ANOVA 
One way Single Factor Models 
KARAN DESAI-11BIE001 
DHRUV PATEL-11BIE024 
VISHAL DERASHRI -11BIE030 
HARDIK MEHTA-11BIE037 
MALAV BHATT-11BIE056
DEFINITION 
Analysis of variance (ANOVA) is a collection of 
statistical models used to analyze the differences 
between group means and their associated procedures 
(such as "variation" among and between groups), 
developed by R.A.Fisher .In the ANOVA setting, the 
observed variance in a particular variable is partitioned 
into components attributable to different sources of 
variation 
2
-Sir Ronald 
Aylmer Fisher 
FRS was an English statistician, 
evolutionary biologist, geneticist, and 
3
Why ANOVA 
• Compare the mean of more than two 
population? 
• Compare populations each containing 
several subgroups or levels? 
4
Problem with multiple T test 
• One problem with this approach is the increasing 
number of tests as the number of groups 
increases 
• The probability of making a Type I error increases 
as the number of tests increase. 
• If the probability of a Type I error for the analysis 
is set at 0.05 and 10 t-tests are done, the overall 
probability of a Type I error for the set of tests = 1 
– (0.95)10 = 0.40* instead of 0.05 
5
In its simplest form, ANOVA provides a statistical test of 
whether or not the means of several groups are equal, 
and therefore generalizes the t-test to more than two 
groups. As doing multiple two-sample t-tests would result 
in an increased chance of committing a statistical type-I 
error, ANOVAs are useful in comparing (testing) three or 
more means (groups or variables) for statistical 
significance. 
6
• Another way to describe the multiple comparisons 
problem is to think about the meaning of an alpha 
level = 0.05 
• Alpha of 0.05 implies that, by chance, there will be 
one Type I error in every 20 tests: 1/20 = 0.05. 
• This means that, by chance the null hypothesis 
will be incorrectly rejected once in every 20 tests 
• As the number of tests increases, the probability 
of finding a ‘significant’ result by chance increases. 
7
Importance of ANOVA 
• The ANOVA is an important test because 
it enables us to see for example how 
effective two different types of treatment 
are and how durable they are. 
• Effectively a ANOVA can tell us how well a 
treatment work, how long it lasts and how 
budget friendly it will be an 
8
CLASSIFICATION OF ANOVA 
MODEL 
1. Fixed-effects models: 
The fixed-effects model of analysis of 
variance applies to situations in which the experimenter 
applies one or more treatments to the subjects of the 
experiment to see if the response variable values 
change. This allows the experimenter to estimate the 
ranges of response variable values that the treatment 
would generate in the population as a whole. 
9
2. Random-effects model: 
Random effects models are used 
when the treatments are not fixed. This occurs when the 
various factor levels are sampled from a larger 
population. Because the levels themselves are random 
variables , some assumptions and the method of 
contrasting the treatments (a multi-variable 
generalization of simple differences) differ from the fixed-effects 
model. 
10
3.Mixed-effects models 
A mixed-effects model contains experimental 
factors of both fixed and random-effects types, with appropriately 
different interpretations and analysis for the two types. 
Example: Teaching experiments could be performed by a 
university department to find a good introductory textbook, with 
each text considered a treatment. The fixed-effects model would 
compare a list of candidate texts. The random-effects model 
would determine whether important differences exist among a 
list of randomly selected texts. The mixed-effects model would 
compare the (fixed) incumbent texts to randomly selected 
alternatives. 
11
ASSUMPTION 
Normal distribution 
Variances of dependent variable are equal in all 
populations 
Random samples; independent scores 
12
One way Single factor ANOVA 
13
ONE-WAY ANOVA 
One factor (manipulated variable) 
One response variable 
Two or more groups to compare 
14
USEFULLNESS 
Similar to t-test 
More versatile than t-test 
Compare one parameter (response variable) 
between two or more groups 
15
Remember that… 
Standard deviation (s) 
n 
s = √[(Σ (xi – X)2)/(n-1)] 
i = 1 
In this case: Degrees of freedom (df) 
df = Number of observations or groups 
16
ANOVA 
ANOVA (ANalysis Of VAriance) is a natural extension used to 
compare the means more than 2 populations. 
Basic Question: Even if the true means of n populations were 
equal (i.e. m1 = m2 = m3 = m4) we cannot expect the sample 
means (x1, x2, x3, x4 ) to be equal. So when we get 
different values for the x’s, 
How much is due to randomness? 
How much is due to the fact that we are sampling from 
different populations with possibly different mj’s.
ANOVA TERMINOLOGY 
Response Variable (y) 
What we are measuring 
Experimental Units 
The individual unit that we will measure 
Factors 
Independent variables whose values can change to affect 
the outcome of the response variable, y 
Levels of Factors 
Values of the factors 
Treatments 
The combination of the levels of the factors applied to an 
experimental unit
Example 
We want to know how combinations of different 
amounts of water (1 ac-ft, 3 ac-ft, 5 ac-ft) and 
different fertilizers (A, B, C) affect crop yields 
Response variable 
– crop yield (bushels/acre) 
Experimental unit 
Each acre that receives a treatment 
Factors (2) 
Water and fertilizer 
Levels (3 for Water; 3 for Fertilizer) 
Water: 1, 3, 5; Fertilizer: A, B, C 
Treatments (9 = 3x3) 
1A, 3A, 5A, 1B, 3B, 5B, 1C, 3C, 5C
Total Treatments 
Fertilizer 
A B C 
1 AC-FT Treatment 1 Treatment 2 Treatment 3 
Water 3 AC-FT Treatment 4 Treatment 5 Treatment 6 
5 AC-FT Treatment 7 Treatment 8 Treatment 9
Single Factor ANOVA 
Basic Assumptions 
 If we focus on only one factor (e.g. fertilizer type in the 
previous example), this is called single factor ANOVA. 
 In this case, levels and treatments are the same thing since 
there are no combinations between factors. 
 Assumptions for Single Factor ANOVA 
1. The distribution of each population in the comparison has a 
normal distribution 
2. The standard deviations of each population (although 
unknown) are assumed to be equal (i.e. s1 = s2 = s3 = s4) 
3. Sampling is: 
Random 
Independent
Example 
The university would like to know if the delivery mode of the 
introductory statistics class affects the performance in the 
class as measured by the scores on the final exam. 
The class is given in four different formats: 
Lecture 
Text Reading 
Videotape 
Internet 
The final exam scores from random samples of students from 
each of the four teaching formats was recorded.
Samples
Summary 
There is a single factor under observation – teaching format 
There are k = 4 different treatments (or levels of teaching 
formats) 
The number of observations (experimental units) are n1 = 7, 
n2 = 8, n3 = 6, n4 = 5 total number of 
observations, n = 26 
Treatment Means : x1 = 76, x2 = 65, x3 = 75, x4 = 
74 
Grand mean (of all 26 observations) : x = 
72
Why aren’t all thex’s the same? 
There is variability due to the different treatments -- 
Between Treatment Variability (Treatment) 
There is variability due to randomness within each 
treatment -- Within Treatment Variability (Error) 
BASIC CONCEPT 
If the average Between Treatment Variability is “large” 
compared to the average Within Treatment Variability, 
we can reasonably conclude that there really are 
differences among the population means (i.e. at least 
one μj differs from the others).
Basic Questions 
Given this basic concept, the natural questions are: 
What is “variability” due to treatment and due to error 
and how are they measured? 
What is “average variability” due to treatment and due 
to error and how are they measured? 
What is “large”? 
How much larger than the observed average 
variability due to error does the observed average 
variability due to treatment have to be before we 
are convinced that there are differences in the true 
population means (the μ’s)?
How Is “Total” Variability 
Measured? 
Variability is defined as the Sum of Square Deviations (from the 
grand mean). So, 
SST (Total Sum of Squares) 
 Sum of Squared Deviations of all observations from the 
grand mean. (McClave uses SSTotal) 
SSTr (Between Treatment Sum of Squares) 
Sum of Square Deviations Due to Different Treatments. 
(McClave uses SST) 
SSE (Within Treatment Sum of Squares) 
Sum of Square Deviations Due to Error 
SST = SSTr + SSE
How is “Average” Variability Measured? 
“Average” Variability is measured in: 
Mean Square Values (MSTr and MSE) 
Found by dividing SSTr and SSE by their 
respective degrees of freedom 
ANOVA TABLE 
# treatments -1 DFT - DFTR 
Variability SS DF Mean Square (MS) 
Between Tr. (Treatment) SSTr k-1 SSTr/DFTR 
Within Tr. (Error) SSE n-k SSE/DFE 
TOTAL SST n-1 
# observations -1
Formula for Calculating 
SST 
Calculating SST 
Just like the 
numerator of the 
variance 
assuming all (26) 
entries come 
from one 
population 
=  
SST (x x) 
ij 
2 2 
2 
82 72) ... (81 72) 4394 
=     =
Formula for Calculating 
SSTr 
Calculating SSTr 
Between Treatment 
Variability 
Replace all entries within 
each treatment by its 
mean – now all the 
variability is between (not 
within) treatments 
76 
76 
76 
76 
76 
76 
76 
=  
SSTr n (x x) 
2 
j j 
75 
75 
75 
75 
75 
75 
65 
65 
65 
65 
65 
65 
65 
65 
2 2 2 2 
=        = 
7(76 72) 8(65 72) 6(75 72) 5(74 72) 578 
74 
74 
74 
74 
74
Formula for Calculating 
SSE 
Calculating SSE (Within Treatment Variability) 
The difference between the SST and SSTr --- 
SSE SST - SSTr 
= = 
4394 - 578 = 
3816
Can we Conclude a Difference Among 
the 4 Teaching Formats? 
We conclude that at least one population mean differs 
from the others if the average between treatment 
variability is large compared to the average within 
treatment variability, that is if MSTr/MSE is “large”. 
The ratio of the two measures of variability for these 
normally distributed random variables has an F 
distribution and the F-statistic (=MSTr/MSE) is 
compared to a critical F-value from an F distribution 
with: 
Numerator degrees of freedom = DFTr 
Denominator degrees of freedom = DFE 
If the ratio of MSTr to MSE (the F-statistic) exceeds 
the critical F-value, we can conclude that at least one 
population mean differs from the others.
Can We Conclude Different Teaching 
Formats Affect Final Exam Scores? 
The F-test 
H0: m1 = m2 = m3 = m4 
HA: At least one mj differs from the others 
Select α = .05. 
Reject H0 (Accept HA) if: 
F =  α,DFTr,DFE = .05,3,22 = 
F F 3.05 
MSTr 
MSE
Hand Calculations for the F-test 
173.45 
578 
= = = 
3816 
22 
SSTr 
SSE 
DFE 
MSE 
192.67 
3 
DFTr 
MSTr 
= = = 
1.11 
192.67 
= = 
173.45 
1.11 3.05 
F 
 
Cannot conclude there is a difference among the μj’s
Excel Approach
EXCEL OUTPUT 
p-value = .365975 > .05 
Cannot conclude differences
REVIEW 
ANOVA Situation and Terminology 
Response variable, Experimental Units, Factors, 
Levels, Treatments, Error 
Basic Concept 
If the “average variability” between treatments is “a 
lot” greater than the “average variability” due to error – 
conclude that at least one mean differs from the 
others. 
Single Factor Analysis 
By Hand 
By Excel

Anova single factor

  • 1.
    ANOVA One waySingle Factor Models KARAN DESAI-11BIE001 DHRUV PATEL-11BIE024 VISHAL DERASHRI -11BIE030 HARDIK MEHTA-11BIE037 MALAV BHATT-11BIE056
  • 2.
    DEFINITION Analysis ofvariance (ANOVA) is a collection of statistical models used to analyze the differences between group means and their associated procedures (such as "variation" among and between groups), developed by R.A.Fisher .In the ANOVA setting, the observed variance in a particular variable is partitioned into components attributable to different sources of variation 2
  • 3.
    -Sir Ronald AylmerFisher FRS was an English statistician, evolutionary biologist, geneticist, and 3
  • 4.
    Why ANOVA •Compare the mean of more than two population? • Compare populations each containing several subgroups or levels? 4
  • 5.
    Problem with multipleT test • One problem with this approach is the increasing number of tests as the number of groups increases • The probability of making a Type I error increases as the number of tests increase. • If the probability of a Type I error for the analysis is set at 0.05 and 10 t-tests are done, the overall probability of a Type I error for the set of tests = 1 – (0.95)10 = 0.40* instead of 0.05 5
  • 6.
    In its simplestform, ANOVA provides a statistical test of whether or not the means of several groups are equal, and therefore generalizes the t-test to more than two groups. As doing multiple two-sample t-tests would result in an increased chance of committing a statistical type-I error, ANOVAs are useful in comparing (testing) three or more means (groups or variables) for statistical significance. 6
  • 7.
    • Another wayto describe the multiple comparisons problem is to think about the meaning of an alpha level = 0.05 • Alpha of 0.05 implies that, by chance, there will be one Type I error in every 20 tests: 1/20 = 0.05. • This means that, by chance the null hypothesis will be incorrectly rejected once in every 20 tests • As the number of tests increases, the probability of finding a ‘significant’ result by chance increases. 7
  • 8.
    Importance of ANOVA • The ANOVA is an important test because it enables us to see for example how effective two different types of treatment are and how durable they are. • Effectively a ANOVA can tell us how well a treatment work, how long it lasts and how budget friendly it will be an 8
  • 9.
    CLASSIFICATION OF ANOVA MODEL 1. Fixed-effects models: The fixed-effects model of analysis of variance applies to situations in which the experimenter applies one or more treatments to the subjects of the experiment to see if the response variable values change. This allows the experimenter to estimate the ranges of response variable values that the treatment would generate in the population as a whole. 9
  • 10.
    2. Random-effects model: Random effects models are used when the treatments are not fixed. This occurs when the various factor levels are sampled from a larger population. Because the levels themselves are random variables , some assumptions and the method of contrasting the treatments (a multi-variable generalization of simple differences) differ from the fixed-effects model. 10
  • 11.
    3.Mixed-effects models Amixed-effects model contains experimental factors of both fixed and random-effects types, with appropriately different interpretations and analysis for the two types. Example: Teaching experiments could be performed by a university department to find a good introductory textbook, with each text considered a treatment. The fixed-effects model would compare a list of candidate texts. The random-effects model would determine whether important differences exist among a list of randomly selected texts. The mixed-effects model would compare the (fixed) incumbent texts to randomly selected alternatives. 11
  • 12.
    ASSUMPTION Normal distribution Variances of dependent variable are equal in all populations Random samples; independent scores 12
  • 13.
    One way Singlefactor ANOVA 13
  • 14.
    ONE-WAY ANOVA Onefactor (manipulated variable) One response variable Two or more groups to compare 14
  • 15.
    USEFULLNESS Similar tot-test More versatile than t-test Compare one parameter (response variable) between two or more groups 15
  • 16.
    Remember that… Standarddeviation (s) n s = √[(Σ (xi – X)2)/(n-1)] i = 1 In this case: Degrees of freedom (df) df = Number of observations or groups 16
  • 17.
    ANOVA ANOVA (ANalysisOf VAriance) is a natural extension used to compare the means more than 2 populations. Basic Question: Even if the true means of n populations were equal (i.e. m1 = m2 = m3 = m4) we cannot expect the sample means (x1, x2, x3, x4 ) to be equal. So when we get different values for the x’s, How much is due to randomness? How much is due to the fact that we are sampling from different populations with possibly different mj’s.
  • 18.
    ANOVA TERMINOLOGY ResponseVariable (y) What we are measuring Experimental Units The individual unit that we will measure Factors Independent variables whose values can change to affect the outcome of the response variable, y Levels of Factors Values of the factors Treatments The combination of the levels of the factors applied to an experimental unit
  • 19.
    Example We wantto know how combinations of different amounts of water (1 ac-ft, 3 ac-ft, 5 ac-ft) and different fertilizers (A, B, C) affect crop yields Response variable – crop yield (bushels/acre) Experimental unit Each acre that receives a treatment Factors (2) Water and fertilizer Levels (3 for Water; 3 for Fertilizer) Water: 1, 3, 5; Fertilizer: A, B, C Treatments (9 = 3x3) 1A, 3A, 5A, 1B, 3B, 5B, 1C, 3C, 5C
  • 20.
    Total Treatments Fertilizer A B C 1 AC-FT Treatment 1 Treatment 2 Treatment 3 Water 3 AC-FT Treatment 4 Treatment 5 Treatment 6 5 AC-FT Treatment 7 Treatment 8 Treatment 9
  • 21.
    Single Factor ANOVA Basic Assumptions  If we focus on only one factor (e.g. fertilizer type in the previous example), this is called single factor ANOVA.  In this case, levels and treatments are the same thing since there are no combinations between factors.  Assumptions for Single Factor ANOVA 1. The distribution of each population in the comparison has a normal distribution 2. The standard deviations of each population (although unknown) are assumed to be equal (i.e. s1 = s2 = s3 = s4) 3. Sampling is: Random Independent
  • 22.
    Example The universitywould like to know if the delivery mode of the introductory statistics class affects the performance in the class as measured by the scores on the final exam. The class is given in four different formats: Lecture Text Reading Videotape Internet The final exam scores from random samples of students from each of the four teaching formats was recorded.
  • 23.
  • 24.
    Summary There isa single factor under observation – teaching format There are k = 4 different treatments (or levels of teaching formats) The number of observations (experimental units) are n1 = 7, n2 = 8, n3 = 6, n4 = 5 total number of observations, n = 26 Treatment Means : x1 = 76, x2 = 65, x3 = 75, x4 = 74 Grand mean (of all 26 observations) : x = 72
  • 25.
    Why aren’t allthex’s the same? There is variability due to the different treatments -- Between Treatment Variability (Treatment) There is variability due to randomness within each treatment -- Within Treatment Variability (Error) BASIC CONCEPT If the average Between Treatment Variability is “large” compared to the average Within Treatment Variability, we can reasonably conclude that there really are differences among the population means (i.e. at least one μj differs from the others).
  • 26.
    Basic Questions Giventhis basic concept, the natural questions are: What is “variability” due to treatment and due to error and how are they measured? What is “average variability” due to treatment and due to error and how are they measured? What is “large”? How much larger than the observed average variability due to error does the observed average variability due to treatment have to be before we are convinced that there are differences in the true population means (the μ’s)?
  • 27.
    How Is “Total”Variability Measured? Variability is defined as the Sum of Square Deviations (from the grand mean). So, SST (Total Sum of Squares)  Sum of Squared Deviations of all observations from the grand mean. (McClave uses SSTotal) SSTr (Between Treatment Sum of Squares) Sum of Square Deviations Due to Different Treatments. (McClave uses SST) SSE (Within Treatment Sum of Squares) Sum of Square Deviations Due to Error SST = SSTr + SSE
  • 28.
    How is “Average”Variability Measured? “Average” Variability is measured in: Mean Square Values (MSTr and MSE) Found by dividing SSTr and SSE by their respective degrees of freedom ANOVA TABLE # treatments -1 DFT - DFTR Variability SS DF Mean Square (MS) Between Tr. (Treatment) SSTr k-1 SSTr/DFTR Within Tr. (Error) SSE n-k SSE/DFE TOTAL SST n-1 # observations -1
  • 29.
    Formula for Calculating SST Calculating SST Just like the numerator of the variance assuming all (26) entries come from one population =  SST (x x) ij 2 2 2 82 72) ... (81 72) 4394 =     =
  • 30.
    Formula for Calculating SSTr Calculating SSTr Between Treatment Variability Replace all entries within each treatment by its mean – now all the variability is between (not within) treatments 76 76 76 76 76 76 76 =  SSTr n (x x) 2 j j 75 75 75 75 75 75 65 65 65 65 65 65 65 65 2 2 2 2 =        = 7(76 72) 8(65 72) 6(75 72) 5(74 72) 578 74 74 74 74 74
  • 31.
    Formula for Calculating SSE Calculating SSE (Within Treatment Variability) The difference between the SST and SSTr --- SSE SST - SSTr = = 4394 - 578 = 3816
  • 32.
    Can we Concludea Difference Among the 4 Teaching Formats? We conclude that at least one population mean differs from the others if the average between treatment variability is large compared to the average within treatment variability, that is if MSTr/MSE is “large”. The ratio of the two measures of variability for these normally distributed random variables has an F distribution and the F-statistic (=MSTr/MSE) is compared to a critical F-value from an F distribution with: Numerator degrees of freedom = DFTr Denominator degrees of freedom = DFE If the ratio of MSTr to MSE (the F-statistic) exceeds the critical F-value, we can conclude that at least one population mean differs from the others.
  • 33.
    Can We ConcludeDifferent Teaching Formats Affect Final Exam Scores? The F-test H0: m1 = m2 = m3 = m4 HA: At least one mj differs from the others Select α = .05. Reject H0 (Accept HA) if: F =  α,DFTr,DFE = .05,3,22 = F F 3.05 MSTr MSE
  • 34.
    Hand Calculations forthe F-test 173.45 578 = = = 3816 22 SSTr SSE DFE MSE 192.67 3 DFTr MSTr = = = 1.11 192.67 = = 173.45 1.11 3.05 F  Cannot conclude there is a difference among the μj’s
  • 35.
  • 36.
    EXCEL OUTPUT p-value= .365975 > .05 Cannot conclude differences
  • 37.
    REVIEW ANOVA Situationand Terminology Response variable, Experimental Units, Factors, Levels, Treatments, Error Basic Concept If the “average variability” between treatments is “a lot” greater than the “average variability” due to error – conclude that at least one mean differs from the others. Single Factor Analysis By Hand By Excel