- A very beautiful lecture explaining the basic concepts underlying an F-ratio and F-distribution. This concept is critical for understanding the ANOVA analyses.
ANOVA (ANALYSIS OF VARIANCE):
- ANOVA – a visual introduction; a very useful video to understand the logic behind interpreting the variances which exist between groups and within groups in the context of the F ratio and these help in understanding the ANOVA results.
- One way ANOVA (single factor analysis) – Part 1; a visual guide. Also explaining the concepts of ‘sum of squares’ and ‘variances’ – both between group and within group.
- One way ANOVA (single factor analysis) – part 2 (understanding the calculation).
- Two way ANOVA (without replication) – Part 1; a visual guide.
- Two way ANOVA (without replication) – Part 1 (understanding the calculation) _ the video beautifully explains how to do a two way ANOVA (without replication) manually in Excel and also elaborates as to where different values come from?
- Two way ANOVA (with replication) – Part 1; a visual guide.
- Two way ANOVA (with replication) – “INTERACTIONS”
- Two way ANOVA (with replication) – “MARGINAL MEANS”
SIMPLE LINEAR REGRESSION:
- The very basics of a simple linear regression (we basically are trying to create a regression line which is best fitting the available data of dependent and independent variables and comparing it to the mean regression line for the dependent variables when there is no independent variable – watch this video to understand this very basic concept forming the basis of a simple linear regression analysis). Enjoy !
- A nice video to comprehend the algebra underlying the simple linear regression line (y = mx + b __ the slope-intercept line equation)
- Click over this link to listen to the explanation of the least square method/criterion employed in formulating the best-fit regression line model.
- Fit of the linear regression line model and the coefficient of determination
- A video on the multiple regression (Part 1) – The very basics – It can be considered as an extension of the simple linear regression. Now we have 2 or >2 independent variables in the regression model attempting to explain the variance of the dependent variable. However, the flipside of the picture is that it could potentially lead to two problem, namely; ‘overfitting’ of IVs into the regression model (when they dont explain the variation in the DV) and ‘multicollinearity’ due to a potential high correlation between the IVs themselves. Enjoy the video !
- A video on the multiple regression analysis (Part 2) – Preparation work required as a pre-requisite to the actual multiple regression analysis. This video talks about running correlations/scatterplots between the IVs and the DV and then between the IVs themselves as well. We should exclude the IVs from putting them into the multiple regression model if, either, the IV is not highly correlated with the DV or if any of the two IVs are highly correlated with each other (multicollinear) because they are redundant. Enjoy the video !
- Multiple regression (Part 3A); Evaluating basic models
- Multiple regression (Part 3B); Evaluating basic models