Document Code FM-STL-013

Saint Louis University

Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 139 of 169

Module 8: Analysis of Variance

This module aims to discuss the method of one-

way analysis of variance to conduct a
hypothesis test of equality of three or more
population means.

Upon completion of this module, you should be able to execute appropriate statistical
methods and tools in managing, organizing, presenting, and interpreting numerical
output.

Engage: T - Test VS ANOVA

Explore: Advantages and Principles of ANOVA
Explain: Hypothesis Testing making use of Single Factor ANOVA as Test Statistic
Elaborate: Demonstration Problem (Excel)
Evaluate:
Answer Laboratory Exercise 11
Summative Quiz at the end of Module 8

ANOVA is an acronym for “Analysis of Variance.” But analyzing Variances

is not its objective. Its objective is to determine whether three or
more of several Means are different from the others by a Statistically
Significant amount. It does this by analyzing Variances.

The t-test and ANOVA examine whether group means differ from one another. The t-
test compares two groups, while ANOVA can do more than two groups.

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.139
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 140 of 169

ANOVA Does ANOVA Does Not

compare several compare several

Means with each Means with the
Other overall Mean
say whether or not say which Means
there is a difference differ
among Means
require Continuous handle Discrete
data data

ADVANTAGES OF ANALYSIS OF VARIANCE (ANOVA)

1.1 The significance of differences between more than two sample means can
be checked SIMULTANEOUSLY. This means an economy in terms of time and
effort. Indeed, if a t-test would be assigned to do the statistical analysis, the
number of t-test needed to pair off all the means would be:

n(n – 1)
2
For example, to pair off 5 sample means, this would amount to

5 (5 – 1) = 10 t-tests
2
With ANOVA, a single operation does the job.

1.2 With ANOVA, we have a high number of df notwithstanding the fact that
there are few subjects involved in the different experimental conditions
Example: 5 experimental conditions with N=10 per experimental condition. With
a t-test, pairing off the conditions, the df = N1 – 1 + N2 – 1 =10 -1 + 10 -1 =
18 df. With ANOVA, the df = N -1 = 50 – 1 = 49 df.

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.140
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 141 of 169

Since the df is much larger, the statistical analysis is more reliable and the “0”
can be rejected more easily.

PRINCIPLE OF ANALYSIS OF VARIANCE

ü The total variance in a group of measurement can be subdivided into two

components and each component has its own specific source of variance.

ü The first component is the variance which exists within the groups. This
variance is caused by pre-existing individual differences and fluctuations
taking place at the time of gathering the data.

ü The second component is the variance which exists between the groups.
This variance is caused by the experimental conditions.

On basis of each component an estimate of the population variance can be

made. Two independent estimates of the population variance are therefore computed:
1) on basis of the variance within the groups; 2) on basis of the variance between the
groups.

* If the different samples (with their respective sample means) originate from
the same population (i.e., there is no significant difference between the different
sample means), the estimate of the population variance on basis of the variance
between the groups will be essentially the same as the estimate of the population
variance on basis of the variance within the groups.

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.141
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 142 of 169

* If the different samples (with their respective sample means) do not come
from the same population (i.e., there is a significant difference between the different
sample means, a difference caused by an experimental factor), the estimate of the
population variance on basis of the variance between the groups will be larger than
the estimate of the population variance on basis of the variance within the groups.

We therefore compare the estimate of the population variance on basis of the

variance between the groups with the population variance on basis of the variance
within the groups by means of an F-test.

ü If the obtained F-value is smaller than the critical F-value, we MAINTAIN

our “0” hypothesis.

ü If the obtained F-value is equal to or greater than the critical F-value, we

reject our “0” hypothesis, it means that we can be reasonably sure that
there is in the population a difference between the groups.

There are 3 types of ANOVA

1-Way ANOVA, aka Single Factor

2-Way ANOVA without Replication
2-Way ANOVA with Replication

Note: We will only cover 1 way or Single Factor ANOVA

Single Factor Design

A Single Factor ANOVA or One-Way ANOVA has only one independent variable, which
can have more than two levels. Each group in the study/experiment represents a
different level of the independent variable.

Single Factor Design for Uncorrelated Means

Sample Problem: A researcher is interested in evaluation the effectiveness of three

methods of teaching a given subject matter. A group of 15 students is considered to
be representative for the population. At random, the 15 students are assigned to
three subgroups of 5 students each, who are exposed to one of the three methods

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.142
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 143 of 169

of teaching. The scores are the result of a test after the students were taught with
method A, B or C.

Data:
Method A Method B Method C
12 11 7
10 9 6
9 8 3
8 7 2
6 5 2

Null Hypothesis: There is no difference in the effectiveness among 3 methods of

teaching.

Step 1 Computations:

Method A Method B Method C

12 11 7

10 9 6

9 8 3

8 7 2

6 5 2

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.143
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 144 of 169

I (; = 45 I (" = 40 I (A = 20 I ( = 105

I (; " = 425 I (" " = 340 I (A " = 102 I ( " = 867

(J; = 9 (J" = 8 (JA = 4

1. (∑ ()" / (n) (k) = 105" / (5) (3) = 735

2. ∑ ( " = ∑(425 + 340 + 102) = 867

3. ∑(∑ (; )" / n = ∑(45" + 40" + 20" )/ 5 = 4025/5 = 805

Step 2: ANOVA TABLE

SOURCE Computational SS df L) F

Formula

BETWEEN 3–1 70 2 35.00 6.77 *

WITHIN 2-3 62 12 5.16

TOTAL 2–1 132 14

Step 3. Check the F-table for the critical values. ( Appendix C)

For 2/12 df, the critical value at 5.>! is 3.88 at

5.>; is 6.93.

Degrees of Freedom:

BETWEEN: k–1

WITHIN: (n) (k) – k

TOTAL: (n) (k) - 1

Step 4. STATISTICAL INTERPRETATIONS AND CONCLUSIONS:

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.144
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 145 of 169

The obtained F value of 6.77 falls beyond critical value 3.88, and is significant at @
.05. Hence, the “0” which says that there is no difference in the effectiveness among
3 methods of teaching has to be rejected.

Step 5. Indicate the direction of the difference if the F-value is significant

In order to situate exactly where the significant differences are, the different
means have to be paired off. This can be done by, among, others, the TUKEY
METHOD. In this method, q-values are computed.

q = 7JK / ,M " L50M56 /n

,M " L50M56 /n = ,5.16/5 = √1.032 = 1.016

q = 7JK /,M " L50M56 /N

,M " L50M56 /N = ,5.16/5 = √1.032 = 1.016

(J; - (J" = 9 – 8 = 1 q = 1/1.016 = .984 NS

(J; - (JA = 9 – 4 = 5 q = 5/1.016 = 4.921 *

(J" - (JA = 8 – 4 = 4 q = 4/1.016 = 3.937 *

For k/N-k df, i.e. 3/15-3 df, the critical value (see APPENDIX G- Tukey Table)) at 5.>! is 3.77,
at 5.>; is 5.05.

The sample means are ranked from the highest to the lowest. Insignificant q-values
determine the sample means which belong to the same cluster, i.e. to the same
population.

A=9

B=8

C=4

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.145
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 146 of 169

Interpretation: Method A is not significantly different from Method B and we therefore

can conclude that both methods are equally good to teach the given course. But both
methods are significantly superior to method c.

Single Factor Design for Correlated Sample Means

Sample Problem: An experiment is carried out to study the influence of the modality of
the stimulus on the Reaction Time (RT). Ten subjects were each exposed to the different
stimuli: visual, auditory, olfactory, taste and touch. The reaction time (RT) is expressed in
hundredths of a second.

Data on Next Page:

Null Hypothesis: There is no difference between the reaction times (RT’s) towards
five different stimuli.

Step 1: Computations

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.146
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 147 of 169

"
1. (∑ () / (n) (k) = 1569" / (10) (5) = 49235.22

2. ∑ (" = 54163
"
3. ∑(∑ (N ) / n = ∑(251" + 190" + 433" + 416" + 433" + 279" / 10

= 53748.70

4. ∑(Q5 )" / k = ∑(149" + 163" + 141" + … + 156" + 161" / 5

= 49353.80

Step 2: ANOVA Table:

SOURCE Computational SS df L) F
Formula

BETWEEN 3- 1 4513.48 4 1128.37 137.36 **

WITHIN 2-3 414.30 45 -

Bet Subs 4-1 118.58 9 -

Residual 2-3-4+1 295.72 36 8.21

TOTAL 2- 1 4927.78 49

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.147
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 148 of 169

Step 3: Check the F-table for the critical values. (Appendix C)

For 4/36 df, the critical value at 5.>! is 2.63, 5.>; is 3.89.

Degrees of Freedom:

BETWEEN: k -1

WITHIN: (n) (k) – k

Between Subjects: n -1

Residual: (n – 1) (k – 1)

TOTAL: (n) (k) - 1

Step 4. Statistical Interpretations and Conclusions:

The obtained F value of 137. 36 falls beyond critical value 2.63, and is significant at
α .05. Hence, the “0” which says that that there is no difference between the reaction
times (RT’s) towards five different stimuli has to be rejected.

Step 5. In order to situate exactly where the significant differences are, the sample
means have to be pained off by means of the Tukey method.

,M " L50M56 /N = ,8.21/10 = √8.21 = .906

(J; - (J" = 25.1 – 19.0 = 6.1 q = 6.1/.906 = 6.73**

(J; - (JA = 25.1 – 43.3 = -18.2 q = -18.2/.906 = -20.08 **

(J; - (JE = 25.1 – 41.6 = -16.5 q = -16.5/.906 = -18.21 **

(J; - (J! = 25.1 – 27.9 = -2.8 q = -2.8/.906 = -3.09 NS

(J" - (JA = 19.0 – 43.3 = -24.3 q = -24.3/.906 = 26.82 **

(J" - (JE = 19.0 – 41.6 = -22.6 q = -22.6/.906 = -24.94 **

(J" - (J! = 19.0 – 27.9 = -8.9 q = -8.9/.906 = -9.82 **

(JA - (JE = 43.3 – 41.6 = 1.7 q = 1.7/.906 =1.88 NS

(JA - (J! = 43.3 – 27.9 = 15.4 q = 15.4/.906 = 16.99 **

(JE - (J! = 41.6 – 27.9 = 13.7 q = 13.7/.906 = 15.12 **

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.148
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 149 of 169

For k/N-k, i.e., 5/50-5 df, 5/40 df, the critical value at 5.>; is 4.04 and at 5.>; is 4.93.

The sample means are ranked from shortest to longest reaction time (RT) and insignificant
q-values are clustered together.

Aud = 19.0

Vis = 25.1

Interpretation: Touch = 27.9

Taste = 41.6

Olf = 43.3

The reaction time (RT) of the Auditory stimulus is significantly shorter than reaction times
(RT’s) of all other stimuli. The RT’s of the Visual and the Touch stimuli are not significantly
different from each other. But both are significantly longer than the RT of the Auditory
stimulus and significantly shorter than the RT’s on the Taste and Olfactory stimuli. These
latter ones are insignificantly different from each other and have significantly longer RT’s
as compared to the other three stimuli.

ANOVA Single Factor Design in Excel

Single Factor Design for Uncorrelated Means

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.149
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 150 of 169

Sample Problem: A researcher is interested in evaluation the effectiveness of three

methods of teaching a given subject matter. A group of 15 students is considered to
be representative for the population. At random, the 15 students are assigned to
three subgroups of 5 students each, who are exposed to one of the three methods
of teaching. The scores are the result of a test after the students were taught with
method A, B or C.
Data:
Method A Method B Method C
12 11 7
10 9 6
9 8 3
8 7 2
6 5 2

Step 1. For this problem, we carry out an Anova: Single Factor since there is
only ONE Independent variable.
In MS Excel: Type in your Data. Go to TOOLS /DATA, then click on DATA ANALYSIS.
Select Anova: Single Factor

Anova: Single Factor

SUMMARY
Groups Count Sum Average Variance
Method A 5 45 9 5
Method B 5 40 8 5
Method C 5 20 4 5.5
ANOVA
Source of Variation SS df MS F P-value F crit

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.150
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 151 of 169

Between Groups 70 2 35 6.77419355 0.010737583 3.885293835

Within Groups 62 12 5.16666667
Total 132 14

Step 2. a) Present your Results. Interpret the statistics. IF the F-value is significant,
proceed to b.
b) generate your conclusions

For Step 2.a.) The obtained F value of ________ falls ___________ critical value ________,

and is (significant/ not significant) at α .05. Hence, the “0” which says that
_______________________________________________________ _________________is
(rejected/accepted).

*With the significant F-value, a post-hoc test (e.g. Tukey Method for ANOVA) is needed in
order indicate the direction of the difference between the means and be able to
generate conclusions.

Note: MS Excel does not a perform post-hoc tests. In order to investigate into the
direction of difference between the means, we carry out a t-test on the means (one
pair at a time).

To do this:
1. Rank the means from best to worst. (in this example, highest to lowest).

Rank 1: Method A; Mean = 9

Rank 2: Method B; Mean = 8
Rank 3: Method C; Mean = 4

2. Perform a t-test on the first PAIR of means. Then on the second pair, third
pair, etc…

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.151
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 152 of 169

Clusters:

A= 9

B=8

C=4

For Step 2.b.) Method A (Mean = 9) and Method B (Mean = 8) yield the same test
results, hence, both are equally effective. Additionally, both methods A & B yield
significantly better results than Method C (Mean = 4). Methods A and B are
significantly more effective than Method C.

Single Factor Design for Correlated Means

Sample Problem:
An experiment is carried out to study the influence of modality of the stimulus on the
Reaction Time. Eight subjects were exposed to five different stimuli: visual, auditory,
olfactory, taste, and touch. They were told to press a lever if they perceive the
stimulus. The reaction time (RT) is expressed in hundredths of a second.

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.152
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 153 of 169

Data:

Visual Auditory Olfactory Taste Touch

19 24 42 37 28
23 28 39 44 25
18 27 37 43 27
19 21 42 38 26
21 23 41 41 23
19 26 40 39 24
15 24 39 45 27
20 23 43 41 26
Step 1

Step 2

Interpret the statistics and generate your conclusions (Note the F-value against the F crit
value)
Anova: Two-Factor Without
Replication
SUMMARY Count Sum Average Variance
1 5 150 30 88.5
2 5 159 31.8 84.7
3 5 152 30.4 94.8
4 5 146 29.2 105.7
5 5 149 29.8 105.2

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.153
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 154 of 169

6 5 148 29.6 88.3

7 5 150 30 144
8 5 153 30.6 113.3
Visual 8 154 19.25 5.3571429
Auditory 8 196 24.5 5.4285714
Olfactory 8 323 40.375 3.9821429
Taste 8 328 41 8.2857143
Touch 8 206 25.75 2.7857143

ANOVA
Source of Variation SS df MS F P-value F crit
Rows 21.775 7 3.1107143 0.5474544 0.7911066 2.35925985
COLUMNS 3138.9 4 784.725 138.1037 5.296E-18 2.714076
Error 159.1 28 5.6821429
Total 3319.775 39
**As we are dealing with ONLY ONE INDEPENDENT VARIABLE and with CORRELATED MEANS, we only
pay attention and interpret the F-value for COLUMNS (Between Subjects).

STATISTICAL INTERPETATIONS AND CONCLUSIONS:

The obtained F value of ___________ falls ____________ the critical value of

_________, and is ________________ at α .05. Hence, the “0” which says that
________________________________________________________________
________________________________________________ is ________________.

With the SIGNIFICANT F-VALUE, we perform a post-hoc test to indicate the

direction of difference of the means. First, we RANK the means from lowest to highest
(a lower number indicates a faster, therefore better Reaction Time). Then, we carry
out t-tests, taking one pair of means at a time. (for CORRELATED means, use t-test:
Paired Two Sample for Means)

FIRST t-test THIRD t-test FIFTH t-test

t-Test: Paired Two Sample for Means t-Test: Paired Two Sample for Means t-Test: Paired Two Sample for Means
Visual Auditory Auditory Olfactory Olfactory Taste
Mean 19.25 24.5 Mean 24.5 40.375 Mean 40.375 41
t Stat -5.1811 t Stat -11.084 t Stat -0.3955
tCrit two-tail 2.36462 tCrit two-tail 2.36462 t Crit two-tail 2.36462

SECOND t-test FOURTH t-test We carry out FIVE t-tests to

t-Test: Paired Two Sample for Means t-Test: Paired Two Sample for Means determine where exactly the
Auditory Touch Touch Olfactory significance of difference of

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.154
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 155 of 169

Mean 24.5 25.75 Mean 25.75 40.375 means lies. Without the ANOVA,
t Stat -1.2127 t Stat -15.497 we would have to carry out a total
tCrit two-tail 2.36462 tCrit two-tail 2.36462 of 10 t tests.

With the t-test, we come up with the following clusters of means.

Visual 19.25
Auditory 24.50
Touch 25.75
Olfactory 40.38
Taste 41.00

Interpretation:

The Reaction Time of VISUAL Stimulus (Mean = 19.25) is significantly the shortest, therefore
fastest among the different stimuli. The reaction times for Auditory (Mean = 24.50) and
Touch (Mean = 25.75) do not differ significantly from one another, but are both significantly
longer than the Visual Stimulus, and significantly shorter than that of the Olfactory (Mean
= 40.38) and Taste Stimuli (Mean = 41.00). The latter two modalities of stimuli are not
significantly different from one another, and yield significantly longer Reaction Times
compared to the three other stimuli. The Modality of a stimulus indeed influences Reaction
Time.

Answer Laboratory Exercise 11

Summative Quiz at the end of Module 8

References

Main Reference

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.155
 Document Code FM-STL-013
Saint Louis University
Revision No. 01
School of Teacher Education and Liberal Arts
Effectivity June 07, 2021

Page 156 of 169

Inferential Statistics Outline. Department of Psychology. Baguio City. Saint Louis

University Printing Press.

Books
Bluman, A.G. (2014). Elementary Statistics: A Step by Step Approach (9th Ed.). New York:
McGraw Hill - Education
Narag, E.R. (2018). Basic Statistics. Manila: Rex Publishing
Cohen, B.H. (2013). Explaining Psychological Statistics (4th Ed.). Hokoben, New Jersey:
John Wiley & Sons, Inc.
Gravetter, F. & Wallnau L. (2019). Statistics for the Behavioral Sciences (10th Ed.).
Philippines: Cengage Learning

Electronic References
http://www.socialresearchmethods.net
http://www.thomsonedu.com/psychology/pagano
http://www.sdecnet.com/psychology/stathelp.htm
http://www.wpcmath.com
http://www.psychnet.apa.org
http://www.allpsych.com/researchmethods/inferentialstatistics.html
http://www. statanalytica.com/blog/uses-of-statistics

Property of and for the exclusive use of SLU. Reproduction, storing in a retrieval system, distributing, uploading or posting online, or transmitting in any form or by any
means, electronic, mechanical, photocopying, recording, or otherwise of any part of this document, without the prior written permission of SLU, is strictly prohibited.156