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.