LUBS5901M01 Quantitative Methods for International Business 2018/2019
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
LUBS5901M01
Quantitative Methods for International Business
Semester One 2018/2019
Section A. There are two questions included in this section. Please answer all of them.
Question1.
A researcher aims to explore whether returnees’ international knowledge affects their entrepreneurial decisions in China. Returnees are defined as individuals who have studied and/or worked in foreign countries for a substantial period of time after graduation and then returned to their home countries. The survey data is collected from participants in the Guangzhou Convention of Overseas Chinese Scholars in Science and Technology in 2011. The convention has been held in Guangzhou every year since 1998 and has grown into one of the largest platforms for Chinese returnees to search for jobs, venture investments and cooperation opportunities from all over China. Through personal connections with the organizer, the researcher obtained a list of 2612 returnees who registered to attend the convention in 2011.
Evaluate the sampling strategy used in this research and propose an alternative method for selecting the samples.
[15 marks]
Question 2.
Your company aims to set up an oversea subsidiary to increase foreign sales. Two countries are considered by the top management team. In the past 12 months, your company has been exporting to these two countries and kept all the sales numbers (in £K):
Country A 25 30 28 40 120 35 33 38 40 30 33 110
Country B 50 55 70 58 65 68 60 66 58 67 68 70
The summary statistics of the two countries’ sale data are listed as follows:
|
Mean |
Median |
Mode |
Range |
Standard deviation |
Country A |
46.8 |
35.5 |
30 |
95 |
32.2 |
Country B |
62.9 |
65.5 |
70 |
20 |
6.5 |
Justify which country you company shall establish the overseas subsidiary in .
[15 marks]
Section B. There are two questions included in this section. Please answer all of them.
Question 3.
You have conducted a research project to investigate the determinants of global knowledge sourcing of overseas R&D subsidiaries (Dependent variable). Three factors are considered:
Vertical administrative embeddedness (measured by joint involvement of HQ and subsidiaries
in key administrative decision-making),
Vertical knowledge embeddedness (measured by joint research projects, transfer of
technology/knowledge, and movement of researchers and engineers between overseas R&D subsidiaries and MNC HQs),
Geographic distance (measured by the distance between the cities in which overseas
subsidiary and HQs were located).
Four hypotheses have been developed and tested based on a sample of 497 Japanese firms:
Hypothesis 1 (H1). Vertical administrative embeddedness (VAE) negatively affects global knowledge sourcing (GKS) by overseas subsidiaries.
Hypothesis 2 (H2). Vertical knowledge embeddedness (VKE) positively affects global knowledge sourcing (GKS) by overseas subsidiaries.
Hypothesis 3a (H3a). Geographic distance weakens the negative relationship between vertical administrative embeddedness (VAE) and global knowledge sourcing (GKS) of overseas subsidiaries.
Hypothesis 3b (H3b). Geographic distance weakens the positive relationship between vertical knowledge embeddedness (VKE) and global knowledge sourcing (GKS) of overseas subsidiaries.
Also, two control variables are included in the project.
Citation (the total number of citations made by a patent)
Inventor (the number of inventors)
The linear regression outcomes are presented in Table 1 (next page).
Table 1 Results
|
Model 1 |
Model 2 |
Model 3 |
Constant
Citation
Inventor
Geographic distance
VAE
VKE
VAE × Geographic distance
VKE × Geographic distance |
0.260 (4.928) -0.003 (0.004) 0.104** (0.049) |
-0.100 (3.535) -0.000 (0.004) 0.086 (0.054) -0.084 (0.433) -0.609*** (0.205) 0.421* (0.234) |
-39.220 (30.469) -0.002 (0.004) 0.080 (0.525) 4.159 (3.320) 1.524 (10.659) 11.918** (5.986) -0.220 (1.161) -1.253* (0.647) |
N |
497 |
497 |
497 |
R-square |
0.377 |
0.455 |
0.456 |
Adj. R-square |
0.305 |
0.410 |
0.414 |
Note. Standard errors in parentheses. * p < 0.05 ** p < 0.01 *** p < 0.001.
You are required to answer the following four questions:
a) Write down the statistical model for Model 3.
[3 marks]
b) Suggest whether the results in Table 1 are consistent with any of the four hypotheses. Please support your answer with the detailed analysis (in terms of the direction and statistical significance of the key variables).
[7 marks]
c) Define R-square and the adjusted R-square values and interpret those values in the three models.
[7 marks]
d) The managers want to know how the results of your project can help them increase their global knowledge sourcing of overseas R&D subsidiaries. How would you respond?
[5 marks]
Question 4:
A research project is conducted to examine the impact of perceived labour regulation obstacle (“LaborReg”) on labour productivity growth (“LPgrowth”) of Emerging Market Enterprises (EMEs). The labour regulation obstacle is measured by a 5-point Likert scale survey question, where 5 suggests the highest level of perceived obstacle.
The researchers consider three moderators that may affect the relationship between labour regulation obstacle and labour productivity growth:
Formal training (“Training”, measured by the years of formal training received by employees), Education (“Education”, measured by the average years of education of employees), Top Management Experience (“TMExperience”, measured by the years of top managers’
experience in the sector).
Two control variables are identified:
Whether the firm was formally registered when it began operation (“Registration”, measured
by a binary variable, where 1 = Yes and 0 = No),
Whether the firm competes against unregistered or informal firms (“InfCompetition”, measured
by a binary variable, where 1 = Yes and 0 = No).
The researchers conduct a linear regression based on 7524 EMEs. The results are illustrated as follows.
Table 2 Descriptive statistics
Table 3 Correlation matrix
Table 4 Model Summary
Table 5 Coefficients
where
LaborReg_Training is the interaction between LaborReg and Training (i.e., LaboTReg × TTaining);
LaborReg_Education is the interaction between LaborReg and Education (i.e., LaboTReg × Education);
LaborReg_TMExperience is the interaction between LaborReg and TMExperience (i.e., LaboTReg × TMExpeTience).
Figure 1 Histogram of residuals
Figure 2 Scatter plot of residuals
Figure 3 P-P plot
Based on the given information, you are required to answer the following question:
a) Evaluate the use of linear regression model by applying the five linear regression assumptions. And what managerial implications can we get from the results?
[34 marks]
The researchers further explore this project by categorizing the dependent variable (i.e., Labour productivity growth) into a binary variable, where LPgrowth=1 if its original value is positive and LPgrowth=0 otherwise. Accordingly, a logistic regression (i.e., logit model) is applied. The SPSS outputs are as follows.
Intercept-Only Model
Table 6 Coefficients for the Intercept-Only Model a
a. -2Log likelihood = 7089.658
Unconstrained Model
Table 7 Coefficients for the Unconstrained Model b
b. -2Log likelihood = 10206.824
You are required to answer the following questions:
b) Write down the statistical model for the unconstrained logit model.
[4 marks]
c) Compare the results in Table 5 (linear model) and those in Table 7 (logit model), explain their differences.
[5 marks]
d) Based on the SPSS outputs, would you recommend the use of the statistical model in b) to predict whether an EME’s labour productivity will increase? Suggest how to improve the existing model. [Note that X0(2).05,9 = 16.919]
[5 marks]
2022-08-06