
biomedical examples and biological data Example exam question: Calculate E(X) and
Var(X) of the variable X, the time to recovery (in days) from the H1N1 flu. Also, each
homework and exam will perform statistical techniques in order to better understand
the distribution of the data and draw conclusion on that. Example exam question:
Marginal densities of the variables in a joint distribution. Plot the histograms of the
marginal densities of X and Y. Are X and Y independent?

2. Prepare report of methods, results, and interpretations from a simulation that investi-
gates the properties of a statistical method

Assessment: At the end of the course the students prepare a report addressing differ-
ent simulation scenarios. In this way, they explore statistical inference and perform
hypothesis testing for the different simulation settings. Example report question: Find
the maximum likelihood estimator of the given function after a simulation of 1,000
sample from a normal distribution. Describe the findings and report what you discov-
ered. Also, the report will include plot, statistical measure, and statistical checking in
order to define all the distributional properties of the simulation scenario considered.
Example report question: Generate a sample of size n = 500 from the distribution
defined. Perform the bootstrap and compute the correlation coefficient for B = 10000
Bootstrap samples. Plot a histogram of the Bootstrap. Discuss your findings.

3. Perform power analysis and sample size calculations to determine the required number
of subjects to carry out scientific studies

Assessment: The homework assignments and exam also focus on test inference to
compute power analysis and the optimum sample size to determine scientific conclusion.
Example exam question: What is the power function of a determined test? Compute
the sample size required for testing this hypothesis with power. Explain your reasoning
and derivation of each step.

Additional Learning Outcomes

1. Apply basic probability concepts such as conditional probabilities, independence, Bayes’

Rule, and combinations and permutations to calculate probabilities of events of prac-
tical interest.
 2. Describe and apply concepts of discrete and continuous random variables, density
functions, joint, marginal, and conditional distributions, and sampling distributions of
statistics such as the mean

3. Apply basic principles of estimation, including maximum likelihood estimation and

method of moments, to calculate estimated statistics for practical health-related prob-
lems

4. Apply hypothesis testing approaches based on maximum likelihood theory to develop

test statistics, including Wald, Likelihood Ratio, and Score tests, and calculate these
test statistics for health-related data

Academic Accommodations

Students seeking accommodations due to a disability or medical condition, should contact

Student and Employee Accessibility Services by phone 401-863-9588. For more information
see [Link]

Academic Code

It is expected that students taking this course abide by the academic code which can be found
at: [Link]

Class Recording and Distribution of Course Materials:

Annotated slides and notes will be provided after each lecture on Canvas. There will be
no recordings of class content due to technical overhead and also to ensure all students feel
comfortable speaking up and partaking in class with some level of privacy.

Diversity and Inclusion

This course is designed to support an inclusive learning environment where diverse perspec-
tives are recognized, respected and seen as a source of strength. It is our intent to provide
materials and activities that are respectful of various levels of diversity: mathematical back-
ground, previous computing skills, gender, sexuality, disability, age, socioeconomic status,
ethnicity, race, and culture.

Technology

This course will be using Canvas as the course management platform. I am committed to
ensuring access to online course resources by students. If you have any concerns or questions
about access please reach out. The IT Service Center ([Link]
provides many IT Services including remote assistance, phones, tickets, and chat. Please
also see the Online and Hybrid Learning Student Guide.