1

____________________________________________________________
Statistical Analysis with Software Application

ANALYSIS OF F test statistic

The value of the test statistic F for an ANOVA
VARIANCE (ANOVA) test is calculated as
The Analysis of Variance, commonly variance between samples MSB
abbreviated ANOVA, is a technique that uses F= =
variance wihin samples MSW
the F test to test a hypothesis concerning the
means of three or more populations. Two To calculate MSB and MSW, first compute the
different estimates of the population variance between-samples sum of squares denoted by
are made. The first estimate is called the SSB and the within-samples sum of the squares
between-group variance. It involves denoted by SSW. The sum of SSB and SSW is
computing the variance by using the means of called the total sum of the squares and it is
the group or between the groups. The second denoted by SST. That is,
estimate, the within-group variance., is made SST = SSB + SSW
by computing the variance using all the data. It
is not affected by differences in the means. The values of SSB and SSW are calculated using
the following formulas.
For a test of the differences among three or
more means, the following hypotheses are The between-samples sum of squares, denoted
used: by SSB,
T1 2 T2 2 Tk 2 (∑ x)2
Ho : μ1 = μ2 = μ3 = ⋯ = μn SSB = ( + + ⋯+ )−
n1 n2 nk n
Ha : Atleast one mean is different from
the others. where:
Ti =1,2,…,k = sum of sample in the ith group
If there is no difference in the means, the ni =1,2,…,k = number of samples in the ith group
between-group variance estimate will
approximately be equal to the within-group The within-samples sum of squares, denoted
variance estimate, and the F test value will be by SSW,
approximately equal to 1. When the means 2
T1 2 T2 2 Tk 2
SSW = ∑ x − ( + +⋯+ )
differ significantly, the between-group n1 n2 nk
variance will be much larger than the within-
group variance. Then the F test value will be The variance between samples MSB and the
significantly greater than 1. Then the null variance within samples MSW are calculated
hypothesis will be rejected. using the following formulas.

The degrees of freedom for the F test are The MSB and MSW are calculated as
𝑑. 𝑓. 𝑁 = 𝑘 − 1 where k is the number of SSB
MSB =
groups, and 𝑑. 𝑓. 𝐷 = 𝑛 − 𝑘 where n is the sum k−1
SSW
of the sample sizes of the groups, 𝑛 = 𝑛1 + MSW =
𝑛2 + ⋯ + 𝑛𝑘 . The sample sizes need not to be n−k
where k – 1 and n – k are respectively, the
equal. The F test to compare the means is degrees of freedoms for the numerator and
always a right-tailed test. denominator for the F distribution.

One-Way Analysis of Variance (One-way ANOVA)

This document is a property of University of Saint Louis Tuguegarao. It must not be reproduced or transmitted in
any form, in whole or in part, without expressed written permission.
2
____________________________________________________________
Statistical Analysis with Software Application

Pos-hoc Analysis
- To determine which of these groups differ from each other, pos hoc analysis is needed.

The Games-Howell test assumes that the groups being compared have unequal variances and
unequal sample sizes. It is designed to make multiple comparisons among means of these groups.

Tukey is only usable when the sample sizes are the same or equal variances.

Levene’s test for Homogeneity Test

Levene's test (Levene 1960) is used to test if k samples have equal variances. Equal variances
across samples is called homogeneity of variance. Some statistical tests, for example the analysis of
variance, assume that variances are equal across groups or samples. The Levene test can be used to
verify that assumption.

P value > 0.05, the variances are not significantly different from each other which means that the
homogeneity assumption of the variances are met. Therefore, use the Fisher Test

P value < 0.05, the variances are significantly different from each other which means that the
homogeneity assumption of the variances are not met. Therefore, use the WelchTest.

How to run the software for One-way ANOVA?

Auditors must make judgments about various aspects of an audit on the basis of their own direct
experience, indirect experience, or a combination of the two. In a study, auditors were asked to make
judgments about the frequency of errors to be found in an audit. The judgments by the auditors were then
compared. Suppose the following data were obtained. Use 5% level of significance, to test whether the
basis for the judgment affects the quality of the judgment?

Direct Indirect Combination

17 16.6 25.2
18.5 22.2 24
15.8 20.5 21.5
18.2 18.3 26.8
20.2 24.2 27.5
16 19.8 25.8
13.3 21.2 24.2

One-Way Analysis of Variance (One-way ANOVA)

This document is a property of University of Saint Louis Tuguegarao. It must not be reproduced or transmitted in
any form, in whole or in part, without expressed written permission.
3
____________________________________________________________
Statistical Analysis with Software Application

1. Using Microsoft Excel, put the values in two columns as shown below. This is done to determine the
grouping variable.

To avoid mistakes in encoding and for data in similar format like in Problem 2: copy the data from the
worksheet and paste in Microsoft Excel. Copy again then in an empty cell, right click and look for
Transpose in Paste Options.

2. Copy and paste in Jamovi.

One-Way Analysis of Variance (One-way ANOVA)

This document is a property of University of Saint Louis Tuguegarao. It must not be reproduced or transmitted in
any form, in whole or in part, without expressed written permission.
4
____________________________________________________________
Statistical Analysis with Software Application

3. Double-click the column header to edit the Data Variable and the measure type (Various aspects of
an audit (nominal) and scores (continuous), respectively).

One-Way Analysis of Variance (One-way ANOVA)

This document is a property of University of Saint Louis Tuguegarao. It must not be reproduced or transmitted in
any form, in whole or in part, without expressed written permission.
5
____________________________________________________________
Statistical Analysis with Software Application

4. Under Analyses Tab, click ANOVA. Then, choose One-way ANOVA.

5. Transfer or Drag the Data, various aspects of an audit to grouping variable box and scores to
dependent variable. At this point, do not mind the default result. Scroll down and check the settings.

One-Way Analysis of Variance (One-way ANOVA)

This document is a property of University of Saint Louis Tuguegarao. It must not be reproduced or transmitted in
any form, in whole or in part, without expressed written permission.
6
____________________________________________________________
Statistical Analysis with Software Application

6. Check the following boxes as seen below. Under the hypothesis, check the constructed hypothesis.
Check the result of the Shapiro Wilk for the Normality Test and the Levene test for the Homogeneity test
to determine the appropriate statistical test. Determine the necessary values to complete the table and
to come up with a sound decision and interpretation.

One-Way Analysis of Variance (One-way ANOVA)

This document is a property of University of Saint Louis Tuguegarao. It must not be reproduced or transmitted in
any form, in whole or in part, without expressed written permission.
7
____________________________________________________________
Statistical Analysis with Software Application

Table 1: Test of Difference among the various aspects of an audit

a. Hypotheses Ha: The basis for the judgment affects the quality of the judgment.
or There is a significant different among the various aspects of an audit.

Ho: The basis for the judgment do not affect the quality of the judgment.
or
There is a significant different among the various aspects of an audit.
b. Level of Significance. 𝛼 = 5% = 0.05
c. Statistical test *Normality test (Shapiro Wilk)
p-value = 0.668
Interpretation: The data is normally distributed.

*Homogeneity Test (Levine’s Test)

p-value = 0.890
Interpretation: The assumption of equal variances is met.

Statistical test: one-way ANOVA or Fisher test

d. Complete the table.

Variables mean standard f-value p-value Decision

deviation
Direct 17.0 2.24 22.2 0.000 0.000<0.05
Indirect 20.4 2.50 Significant
Combination 25.0 2.00
e. Interpretation.

The basis of judgment (Direct, Indirect, or Combination) has a statistically significant effect on the
quality of the judgment at the 5% level of significance.

Since the decision is significant, conduct the post-hoc test to determine which aspect affects the quality
of the judgment.

Choose Tukey since the variances are equal and the sample size is also equal. Otherwise, choose
Games-Howell test.

One-Way Analysis of Variance (One-way ANOVA)

This document is a property of University of Saint Louis Tuguegarao. It must not be reproduced or transmitted in
any form, in whole or in part, without expressed written permission.
8
____________________________________________________________
Statistical Analysis with Software Application

Table 1.1: Post-hoc Analysis of the various aspects of an audit

Tukey Post-Hoc Test – Scores
Direct Indirect Combination

Direct Mean difference — -3.40* -8.00***

t-value — -2.82 -6.63
df — 18.0 18.0
p-value — 0.029 <.001

Indirect Mean difference — -4.60**

t-value — -3.81
df — 18.0
p-value — 0.003

Combination Mean difference —

t-value —
df —
p-value —

Note. * p < .05, ** p < .01, *** p < .001

One-Way Analysis of Variance (One-way ANOVA)

This document is a property of University of Saint Louis Tuguegarao. It must not be reproduced or transmitted in
any form, in whole or in part, without expressed written permission.
9
____________________________________________________________
Statistical Analysis with Software Application

Interpretation: There are significant differences in all pairwise comparisons (Direct vs. Indirect, Direct
vs. Combination, and Indirect vs. Combination). Combination judgments have the highest mean score,
followed by Indirect, and then Direct. The Combination approach leads to the highest quality judgments,
while the Direct approach results in the lowest quality judgments

One-Way Analysis of Variance (One-way ANOVA)

This document is a property of University of Saint Louis Tuguegarao. It must not be reproduced or transmitted in
any form, in whole or in part, without expressed written permission.