### Main Difference

ANCOVA and ANOVA are two statistical techniques for equating samples or groups on one or more than one variables. They are used to perform the same function but the method adopted is different. ANCOVA is more robust and unbiased as compared to ANOVA. ANCOVA is exactly like ANOVA, except the effects of a third variable are statistically “controlled out”. ANCOVA only uses a general linear model while ANOVA uses linear as well nonlinear models.

### What is ANCOVA?

ANCOVA is a statistical technique used to equate samples or groups on one or more than one variables. ANCOVA stands for “Analysis of Covariance”. It is an analysis technique which has two or more variables. Variables involved in ANCOVA should be at least one continuous and one categorical predictor variable. It is a test method for testing the effect of outcome variable after removing the variance. It uses covariant to improve its statistical power. ANCOVA implies that there is a linear relationship between dependent and independent variables.

### What is ANOVA?

ANOVA is a statistical technique used to equate samples or groups on one or more than one variables. ANOVA stands for “Analysis of Variance” in statistics. It is tested to check the presence of common mean among various groups. It is quite a useful test as compared to t-tests for such purposes. There are different types of ANOVA including ½ One-way ANOVA, ½ Factorial ANOVA, ½ Repeated measure ANOVA and MANOVA.

### Key Differences

1. ANCOVA uses covariant while ANOVA doesn’t use covariant.
2. A Distinguished feature of ANOVA is BG while in a case of ANCOVA, BG is divided into TX and COV variation.
3. Both ANOVA and ANCOVA use WG variation. In ANCOVA WG variation is divided by individual differences as COV while ANOVA uses it for individual features only.
4. ANCOVA is more robust and unbiased as compared to ANOVA.
5. ANCOVA is exactly like ANOVA, except the effects of a third variable are statistically “controlled out”.
6. ANCOVA only uses a general linear model while ANOVA uses linear as well nonlinear models.  