Section 2 :
Analyze factorial
The factorial analysis is a geometrical principle making it
possible automatically to convert a table of figures into synthetic images,
which release the principal structures of them. The first stage consists in
translating the table in a space in the shape of a group of dots, the second
stage acts to seek to visualize the clouds obtained on best possible plane
projections, one thus chooses, best projection to form the factorial axes.
The variable points all are in this cloud materialize well all
the phenomena of correlation between the variables of the table.
Our research releases the following results ;
It is noticed that KMO = 0.638 >0.6 thus these items can
translate the phenomenon to be studied.
While referring in the table of original variance (see
appendix n°5) one retained two axes according to the criterion eigenvalue
> 1, the first axis explains 40.428% of the original variance of the sample
and the second axis explains 21.706% of the original variance.
The system of axes formed explains 62.133% of the original
variance.
The graph shows well that the first two axes are those whose
eigenvalue is higher than 1 and thus they form the factorial design.
It is noticed that the first axis is formed by the number of
employment with 84.5% and the authorized capital with 81% whereas the second
axis represents nationality with a factorial contribution of 89.1%.
The rate of participation can improve, therefore one carries
out a rotation varimax to optimize the result and to seek a better positioning
of the variables around the axes.
After rotation, one notices that the year of presence can also
contribute to form the first axis since its factorial contribution is higher
than 50% (67.1%), in the same way for the rate of participation which can in
its turn form the second axis with a coefficient of structure = 55%.
Thus the first axis expresses the size and the experiment of
the company which are the structural factors and the second axis expresses in
its turn the level of control and nationality.
According to this table one can conclude that the two axes are
not correlated (see appendix n°5).
Section 3 :
Analyze of Khi two
1 it relation between the nationality and the mode
of establishment
The test of ÷2 shows well the existence of a relation
between the variable nationality and the rate of foreign participation since
÷2observ? (24.148) > ÷2tabul? (6.304) to 12 degrees of freedom and
when á =10% thus the two variables are dependant (see appendix
n°6).
2 it relation between the size of the company and
the mode of establishment
2.1 it relation between the capital of the company
and the mode of presence.
According to the result of the test of ÷2 one
notices that ÷2observ? (with 12 degrees of freedom) is 30.823 whereas
÷2tabul? is only 6.304 with the threshold of á =10%, one thus finds
that ÷2observ? higher than one is ÷2tabul? then rejects H0 (the two
variables are independent) and one accepts H1 our assumption of research which
means the existence of relation between the two studied variables (see appendix
n°6).
2.2 it relation between the number of employees and
the mode of presence
One notices that ÷2observ? (22.494) > ÷2theoric
(6.304) to 12 degrees of freedom thus one rejects H0 (not relation between two
variables), which leads us to conclude that there is a connection between the
number of employees and the mode of presence, therefore that confirms our
assumption of research (see appendix n°6).
3 it relation between the number of years of
presence and the mode of establishment
According to table Ci above one notices that ÷2observ?
who is 16.976 is higher than ÷2theoric which is 6.304 with the threshold
of á =10% and with 12 degrees of freedom then one rejects the H0
assumption and one accepts H1, that confirms our assumption of research which
supposes that there is a relation between the experiment (calculated by the
number of years of presence) and the choice of the mode of presence which
expresses at the same time the level of control and the level of investment
(calculated by the rate of foreign participation) (see appendix n°6).
Synthesis
At the time of this empirical study we could confirm our
assumptions of research by an analysis of the frequencies, a factorial analysis
and an analysis of Khi two.
The size of the company that we studied through the variables
capital and a number of employees strongly explains the selected mode and
influences in great part the level of control held by the company.
The rate of foreign participation reflects the capital invests
and the level of control, one noticed that there is a strong relation between
control and the TPE.
The experiment of the company measured by the number of years
of presence expresses the training obtained in the field of the IDE and which
influences the behavior of the choice of the mode of presence.
| 
