DENTAL (DIA 2006)

2018/2019

Basic statistics for pharmacology

Practicals
A/P Dr. Mohammed Abdullah, PhD

Department of Pharmacology
Faculty of Medicine
University of Malaya
alshaweshmam@[Link]

14th September 2018

[Link]
 Lecture outlines

• Introduction to statistics
• Statistic in Pharmacology
• Types of data
• Summarizing data
• Describing data & Normality
• Selection tests of significance
• Hypothesis testing & Significance level (p value)
• Student’s t-test
• Data presentation
 Biostatistics

Biostatistics is the application of statistics to a wide range

of topics in biology, and medically related data. The

statistics methods include:

• Collection of data
• Analysis of the data
• Presentation of the collected data
• Interpretation of the results
• Making decisions on the basis of such analysis
 Statistic in Pharmacology

Biological experiments & measurements are associated

with variations, e.g. routs of drugs administration
• Statistics are essential for dealing with these variations
• Summarising & describing data
• Comparing groups – using tests of significance to calculate
probabilities associated with experimental findings
• Can help you to interpret your results
• To help you draw scientific conclusions from your data
 Types of data

• Quantitative data (numerical): data can be measured e.g. weight

(g), onset of convulsion (seconds), sleeping time (min), levels of…
• Qualitative data (categorical): data can be observed but not
measured e.g. race, sex

Sources of
data

Records/previous studies Surveys Experiments

(secondary data) (primary data) (primary data)
 Summarizing data

• Measures of central tendency: a measure of central tendency

is a measure which indicates where the middle (center) of the

data is.

• Mean

• Median

• Mode
 Mean

• The sum of a series of numbers (x) divided by the number of

values (n).
• The magnitude of each value is important, SO, easily affected
by extreme values (Disadvantage)
 Describing data & Normality

• Measures of dispersion (Variation): distribution of values from

the mean
• Range
• Variance (S2)
• Standard deviation (SD)
• Standard error of the mean (SEM)
• Confidence interval
• Coefficient of variation (CV)
 Variance and Standard deviations (SD)

• Provides information on the variability of the observations

around the mean

• A measure of the variability or dispersion of a population, or a

data set. A low SD indicates the data points tend to be very

close to the same value (the mean), while high SD indicates

that the data are spread out over a large range of values.
 Variance and Standard deviations (SD)

No. x
• Variance (S2): the average of the squares of the 1 540
2 480
3 290
difference of each observation from the mean. 4 300
5 480
• Standard deviation (SD): the square root of the 6 120
7 300
sample variance. 8 260
Mean
STD
SEM
Standard error of the mean (SEM)
 Normal Distribution (Normality)

• An important aspect of the "description" of a variable

• Normal distribution is important, because It indicates the

type of analysis that you can run on the data (selection tests of
significance)

• Normal distribution is represented by a group of curves (mean

& SD).

• The curves are - Symmetrical

Normal Distribution
 Normal Distribution

• 1 SD: 68.2 %
• 2 SD: 95.4 %
• 3 SD: 99.6 %
 Shape of curve distribution

Skewness – measure of symmetry of a distribution. In most

instances the comparison is made to a normal distribution
I. Symmetrical – bell-shaped distribution
II. Positively skewed distribution – elongated tail to the right
III. Negatively skewed distribution – elongated tail to the left
 Selection tests of significance

• Normally distributed data Parametric tests: e.g.

Student’s t-test, One way ANOVA
Normally distributed data (mean ±SD or SEM)

• Non-normally distributed (skewed) data Non-parametric

tests: e.g. chi-square test, Mann Whitney test, Wilcoxon
signed-rank test
Non-normally distributed data (median ± Range
or Interquartile range)
 Hypothesis testing

To answer a statistical question, the question is translated

into a hypothesis
• The hypothesis tested is known as null hypothesis (H0)
• The alternative hypothesis is HA which is contradicted to
the null hypothesis

Example:
– H0 : There are no difference between the mean of two groups
– HA : There are difference between the mean of two groups
• Hypothesis may be accepted or rejected based on the
probability (p) value at certain significance level
• To accept or reject the hypothesis the significance level
must be set
 Significance level (p value)

• p- value is the probability of getting the output observed,

assuming the null hypothesis to be true.
– P=0.05 (probability 1 in 20); significant
– P=0.02 (probability 1 in 50); very significant
– P=0.01 (probability 1 in 100); extremely significant

• For significance level at 0.05:

H0 = There are no difference
HA = There are difference
• P- value ≤ 0.05, significant different between the two mean
values, strong enough evidence to reject the null hypothesis

• P- value > 0.05, no different between the two mean values, so

accept null hypothesis
 Student’s t-test

• Introduced by William Gosset in 1908

• “Student” his pen name
• A test of significance to calculate the
probabilities associated with experimental
finding
• Student’s t-test is used to evaluate the
level of confidence for the difference
obtained between two normally
distributed mean values
 Student’s t-test

During calculations of the t value, the significance/

insignificance of the difference between the two mean
values depend on the following factors:
• The number of observations (n)
• The degree of differences between two mean values
• The dispersion of the individual value from the mean value
 Student’s t-test

Type of t test:
i. Independent t- test (two-sample t-test): to test the
difference between two independent variables, (e.g.
case/control)
ii. Paired t- test: to test the difference between two
dependent variables or paired populations (before/after ,
2 methods used on the same animal), e.g. glucose level
before & after treatment
How to calculate t-test?
 How to calculate t-test?

• Refer the calculated t value with the t value given in the student
t-test table (at the appendix of the practical handout) which
shows the t value distribution
• Find the row for the degree of freedom (df); df = n-1; and in the
same row, seek the t value that is closest to the calculated t
value.
• – For two sample t test, df = (n-1)1 + (n-1)2
• Look at the probability (P) value corresponds to the calculated
t value
 How to calculate t-test?

• The difference is considered significant If the difference

between the mean values occurs at the rate of p value obtain
≤ 0.05
• the null hypothesis can be rejected (the two means are
significantly difference)
- Rejecting the null hypothesis; the two means are
significantly difference.

- Accepting the null hypothesis; the two means are not

significantly difference
 Example

30.15
 Pharmacology Practical

For each practical experiment:

• Record the data
• Statistically analyse the data (calculate: mean, SD, SEM
for each group)
• Using Student’s t-test determine the significant difference
between the means of each two groups.
• The result should be presented in suitable form (tables,
graphs…etc)
Practical example
 Data presentation

• Numerical presentation: Tables

• Graphical presentation: Figures

Figures
 Histogram  Bar Chart
 Line graph  Pie Chart
 Scatter Diagram
Tables

Table title

Unit
Histogram
Histogram
Bar chart

- No class boundaries
- Spaces (gaps)
Line Graph
Scatter Diagram (plot)
Pie Chart
 References

1. Practical handout, DIA 2006, Session 2017/2018, Department of

Pharmacology, Faculty of Medicine, University of Malaya

2. Also I would like to thank Dr. Zamri Chik and A/P Dr. Hesham Al-Mekhlafi for
using some of theirs slides .

• Images taken from the online materials with website link as shown.
Thank You