2014.3.10 1 medical statistics medical statistics tao yuchun tao yuchun practice 1
TRANSCRIPT
2014.3.101
Medical StatisticsMedical Statistics
Tao YuchunTao Yuchun
Practice 1Practice 1
http://cc.jlu.edu.cn/ms.html
2014.3.102
Review Review
1.1. Population and Sample
•Randomization
2.2. Probability and Frequency
I.I. Basic conceptsBasic concepts
3.3. Parameter and Statistic
2014.3.103
II. Types of dataII. Types of data
1.1. Numerical Variable (Measurement Data)
•quantitatively
2.2. Categorical Variable (Enumeration Data)
•qualitatively •Nominal Variable •Count Data
3.3. Ordinal variable ( Rank data )
2014.3.104
III. The Basic Steps of Statistical WorkIII. The Basic Steps of Statistical Work
1.1. Design of study Design of study
2.2. Collection of data Collection of data
3.3. Data SortingData Sorting
4.4. Data Analysis Data Analysis
• Descriptive statistics
• Inferential statistics
2014.3.105
IV. Statistical Description for MeasurementIV. Statistical Description for Measurement
DataData
1.1 1.1 Frequency Distribution
(1)(1) Steps of establishing a frequency tableSteps of establishing a frequency table
((2)2) Frequency plot ---histogram Frequency plot ---histogram
((3)3) The use of frequency table The use of frequency table
• central tendency
• measure of dispersion
2014.3.106
1.2 1.2 Measures for Average ★
(1)(1) Arithmetic Mean (Arithmetic Mean (meanmean))
μ X
n
X
n
X
n
XXXX
n
ii
n
121 ...
•Suitable to symmetric distribution.
2014.3.107
(2)(2) Geometric Mean (Geometric Mean (GG))
)lg
(lg)lg...lglg
(lg 1211
21
n
X
n
XXXG
XXXG
n
nn
•lg-1 =10x
•Suitable to positive skew distribution,
like geometric progression.
2014.3.108
(3)(3) Median (Median (MM))
a. For raw data
•Ranking the data, finding the middle value.
For odd number, it is; for even number, it is
mean of two middle values.
b. For frequency table
)2
( LM
fn
f
iLM
•Suitable to all kinds of data, but usually to
positive skew distribution.
2014.3.109
Add:Add: PercentilePercentile --- Px
P Li
fn x f
x
x
L ( % )
•Symmetric •Positive skew
1.3 1.3 Measures for variability ★
(1)(1) Range (Range (RR))
R = Maximum - Minimum
•Suitable to all kinds of data, but no useful.
2014.3.1010
(2)(2) InterQuartile Range (InterQuartile Range (IQRIQR))
257513 PPQQIQR
•Suitable to all kinds of data, but usually skew
distribution.
(3)(3) Variance Variance andand Standard Deviation ( Standard Deviation (SS2 2 and and SS))
1
)( 22
n
XXS
1
/)(
1
)( 222
n
nXX
n
XXS
2014.3.1011
•Suitable to symmetric distribution.
(4)(4) Coefficient of Variation (Coefficient of Variation (CVCV))
CVS
X 100%
• Comparison of the variation of two variables
with different dimensions oror bigger difference
of means .
2014.3.1012
Calculative tools Calculative tools
I.I. Scientific calculatorScientific calculator
• The scientific calculator with statistical
function. (like CASIO fx-3600PV or CASIO fx-82TL)
• Calculative method is often:
Input all raw data, press the special
button under statistical mode, you will
get the result (like , S) directly. X
2014.3.1013
II. Excel II. Excel ★
• Many statistical function.
• The macro of statistical analysis tool.
• Using any expression directly.
• Many statistical graphs.
•See the example (stat1(English).xls)
2014.3.1014
III. Statistical softwareIII. Statistical software
• Professional statistical analysis tool.
• The special data management and
statistical analysis procedure.
• The special executive commands.
• Include almost all statistical methods.
• SAS, SPSSSPSS, Stata, BDMP, …
• If you want use it, you should have tolearn another lesson.
2014.3.1015
Practice in class Practice in class Exercise 1Exercise 1: the blood-glucose(mmol/L) values from
12 randomly selected patients.
5.31, 6.12, 6.53, 6.53, 6.65, 6.66, 6.71, 6.93, 7.05,
7.15, 7.21, 7.35
Please calculate the arithmetic mean, geometric
mean and median; range, quartile range and standard
deviation.
2014.3.1016
Exercise 2Exercise 2: the frequency table of latent period (day)
from 110 certain infectious disease patients.
(1) (2)
2~ 26 26 23.644~ 48 74 67.276~ 25 99 90.008~ 6 105 95.45
10~ 3 108 98.1812~14 2 110 100.00
total 110 - -
(4)=(3)/n
tab2 the frequency table of latent period (day) from some infectious disease patients
latentperiod
CumulativeFrequency(∑ f )frequency(f )
CumulativeFrequency(%)
(3)
2014.3.1017
1) Please calculate the arithmetic mean, geometric mean and
median, which one better reflects the average level ?
2) Please calculate the range, quartile range and standard
deviation, which one better reflects the variation ?
Answer Answer
•See the Excel file (practice1key.xls)
2014.3.1018
HomeworkHomework There are raw data of temperature (℃) for 102 female
students from certain college(see below).
37.05 36.90 37.20 37.10 37.00 36.85 36.85 37.40 37.05 36.8537.20 37.00 37.00 36.90 36.85 37.15 37.10 36.80 37.40 37.4037.30 37.40 37.25 37.10 37.10 36.85 36.80 37.05 37.00 36.9037.35 37.25 36.95 37.05 36.80 37.15 37.05 37.15 37.15 37.2537.50 37.00 37.35 37.05 37.10 37.00 37.05 37.35 37.10 37.1037.25 37.20 36.95 37.00 37.10 37.00 36.90 37.05 37.00 36.9036.55 36.80 37.05 36.60 37.05 37.20 36.70 37.20 36.90 37.3036.85 36.70 37.15 37.10 37.05 36.95 37.25 36.90 37.05 36.7536.90 36.85 36.70 36.95 37.15 36.90 37.05 36.90 37.35 37.0537.05 37.00 37.35 37.10 37.20 36.65 36.65 36.90 36.95 36.9036.70 36.80
2014.3.1019
CC
1) Please work out a frequency table and a histogram.
2) Please calculate the arithmetic mean, median, quartile
range, standard deviation and coefficient of variation.
(http://en.wikipedia.org/wiki/Great_Wall_of_China)
2014.3.1020
CASIO
2014.3.1021
CASIO fx-82TL
2014.3.1022