Transcript Medical Statistics

Postgraduate books recommended by Degree
Management and Postgraduate Education
Bureau, Ministry of Education
Medical Statistics
(the 2nd edition)
孙振球
主
编
徐勇勇
副主编
Arrangement: total 72 class hours,
two classes each week
chapter 1
Introduction
1. Key definitions
2. the steps for medical statistics
3. Brief history of Statistics
Statistics

The science for data collection,
sorting, and analysis.
Medical Statistics
Definition：the science that study the
collection, sorting and analysis of medical
data.
Characteristics：
1、Using the quantity to reflect the quality
2、Using chance events (uncertainty) to
reflect the inevitability (rules)
Medical Statistics
Learning objectives：
1、Basic principles and methods of Statistics
（Learning Emphasis）
2、Application Statistics——(Clinical
Medicine, Preventive Medicine, and Health
Care Management)
Medical Statistics
Purpose：a tool for medical research
Emphasis: statistical indicators used for calculating
or comparing the quantitative characteristics of
population
Example: health expectation
infant mortality
Section 1. Key definitions
Ⅰ variable, individual, sample
and population
individual（observatory unit）：the basic unit in
statistical research, it depends on the purpose.
variable（indicator）：individual characteristics
examples: height、weight、gender、blood type、
treatment effect etc.
Variable value：the value of variables
Examples:
height 1.65 meters weight 52 kg
gender female
laboratory test
blood type “O”
negative
treatment effect better
Data: composed of a lot of variable values.
Example: Data for blood glucose
homogeneity：common characteristics for the given
individuals
example： the heights of the boys with the age of 7
living in Changsha 2004
variation: difference existing among the variable
values of homogeneity individuals
example: the different heights of the boys with the age
of 7 living in Changsha 2004
population
Definition：the whole homogeneity individuals determined
by specific purpose.
example：all the heights of boys at 7 that lived in
Changsha 2004
finite population：the space, time and population for a
specific population have been limited.
infinite population： no time and space limits for the
population. Such populations only exist in imagination, so it
is called infinite population.
sample
definition：the set of variable values of some
individuals sampled from the population at random.
Example: the heights of 200 boys at 7 from
Changsha.
Sampling study
Sample information
(statistic)
inference
Population
characteristics
（parameter）
note：sampling is only the way to get
information, inferring the population is our
purpose
Ⅱ、variable and data
measurement data: it is also called as
quantitative or numerical data. Its value is
quantitative. Measurement data always
has measurement units.
example：height data, weight data
enumeration data: qualitative or count data.
For such data, it needs to classify the
observation units before and count them.
Its value appear different characteristics
and sorts.
Binomial: gender, live or death, yes or no.

Multiple：blood type, A、B、O、AB.

ranked data: ordinal or semi-quantitative data. It need to
classify observatory units into different classes according
the extent before calculate the frequencies of each
groups. There exists obvious differences among different
classes.
example: to evaluate the treatment effect of one drug on
heart failure, we use the indicator (cured, better, worsen,
dead) to assess the treatment effect.
Choosing of statistical methods depends on the data
type to a great extent。
Data transformation
Quantitative data
ranked data（multiple）
binomial data
example：WBC（1/m3）count of five persons：
3000
6000
5000
8000
lower
normal normal normal
12000
quantitative variable
higher
qualitative variable
Binomial data: normal 3 persons; abnormal 2 persons
Multiple category data: lower 1 person; normal; 3 persons;
higher 1 person
Ⅲ error
definition：the difference between measurement value and
true value.
1、rand error：unstable and changing at random errors that
caused by uncontrolled factors. Commonly, rand errors are
referred to those errors appearing during repeated
measurements and sampling.
Often, measurement error is extremely lower than sampling
error. In Statistics, sampling error is the main study contents.
2. Nonrandom error is divided into systematic
error and non systematic error:
Systematic error: it is produced in experiment
and keeps constant or changes according
certain rules. Usually, its reasons are known and
controllable.
Nonsystematic error（gross error）: it is always
caused by obvious grosses.
Ⅳ、frequency and probability
1．Frequency
Given the same condition, repeat a trial for n
times independently. Among n trials, A
appears for m times，so the ratio of m/n is
called the frequency of random event A
among n trials.
P(A)  m n
2．probability: the likelihood of random events.
Given the same condition, repeat a trial for n times
independently. Among n trials, A appears for f times，so
the ratio of
f / n is called the frequency of random event
A. As n increases gradually, the frequency
f / n will
approach a constant. The constant is called the probability
of random event A and expressed in P (A) . In common, it
is abbreviated as
P
.
（A） 1
Range：0  P


 nonr andom event ;
if P
（A）
=0，A is absolutely nonoccurrence event 
if P
（A）
=1， A is absolute event.
i f 0 P
（A） 1， A is random event.
Frequency
is used to describe the sample,
while the probability for the population. m/n is
the estimation of
P (A) . As trials increases, the
estimation value is more reliable.
small probability event: Because the
conclusions are made based on a certain
significance level, statisticians always select
P(A)  0.05或 P(A)  0.01 as judge criterion. So
such events with P(A)  0.05或 P(A)  0.01 are
called small probability events. It means that
such events happen rarely and can be
regarded as nonoccurrence.
Section 2 the steps for
statistical work
Ⅰ design
Here, it means statistical design, the most
important factor for a successful research.
It involves the arrangements for the
process of data collection, sorting and
analysis.
Three principles for experiment design
1．randomization
2. Replication
3．control
Ⅱ Data collection
objective：to gather accurate and reliable raw data
data sources：
① statistical reporting
② routine records
③ purposive surveys or experiments
④ statistical yearbook and special data book
requirements：1、randomization
2、sufficient sample size
Ⅲ．Data sorting
It is the process that cleans and systematizes
raw data. Data sorting prepares the required
data for next step, data analysis.
Ⅳ Data/statistical analysis
objective ： to illustrate the rules hidden in the data. It
includes two aspects:
1. statistical description：it is the process of describing the
characteristics of data using statistical indicators, statistical
charts and statistical tables.
2. statistical inference ： the process of using sample
statistic to infer population parameter.
It consists of:
parameter estimation and hypothesis testing.
indicator
Statistical
description
Table and chart
Statistical
analysis
Parameter
estimation
Statistical
inference
Hypothesis
testing