Parametric tests: t-test,
F-test, Karl pearson
coefficient of
correlation
MODEL 5 - UNIT 2
Presentation by Nandhana A
msy2518
Introduction
Parametric tests are a branch of inferential statistics that assume the sample
data comes from a population that follows a specific distribution, usually a
normal distribution. These tests are generally more powerful than non-
parametric tests, meaning they are more likely to detect a significant effect if
one truly exists.
Assumptions:
Normality: Data should be normally distributed in a bell-shaped curve.
Homogeneity of Variance: The variance within the groups being compared
should be approximately equal.
Independence: Observations must be independent of each other.
Interval or Ratio Scale: Data should be measured on a continuous scale.
(e.g, scores, time, age etc).
These tests offer higher statistical power to detect true effects in fields like
clinical trials or psychological interventions (Field, 2018).
t-test
Developed by William Sealy Gosset, the t-test is used to compare
the means of two groups. A t-score measures the size of the
difference between a sample mean and the population mean
relative to the variation in the sample data. It is especially useful
when the sample size is small (n<30) and the population standard
deviation is unknown.
One sample t-test:
Compares the mean of a single group to a known or hypothesized
population mean. For e.g, a manufacturer claims energy bars have
20g of protein so you test 31 bars (sample) to see if their average
protein content is actually 20g.
Two sample t-test (Independent sample t-test):
Compares the means of two unrelated groups. For e.g, a study
Comparing the cognitive test scores of individuals with Chronic
Liver Disease (CLD) against a healthy control group to identify
Minimal Hepatic Encephalopathy (Grover & Sarkar, 2012).
Paired t-test (Dependent sample t-test):
Compares means from the same group at different times. For e.g,
a study evaluating a 12-week Mindfulness-Based Stress Reduction
(MBSR) program measures participants' cortisol levels at baseline
and again at 12 weeks. A paired-samples t-test is used to see if the
intervention significantly lowered stress hormones in the same
individuals.
F-test (Analysis of Variance
- ANOVA)
The F-test was developed by the British statistician and geneticist
Sir Ronald A. Fisher (R.A. Fisher) in the 1920s. While a t-test compares two groups,
the F-test is used to compare the means of three or more groups simultaneously,
controlling the Type-I error (false positives). It determines whether the between-
group variance is significantly larger than the within-group variance.
One-Way ANOVA: Testing the effect of one independent variable (e.g., three
different types of therapy) on a dependent variable.
Two-Way ANOVA: Testing the effect of two independent variables (e.g.,
therapy type and dosage) and their interaction.
Equality of Variances: The F-test is also used specifically to check if two
populations have equal variances before performing other tests.
A researcher investigates three different therapeutic approaches for Alcohol Use
Disorder: Motivational Interviewing (MI), Cognitive Behavioral Therapy (CBT), and
a Control Group. A One-Way ANOVA is used to determine if at least one therapy is
more effective in increasing days of abstinence (Sollenberger & Chang, 2025). If
additionally, another independent variable of Gender (Males and Females) is
introduced, Two-way ANOVA is used to see the effect.
Key advantage of using two-way anova is Interaction Effect which reveals if the
effect of one factor changes depending on the level of the other factor.
The F-ratio Logic:
F=Variance Between Treatments/ Variance Within Treatments
If F>1 significantly, it suggests the treatment had a real effect.
Post Hoc Tests: Gravetter and Forzano emphasize that if a significant F is
found for 3+ groups, post hoc tests are to be performed (like Tukey’s HSD) to
determine exactly which pairs of means are different.
TYPES OF ANOVA
Test Explanation Simple Example Key Differentiating Point
Compares the means of 3+ Comparing the test scores of
Only one factor is changed; groups
One-Way ANOVA independent groups on one students using 3 different study
are independent.
outcome. methods (A, B, or C).
Examines the effect of two factors
Comparing test scores based on Tests for an Interaction Effect (e.g.,
Two-Way ANOVA simultaneously and how they
Study Method and Gender. Method A works better for girls).
interact.
Compares means from the same Measuring a patient’s anxiety Uses the same people in every
Repeated Measures ANOVA participants across different before, during, and after a 6-month group; reduces individual
conditions or time. therapy. difference "noise."
An ANOVA that controls for a third Comparing 3 diets on weight loss,
Statistically removes the influence
ANCOVA variable (Covariate) that might bias but controlling for the participants'
of an unwanted variable.
results. starting weight.
Testing a new drug’s effect on both
Used when you have 2 or more Protects against false positives
MANOVA Heart Rate and Blood Pressure at
related Dependent Variables. when you have multiple outcomes.
once.
Testing a drug on Heart Rate and The most complex: handles
A MANOVA that also controls for a
MANCOVA BP, while controlling for the multiple outcomes and controls for
Covariate.
patient’s Age. noise.
TYPES OF POST HOC
TESTS
Test Name Use Case
The gold standard. Used when group sizes are equal and
Tukey’s HSD
variances are similar.
Very conservative and strict. Great for a small number of
Bonferroni
comparisons.
The most flexible but least powerful. Often used for
Scheffé
complex, unplanned comparisons.
Used when the assumption of Homogeneity of Variance is
Games-Howell
violated (groups have different spreads).
Karl pearson coeffecient of
correlation
The correlational research strategy is used to describe the relationship between
two variables as they exist naturally, without attempting to explain cause and
effect. This statistical technique is used in such studies. It is also known as the
Pearson Product-Moment Correlation and this test measures the strength and
direction of the linear relationship between two continuous variables. It is
represented by r.
Interpretation of (r):
The value of r always ranges from -1.0 to +1.0.
+1.0: Perfect positive correlation (as X increases, Y increases).
0: No linear relationship.
-1.0: Perfect negative correlation (as X increases, Y decreases).
Assumptions Specific to Pearson’s r:
Linearity: The relationship between variables must be a straight line.
Homoscedasticity: The scatter of the data points should be roughly the same
across all values of the variables.
Example: Investigating the link between psychological resilience and quality of
life (QoL) in liver transplant recipients. Research by Meşe and Harmancı (2025)
suggests a strong positive correlation (high r value), indicating that as a patient's
awareness of a healthy life increases, their psychological well-being also
improves.
Conclusion
Parametric tests like the t-test, F-test, and Pearson’s r provide a systematic,
objective method for researchers to determine if observed patterns represent
real effects or simple chance (Gravetter & Forzano, 2012). By evaluating mean
differences and variable relationships, these tools allow behavioral scientists to
move beyond mere description toward meaningful statistical inference.
Decision Tree:
2 groups? t-test.
3+ groups? F-test (ANOVA).
Relationship between 2 variables? Pearson’s r
References
Field, A. (2018). Discovering statistics using IBM SPSS statistics (5th ed.). SAGE Publications.
Gravetter, F. J., & Forzano, L. B. (2012). Research methods for the behavioral sciences (4th
ed.). Wadsworth, Cengage Learning.
Grover, S., & Sarkar, S. (2012). Liver transplant—Psychiatric and psychosocial aspects.
Journal of Clinical and Experimental Hepatology, 2(4), 382–392.
[Link]
Meşe, S., & Harmancı, P. (2025). Examining the relationship between healthy life awareness
and psychological well-being in liver transplantation patients: A structural equation model.
Nursing & Health Sciences, 27. [Link]
Romeo, M., Dallio, M., Cipullo, M., Coppola, A., Mazzarella, C., Mammone, S., Iadanza, G.,
Napolitano, C., Vaia, P., Ventriglia, L., & Federico, A. (2025). Nutritional and psychological
support as a multidisciplinary coordinated approach in the management of chronic liver
disease: A scoping review. Nutrition Reviews, 83(7), 1327.
[Link]
Sollenberger, C. H., & Chang, H. R. (2025). Psychosocial therapy improves outcomes for
alcohol use disorder in people with chronic liver disease: A systematic review and meta-
analysis. Psychosomatic Medicine.
Thank You
So Much!