3 Practical Part
3.2 Analysis
45 2016, as a result of the acquisition of United Biscuits by the Turkish conglomerate Yildiz Holding. The company has 34 factories in 13 countries and it reaches approximately 4 billion people worldwide in 120 countries across Europe, the Middle East, Africa, Asia, North and South America. (Pladis Global, 2019)
Abdi Ibrahim is one of the oldest pharmaceutical companies in Turkey. The company today has more than 2.000 employees and continues its activities in Turkey as well as in oversees. It has partnerships in foreign markets such as Abdi Ibrahim Otsuka in Japan, Abdi Ibrahim Global Pharm in Kazakhstan and Abdi Ibrahim Remede Pharma in Algeria. These joint ventures are active in several markets such as Abdi Ibrahim Otsuka has been acitve in the markets Tunisia, Turkey and Algeria. (Abdi İbrahim, 2019)
ABC Detergents is one of the firms of ABC Holding. Today ABC Detergents is one of the well-known hygiene companies in the world. In fact the market share of the firm in some of the foreign markets is higher than its market share in Turkey.
The firm is specialized in the manufacture of powder detergent, chemical semi-finished goods, liquid products, cosmetics and bottle. (ABC Detergents, 2019)
46 Figure 6: Statistical Diagram Model 1, Hayes (2013)
As it is seen the statistical diagram there are three pathways: b1, b2 and b3.
B1, which is the first pathway, represents the direct effect of independent variable on the dependent variable. In this case the direct effect of uncertainty on the success of Turkish firms in foreign markets. B2, which is the second pathway, represents the direct effect of the moderator on the dependent variable. In this case the direct effect of networking of Turkish firms on the success in foreign markets. B3 is the interaction pathway, which is the interaction of independent variable and the moderator and their combined effect on the dependent variable.
The interaction can be considered as the multiplication of the paths b1 and b2.
Briefly, in the moderator analysis the conditional effect of independent variable X is seen on the dependent variable Y so in this case the conditional effect of uncertainty on the success of Turkish firms. The calculation of the effect is given in figure 7.
Y= 𝑏 ! + 𝑏 ! M
Figure 7, the conditional effect of X on Y
In the analysis of the responses, the size and the revenue of the company, which the respondent works for, were used as control variables. In the moderator analysis, the conditional effect of each one of the uncertainty environments through networking was calculated. In the following figures the effect of
47 networking along with each type of uncertainty separately and the interaction of both are given. Each model is given with its summary.
R R-sq MSE F
0.6524 0.4257 1.1131 12.5992
Figure 8, the model summary 1
In the following figure 9, the conditional effect of FDI-related environment uncertainty on the performance under the moderation of networking is given.
Coefficient Standard Error
t p LLCI ULCI
FDI 0.4408 0.3386 1.3019 0.1988 -0.2389 1.1205
Networking 0.7047 0.2641 2.6680 0.0102 0.1745 1.2350
Interaction -0.423 0.0626 -0.6765 0.5018 -0.1680 0.0833
Figure 9, the conditional effect of FDI-related environment uncertainty
As it seen from figure 9 the FDI-related environment uncertainty has P value of 0.1988 with zero between the levels of confidence interval, which means FDI-related environment uncertainty effect on performance is insignificant along with the interaction with networking. Interaction also has a high P value of 0.5018 with zero between the levels of confidence interval. While the FDI-related environment uncertainty and its interaction with networking have an insignificant effect on the performance, the networking alone has a significant effect on the performance with a low P value of 0.0102 and zero not being between the levels of confidence interval.
48 In the following figures of figure 10 and 11 the summary of the model and the conditional effect of macroeconomic environment uncertainty on the performance under the moderation of networking are given respectively.
R R-sq MSE F
0.6665 0.4443 1.0771 13.5895
Figure 10, the model summary 2
Coefficient Standard Error
t p LLCI ULCI
Macroeconomic 0.3673 0.3437 1.0684 0.2904 -0.3229 1.0574
Networking 0.5800 0.2325 2.4948 0.0159 0.1133 1.0468
Interaction -0.234 0.613 -0.3824 0.7038 -0.1465 0.0996
Figure 11, the conditional effect of macroeconomic environment uncertainty
In figure 11, the numbers display the result of macroeconomic environment uncertainty and its interaction with networking have an insignificant effect on the performance. Macroeconomic environment uncertainty has the P value of 0.2904 with zero between the levels of confidence interval. Interaction has the P value of 0.7038 with zero between the levels of confidence interval. When the networking is examined independently from macroeconomic environment uncertainty, it has a significant effect on the performance with the P value of 0.0159.
R R-sq MSE F
0.6617 0.4379 1.0894 13.2422
Figure 12, the model summary 3
49 Coefficient Standard
Error
t p LLCI ULCI
Material and Infrastructure
0.6114 0.3534 1.7302 0.0896 -0.0980 1.3208
Networking 0.7076 0.2772 2.5524 0.0137 0.1510 1.2641
Interaction -0.0630 0.637 -0.9879 0.3278 -0.1909 0.0650
Figure 13, the conditional effect of material and infrastructure environment uncertainty
In the figures of figure 12 and 13 the summary of the model and the conditional effect of material and infrastructure environment uncertainty on the performance under the moderation of networking are given respectively.
In figure 13, the conditional effect of material and infrastructure environment uncertainty on performance through networking is seen. As it looks from the table the interaction of material and infrastructure environment uncertainty and networking has an in significant effect on performance with the P value of 0.3278 and zero between the levels of confidence interval. The networking however, has a direct significant effect on the performance with the P value of 0.0137 and level of confidence interval not containing zero.
In the following figures of figure 14 and 15 the summary of the model and the conditional effect of consumption environment uncertainty on the performance under the moderation of networking are given respectively.
R R-sq MSE F
0.6639 0.4407 1.0839 13.3965
Figure 14, the model summary 4
50 Coefficient Standard
Error
t p LLCI ULCI
Consumption 0.7545 0.3606 2.0922 0.0414 0.0305 1.4785
Networking 0.7339 0.3064 2.4136 0.0194 0.1244 1.3549
Interaction -0.0790 0.672 -1.1762 0.2450 -0.2139 0.0558
Figure 15, the conditional effect of consumption environment uncertainty
By looking at the figure 15, the interaction of consumption environment uncertainty and networking has an insignificant effect on the performance with the P value of 0.2450. The consumption environment uncertainty alone has a significant effect but the P value of 0.0414 is very close 0.05, so it can be said that the effect is barely significant. Again the networking alone has a significant effect on the performance with the P value of 0.0194 and the level of confidence interval not containing zero.
R R-sq MSE F
0.6485 0.4205 1.1230 12.3376
Figure 16, the model summary 5
Coefficient Standard Error
t p LLCI ULCI
Competitive 0.6363 0.3754 1.6950 0.0962 -0.1174 1.3900
Networking 0.8193 0.2757 2.9719 0.0045 0.2658 1.3727
Interaction -0.0786 0.675 -1.1650 0.2494 -0.2140 0.0568
Figure 17, the conditional effect of competitive environment uncertainty
51 In the figures above, figure 16 and 17, the summary of the model and the conditional effect of competitive environment uncertainty on the performance under the moderation of networking are given respectively.
Figure 17 displays the insignificant effects of competitive environment uncertainty and its interaction with the networking with the P values of 0.0962 and 0.2494 respectively and level of confidence intervals, which are containing zero. Likewise in the previous models, networking alone has a significant effect on the performance with the P value of 0.0045 and level of confidence interval, which isn’t containing zero.
In the following figures of figure 18 and 19 the summary of the model and the conditional effect of total environment uncertainty on the performance under the moderation of networking are given respectively. These two figures represent the total outcome of the other models and the total conditional effect of the types of uncertainty environments on performance under the moderation of networking.
R R-sq MSE F
0.6831 0.4666 1.0338 14.8705
Figure 18, the model summary 6
Coefficient Standard Error
t p LLCI ULCI
Total Uncertainty
0.7474 0.3723 2.0072 0.0500 -0.001 1.4949
Networking 0.6449 0.2709 2.3806 0.0211 0.1010 1.1887
Interaction -0.0623 0.657 -0.9492 0.3470 -0.1942 0.0695
Figure 19, the conditional effect of total uncertainty
52 In figure 19, the total conditional effect of uncertainty on performance through networking is given with the numbers. Starting from the direct effect of total uncertainty on performance, the numbers show the result of the effect is barely significant with the P value of 0.005 and the levels of confidence interval involves zero on the edge. The direct effect of networking alone on performance is significant, likewise in every previous model, with the P value of 0.0211 and levels of confidence interval don’t contain zero. The effect of the interaction of total uncertainty and networking has an insignificant effect with the high P value of 0.3470 and the level of confidence interval containing zero.