M113
Advanced
Statistics
Introduction to Research ◆ Overview of Non-parametric
Statistics ◆ Importance of Non-parametric Statistics
Preliminary Activity: Round Table Discussion
● What is statistics?
● What is Research?
● Importance of Statistics?
● What is parametric statistics?
● What is non-parametric statistics?
01
What is Statistics?
Statistics
● The science of collecting, organizing,
presenting, analyzing, and interpreting
numerical data for the purpose of
assisting in making a more effective
decision.
Statistics
• refers to the processes that are used in
Descriptive presenting and describing data.
Statistics
• refers to the processes of making inferences
about a large group, called population, based
Inferential on observations of a smaller group called
Statistics sample.
02
Introduction to
Research
What is Research?
“A systematic investigation (i.e., the
gathering and analysis of information)
designed to develop or contribute
to generalizable knowledge.”
What is Research?
Research is different than other forms
of discovering knowledge (like reading
a book) because it uses a systematic
process called the Scientific Method.
03
Role of Statistics in
Research
Role of Statistics in Research
● Establishing Sample Size
● Testing of Hypothesis
● Data Interpretation through Analysis
● Decision making based on findings
04
Types of Statistical
Tests
Types of Statistical Tests
• Used with interval and ratio data and usually with data that
were obtained from groups randomly assigned, normally
Parametric distributed, and equal variability between group.
Statistics
• Used with nominal and ordinal data and sometimes with
interval and ratio data when other assumptions cannot be
Nonparametric met.
Statistics
What is a Parameter?
● A parameter is a useful component
of statistical analysis. It refers to the
characteristics that are used to define a
given population. It is used to describe a
specific characteristic of the entire
population.
Parameter vs. Statistic
● When making an inference about the
population, the parameter is unknown
because it would be impossible to collect
information from every member of the
population. Rather, we use a statistic of a
sample picked from the population to
derive a conclusion about the parameter.
Parameter vs. Statistic
● When making an inference about the
population, the parameter is unknown
because it would be impossible to collect
information from every member of the
population. Rather, we use a statistic of a
sample picked from the population to
derive a conclusion about the parameter.
Overview of
Non-parametric Statistics
What is a Non-parametric Test?
● Non-parametric tests are the mathematical methods
used in statistical hypothesis testing, which do not
make assumptions about the frequency distribution
of variables that are to be evaluated. The non-
parametric experiment is used when there are
skewed data, and it comprises techniques that do
not depend on data pertaining to any particular
distribution.
Advantages of Non-parametric Test?
• Easily understandable
• Short calculations
• Assumption of distribution is not required
• Applicable to all types of data
• Not susceptible to outliers
Disadvantages of Non-parametric Test?
• Less efficient as compared to parametric
test
• The results may or may not provide an
accurate answer because they are
distribution free
Some Common Non-parametric Tests
• Chi-square
• Mann Whitney U Test
• Sign Test
• Kruskal Wallis Test
• Wilcoxon signed-rank Test
Activity: Venn Diagram
● Compare and Contrast Parametric and Non-
parametric Statistics
06
Variables, Measurements of Data
Indicate whether each of the following statements is a descriptive
or inferential statistics.
● Last semester, the ages of the students at a certain college ranged from 16 to 25
years old.
A. Descriptive
B. Inferential
Indicate whether each of the following statements is a descriptive
or inferential statistics.
1. Based on the survey conducted by the National Statistics Office, it is estimated
that 24% of unemployed people are women.
A. Descriptive
B. Inferential
Indicate whether each of the following statements is a descriptive
or inferential statistics.
1. A survey says that 1 out of 10 Filipinos is a member of a fitness center.
A. Descriptive
B. Inferential
Indicate which of the following examples refer to population or sample.
● A group of 25 students selected to test a new teaching technique
A. Population
B. Sample
Indicate which of the following examples refer to population or sample.
● The total number of machines produced by a factory in one week
A. Population
B. Sample
Indicate which of the following examples refer to population or sample.
The yearly expenditures on food for 10 families
A. Population
B. Sample
Variables are the quantities measured in a
sample. They may be classified as:
Quantitative Categorical
Continuous (e.g. pH Nominal (e.g. gender,
blood group)
of a sample, patient
cholesterol levels)
Ordinal (ranked e.g. mild,
Discrete (e.g. moderate or severe illness).
Often ordinal variables are re-
number of bacteria coded to be quantitative.
colonies in a culture)
Classify each variable as categorical or quantitative.
The height of giraffe living in India
A. Categorical
B. Quantitative
Classify each variable as categorical or quantitative.
The religious affiliation of the people in the Philippines
A. Categorical
B. Quantitative
Classify each variable as categorical or quantitative.
Favorite movie
A. Categorical
B. Quantitative
Classify each variable as categorical or quantitative.
The daily intake of proteins
A. Categorical
B. Quantitative
Classify each variable as categorical or quantitative.
Nationality
A. Categorical
B. Quantitative
Measurement
The assignment of numbers to properties of objects, events or
persons according to logically accepted rule.
Nominal • Refers to the data that can only be classified into
categories
Level
Ordinal • Presumes that one category is higher than other.
Level
Interval • Includes the ranking characteristics of the ordinal level of
measurement and specifies that the distance between
Level numbers is the same.
• All the characteristics of the interval level of
Ratio measurement.
Examples
VARIABLE QUALITATIVE QUANTITATIVE DEPENDENT INDEPENDENT MEASUREMENT
S VARIABLE VARIABLE VARIABLE VARIABLE
GENDER GENDER NUMBER OF (depends on (depends on the Nominal
EMPLOYEES the problem) problem) Color, education
IQ RELIGIOUS HEIGHT AND Ordinal
AFFILIATION WEIGHT Socio-economic
COLOR PLACE OF SCORES IN Interval
BIRTH EXAMS Temp., IQ test
YEAR BLOOD TYPE GRADES OF Ratio
LEVEL STUDENT H/W, income
Hypothesis Testing
What is a Hypothesis
● It is an assumption or conjecture about a
population parameter which may or may not be
true.
● Two-types: Null and Alternative
● One-tailed (Biased)
● Uses words like “greater than” or “less than”
● Two-tailed (Unbiased)
● Uses words like “unequal”, “differ from” or
“not/none”
Hypothesis
● Null hypothesis: A hypothesis put forward to argue that a
relationship or pattern does not exist.
○ Cholesterol study example: In a Randomized Control Trial, the
control group and the treatment group have equal levels of
cholesterol at the end of a study.
■ Null hypothesis: Groups A and B are equal.
○ Denoted by Ho:
The null hypothesis will be true if the findings are insignificant.
The null hypothesis will be false if the findings are significant.
Hypothesis
● Alternative Hypothesis: Statement of what study is set
to establish.
○ Alternative Hypothesis: Groups A and B have different levels
of cholesterol.
○ Denoted by H1:
The null hypothesis will be false if the findings are significant.
Hypothesis Testing
● Is a decision-making process for evaluating
claims about a population
● It is basically testing an assumption that we
can make about a population
Level of Significance
● Significance is defined as the quality of being
statistically significant. Usually serves as basis for the
decision about the hypothesis.
● It is denoted by alpha or α refers to the degree of
significance in which we accept or reject the null
hypothesis.
● In public health research, alpha is usually 0.01 or 1%.
In social science, alpha is usually 0.05 or 5% and 0.1 or
10% in other studies.
What Statistical
Treatment to use?
Types of Statistical Test
• Used with interval and ratio data and usually with data
that were obtained from groups randomly assigned,
Parametric normally distributed, and equal variability between group.
Statistics
• Used with nominal and ordinal data and sometimes with
interval and ratio data when other assumptions cannot be
Nonparametric met.
Statistics
Ask yourself the ff. questions
● 1. What type of data do I have?
● 2. How many samples do I have?
● 3. What do I want to determine?
What type of data do you have?
Categorical/Nominal- Continuous/Interval-
Ordinal/ Qualitative Ratio/ Quantitative
Gender Age
Year Level Grades
Number of Family
Civil Status
Members
Yes/No Responses Rating Scale
How many samples do you have?
● One Sample Two Samples (Special Case)
○ Determining the perception of a group ○ One sample, but different
measurements made
Comparing a group against
○
information from outside source ○ Two samples, but each of them is
● Two Samples being compared to its correspondence
○ Control group vs. test group in an ■ Grade level Teachers and their
experiment students
○ Comparing one group to the other ■ Parents and their child
group ■ Wife and their husbands
● Three or More Samples
○ Comparing the three groups to each
other
What do you want to determine?
Comparison (Difference) Relationship (Connection)
Male vs. Female Students Score in math and
science
Control vs. Experimental Service vs. Satisfaction
Grouped by categories Medication vs. Recovery time
Three Main Families of Statistical Test
Chi-square T-test Correlation
Categorical and/or Continuous
Categorical Data Continuous Used for relationship
Used for comparison Used for comparison
Examples:
Examples: Examples: Pearson
Test of homogeneity 1 sample t-test Spearman (If having ordinal
Test of Independence 2 sample t-test data)
Analysis of Variance Regression
Purpose Comparison Relationship
Number of Samples Chi-square T-test Correlation
One Sample 1 sample t-test
Two Samples 2 sample t-test 1 IV and 1 DV
Any number of levels or
samples Pearson’s Correlation
2 sample t-test Regression
Two Samples Special
Paired sample t-test
Three or More Samples One-way ANOVA
Non-parametric Counterparts
Purpose Comparison Relationship
Number of Samples Chi-square T-test Correlation
1 sample Wilcoxon
One Sample
signed rank test
Two Samples 1 IV and 1 DV
Any number of levels or Wilcoxon
samples Mann Whitney Spearman’s Correlation
Two Samples Special Regression
Three or More Samples Kruskal-Wallis Test
Purpose of statistical
Start Comparing Groups Testing for Relationship
Test
Dependent Samples Independent Samples Parametric? No
Yes Yes Yes
Chi-square
Parametric? Parametric?
Pearson r
Yes Yes
No
T-test for dependent Z-test for two sample Spearman Rho
means means
No Mann-Whitney U test
T-test for independent
means
Wilcoxon signed rank test
ANOVA (Analysis of
Kruskal Wallis
Variance)
STATISTICAL TREATMENT ROADMAP
Try this out!
● Identify which statistical treatment is appropriate.
● 1. Difference in the board rating of teacher when grouped according to sex.
● A. t-test for independent means/samples
● B. z-test
● C. t-test for dependent means/samples
● D. ANOVA
Try this out!
● Identify which statistical treatment is appropriate.
● 2. Relationship between the type of school and job performance. Job
performance is rated as poor, fair, good, very good, excellent.
● A. Pearson r
● B. Chi-square
● C. Spearman rho
Try this out!
● Identify which statistical treatment is appropriate.
● 3. Difference between the pre-test and post test performance of the single class.
● A. t-test for independent means/samples
● B. z-test
● C. t-test for dependent means/samples
● D. ANOVA
Try this out!
● Identify which statistical treatment is appropriate.
● 4. Relationship between teachers’ board rating and job performance rating.
● A. Pearson r
● B. Chi-square
● C. Spearman rho
Try this out!
● Identify which statistical treatment is appropriate.
● 5. Difference among the performances of students in online, modular and
blended approach.
● A. t-test for independent means/samples
● B. z-test
● C. t-test for dependent means/samples
● D. ANOVA