Course Description
This course provides students with a step-by-step guide to time series analysis using the SAS system. We will walk through all of the statistical methodology required to perform a proper time series analysis and will provide students with sample syntax to help them along the way.
What am I going to get from this course?
- State the basic reasons for
why data are analyzed in the healthcare setting
- Demonstrate an understanding of the nature of health-related data
- Properly clean and organize data for statistical analysis
- Arrange a data set in time-series format
- Construct a statistical model to analyze time-series data
- Report results in an appropriate format
Prerequisites and Target Audience
What will students need to know or do before starting this course?
Students should have a basic understanding of the principles of
data analysis and should be familiar (not expert) in SAS.
Who should take this course? Who should not?
Health researchers who are familiar with SAS and who want to learn how to perform time-series analysis or who need a refresher.
Curriculum
Module 1: Introduction
11:03
Lecture 1
Course Overview and Objectives
11:03
This "lecture" introduces the instructor and provides an outline of the course. It sets expectations for previously acquired knowledge and expertise and outlines what the learner will achieve throughout the course.
Module 2: Basics of Healthcare Data
13:25
Lecture 2
Basics of Healthcare Data
13:25
This module gives an overview of the basic principles of healthcare data. By the end of this section, the learner should be able to:
- Identify the origins of healthcare data;
- Identify the basic elements of healthcare data;
- Understand the reasons why healthcare data are analyzed;
- Visualize the structure of a finished data product.
Module 3: Count Data
12:46
Lecture 3
Count Data
12:46
This lecture describes count data and its usefulness (and necessity) in a time-series analysis.
By the end of this section, the learner should be able to:
- List the basic components of a count data set;
- Understand how to organize count data;
- Describe the interplay between count data and time measurements in a time series analysis.
Module 4: Person-time and Rates
23:22
Lecture 4
Person-time and rates
23:22
By the end of this lecture, the learner will be able to:
- Define the concept of person-time
- Enumerate person-time for individuals and groups;
- Differentiate between rates and risk calculations;
- Calculate crude rates;
Describe the utility of rates in a time series analysis.
Module 5: The Poisson Distribution
18:37
Lecture 5
The Poisson Distribution
18:37
This lecture explores the principles of the Poisson distribution and explains its use in time series analysis.
By the end of this lecture, the learner should be able to:
- Describe the nature of the Poisson distribution
- List the assumptions involved in Poisson modeling
- Describe which types of outcomes follow the Poisson model
- Describe how the Poisson distribution relates to count data
- Calculate/estimate probability in Poisson distributed data.
Module 6: Organizing Time Series Data
39:17
Lecture 6
Getting Started with SAS
15:51
Lecture 7
Organizing Time Series Data
23:26
This lecture teaches the learner coding strategies for organizing and checking time series data.
By the end of this lecture, the learner should be able to:
- Write code to organize time series data
- Create appropriate variables for time series analysis
- Perform An exploratory analysis and explain the importance of doing so using
- Use the frequencies, univariate, means, and correlation procedures (PROCS).
Module 7: Time Series Modeling
02:02:58
Lecture 8
Time Series Modeling Part 1
22:54
This is the first part of a two-part lecture in time series modeling.
By the end of this lecture, the learner should be able to:
- Describe the basic principles behind time series modeling;
- Successfully produce code that properly edits your analysis data preparatory to performing a time series analysis;
- Produce code to identify the proper link, distribution, and offset for time series modeling using PROC GENMOD.
Lecture 9
Time Series Modeling: Part 2
24:11
This is lecture 2 of 2 in time series modeling.
By the end of this lecture, the learner should be able to:
- Add a predictor variable to the PROC GENMOD syntax and interpret its use;
- Estimate unadjusted rates using PROC GENMOD;
- Use PROC TIMESERIES to graph and estimate trends in your data;
- Understand the basics of PROC ARIMA.
Lecture 10
Guided Exercise 1
15:02
Lecture 11
Adjusting the Model
29:44
This lecture teaches the statistical principles behind adjusting a statistical model and provides examples of syntax that is to be used to adjust models,
By the end of the lecture, the learner should be able to:
- Understand why statistical models may need adjusting;
- Define the concept of "statistical influence";
- Explain the "change-in-estimate" criteria;
- Explain the concept of effect modification and how to code it.
Lecture 12
Reporting Results
21:50
This lecture explores the tenants of reporting results in an organized and professional manner.
By the end of this lecture, the learner will be able to:
- List the sections involved in a manuscript/white paper;
- Properly formulate a Methods section;
- Understand the order in which Results should be reported;
- Properly formulate tables for manuscripts;
- Organize an abstract properly.
Lecture 13
Guided Exercise 2
09:17
