Question

Difference between mediating, moderating, and confounding variable in research methodology ? Please give examples ( Explain in detail)

Answer 1

Mediating variable : It causes conciliation between dependent and independent varibales. It explains the relationship between dependent and independent variable.

Moderating Variable : It helps in influencing the strength of a relationship between two varibales.i.e. dependent and independent variables.

Confounding Variable : It is an external influence affecting the relationship between the dependent and independent variables, thus resulting in false association between them . It can ruin the outcome , and hence can produce false results .

Examples :

Moderator variable is often denoted as M in statistics. It is a third variable, thus affecting the strength of relationship between 2 variables; i.e. dependent and independent varibale which is found in correlation . If x is the predictor and y is the outcome,M is considered to be the moderator . Moderator can be quantitative or qualitative variable as well, affecting the direction/strength of the relationship between two variables. Independent varibale is the predictor and the dependent variable is the outcome variable.

If a researcher wants to know the effect of a new cholestrolol drug . Prior to this , the participants have done exercise . Accordingly, the researcher takes a note after a month , on how they are fairing and what effects are they having after having the new cholestrolol drug.  

Hence , the researcher found that at low drug dose , there is a small association between exercise and cholesterol levels. Whereas in case of high drug dose , there is a remarkable increase in the association between exercise and cholestorol level. In this case , the dosage of drug moderates between exercise and level of cholesterol .

Mediating variable :

In case Data is provided , regarding the age group and frequency of running capacity of the people among those age groups. We need to find a relation between them , and hence showcase that a relationship do exist between them. Like the age group between 20-30 , the frequency is more as compared to the age group between 60-70.

Confounding variable : Lack of exercise leads to weight gain. Here lack of exercise is the independent variable and weight gain is the dependent variable . Confounding variables are like another kind of variable which affects the dependent variable. They have two problems. They introduce biasness and increases variance .

In this case particularly the confounding variables which are hidden , they are :

1. How much do people eat?

2. Sex (i.e. Male/Female) can also be a confounding variable. Men can even gain weight due to lack of exercise , not necessarily it has to be women. The researcher cannot be gender specific, without having full information.

3. No records of weight, age, occupation has been mentioned.

Hence , random data could lead to many problems and can produce false outcomes.