Inferential
Statistics
What is inferential
statistics?
Inferential statistics is the branch of statistics that uses data collected
from a sample to make inferences, generalizations, and predictions
about a larger population from which the sample was drawn.
It uses probability theory and statistical models to estimate population
parameters and test population hypotheses based on sample data.
[Link]
Inferential statistics is more subjective as it involves generalizing and
estimates about the population using the sample.
Hypothesis
testing
Significanc
Levels
e
Hypothesis • Null Hypothesis (H0): This is a
statement of no effect, no
testing difference, or no relationship. It
represents the status quo or a
Is a fundamental statistical
commonly accepted belief.
method used to make
inferences about a population
based on sample data. It • Alternative Hypothesis (H1 or Ha):
involves formulating two This is the statement that you are
competing hypotheses about trying to find evidence for. It
a population parameter: the contradicts the null hypothesis.
null and the alternative
hypothesis.
Siginificance
Levels
The significance level—denoted by alpha (α)—is the threshold a researcher
sets to determine whether an observed result is statistically significant. It
represents the probability of rejecting the null hypothesis when it is actually
true, also known as a Type I error.
The significance level helps you decide how strong your evidence must be
before you can conclude that a difference or effect exists in your study.
Common
Significance
Levels
• Interpretation: If you set α=0.05, it
means you are willing to accept a 5%
chance of incorrectly rejecting the null
hypothesis.
• Impact on Decision: A smaller α makes
it harder to reject the null hypothesis,
requiring stronger evidence.
Conversely, a larger α makes it easier
to reject the null hypothesis, but
increases the risk of a Type I error.
• Convention: The choice of α is often
based on the field of study and the
potential consequences of making a
Type I error. For many research areas,
0.05 is a common default.
T-test
(Independent and
Paired)
What is t-test?
A t-test is a type of inferential statistical test used to determine
whether there is a significant difference between the means
(averages) of two groups or two conditions.
It helps researchers make decisions about whether observed
differences in data could be due to chance or reflect a true effect in
the population.
Types of T-
tests
Independent T-test Formula:
• t = Student's t-test
• x1 = mean of first group
• x2= mean of second group
• s1 = standard deviation of group 1
• s2 = standard deviation of group 2
• n1= number of observations in group
1
• n2= number of observations in group
2
Research Independent
Hypotheses Study Design Interpretation
Question Samples T-Test
(H0): There is no An independent
significant difference in A researcher identifies samples t-test is
performed to compare The researcher
the mean literary two distinct groups of
the mean literary would conclude that
analysis scores between English literature
Is there a analysis score of there is a
students who primarily students:
significant Group 1 with the mean statistically
read print books and
difference in the literary analysis score significant
students who primarily Group 1: 40 students
literary analysis of Group 2. difference in literary
read e-books (μprint who report primarily
analysis scores
scores between =μe-book). reading print books.
The t-test calculates a between students
English literature
t-statistic and a p- who primarily read
students who (H1): There is a Group 2: 40 students
value. print books and
primarily read print significant difference in who report primarily
those who primarily
books versus those the mean literary reading e-books. If, for example, the p- read e-books.
analysis scores between
who primarily read value is 0.03 (and the Further examination
students who primarily Both groups take the significance level α
e-books? of the means would
read print books and same standardized was set at 0.05), the reveal which group
students who primarily literary analysis researcher would scored higher.
read e-books (μprint examination. reject the null
=μe-book). hypothesis.
Research Independent
Hypotheses Study Design Interpretation
Question Samples T-Test
If your significance
Two classes of Grade level is α = 0.05:
11 students:
Since 0.02 < 0.05,
Group A: Taught You compute the mean you reject H₀
(H0): There is no vocabulary score for
Does using poetry vocabulary using
difference in vocabulary each group.
poetry. There is a
in English class scores between the two
statistically
improve students' groups. Run a t-test to
Group B: Taught the significant
vocabulary more compare the means.
same vocabulary difference between
than not using (H1): There is a
using textbook drills. the two teaching
significant difference in Suppose the p-value =
poetry? methods.
vocabulary scores. 0.02
After 4 weeks, both
groups take the same You conclude that
vocabulary test. using poetry leads
to better vocabulary
acquisition.
Paired T-test Formula:
• d_bar: is the mean difference between the
paired samples.
• s_d: is the standard deviation of the
differences.
• n: is the number of pairs.
Research Paired Samples T- Interpretatio
Hypotheses Study Design
Question Test n
(H0): There is no A researcher recruits 30 For each student, the
significant change university students researcher calculates the
in students' mean enrolling in a creative difference between their
perceived writing writing workshop. "after" score and their
Does "before" score. The researcher
confidence scores
participation in after participating Before the workshop would conclude
a semester-long A paired samples t-test is that participation
in a creative writing begins, each student
creative writing performed on these in the creative
workshop (μafter completes a survey
differences. Essentially, it
workshop =μbefore). measuring their writing workshop
checks if the mean of these
improve perceived writing significantly
differences is significantly
students' (H1): Students' confidence (e.g., on a 1- improved
different from zero.
perceived mean perceived 10 scale). students'
writing writing confidence The t-test calculates a t- perceived writing
scores significantly After the workshop
confidence statistic and a p-value. confidence
increase after concludes, the same 30
scores? scores.
participating in a students complete the If, for example, the p-value is
creative writing identical writing 0.001 (and α was 0.05), the
workshop (μafter confidence survey researcher would reject the
>μbefore). again. null hypothesis.
Research Paired Samples
Hypotheses Study Design Interpretation
Question T-Test
Participants: 25 senior
high school students Collect two sets of
(H0): There is no scores: Pre-test and
Post-test Since p = 0.01 <
difference in Pre-test: Vocabulary
0.05, the difference
vocabulary scores test administered
Use statistical software is statistically
before and after before the poetry
(SPSS, Excel, R, etc.) to significant.
Does reading the poetry reading sessions
run the paired samples
poetry for four intervention.
t-test The poetry reading
weeks improve Intervention: Students
sessions
students’ (H1): There is a read and discuss 10
Example result: significantly
vocabulary skills? significant poems over 4 weeks
Mean pre-test score: improved the
improvement in 65 students’
vocabulary scores Post-test: Vocabulary Mean post-test score: vocabulary
after the test administered 75 performance.
intervention. again after the p-value: 0.01
intervention
ANNOVA
Analysis of Variance
MANOVA
Multivariate Analysis of
Variance
ANOVA
ANOVA, or Analysis of Variance, is a statistical test used to determine if there
are any statistically significant differences between the means of three or
more independent (unrelated) groups on a single continuous dependent
variable.
Types of
ANOVA
• One-way ANOVA is a statistical test
used to determine if there are
significant differences between the
means of three or more groups of
a single independent variable.
• Two-way ANOVA is used to
examine the effect of two
independent variables on a
dependent variable and also to
assess the interaction between
those two variables.
Example
Does the learning environment (traditional classroom, online
Research Question self-paced, or blended learning) affect university students' final
scores on a grammar proficiency exam?
• Independent Variable (Factor): Learning Environment
(Categorical, with 3 levels: Traditional Classroom, Online Self-
Variables Paced, Blended Learning)
• Dependent Variable: Grammar Proficiency Exam Score
(Continuous, e.g., percentage score)
ANOVA One-Way ANOVA.
120 students
40 students - Traditional Classroom
40 students - Online Self-Paced
= Grammar Proficiency Exam Score
40 students - Blended Learning
Example
Do different English language teaching methods (e.g.,
Communicative Language Teaching, Task-Based Language
Research Question
Teaching, Grammar-Translation Method) lead to significant
differences in the oral fluency scores of advanced ESL learners?
• Independent Variable: Teaching Method (Categorical, 3
levels: Communicative, Task-Based, Grammar-Translation)
Variables
• Dependent Variable: Oral Fluency Score (Continuous, e.g.,
assessed by a standardized speaking rubric from 0-100).
ANOVA One-Way ANOVA.
Example
Do the narrative perspective of a short story (e.g., First-Person
Research Question vs. Third-Person Limited) and the main character's gender (Male
vs. Female) affect readers' self-reported empathy scores?
• Independent Variable 1: Narrative Perspective (Categorical, 2
levels: First-Person, Third-Person Limited)
• Independent Variable 2: Character Gender (Categorical, 2
Variables
levels: Male, Female)
• Dependent Variable: Empathy Score (Continuous, e.g., score
on a validated empathy questionnaire)
ANOVA Two-Way ANOVA.
Example
Do the teaching approach (e.g., Explicit Strategy Instruction vs.
Implicit Strategy Exposure) and the student's initial motivation
Research Question
level (e.g., Low vs. High) affect their reported use of reading
comprehension strategies?
• Independent Variable 1: Teaching Approach (Categorical, 2
levels: Explicit Instruction, Implicit Exposure)
• Independent Variable 2: Student Motivation (Categorical, 2
Variables
levels: Low, High)
• Dependent Variable: Reading Comprehension Strategy Use
Score (Continuous, e.g., score on a self-report questionnaire).
ANOVA Two-Way ANOVA.
MANOVA
MANOVA, or Multivariate Analysis of Variance, is an extension of
ANOVA. It is used when you want to compare the means of two or
more independent groups on two or more related continuous
dependent variables simultaneously
Example
Do different types of AI-powered writing tools (e.g., Grammar
Checker, Style Suggestion, or Content Generation Aid) affect
Research Question
university students' grammatical accuracy, stylistic sophistication,
and argument coherence in their academic essays?
• Independent Variable: Type of AI Writing Tool (3 levels: Grammar
Checker, Style Suggestion, Content Generation Aid).
• Dependent Variables:
⚬ Grammatical Accuracy Score (continuous, e.g., errors per 100
Variables words)
⚬ Stylistic Sophistication Score (continuous, e.g., rated by expert
rubric)
⚬ Argument Coherence Score (continuous, e.g., rated by expert
rubric)
MANOVA MANOVA
Example
Do specific components of oral presentation training (e.g., focusing
on Voice Modulation, Body Language, or Content Delivery) collectively
Research Question
affect L2 speakers' overall speaking performance (e.g., rubric score)
and their public speaking anxiety levels?
• Independent Variable: Presentation Training Component (3 levels:
Voice Modulation Focus, Body Language Focus, Content Delivery
Focus).
• Dependent Variables:
Variables
⚬ Overall Speaking Performance Score (continuous, e.g., rubric
rating by judges)
⚬ Public Speaking Anxiety Score (continuous, e.g., self-report
anxiety scale)
MANOVA MANOVA
Chi-square
test
CHI-SQUARE TEST
The Chi-Square (χ2) test is a non-parametric statistical test used to
examine the relationship between categorical variables. Unlike ANOVA
and MANOVA, which deal with continuous dependent variables, the
Chi-Square test is designed for situations where you have frequency
counts or proportions of observations falling into different categories
The Chi-Square Where:
• O = Observed frequency (actual count in a
statistic is cell)
calculated • E = Expected frequency (count you would
expect if there were no association)
using the
• A large χ2 value (and a small p-value)
formula: indicates a significant difference between
observed and expected frequencies,
leading to the rejection of the null
hypothesis of independence.
• Chi-Square Goodness-of-Fit Test: Used to
determine if an observed frequency
distribution for a single categorical Types of
variable differs significantly from a
hypothesized or expected distribution. Chi-Square
• Chi-Square Test of Independence: This is Test
the more commonly used type. It
determines whether there is a statistically
significant association between two (or
more) categorical variables. The null
hypothesis is that the variables are
independent (not related), and the
alternative hypothesis is that they are
dependent (related)
• Categorical variables are those that place Identifying
individuals or items into distinct groups or
categories. They don't have a numerical
Variables
meaning that allows for arithmetic
operations (like averaging).
for Chi-Square
⚬ Nominal variables: Categories with no
Test
intrinsic order (e.g., Gender: Male,
Female; Language: English, Spanish).
⚬ Ordinal variables: Categories with a
meaningful order, but uneven or
unknown intervals between them (e.g.,
Satisfaction: Low, Medium, High; Grade
Level: Freshman, Sophomore, Junior).
Example
Is there a significant association between a student's gender
Research Question and their preferred learning style for acquiring new English
vocabulary (e.g., Visual, Auditory, Kinesthetic)?
• Independent Variable: Gender (Categorical: Male, Female,
Non-binary)
Variables
• Dependent Variable: Preferred Vocabulary Learning Style
(Categorical: Visual, Auditory, Kinesthetic)
Example
Is there a significant association between the instructional
method used to teach literary analysis (e.g., traditional
lecture, collaborative group work, multimodal approach) and
Research Question
students' reported level of engagement with the literary text
(e.g., highly engaged, moderately engaged, minimally
engaged)?
• Independent Variable: Instructional Method (Categorical:
Traditional Lecture, Collaborative Group Work, Multimodal
Approach)
Variables
• Dependent Variable: Student Engagement Level
(Categorical: Highly Engaged, Moderately Engaged,
Minimally Engaged)
References:
[Link]
[Link]
[Link]
[Link]
[Link]
[Link]
[Link]
Thank You!