Scribd
Upload
21 views22 pages

Analysisofcovariance 140715173812 Phpapp02

Analysis of Covariance (ANCOVA) is a statistical test that assesses differences between groups while controlling for the variance explained by a covariate, which is a continuous variable related to the dependent variable. It requires specific hypotheses, assumptions, and adjustments to the sum of squares and sum of products to account for the covariate's influence. ANCOVA can be performed in statistical software like SPSS, allowing for the analysis of both categorical and continuous independent variables.

Uploaded by

Prof. V. M. Gawde
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
21 views22 pages

Analysisofcovariance 140715173812 Phpapp02

Analysis of Covariance (ANCOVA) is a statistical test that assesses differences between groups while controlling for the variance explained by a covariate, which is a continuous variable related to the dependent variable. It requires specific hypotheses, assumptions, and adjustments to the sum of squares and sum of products to account for the covariate's influence. ANCOVA can be performed in statistical software like SPSS, allowing for the analysis of both categorical and continuous independent variables.

Uploaded by

Prof. V. M. Gawde
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 22

Analysis of Covariance

David Markham
djmarkham@bsu.edu
Analysis of Covariance
 Analysis of Covariance (ANCOVA) is a
statistical test related to ANOVA
 It tests whether there is a significant
difference between groups after controlling
for variance explained by a covariate
 A covariate is a continuous variable that
correlates with the dependent variable
So, what does all that mean?
 This means that you can, in effect, “partial
out” a continuous variable and run an
ANOVA on the results
 This is one way that you can run a
statistical test with both categorical and
continuous independent variables
Hypotheses for ANCOVA
 H0 and H1 need to be stated slightly
0
differently for an ANCOVA than a regular
ANOVA
 H0: the group means are equal after
controlling for the covariate
 H1: the group means are not equal after
controlling for the covariate
Assumptions for ANCOVA
ANOVA assumptions:
 Variance is normally distributed
 Variance is equal between groups
 All measurements are independent
Also, for ANCOVA:
 Relationship between DV and covariate is
linear
 The relationship between the DV and
covariate is the same for all groups
How does ANCOVA work?
 ANCOVA works by adjusting the total SS,
group SS, and error SS of the independent
variable to remove the influence of the
covariate
 However, the sums of squares must also
be calculated for the covariate. For this
reason, SSdv will be used for SS scores for
the dependent variable, and SScv will be
used for the covariate
Sum of Squares


SSxtotal (xij  x) 2


SSxgroup nj (x j  x)2


SSxerror (xij  xj ) 2

SSxerrorSSxtotal SSx group


Sum of Products
 To control for the covariate, the sum of products
(SP) for the DV and covariate must also be
used
 This is the sum of the products of the residuals
for both the DV and the covariate
 In the following slides, x is the covariate, and y
is the DV. i is the individual subject, and j is the
group.
Total Sum of Products

SPxytotal
j i
(xij  x)(yij  y)

 This is just the sum of the multiplied residuals


for all data points.
Group Sum of Products

SPxygroup
 j
nj(xj  x)(yj  y)

 This is the sum of the products of the group


means minus the grand means times the group
size.
Error Sum of Products
SPxyerror
j i
(xij  xj)(yij  yj)

SPxyerrorSPxytotal SPxygroup

 This is the sum of the products of the DV and


residual minus the group means of the DV and
residual
 This just happens to be the same as the difference
between the other two sum of products
Adjusting the Sum of Squares
 Using the SS’s for the covariate and the
DV, and the SP’s, we can adjust the SS’s
for the DV
Sum of Squares

SSy adjtotal
SSytotal
SP total
xy
2

SSxtotal

SS adjgroup
y SS group
 
SP error


SP total
xy
2
xy
2

y
SSxerror SSxtotal
2
(SPxyerror
)
SSy adjerror
SSy error
SSxerror
Now what?
 Using the adjusted SS’s, we can now run
an ANOVA to see if there is a difference
between groups.
 This is the exact same as a regular
ANOVA, but using the adjusted SS’s
instead of the original ones.
 Degrees of freedom are not affected
A few more things
 We can also determine whether the
covariate is significant by getting a F score

 (SPxytotal)2 
  N  2
 SSxtotal 
F(1, N  2)   
SSy adjtotal
A few more things
 The group means can also be adjusted to
eliminate the effect of the covariate

 SPxyerror
adjy j y j   
 xj  x


 SSxerror
Post-hocs for ANCOVA
 Post-hoc tests can be done using the
adjusted means for ANCOVA, including
LSD and Bonferroni
Example of ANCOVA
 Imagine we gave subjects a self-esteem
test, with scores of 1 to 10
 Then we primed subjects with either
positive or negative emotions.
 Then we asked them to spend a few
minutes writing about themselves.
 Our dependent measure is the number of
positive emotion words they used (e.g.
happy, good)
Example of ANCOVA, cont.
 The null hypothesis is that the priming
doesn’t make a difference after controlling
for self-esteem
 The alternative hypothesis is that the
priming does make a difference after
controlling for self-esteem
Data
Subject # Priming Self-Esteem Positive Words
1 Positive 1 7
2 Positive 5 10
3 Positive 7 11
4 Negative 8 7
5 Negative 3 4
6 Negative 6 5
ANCOVA in SPSS
 To do ANCOVA in SPSS, all you need to
do is add your covariate to the “covariate”
box in the “univariate” menu
 Everything else is the exact same as it is
for ANOVA

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy