ARDENT MDS CLASS

Subject: Public Health Dentistry

Bio-statistics
Dr. Karunanidhi K, M.D.S
Assistant Professor
Meenakshi Ammal Dental College and Hospital,
Meeankshi Academy of Higher Education and Research (MAHER)

Date: 22.11.2024

Time: 6.00 PM-9.00 P.M

WHY EDUCATION IS IMPORTANT
 What is STATISTICS ?
• Statistics is the science of compiling, classifying
and tabulating numerical data and expressing the
results in a mathematical or graphical form.
 What is BIOSTATISTICS ?
• Biostatistics is a branch of statistics
concerned with mathematical facts and data
related to biological events.
TYPES OF VARIABLES
 What is A VARIABLE?
• A variable is a state, condition, concept or event
whose value is free to vary within the population.

Dependent Independent
variable variable
INDEPENDENT VARIABLE

• Independent variable is a cause

which changes the effect,
dependent variable.
DEPENDENT VARIABLE

• Dependent variable is an effect,

or outcome which tends to be
changed due to an independent
variable.
 QUESTION SESSION
I want to assess the effect of lecture classes on student’s
performance.
WHAT ARE THE INDEPENDENT, DEPENDENT
VARIABLES ?
 QUESTION SESSION
I want to assess the effect of a good story on a film’s
success..
WHAT IS THE DEPENDENT VARIABLE ?
WHAT IS THE INDEPENDENT VARIABLE ?
CONFOUNDING VARIABLE

• A confounding variable is that

variable which is independently
associated with both effect and
cause.
TYPES OF DATA
 TYPES OF DATA

QUALITATIVE DATA
QUANTITATIVE DATA
 TYPES OF DATA
Qualitative Quantitative
data data

Categorical Continuous Discrete

Nominal Ordinal Interval Ratio

TYPES OF QUALITATIVE DATA
NOMINAL DATA ORDINAL DATA

CATEGORIES WITH
CATEGORIES WITH
ORDER
NO ORDER
 QUALITATIVE DATA
NOMINAL DATA ORDINAL DATA

CATEGORIES WITH CATEGORIES WITH

NO ORDER ORDER
QUANTITATIVE DATA
INTERVAL
RATIO

HOLDS VALUE BELOW NEVER HOLD VALUE

ZERO BELOW ZERO
 QUESTION SESSION
I want to categorize students as good, average
and poor based on their marks scored in their last
internal test?
WHAT TYPE OF DATA IS THIS?
QUALITATIVE DATA- CATEGORICAL -ORDINAL
 QUESTION SESSION
I want to record length of the desks present in a
lecture hall.
WHAT TYPE OF DATA IT IS ?
QUANTITATIVE DATA- CONTINUOUS DATA (RATIO)
 QUESTION SESSION
I want to categorize you all based on your choice
of the speciality of MDS.
WHAT TYPE OF DATA IS THIS?
QUALITATIVE DATA- CATEGORICAL DATA -NOMINAL
CONVERSION OF Quantitative DATA TO Qualitative DATA
WHAT TYPE OF DATA IT IS ?

Number of missing,
decayed and filled teeth
WHAT TYPE OF DATA IT IS ?

Russell’s index
WHAT TYPE OF DATA IT IS ?
WHAT TYPE OF DATA IT IS ?

DICHOTOMOUS DATA
 NEET MODEL QUESTIONS
EASY

When the frequency is given as mild, moderate

and severe, the data scale used is
a) Ordinal b) Nominal c) Variance d) Interval
 NEET MODEL QUESTIONS
EASY

When the frequency is given as mild, moderate

and severe, the data scale used is
a) Ordinal b) Nominal c) Variance d) Interval
 NEET MODEL QUESTIONS
MODERATE

Angle’s class of malocclusion (Class I, Class II,

Class III) is what type of data?
a) Ordinal b) Nominal c) Ratio d) Interval
 NEET MODEL QUESTIONS
MODERATE

Angle’s class of malocclusion (Class I, Class II,

Class III) is what type of data?
a) Ordinal b) Nominal c) Ratio d) Interval
 NEET MODEL QUESTIONS
DIFFICULT

When the variables are not measurable, which of

the following is used to be represented ?
a) Variation b) Ordinal c) Nominal d) Ratio
 NEET MODEL QUESTIONS
DIFFICULT

When the variables are not measurable, which of

the following is used to be represented ?
a) Variation b) Ordinal c) Nominal d) Ratio
MEASURES OF CENTRAL TENDENCY
MEASURES OF CENTRAL TENDENCY
MMM

Mean Median

Mode
Arithmetic Mean

• Arithmetic Mean is obtained by

adding all the individual
observations and then divided by
the total number of observations.
MEDIAN

• Median is the middle value in a

distribution after arranging all the
observations in ascending value.
MODE

• Mode is the value in a series of

observation that occurs with
greater frequency.
Find MEAN, MEDIAN & MODE FOR THE FOLLOWING DATA

Data set: 4,5,6,7,8

• Mean = (4+5+6+7+8)/5 =6
• Median = 6
• Mode= ?

• Mode = 3 Median – 2 Mean

• Mode = (3x6)-(2x6)
• Mode= 18-12=6
 FIND THE MEAN & MEDIAN

92 91
Mean = 92
STUDENT A
88 95 Median = 92

94
 FIND THE MEAN & MEDIAN

61 91
Mean = 70.6
STUDENT B
56 62 Median = 62

83
MEASURES OF DISPERSION
MEASURES OF DISPERSION
MSV-RC

Mean Standard
Variance
Deviation Deviation

Inter
Range Quartile
Range
MEAN DEVIATION
• The mean deviation is defined as

a statistical measure that is used to

calculate the average deviation from

the mean value.

STANDARD DEVIATION

• Standard Deviation is the most

used measure of studying dispersion.
• Greater the standard deviation,
greater will be magnitude of
dispersion from mean.
FIND THE STANDARD DEVIATION
Mean (X) = 92

MD=
92 91
X X I(X-X)I (X-X)²
92 92 0 0
88 95 91 92 1 1
88 92 4 16
95 92 3 9

94 94 92 2
10
4
30

M.D = 10/5 = 2 S.D = √30/5 = 2.44

FIND THE STANDARD DEVIATION
Mean (X) = 80
60 90

100 80

70
FIND THE STANDARD DEVIATION
Mean (X) = 80
60 90
X X I(X-X)I (X-X)²
60 80 20 400
100 80 90 80 10 100
100 80 20 400
80 80 0 0

70 70 80 10
60
100
1000

M.D = 60/5 = 12 S.D = √1000/5 = 14.14

 MD vs SD
• Mean deviation takes an absolute value, hence doesn’t
predict deviations accurately in the data sets having
outliers. Hence, Mean deviation is not sensitive to
data having outliers.

• Standard deviation obtained by squaring the differences

predicts the deviation in the data set more accurately.
Standard Deviations are sensitive to data having
outliers.
VARIANCE

• The value of variance is equal

to the square of standard

deviation, which is another

measure of dispersion.
 INTER QUARTILE RANGE
Marks scored by five students: 30,35,40,60,65.

25th 50th 75th 100th

1st Percentile
Percentile Percentile percentile percentile
30 35 40 60 65

50th percentile is Median

IQR: 75 percentile-25
th th

percentile
IQR: (60-35)
 QUARTILE
• Quartile divides the observations in to four equal
parts.
REPRESENTATION OF DATA
Qualitative n (%)
data

Mean ±
Quantitative S.D
data
Median ± IQR
NORMAL DISTRIBUTION
WHEN TO USE MEAN ? AND WHEN TO USE MEDIAN ?

STUDENT A STUDENT B

92 91 88 61 91 56

95 94 62 83
 MEAN & MEDIAN

88 91 92
STUDENT A

94 95

Mean = 92 Median = 92
 FIND THE MEAN & MEDIAN

61 91
STUDENT B
56 62

83
 FIND THE MEAN & MEDIAN

56 61 62

STUDENT B

83 92

Mean = 70.6 Median = 62

NORMAL DISTRIBUTION

• Data near the mean are more frequent in

occurrence than data farther than mean.

• Normal distribution produces symmetrical

curve.
NORMAL DISTRIBUTION
SKEWNESS
Is the following data normally distributed ?

10 10 10

20 40 70

90
 NEET MODEL QUESTIONS
EASY

Best measure of central tendency in highly

skewed data
a) Mean b) Median c) Mode d) Range
 NEET MODEL QUESTIONS
EASY

Best measure of central tendency in highly

skewed data
a) Mean b) Median c) Mode d) Range
 NEET MODEL QUESTIONS
DIFFICULT

Which of these are not closest to the median

a) 50 percentile b) Second Quartile
th

c) 2nd tertile d) 5th Decile

 NEET MODEL QUESTIONS
DIFFICULT

Which of these are not closest to the median

a) 50 percentile b) Second Quartile
th

c) 2nd tertile d) 5th Decile

A KUTTY STORY !!
HYPOTHESIS & TESTING OF HYPOTHESIS
 HYPOTHESIS
• Hypothesis is a predictive tentative statement on
association between two variables.

RESEARCH/ALTERNATE
NULL HYPOTHESIS
HYPOTHESIS
 HYPOTHESIS
RIYAZ BIRIYANI IS BETTER
RESEARCH/ALTERNATE
THAN SS HYDERABAD
HYPOTHESIS
BIRIYANI

RIYAZ BIRIYANI IS NOT

NULL HYPOTHESIS BETTER THAN SS
HYDERABAD BIRIYANI
TYPE I and TYPE II ERRORS
PROBABILITY OF ERROR (P-VALUE)

• Probability of error is denoted by

p-value.
• It tells whether the results are
chance results or true results.
 PROBABILITY OF ERROR

Probability of
p-value Statistically
error is less
<0.05 significant
than 5%

ACCEPT
TRUE RESULTS RESEARCH
HYPOTHESIS
 PROBABILITY OF ERROR

Probability Statistically
p-value=0.23 of error is non-
23% significant

CHANCE ACCEPT NULL

RESULTS HYPOTHESIS
 PROBABILITY OF ERROR
50

54% p-value= 0.34 46%

 PROBABILITY OF ERROR
300

74% 26%
p-value= 0.04
 Comparisons
• The table showing comparison of gingival bleeding scores
before and after use of two mouthwashes

Before After 30 days

Chlorhexidine 2.4+0.9 0.6+0.2

mouthwash
(n=20)
Listerine 2.2+0.7 1.6+0.4
mouthwash
(n=20)
 HYPOTHESIS TESTING

Non-parametric
Parametric tests
tests of
of significance
significance
PARAMETRIC AND NON-PARAMETRIC TESTS

Parametric tests Non-parametric

tests
Data Quantitative Qualitative or
Quantitative
Distribution Normal distribution Non-normal or
skewed
Sample size Large > 30 Small < 30
Sampling method Random Non-random
Expressed in Mean and Median and IQR or
standard deviation number (n) and
percentage (%)
PARAMETRIC AND NON PARAMETRIC TEsts

Non-parametric
Number of Type of Parametric tests
tests of
groups comparison of significance
significance
Intra group Wilcoxon sign rank
Paired t-test
comparison test
2 Groups Un-paired t-
Inter group
test/Independent Man Whitney test
comparison
sample t-test
Intra group Repeated
Friedman test
comparison Measures ANOVA
3 Groups
Inter group
One Way ANOVA Kruskal Wallis test
comparison
 QUANTITATIVE DATA

Tests for 2 groups & 3 groups

Before After 30 days Before After 30 days

Chlorhexidine 2.4+0.9 0.6+0.2 Chlorhexidine 2.4+0.9 0.6+0.2

mouthwash mouthwash
(n=20) (n=20)
Listerine 2.2+0.7 1.6+0.4 Listerine 2.2+0.7 1.6+0.4
mouthwash mouthwash
(n=20) (n=20)
Salt water 2.5+0.8 2.1+0.9
(n=20)
t-test

ANOVA
 CHI-SQUARE TEST
Anemic Non- Total
• Used to find association between
(n) anemic (n)
Male 10 40 50 two variables (when both
Female 40 10 50 variables are categorical)
• For small sample size (Fischer
Total 50 50 100
Exact test is used)
 NEET MODEL QUESTIONS
EASY
Study that is used to compare serum cholesterol levels
in obese and non-obese women. To find the relation of
significance what test to be used?
a) Chi-square test b) Paired t-test
c) Un-paired t-test d) One Way ANOVA
 NEET MODEL QUESTIONS
EASY
Study that is used to compare serum cholesterol levels
in obese and non-obese women. To find the relation of
significance what test to be used?
a) Chi-square test b) Paired t-test
c) Un-paired t-test d) One Way ANOVA
 NEET MODEL QUESTIONS
EASY
Study that is used to compare gingival index scores
between three different types of mouthwash. To find the
relation of significance what test to be used?
a) Chi-square test b) Paired t-test
c) Un-paired t-test d) One Way ANOVA
 NEET MODEL QUESTIONS
EASY
Study that is used to compare gingival index scores
between three different types of mouthwash. To find the
relation of significance what test to be used?
a) Chi-square test b) Paired t-test
c) Un-paired t-test d) One Way ANOVA
 CORRELATION
POSITIVE
CORRELATION
• Correlation is denoted by "r".
AGE SALARY
• It takes values from -1 to +1.
• Tells whether increase in one
NEGATIVE
CORRELATION variable increases or decreases
OXYGEN
ALTITUDE another variable.
LEVEL
CORRELATION

• Pearson correlation: Used to

see correlation between two
continuous variables.
• Spearman correlation: Used
to see correlation between two
categorical variables.
 NEET MODEL QUESTIONS
MODERATE

Which of the following has highest correlation

coefficient?
a) 0.56 b) 0.64 c) 0.32 d) -0.75
 NEET MODEL QUESTIONS
MODERATE

Which of the following has highest correlation

coefficient?
a) 0.56 b) 0.64 c) 0.32 d) -0.75
 DEGREE OF FREEDOM

• Degree of freedom (df) is Number of independent numbers

in a sample.

• For paired t-test: df= n-1

• For unpaired t-test: df= (n1+n2) -1
• For chi-square test: df= (r-1) (c-1) where r = no. of rows and
c= number of columns
REPRESENTATION OF DATA
BAR DIAGRAM LINE DIAGRAM
REPRESENTATION OF DATA
FREQUENCY
HISTOGRAM
POLYGON
REPRESENTATION OF DATA
PIE DIAGRAM HEAT MAP
CORRELATION
SCATTER PLOT
DIAGRAM
 NEET MODEL QUESTIONS
EASY

Diagrams used to show percentage breakdowns for

qualitative data
a) Frequency polygon b) Pie diagram c) Line diagram d)
Bar diagram
 NEET MODEL QUESTIONS
EASY

Diagrams used to show percentage breakdowns for

qualitative data
a) Frequency polygon b) Pie diagram c) Line diagram
d) Bar diagram
 NEET MODEL QUESTIONS
MODERATE

Histogram is a
a) Frequency polygon b) Pie diagram c) Line diagram d)
Bar diagram
 NEET MODEL QUESTIONS
MODERATE

Histogram is a
a) Frequency polygon b) Pie diagram c) Line diagram
d) Bar diagram