Lecture Slides

Semester I, Academic Year 2024-2025

International University, Vietnam National University – Ho Chi Minh City

Statistics for Health Science

Lecture 2: Describing Data: Measres of Central Tendency, Dispersion,
Frequency, Graphs, Normal Distribution, Sampling Distributions of the
Sample Means, Stadard Error, Confidence Interval
Tan-Nhu Nguyen, Ph.D.
Department of Biomedical Instrumentation
School of Biomedical Engineering
International University; Vietnam National University Ho Chi Minh City
Email: [email protected]
Facebook: https://www.facebook.com/michealnguyentannhu
Google scholar: https://scholar.google.com/citations?user=lvCf2XAAAAAJ&hl=en
ORCID: https://orcid.org/0000-0003-3343-0886
----------------------------------------------------------------------------------------------------------------------------- -------------------------------------------------------------------------------------------------------------------------------- --------------------------------
 Contents

1. Statistical Tables 3. Statistical Charts

1.1. Definition of Statistical Table 3.1. Pictorial Representation of Data
1.2. Purpose and Properties of Tables 3.2. Purposes of Pictorial Representation
1.3. Parts of a Table 3.3. Properties of Pictorial Representation
1.4. Guidelines for Tabulation 3.4. Pictograms
1.5. Types of Tables 3.5. Bar Charts
2. Frequency Tables 3.6. Pie Charts
2.1. Frequency Tabulation of Qualitative Data 3.7. Statistical Maps
2.2. Array 4. Statistical Graphs
2.3. Stem and Leaf Table 4.1. Histograms
2.4. Frequency Distribution 4.2. Polygons
2.5. Frequency Distribution of Quantitative Data 4.3. Ogives
2.6. Frequency Distribution of Qualitative Data 4.4. Line Graphs

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 2
 1. Statistical Tables - Definition of Statistical Table

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 3
 1. Statistical Tables - Purpose and Properties of Tables

Tables are used:

(1) To present the original data, in an orderly manner
(2) To show any distinct pattern in the data
(3) To summarise data in a small space so that the eye can easily take them in
(4) As a basis for further statistical analysis and interpretation of data
Tables have Properties:
(1) Exclusiveness
(2) Inclusiveness

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 4
 1. Statistical Tables - Parts of a Table

TABLE number
Title

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 5
 1. Statistical Tables - Guidelines for Tabulation

Basic Principles Model Table Title Columns and Rows Unit of Measurement

• Objective • Rough draft • Title all tables • Data into columns & rows • Include unit in the title
• Effective • Layout rows and • Write informative • Name all columns & rows if for whole data
• What is the Purpose columns titles • Short & clear names • If not for whole data ->
• Number is large ->
of this Table? • Do not place number them
write them all
punctuation after the • Minimize the V & H lines
title

Table Cell Total Source of References Footnote General Issues

• Fill all table cells • Indicate column • Give the source of • Bottom of the table • Number tables in Arabic
• Indicate the empty and/or row totals data • Lower-case letters numerals
cells • Indicate main totals • Sources: title, • Put the title above the
• “…” for not available table
and sub totals compiler, place, date • Use “double number”
• “-” for nil or less than • Bold all totals • Come immediately only in long reports that
half unit used after the table contain chapters or
• “X” for not applicable sections
• Do not enclose a table
with a box
Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 6
 1. Statistical Tables - Guidelines for Tabulation

• Prepare a table having the following

data in your computer

Name Age Occupation

Software
John Doe 35
Engineer
Data
Jane Smith 28
Scientist
Alice Product
50
Johnson Manager

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 7
 1. Statistical Tables - Types of Tables

Types of Table

Quantitative Tables Frequency Tables

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 8
 2. Frequency Tables - Frequency Tabulation of Qualitative Data

Example 4.1: Suppose the new Director of a hospital decides to give monthly
supplementary allowances to all the health workers in the hospital for working extra
hours each day. The set of numbers below shows the allowances in VND million of a
sample of 50 members of staff.

=> How can we make conclusions relating to these data?

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 9
 2. Frequency Tables - Array

Advantages Disadvantages
• Simplest arrangement • Considerable amount of time
The array from the example 4.1: • Easiest to note the lowest and • Not easy to note quickly and
highest values easily the content of data
• Support other operations in • Difficult to estimate the single
effective calculation value to represent the dataset
• Difficult to work with large
amount of number

Compute Array in Excel Example of Large Array in

Examples Excel Examples

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 10
 2. Frequency Tables - Stem and Leaf Table

(1) Determine (2) Rank the Observations (3) Place all stems on left (4) Place the leaves
the steam & the leaf Find largest & smallest Smallest at top on right

=> Determine how many times a number in the 230’s occurs

Advantages Disadvantages
• Easy to construct • Require
• Give the general shape of the
distribution
• Easy to put data into intervals

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 11
 2. Frequency Tables - Frequency Distribution

Ungrouped Data Grouped Data

(1) Determine values (2) Count the number (3) Add up to obtain (2) Count number of (3) Add up to obtain
(1) Select groups
of categories of times the total number observations the total number

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 12
 2. Frequency Tables - Frequency Distribution of Quantitative Data

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 13
 2. Frequency Tables - Frequency Distribution of Quantitative Data –
Grouped Frequency Distribution

𝑅
ℎ=
𝑙

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 14
 2. Frequency Tables - Frequency Distribution of Quantitative Data

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 15
 2. Frequency Tables - Frequency Distribution of Quantitative Data

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 16
 2. Frequency Tables - Frequency Distribution of Qualitative Data

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 17
 2. Frequency Tables - Frequency Distribution of Qualitative Data –
Norminal Data

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 18
 2. Frequency Tables - Frequency Distribution of Qualitative Data –
Ordinal Data

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 19
 3. Statistical Charts - Pictorial Representation of Data

Pictorial Representation

Charts
Graphs
(Diagrams)

• Suitable replacement for hundreds of data

• Much easily read than statistical data
• Suitable for untrained readers to be
understood

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 20
 3. Statistical Charts – Purposes & Properties of Pictorial
Representation

Purposes Properties

• Easier to hold in mind • Easier to read and understand

• Show more clearly any relations from data • Clear in its meaning
• Show degrees of difference between items • Carry all the necessary information
• Make quick, lasting, and accurate impression
of relation

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 21
 3. Statistical Charts - Pictograms

• Suitable for simple and less accurate

information

• Requires artistic ability to draw

• Manual drawing extract symbol is difficult
• Difficult to draw accurately fractions of a
symbol

(2) Give the total value (3) Give the source

(1) Choose the symbol
of the pictograms of data

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 22
 3. Statistical Charts - Bar Charts
Usages:
(1) To compare quantities of one year with another
(2) To compare quantities of one place with another
(3) To compare frequencies of categories of qualitative
(4) To compare frequencies of a discrete quantitative
variable
Guidelines:
(1) Scale
✓ All data can appear easily in the space available
✓ No scale on the axis of positioned bars
(2) Bars
✓ Have the same width
✓ The length/height should be suitable with
numbers
✓ Control scale if more differences are needed
✓ Clearly labelled
✓ Should have space
✓ Re-arrange the bars in order
Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 23
 3. Statistical Charts - Bar Charts – Types of Bar Charts
Simple Bar Charts Component Bar Charts Percentage Bar Charts Multiple Bar Charts

• Easy to draw, farily accurate, very informative, userful for comparison

• Comparative size
• Values can be read from the scales of values
• Significance of the information

• Difficult for items whose sizes are not comparative (some bar very
long, some bar very short)

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 24
 3. Statistical Charts - Pie Charts

Usages:
(1) Give visual comparisons between different components
(2) Give visual impressions of the proportion each component
contributes to the overall total
Steps:
(1) Find the total 𝑇
(2) Divide each component by 𝑇 to obtain proportion 𝑟 of the
components
(3) Multiply each 𝑟 with 360𝑜 to obtain the number of degrees
(4) Use the degrees to draw the pie chart
Guidelines:
(1) Summarise the data into only a few categories (<5)
(2) Order the pie chart
(3) Colour and shading should be used judiciously
(4) Use pie charts for displaying percentages

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 25
 3. Statistical Charts - Pie Charts

• Dramatic and appealing way of representing data

• The contribution of significances are more important than the items
• Show relation between the parts and the whole better than bar chart
• Component parts are easily be compared

• More difficult to be drawn in comparsion with bar chart

• Values cannot be read off
• Cannot compare different sets of data
• Cannot be userful to give quick picture of a general situation
• Little used when the number of components are large

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 26
 3. Statistical Charts - Statistical Maps

• Difficult to choose the units

• Can be drawn with the help of PC software

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 27
 4. Statistical Graphs – Histograms – Frequency Histograms

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 28
 4. Statistical Graphs – Histograms – Relative and Percentage
Frequency Histogram

• The are of the histogram shows the proportion of the total number of
observations inside intervals
• Reactangle clearly shows each separate class in the distribution

• Sample histographm cannot infer a population histogram

• Difficult to measure the similarity between two histograms
• Cannot be used for composite datasets

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 29
 4. Statistical Graphs - Polygons

• Simpler than its histogram

• Outline data more clearly
• The polygon becomes smoother when increasing the number of classes
and observations
• Several distributions can be compared on the same graph

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 30
 4. Statistical Graphs – Polygons – Frequency Curves

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 31
 4. Statistical Graphs - Ogives

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 32
 4. Statistical Graphs - Line Graphs – For Trend Analyzing

• Preferred for time series

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 33
 Homework

Conduct your homeworks

in the BlackBoard Website

Lecture Slides – Statistics for Health Science – Semester I – Academic Year: 2024 - 2025 34