HYPOTHESIS TESTING

HYPOTHESIS TESTING

Hypothesis Testing is a statistical method used to make decisions or

inferences about a population based on a sample of data.

• Inmidwifery, we may use hypothesis testing to determine whether an

observed difference in outcomes (e.g., birth weights between two groups
of mothers) is statistically significant or due to chance.
 KEY TERMINOLOGIES

Term Definition
A statement or assumption about a
Hypothesis
population parameter
The default claim—no effect, no
Null Hypothesis (H₀)
difference, or status quo
The claim we test for—an effect or a
Alternative Hypothesis (H₁)
difference
 The probability of rejecting H₀ when
Significance Level (α) it's actually true (Type I error),
commonly set at 0.05
The probability of observing the data
p-value
(or more extreme) if H₀ is true
A value calculated from sample data
Test Statistic
used to decide whether to reject H₀
The range of values where H₀ is
Critical Region
rejected
The probability that the true population
Confidence Level parameter lies within a confidence
interval (e.g., 95%)
STEPS IN HYPOTHESIS TESTING

1. State the Hypotheses

o Null hypothesis (H₀)
o Alternative hypothesis (H₁)
2. Choose the Significance Level (α)
o Common values: 0.05, 0.01
3. Select the Appropriate Test
o Depends on data type and study design (e.g., t-test, chi-square)
4. Calculate the Test Statistic
o Use formulas or statistical software (e.g., SPSS, R)

5. Find the p-value or Compare to Critical Value

6. Make a Decision
o If p-value ≤ α, reject H₀ (significant)
o If p-value > α, fail to reject H₀ (not significant)

7. State the Conclusion in Context

WHAT IS A HYPOTHESIS?

A hypothesis is a statement about a population parameter, such as:

• A mean (e.g., average blood pressure)
• A proportion (e.g., % of smokers in a population)
• A difference between groups (e.g., treatment vs. control)
Null Hypothesis (H₀)

•The default or status quo assumption.

•States no effect, no difference, or no association.
•It is what we assume to be true unless evidence
suggests otherwise.
Think: "Nothing is happening" or "No difference
exists."
Alternative Hypothesis (H₁ or Ha)

•The research or investigative claim.

•Suggests there is an effect, a difference, or a
relationship.
•What we are trying to provide evidence for.
Think: "There is something happening."
STEPS IN FORMULATING HYPOTHESES

•Identify the research question.

•Determine what parameter is being tested (mean, proportion, etc.).
•State the null hypothesis (H₀: no effect or difference).
•State the alternative hypothesis (H₁: what the researcher expects to
find).
•Choose the direction of the test:
•Two-tailed: if you're testing for any difference.
•One-tailed: if you're testing for a specific direction (e.g., increase or
decrease).
EXAMPLE 1: NEW DRUG EFFICACY

Scenario: A pharmaceutical company claims that a new drug lowers blood

pressure more than the current standard medication.
• Parameter: Mean blood pressure
• H₀: The new drug has the same effect as the standard (μ_new = μ_standard)
• H₁: The new drug has a greater effect (μ_new < μ_standard)
EXAMPLE 2: SMOKING CESSATION PROGRAM

Scenario: A public health agency wants to test if a new counseling program

increases the quit rate among smokers.
• Parameter: Proportion of smokers who quit (p)
• H₀: The quit rate is the same as before (p = 0.25)
• H₁: The quit rate is higher with the new program (p > 0.25)
EXAMPLE 3: DIET AND WEIGHT LOSS

Scenario: A dietitian claims that a new low-carb diet results in more than 5 kg
average weight loss in 3 months.
• Parameter: Mean weight loss (μ)
• H₀: μ = 5 kg
• H₁: μ > 5 kg
EXAMPLE 4: HOSPITAL WAIT TIMES

Scenario: A hospital administrator believes that a new check-in system has

reduced patient wait times.
• Parameter: Mean wait time ( 30 minutes)
• H₀: μ = 30 minutes
• H₁: μ < 30 minutes
EXAMPLE 5: VACCINE EFFECTIVENESS

• Scenario: Researchers are testing whether a new vaccine is more effective

than the current one.
Ho:?
• Ha:?
EXAMPLE 6: PHYSICAL THERAPY INTERVENTION

Scenario: A physical therapist wants to know if a new rehabilitation technique

improves mobility scores.
• Parameter: Mean mobility score
• Ho=?
• Ha=?
EXAMPLE 7: INFECTION RATE IN TWO WARDS

Scenario: You want to test whether Ward A and Ward B have the same
hospital-acquired infection rate.
• Parameter: Proportion of infected patients
• Ho
• Ha
STATISTICAL TESTING

Statistical testing helps researchers determine whether the findings in their study
are likely due to chance or represent a real effect.
Why it's important:
• Supports evidence-based midwifery practice.
• Helps make decisions in maternal and newborn care.
• Ensures validity and reliability of research findings.
TYPES OF DATA

Understanding your data type is crucial.

A. Categorical (Qualitative) Data
• Nominal: Categories without order (e.g., blood type, religion, type of
delivery).
• Ordinal: Categories with a meaningful order but no consistent interval (e.g.,
pain scale, stages of labor).
B. Numerical (Quantitative) Data
• Discrete: Countable values (e.g., number of ANC visits).
• Continuous: Measurable values with decimals (e.g., birth weight, gestational
age).
KEY QUESTIONS BEFORE CHOOSING A TEST

1.What is your research question?

2.What type of variables are involved?
3.How many groups are being compared?
4.Are the samples related or independent?
5.Is the data normally distributed?
PARAMETRIC VS NON-PARAMETRIC TESTS
Feature Parametric Tests Non-Parametric Tests
Continuous, normally Categorical or non-
Data Type
distributed normal data
Chi-square, Mann-
Examples t-test, ANOVA
Whitney U
Yes (if assumptions Less powerful but
More powerful?
met) safer
CHOOSING A PARAMETRIC TEST: REGRESSION, COMPARISON, OR CORRELATION

Parametric tests usually have stricter requirements than nonparametric tests,

and are able to make stronger inferences from the data. They can only be
conducted with data that adheres to the common assumptions of statistical
tests.

• Themost common types of parametric test include regression tests,

comparison tests, and correlation tests.
REGRESSION ANALYSIS

• To examine relationships to predict outcome, understand association and

adjust confounding variables.
 Predictor variable Outcome variable Research question
example
Simple linear regression • Continuous • Continuous What is the effect
• 1 predictor • 1 outcome of income on longevity?

Multiple linear regression • Continuous • Continuous What is the effect

• 2 or more predictors • 1 outcome of income and minutes of
exercise per
day on longevity?
Logistic regression • Continuous • Binary What is the effect of drug
dosage on the survival of a
test subject?
 COMPARISON TESTS

Comparison tests look for differences among group means. They can be used
to test the effect of a categorical variable on the mean value of some other
characteristic.

• T-tests are used when comparing the means of precisely two groups (e.g.,
the average heights of men and women). ANOVA and MANOVA tests are
used when comparing the means of more than two groups (e.g., the average
heights of children, teenagers, and adults).
 Predictor variable Outcome variable Research question
example
Paired t-test • Categorical • Quantitative What is the effect
• 1 predictor • groups come from of two different test
the same prep programs on
population the average exam
scores for students from
the same class?

Independent t-test • Categorical • Quantitative What is the difference

• 1 predictor • groups come from in average exam
different scores for students
populations from two different
schools?
ANOVA • Categorical • Quantitative What is the difference
• 1 or more predictor • 1 outcome in average pain
levels among post-
surgical patients
given three different
painkillers?

MANOVA • Categorical • Quantitative What is the effect

• 1 or more predictor • 2 or more outcome of flower species on petal
length, petal width,
and stem length?
CORRELATION TESTS

Correlation tests check whether variables are related without hypothesizing a

cause-and-effect relationship.

• These can be used to test whether two variables you want to use in (for
example) a multiple regression test are autocorrelated.
 Variables Research question
example
Pearson’s r • 2 continuous How
variables are latitude and temp
erature related?
CHOOSING A NONPARAMETRIC TEST

• Non-parametric tests don’t make as many assumptions about the data, and
are useful when one or more of the common statistical assumptions are
violated. However, the inferences they make aren’t as strong as with
parametric tests.
 Predictor Outcome Use in place
variable variable of…
Spearman’s r • Quantitative Quantitative
• Pearson’s r
Chi square test Categorical
• Categorical
• Pearson’s r
of
independence
Sign test • Categorical Quantitative
• One-sample t-
test
Kruskal– • Categorical • Quantitative ANOVA
Wallis H • 3 or more
groups
ANOSIM • Categorical • Quantitative MANOVA
• 3 or more groups • 2 or more outcome
variables

Wilcoxon Rank-Sum test • Categorical • Quantitative Independent t-test

• 2 groups • groups come from
different populations

Wilcoxon Signed-rank test • Categorical • Quantitative Paired t-test

• 2 groups • groups come from the
same population
Choosing the right statistical test