“ A Number speaks more than thousand words.

Statistical Methods and Data Analysis

Bijay Lal Pradhan, PhD

Associate Professor
Tribhuvan University
 “Most people use statistics like a
drunk man uses a lamppost; more
for support than illumination”

― Andrew Lang
Research Method
Errors using inadequate data are much less than those using no data at all.
-Charles Babbage
● Computer technology: Changes in teaching learning statistics
● Researcher if not aware: affect in research capabilities
● This course is designed to enhance research capability of students by
analysis of data using technology.
Research Environment
● Paper-and-pencil to point-and- click
● Small set of data to large set of data
● Simple research questions to complex research questions
● Ad-hoc decision making process to research-based decision making
process
● More and more emphasis on interdisciplinary nature of research
● Development of computer based information system for formulating
research-based policies
● Development of research for monitoring and evaluation of development
works
Quantitative Research
Methods and materials for quantitative research.
Within the domain of social sciences, quantitative research relates to
empirical methods originating from thought - positivism – and its
development parallels with the development of the subject statistics as
well as the development of computer technology.
A quantitative research - clearly defined set of research questions
including hypotheses - objective answers - data.
Research Process
Source of New Idea Heart and Soul The Blue Print

Prior Research
Statement of Research Design
Own Experience Problem/ Research
Question
Population Sample
Curiosity

Instrumentation

End Product Data Collection

Findings, Conclusions
Recommendation and Data Analysis
Implication
Statistical Methods
Statistical methods for collecting, organizing, analyzing, interpreting and
presenting data and
generalizing the sample information to a population from which the
sample was drawn.
Definition of Statistics
In modern days, the term "statistics" is used in two senses.
● In plural sense the word refers to a set of numerical facts or estimates
generally presented in tabular or graphical form. It is in this sense that
the public usually think about statistics, but it follows the different
characteristics: aggregate, affected by multiplicity of causes,
enumeration/estimation, comparable, purpose.
● In singular sense the term refers to a subject of study that deals with
scientific methods of collection, organization, analysis, interpretation
and presentation of data. In singular sense the term is often
understood as statistical methods.
Statistics
Statistics is a methodological discipline. It exists not for itself but rather to
offer to other fields of study a coherent set of ideas and tools for dealing
with data. Statistics requires a different kind of thinking, because data are
not just numbers, they are numbers with a context (Cobb and Moore;
1997). Context provides meaning.
 Scopes & Limitations of Data Analysis

Collection Organization Presentation Analysis Reporting

If the data were collected from a random sample drawn from a

well-defined population in such a way that every unit in the population
has a known non-zero probability of being included in the sample, then
the information derived from such sample can be generalized to the
population (inferential statistics). If the data were collected from a
non-random sample, then the information derived from sample cannot be
generalized (descriptive statistics).
Research Steps: Data
After identifying research problem, setting research objectives, and preparing a
list of research questions and variables - a researcher’s effort turns to the
process of collecting data.
If the data required for answering research questions are readily available in the
secondary sources (e. g. reports and electronic data files), then further data
collection process stops here.
In practice, secondary sources often fail to provide adequate data, as a result
researcher’s effort turns for collecting primary data.
•Sample survey is the most widely used method for collecting primary data in
quantitative research.
 Data types
Variables

Scale Categorical

Continuous Ordinal: Nominal:

Discrete:
Measurements Categories, order no meaningful order
Counts/ integers
takes any value within Categories
Whole Number value: Eg: 1/2/3/4/5
the range
typically Count Eg. Male/Female
Height of tree, Qty of Min/Moderate/High
Number of tree cutdown Smoker/ Non Smoker
fuel consumption.
 Descriptive Analysis
Variables

Scale Categorical

Ordinal: Nominal:
Continuous Discrete:

Mean Median Mode Mean Median Mode Median, Quartiles,,

Standard Deviation Standard Deviation Percentage, Mode
Percentile, Q.D.
skewness Kurtosis skewness Kurtosis
 Test of Hypothesis to be used
Data

Scale Categorical

Ordinal: Nominal:
Normally
Skewed data Non-parametric Chi-squared
distributed
Non-parametric
Parametric test
 What data types relate to following questions?

Q1: What is your favourite subject?

Management English Statistics Energy Environment
Q2: Gender:
Male Female
Q3: I consider myself to be good at environment:
Strongly Disagree Disagree Not Sure Agree Strongly Agree

Q4: Score in a Qualifying Examination.

Score between 0% and 100%

Q5: How many members are there in your house.

Number of People in house
 Data Types (form)
Data Types
Information, that necessary for any study are achieved in different forms. The main forms
of the information available are as following.
Cross Section Data

Time Series Data

Panel Data

Spatial Data
Populations and samples
Taking a sample from a population

Sample data ‘represents’ the whole population

 Descriptive & Inferential Statistics
Statistics
Descriptive Inferential

Tabular
Hypothesis
Numerical Estimation Testing

Graphical
Point Parametric Non-Parametric
Interval

Methods of inferential statistics are applicable when results are obtained from a random.
Uncertainty always remains while generalizing results from a sample to a
population. The degree of uncertainty is measured in terms of probability in
inferential statistics.
Course detail
Contents LH 1. Pro
Proba
1. Descriptive Statistics, Exploratory data analysis, Foundations of data analysis 6
Distri
2. Probability and Probability distributions (Binomial, Poisson, Normal). 6 2. Es
Hypo
3. Sampling techniques 3
3. An
4. Sample design, Sampling distribution of mean and proportions. 3 Varia
4. Sim
5. Manage research data using software (SPSS) 3
Regr
6. Inferential Statistics: Estimation of mean and proportion, Z test and 6 Analy
5. Mu
Regr
Course detail
Contents LH 1. Pro
Proba
7. Hypothesis Testing for small sample t-test for one and two sample 3
Distri
8. Analysis of variance 3 2. Es
Hypo
9. Review of linear models: correlation, linear regression & residual analysis using software 6
3. An
10. Hypothesis testing for categorical data: chi-square test 3 Varia
11. Exploratory Factor Analysis (EFA) using software 3 4. Sim
Regr
12. Discriminant Analysis using software 3
Analy
5. Mu
Regr
Type of statistics

Descriptive statistics – Methods of organizing,

summarizing, and presenting data in an informative way
Inferential statistics – The methods used to determine
something about a population on the basis of a sample
Population –The entire set of individuals or objects of interest or
the measurements obtained from all individuals or objects of
interest
Sample – A portion, or part, of the population of interest
Statistics

Statistics
A Bunch of
Numbers Gives
the information
Descriptive Statistics
Probability and Probability Distribution
Patterns, trends, and characteristics identified through studies of past data
can be of assistance in establishing the likelihood of future events.
This course begins with the study of probability, random variables, discrete
and continuous distributions, and the use of calculus to obtain expressions
for parameters of these distributions such as the mean and variance.
Concept and importance of probability, Types of events, Approaches to
probability, Theorems of probability, Conditional probability, Baye’s
theorem and mathematical expectation. Concept of theoretical
distribution: Binomial, Poisson and Normal distribution, Fitting of Binomial
and Poisson distribution.
Sampling Estimation and Hypothesis Test
By a small sample, we may judge of the whole piece. -Miguel de Cervantes,

sampling, Methods of sampling, Sampling distribution, Standard error and its utility in
testing of hypothesis.

Estimation and estimators, types of estimates, Criteria of a good estimator, Relationship

among the errors, the risk and the sample size, Estimation of sample size, Interval
estimation

Parametric tests for variables and attributes, Tests of significance of mean, proportion,
difference of means and difference of proportion, Small sample test: Student’s – test, Test
of significance of mean and difference of means, paired t – test, F – test (ANOVA).

Test of significance of attributes along with its relationship (chi square)

Regression Analysis
“Regression analysis is the hydrogen bomb of the statistics arsenal.” ―
Charles Wheelan

Identification of the factors that have an impact on a particular issue can be

done with regression analysis. Regression analysis enables you to accurately
identify the factors that are most important and those that can be ignored, and
identifies the relationships between these variables.
Multivariate Analysis
Factor analysis groups attributes that are alike.
This technique can be used to examine interrelationships among many
variables and to explain these variables in terms of their common underlying
and unobservable dimensions (called “factors”).
Researchers use factor analysis to reduce the information contained in
several original variables into a smaller, more manageable set of variables
while losing as little information as possible.

Discriminant Analysis is a classification technique that deals with the data

with a response variable and predictor variables. It is mainly used to classify
the observation to a class or category based on the independent variables of
the data. The two types of Discriminant Analysis.
Statistical
Softwares
Bijay Lal Pradhan, PhD
Associate Professor
Amrit Campus
SPSS

SPSS stands for Statistical Package for the Social Sciences. It was one of the
earliest statistical packages with Version 1 being released in 1968

● Very easy to learn and use

● Can use either with menus or syntax files
● Quite good graphics
● Descriptive statistics, basic regression analysis, analysis of variance, and
some newer techniques such as Classification and Regression Trees (CART)
● Has its own structural equation modelling software AMOS, that integrated
with SPSS
● Focus is on statistical methods mainly used in the social sciences, market
research and psychology
● Difficult to use for panel data analysis

Availability: We should buy for its genuine software.