Business Research Methods: Assignment-3 Mediation Analysis
Business Research Methods: Assignment-3 Mediation Analysis
Business Research Methods: Assignment-3 Mediation Analysis
Assignment-3
Mediation Analysis
Path Diagram
Equations
Equation-1
Income = α + β Education
Equation-2
Skills = α + β Education
Equation-3
Income = α + β Education
The above regression equation shows that income is the dependent variable and education is an
independent variable. The regression equation shows that if education increases by 1% income
will increase up to 3740.554076 times (Education impact on Income).
On the basis of t-Statistics, 3.243359 shows that effect of education on income shown on the
regression equation is significant and valid as this t-value is greater than two. On the other hand
if we look at P-value, 0.001278 shows that the impact of education on income explained in the
above equation is valid and significant as p-value lies between 0-10, considered as three stars and
chances of errors are extremely less. So the first equation is significant!
2. Regress x on M (Regress Education on Skills)
Skills = α + β Education
The above regression equation shows that skills are the dependent variable and education is an
independent variable. The regression equation shows that if education increases by 1% skills will
increase up to 0.012806078 times (Education impact on Skills).
If we look at the result of t-Statistics, 12.1985141 shows that the impact of education on skills in
the above equation is authentic, significant and valid too, because t-value is greater than two.
Moreover if we focus on P-value then 2.0525E-29 shows that the above regression equation is
significant and valid as chances of errors are very close to zero. Thus or second equation is
significant and valid too!
3. Regress x and M on Y (Regress Education and Skills on Income)
The above regression equation shows that income is the dependent variable and education and
skills are the independent variables and most importantly Skills are acting as an mediator. The
regression equation shows that if education increases by 1% then income will increase up to
1495.489461 times (Education impact on Income). Furthermore, if skills increase by 1% then
education will increase upto 175312.423 time (Skills impact on Income).
If we look at the result of t-Statistics of education, 1.123766 shows that the impact of education
on income in the above equation is insignificant and invalid too, because t-value is less than two.
On the other hand if we look at the t-Statistics of Skills, then 3.269221 shows the impact of skills
on income explained in the regression equation is authentic, significant and valid too as it is
greater than 2.
And if we focus on P-value of Education then 0.26177 or 26.177% shows that above regression
equation is insignificant and invalid and chances of errors are greater than 10. The P-value of
skills is 0.00116965 or 0.11% which shows that the relation of skills with income explained in
the above regression equation is valid and significant as P-value is close to zero and considered
as 3 stars!
Conclusion
Yes Mediation exists! On the basis of Mediation Analysis we can say that it is full mediation as
the x variable in the 3rd step is insignificant and mediator effect on Income is significant. We can
now say that education significantly effects and influences income through the channel of skills.