Investigation of Judicial Service Quality and Customer Satisfaction: The Case of Dire Dawa City Courts, Ethiopia

This study is conducted to examine Courts Quality Service, Service Delivery and Customer Satisfaction in City Courts. The study was used the primary and secondary source of data and data was collected from 282 sample through a survey questionnaire. Different descriptive statistics, Chi-Squared test, SERVQUAL Model, and logistic regression were used to carry out the research. From 282 study participants, 216 (77%) were males and 66 (23%) were females. The SERVQUAL result depicted that responsiveness is the most important factor to have a positive and significant effect on customer satisfaction followed by reliability, assurance, empathy, and tangibility. The overall satisfaction analysis result showed that 73% customers were satisfied by the City courts service. From the logistic regression, the City Courts are provide more service on the reliability dimension, assurance dimension, and minimizing of visiting frequency to the customer would increase the satisfaction of customer. Conferring inclusive jurisdictional power that stands by itself and Establishing Cassation Bench in Appellate Court of the City resolves the problems of dissatisfaction of Courts service delivery.


Introduction 1.Background of the Study
The justice sector includes the institutions for the administration of justice, and it should ensure that laws are enforced, the legal system is implemented, social security is promoted, and that social cohesion is guaranteed. A well-functioning justice sector promotes legal certainty, resolves conflicts and regulates the exercise of executive authority. On the other hand, a dysfunctional justice sector encourages human insecurity and impedes economic development.
The 1995 Federal Democratic Republic of Ethiopia (FDRE) Constitution of Ethiopia provides the legal basis for establishment of Federal state structure that consists of nine Regional States and two city administrations. The Federal, Regional States and city administrations have their own legislative, judiciary and executive branches of government. The judicial system can be defined as the governmental branch responsible for enforcing the law, interpreting, adapting to concrete cases and solving disputes. Judicial power is shared between the Federal Government and the Regional States. Article 78 of the Constitution creates three levels of courts in the Federal and Regional States i.e. First Instance Court (FIC), High Court (HC), and Supreme Court (SC) respectively. In the Charter of Dire Dawa Administration Proclamation No.416 /2004, a two level Municipal Courts are established, namely First Instance Court (FIC) and Appellate Court (AC) that exercise a municipal jurisdiction. There is a cassation bench within the Appellate court. Besides there are social courts whose function is based on traditional social values. District Social Courts shall have the jurisdiction over cases regarding property and money claims of an amount not exceeding 100 US dollars, and other civil jurisdiction which shall be determined by law according to Article 41 of the Charter. Any person who is dissatisfied by the decision of social courts may appeal to the First Instance Court of the City.
This study discusses only some of those concerns in its focus on the delivery of services in the Dire Dawa City Court and not in law. An essential component of this service quality enquiry is the delivery of on-time services in the First Instance Court (FIC) and Appellate Court (AC) that exercise what modes of service delivery should be utilized. Dissatisfied customer in the private sector can simply reject the unsatisfactory provider and look to someone else to provide the desired services. In this environment, the laws of the market ensure that providers constantly meet the needs and desires of consumers. Many of the deficiencies in the administration of justice may be addressed by proper allocation of sufficient resources. According to World Bank (2013), in the last two decades, the Ethiopian Public Service institutions improved performance and service delivery by implementing various reform tools and restructuring instruments including Business Process Reengineering (BPR) at federal, regional, and city administration levels in almost all public institutions. The Business Process Reengineering (BPR) was instrumental in clarifying organizations' strategic visions, missions, mandates, and products, and enhancing their top leadership's appreciation of strong institutions simplifying work processes and requirements, which resulted in substantially shorter transaction times for most services; developing service standards establishing important flat structures and offices as well as decentralizing tasks to lower levels of government, establishing one-stop shops

Significant of the study
This study has paramount significance to decision makers, courts manager, stakeholders and practitioners in different ways.
· To management of courts', the findings of the study provide viewpoint for evaluating the level of customer satisfaction associated with the services they are delivering. · The results of the study also provide empirical support for management decisions in the service delivery processes and develop a justifiably valid and reliable guide to designing workable service delivery improvement strategies to enhance customer satisfaction. · The results of the study provide invaluable insights to the future successful service delivery by identifying the major determinant factors and the root causes of the problems. Accordingly it provides policy makers and managers of courts' with feedback that was allowing them to prioritize their efforts and it will serve as a basis for further studies on the issue.

Limitations of the Study
Conducting successful and unbiased research is a challenging task. It requires the commitment of the researcher and the respondent and adequate resources. The research is also geographically limited to Dire Dawa City Courts; the service delivery of the courts' have impact on those customers served in the City Courts' rather than those served in other Courts' in the same City . Moreover, the outcome of the study is mainly dependent on the individual responses of the respondents who participate in the study. So the result may not be generalized beyond the City Courts customer. The limitations of this study results delimited to the five service delivery dimensions of SERVQUAL model to measure customer satisfaction in the City Courts.

LITERATURE REVIEW
These several reports mentioned itemized a damning list of complaints about the inefficiencies and dysfunction of the justice system without characterizing any of them as service quality issues. Service quality issues are included an academic literature, but the literature has concentrated on hospitality management (Crick & Spencer, 2011;Gilbert et al., 2004), information systems (Golding, 2005;Golding, Donaldson, Tennant, & Black, 2008), and institutional change (Schoburgh, 2014).

Service Quality
Three decades ago, when  first attempted to develop a conceptual model of service quality, they complained that "… the literature on service quality is not yet rich enough to provide a sound conceptual foundation for investigating service quality." This is certainly no longer the case.  developed the conceptual model and refined the SERVQUAL scale (Parasuraman et al., 1991b1993). Adapting the SERVQUAL scale to the measurement context improves its predictive validity; conversely, the predictive validity of SERVPERF is not improved by context adjustments. In addition to, measures of services quality gain predictive validity when used in less individualistic cultures, non-English speaking countries, and industries with an intermediate level of customization (Carrillat et al., 2007). The original SERVQUAL scale was developed in five service industries in the private sector.  sought originally to capture ten dimensions of service quality, but in later studies reduced these to five Zeithaml et al., 1990). Identifying five dimensions may not be suitable or possible in all cases. Some research projects have relied on fewer dimensions (Chakrapani, 1998). It has been suggested that service quality "… is best represented as an aggregate of the discrete elements from the service encounter," which may include these five dimensions as well as others (Carrillat et al. 2007, Cronin and Taylor, 1992. Lam and Woo (1997) classified them as the American school and the Scandinavian school of thought. The Nordic model, or Scandinavian school, is associated with Grönroos (1984Grönroos ( , 1990Grönroos ( , 2000Grönroos ( , 2001) and the American model or school, is associated with ; Berry et al., (1988) and Zeithaml et al., (1996). The American model also incorporated two different approaches to service quality measures. The first is based on the disconfirmation paradigm where satisfaction was seen as, "a function of the disconfirmation of performance from expectation" (Oliver, 1980, cited by Lee, Lee, & Yoo, 2000. In  application of this model, service quality is assessed as the difference between the expectation of the service and the perception of the service. This is the gap model and it is described as the "confirmation and disconfirmation of expectations approach" With this approach, the more one's perceptions exceed that which was expected, the better rated is the service; conversely, the more one's expectations exceed that which is perceived, the worse the service is considered to be . SERVQUAL is built on this paradigm (Parasuraman et al., 1991a, and it has found numerous applications as a service quality measure (Babakus & Boller, 1992;Brysland & Curry, 2001;Donnelly, Mike, & Shiu et al., 1999;Donnelly et al., 2006;Donnelly, Wisniewski, Dalrymple & Curry, 1995;Dyke et al., 1999;Lam & Woo, 1997;Parasuraman et al., 1991bWeekes, Scott, & Tidwell, 1996;Wisniewski, 2001aWisniewski, , 2001b. Moreover, it continues to be very popular as a service quality measure across many cultures (Aghamolaei et al., 2014;Musah et al., 2015;Ogunnowo, Olufunlayo, & Sule, 2015;Roslan, Nor, & Wahab, 2015). The instrument has been described as the most commonly used tool for measuring service quality.

Measuring Service Quality
The second approach of the American model to service quality is based on the performance paradigm. The assumption is that the expectation of service is always qualified or conditioned by the service received and therefore in measuring service quality expectations are ignored, and only the perception or performance questions are considered (Cronin & Taylor, 1992). The SERVPERF instrument is built on this paradigm, and this too has its full share of adherents (Al Khattab & Aldehayyat, 2011;Bayraktaroglu & Atrek, 2010;Carrillat, Jaramillo, & Mulki, 2007;Cronin & Taylor, 1994;McAlexander, Kaldenberg, & Koenig, 1994). Both frameworks are similar, SERVQUAL measuring expectations and perceptions, each on a 22-item scale, while SERVPERF measures perceptions, using only the perception battery of questions.

Multidimensionality
There is an additional difference between SERVQUAL and SERVPERF. The latter treats service quality as one dimensional. This is referred to as overall service quality (OSQ). SERVQUAL, on the other hand measures OSQ, and further breaks down the scale into several sub-categories or service quality dimensions (SQDs). There is still debate whether it is necessary to rely on the multidimensional formulation of service quality, or even whether it is desirable to use the gap between expectation and perception as a measure (Cronin & Taylor, 1992;Finn & Lamb, 1991). The literature provides support for both sides of the debate (Durvasula, Lysonski, & Mehta, 1999), and some studies have treated SERVPERF as a multidimensional measure (Al Khattab & Aldehayyat, 2011) Rhee and Rha (2009) argued that since the 1990s, service quality has been regarded as critical for reinventing the public sector. At the same time it was recognized that, "Adapting quality management approaches to the public sector has not been particularly easy" (Brysland & Curry, 2001). Assessing the quality of public services, by the nature of the public services environment, is just difficult (Donnelly et al., 2006). However, there is now a rich body of scholarship applying service quality measures to the delivery of public services (Carvalho, Brito, & Cabral, 2010;Wisniewski, 2001b;. The early applications were not all satisfactory. More recent applications of SERVQUAL to the public services have been much more successful. These include citizens service centers in Greece (Chatzoglou et al., 2013), public health care facilities in Brazil, Greece and Nigeria (Ogunnowo et al., 2015;Rocha et al., 2013), public Hospitals in Iran and Saudi Arabia (Aghamolaei et al., 2014), public physiotherapy service in Brazil (Almeida et al., 2013), and public transport in South Africa and Turkey (Govender, 2016). Rhee and Rha (2009) complained that for the most part, the literature directly applied the five SERVQUAL dimensions of service quality without exploring or validating these dimensions in the public sector and that only a few studies identified other dimensions such as equity and feedback, which are important to public services. Public services relationships are unique and distinct from private services relationships. However, the issues identified by Rhee and Rha (2009) are relevant to where members of the public directly interact with the public service agency to receive a benefit.

Understanding the Gaps Model
The SERVQUAL gaps model seeks to measure service quality as "…the calculated difference between customer expectations and performance perceptions of a service encounter" (Carrillat et al. 2007). As such, the model identifies five gaps that account for unsuccessful service quality delivery. Since the development  and refinement (Parasuraman et al., 1991b) of the conceptual service quality model, it has assumed increasing importance in identifying service quality gaps in the delivery structure of many types of services providers (Deshwal, 2011;Johns & Tyas, 1996;Ogunnowo et al., 2015;Xianying & Qinhai, 2007). The conceptual model of service quality, Zeithaml et al. (1990) identified five gaps. These service quality gaps are Gap 1, the management perception gap; Gap 2, the quality specification gap; Gap 3, the service delivery gap; Gap 4, the marketing communication gap; and Gap 5, the perceived service quality gap (Arokiasamy & Abdullah, 2013).
Gap 1 identifies the difference between the manager's perception of the customers' expectations, on the one hand, and the customers' actual needs and desires, on the other. It is important that managers accurately perceive customers' expectations.
Gap 2 identifies divergences in the specifications of service quality, where the manager may accurately identify the customers' needs but cannot translate that into the appropriate service specifications.
Gap 3 identifies those cases where quality specifications may be accurately identified but these specifications are not met in the delivery process.
Gap 4 identifies those cases where market communications promise services that are not delivered, and Gap 5 identifies cases where the perceived service quality falls short of the customers' expectations (Zeithaml et al. 1990). All these gaps are important to a full understanding of service quality delivery in an organisation and, as can be seen from Figure 3, these are all intimately related. This study focused on Gap 5 and in keeping with the conceptual framework identified in statement of problem. It sets out to identify where the perceived service quality fell short of the expectations of the customers whose was registered in the City Court.
A study conducted by Murillo et al (2014), entitled "Understanding The Service Quality Perception Gaps Between Judicial Servants And Judiciary Users" SERVQUAL's results have elicited that the main sources of users' dissatisfaction (defined as the difference between service users´ expectations and perceived service quality) for the entire Costa Rican population of Judiciary service users were speed, accurate reporting upon service conclusion, service delivery within promised time, constant willingness to answer queries, performing well since the very beginning and genuine interest to provide a solution. Findings herein were found useful in further extending the latter reported results, as they actually suggest that Court´s staff priorities are not focused on factors comprising the Capability and Response Speed Service Dimensions.
Research has consistently shown that the perceptions of those using the courts are influenced more by how they are treated and whether the process appears fair, than whether they received a favorable or unfavorable result. Thus, one of the important aspects of the quality approach and the 'search for excellence' is that it takes the needs and perceptions of Court customers into account. Court users include but may not be limited to members of the public and businesses making use of the services of the courts (e.g., litigants, witnesses, crime victims, those seeking information or assistance from court staff, etc.) and professional partners (lawyers, public prosecutors, enforcement agents, legal representatives of governmental agencies, court experts, and court interpreters). Accordingly, measures must address not only the level of satisfaction with the outcome of the court proceeding, but also the level of satisfaction with how the parties, witnesses, and lawyers were treated by the judges and the court staff; the (perceived) expertise of the judges and staff; and the fairness and understandability of court procedures and decisions Justice, as a public service includes positive characteristics, as well as flaws, that are obvious to all users. Users develop a series of service delivery perceptions based on their personal experiences, information gathered from their surroundings and reactions to observed judicial system dysfunctions (López & Zúniga, 2014, cited in Rodrigo, M., 2014.) Moreover, given the multiple and diverse stakeholders involved in judicial service delivery, i.e., plaintiffs and defendants, witnesses, victims, users and law professionals, there are different levels of expectations in terms of service quality. In this sense, Judiciaries must take on the responsibility for developing and promoting flexible and informal channels for judicial users to freely make suggestions on and complaints against, system or staff performance. This way, the judicial system would have a direct feedback mechanism supporting of sustainability and continuous improvement. If Judiciary service quality level is linked to users' satisfaction provided their real needs and expectations are taken into account, judicial system justices are responsible for reinforcing and assuring delivery of an 'acceptable' service quality level. Moreover, through provision of effective and efficient service to meet the needs of the public at large the government introduced the civil service reform program.

METHODOLOGY OF THE STUDY 3.1. Research Design
The study would be carried on the descriptive design in which both quantitative and qualitative data analysis were used to produce richer and more complete information. The reason behind using descriptive study design is because the researcher was interested in describing the existing situation under study. This study was also use explanatory study design, to explaining, understanding, predicting and controlling the relationship between variables and solve the research problem at hand, the study was used a SERQUAL model and some different statistical method to analyze the data. To determine the service quality & customers' satisfaction levels related to the FIC & AC, the study are used SERQUAL model; and to describe as well as to explore the satisfaction of the customer the study are uses summery statistics and logistic regression.

Target Population
The population of the study was all the customers served on the Dire Dawa City Courts First Instance Court and Appellate Court (FIC &AC).

Sampling Technique
This study possessed the probability sampling techniques since the sampling units under the study have equal chance being a sample. There are four types of probability sampling technique, and among of those this study was used the stratified probability sampling techniques. Under stratified sampling, the population is divided into two subpopulations (strata) that are individually more homogeneous than the total population and then we select items from each stratum to constitute a sample.

Sample Size determination
In conducting researches that require taking a sample, we always have the stage of deciding the sample size. Based on the above information, there are several formulas developed for sample size calculation that conform to different research situations. To determine the sample size and representative of the target population, the study was used statistical instrument formula. The following formula used for the calculation of the sample size since it is relevant to studies where a probability sampling method is used (Watson, 2001).The sample size of study computed as follows. Where, n = sample size required = 290 N = number of population = 1,400 P = estimated variance in the population = 50% A = margin of error = 5% Z = confidence level = 1.96 for 95% confidence R = estimated response rate = 96% The proportional stratified sampling for each stratum will be determined by n /N = 290/1,400 = 0.20 which means that 20 % of each clusters as calculated in the table.

Data Source
Primary Data source are the actual customer those which would be get service on the Dire Dawa City courts. For instance, the actual customer is the person that would be getting the service in the First instance court and appellate courts from September, 2017 up to April, 2018. In addition to this, judges and registrar for federal courts and city courts, attorneys, public prosecutors and police were another primary source of data for the study. The secondary data sources included in the study are legislation, books & journal, different organizational document, and articles. Particularly the literature for principles of judicial independence, principle case flow management and their essentiality for the efficient quality of courts, and access to justice were as a source of data.

Type of Data and Collection Technique
In order to achieve the intended objectives the data collection instruments was very important. The data of the study was quantitative (expressed in numerical form) and qualitative (expressed in the form of verbal descriptions rather than numbers).

Variable of the study 3.5.1. Dependent variable (response variable):
It is a variable which is depends upon or is a consequence of the other variable (variables). Ø Customer satisfaction 3.

Independent variable (explanatory variable):
It is the variable that is antecedent to the dependent variable and also a variables that stand alone and not changed by other variable.
Ø Demographic profiles of customers' (Age, Sex, Marital status, Educational level and Frequency for cases) Ø Tangibility Ø Reliability Ø Empathy Ø Responsiveness Ø Assurance

Method of Data Analysis
After gather, all necessary information the intended tools data processing and analysis would be properly implemented. The data was then processed using Statistical Packages for Social Sciences (SPSS) Software Version 23 and analysis methods included descriptive and inferential statistical techniques. Statistics analysis entails organizing data for description and decision-making. Depending on the data type, both descriptive and inferential statistics were used to analyze it.

Inferential Statistics
It describes the data with making inference or conclusion and summarizing sources of numerical data in to meaningful form. In this study chi-square test, SERVQUAL model and logistic regression were used.

Logistic Regression
The independent or predictor variables in logistic regression could be discrete, continuous or a mix of both. Logistic regression makes no assumption about the distribution of the independent variables. They do not have to be normally distributed, linearly related or have equal variance within each group. The model for logistic regression analysis assumes that the outcome variable Y is categorical. When we assume that Y is dichotomous, taking a values of 1 (that is, success, or satisfied customer) and 0 (that is, failure, or not satisfied customer).Then the conditional probability that the respondents are satisfaction of customer with judicial service delivery given the X set of predictor variables is denoted by Prob (Yi =1│X) =Pi. The expression Pi has the form: ( 1│X) pr Pi = the probability of customer satisfaction on the service delivery for city court Yi = the observed satisfaction of customer i β is a vector of unknown coefficients. The model given in the above is logistic regression model. The relationship between the predictor and response variables is not a linear function in logistic regression; instead, the logarithmic transformation of equation yields the linear relationship between the predictor and response variables. The logit transformation of Pi given as follows: Logistic regression is a form of regression, which is used when the dependent variable is dichotomous, such as satisfied (when the response is very high, high, and moderate) and not satisfied (if the response is low and very low). The logistic model is used to analyze and measure the level of satisfaction of customer towards on the judicial service delivery in probability (Agresti, 2007).

Chi-square Test of Association
Chi-square test of association are used to test the relationship between two categories exists or not. Which means it deals there is a significance association between the variable or not. Model: Where: O=observed count E= expected count The hypothesis: H0: There is no significant relationship between service delivery and customer satisfaction. H1: There is a significant relationship between service delivery and customer satisfaction.

Response rate
The survey was conducted between September, 2017 -April, 2018 questioners were distributed to 290 randomly chosen customers of First instance court and appellate courts of Dire Dawa City courts.

Demographic Profile
From 282 study participants, 216 (77%) were males and 66 (23%) were females. This indicates that the males' respondents were more than that of female respondents. When we look at the age distribution of respondents, 227 (80%) of the respondents were between the age of 18 to 40 years, 55(20%) of them were above 40 years which implies majority of the respondents were among economically active population. (See Appendix 1)

Reliability
Cronbach's alpha reliability analysis was conducted on the independent variables in order to determine the reliability of the instrument used. Nunnally (1978) has suggested 0.70 as the acceptable level for reliability measure. Alpha values ranged from 0.767 to 0.867, thus indicating an acceptable level of reliability.

Factor Analysis
According to Coakes and Steed (2007), Factor Analysis is a data reduction technique used to reduce a large number of variables to a smaller set of underlying factors that summarize the essential information contained in the variables. Factor Analysis is more frequently used as an exploratory technique when to achieve the objective of summarizing the structure of a set of variables. Before conducting factor analysis, items had been classifying into five domains, namely Tangible, Reliability, Responsiveness, Assurance and Empathy. Factor analysis was to run by using principle component analysis according to items in each domain across and components extracted by way of varimax rotation. Kaiser-Meyer-Olkin (KMO) is measure of sampling adequacy also examined to validate factor analysis, between the ranges of 0 to 1, with 0.6 as the minimum level. The results of the Barlett's test were significant for all variables and KMO is the above-recommended value of 0.6 for all variables meaning that it indicates sampling adequacy. .000 .000 .000 .000 .000 Source: Survey Result (2018), Tangible (TANGI), Reliability (RELI), Responsiveness (RESPO), Assurance (ASSU), Empathy (EMPA)

Result of SERVQUAL Composite Score Analysis
The various perceptions under each service quality dimension were grouped into five constructs and calculated its individual average to get composite scores. The modified version of SERVQUAL as proposed by Parasuraman et al., (1998) which involve five dimension of service quality is used to group the various perceptions of items and the composite scores were calculated. The table below shows the result of the service quality dimensions composite score.

. Service Quality Gap Analysis (Perceptions -Expectations)
Expectations and perceptions were both measured using the 5-point Likert Scale whereby the higher numbers indicate higher level of expectation or perception. In general, consumer expectation exceeded the perceived level of service shown by the perception scores. This resulted in a negative gap score (Perception -Expectation).
According to  it is however common for consumer's expectation to exceed the actual service perceived and this signifies that there is always need for improvement. From the table one can understand that the average score of customer expectations was mean value of 4.414 and the overall perception score was 3.548. The score shows a -0.4496 difference. This means that the expectations are higher than the perceived service quality. In general, the negative score for the SERVQUAL dimension were showed that there is a room for improving the City Courts service quality. As can be seen, the highest gap score were tangibility (-1.0375) followed by assurance (-0.865), responsiveness (-0.855), reliability (-0.798) and empathy (-0.7675) of which the gap score in all dimension is pronounced and need attention from the service provider point of view.

General Satisfaction Level of Customers
On the other hand, customer were asked to rate their general level of satisfaction on service received at the City Courts in a single question and as the table 4.5 shows about 29.7 % of respondents who rate their level of satisfaction as very high plus 33.36% of respondents rate their level of satisfaction as high and 9.576% respondents rate their level of satisfaction as moderate, the respondents think that the service meets their overall expectation and satisfaction 73 %. On the contrary, the rest 25% of respondents did not think that the service meet their expectation and unsatisfied; that is 10.64% of respondents rate that their level of satisfaction from the service they got is low plus 14.56 % of respondents rate their level of satisfaction from the service they got is as very low.

The Relationship between Service Delivery and Customer Satisfaction
The Chi-square test for association is used to discover if there is a relationship between two categorical variables. In other words, it tests whether or not a statistically significant relationship exists between a dependent and an independent variable. Where the dependent variable is the customer satisfaction in the City courts and the in dependent variable are the service delivery dimensions. The Chi-square statistics determines whether the observed values from the sample and expected values from the specified distribution are statistically different compared to the p-value (sig. value) to the significance level. Usually, a significance level (denoted as α or alpha) of 0.05 works well. A significance level of 0.05 indicates a 5% risk of incorrectly rejecting the null hypothesis.
Thus, P-value (sig.) ≤ α: The observed data are statistically different from the expected values (Reject H0). In other word, If the p-value is less than or equal to the significance level, reject the null hypothesis and conclude that there was a significant association between the service delivery and the customer satisfaction in the City Courts. Therefore H1 is accepted. This implies that, the service deliveries of City Courts in the basis of five dimensions which are tangibility, reliability, responsiveness, assurance and empathy have significant association with the satisfaction of customer.

The Logistic Regression
Logistic regression is accepting quantitative, binary or categorical predictors and codes the latter two in various ways. Here's a simple model including a selection of variable types the criterion variable is satisfaction of customer on the service delivery of Dire Dawa City Courts and the predictors are responsiveness, tangibility, reliability, assurance, empathy, age, sex, marital status, education level and court frequency of delivering service in the city courts. But the following logistic regression table are includes some predictors which are uses for interpretation. The table also includes the test of significance for each of the coefficients in the logistic regression model. For small samples the t-values are not valid and the Wald statistic should be used instead. Wald is basically t² which is Chi-Square distributed with df =1. However, SPSS gives the significance levels of each coefficient. As we can see, reliability, assurance, frequency of the customer come in the court has significant effect and other variables are not.
These estimates tell you about the relationship between the independent variables and the dependent variable, where the dependent variable is on the logit scale. These estimates tell the amount of increase (or decrease, if the sign of the coefficient is negative) in the predicted log odds of satisfaction = 1 that would be predicted by a 1 unit increase (or decrease) in the predictor, holding all other predictors constant. For the independent variables which are not significant, the coefficients are not significantly different from 0, which should be taken into account when interpreting the coefficients. (See the columns labeled Wald and Sig. regarding testing whether the coefficients are statistically significant). Because these coefficients are in log-odds units, they are often difficult to interpret, so they are often converted into odds ratios. You can do this by hand by exponentiation the coefficient, or by looking at the right-most column in the Variables in the Equation table labeled "Exp(B)".
Reliability effect is positive, which means with a one unit increase on the five-point reliability scale being associated with the odds of customer satisfaction increasing by a multiplicative factor of 1.601. In other word, for every one-unit increase in reliability score (so, for every additional scale on the reliability), we expect a 1.601 increase in the log-odds of satisfaction of the customer, holding all other independent variables constant. Similarly, the effect of assurance on the satisfaction of customer in the city courts service delivery has positive effect. Thus, for every one-unit increase on the five-point assurance score, we expect a 1.851 increase in the log-odds of satisfaction, holding all other independent variables constant. For instance, when an increasing of the assurance dimension delivery in the city courts with one unit are yields 85% increment in the satisfaction of the customer. In addition, the frequency customers observed the City Court to get the service has also a positive effect for satisfaction of customer. Thus, the customer who observed the city courts up to two times to get services are16.489 more likely to satisfy than frequency of served in the City Courts more than two times. In other word, when the customer have come to the Courts 'to get service more frequently yields a more un-satisfaction. Therefore, the logistic regression using 10% significance level (p-value<0.01) the reliability, assurance and frequency to visit the city courts of the customer have positive effect on the satisfaction of customer. Thus, if the City Courts are provide more service on the reliability dimension, assurance dimension, and minimizing of visiting frequency to the customer would increase the satisfaction of customer.

Conclusion
The study was conducted to analyze service delivery and customer satisfaction in Dire Dawa City Courts. Gap analysis (disconfirmation approach) was employed to measure the service quality. Accordingly, a five dimensional instrument comprising of tangibility, reliability, responsiveness, assurance and empathy have been used for the study. The finding of the study showed a room for improvement for all dimension. The SERVQUAL result indicated that among the various service quality dimensions in the Courts responsiveness was the most important determinant of service quality followed by assurance, reliability, tangibles and empathy. Moreover, the SERVQUAL result depicted that responsiveness is the most important factor to have a positive and significant effect on customer satisfaction followed by reliability, assurance, empathy, and tangibility. The Chi-square test showed that, the five service quality dimensions (tangibility, responsiveness, assurance, empathy and reliability) were positively and significantly related with customer satisfaction. The overall satisfaction analysis result showed that 73% customers were satisfied by the City courts service. From the logistic regression, the City Courts are provide more service on the reliability dimension, assurance dimension, and minimizing of visiting frequency to the customer would increase the satisfaction of customer.

Recommendation
The following are recommendations from the research outcome § The City court judges require supporting training (like on customer handling, jurisdictional power), seminars about cassation bench decisions, experience sharing with concerning body and so on are important to improve the satisfaction of customer and the competence of the judges. Similarly, registrar of the courts should be provided training on customers handling and management. § Since customers have the highest expectations on responsiveness dimension of SERVQUAL model which shows it is the most important dimension of service quality for customers followed by assurance, the City Courts should work on responsiveness by giving trainings to employees to improve their ability to give prompt service, and willingness of the employees to help customers. It also improves the empathy dimension by training its employees to make them competent enough to deliver the service and to have the customer's best interest at heart.