statistics in nursing

Introduction
To introduce application of Statistics in
medicine, and some of the statistical methods
commonly used in health sciences research.
Application of Statistics in
Research/Measurements

Objectives
To appreciate the role of Statistics in medical
research
To understand some of the statistical principles of
good practice in medical investigation
To understand how to use and interpret some of
the statistical techniques used in medical data
analysis

Statistics may be defined as the
collection, presentation, analysis and interpretation
of results based on numerical (or) observational
data – Croxton & Cowden
Bio- Statistics may be defined as the
statistical methods applied to the biological
sciences
What is Statistics?

 Statistics is the only subject has the
capability of making inferences based on the
results of observational data.
 Example: Thermometer and Physician
Why Statistics is in research?

Development of Statistics
Francis Galton Regression theory and the use
of Statistical methods in
biometry
Karl Pearson The theory of distribution,
correlation analysis and
chi-square test
W.S. Gosset Students t-test
R.A. Fisher The theory of estimation, the
fiducia inference and the design
of experiments

Function of Statistics
Collection of data
Presentation of data
-Tabulation
-Diagrams and Graphs
Analysis of data
Interpretation of results

Collection of data
Survey method
Simple random sampling
Systematic sampling
Stratified random sampling
Cluster sampling
Multistage sampling
Experimental method
Another way of collecting data in by experimental
is an actual experiment is conducted in certain
individuals or unit about which the inference is to be
drawn.

Types of data
Quantitative data
- The data related to exact measurements
such as height, weight and age
Qualitative data
- The data concerned with qualitative
aspects like opinion, attitude and awareness

Types of variables
Continuous : Temperature, heart rate
Discrete : Categorical
Ordinal : Severity of colic, tumour size
Nominal : Breed, sex
Binomial : Yes or No, absent or present

Statistical analysis of data
Descriptive statistical analysis
Inferential statistical analysis

Descriptive measure in statistics
 Mean – An arithmetic average of given
observations
 Median – The value in which it cuts the
distribution into two equal halves
 Mode – The more number of times repeated
observation
 Mean deviation – The average value of the
observation deviated from one of its central values
either mean or median
 Standard deviation – The square root of the average
of squared deviation from it mean

What is correlation analysis?
Correlation is a characteristic that is found two
variables, which shows a sort of relationship
between them
Height and weight, Treatment and Response are
certain pairs of characteristics that exhibit
relationship
The relationship could be linear or non linear and
can be observed from statistical data
Correlation coefficient is a measure of linear
relationship between two variables and denoted by
r, known as Karl Pearson’s correlation coefficient
The value of r lies in between –1 and +1

A positive value indicates the increase in one
variable is accompanied by the proportionate
increase in the other variable, positive correlation
A negative value indicates the increase in one
variable causes the proportionate decrease in the
other variable, negative correlation
A zero value represents there is no relationship
between the variables, zero correlation
Correlation analysis is a statistical method that
explains the correlation among many variables,
which are possibly interrelated
The researcher should ensure that the relationship
is at least nearly linear before using the correlation
coefficient
The scatter diagram is a basic observation to be
made before examining the correlation

It is a statistical technique to study the cause and
effect relationship between two variables
One variable (BP) is identified as dependent variable
(effect) known to be influenced by one or more
variables(like body weight, age and heart rate) called
independent variables (causes)
Regression analysis is used to estimate a linear
relationship between the variables and hence it is
called linear regression given by the model
Y=a+bX+e, where a, b are constants and e is called
error component
Regression

The regression coefficient represents the marginal
change in Y due to a unit change in X
The study of regression between two variables is
known as simple regression
Multiple regression refers to the case of one
dependent variable and several independent variables
The goodness of the regression is usually measured
in terms of an index R2 called the coefficient of
determination
Its value lies between 0 and 1. Higher the value of R2 ,
stronger is the relationship

Inferential Statistics
Point estimation – To estimate the actual
value of the parameter of a distribution, i.e.,
Mean, S.D
Testing of hypothesis – This involves making
an assumption about the parameter and
checking the plausibility of that assumption
using sample data

Why we need to study the Testing of hypothesis?
What is hypothesis testing?
When to use Testing of hypothesis?

Various items involved in the testing of
hypothesis
Null hypothesis (H0)
Assumption that there is no difference between
the population parameters
Alternative hypothesis (H1)
Making contradiction with null hypothesis
Types of error
Type I error – Rejecting H0 when it is true
Type II error – Accepting H0 When it is false

Level of significance
The probability of type I error is called as level
of significance
Test Statistic
Z= [t-E(t)/SE(t)] ~ N(0,1), where t is the
sample statistic

P- Value concept
Another approach is to find out the P-value at
which H0 is significant, i.e., to find the smallest level
α at which H0 is rejected. In this situation , it is not
inferred whether H0 is accepted or rejected at level
0.05 or 0.01 or any other level. This facilitates an
individual to decide for himself as to how much
significant the data are. This approach avoids the
imposition of a fixed level of significance. About the
acceptance or rejection of H0 , the experimenter can
himself decide the level α by comparing it with the
P-value. The criterion for this is that if the P-value is
less than or equal to α reject H0 otherwise accept
H0

Finding out the type of problem and the question to be
answered
Stating the null hypothesis
Determining the correct sampling distribution and
calculating the standard error of the statistic used
Calculate the test statistic
Z= [t-E(t)/SE(t)] ~ N(0,1)
Comparison with the predetermined significant level
given by the table
Making inferences
Test Procedure

 Large sample – Sample size n is greater than 30
 Small sample - Sample size n is less than or equal to
30
Tests based on large samples
 Comparison of sample proportions
 Comparison of sample means
Type of Sample

Tests based on small samples
t-test – To test the equality of means and also
correlation coefficient
Chi-Square test – To test the population variance and
extensively used to study the independence or
association between the attributes
F-test – To test the equality of variance

 The importance and applications
of statistics have been illustrated
through examples from the medical
sciences. An effort has been made
to create better statistics skill
among the health professionals
Conclusion

statistics in nursing

More Related Content

What's hot (20)

Similar to statistics in nursing (20)

More from Pratibha Srivastava (9)

Recently uploaded (20)

statistics in nursing