CHAPTER IV DATA ANALYSIS AND RESULTS
This chapter provides the results of the data analysis in this study. It provides the correlation of the variables and the results of the statistical analysis used to test the hypotheses.
Correlation
Pearson’s Correlation coefficient was used to examine the association between the demographics, the control variables, the independent variables (motivational factors, networks, human capital and environmental factors) and the dependent variable (performance). Table 4.1 shows the correlation analysis among the variables. The results indicate that several variables display significant correlation with each other.
As was expected almost all of the stated hypothesis, showed significant correlations.
Of the hypotheses related to motivations, economic motivation was positively correlated with performance (r=0.363, p<0.01), personal motivation also had a positive correlation with performance (r=0.233, p<0.01) and social motivation as well (r=0.345, p<0.01). Of the network related hypothesis, only mentors was positively correlated with performance (r=0.338, p<0.01). Of the human capital factors, business skills and the influence of the area of education were both positively correlated with performance (r=0.479, p<0.01) and (r=0.464, p<0.01) respectively. Level of education was also positively correlated with performance (r=0.317, p<0.01). Of the environmental dimensions, only perceived support from government was positively correlated with performance (r=0.167, p<0.05).
The current number of employees (size of the venture), the control variable, was positively correlated with performance (r=0.229, p<0.01). Even though this was not hypothesized in this study, Kalleber and Leicht (1991) showed that these two have a
relationship. This could indicate that as the female entrepreneurs increase their number of employees, their performance becomes better.
Regarding the correlations among the independent variables, none had an extraordinary high coefficient therefore there is no serious concern of multicollinearity among them. Economic motivation is positively correlated with personal (r=0.353, p<0.01) and social (r=0.364, p<0.01) motivation. It is also correlated with perceived support from women associations (r=0.310, p<0.01) and mentors (r=0.246, p<0.01). This could mean that the more economically motivated these women are, the more aware they are of their problems and weaknesses, thus they seek for support from women associations and mentors.
Results also show economic motivation negatively correlated with country insecurity (r=-0.192, p<0.05).
Personal motivation is positively correlated with social motivation (r=0.517, p<0.01), perceived support from women associations (r=0.226, p<0.01) and mentors (r=0.397, p<0.01) and negatively correlated with country insecurity (r=-0.248, p<0.01). Social motivation is positively correlated with perceived support from women associations (r=0.261, p<0.01), mentors (r=0.392, p<0.01), business skills (r=0.236, p<0.01) and the influence of the area of education (r=0.276, p<0.01).
Perceived support from women associations is positively correlated with mentors (r=0.515, p<0.01), business skills (r=0.199, p<0.05), the influence of the area of education (r=0.184, p<0.05). Perceived support from mentors has a positive correlation with business skills (r=0.209, p<0.05) and the influence of the area of education (r=0.214, p<0.01). These results could mean that the better these female entrepreneurs perceive the support from women associations and mentors the better their business skills and the influence of their area of education.
Business skills is positively correlated with the influence of the area of education (r=0.651, p<0.01) and government support (r=0.184, p<0.05). Level of education was positively correlated with economic motivation (r=0.207, p<0.05), social motivation (r=0.284, p<0.01), business skills (r=0.169, p<0.05) and influence of area of education (r=0.366, p<0.01). Since in Honduras, the majority of respondents in this study have a higher level of education, these results could mean that the higher the level of education the more economically and socially motivated these female entrepreneurs are. Also since they have higher levels of education, they possess more business skills and perceive a greater influence from their area of education. Work experience was positively correlated with the age of the business (r=0.29, p<0.01).
Of the demographics that are part of the independent variables, age has a significant positive correlation with work experience (r=0.388, p<0.01), age of the business (r=300, p<0.01) and a negative correlation with country insecurity (r=-0.215, p<0.01). This suggests that the older the female entrepreneur the more work experience they possess, the longer they have had their business and the less they feel the insecurity of the country affects their business.
The size of the family is positively correlated with level of education (r=0.221, p<0.01) and country insecurity (r=0.241, p<0.01). The age of the business is positively correlated with number of employees (r=0.248, p<0.01).
The control variable, size of the venture was positively correlated with business skills (r=0.191, p<0.05) and the influence of the area of education (r=0.172, p<0.05).
Table 4.1.
Means, Standard Deviations and Correlation Coefficients
Correlations
Variable Mean S.D 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1.Agea 3.44 1.08
2.Number of dependents b
2.79 1.34 .020
3.Level of education c
5.14 2.02 .035 -.221**
4.Work experience
11.80 9.10 .388** -.033 .006
5.Age of business
6.13 6.90 .300** .021 -.064 .289**
6.Number of employees when business started
1.88 2.07 .115 -.059 .084 .066 .248**
7.Number of employees currently
3.24 4.33 .090 .041 .043 .086 .079 .189*
(continued)
Correlations
Variable Mean S.D 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
16.Country insecurity
2.33 1.73 -.215**
.241** -.150 -.153 .023 .177* -.045 -.192* -.248**
-.118 .072 -.053 .029 -.023 .141 (0.82)
17.Performance 3.73 1.33 -.004 -.001 .317** -.044 -.102 .077 .229** .363** .233** .345** .151 .338** .479** .464** .167* -.043 (0.84)
c. (1) elementary school, (2) middle school, (3) Technical, (4) High school, (5) University, (6) Master degree, (7) Doctorate degree Note. The numbers in parenthesis correspond to the Cronbach’s alpha for each variable.
a. (1) below 20 years, (2) 21-30, (3) 31-40, (4) 41-50, (5) More than 50 years b. (1) 0, (2) 1, (3) 2, (4) 3, (5) 4+
Table 4.1. (continued)
Hypotheses Testing Partial Least Square (PLS)
Structural equation modeling (SEM) is a second generation technique that allows a simultaneous modeling of the relationships among different independent and dependent constructs (Gefen, Straub, & Boudreau, 2000). There are two approaches to estimate the parameters of SEM, the covariance based approach and the variance based approach. Partial least squares structural equation modeling (PLS-SEM) is a variance based approach and focuses on maximizing the variance of the dependent variables explained by the independent ones (Haenlein & Kaplan, 2004).
After the measurement scales were tested for validity and reliability, the proposed hypotheses were tested by using partial least squares (PLS) analysis. The model was tested by using each dimension as a separate variable instead of as a global variable. To test it, bootstrapping was run to see the path coefficients’ significance and t-values. According to Hair et al. (2011) it should be run with a minimum of 5,000 bootstrapping samples.
The R2 or coefficient of determination was taken into consideration; this explains how much the variance of the dependent variable is explained by the variance of the independent.
Another thing that was observed was the path coefficients to see the relationship between the dependent and independent variables, that is, whether they are positive, negative and significant. To determine whether a path coefficient is significant, the critical t-value is examined. For two-tailed test the critical t-value should exceed 1.65 at 90% of confidence level, 1.984 at 95% of confidence level and 2.626 at 99% of confidence level (Moore et al., 2009). Refer to table 4.2 to see the path coefficients, errors and t-values of the control variables and the main variables.
Table 4.2.
Path Coefficients, Error, t-values and R square
Economic -> Performance 0.3338 0.3225 0.0625 0.0625 5.3365***
Personal -> Performance 0.1918 0.1989 0.0763 0.0763 2.5121**
Social -> Performance 0.0082 0.0659 0.0499 0.0499 0.1646 Networks
Women associations-> Performance -0.2454 -0.2364 0.0891 0.0891 2.7546***
Mentors-> Performance 0.2525 0.2379 0.0882 0.0882 2.8622***
Human Capital
Business skills -> Performance 0.1755 0.1845 0.0755 0.0755 2.3238**
Influence of area of education -> Performance 0.2285 0.227 0.0854 0.0854 2.6771***
(continued)
Original Sample (O)
Sample Mean
(M)
Standard Deviation (STDEV)
Standard Error (STERR)
T Statistics (|O/STERR|)
R Square
Level of education -> Performance 0.0274 0.0562 0.0424 0.0424 0.6473 Previous work experience-> Performance 0.0679 0.0694 0.0506 0.0506 1.342 Environmental
Government support -> Performance -0.0307 -0.0538 0.0432 0.0432 0.7096 Country insecurity -> Performance -0.0294 -0.0702 0.0503 0.0503 0.5855 Demographics
Age -> Performance -0.0827 -0.0928 0.0566 0.0566 1.4613 Family size -> Performance 0.0003 0.051 0.0388 0.0388 0.009
Performance 0.5634
Table 4.2. (continued)
Note. * t> 1.65 at 90% of confidence level, ** t>1.984 at 95% of confidence level, *** t> 2.626 at 99% of confidence level
Effect of Motivations on Performance
The hypotheses (1a-1c) regarding the effects of economic, personal and social motivations have on performance were tested using PLS. As was expected, for economic motivation the results show that the hypothesis (1a) stating that economic motivations will positively affect the performance of female entrepreneurs in micro and small enterprises in Honduras was supported (â=0.334, t>2.626). As Lerner et al. (1997) showed in their study in Israel, those female entrepreneurs that are highly motivated, especially for economic reasons are more likely to achieve high profitability. In Honduras this motivation had the most significant effect on performance, probably due to the lack of jobs in the country, so women are more motivated by economic reasons to become an entrepreneur. It was found that the most important economic reasons for starting their business were economic necessity and the need to increase their income.
Hypothesis 1b stating that personal motivations will positively affect the performance of female entrepreneurs in micro and small enterprises in Honduras was also supported (â=0.192, t>1.984). This was expected as was shown in previous studies that personal motivations such as independence, opportunities and desire for achievement positively affect the success of female entrepreneurs (Lee & Stearns, 2012; Naser et al., 2009). In this study, results show that the most common personal motives these female entrepreneurs had for starting their business are the need for independence, wanting to be their own boss and a desire for achievement.
Contrary to what was expected by the researcher hypothesis 1c, social motivations will positively affect the performance of female entrepreneurs in micro and small enterprises in Honduras was not supported (â=0.008, t<1.65). In this research, results show that most of the respondents didn’t feel they were motivated to start their business due to social
motivations. According to the results, they didn’t feel too dissatisfied with their previous job conditions or payment, even though they did feel they have family responsibilities. This was probably why they didn’t feel social motivations affected their performance. In a study conducted by Naser et al. (2009), they also found family responsibilities to be one of the major social motivations, but contrary to the findings in this study they did find it to be an influence on the growth of female entrepreneurs businesses.
Effect of Networks on Performance
Hypothesis 2a was not supported since it seems that the support of women associations has a negative effect on performance (â=-0.245, t>2.626), contrary to what was hypothesized. In previous studies, it was found that those women who belonged to a women association had higher profitability (Lerner et al., 1997). In this study, the researcher expected that perceived support from women associations would have a positive effect on performance, as was established by previous studies, but instead the impact was negative. This could be due to the fact that most of these female business owners have had their business for approximately one year, so their business is quite new. They probably decided to become part of a women association due to startup problems so they have not yet seen so many positive results from these associations.
On the other hand, hypothesis 2b related to support provided by mentors will have a positive effect on female entrepreneur performance was supported (â=0.253, t>2.626). Like in previous studies, mentors were found to have a positive effect on performance, especially on revenues (Lerner et al., 1997). In this study in Honduras, results show that most of the respondents have a good perception of the support provided by mentors.
The significant result of hypothesis 2b helps to explain the negative influence of
the majority of the female entrepreneurs in this study belong to the Chamber of Commerce, maybe the negative result was due to the fact that even though the Chamber of Commerce offers more services than the other organization, its mentorship programs is not as strong. For Vital Voices the main service or program it offers is the use of mentors, while the Chamber of Commerce focuses on more than one service.
Effect of Human Capital on Performance
Hypothesis 3a was not supported since it seems that a high level of education does not have an effect on female entrepreneur business performance (â=0.027, t<1.65). These results are contrary to what was expected and to what previous studies have shown that women who have a higher level of education have businesses that grow faster (Biggs & Shah, 2006). Even though in this study almost half of the respondents have university studies, still they don’t feel their level of education has any effect on their businesses performance.
Hypothesis 3b, stating that the influence of area of education will be positively related to female entrepreneur business performance was supported (â=0.229, t>2.626). Like in other countries, in Honduras women have the opportunity to any course of study, and this can have a positive effect on their performance (Lerner et al., 1997). In another study they also found that education was a factor that affected growth of female entrepreneur’s ventures (Arasti et al., 2012). Since in this study the majority of the respondents have studies related to business, this may be the reason why they feel it has had a positive effect on their performance.
Hypothesis 3c was not supported since it shows that previous work experience does not have an effect on female entrepreneur business performance (â=0.068, t<1.65). This was also contradictory with what past researchers have found that having work experience helps ensure business growth (Tundui & Tundui, 2012). These results could be because the female entrepreneurs in Honduras probably had previous work experience not related to their
business. It has been found that in another study that having previous work experience not related to their business could be a constraint of growth (ILO, 2008).
Finally the last hypothesis 3d was also supported since it seems that business skills has a positive effect on female entrepreneur business performance (â=0.176, t>1.984). These results were expected, since literature shows that business skills are related to business performance, specifically revenues and profitability (Lernet et al., 1997).
Effect of Environmental Factors on Performance
None of the hypothesis relating to the effect of environmental factors with performance was supported. Both hypothesis 4a, government support will have a positive effect on performance (â=-0.031, t<1.65), and hypothesis 4b relating to the insecurity of the country will have a negative effect on performance (â=-0.029, t<1.65) showed no significance, thus were not supported.
Regarding the perceived government support, literature says that the amount of support received by female entrepreneurs affects performance (Singh & Belwal, 2008). In this study results showed that most of the respondents don’t feel the government provides them the desired support, this could be seen as the path coefficient has a negative sign (â=-0.031), but still it had no significant effect on performance. This could be due to the fact that the majority of the respondents are not aware of the trainings, loans and other support the government provides.
Country insecurity showed similar results, most of the respondents said they didn’t feel it affected their performance too much. This contradicts what a previous study (Callejas
& Yeh, 2013) found in Honduras. This could be due to the fact that when that previous study was conducted, the country was suffering its greatest insecurity problem.
Effect of Demographics on Performance
Of the three demographic hypotheses, 5a and 5c were tested using PLS and hypothesis 5b was tested using one-way ANOVA. Hypothesis 5a, stating that the age of the entrepreneur will have a positive effect on female entrepreneur performance was not supported (â=-0.083, t<1.660). In this study it didn’t show any effect on performance, which contradicts what previous studies showed that it affects performance, specifically growth (Arasti et al., 2012, Teoh & Chong, 2007). Even though in this study the female entrepreneurs are between the ages of 31 and 40, and are supposed to have more work experience and knowledge, still age is not significant with performance.
Hypothesis 5c which states that family size will have a negative effect on female entrepreneur performance was also rejected (â=0.0, t<1.660). This also contradicts what other researchers have found that the number of dependents affects female entrepreneurs performance due to the responsibilities it conveys (Holmén et al., 2011). The results of this study may be contrary due to the fact that most of the respondents had a few dependents, so they have more time and flexibility to manage their business.
Results of one-way ANOVA (table 4.3) indicate that there was a significant effect of the business sector on performance at the p<0.05 level for the different types of sector [F (3,141) = 2.928, p=0.036]. Therefore it can be concluded that hypothesis 5b which states that business sector has a positive effect on female entrepreneur performance was supported.
Table 4.3.
Results of One-Way ANOVA
Sum of Squares df Mean Square F Sig.
Between Groups 14.964 3 4.988 2.928 .036
Within Groups 240.179 141 1.703
Performance
Total 255.143 144
Even though the business sector has a positive significant effect on business performance, not a specific industry sector showed significance. These results can be seen in table 4.4 where the post hoc results of Sheffe’s test are shown.
Table 4.4.
Post hoc Results of Scheffe’s Multiple Comparison Test for Inner Group Differences
(I) Industry sector (J) Industry sector
Mean Difference
Table 4.4. (continued)
Mean Difference Std.
(I) Industry sector (J) Industry sector (I-J) Error Sig.
Service Hand craft .768 .603 .655
Production/manufacturing 1.164 .620 .321
Tourism .301 .635 .974
Summary of Hypothesis Testing Results
Figure 4.1 illustrates the path coefficients and r squares of the 13 hypotheses tested in PLS model. The factors in the PLS model explain 56.3% of the model variance (R2 =0.563).
Figure 4.1. Hypotheses testing, path coefficient and R squared
The following figure 4.2 shows the significance levels of the hypotheses.
Economic
Figure 4.2. Significance levels of hypothesis H2b+ (0.253) ***
Table 4.5 provides a summary of the results of the hypotheses testing, overall from the 14 hypotheses, 7 were supported and 6 were not supported.
Table 4.5.
Overview of the Results of Hypotheses Testing in PLS
Hypothesis Results Hypothesis 1a: Economic motivations will positively affect the
performance of female entrepreneurs in micro and small enterprises in Honduras.
Supported
Hypothesis 1b: Personal motivations will positively affect the performance of female entrepreneurs in micro and small enterprises in Honduras.
Supported
Hypothesis 1c: Social motivations will positively affect the performance of female entrepreneurs in micro and small enterprises in Honduras.
Not supported
Hypothesis 2a: The support of women’s associations will have a positive effect on performance of female entrepreneurs in micro and small enterprises in Honduras.
Not supported
Hypothesis 2b: Support from mentors will have a positive effect on performance of female entrepreneurs in micro and small enterprises in Honduras.
Supported
Hypothesis 3a: A high level of education will be positively related to business performance of female entrepreneurs in micro and small enterprises in Honduras.
Not supported
Hypothesis 3b: The influence of the area of education will be positively related to business performance of female entrepreneurs in micro and small enterprises in Honduras.
Supported
Table 4.5. (continued)
Hypothesis Results Hypothesis 3c: Previous work experience will have a positive effect on
business performance of female entrepreneurs in micro and small enterprises in Honduras.
Not supported
Hypothesis 3d: Business skills will have a positive effect on business performance of female entrepreneurs in micro and small enterprises in Honduras.
Supported
Hypothesis 4a: Government support will have a positive effect on performance of female entrepreneurs in micro and small enterprises in Honduras.
Not supported
Hypothesis 4b: The insecurity of the country will have a negative effect on performance of female entrepreneurs in micro and small enterprises in Honduras.
Not supported
Hypothesis 5a: The age of the entrepreneur will have a positive effect on performance of female entrepreneurs in micro and small enterprises in Honduras.
Not supported
Hypothesis 5b: Business sector will have a positive effect on performance of female entrepreneurs in micro and small enterprises in Honduras.
Supported
Hypothesis 5c: Family size will have a negative effect on performance of female entrepreneurs in micro and small enterprises in Honduras.
Not supported