Introduction to the module
Module Descriptions
Site map
Welcome
This is the home page of the web site for the Statistics Module for the M.Sc. course 20062008, organized by the Dept. O&G, CUHK.
The contents were initially developed for the statistics module of the MSc. Midwifery 20142016, improved and expanded in response to feedback from that module. The exercises and assessments are further expanded with concepts and terminologies that are more familiar to students from Reproductive Medicine.
The module begins in May, 2017, based on selflearning by the students between May and November, and a multiple session workshop in November 2017. This web site contains the teaching material and exercises required in the module.
The remainder of this page are module descriptions and the site map.
The module description orientate students to what the module hopes to achieve and provides a conceptual outline to the contents.
The site map echoes the web pages listed in the menu bar, and provides a brief description of what each page on the web site provides.
Objective
The objective of the module is to enable students to use simple statistical procedures in clinical research and quality evaluation.
Statistics is a subbranch of mathematics. A thorough understanding of statistics requires comprehension and skills in calculus, and is beyond what can be achieved in a small module as designed in the MSc course from the department. The objective of the module is therefore not so much learning statistics than how to use statistics in the student's future professional duties.
The module will cover statistical procedures frequently used in clinical research and quality evaluation. Theories are covered only briefly and superficially, sufficient for the students to understand the concepts underlying the procedures. Most of the module will cover statistical procedures, when they are used, reasons for their use, how they are use, how the results should be interpreted, and what are the pitfalls to avoid.
Contents
Theories will be covered in one web page, Contents_1. Probability. and one workshop.
The other web pages and workshops aim to enable students to carry out the statistical procedures associated with the following research processes.
 In Survey research such as epidemiological studies and quality evaluations. Useful statistical procedures are single group methods such as population mean, proportion, and standard Deviation, and two group methods such are difference between means and proportions.
 In clinical trials, using two group methods such are difference between means and proportions.
 In seeking relationships between clinical parameters, using correlation and regression methods
 In evaluating predictions, tests, and diagnosis, using True and False Positives and Negatives, Bayesian conversion of pretest to posttest probabilities, and the Receiver Operator Characteristics
 In seeking Evidence to support clinical practice, using metaanalysis
The module also introduces students to a detailed approach to produce two dimensional graphics frequently used to present data, including the pie chart, the bar chart, the Forest plot, the scatter plot, the data in groups plot, and the Receiver Operator Characteristics plot
Method of teaching and study
The Module begins in May 2017, when this web site goes live, and ends in April 2018, at the end of the MSc. course. There are 4 modes of teaching and studies
 This web site, which contains all the course material in the module
 Web based learning by individual students, consisting of
 Contents of this web site. The explanations, the computer programs, and the exercises, designed for selfdirected learning
 Direct communication with Prof. Chang by email asking for further explanations
or additional information
 The Q and A page listing questions other students have asked and accompanied by the answers.
 Suggestions sent to students at intervals of topics to study and exercises to be carried out to improve their
understanding and ability
 Assessment. All information and links related to assessment are provided in the
Guidelines to Assessment page
 Workshops. Professor Chang will be in Hong Kong to conduct the workshops in November, 2017. The workshops are planned to be
 2Nov (Thu): 18302130
 4Nov (Sat): full day
 6Nov (Mon): 18302130
 9Nov (Thu): 18302130
 11Nov (Sat): full day
 13Nov (Mon): 18302130
The workshops will cover the following subjects, in order
 Introduction to the web site
 Introduction to StatPgm set of statistical programs
 Introduction to producing graphics using MacroPlot
 Introduction to probability
 Introduction and exercises in Survey, Quality Assurance, fact finding. Statistics using one group
 Introduction and exercises in using Correlation and Regression
 Introduction and exercises in using statistics comparing two groups.basic metaanalysis procedures
 Introduction and exercises in basic metaanalysis procedures.
 Introduction and exercises in evaluating tests and diagnosis, with basic Bayesian Probability procedures.
 Discussions on assessment
Please note that some of the subjects are easy and can be covered quickly. Other subjects are more complex, with students asking many questions, and the exercisesd may take some time to complete.
Consequently, all the sessions are merely time slots, each session does not cover one subject. Rather the subjects flow from one session to another depending on the time required.
Going by the experience of the 2015 module, the first 4 introductory workshops are basic knowledge and skills, and likely to take 3 to 4 sessions. The statistics of one group, two group, and metaanalysis are relatively straight forward and may take perhaps two sessions or less between them. Tests, diagnosis, and Bayesian are usually new to nonstatisticians, so may take one and perhaps one and half session. By the time of the workshop in 2015, a third of students have already completed their assessments, and the rest were in progress without difficulties, so not too much discussions on assessment was required.
Introductory Pages
 Course Preparation. All students should read this at the commencement of the module in May,
2017. It lists the resources the students should acquire and preparation students should undertake before studying the
statistics module.
 Introduction to StatPgm. This describes the computer programs provided to help the students
to carry out the calculations involved
 Graphic Editor and MacroPlot : Explanation and Help. This provides an explanation of the
MacroPlot system of graphics production, and templates of MacroPlot codes
 Exercises Explained. This provides the introduction and explanations of the
exercises, the basis of developing skills in data analysis
Help to Students
 Email to Professor Chang. This activates the email program on the
student's computer so that questions can be put to Professor Chang to explain or enlarge on any part of the module
 Q and A . This provides a posting of interesting questions asked by students (with identity of the
student removed), and answers provided. These includes Q&A from previous courses, and new ones as the module progresses
 the Notice Board. Students should visit this page regularly, as any news or changes to the module
are posted on this notice board
 Dictionary of Medical and Midwifery Terms. Containing brief explanations of medical and
midwifery terms used in this web site, to help students from Reproductive Medicine that may not be familiar with these terms
Course Contents
 Contents_1. Probability. This page contains the theoretical basis of the statistics
module, with the following contents
 Concepts of probability
 Normal, t, and log normal distribution.
 Types of data, and transformation to normal distribution
 Mean, Standard Deviation, Standard Error, 95% confidence interval,
 Null hypothesis, Type I and II Error, and statistical significance
 Sample size
 Contents_2a. Survey for Means & Proportions. This page explains the statistics
of survey research, with the following contents
 Estimation of population means, Standard Deviation, and Proportions
 Sample size estimation during planning, and analysis of precision of the results
 Contents_2b. Correlation and Regression. This page provides an extension of
survey research, using correlation and regression analysis
 Contents_3. Comparing Two Groups. This page provides explanations of statistics
for comparing two groups of data, commonly used in clinical trials and epidemiological studies, with the following contents
 Comparing two parametric means, the t test, the 95% confidence interval of the difference
 Comparing two nonparametric measurements, the Mann Whitney U Test
 Comparing two proportions, chi square, Fisher's Exact probability, Risk Difference, Number Needed to Treat, Risk Ratio,
Odds Ratio
 Comparing two regression lines, introduction to covariance analysis
 Contents_4. Metaanalysis. This page covers the basic concepts of metaanalysis, with
the following contents
 Effect Size, Standard Error, Variance
 Heterogeneity
 Publication Bias
 Combining results of multiple studies to produce summary Effect Sizes
 Contents_5. Predictions and Tests. This page covers the statistics of evaluating the
quality of tests, diagnosis, or predictions, and covers the following topics
 True and False Positives and Negatives
 True and False Positive and Negative Rates, Sensitivity, Specificity, Youden's Index
 Likelihood Ratio for Test Positives and Test Negatives, and their use in Bayesian Calculations of Pretest and Posttest
Probabilities
 Receiver Operator Characteristics and its clinical use
Computer Programs
 Introduction to StatPgm. This describes the computer programs provided to help the students
to carry out the calculations involved
 Graphic Editor and MacroPlot : Explanation and Help. This provides an explanation of the
MacroPlot system of graphics production, and templates of MacroPlot codes
 StatPgm 1. Probability of z and t, including
 Calculate Probability from z and t
 Calculate z and t from Probability
 Calculate percentiles and confidence intervals from sample size, mean, and Standard Deviations
 StatPgm 2a. One Group : Survey for Means and Proportions, including
 Calculations for sample size and precision for population survey of Standard Deviation
 Calculations for sample size and precision for population survey of mean
 Calculations for sample size and precision for population survey of proportions
 StatPgm 2b. One Group : Correlation and Regression, including
 Calculating sample size requirements for correlation analysis
 Calculation for Pearson's Correlation Coefficient, Fisher's Z transformation, 95% confidence interval
 Calculation for Spearman's Correlation Coefficient
 Calculation for regression analysis
 Data plot for correlation and regression
 StatPgm 3a. Compare Two Measurements, including
 Calculating sample size requirements for comparing two means
 Calculations for comparing two parametric means, t test, 95% confidence interval of the difference
 Calculations for comparing two nonparametric measurements, MannWhitney U Test
 Data plotting for two groups of measurements
 StatPgm 3b. Compare Two Proportions, including
 Calculating sample size requirements for comparing two proportions
 Calculations for Chi Squares Test and Fisher's Exact Probability for 2x2 contingency tables
 Calculations for Risk Difference and Numbers Needed to Treat
 Calculations for Risk ratios
 Calculations for Odds ratios
 StatPgm 3c. Compare Two Regressions, including
 Calculations for comparing two regression coefficients
 Calculations for comparing two means adjusted for a common slope, simple covariance analysis
 Data plot for two regression lines
 StatPgm 4. MetaAnalysis, including
 Calculations for heterogeneity, Q Test and the I^{2} Test
 Calculations for Publication Bias, the Rank Correlation Test
 Calculations to combine results of multiple studies, the Fixed and Random Effect Models
 Forest Plot to display effect size of multiple studies
 StatPgm 5a. Tests : Binary Tests, including
 Calculations for True and False Positive and Negative Rates
 Calculations for Likelihood Ratios
 Calculations to produce posttest probability from pretest probability and likelihood ratio
 StatPgm 5b. Tests : Receiver Operator Characteristics (ROC), including
 Converting raw data to enable data entry for ROC calculations
 Calculations for Receiver Operator Characteristics (ROC)
 Tabulation of test statistics throughout range of predictor values
 Plot of ROC
 StatPgm 6. Graphic Editor, including
 Pie chart
 Bar plots, fixed and variable intervals for the x axis
 Scatter plot
 Forest Plot
 Data in groups plot
 StatPgm 7. Supportive Utilities, including
 Converting raw data to mean and Standard Deviation
 Converting raw data to table of counts
Exercises
Assessment
 Guidelines to Assessment : Sets out what the assessment is all about
 For Midwifery students
 For Reproductive Medicine students
