Business Research Methods: Regression and Significance Analysis

This assignment solution focuses on business research methods, specifically addressing statistical significance through regression and logistic analysis. The first question analyzes an OLS regression of US real GDP growth rates on oil prices, interest rates, and inflation, discussing the statistical significance of the parameters, interpreting the sign and magnitude of the estimates, and evaluating the overall fit of the model. It also assesses whether the results align with theoretical predictions. The second question examines a logistic regression model predicting promotion probability based on age, experience, and sex, evaluating the significance of each parameter and the model's usefulness for prediction. The solution provides detailed interpretations of the coefficients and their implications, concluding with an assessment of the model's overall fit and predictive power. Desklib provides similar solved assignments for students.

BUSINESS RESEARCH METHODS
STUDENT ID:
[Pick the date]

Question 1
(a) The statistical significance of the parameters can be estimated from the p value associated
with the slope coefficients of the different independent variables. Assuming a significance
level of 5%, it is apparent that p value of slope cficients of all the parameters except inflation
is lower than 0.05. This implies that all parameters except inflation are statistically significant
as the underlying slope cannot be assumed to be zero (Flick, 2015).
The slope coefficient for oil is negative as there tends to be an adverse relationship between
real GDP and oil prices. The slope magnitude is 0.037 which implies that a change in the oil
prices by $1 per barrel would tend to change the real GDP growth rates is US by 0.037 units.
The slope coefficient for interest rate (INTERSTRATE) is negative which is on expected
lines as higher interest rates tend to make credit expensive and adversely impact economic
growth. The slope magnitude is 0.012 which implies that a change in the interest rate by 1%
would tend to change the real GDP growth rates is US by 0.012 units. The slope coefficient
for inflation is negative as higher inflation would lead to lower real GDP growth. The slope
magnitude is 0.004 which implies that a change in the inflation rate by 1% would tend to
change the real GDP growth rates is US by 0.004 units (Hair et. al., 2015).
The R2 value for thee given regression model is 58% which implies that the given
independent variables can jointly explain 58% of the variation in the real GDP growth rate in
US. Considering that two of the three slope coefficients are significant coupled with a
moderately high R2 value would imply that the model is a good fit (Hillier, 2016).
(b) The results are broadly in line with the theory considering that higher inflation, interest
rate or oil prices would tend to have adverse impact on growth. The increase in all these
factors tend to lead to lower demand and potentially have adverse impact on consumption
owing to which adversely impact real GDP growth. Inflation for instance would increase
nominal GDP growth but since higher prices would adversely impact consumption volume,
hence real GDP growth would be lower. Similarly, higher interest rates tend to lower lending
owing to which there is a slowdown in demand and hence real GDP growth is adversely
impacted.
Question 4

(a) The slope coefficient of age variable is positive in sign which implies that the probability
of promotion increases with age which is quite natural as it implies more experience. The
magnitude of slope coefficient of age variable is 0.035 which implies that as there is change
in age by 1 unit, there would be a change in the promotion probability by 0.035. The slope
coefficient of experience variable is positive in sign which implies that the probability of
promotion increases with experience which is quite expected (Eriksson and Kovalainen,
2015). The magnitude of slope coefficient of experience variable is 0.148 which implies that
as there is change in experience by 1 unit, there would be a change in the promotion
probability by 0.148. The slope coefficient of sex variable is negative in sign which implies
that females are less likely to be considered for a promotion than their male counterparts. The
magnitude of slope coefficient of sex variable is 0.986 which implies that there is significant
difference in the promotion probabilities associated with the two genders assuming other
variables are same (Flick, 2015).
With regards to statistical significance, all the parameters in the given logistic regression are
non-significant assuming the underlying significance level of 5%. This is because the p
values for the respective coefficients for all the parameters are greater than 5% and hence the
underlying coefficients can be considered as zero (Hillier, 2016).
(b) Thee given logistic regression model would not be considered as a useful model for
predicting the chances of promotion considering the fact that neither of the three slope
coefficients is statistically significant. This implies that the predictive power of the given
model is quite low owing to which the R2 for this model would be quite low. As a result, the
estimates produced from this model would have a significant deviation from the actual model
and hence the model is a bad fit (Hair et. al., 2015).

References
Eriksson, P. and Kovalainen, A. (2015). Quantitative methods in business research (3rd ed.).
London: Sage Publications, pp. 132-134
Flick, U. (2015). Introducing research methodology: A beginner's guide to doing a research
project (4th ed.). New York: Sage Publications, pp. 67-72
Hair, J. F., Wolfinbarger, M., Money, A. H., Samouel, P., and Page, M. J. (2015). Essentials
of business research methods (2nd ed.). New York: Routledge, pp. 103-106
Hillier, F. (2016). Introduction to Operations Research. (6th ed.). New York: McGraw Hill
Publications, pp. 74-77