CHAPTER FOUR
4. DATA PRESENTATIONS AND ANALYSIS
4.1. Introduction
This chapter will present the major findings of the research survey conducted on
Gambella National Park on two groups the management of the park and tourists who
visit the park. This section starts with explaining the profiles of the survey
respondents, it continues with a descriptive analysis of the study. In the quantitative
analysis part of this chapter, the performance of Gambella National Park before the
study period and in the study period are briefly overviewed. The reliability test
outcomes of the study survey questionnaire were also examined. Preliminary
regression and regression analysis tests and their results are presented in a more
detailed manner. Finally, a summary of the major findings of the research is presented
shortly.
4.2 Response Rate
A total of 171 questionnaires were distributed to the target population of the study
area by using strafied proportionate sampling method and purposive sampling method
for concerned department those their tasks is related to national park. Out of the total
171 questionnaires, 150 of the questionnaire were obtained from the respondent with
complete answered of each dimensions of the questionnaire.
Thus, the number of uncollected questionnaires from the respondents is insignificant
and can’t affect the finding of the study. The remaining 21 questionnaires were not
collected due to refusal to give response for the questionnaires and that there were
problems of clarity, and were not filled properly. Therefore, a response rate can be
87.7%. To effectively identify and analyze the respondents who participated in the
study, the analysis of the response rate was carried out as shown in the table below;
Table 4.1: Analysis of the Response Rate
Category Response Percentage
Response 150 87.7
Non-response 21 12.3
Total 171 100
Source: Own survey, 2024
4.3 Demographic background of the respondents
The analysis of the demographic background of the respondents in the study
conducted on the bases of; age, gender, marital status, educational level and
occupational of the respondents. To describe this demographic information, the
researcher used descriptive statistics such as, table percentage and Bar graph.
Figure 4.1 Gender of the respondents
Source: Own survey, 2025
The above figure indicates that, out of 150 respondents, 59 were female which is
equivalent to 39.3% have been represented in the study and 91 were male which is equal
to 60.7%. This implies that the study is being dominated by male.
Figure 4.2 Age of the respondents
Source: Own survey, 2025
As clearly seen on the above figure 4.2, the majority of the respondents (28.67%)
were between the ages of 26-35, 26.0% of the respondent were the age between 26-
35, 25.33% of the respondent were the age between 36-45 and the remaining 20% of
the respondent were the age above 46 years. This shows that the majorities of the
respondents were on working age and are suitable for the study.
Figure 4.3 Marital Status of the respondents
Source: Own survey, 2025
As indicated in the above table, 24.0 % % of the respondents were single, 54.7% were
married, 14.7% of respondent were widowed and the remaining 6.7% indicated
divorced. According to the finding of marital status of the respondents, the majority of
the respondents under this study were married.
Figure 4.4 Educational level of the respondents
Source: Own survey, 2025
In terms of educational background, 15 of the respondents (10%) are in high school
level, 20 (13.3%) of the respondents are certificate holders, 24(16.0%) of the
respondents were diploma holders and the remaining 91(60.7%) respondents were
bachelor degree and above.
Figure 4.5 Occupation of the respondents
Source: Own survey, 2025
The above figure shows that; majority of the respondents 82(54.67%) are civil
servants followed by teachers with 24(16.00%) where as 13.33% are farmers who
resided around park. The number of tourists was only 11 or 7.33% of the total
respondents. This indicates that, majority of the respondents are civil servants. These
individuals are very relevant for this study because they are the ones with great
knowledge of overall park performance.
4.4 Descriptive Analysis of the study
4.4.1 Mean Score Interpretation in the study
Mean score and standard deviation are the most frequently used in descriptive
research study. Thus, the study used both qualitative and quantitative approach, the
researcher employed mean score interpretation to support the inferential statistic result
by rating the perception of participants about their national park performance.
Accordingly, the mean interpretation in this study was conducted based on Maugo
(2013) mean score interpretation. Maugo, interpreted the attitudes and perception of
respondents by mean sore on Likert scale that the range was strongly disagree (1) to
strongly agree (5). Accordingly, the scores of strongly disagree and disagree (low
performance) have been taken to represent a variable which had a mean score of (0 to
2.5). The scores of moderate agree (medium performance) have been taken to
represent a variable with a mean score of (2.5 to 3.4) and the score of both agree and
strongly agree (high performance) have been taken to represent a variable which had a
mean score of (3.5 to 5).
And also a standard deviation of >0.9 implies a significant difference on the impact of
the variable among respondents.
Table 4.2: Over all Descriptive Analysis of the Explanatory Variables over Response
Variable in the study
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
National Park Performance 150 1.00 4.40 2.7693 0.81388
Environmental factors 150 1.00 4.25 2.8133 1.25956
Socioeconomic factors 150 1.00 4.25 2.7392 0.81576
Government policies 150 1.50 4.17 2.7322 1.23661
Valid N (listwise) 150
Source: Own Survey, 2025
In this study the summary of the descriptive statistics is shown in the above table, all
variables were conducted based on a 5-point Likert scale. Accordingly, the results
show that the mean for national park performance was 2.7693 with SD = 0.8138
which indicates the park performance moderate. And also the mean score shown in
the above table infer that the park wasn’t not highly efficiencies in the attractive. The
average performance of the national park, as rated on a scale from 1 to 4.40, is
approximately 2.7693. The relatively low standard deviation indicates that there isn’t
a large variability in responses, suggesting a moderately consistent perception of park
performance among the respondents.
Environmental factors scored an average of 2.81, indicating that these factors have a
moderate effect on park performance. The higher standard deviation (1.26) implies
variability in the impact of environmental factors, which could include issues like
climate change, habitat degradation, and biodiversity loss. Studies confirm that
environmental factors are crucial to park performance, as they directly impact wildlife
habitats and ecosystem health (Convention on Biological Diversity, "Global
Biodiversity Outlook" (2020).
Socioeconomic factors have a mean of 2.74, close to that of national park
performance, suggesting a moderate perceived impact. The standard deviation here is
relatively low, indicating that most respondents had similar views on how
socioeconomic factors affect the performance of the national park.
Government policies have a mean of 2.73, which is very close to the mean scores of
national park performance and socioeconomic factors. The relatively high standard
deviation indicates a wide range of responses, suggesting that perceptions of the
impact of government policies on park performance vary greatly among respondents.
Generally, each of these factors (environmental, socioeconomic, and government
policies) has a mean score close to the mean score of national park performance
(2.77), suggesting that respondents view all three factors as moderately influencing
park performance. However, the variability (standard deviation) in responses for
environmental factors and government policies indicates diverse perspectives among
respondents on how these factors affect the park.
Table 4.3: Correlations
Correlations
Environmenta Socioeconomi Government Park
l factors c factors policies Performance
** **
Environmenta Pearson Correlation 1 .052 .064 .759**
l factors Sig. (2-tailed) .000 .000 .000
N 150 150 150 150
** **
Socioeconomic Pearson Correlation .052 1 .632 .516**
factors Sig. (2-tailed) .000 .000 .000
N 150 150 150 150
** **
Government Pearson Correlation .064 .632 1 .533**
policies Sig. (2-tailed) .000 .000 . .000
N 150 150 150 150
** ** **
Park Pearson Correlation .759 .516 .533 1
Performance Sig. (2-tailed) .000 .000 .000
N 150 150 150 150
**. Correlation is significant at the 0.01 level (2-tailed).
Source: Own Survey, 2025
The table's correlation coefficients show that, at the typical conventional level of
correlation, the correlation coefficient's maximum value of 0.7 is safe. When two
independent variables have correlation coefficients more than 0.8, it indicates that the
coefficient is getting close to one and suggests that the independent variables are
strongly correlated, which raises concerns about the model's predictive ability. With a
value of 0.759, the accompanying table shows the highest correlation between
environmental factors and park performance. The correlation coefficient between
environmental factors and park performance is 0.759. The correlation coefficient
between socioeconomic status and park performance is 0.516. The remaining
associations between Park and government policies
The findings suggest that, from all variables, environmental factors have significant
associations with park performance.
4.5 Regression Analysis
Regression analysis is the inferential statistics model used to conduct to know by how
much the explanatory variable explains the response variable (Bekele, 2014). And it
helps to understand by how much each explanatory variable explains the response
variable. Accordingly, in this study the regression analysis of park performance was
conducted and the results of the regression analysis were presented here. Because of
the study contained more than two explanatory variables; the researcher used multiple
linear regressions in the study. Therefore, regression analysis for predictor variable
(environmental factors, socioeconomic factors and government policies) and
(response variable) park performance was conducted by multiple linear regressions
4.5.1 Homoscedasticity
The test of this assumption of homoscedasticity is the central to linear regression
model. It describes a situation in which the error term (random disturbance in the
relationship between the independent variables and the dependent variables) is the
same across all values of the independent variables. Assumptions can be checked by
scatter P-plot diagram. The result plots the values the model would predict, against
the residuals obtained.
As the predicted values increase, the variation in the residuals should be roughly
similar. The graph observed value moves around the increasing predicted averages
values of variance. So, the model is homoscedasticity.
Figure 4.1 Homoscesdasticity test
4.5.2 Autocorrelation
In case of auto co relation, the test of Durbin-Watson statistic is used to test for the
presence of serial correlation among the residuals. Durbin-Watson- the Durbin–
Watson statistic expresses that whether the assumption of independent errors is
acceptable or not. As the conservative rule suggested that, values less than 1 or greater
than 3 should definitely raise alarm bells (Field, 2009). So that the desirable result is
when the value is closer to 2.The residuals are not correlated if the Durbin-Watson
statistic is approximately 2, and an acceptable range is 1.50 - 2.50.For this study the
condition existed between those intervals which means that 1.986 and not correlated
thus it satisfied the expected result.
Table 4.4 Autocorrelation test for of the study
Model summary
Model Summaryb
Model Durbin-Watson
1 1.888
a. Predictors: (Constant), Governemt policies, Environmental factors, Socioeconomic factors
b. Dependent Variable: National Park Performance
Predictors: (Constant), Government policies, Environmental factors, Socioeconomic
factors According to above table on durbin-watson, when there is serial correlation
the expected value of the durbin-Watson test statistics is 1.888 this is a below 2.5.
According the science, the value of durbin-watson is 1.5 to 2.5 was normal if the
value of durbin-watson is below 2 what called positive correlation. And if the value of
durbin Watson was above 2 it is negative correlation. So, the value of this study is
normal correlation.
4.5.3 Normality test
The normality test according to Field (2009), the assumption of normality is important
in research while using regression and helpful to generalize the result of the analysis
beyond the Normality test assumes that the residual has zero mean and constant
variance. Thus, the result on Figure 4.2 indicates the mean of the residual was
between -1 and +1 that means 3.198E-16 as well as its variance is 0.990
approximately 1 which implies that the distribution of the error term is normally
distributed.
Figure 4.2 Normality Test
From the figure above, if the dots created a pattern, this would indicate the residuals
are not normally distributed, the residual is correlated with the independent variables,
and/or the variances of the residuals are not constant. Here, there is a worrying effect
of larger residuals for larger fitted values. This is called ‘heteroscedascity’ meaning
that not only is variance in the response not equal across groups, but that the variance
has some specific relationship with the size of the response. And also the residual is
not relatively uncorrelated with the linear combination of predictors. As we only have
a small number of sample sizes in this study, the graph can be difficult to read, but as
it generally appears more random than funneled, this assumption is probably ok.
4.5.1Multi-co linearity
In this study, the researcher was checked multi-co linearity assumption with tolerance
and VIF statistics. A tolerance value less than 0.1 certainly indicate a serious co
linearity problem (Andy, 2009). Similarly, Yan and Gang Su (2009) suggested that in
multiple regressions a VIF value greater than 10 is an indication of potential co
linearity problems and importantly take in to consideration. However, in this study all
the predictors’ variable values were below10. This value indicates that there is not an
issue of co linearity problem between predictor variables in the study. Which means
the derived model is likely to be unchanged by minor changes in the measured
variables.
Table 4.5: Multi co linearity of the Explanatory Variable in the study
Collinearity statistics
Explanatory variables Tolerance VIF
Environmental factors 0.246 4.068
Socioeconomic factors 0.132 5.587
Government policies 0.208 4.812
Source: 0wn Survey, 2025
The result in table above shows that, the collinearity between independent variables
there was no series problem. Since the value of tolerance for all independent variable
is greater than 0.1 and all VIF is less than ten (VIF<10). Therefore, the assumptions
for regression analysis for the majority variables are met.
4.5.2 Coefficient of determination
The coefficient of determination clarifies the extent to which changes in the
dependent variable can be explained by the change in the independent variable or the
percentage of variation in the dependent variable (Park performance) that is explained
by all the four independent variables (environmental factors, socioeconomic factors
and government policies).
Table 4.6: Model Summary
Model Summaryb
Adjusted R Std. Error of
Model R R Square Square the Estimate
1 0.824 0.680 0.667 0.46006
a. Predictors: (Constant), Government policies,
Environmental factors, Socioeconomic factors
b. Dependent Variable: National Park Performance
Source: Own Survey, 2025
As indicated above from model Summary table, The R value of 0.824 indicates very
strong correlation between national park performance and the three predictor variables
(environmental factors, socioeconomic factors and government policies) which shows a
good level of prediction. The R2 value (also called the coefficient of determination),
which is the proportion of variance in the dependent variable that can be explained by
the independent variables. As shown from the table, R 2 value of 0.68 indicates that, 68%
of the variation in the national park performance of the study area can be explained by
the independent variables included in the model. The remaining 32% variation is
explained by stochastic error term (ε i) meaning that 32% of national park performance
changes were explained by factors that are not explained in the model.
4.5.3 Autocorrelation
In case of auto co relation, the test of Durbin-Watson statistic is used to test for the
presence of serial correlation among the residuals. Durbin-Watson- the Durbin–
Watson statistic expresses that whether the assumption of independent errors is
acceptable or not. As the conservative rule suggested that, values less than 1 or greater
than 3 should definitely raise alarm bells (Field, 2019). So that the desirable result is
when the value is closer to 2.
Table 4.7: ANOVA Results of the regression
ANOVAa
Sum of Mean
Model Squares Df Square F Sig.
1 Regression .415 3 42532 114.100 .001b
Residual 44.311 146 0.373
Total 44.726 149
a. Dependent Variable: National Park Performance
b. Predictors: (Constant), Government policies, Environmental factors,
Socioeconomic factors
Source: Own Survey, 2025
From the ANOVA statistics in table 4.11, the study established the regression model
had a
significance level of 0.001 which is an indication that, the data were ideal for making
a
conclusion on the population parameters. The calculated value was greater than the
critical value (114>0.389) an indication that independent variable significantly
influences the park performance, Gambella Southwest, Ethiopia.
The significance value was less than 0.05 indicating that the model was significant.
The ANOVA table tested the significance of the overall regression model. Four
independent variables identified in the study found to be statically significant
(p=0.001, F=114.100). This implies that the predicting ability and probability of the
independent variables (environmental factors, socioeconomic and government). To
dependent variable (the park performance) is high and significant. Hence, the
regression model is statically suitable to explain the variable which is a good fit for
the data.
Model Regression Coefficients
The study conducted a multiple regression analysis to determine the relationship
between independent variables and dependent variable. The regression equation was
Y = β0 + β1X1 + β2X2 + β3X3 + 0 Where:
Y is the dependent variable (Park performance),
β0 is the regression coefficient/constant/Y-intercept,
β1, β2, β3 are the slopes of the regression equation,
X1 = the Environmental factors
X2 = the Socioeconomic factors
X3 = the Government policies
α is an error term at 99% confidence level
Standardized Coefficients
In the below table 4.7, all predictor’ variables of Beta value or standardized
coefficients in the analysis were positive. Thus, the positive value of independent
variables on the dependent variable indicates there is direct relationship between Park
performance and independent variables of all independent variables. The ᵦ coefficient
of independent variables in the analysis were Environmental factors (ᵦ =0.080, t
=1.001& p =.0.971); Socioeconomic factors (ᵦ=0. .008, t =.090& p=0. .005), and
Government policies (ᵦ=0.053, t =0.623& p=<0.014. According to the above table, all
explanatory variables are statistical significant and the regression model is valid.
Therefore, according to the above regression analysis result, it can be possible to
conclude based on the Beta value that it helps to measure how strongly predictor
variables influence the response variable (Park performance). So that, the Beta values
of independent variables in the study interpreted and concluded as follows:
Unstandardized Coefficients
Unstandardized coefficient denotes the change in the dependent variable with a unit
change in the independent variables. But they are not comparable in terms of impact
on the dependent variable. The study used the following multiple regression model to
establish the statistical significance of the independent variables on the dependent
variable.
OP i=α + β 1 Ef i+ β 2 Sei + β 3 Gpi +ε i
Where;
PPi: Dependent variable (Park Performance)
α : Constant
β i: Beta value (Coefficient of slop of regression model)
Ef i: Environmental factors
Sei : Socioeconomic factors
G pi: Government policies
ε i: Error term
Therefore:
PPi=2.643+0.077 (Ef i)+0.009 (Se ¿¿ i)+0.054 (Gp¿¿ i)+ ε i ¿ ¿
Table 4.8: Model Regression Coefficients
Coefficientsa
Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 (Constant) 2.643 .418 6.327 .000
Environmental
.077 .080 .080 .971 .003
factors
Socioeconomic
.009 .098 .008 .090 .005
factors
Government
.054 .087 .053 .623 .014
policies
a. Dependent Variable: National Park Performance
Source: Own Survey, 2025
Discussion of the interview analysis
The interview was one of data collection method in addition to the questionnaire.
Interview was designed to in order to obtain data that has not in the questionnaire and
support answers on the results of the questionnaire. The interview was held
specifically with the Park Manager Gambella National Park and other concerned
bodies in the surrounding community. The purpose of an interview is to gather in-
depth information through direct interaction, often focusing on understanding the
interviewee’s perspectives, experiences, knowledge, or competencies. Here are the
results of interview conducted during data collection period.
What are the causes of challenges affecting the performance of Gambella national
park?
According to interviewees’ response, the performance of Gambella National Park,
located in Ethiopia, faces several challenges that hinder its conservation efforts,
management, and tourism potential. Some primary causes of these challenges include:
limited funding, climate change human conflict and infrastructure development. The
interviewees responded that, by addressing these challenges, it requires a combination
of government commitment, sustainable funding, community-based conservation
approaches, and international support for biodiversity protection.
Is the cause of challenges affecting the Gambella national park
performance is it socio-economic/environmental/cultural factor?
The challenges affecting the performance of Gambella National Park arise from a
combination of socio-economic, environmental, and cultural factors. Here's how each
of these factors plays a role: according to interviewee, one of the most common
factors is socioeconomic factor. Among socioeconomic factors, poverty and
livelihood dependence is the main factor because local communities rely on natural
resources for their livelihood leading to activities like poaching, logging, and
agricultural encroachment. They also added that, the challenges facing Gambella
National Park are intertwined across socio-economic, environmental, and cultural
dimensions, with each factor contributing to the complexity of managing and
conserving the park effectively. Addressing these challenges will require a holistic
approach that considers all these factors and promotes sustainable practices,
community involvement, and ecosystem resilience.
What are some efforts that can be taken to overcome the problem of
challenges?
This is one of the important questions during interview with the management bodies
and their responses was discussed here below. To overcome these challenges facing
Gambella National Park, a combination of conservation, community engagement, and
sustainable management practices is needed.
These efforts require a comprehensive approach that involves conservation actions,
community collaboration, infrastructure improvements, and policy support. By
addressing these issues holistically, Gambella National Park can work towards
sustainable performance in conservation, tourism, and community development,
thereby securing its ecological and economic.
