MedI7 Intro 2011 Mortalitate Alcool

Embed Size (px)

Citation preview

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    1/48

    1

    Research design and methodology,statistics and data handling (MedIs7)

    Biostatistics II (KVT3)

    Lecturers:

    Alina Zalounina (AZ)Carsten Dahl (CD)Dan Karbing (DK)

    Department of Health Science and TechnologyAalborg University

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    2/48

    2

    Aim of the course

    Learn to use and understand normally appliedstatistics in medical research (advanced level)- Software tool: SPSS

    Introduce important aspects of research work(only for MedIs7)

    - Funding and intellectual property rightsMedical/research writingPlanning studies

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    3/48

    3

    ProgrammeDate Topic Lecturer

    Sept 2 Introductory statistics AZ

    Sept 5 Funding academic research DKSept 9 Developing and testing a hypothesis AZ

    Sept 12 Medical writing DK

    Sept 16 Contingency tables CD

    Sept 21 Parametric analysis DK

    Sept 23 Non-parametric analysis CD

    Sept 26 Intellectual Property Right (IPR) and patents DK

    Sept 30 Regression analysis CD

    Okt 5 Study design AZ

    Okt 7 Survival analysis CD

    Okt 19 Meta-analysis. Evidence-based medicine AZ

    = only for MedIs7

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    4/48

    4

    Learning material

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    5/48

    5

    Software: SPSS

    http://spss.software.aau.dk/

    The students are expected to bring a laptop.

    SPSS should be installed before the coursestarts using guidelines at:

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    6/48

    6

    Examination

    Written exam - 4 hours, in January

    Pensum: slides + learning material specifiedfor each lecture

    Exam questions will reflect lectures and course

    assignments

    Some questions will require use of a software (SPSS)

    Hjlpemidler: everything, but not Internetand communication with others

    More detailed Exam info can be found at

    http://person.hst.aau.dk/az/MedIs7

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    7/48

    7

    Today:Introductory statistics

    Alina Zalounina

    Center for Model-basedMedical Decision Support

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    8/48

    8

    Learning material:

    Chapter 1: DataChapters 2-5: Descriptive StatisticsChapters 7-8: Statistical Inference

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    9/48

    9

    Learning Objectives

    Identify the type of data

    Define and understand the main termsof Descriptive Statistics

    Understand the purpose of InferentialStatistics

    Outline the major measures of risk

    Outline the basic operations in SPSS

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    10/48

    Type of data

    Categoricaldata

    Metricdata

    Nominal

    etnicitygendermarital status

    type of operationsmoking status

    Ordinal

    score

    Discrete

    number ofchildren

    Continuous

    weightheighttemp.

    ageblood pressuretimecholesterolbody mass index

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    11/48

    11

    Samples and populations

    Sample = collected dataPopulation = all possible data

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    12/48

    12

    Type of Statistics

    Descriptive used to organize and

    describe a sample

    Inferential used to extrapolate from asample to a larger population

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    13/48

    13

    Learning Objectives

    Identify the type of data

    Define and understand themain terms of Descriptive Statistics

    Understand the purpose of InferentialStatistics

    Outline the major measures of risk

    Outline the basic operations in SPSS

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    14/48

    14

    Descriptive Statistics.Issues for today

    Frequency

    Measures of Central Tendency- Mean

    - Median

    Measures of Variability- Variance

    - Standard deviation- Standard error

    Descriptive Plots- Boxplot- Histogram

    - Q-Q plot

    Data distibutions- Normal- Binomial

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    15/48

    15

    Frequency table

    Relative frequency

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    16/48

    16

    Measures of Central TendencyMean(average)

    Sample Mean

    PopulationMean

    1

    n

    i

    ix

    n

    x

    1

    N

    i

    i

    N

    x

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    17/48

    17

    Median(middle)

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    18/48

    18

    Measures of VariabilityVariance

    Sample Variance Population Variance

    2

    2 1

    -1

    ( )n

    i i

    n

    xxs

    2

    2 1( )

    N

    ii

    N

    x

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    19/48

    19

    Standard deviation, Standard error

    Sample SD Population SD

    2

    1

    ( )N

    i

    i

    N

    x

    2

    1

    -1

    ( )n

    i

    i

    sn

    xx

    sse=

    nStandardError

    D i i Pl

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    20/48

    20

    Descriptive PlotsBoxplot

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    21/48

    21

    Histogram

    The histogram shows the frequency distribution across aset of measurements as a set of physical bars.

    Overall shape curve shows distribution

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    22/48

    22

    Bell-shaped

    In a perfect normal frequency distribution, the mean and

    median are equal. The data is continuous andsymmetrically distributed around the central point.

    Variability is represented by the width of the distribution.

    Normal distribution

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    23/48

    Normal distribution: formulae

    X~N(, 2) =>

    b

    a

    f(x)dxb)XP(a

    2

    2

    1

    x-

    2

    1f(x) e

    Note: about 95% of observations liewithin 1.96*standard deviation ofthe mean

    +1.96*-1.96*

    95%

    X = a continuous variablef(x) = probabilitydistribution function of X= mean= standard deviation

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    24/48

    24

    Without inspecting the data it is risky to assume a normaldistribution.There are a number of graphs that can be used to check thedeviations of the data from the normal distribution:

    A histogramshould reveal a bell shaped curve.

    QQ plot: Curvature of the points indicates departures of

    normality

    Check normality

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    25/48

    25

    Skew distribution

    This population is skewed to the right(i.e. it has a long right hand tail)

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    26/48

    26

    Binomial distribution

    o There are nidentical independenttrialso Each trial can have only 2 outcomes: success or failureo Probability pof success in each trial is constanto Variable of interest is X=the number ofsuccesses in ntrials

    binomial variable

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    27/48

    27

    Binomial distribution: formula

    x n-xn!P(X=x)= (1 p)px!(n-x)!

    Note: n!=n(n-1)(n-2)1

    X~Bin(n,p) =>

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    28/48

    28

    Binomial distribution: Example

    The probability that a student is accepted tothe Department of Medicine is 0.3.

    If 5 students from the same school apply,what is the probability that 2 are accepted?

    P(X=2)=?

    p=0.3

    n=5

    5-225!=> P(X=2)= (1 0.3) 0.310.32!(5-2)!

    probabilitydistribution

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    29/48

    29

    Learning Objectives

    Identify the type of data

    Define and understand themain terms of Descriptive Statistics

    Understand the purpose ofInferential Statistics

    Outline the major measures of risk

    Outline the basic operations in SPSS

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    30/48

    30

    Inferential Statistics

    Can your experiment make a statement about

    the general population?

    Two types of tests:

    1. Parametricassume that the variable in question has a knownunderlying mathematical distribution that can bedescribed (normal, binomial, etc.)

    2. Non-Parametricare considered distribution-free methods becausethey do not rely on any underlying mathematicaldistribution.

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    31/48

    31

    Learning Objectives

    Identify the type of data

    Define and understand themain terms of Descriptive Statistics

    Understand the purpose ofInferential Statistics

    Outline the major measures of risk

    Outline the basic operations in SPSS

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    32/48

    Risks and Odds.Issues for today

    Risk(probability) = a measure of the chance ofgetting some outcome of interest (e.g., disease) fromsome event (e.g., exposure to a risk factor)

    Absolute risk Relative risk Odds Odds ratio

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    33/48

    Mother smoked duringpregnancy

    Yes No Totals

    Apgarscore

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    34/48

    Relative risk(Risk Ratio, RR)= the risk for theexposed group compared to the risk for the non-exposedgroup.

    The risk of low score among those having smoked

    compared to those who did not smoke is

    RR= p1/p2 = 80%/15% = 5.3

    Interpretation of RR:

    Mothers who smoked during pregnancy had more than 5 timesthe risk of getting low Apgar score as those who did not smoke.

    Risk (low score | smoking)= p1= 80 %Risk (low score | no smoking)= p2= 15 %

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    35/48

    Apgar score

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    36/48

    Odds ratio (OR)

    The ratio between the odds is the odds ratio for smokingamong mothers with low score compared to mothers withhigh score:

    OR= odds1/odds2 = 22.67

    Interpretation of OR:Mothers with low Apgar score were more than 22 times aslikely to have smoked during pregnancy as those with highApgar score.

    Odds (mothers with low score smoked) = odds1= 2.7Odds (mothers with high score smoked) = odds2= 0.12

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    37/48

    RR versus OR

    A B

    C D

    Outcome No Outcome

    Exposed

    Non-Exposed BC

    ADOR

    A B

    C D

    Exposed Non-Exposed

    Outcome

    No Outcome C)B(AD)A(BRR

    RR=1 or OR=1 => there is no association between the outcomeand exposure to risk factor

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    38/48

    38

    Learning Objectives

    Identify the type of data

    Define and understand themain terms of Descriptive Statistics

    Understand the purpose ofInferential Statistics

    Outline the major measures of risk

    Outline the basic operations in SPSS

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    39/48

    39

    Example

    Introduction to SPSS

    D t i

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    40/48

    Data view

    Variable view

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    41/48

    41

    Variable view

    Smoking

    LowApgarScore

    Frequences

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    42/48

    42

    Frequences

    Cross - Tabulations

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    43/48

    43

    Risk estimate

    Odds ratio: 22.667=(17/2)/(3/8)

    Relative risks:4.25=(17/20)/(2/10)0.188=(3/20)/(8/10)

    odds ratio for non-smoking among mothers withhigh score compared to mothers with low score:

    risk of low score among those who did notsmoke compared to those having smoked

    risk of high score among those who did notsmoke compared to those having smoked

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    44/48

    44

    Box-plot

    Descriptives

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    45/48

    45

    Histogram

    Q Q Plot

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    46/48

    46

    Q-Q Plot

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    47/48

    47

    Learning Objectives

    Identify the type of data

    Define and understand the main termsof Descriptive Statistics

    Understand the purpose of InferentialStatistics

    Outline the major measures of risk

    Outline the basic operations in SPSS

    E i

  • 7/23/2019 MedI7 Intro 2011 Mortalitate Alcool

    48/48

    48

    Exercises:http://person.hst.aau.dk/az/MedIs7