Unit- 4 Biostatistics & Research Methodology.pdf

BIOSTATICS & RESEARCH METHODOLOGY
UNIT-4

CONTENTS
Blocking and confounding (when a third variable, or confounder, in
fl
uences both the exposure and the
outcome) system for Two-level factorials (a type of experimental design where each factor (independent
variable) is investigated at only two levels, typically denoted as "high" and "low" or "+1" and "-1")
Regression modeling (statistical model that estimates the relationship between one dependent
variable and one or more independent variables using a line): Hypothesis testing in Simple and
Multiple regression models
Introduction to Practical components of Industrial and Clinical Trials Problems: Statistical Analysis
Using Excel, SPSS, MINITAB®, DESIGN OF EXPERIMENTS, R - Online Statistical Software to
Industrial and Clinical trial approach

Blocking and Confounding System for Two-level Factorials
These techniques are crucial for ensuring the validity and reliability of statistical experiments.-
Blocking
❖ Blocking is a technique used in experimental design to control variability (the tendency to shift or change) that is not of
primary interest.
❖ It involves grouping experimental units (subjects, materials, or conditions) into similar categories before applying
treatments.
❖ The goal is to reduce the effect of uncontrolled factors, making it easier to detect the real effect of the factors being
studied.
• Example:
❖ Suppose a pharmaceutical company is testing a new drug. The drug is produced using different batches of raw materials.
Some batches might be of higher purity, while others may have minor impurities.
❖ If we don't account for this variation, it could introduce unwanted noise in the experiment.
❖ By grouping (blocking) experimental runs based on batch numbers, we ensure that the effect of batch differences is
minimized, making the drug's effectiveness clearer.

Block is a group with contain elements
Blocking is the process of dividing the experimental units into homogenous group
Research - Hypothesis- Test on Population - Sample
+
Or The
process of
separating
the
experimental
unit into
group

Drug- A
+
Drug B
Drug- A
+
Drug C
Drug- A
+
Drug D
Drug- A
+
Morning
Drug- A
+
Afternoon
Drug- A
+
Night
To check effect
Homogenous Condition
B.L.O.C.K.I.N.G

Confounding
❖ Confounding occurs when two or more factors cannot be separated in their effects on the
response variable.
❖ In other words, the impact of one factor is mixed (confounded) with the impact of another
factor, making it dif
fi
cult to determine which factor is responsible for changes in the outcome.
Example:
❖ In a factory, we are testing two new machine settings (A and B) to improve product quality.
❖ At the same time, different operators (X and Y) are running the machines. If operator X
always works with setting A and operator Y always works with setting B, we cannot tell
whether the machine setting or the operator’s skill is responsible for any improvement in
quality.
❖ This is an example of confounding, as the effects of machine settings and operator skill are
mixed.

Confounding in 2k Factorial Design
❖ Confounding is a design technique where a complete factorial experiment is
arranged in a block.
❖ 22 Design - 22 = 4
4 - Combination (treatment)
Treatment Factorial Effect
Combination A B AB Block
1 - + - 1
2 + + + 2
3 - - + 2
4 + - - 1
- = 1 ,
+ = 2

Why is it Important?
❖ Improves experiment precision-
❖ Blocking helps reduce unwanted variability, making the results more
reliable.
❖ Confounding must be avoided to ensure that the results correctly attribute
effects to the right factors.
❖ Reduces noise in data-
❖ By properly controlling external factors, we can isolate the real effects of
the treatment being tested.

❖ Under homogeneous conditions, there are many conditions in which it is impossible to
perform all runs in a 2k
factorial experiment.
❖
For example, a single group of raw materials may not be suf
fi
cient to make other requirements
in process.
❖ There are many case where it may be desirable to vary the experimental conditions to ensure
that treatments are equally effective in many situations, likely to come into practice.
❖
For example, an engineer may run a pilot plant experiment with multiple batches of raw
materials as he knows quality grades of raw materials that are likely to be used in the process.
❖ Bloacking is a design technique that can be used in these situations. It is a method of
dealing controllable nuisance (Strong matches) variables.
❖

Regression modeling (statistical model that estimates the
relationship between one dependent variable and one or more
independent variables using a line): Hypothesis testing in Simple
and Multiple regression models

Regression Modelling
❖ Test hypothesis in simple and multiple regression model
❖ Regression is an estimation of the unknown value of one variable from the
known value of the related variable.
❖ Regression Modelling- It is the process of determining a relationship
between one or more independent variables to and one dependent variable.
❖ Ex- Predicting the sales of the product based on their quality.

Y= a+bx
Linear Regression
X
Y
Y= a+bx
Dependent
Independent
a = intercept
b = slope/ trend line
Hypothesis
The hypothesis is an assumption that we
make about the population parameter or
research
Hypothesis testing is the process by which
we decide the acceptance or rejection of the
hypothesis
- Null hypothesis and Alternate hypothesis

Hypothesis Testing in Simple Regression Modelling
❖ Study the relationship between the
two continuous variable
❖ One is the independent variable x
and other is the dependent variable y
❖ X is regarded as predictor and Y is
regarded as response variable
❖ For this equation Y = a0 + a1x
❖ Where, Y= dependent , x=
independent, a0 = intercept, a1= slope
X
Y
Dependent
Independent
a = intercept
b = slope/ trend line
Y = a0 + a1x
a1
a0

Hypothesis Testing in Multiple Regression Modelling
❖ In multiple regression equation Y on X1, X2, X3 ____ Xn is given by
❖ Y = a0 + a1x1 + a2x2 + a3x3 + anxn
❖ Where,
❖ Y= Dependent variable
❖ x1, x2, x3,_ _ _ _ _xn independent variable
❖ a0 = intercept
❖ a1 = slope
❖ For accept hypothesis, f calculated smaller than f tabulated

STATISTICAL ANALYSIS
❖ These are special program software that are used for complex analysis.
❖ These programs provide tools that can be used to organized, interpret and
present data set
❖ Statistical Analysis software -
❖ MS Excel
❖ SPSS
❖ Minitab

STATISTICAL ANALYSIS USING MS Excel
❖ Ms Excel (Microsoft excel) is one of the most powerful and commonly used
software for data analysis and to store records.
❖ It is box sheet software that contains small mini-software
❖ Eg- used in making result, mark sheet etc.
❖ This tool is cost-effective and also convenient to use
❖ It contains lots of features-
❖ The 3D effects, charts, and graphs, which help in the presentation of data
❖ It can arrange the large data in apprioprate manner
❖ It provide various formula helps in calculation and accurate data

❖ Application-
❖ For Pivot table- used to summarize a list
into simple format
❖ it summarize, analyze and present the data
❖ This tool can be used to make better
business decision
❖ For descriptive statistics - It helps to
fi
nd
mean, median, mode and range.
❖ Variance and Standard deviation
❖ Skewness
❖ Sample Variance
❖ Helps in regression

STATISTICAL ANALYSIS USING SPSS
❖ SPSS- Statistical data for the social science
❖ It is developed by Norman H, C. Hadlai and Dale H in 1968
❖ It is a software which is used to quick analysis of high volumne of social
science data, collected from different method of research
❖ SPSS is a computer program that is used for survey authoring and
development, data mining, text analytics, statistical analysis, and
collaboration.
❖ It basically convert the data into different presentable form such as graph, pie
charts, comparative bar, charts etc.

❖ Working of SPSS
❖ After installing, click on the window start button situated on left hand side of computer
screen. Next, click on program and select “ IBM SPSS Statistics 20”.
❖ Or by directly clicking the icon on desktop
❖ The 1st SPSS screen is shown following-
1. Run the tutorial
2. Type in data
3. Run an existing query
4. Create new query using database wizard
5. Open an existing data source
6. Open another type of
fi
le

❖ The researcher visited each store and recorded information on following
variables -
❖ Participant number
❖ Types of participant department
❖ Rating the quality
❖ Ownership of the store
❖ Uses
❖ The main aim of using the SPSS package is to de
fi
ne research problem,
formulated the research design, collect and analyze data and to assist the
research in the report preparation and presentation.

❖ It is used to quickly and accurately analyze the research data
❖ It helped by quantitative technique helps in decision-making
❖ It is used by researchers in
fi
eldwork and data collection
❖ Helps in tabulation of result
❖ It helps in cross tabulation and chi-square test.
❖ It also help in report writing

Minitab
❖ Minitab is a specialized software which is used to perform various statistical,
numerical and graphical calculations.
❖ It offers data to be represented graphically in the form of plots
❖ The major types of commands used in this software are menu commands and
subcommands
❖ Minitab is statistical software that helps a user to enter data quickly and then analyze
the data
❖ This data can be entired in the form of spread sheet using this program, it is very
easy to prepare charts and calculate regression.

Practical Components of Industrial and Clinical Trial Problems
❖ Main Aim is to improve the quality of life of patients
❖ Every year 10 to 12 % of clinical trial increases
❖ The signi
fi
cant component of clinical trials-
1. Involvement of human subject
2. Moving forward in time
3. Comparison with control
4. Method to measure intervention
5. Focusing on the unknown effects of medication
6. Conduct early in the development of study
7. Conduct hypothesis
8. Study Protocol

QUESTIONS
❖ Elaborate on any pharmaceutical example for data analysis using SPSS.
❖ Explain Regression Analysis
❖ What is regression analysis? Differentiate simple and multiple regression.
❖ What is the blocking and confounding effect in factorials?
❖ Distinguish between correlation and regression analysis and indicate the utility of regression
analysis in pharmaceutical research.
❖ Explain the importance of statistical software in pharmacy & describe SPSS
❖ Enumerate Statistical Software
❖ Write application of SAS and Minitab
❖ Discuss the importance of SPSS, Minitab and R software used for statistical analysis in research

Unit- 4 Biostatistics & Research Methodology.pdf

More Related Content

What's hot

Similar to Unit- 4 Biostatistics & Research Methodology.pdf

More from KRUTIKA CHANNE

Recently uploaded

Unit- 4 Biostatistics & Research Methodology.pdf