ANOVA MODEL
Parametric Test
Test Description
Independent Samples t-test interested in detecting differences
between the means of two independent groups are called
Independent Samples independent samples test.
t-test Example - A labor union wants to compare the productivity levels
of workers for two different groups.
The ANALYSIS OF VARIANCE (ANOVA) is used to determine
ANALYSIS OF VARIANCE whether there is any statistical significant difference between the
(ANOVA) means of three or more independent (unrelated groups)
.
ANOVA
ANOVA or Analysis of Variance is used to compare the means of more
than two populations. It uncovers the main and interaction effects of
classification or independent variables on one or more dependent
variables.
Also, known as f-test, which is based on f distribution
Applicability of ANOVA
• Consumer behavior: A researcher wants to investigate the impact of
three different advertising stimuli on the shopping propensity of
males and females as well as consumers of different age brackets. The
dependent variable here is shopping propensity and independent
variables or the factors are advertising stimuli, gender, and age
brackets.
• Marketing management: A marketing manager wants to investigate
the impact of different discount schemes on the sale of three major
brands of edible oil.
ANOVA Procedure
ANOVA analysis uses the F-statistic, which tests if the means of the groups, formed
by one independent variable or a combination of independent variables, are
significantly different.
It is based on the comparison of two estimates of variances—one representing the
variance within groups, often referred to as error variance and the other
representing the variance due to differences in group means.
ANOVA is ratio between “Mean sum of Square between (MSSB) and “ Mean sum of
Square within (MSSW)
Between Groups
Group 1 Group 2 Group 3
19 Variance Variance 20
14
25 Between Between 22
Groups 27 Groups
32 39 33
58 51 50
59 66 52
94 70 55
56 60 53
Between Groups
Group 1 Group 2 Group 3
19 14 20
25 27 22
32 39 33
58 51 50
59 66 52
Variance Variance
94 70 55
within within
56 Groups 60 Groups 53
ANOVA Procedure
If the two variances do not differ significantly, one can believe that all the group
means come from the same sampling distribution of means and there is no reason
to claim that the group means differ.
If, however, the group means differ more than can be accounted for due to random
error, there is reason to believe that they were drawn from different sampling
distributions of means.
The larger the F-ratio, the greater is the difference between groups as compared to
within group differences. An F-ratio equal to or less than 1 indicates that there is
no significant difference between groups and the null hypothesis is correct.
ANOVA Assumptions
Independent observations: each record in the data must be a distinct and
independent entity.
Normality: the dependent variable is normally distributed in the population.
Normality is not needed for reasonable sample sizes, say each n ≥ 25.
Homogeneity: the variance of the dependent variable must be equal in each
subpopulation. Homogeneity is only needed for (sharply) unequal sample
sizes. In this case, Levene's test can be used to see if homogeneity is met.
Example- Using SPSS
One-Way ANOVA is the generalization of the t-test for independent
samples to situations with more than two groups. It is also known as
single classification ANOVA or one-factor ANOVA. It is used to test the
difference in a single dependent variable among two or more groups
formed by a single independent or classification variable.
Example
An oil company has introduced a new brand of gasoline in its outlets in
three major metro cities. However, they are not sure how the new
brand is selling at the three places since there is a lot of difference in
the driving habits of people in the three metros. The company selected
10 outlets in each city and tabulated the data on an average daily sale
at each of the selected outlets. The items 1, 2, and 3 in the table
represent the three metros.
The null hypothesis in this case is:
• H01: The average sale of the new brand of gasoline is same in all the
metro cities.
We will also explain multiple comparisons in this example by testing for
the following null hypotheses:
• H02: The average sale of the new brand of gasoline in city 1 is same as
that in city 2.
• H03: The average sale of the new brand of gasoline in city 1 is same as
that in city 2 and city 3 together.
NOTE
Please note the difference between H02, H03, and H01. The F-statistics
obtained from ANOVA only tells us whether there is any significant
difference in the mean values of the three groups, which is our H01.
From this, we cannot conclude anything about specific hypotheses as
stated in H02 and H03.
Using SPSS
The variables are labeled as outlets, metro, and volume respectively. Click on
Analyze, which will produce a drop down menu, choose Compare Means
from that and click on One-Way ANOVA.
In this case Volume is the
dependent variable and
should be transferred into the
Dependent List box by
clicking on the first arrow in
the middle of the two boxes.
Metro is the factoring variable
and should be transferred into
Factor box by clicking on the
second arrow. One can select
a number of additional
statistics from Options,
Result
The results are given in three rows. The first row labeled Between Groups
gives the variability due to the place of sale (between-groups variability),
the second row labeled Within Groups gives variability due to random error,
and the third row gives the total variability. In the given example,
F-value is 35.524 and the corresponding p-value is given as <0.000.
Therefore we can safely reject the null hypothesis (H01) and conclude that
the average sale of the new brand of gasoline is not the same in all three
metros.
Thank You
