Analysis Of Covariance (ANCOVA)

ANCOVA is a blend of ANOVA (analysis of variance) and regression. It is the same as ANOVA, but the difference is that it estimates coalition between independent variables. It is used to determine the differences in the mean values of three or more independent groups while taking into account the influence of the uncontrolled independent variables. To cheque uniformity invariance/regression slopes, ANCOVA heavily uses assumption testing.

It is used widely in businesses to determine the variation in the intention of customers buying a particular brand rather than of another, concerning different levels of price and the consumer’s attitude towards that brand.

It also shows the stats about how a little change in the price level of a particular commodity will affect the consumption of that commodity by the consumers. ANCOVA consists of a minimum of one categorical independent variable and one interval natured independent variable. The categorical independent variable is known as a factor while the interval natured independent variable is termed as a covariate. The independent variable covariate helps in removing the extraneous variation from the dependant variable. This analysis is done while the independent variable has a powerful correlation analysis with the dependant variable. It is majorly applied in those cases where the balanced independent variable is measured on a continuous scale.

During the analysis, ANCOVA assumes the following assumptions:-

The variance in the Analysis of covariance that is being analysed must be independent.
The variance in the Analysis of covariance (ANCOVA) must be homogeneous within each cell, in the case of more than one independent variable.
The data for the Analysis of covariance is received from the random sampling.
This process of ANCOVA is done through linear regression analysis. It indicates that ANCOVA assumes a linear relationship between the independent variable and the dependant variable
In the Analysis of covariance, the independent variables are assumed to be drawn from the normal population having a mean of zero.
It assumes that the multiple-regressions analysis coefficients in every group of the independent variable must be homogeneous.

Purposes of ANCOVA

In experimental designs, ANCOVA is used to control for factors that cannot be randomised but can be measured on an interval scale.
In observational designs, ANCOVA helps remove the effects of variables and thus improve the relationship of the categorical independents to the interval dependant.
In regression models, ANCOVA itself fits regressions where there are both categorical and interval independents.

Steps for ANCOVA

Run a regression between the independent and dependant variables.
Identify the residual values from the results.
Run an ANOVA on the residuals.

We at SPSS tutor helps the researcher to determine the effect of in-store promotion on sales revenue. For this, the appropriate technique of ANCOVA is used, where the dependant variable will be the sales revenue of the store and the independent variable will be the attitude of the consumer.