Introduction

Statistical Methods

“Most data cannot be efficiently

utilized without statistical method and that
statistical method is futile unless applied to a
data”

Snedecor & Cochran, 1989

 Research and Statistics
• Research is an organized and scientific
way of finding answers to questions
• collection of methods and
methodologies that researchers apply
systematically to produce scientifically
based knowledge
 Scientific Research
• A methodical study in order to prove a
hypothesis or answer a specific question
• Provides scientific information and theories
for the explanation of the nature and
properties of the world
• Requires the use of the scientific method
 Types of Research
 Basic  Comparative
 Applied  Exploratory
 Quantitative  Explanatory
 Qualitative  Causal
 Longitudinal  Theory-testing
 Descriptive  Theory-building
 classification  Action
 Research methods
 Observational
Units are observed in the normal
environment that they exist
Experimental
Units are observed in an environment that is
highly controlled by the researcher
 Experiment
 Methodical trial and error procedure carried
out with the goal of verifying, falsifying or
establishing the validity of the hypothesis

 Relies on repeatable procedure and logical

analysis of the results
 Experiment
• Discovering new facts and their correct
interpretation
• Revising accepted conclusions, theories or
laws in the light of the newly discovered facts
or demonstrate the practical applications of
such new or revised conclusions
• Repeating results of earlier experiments and
extend them to more complicated and general
situations
 Statistics
• Scientific discipline consisting of theories and methods
for processing numerical information that one can use
when making decisions in the face of uncertainty.
– Refers to the methods in collection, presentation, analysis
and interpretation of data.
 collecting and summarizing data
 describing data and variables
 Designing experiments and surveys
 estimating population parameters
 Testing hypothesis about populations
 Studying relationships among variables
Roles of Statistics in Research

• “tool kit” for research workers in the planning,

analyzing and interpreting the results of his
investigations using methods or techniques
that were proved useful in many fields of
inquiry
• Statistics in research
– Problems in real life
– Formulations of specific problems
Data collection
• design of experiments
• sample surveys
– Historical
– record measurements
– prior information
Data examination
• Data cleaning and test of prior information
Data analysis
• Model selection and validation
• Hypothesis testing, estimation, prediction and
decision-making
Steps in conducting experiments
• State the problem
• Specify the objective of the study
• Formulate hypothesis
• Define the data to collect, the statistical analysis of the data
• Determine the exptl technique and design
• Construct the exptl layout
• Conduct the experiment and collect the necessary data
• Analyze the data obtained from the experiment
• Summarize the results, make conclusions and inferences
• Evaluate the whole study in relation to similar investigation
 Experimental Design
• Involves the assignment of treatments to the
experimental units

• Thorough understanding of the analysis to be

performed once the data become available
 Relationship between statistics and
experimental designs
 Experiments involves probability samples
 Experiments lead to inference that must be
accompanied by probability statements
 Experiments should be designed in accordance with
the principles of statistics for the probability statement
to be valid
 Importance of Experimental Designs
 Ensure cost effective collection of appropriate
data
 Provides a roadmap for appropriate and valid
analysis of data
 Allows for drawing valid conclusions using
statistical inference
Terms:
• Treatment or Factor. Set of experimental
procedures or conditions whose effects are to be
measured and compared.
• Experimental unit (eu). Unit or group of units
(experimental materials or individual) to which a
single treatment is applied
• Sampling unit. Portion of the experimental unit on
which the response variable is observed and
measured.
• Response variable. Characteristic used to measure
the effect of a treatment.
• Experimental error. Variations in the observed
values of the response variable from the eu’s
treated alike
– Failure of two eu’s to yield the same response
Sources:
• inherent variability of the experimental materials
used
• Error in experimentation
• Errors in observations and measurements
• Combined effects of all extraneous uncontrolled
factors
• Sampling error. Measure of variation among
sampling units within experimental units
• Levels. Pre-set quantity of a quantitative
factor/treatment or categories of a
factor/treatment under study
• Layout. Final arrangement of treatment levels
over the entire set of eu’s
Fixed model vs. Random effects
• Fixed model. All features under test are fixed factors
– A factor is considered fixed when its levels are selected on
purpose
– When all levels are present in the experiment
– Only applies to the set of population where the test is
applied
• Random model. All factors under test are random
factors
– When the levels of the factors tested are random samples
from a population of levels
– Applies to the general population
• Mixed model. E.g. Varieties and location of trials
• Precision. The ability of a measurement to be
consistently reproduced
– Measured by the variance
– Ways to increase:
• Increase no. of samples
• Skilful grouping of exptl materials
• Proper selection of treatments
• Accuracy. Closeness of the observed values to the
true values
– Measured by the bias (diff b/n the average of the ov
and the tv)
– Ways to increase:
• Refinement of exptl technique
 Principles of Experimentation
 Randomization
– Methods of allocating the treatment levels to the eu using some chance
mechanism such that eu’s equally likely to be assigned to a treatment level
 Functions:
– To have a valid measure of experimental error
– To provide a random sample of observation
– To eliminate systematic bias in assigning the treatments to the eu’s
– To satisfy assumption of independent errors
– To minimize errors associated with experimental units that are adjacent in
space and time
 Principles of Experimentation
 Local Control ( Error Control)
– All practices or techniques used to minimize the experimental
error which includes balancing, blocking and grouping of the
experimental units
– Making the design more efficient making the test of significance
more sensitive or the test procedure more powerful
 Common techniques of local control:
– Use of most appropriate experimental design
– Use of proper shape and size of eu
– Use of concomitant variable
– Refinement of experimental technique
 Principles of Experimental Design
 Grouping – placing homogeneous eu’s into a groups
and comparing the treatments in each group
 Blocking – grouping the experimental units into
blocks such that the units within a block are
relatively homogeneous
 Balancing – grouping, blocking and assignment of
the treatments to the eu in such a way that a
balanced configuration results
 Structure of Experimental Design
 Treatment Structure – consists of the Design Structure – refer to
set of treatments, treatment grouping of the experimental
combinations or populations that the units into homogeneous
researcher has selected to groups or blocks
study/compare • Completely randomized
– One-way treatment – treatments design
consist of levels of a single factor
• Randomized complete block
– Factorial – treatments consists of all
possible combinations of one levels of
design
two or more factors being tested • Latin square design
– Fractional factorial – only a fraction of • Split-plot design
the full set of treatment combinations • Strip-plot design
in a factorial treatment combination
• Strip-split plot design
– Factorial arrangement with one or
more controls • Incomplete block design
– Nested treatment structure
 Considerations in choosing the
experimental design
 Objectives of the research
 Description of responses or animal/farm performance to be evaluated
– Quantitative vs qualitative
– Univariate vs multivariate
 Descriptions of factors hypothesized to influence animal on farm performance
[independent variables]
– Primary factors, nuisance factors (blocking factors/factor levels)
– Fixed effects vs random effects
 Practical restrictions (cost, time, allocations of limited resources)
•
Approaches in data analysis
Exploratory data analysis – process of looking at raw
data to decide on its important features
– Rounding-off data
– Grouping of data in a convenient form
– Identifying outliers
– Finding mean and quantile values
• Determining violations of assumption of anova
– Non-additivity
– Heterogeneity of variance
– Non-normality
– Non-linearity
 Approaches in data analysis
 Model Building – mathematical (linear or non-linear)
model to describe to describe biological processes
or relationship.
– Represent theoretical relationships that are observed to
exist the real life process, situation or problem
 Analysis of variances
 Comparison among treatment means
 Assumptions of ANOVA
 homogeneity of variance of the expl errors (i.e., erratic
effects of treatments outliers, non-normal or skewed
distribution of data)
 Tests:
 Hartley’s F-max; Ho: σ21 =σ22 = σ23; DR: Reject Ho if Fmaxc >
Fmax(0.05). Should not be used for heavily skewed distribution.
 Bartlett’s test (U); Ho: σ21 = σ22 = σ23. Most powerful when all
treatments are ND.
 Levene’s test. Uses absolute deviation from the mean to perform ANOVA.
The DR is based on the comparison of Fc and Ft (a=0.05).
 Brown-Forsythe test.
 Cochran’s test.
 Assumptions of ANOVA
 Normality of expl errors (cause extremely skewed
distribution which can be reduced by increasing
sample sizes)
 Tests;
 Chi-square goodness of fit
 Kolmogorov-Smirnov
 Wilk-Shapiro W test
 Normal probability plots
 Assumptions of ANOVA
 Independence of errors
 Tests:
 Plot of the predicted Yij vs Residual. No pattern means independent.
 Durbin-Watson
 Runs test
 Additivity of effects
 No interaction b/n treatment and environmental effects
 Effects of treatments across
 Require transformation if violation occurs
 Test: Tukey’s test for non-aditivity
 Model Building
Weight gains (kg) of pigs to varying levels probiotic The model will be:
Natuphos (phytase)
Level (gm/kg Initial
feeds) Sex Weight Weight gain yijk = µ + Leveli + Sexj + Initial_weightk +
2 M 12.0 24
4 F 10.0 20
ɛ(ijk);
6 F 11.0 21
2 M 13.0 25 where
4 M 11.0 23 y is the observed values of
6 F 10.0 22 weight gain
2 F 9.5 19.5
4 M 10.5 21
µ is the overall mean of weight
6 M 11.5 23 gain
2 F 12.0 25.5 ɛ(ijkl)m is the random error
4 F 10.5 21.5 associated with the ith level given to
6 M 12.5 26.5
2 F 11.5 23
jth sex with the kth initial weight
4 F 9.5 20.5
6 M 10.5 21