MEV 019: Unit 12 - Analysis of Variance Tests
UNIT 12: ANALYSIS OF VARIANCE TESTS
12.1
Introduction
When comparing two groups, t-tests are appropriate.
But when comparing three or more groups, performing multiple t-tests
increases the risk of committing Type I error.
Analysis of Variance (ANOVA) overcomes this problem by testing for
differences among several group means simultaneously using variance
estimates.
Developed by R.A. Fisher, ANOVA is one of
the most widely used statistical tools in scientific research, including
agriculture, medicine, social sciences, and environmental studies.
12.2
Objectives
After studying this unit, you will be able to:
- Explain
the concept and purpose of ANOVA.
- Understand
degrees of freedom in ANOVA.
- State
the significance and uses of ANOVA.
- Perform
one-way ANOVA and two-way ANOVA.
- Recognize
the assumptions underlying ANOVA.
12.3
Analysis of Variance (ANOVA)
ANOVA is a statistical method used to test whether
the means of k groups are equal.
It partitions the total variability in data into components:
- Between-group
variability – due to differences in group means.
- Within-group
variability – due to random error or variation within
groups.
The F-test is used to compare these
variances.
12.3.1 Significance of Analysis of Variance
- Determines
if observed differences between sample means are statistically
significant.
- Avoids
multiple t-tests and controls the overall error rate.
- Can
handle multiple factors simultaneously.
12.3.2 Degrees of Freedom (df)
In ANOVA:
- Between-groups
df = k−1k - 1k−1
- Within-groups
df = N−kN - kN−k
- Total
df = N−1N - 1N−1
Where: - kkk =
number of groups
- NNN =
total number of observations
12.3.3 Uses of ANOVA
- Comparing
mean yields of different crop varieties.
- Assessing
effects of different treatments in experiments.
- Studying
variation due to different environmental conditions.
- Evaluating
multiple teaching methods in education research.
12.4 One-way
Analysis of Variance (ANOVA)
In one-way ANOVA, there is one factor
(independent variable) with two or more levels (groups).
We test whether the means of these groups are equal.
12.4.1 Basic Assumptions of One-Way ANOVA
- The
populations are normally distributed.
- The
populations have equal variances (homogeneity of variance).
- Samples
are independent.
12.4.2 Test of Hypothesis in One-Way ANOVA
Hypotheses:
- H0H_0H0:
All group means are equal (μ1=μ2=⋯=μk\mu_1
= \mu_2 = \dots = \mu_kμ1=μ2=⋯=μk)
- H1H_1H1:
At least one mean is different.
Steps:
- Calculate
group means and overall mean.
- Compute
Sum of Squares Between (SSB) and Sum of Squares Within
(SSW):
SSB=∑i=1kni(Xˉi−Xˉ)2SSB = \sum_{i=1}^{k} n_i (\bar{X}_i -
\bar{X})^2SSB=i=1∑kni(Xˉi−Xˉ)2 SSW=∑i=1k∑j=1ni(Xij−Xˉi)2SSW = \sum_{i=1}^{k}
\sum_{j=1}^{n_i} (X_{ij} - \bar{X}_i)^2SSW=i=1∑kj=1∑ni(Xij−Xˉi)2
- Find Mean
Squares:
MSB=SSBk−1,MSW=SSWN−kMSB = \frac{SSB}{k-1}, \quad MSW =
\frac{SSW}{N-k}MSB=k−1SSB,MSW=N−kSSW
- Compute
F-ratio:
F=MSBMSWF = \frac{MSB}{MSW}F=MSWMSB
- Compare
F with the critical value from F-distribution table (df₁ = k – 1,
df₂ = N – k).
- Draw
conclusion: Reject H0H_0H0 if calculated F > critical F.
12.5 Two-way
Analysis of Variance (ANOVA)
Two-way ANOVA is used when there are two
factors and we want to study:
- The
effect of each factor individually (main effects).
- The
combined effect of both factors (interaction effect).
Example: Studying crop yield influenced by fertilizer
type (Factor A) and irrigation level (Factor B).
12.5.1 Basic Assumptions of Two-way ANOVA
- Each
cell (combination of factors) is normally distributed.
- Equal
variances across all cells.
- Independence
of observations.
12.5.2 Test of Hypothesis in Two-way ANOVA
Hypotheses:
- H0AH_{0A}H0A:
No difference between levels of Factor A.
- H0BH_{0B}H0B:
No difference between levels of Factor B.
- H0ABH_{0AB}H0AB:
No interaction between Factor A and B.
Steps:
- Arrange
data in a two-factor table.
- Calculate
Sum of Squares for:
- Factor A (SSA)
- Factor B (SSB)
- Interaction (SSAB)
- Error/Within (SSE)
- Find
respective Mean Squares by dividing SS by df.
- Compute
F-ratios for each source:
FA=MSAMSE,FB=MSBMSE,FAB=MSABMSEF_A = \frac{MS_A}{MS_E}, \quad F_B =
\frac{MS_B}{MS_E}, \quad F_{AB} =
\frac{MS_{AB}}{MS_E}FA=MSEMSA,FB=MSEMSB,FAB=MSEMSAB
- Compare
with F-critical values for each df set.
12.6 Let Us
Sum Up
- ANOVA
compares means of three or more groups using the F-test.
- One-way
ANOVA deals with a single factor; two-way ANOVA handles two factors.
- Both
types require assumptions of normality, equal variances, and independent
samples.
- The
F-ratio compares between-group variance to within-group variance to decide
significance.
12.7 Key
Words
- ANOVA –
Statistical method to compare means of multiple groups.
- One-way
ANOVA – Analysis with one factor.
- Two-way
ANOVA – Analysis with two factors, possibly
including interaction effects.
- Between-group
variance – Variation due to differences among group
means.
- Within-group
variance – Variation within groups due to random
error.
- F-ratio – Test
statistic in ANOVA.
Comments
Post a Comment