This chapter presents the results of data analysis. The mean and standard deviation of major research variables was calculated along with the correlations between each pair of variables.
Later on the hypotheses were tested using hierarchical regression in SPSS software.
Correlation Analysis
A Pearson product-moment correlation coefficient was computed to assess the relationship among all study variables. There was a significant correlation between PSS and OCB (r=.163, p
< .05). Impression management motives, choice of contract, time remaining and expectation of contract continuation had no significant correlations with OCB .
Job satisfaction as a control variable had a positive and significant correlations with OCB (r=.265, p < .01), altruism (r=0.23, p <.01), conscientiousness (r=0.466, p < .01) and compliance (r=.215, p < .01). None of the other control variables had significant correlations with the other main variables.
Perceived supervisor support was positively correlated with altruism (r=.190, p< .05), conscientiousness (r=.907, p<.01). Conscientiousness was positively correlated with impression management motives (r=.210, p <.01)) and perceived supervisor support (r=.253, p <.01). Choice of contract was positively correlated with conscientiousness (r=.171, p <.05) and negatively correlated with attendance (r=-.151, p<.05).
Previous temporary work experience had a positive relationship with altruism (r=.172, p<.05).
Job satisfaction was negatively correlated with time remaining on contract (r= -.161, p <.05).
Expectation of contract continuation was negatively correlated with previous temporary work experience (r= -.156, p <.05)
Table 4. 1. Correlational Analysis
44
Variables Mean S.D 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Age 2.52 0.696 1
2. Gender 0.52 0.501 .209**
3. Expectation 0.15 0.363 -.059 .040
4. Education 2.12 1.687 -.016 -.157* -.067
5. Tenure 4.54 1.376 .250** .178* -.089 -.129
6. Job satisfaction 5.02 1.087 .090 .027 -.088 .001 -.066
7. Previous temp. work experience 2.69 2.004 .217** .190* -.156* .029 -.056 .138 8. Contract renewed how
many times
8.8 5.427 .199** .132 -.004 -.189* .718** .037 .061 9. Contract Duration 2.42 1.687 .019 -.091 -.098 .119 -.212** -.127 -.012 -.484**
10. Time remaining 74.6 83.84 -.042 -.193** -.024 .237** -.164* -.161* -.045 -.357** .663**
11. Impression Mgt Motives 4.51 1.602 -.054 .168* .046 -.094 .051 .189* .107 .105 -.100 -.127 (0.85) 12. Perceived sup.
Support
4.86 1.357 .033 .025 -.066 .069 -.108 .577** .142 -.032 .007 -.175* .253** (.93) 13. Choice of contract 3.26 1.482 -.045 .014 -.031 .025 -.257** .074 .029 -.207** -.017 -.003 .318** .099 (.67)
14. OCB 4. 8 1.242 .002 -.054 -.088 .093 -.049 .265** .143 -.012 .071 .049 .140 .163* .001 (.87)
15. Altruism 4.78 1.204 -.002 -.023 -.099 .045 -.078 .230** .172* -.043 .081 .023 .122 .190* -.021 .919** (.91)
16. Conscientiousness 4.73 1.524 .040 .059 -.081 .092 -.141 .466** .121 -.064 .020 -.134 .210** .907** .171* .191** .196** (.85)
17. Attendance 5.66 0.877 .027 .013 .008 .140 .081 .146 .060 .091 .068 .134 .095 .001 -.151* .501** .396** -.012 (.58)
Note. Numbers in the brackets represent the Cronbach’s Alpha values of the variables. Expectation is coded as dummy variable yes=0, no=1. Gender is coded as dummy variable female=0, male =1. Age, education and tenure are coded as categorical variables.
*p < .05. **p < .01.
45
Hierarchical Regression Analysis
After the model was validated and estimated as reliable, hypothesis testing was conducted in order to answer the research questions. The research hypothesized that organizational citizenship behavior antecedents identified from literature would influence the OCB of contingent workers in the non-profit sector in Haiti. Regression was run four times on OCB and its three dimensions in order to test the hypotheses. Impression management motives, perceived supervisor support, choice of contract were entered in regression as independent variables and OCB or its dimensions as dependent variable while taking into consideration the possible moderating effect of time remaining on contract and expectation of contract continuation on the relationships between impression management motives and OCB.
OCB as Dependent Variable
Hierarchical regression analysis was used to test if impression management motives, perceived supervisor support, choice of contract significantly predicted OCB. Hypotheses 1 to 5 were tested and the analysis also explored a possible interaction between time remaining on contract, expectation of contract continuation and impression management to determine whether time remaining and expectation of contract continuation moderate the relationships between impression management motives and OCB. In the hierarchical regression reported in Table 4.2, age, gender, education, tenure and job satisfaction were entered as control variables into the analysis first. In model 2, IMM, PSS and CC were entered. In model 3, time remaining on contract and expectation of contract continuation were entered, followed by the IMM-time interaction in model 4. In the last model, the IMM-expectation interaction was entered.
Hypothesis one stated that: Perceived supervisor support has a positive influence on OCB. As depicted in table 4.2 under model 5, PSS had no significant impact on OCB (β=-0.016) and had a negative sign which is contrary to what was expected. Hypothesis one is therefore rejected. Hypothesis 2 stated that: Impression management motives have a positive influence on OCB. The results showed that IMM does not significantly predict OCB (β= 0.143); hypothesis two is also rejected.
46 Table 4.2.
Summary of Hierarchical Regression Analysis for Variables Predicting OCB (N=180)
Note: 1. IMM=Impression Management Motives; PSS= Perceived Supervisor Support; CC=
Choice of Contract; OCB= Organizational Citizenship Behavior; Expect= Expectation of contract continuation; Time= Time Remaining on Contract; JS= Job Satisfaction 2. The variables involved in the interaction were centered at their mean
3. Expectation coded as dummy variable yes=0, no=1 4. Gender coded as dummy variable female=0, male =1 † p< 0.1 *p < .05 **p < .01 ***p < .001
Organizational Citizenship Behavior
Model 1 Model 2 Model 3 Model 4 Model 5
Age -.009 .008 .004 .008 .010
Gender -.045 -.063 -.051 -.050 -.047
Education .084 .094 .075 .094 .092
Tenure -.010 -.039 -.038 -.016 -.020
JS .267*** .255** .255** .227* .233*
IMM .142† .149† .122 .143
PSS -.022 -.014 -.014 -.016
CC -.072 -.076 -.082 -.080
Time .071 .059 .063
Expect -.070 -.076 -.071
IMM x time .150† .142†
IMM x Expect -.057
R2 .082 .098 .107 .127 .129
ADJ.R2 .055 .056 .054 .070 .067
∆R2 0.016 .009 .020† .003
F 3.088* 2.319* 2.031* 2.218* 2.069*
47
Hypothesis three stated that choice of contract would have a negative relationship with OCB.
The results showed CC to have a negative but not significant relationship with OCB (β=- 0.080) leading to the rejection of this hypothesis as well. It was hypothesized that shorter time remaining on contract would moderate the relationship between IMM and OCB such that less time remaining would lead to a stronger relationship between IMM and OCB. The results showed the interaction as being significant (β=0.142, p<.1) but with a positive influence causing hypothesis four to be rejected. Figure 4.1 shows the possible influence of time remaining on contract on OCB with more time remaining strengthening the relationship between IMM and OCB.
Figure 4.1. Moderating effect of time remaining between IMM and OCB
Hypothesis five stated that the interaction of negative expectation of contract continuation would strengthen the relationship between IMM and OCB. The results do not support this hypothesis as the interaction was not significant (β=-0.57). The regression shows an acceptable power of the model to explain OCB (F=2.069, p<.05), although it only explained 12.9% of the variance of OCB (R2 = 0.129). There was a significant R squared change in model 4 of the table (∆R2=0.020, p< .1) with the addition of the IMM-time interaction.
48
Altruism as Dependent Variable
Hierarchical regression analysis was used to test the impact of impression management motives, perceived supervisor support, choice of contract on altruism. The five hypotheses were tested while considering a possible interaction between time remaining on contract and impression management as well as expectation of contract continuation and IMM to determine whether time remaining and expectations moderate the relationships between impression management motives and altruism. In the hierarchical regression reported in Table 4.3, age, gender, education, tenure and job satisfaction were entered as control variables into the analysis first. In model 2, IMM, PSS and CC were entered. In model 3, time remaining on contract and expectation of contract continuation were entered, followed by the IMM-time interaction in model 4. In the last model, the IMM-expectation interaction was entered.
As altruism was considered as dependent variable, this regression showed the models as providing very low fit to the data as portrayed in table 4.3. Only job satisfaction had some impact (β=0.154, p<.10) in model 1.
Table 4.3.
Summary of Hierarchical Regression Analysis for Variables Predicting Altruism (N=180) Altruism
Model 1 Model 2 Model 3 Model 4 Model 5
Age -.006 .010 .005 .009 .010
Gender -.012 -.029 -.017 -.017 -.015
Education .036 .040 .023 .039 .038
Tenure -.055 -.084 -.089 -.071 -.074
JS .227** .176† .173† .150 .154†
IMM .119 .129 .106 .119
PSS .056 .060 .060 .059
(continued)
49
Note: 1. IMM=Impression Management Motives; PSS= Perceived Supervisor Support; CC=
Choice of Contract; OCB= Organizational Citizenship Behavior; Expect= Expectation of contract continuation; Time= Time Remaining on Contract; JS= Job Satisfaction 2. The variables involved in the interaction were centered at their mean
3. Expectation coded as dummy variable yes=0, no=1 4. Gender coded as dummy variable female=0, male =1 † p < 0.1 *p < .05 **p < .01 ***p < .001
Hypothesis 1a predicting influence of PSS on altruism was rejected (β=0.059) and so was hypothesis 2a (β=0.119). Hypothesis 3a stating that CC would have a negative influence on altruism was not supported (β=-0.110). The moderators did not have any significant impact but rather caused the model to become insignificant (F=1.575) when they were added to the model consequently rejecting hypotheses 4a and 5a. The predictors in model 2 explained 7.7% of the variance (R2 =0.077) with F= 1.776, p<0.10.
Conscientiousness as Dependent Variable
Hierarchical regression analysis was used to test if impression management motives, perceived supervisor support, choice of contract significantly predicted conscientiousness. The analysis Table 4.3. (continued)
Altruism
Model 1 Model 2 Model 3 Model 4 Model 5
CC -.099 -.106 -.111 -.110
Time .052 .042 .044
Expect -.093 -.098 -.095
IMM x time .123 .119
IMM x Expect -.033
R2 .058 .077 .088 .101 .102
ADJ.R2 .031 .034 .034 .042 .037
∆R2 .018 .011 .013 .001
F 2.152† 1.776† 1.623 1.712† 1.575
50
also explored interactions between time remaining on contract and impression management and expectation of contract continuation and IMM to test whether time remaining and expectation moderate the relationship between impression management motives and conscientiousness.
In the hierarchical regression reported in Table 4.4, age, gender, education, tenure and job satisfaction were entered as control variables into the analysis first. In model 2, IMM, PSS and CC were entered. In model 3, time remaining on contract and expectation of contract continuation followed; in model 4 the IMM-time interaction was entered and in the last model, the IMM-expectation interaction. Hypothesis 1b stated that: Perceived supervisor support has a positive influence on conscientiousness. Table 4.4 model 5 shows that perceived supervisor support has a significant relationship with conscientiousness (β= 0.955, p<0.001). Hypothesis 1b is therefore supported as perceived supervisor support is found to relate to conscientiousness.
Table 4.4.
Summary of Hierarchical Regression Analysis for Variables Predicting Conscientiousness Conscientiousness
Model 1 Model 2 Model 3 Model 4 Model 5
Age .014 .016 .014 .015 .015
Gender .078 .049 .052 .052 .052
Education .089 .024 .020 .023 .023
Tenure -.117 -.028 -.029 -.025 -.025
JS .455*** -.086* -.087* -.092* -.091*
IMM -.047 -.045 -.049 -.049
PSS .954*** .955*** .955*** .955***
CC .091** .089* .088* .088*
Time .014 .012 .012
Expect -.024 -.024 -.024
IMM x time .024 .024
(Continued)
51
Note: 1. IMM=Impression Management Motives; PSS= Perceived Supervisor Support; CC=
Choice of Contract; OCB= Organizational Citizenship Behavior; Expect= Expectation of contract continuation; Time= Time Remaining on Contract; JS= Job Satisfaction 2. The variables involved in the interaction were centered at their mean
3. Expectation coded as dummy variable yes=0, no=1 4. Gender coded as dummy variable female=0, male =1 † p < 0.1 *p < .05 **p < .01 ***p < .001
Hypothesis 2b was not supported by the results as there was no significant positive relationship between IMM and conscientiousness. The model also shows a significant relationship between choice of contract and conscientiousness (β= 0.088, p<0.05) but it was not as hypothesized. Since it was predicted that CC would have a negative relationship with conscientiousness, hypothesis 3b was not supported although CC was significant. The interactions IMM-time (β=0.024) and IMM-expectation were not significant (β=-0.001) and did not add any value to the model. Hypotheses 4b and 5b were rejected. The regression shows that these predictors in model 5 explained 84% of the variance (R2=.841, F=73.644, p<0.001).
Model 2 showed the most significant change in R2 compared to the other models (∆R2 =0.598, p<0.001).
Attendance as Dependent Variable
A hierarchical regression was conducted with attendance as dependent variable while exploring possible interaction between time remaining on contract and expectation of contract Table 4.4. (continued)
Conscientiousness
Model 1 Model 2 Model 3 Model 4 Model 5
IMM x Expect -.001
R2 0.242 0.84 0.841 0.841 .841
ADJ.R2 0.220 0.832 0.831 0.831 .830
∆R2 0.598*** 0.001 0.002 .000
F 11.124*** 112.071*** 89.092*** 81.819*** 73.644***
52
continuation and IMM. In the hierarchical regression presented in Table 4.5, age, gender, education, tenure and job satisfaction were entered as control variables into the analysis first. In model 2, IMM, PSS and CC were entered. In model 3 the moderators were entered, followed by the IMM-time interaction in model 4 and IMM-expectation interaction in model 5.
Hypothesis 1c stated that: Perceived supervisor support has a positive influence on attendance.
The results show that PSS had a negative significant influence on attendance (β=-0.138, p<.05) therefore rejecting hypothesis 1c. Hypothesis 2c stated that IMM would have a positive influence on attendance. Model 5 of table 4.5 shows significant influence of IMM on attendance (β=0.214, p<.05). There is therefore support for this hypothesis.
Table 4.5.
Summary of Hierarchical Regression Analysis for Variables Predicting Attendance (N=180) Attendance
Model 1 Model 2 Model 3 Model 4 Model 5
Age -.015 .005 .000 .000 .004
Gender .016 .005 .024 .024 .029
Education .157* .177* .148† .148† .144†
Tenure .113 .040 .065 .066 .057
JS .154* .221* .235** .234+ .247**
IMM .170* .171* .170* .214*
PSS -.158† -.134 -.134 -.138*
CC -.200* -.195* -.195* -.191*
Time .150† .150† .158*
Expect .024 .024 .035
IMM x time .003 -.012
IMM x Expect -.114
R2 .053 .106 .126 .126 .137
ADJ.R2 .026 .065 .074 .069 .075
(continued)
53
Note: 1. IMM=Impression Management Motives; PSS= Perceived Supervisor Support; CC=
Choice of Contract; OCB= Organizational Citizenship Behavior; Expect= Expectation of contract continuation; Time= Time Remaining on Contract; JS= Job Satisfaction 2. The variables involved in the interaction were centered at their mean
3 Expectation coded as dummy variable yes=0, no=1 4. Gender coded as dummy variable female=0, male =1
† p < 0.1 *p < .05 **p < .01 ***p < .001
It was hypothesized that choice of contract would have a significant negative influence on attendance. In models 4, CC has a significant negative influence on attendance (β= -0.191, p<.05). Hypothesis 3c is also supported because CC had a negative influence on this dimension of OCB. Hypothesis 4c stated that less time remaining would make the IMM-attendance relationship stronger. The results showed that the interaction did not have a significant impact (β=-0.012). Hypothesis 4c is therefore rejected. It was hypothesized that negative expectation of contract continuation would strengthen the relationship between IMM and attendance.
The results showed the interaction was not significant (β=-0.114) hence rejecting hypothesis 5c. This regression shows that these predictors in model 5 shared 13.7% of the variance (R2=0.137). There was a significant change in R squared of model 2 (R 2 = 0.053, p < .05).
Table 4.6.
Overview of Hypotheses Testing
Hypothesis Result
Hypothesis 1: Perceived supervisor support has a positive relationship with contingent workers’ OCB in the nonprofit sector in Haiti.
Not supported
(continued) Table 4.5. (continued)
Attendance
Model 1 Model 2 Model 3 Model 4 Model 5
∆R2 .053* .020 .000 .011
F 1.958† 2.546* 2.439* 2.204* 2.207*
54 Table 4.6. (continued)
Hypothesis Result Hypothesis 1a: Perceived supervisor support has a positive relationship with
contingent workers’ altruism in the nonprofit sector in Haiti
Not supported
Hypothesis 1b: Perceived supervisor support has a positive relationship with contingent workers’ conscientiousness behaviors in the nonprofit sector in Haiti.
Supported
Hypothesis 1c: Perceived supervisor support has a positive relationship with contingent workers’ attendance behaviors in the nonprofit sector in Haiti.
Not supported
Hypothesis 2: Impression management motives have a positive relationship with OCB of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 2a: Impression management motives have a positive relationship with the altruism of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 2b: Impression management motives have a positive relationship with conscientiousness behaviors of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 2c: Impression management motives have a positive relationship with attendance behaviors of contingent workers in the non-profit sector in Haiti.
Supported
Hypothesis 3: Involuntary choice of contract has a negative relationship with OCB of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 3a: Involuntary choice of contract has a negative relationship with altruism of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 3b: Involuntary choice of contract has a negative relationship with conscientiousness of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 3c: Involuntary choice of contract has a negative relationship with attendance of contingent workers in the non-profit sector in Haiti.
Supported
(continued)
55 Table 4.6. (continued)
Hypothesis Result
Hypothesis 4: Time remaining on contract negatively moderates the relationship between IMM and OCB with shorter time remaining strengthening the relationship between IM motives and the OCB of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 4a: Time remaining on contract negatively moderates the relationship between IMM and altruism with shorter time remaining strengthening the relationship between IM motives and the altruism of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 4b: Time remaining on contract negatively moderates the relationship between IMM and conscientiousness with shorter time remaining strengthening the relationship between IM motives and the conscientiousness of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 4c: Time remaining on contract negatively moderates the relationship between IMM and attendance with shorter time remaining strengthening the relationship between IM motives and the attendance of contingent workers in the non-profit sector in Haiti.
Not supported
Hypothesis 5: Expectation of contract continuation negatively moderates the relationship between IMM and OCB with negative expectation strengthening the relationship between Impression management motives and OCB
Not supported
Hypothesis 5a: Expectation of contract continuation negatively moderates the relationship between IMM and altruism with negative expectation strengthening the relationship between impression management motives and altruism.
Not supported
(continued)
56 Table 4.6. (continued)
Hypothesis Result
Hypothesis 5b: Expectation of contract continuation negatively moderates the relationship between IMM and conscientiousness with negative expectation strengthening the relationship between impression management motives and conscientiousness.
Not supported
Hypothesis 5c: Expectation of contract continuation negatively moderates the relationship between IMM and attendance with negative expectation strengthening the relationship between impression management motives and attendance.
Not supported
57