Table Employees’ job satisfaction at Boma hotel
Statements related to employees’ job
satisfaction
Responses
(%)
Standard
deviation
Mean Ranking
Agree Strongly
agree
Employees strive to give their best
performance at Boma hotel
44.9 34.7 1.235 3.88 1
There is high employee retention at
Boma hotel
36.7 26.5 3.51 3.51 3
The employees’ commitment at Boma
Hotel is high
42.9 30.6 3.76 3.76 2
n = 49
Diagnostic tests
The data was subjected through a thorough data screening process which included tests to
ascertain that the data met the general assumptions of the regression analysis which was a key
analytical model for the study at hand.
Test for normality
Tests for normality are undertaken for purposes of comparing the selected sample distribution
shape of a normal curve. The assumption is that if the sample has a normal shape, the population
too will be distributed normally. Normality is then assumed. A significant test insinuates that the
sample distribution is not shaped like a normal curve. Shapiro -Wilk's W is recommended for
small and medium samples up to n = 2000. Since the sample is only 196 respondents, Shapiro
Wilks W test was used. The following normality hypotheses are assumed:
H0: the observed distribution fits the normal distribution.
Ha: the observed distribution does not fit the normal distribution. If we accept the H0, we
accept/assume normality
Table Shapiro – Wilk test of Normality
Tests of Normality
Kolmogorov-SmirnovaShapiro-Wilk
Statistic Df Sig. Statistic df Sig.
Job satisfaction .410 365 .068 .618 365 .068
a. Lilliefors Significance Correction
H0 = Normality
Since the sig. or the p value of the Shapiro-wilk test of normality is more than 0.05 for job
satisfaction, standing at 0.068, then the researcher failed to reject H0 (the data does not deviate
from a normal distribution. The interpretation was guided by conventional wisdom presented by
Shapiro & Wilk (1965) and Razali & Wah (2011).
Test for Auto correlation
The Durbin Watson statistics as generated using SPSS with job satisfaction is presented below.
Table 4.8: Durbin Watson Test for auto correlation
Model Summaryb
Mode
l
R R Square Adjusted R
Square
Std. Error of the
Estimate
Durbin-
Watson
1 .992a.983 .982 .106 1.930
a. Predictors: (Constant), Working conditions, Remuneration, Relationship with supervisor, and
Co-worker relations.
b. Dependent variable: Job satisfaction
The Durbin-Watson in the Model summary, d = 1.930 lies between the two critical values of
1.5<d<2.5. Following the conventional wisdom presented by the authors, Durbin & Watson
(1971), the researcher therefore made an assumption that there was no first order linear auto-
correlation in the multiple linear regression data.
Test for multi Collinearity
The statistics on the multicollinearity test employed using SPSS tools are presented below.
Table Test for multi collinearity using Tolerance and VIF
Model Collinearity Statistics
1(Constant) Tolerance VIF
Working conditions 0.118 8.47
Remuneration 0.158 6.33
Relationship with supervisor 0.216 4.63
Co-worker relations 0.112 8.93
a. Dependent Variable : Job satisfaction
According to Liu, Kuang, Gong & Hou (2003), extremely small values would indicate that a
predictor is redundant. This means that values that are less than 0.1 merits further investigations.
In this output, tolerance values stand at 0.118, 0.158, 0.126 and 0.112 for working conditions,
remuneration, relationship with supervisor, and co-worker relations. In addition, Variance
Inflation Factor (VIF) stands at 8.47, 6.33, 4.63 and 8.93 for working conditions, remuneration,
relationship with supervisor, and co-worker relations. All this values are below the maximum cut
off point of 10. This implies that the researcher made a conclusion on the absence of multi
collinearity problem in the data set.
Regression Results
Regression Coefficients
The regression coefficients are presented in table 4.10.
Table Coefficients table
Model Unstandardized
Coefficients
Standardized
Coefficients
t Sig.
B Std. Error Beta
1 (Constant) .339 .449 0.933 .535
Working conditions .392 .254 .205 1.933 .027
Remuneration .324 .239 .235 1.827 .008
Relationship with supervisor .141 .166 .196 1.555 .052
Co-worker relations .181 .156 .136 1.437 .061
a. Dependent Variable: Customer choices
The B values, and the p-value to check for significance are depicted in table 4.10 above. We
reject Ho if p < .05. This means the relationship is reliable and can be used to make predictions.
The findings also show the contribution of each variable in explaining relationships between the
research variables as depicted by unstandardized. The study utilized the equation below.
Y = β0 + β1X1 + β2X2 + β3X3+ β4X4+έ
Where:
Y = Job satisfaction
X1 = Working conditions
X2 = Remuneration
X3 = Relationship with supervisor
X4 = Co-worker relations
έ = Error Term.
β0 = Point of intercept on the y axis
Y = 0.339 + (0.392 X1) + (0.324 X2) + (0.141 X3) + (0.181 X4). The implication is that even by
excluding the four study variables, job satisfaction would be 0.339. The findings also indicate
that a unit change in working conditions would result in 0.392 change in job satisfaction, a unit
change in remuneration would result in 0.324 change in job satisfaction, a unit change in
relationship with supervisor would result in 0.141 change in job satisfaction and a unit change in
co-worker relations would result in 0.181 change in job satisfaction.
Findings in the table also show that factors attributed to working conditions (p=0.027) and
remuneration (p=0.008) were significant in statistical terms. The error term (0.02) The error term
(0.02) insinuates lack of complete accuracy and result in the results in actual situation will be
different. Findings of the study indicate that, overall, the order of ranking in terms of relative
influence is as follows: (1) remuneration = (0.008); (2) working conditions = (0.027);
relationship with supervisor = (0.052); and co-worker relations = (0.061).
Coefficient of determinant Table Regression Model
Summary
Presented in table 4.11 is the model summary that shows values of R and R Square, which
includes information on variance quantity which is explained by the predictor variables. The first
statistic, R, is the multiple correlation coefficients between all of the predictor variables and the
Model Summaryb
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .280a.784 .689 0.02
a. Predictors: (Constant), Working conditions, Remuneration, Relationship with supervisor,
Co-worker relations
b. Dependent
Variable: Job satisfaction
dependent variable. In this model, the value is 0.280, which indicates that there is a great deal of
variance shared by the independent variables and the dependent variables. This is frequently used
to describe the goodness-of-fit or the amount of variance explained by a given set of predictor
variables. In this case, the value is 0.784, which indicates that 78.4% of the variance in the
dependent variable is explained by the independent variables in the model.
Analysis of Variance (ANOVA) Table F- test on
ANOVA
It can be concluded from the ANOVA that at 5 % significance level, there exist sufficient
evidence to justifiably conclude that that the slope of regression line is not zero. As Such, job
satisfaction influencing factors included in the model are useful predictors of job satisfaction
since the p value is 0.000 which is less than 0.05. Sig = 0.000 suggests that the model adopted
for this research is significant for predicting job satisfaction in Boma hotel.
Limitations of the Study
Limitations include: Restrictive organizational confidentiality hindering provision of full
responses to the study-this was addressed by assuring respondents of utmost confidentiality and
disclosing the academic purpose and intention of the study; loss of questionnaires by the
respondents, incomplete questionnaires and failure to provide objective responses by the
respondents which was overcome by organizing meetings outside working hours and seeking for
ANOVAa
Model Sum of Squares Df Mean Square
F
Sig.
1 Regression 2.687 5 .537 22.63 .000
Residual 10.410 25 .416
Total 13.097 30
a. Predictors: (constant), Working conditions, Remuneration, Relationship with supervisor,
Co-worker relations
b. Dependent
variable: Job satisfaction
personal contacts of would be respondents; some issues being misunderstood by
the responses; unexpected occurrences like respondents proceeding on leave
before completing the questionnaires which was mitigated through constant
reminders to the respondents during the period they were expected to complete
the questionnaire.
