Pilot Test Descriptive Statistics
A pilot study was conducted in advance before collecting the sample for the main study. This test was conducted to test and confirm whether the instrument was valid and reliable for use it as the quantitative tool to accomplish the research. However, for the main study, a new dimension was included and therefore, the measurement instrument was modified including the necessary elements for covering the new dimension.
Table 4.1 shows the demographic characteristics of the sample collected for running the pilot study. The biggest difference with the main sample relies on the inclusion of participants from different industries in the sample. Although 60% of the participants included in the pilot sample belonged to the manufacturing industry, the researcher with the objective of testing the validity and reliability in a wider scope, the sample was not limited to the manufacturing sector.
Moreover, the pilot test included different rank employees, creating a good representation of the inner body from the organizations surveyed.
58 Table 4.1
Pilot Test Demographic (N=32)
Variable Ranges Frequency Percent
Age
The descriptive statistics of the sample collected for running pilot test analyses are shown in Table 4.2. From this small size simple the statistics named the variables GCL and GCAC as the ones with highest means (3.99). On the other side, the factor BPGLP showed up as the variable which presents the lowest mean (3.16).
59 Table 4.2
Pilot Test Descriptive Results (N=32)
Variable Mean Std.
Deviation Variable Mean Std.
Deviation
GCHR 3.85 .78 CSRREM 3.24 .65
GCL 3.99 .64 CSRSM 3.48 .64
GCE 3.59 .74 CSRCR 3.32 .91
GCAC 3.99 .82 BPFP 3.59 .68
CSRV 3.78 .61 BPCP 3.47 .79
CSRMS 3.44 .71 BPPP 3.23 .75
CSRPS 3.55 .68 BPLGP 3.16 .79
Pilot Test’s Validity and Reliability
In Table 4.3 are shown the results of the exploratory factor analysis conducted with the pilot test sample. These results expressed that the factors included in the measurement instrument are appropriate, with KMO values between 0.5 and 1.0 confirming the acceptance of using the factors according on how they are distributed.
Furthermore, the Barlett’s Test of Sphericity confirms what revealed by KMO analysis.
On the other side, the factor loading tested for the confirmatory factor analysis support each of the factors included in the GCSR Model. Moreover the Cronbach’s Alpha, Composite reliability and Average Variance Extracted assure the validity and reliability of the dimension and variables.
The pilot test conducted, showed the significance of the factors included in each of the dimensions as well as the significance of the path coefficients that link the dimensions.
This GCSR Model developed by (Nguyen, 2014) settled the base for creating the Glocal CSR model in which the main study is conducted (see Appendix D).
60 Tabla 4.3
Pilot Test validity and reliability analysis
KMO
Note. *** Significance level p < .01
As gathered in table 4.3, all the factors loadings were higher than 0.6 assuring the internal consistency of the factors included in each of the dimensions.
In the case of Cronbach’s Alpha statistics, the values of each of the dimensions are higher than 0.7 and the Composite reliability higher than 0.8 confirming the reliability of the model.
For checking the validity of the model in a confirmatory way, the Average Variance Extracted was tested giving values higher than 0.5 in each of the dimensions and hence supporting the validity of the instrument and structural equation model.
61 Table 4.4
Pilot Test Path Hypotheses Results
Note: Path Coefficients & T-values: * p < .1, ** p < .05, *** p < .01
From the pilot test, the researcher rejected the null hypothesis after the results exposed by SmartPLS software (see Apendix D, Figure 4.1). As written in Table 4.4 there are positive and direct effects between the United Nations Global Compact and the Corporate Social Responsibility dimensions. Besides, the results showed the positive and direct effects of Corporate Social Responsibility on Business Performance.
The results support the studies from Carrol & Shabanna (2010), Mishra & Suar, (2010), Nguyen (2014), UNGC, DNVGL & MM&S (2015) which findings showed the positive influence of CSR on the Business Performance of organizations.
Main Study Descriptive Statistics
The main research was conducted approaching two different populations. The overall sample formed by the Spanish and Taiwanese participants is composed by N=144, and the characteristics and demographic of the sample are shown in Table 4.5, while Table 4.6 shows the correlations between the independent variables, the dependent variables and the demographic variables included in the research.
The characteristics of the industry included in the sample, the participants considered all their companies belong to the manufacturing sector, even if the companies included in the research belong to different kind of manufacturing organizations.
Path Hypothesis β-path Adj.
t-value Sig. Direction Null Hypotheses
UNGCCSR Ho2 0.666 7.507 *** + Reject
CSRBP Ho4 0.713 9.047 *** + Reject
62 Table 4.5
Main Study Demographics (N=181)
Variable Ranges Frequency Percent
Age
Industry Manufacturing 181 100
From the Table 4.5, we observe the demographic characteristics of all the samples included in the research. From this data and regarding the Level of Studies we remark that 48% of the participants are Master/Graduate and therefore we can assume that they are great connoisseurs of the terms included in the questionnaire.
Besides the level of studies, the participants included in the research are composed by 65% of employees, 30% of middle managers and 5% of top managers or executives. From these numbers, we can create a good representation of all the stratus that conform an organization, and therefore, more accurate findings can be extracted from this research.
Fitting the research purposes, it is important to highlight that gathering self-reports from employees who represent the base of organizations, middle managers that are in contact with employees and top management, and count with samples that
63
represent top managers of organizations, make this research to better represent the reality of the targeted population.
From the correlations shown in table 4.6, we observe that the variables CSRMS and the variable CSRV present a correlation of .83, while the variables CSRPS and CSRMS present a correlation of .80. However, the previous exploratory factor analysis discarded that multicollinearity will affect the effective predictive power of the Glocal CSR Model as a whole (Nguyen, 2014); it may only affect when the variables are taken from individual correlation tests.
64
Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
1 Age -
2 Level of Studies .04 -
3 Experience .85** .07 -
4 Rank Employee .55** .02 .57** -
5 GCHR .06 -.06 .07 .12 -
6 GCLR .08 .03 .109 .17* .59** -
7 GCE .21* -.05 .221 .20* .67** .45** -
8 GCAC .05 .15 .075 .06 .51** .28** .58** -
9 LIE .09 .00 .154 .05 .41** .35** .52** .48** -
10 LIS .08 .09 .116 -.04 .48** .27** .51** .51** .59** -
11 LIMLB .16 .04 .28* .03 .39** .28** .47** .44** .60** .53** -
12 CSRV .11 .07 .14 .13 .53** .39** .60** .65** .66** .63** .71** -
13 CSRMS .13 .11 .11 .18* .55** .51** .57** .61** .65** .55** .61** .83** -
14 CSRPS .05 .07 .01 .08 .41** .39** .47** .52** .56** .54** .50** .75** .80** -
15 CSRREM .05 -.06 .05 .15* .44** .27** .55** .46** .46** .50** .41** .54** .48** .46** -
16 CSRSM .17* .04 .18* .21** .41** .40** .54** .44** .54** .56** .54** .71** .75** .73** .45** -
17 CSRCR .02 .01 .01 .10 .44** .37** .57** .48** .66** .61** .63** .74** .72** .70** .60** .72** -
18 BPFP .03 -.07 -.01 .12 .27** .18* .29** .33** .39** .30** .19** .39** .46** .44** .18** .44** .37** -
19 BPCP .11 .00 .06 .09 .31** .22** .28** .41** .39** .34** .31** .44** .48** .41** .14 .46** .37** .65** -
20 BPIBP .21** -.19* .17* .17* .39* .27** .47** .32** .45** .43** .41** .52** .54** .57** .44** .57** .53** .59** .17** -
21 BPLGP .24** -.15 .22** .30** .43** .42** .52** .45** .48** .39** .45** .57** .63** .57** .41** .63** .54** .57** .60** .73** -
Note: Correlations significance level at *p<0.05 **p<0.01
Table 4.6
Main Study Correlation Results (N=181)
65
As it can be appreciated in Table 4.6, the variables used to create the Glocal CSR Model correlate between them. Therefore, after the exploratory factor analysis was done, we can affirm once again the consistency of the dimensions and variables extracted from the literature review.
In the Table.4.7 we can appreciate the means and standard deviations of the variables included in the research. Analyzing the descriptive data, the results show how GCLR (Global compact labour rights) has the highest mean (4.08) among the rest of variables, while the variable which presents the lowest mean is the CSRREM (Corporate Social Responsibility Resources and Environment Management) with a value (3.25).
Table 4.7
Descriptive Statistics of Main Study (N=181)
Variable Mean Std.
Deviation Variable Mean Std.
Deviation
GCHR 3.99 .64 CSRPS 3.60 .82
GCLR 4.09 .66 CSRREM 3.31 .97
GCE 3.64 .81 CSRSM 3.52 .72
GCAC 3.90 .86 CSRCR 3.47 .92
LIE 3.68 .71 BPFP 3.58 .77
LIS 3.59 .82 BPCP 3.62 .82
LMLB 3.27 .84 BPIBP 3.46 .80
CSRV 3.65 .86 BPLGP 3.43 .90
CSRMS 3.43 .88
Spain’s Descriptive Statistic
In the case of Spain, the sample was composed by N=99, however, as explained in Chapter 3, after statistical analysis the sample dropped to N= 81 which were the number of questionnaires considered valid for analysis. From this 81 the demographic and sample characteristics are described and represented in Table 4.8.
66 Table 4.8
Spanish Sample Demographics (N=81)
Variable Ranges Frequency Percent
Age
Industry Manufacturing 81 100
Table 4.9 represents the descriptive statistics of the variables included in the Glocal CSR Model from the Spanish surveyed organizations. As the results have shown, in the Spanish sample it can be appreciated higher variability of the means when comparing them with the Taiwanese ones (Table 4.11). Variables as GCE, GCAC, LIS, CSRV, CSRMS, CSRREM, CSRCR, BPFP, BPCP and BPLGP present standards deviations higher than 0.9 or 1.0, which express the great variability of the self- perceptions of the variables tested.
From these results, we can realize that the Spanish sample can be defined as a heterogeneous sample, with participants having wider opinions about the organizations in where they work.
On the other side, the variables containing the higher means are GCHR (4.06) and GCLR (4.11), which also present the lowest std. deviations, and therefore shows
67
how Spanish organizations present homogeneous considerations about these variables.
On the other side, the variable that has the lowest mean is LMLB (3.00) Table 4.9
Spanish Sample Descriptive Statistics (N=81)
Variable Mean Std.
Deviation Variable Mean Std.
Deviation
GCHR 4.06 0.63 CSRPS 3.56 0.85
GCLR 4.11 0.68 CSRREM 3.46 1.03
GCE 3.64 0.92 CSRSM 3.39 0.79
GCAC 3.74 0.95 CSRCR 3.48 0.99
LIE 3.61 0.74 BPFP 3.64 0.90
LIS 3.51 0.96 BPCP 3.49 0.94
LMLB 3.00 0.89 BPIBP 3.54 0.82
CSRV 3.50 1.03 BPLGP 3.41 1.01
CSRMS 3.30 0.98
Taiwan’s Descriptive Statistics
In the case of Taiwan, the demographic characteristics of its sample N=100 are described in Table 4.10.
With 60% of master and graduate students included in the sample, 58% of employees and more than 40% of middle and top managers, the sample reflects a representation of the whole body of the organizations surveyed
Furthermore, 79% of the respondents have at least 5 years of working experience in the current companies, which can also reflect higher validity and reliability due to their knowledge about their organizations. As shown in the demographic variable industry, all the samples collected belong to the manufacturing industry.
68 Table 4.10
Taiwanese Sample Demographics (N=100)
Variable Ranges Frequency Percent
Age
Industry Manufacturing 100 100
In Table 4.11 the mean and standard deviations of the Taiwanese sample are shown. As it happened with the main study, in the case of Taiwan the highest mean is the GCLR (4.07) having the lowest standard deviation from all the variables included in this research.
On the other side, the lowest mean presented by the variable CSRREM (3.18) it also concurs as the lowest mean of the main study, while its standard deviation with a value of .91 represents the highest variability among all the factors. Therefore and as the numbers show, the variable CSRREM can be considered the most diverse factor among the surveyed organizations.
69 Table 4.11
Taiwan’s Sample Variables Descriptive Statistics (N=100)
Variable Mean Std.
Deviation Variable Mean Std.
Deviation
GCHR 3.94 .64 CSRPS 3.62 .79
GCLR 4.07 .63 CSRREM 3.18 .91
GCE 3.64 .71 CSRSM 3.61 .64
GCAC 4.03 .75 CSRCR 3.45 .86
LIE 3.78 .67 BPFP 3.52 .65
LIS 3.70 .64 BPCP 3.72 .69
LMLB 3.59 .67 BPIBP 3.39 .78
CSRV 3.77 .68 BPLGP 3.44 .81
CSRMS 3.53 .77
Different Rank Employees Descriptive Statistics
To contrast null hypothesis 8 (Ho8), the research segmented the main sample (N=181) into different rank employees -employees, middle managers and top managers-.
The descriptive statistics of the perception from the different employees’ categories are introduced in Table 4.12, 4.13 and 4.14.
Due to further analysis, in this section there are only presented the descriptive statistics, meanwhile in Chapter 5, the results from the comparison of the average between different rank employee groups are introduced after the completion of the pair t-test analysis, and therefore, these differences are explained in detail.
70 Table 4.12
Employee Perceptions Descriptive Statistics (N=111)
Variable Mean Std.
Deviation Variable Mean Std.
Deviation
GCHR 3.98 0.65 CSRPS 3.62 0.78
GCLR 4.04 0.67 CSRREM 3.25 1.04
GCE 3.57 0.81 CSRSM 3.44 0.70
GCAC 3.90 0.85 CSRCR 3.45 0.90
LIE 3.66 0.76 BPFP 3.53 0.80
LIS 3.63 0.82 BPCP 3.59 0.88
LMLB 3.24 0.89 BPIBP 3.40 0.86
CSRV 3.62 0.84 BPLGP 3.27 0.97
CSRMS 3.38 0.88
Table 4.13
Middle Manager Perceptions Descriptive Statistics (N=55)
Variable Mean Std.
Deviation Variable Mean Std.
Deviation
GCHR 3.96 0.65 CSRPS 3.42 0.88
GCLR 4.13 0.61 CSRREM 3.34 0.84
GCE 3.66 0.77 CSRSM 3.58 0.75
GCAC 3.81 0.91 CSRCR 3.45 0.95
LIE 3.76 0.60 BPFP 3.57 0.71
LIS 3.48 0.91 BPCP 3.63 0.73
LMLB 3.39 0.76 BPIBP 3.56 0.67
CSRV 3.61 0.92 BPLGP 3.64 0.65
CSRMS 3.38 0.83
71 Table 4.14
Top Management Perception Descriptives Statistics (N=9)
Variable Mean Std.
Deviation Variable Mean Std.
Deviation
GCHR 4.38 0.32 CSRPS 4.37 0.45
GCLR 4.53 0.65 CSRREM 3.81 0.84
GCE 4.49 0.44 CSRSM 4.11 0.39
GCAC 4.49 0.35 CSRCR 3.71 1.00
LIE 3.72 0.88 BPFP 4.19 0.53
LIS 3.79 0.43 BPCP 3.86 0.40
LMLB 3.00 0.81 BPIBP 3.63 0.75
CSRV 4.28 0.53 BPLGP 4.22 0.49
CSRMS 4.35 0.56
72
73