This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
This study examines the dimensions of service operations flexibility and its relationship with service firms' performances. Service operations flexibility is operationally defined as the capability of service firms to adapt to internal and external driven changes. It is further subdivided into six dimensions: design, package, variety, delivery, service recovery, and system robustness. Service firms' performances consist of cost, customer service and financial. Respondents are operations managers or equivalents who are involved in the decision making process in operations. All items used to measure the studied factors are found to be reliable and valid. The preliminary study was based on a large scale survey that involved two service industries; private hospitals and auto repairs, in four emerging economies countries namely Malaysia, Indonesia, India, and South Africa. The two industries are considered as service shop, characterized by high degree of customer interaction and low labor intensity (Schmenner, 1986). The structural relationship between service operations flexibility and service firms' performances is tested using Structural Equation Modeling. The final model fits the data with substantial changes from the original hypothesized model. As we have not found any moderation effects of industry type, we believe that the model is suitable for both hospitals and auto repairs. We discuss the findings and suggested implications for theory and practices.
Keywords: Service operations flexibility, firm performance
High persevering firms prove themselves to be able to withstand challenges during the time of crisis and challenges. Due to the lack of this quality, many service firms struggle to maintain business operations during economic turbulence. The ability to be flexible could be the answer to these challenges. This scenario leads to the call for greater operations flexibility to strengthen organizations. With the ever-changing nature of a global business operation that requires firms to adjust rapidly, world-class service organizations rely on the right strategies and practices to enhance their operational flexibility. In Malaysia, for instance, the world's best budget airline, AirAsia, applies certain principles, practices and procedures that align with its operations objectives to achieve appropriate level of flexibility in its operations that suited its market segments' requirements (Idris, 2007; Idris, 2008). On the other hand, due to lack of operational flexibility, there was a lot of confusion on the part of the passengers and employees of Jet Blue airlines in New York when most flights were delayed due to bad weather conditions (Strickler and Chung 2007).
Meanwhile, it can be said that operational flexibility in service organizations is the least researched topic as compared to other operational priorities such as cost and quality. One of the reasons held to be valid by some authors is that, this competitive operational objective is the means to an end rather than the end itself (Buzacott and Meandelbaum, 2008; Slacks, 2005). Buzacott and Meandelbaum (2008) review the development of flexibility studies from different perspectives and offer three paths of flexibility thinking: prior flexibility, state flexibility and action flexibility. What is nonetheless clear, in their review is the limited coverage of the service flexibility as compared to the manufacturing flexibility. Slack (2005) revisits his own article written in 1987, offering a recent development in the field mostly related to the manufacturing context. But, the attention to service flexibility is only highlighted at the end of the paper with few noted citations. Therefore it is compelling that researchers scrutinize what constitutes operational flexibility of service-based organizations. The importance of this issue is aggravated by the volatile conditions of the present business environment.
Quite recently, Bernardes and Hanna, (2009) have clarified differences between the terms flexibility, agility and responsiveness in the operation management literature. They propose that the term flexibility is most commonly associated with internal characteristics of the systems to change within pre-established parameters. Agility is used to describe a systematic approach to organizing quick system reconfiguration in the time of unforeseeable changes; and responsiveness commonly refers to the behavior of the system to react in the presence of modulating stimuli.
In any circumstances, organizations must strategize on how to deal with the impact of changing events that will affect their operations. Some of the pressure of changes must be dealt with at the source through standardizations of products, services and process delivery. The remaining problems must be dealt with at the point of impact using robust structural and infrastructural resource deployment strategies. One of the most essential moves to establish and eventually enhance the operational flexibility is the use of technology, especially the IT, to better communicate internally within organizational units and externally with their customers; thus providing flexibility in their operations. Others may rely on smart networking with clients and suppliers so that they will be able to handle the uncertainties together as a group. At the same time, having a flexible workforce will ensure a certain level of variability which will be reduced by tactically reassigning the workforce. In summary, the changing nature of the environment requires flexibility to be one of the primary competitive components to be applied and considered seriously. Studies suggest that capability in flexibility is both internally and externally driven. Internally, organizations must have a robust mechanism to produce consistent results and avoid confusion. Externally, organizations must be flexible with regards to the needs of the customers.
In summary, the changing nature of the environment requires flexibility to be one of the primary competitive components to be applied and considered seriously. In this paper, we will evaluate the dimensions of service operations' flexibility in 'service shops'. Our first objective is to validate the measures of service operations' flexibility (SOF) and service firms' performances. Secondly, we will test the structural relationship between service operations' flexibility (SOF) and company performance. The third objective will be the testing of the influence of industrial difference to the structural model.
2.0 LITERATURE REVIEW
Operational flexibility refers to the managerial capabilities that can be set up quickly in order to provide a rapid response to environmental changes that are familiar or routine in an organization (Verdu-Jovre et al., 2004). The term flexibility has been used in many areas of management including finance, automation, manufacturing, health care and human resources; however, in each area the definition differs depending on the types of product or service created (Chanopas et al., 2006). Research on operations' flexibility cover several broader areas. Zhang et al. (2002, 2003) describe flexibilities relating to machine, labor, material handling and routing as internal competencies, and flexibilities made of mix and volume flexibility as external. Originating from the manufacturing sector, researchers have expanded the concepts to the service sector and supply chain. While manufacturing and service flexibility mostly concern with the organization within, the supply chain flexibility and agility cover the entire chain of external suppliers. In general, the essence of the manufacturing flexibility, service flexibility and supply chain flexibility converge to two important groups, internal driven flexibility and external driven flexibility with variations on the dimensions of each category.
2.1 Manufacturing flexibility
Earlier works on this field have been done by many researchers. For example, Sethi and Sethi (1990) have brought to light 11 dimensions of manufacturing flexibility. Gerwin's (1993) propose taxonomy of seven dimensions of the same type of flexibility. D'Souza and Williams (2000), to add, have argued for more parsimonious four dimensions of manufacturing flexibility. They further categorize them into externally-driven dimensions that include volume flexibility and variety flexibility and internally-driven dimensions of process flexibility and material-handling flexibility.
More recently, Hallgren and Olhager, (2009) investigate the relationship between volume and product mix flexibility. They investigate how different flexibility configurations might be related to various manufacturing practices. They find that based on high or low levels of volume and mix flexibility combinations, flexibility configurations have demonstrated significant differences both in terms of operational performance, and in the emphasis that was put into different flexibility-sourced factors.
CamisoÂ´n, and LoÂ´pez, ( 2010) conduct a study to test the relationship between manufacturing flexibility and performance with the mediating roles of product, process, and organizational innovation. They conclude that firms that pursue manufacturing flexibility should build up innovation capabilities to obtain an improvement in the organizational performance, because simply using a flexible manufacturing system will not guarantee improvements in the firm's performance.
In another study, Wilson and Platts (2010) address the question of how mix flexibility is achieved during day-to-day operations. Frameworks from the coordination theory are used to explain day-to-day-operations in coordinating manufacturing resources. The authors find that mix flexibility requirements influence how a company achieves mix flexibility. At an operational level, mix flexibility is achieved through shared resources, floor dependency, and simultaneity constraints.
Hutchison and Das (2007) use a case study approach to examine and analyze the decision process that a firm undergoes for acquiring an advanced manufacturing system to obtain manufacturing flexibility for its operations. The guiding variables of a firm's decisions on the selection of options of manufacturing flexibility are exogenous by nature - strategy, environmental factors, organizational attributes, and technology.
Cousens et al. (2009) come out with a new conceptual framework to improve manufacturing flexibility through developing and testing a procedure, to enable organizations to establish a competitive capability and progress towards manufacturing flexibility.
Overall, we have been made aware of the facts that external flexibility is associated with mix and volume flexibility while internal flexibility is more associated with machine, labor, material handling and routing (Zhang et al. ,2002; 2003)
2.2 Supply chain Flexibility
Supply chain flexibility also relates with internal and external factors but is seen from a wider perspective of suppliers collaboration. For example, Braunscheidel and Suresh (2009) define a firm's supply chain agility (FSCA) as the internal capability of the firm including the ability to cooperate with its key suppliers and customers, to adapt or respond quickly and perhaps, maturely to marketplace changes. They propose that the organizational orientation consists of market orientation and leaning orientation which will influence the organizational practices in terms of internal integration, external integration and external flexibility. The three dimensions of organizational practices will lead to improvement in the firms' supply chain agility. The external flexibility elements considered are volume and mix flexibility. They have come to the conclusion that all three organizational practices significantly have a positive impact on the firm's supply chain agility.
Stevenson and Spring (2009) study specific inter-firm practices that are used to achieve increased flexibility in buyer-supplier pairs and in the wider supply chain or network. Supply chain flexibility is grouped into ten categories and relates to two distinctive themes; configuration flexibility and planning and control flexibility. They observe that to reduce their own needs for internal flexibility, firms have the tendency to use various forms of outsourcing and subcontracting. To be externally flexible, firms ought to fully commit with their counterparts along the full chain.
Squire et al (2009) study the relationships between supplier capabilities, supply chain collaboration and buyer responsiveness and found that those suppliers' capabilities in terms of flexibility, responsiveness and modularity directly affect buyer responsiveness with the level of buyer-supplier collaboration moderating this relationship.
Tachizawa and Thomsen (2007) focus on the reasons for firms increase supply flexibility which is related to the upstream supply chain. They have found that firms require flexibility for several vital reasons such as manufacturing schedule fluctuations, JIT purchasing, manufacturer slack capacity, low level of parts commonality, demand volatility, demand seasonality and forecast accuracy. They also observe that companies increase this type of flexibility by improving supplier responsiveness and having a flexible sourcing. The results also suggest that the supply flexibility strategy being selected would depend on the type of uncertainty (mix, volume or delivery).
2.3 Service operations' flexibility
Harvey et al. (1997) suggest that a flexible service firm is the one that can handle variability with minimum penalty and highlight the possibility of the difference between internal robustness and external flexibility. Internal robustness must be dealt with minimum efforts due to the fact that it will not create value to customers. Harvey et al. (1997) also suggest that in order to deal with the internal variability, firms may be required to have some specified organizational arrangements, such as cross-functional teams, empowering contact personal, and a flat organization, which factors are related to infrastructural elements of operations, as well as the modification of the structural elements, like the networking capability. It is the external flexibility that must be managed carefully in order to gain competitive advantage. Harvey et al. (1997) suggest the use of structural element, mainly the manipulation of the IT technology in order to manage flexibility.
Customer interaction and customization requires a high degree of flexibility where the main input flow is information. For this reason, process equipment in service industries are mainly based on information technologies (Gouillart and Sturdivant, 1994). Besides, managers must also tackle personnel flexibility and versatility as these factors assume a higher role in the service delivery compared to machines in manufacturing industries (Harvey et al., 1997). Some researchers observe that flexibility in services involves quick introduction of newly designed services into the service delivery system, handling changes in the service mix and variations in customer delivery schedules, rapid adjustment of capacity and customized services (Aranda,2003;Suarez et al., 1996). Guided by studies from the manufacturing flexibility context,
Correa and Gianesi (1994) link uncertainty and variability with unplanned changes, which require firms to understand the concept behind these non-expectations. Managing unplanned change can be divided into two dimensions. One is categorized as the flexibility in dealing with change after the unplanned change has occurred. The second dimension is the ability to reduce with a certain amount of change and reducing the effect of change. This can be done by finding ways to control the changes by implementing strategies like the forecasting technique, system maintenance, parts standardization, and manufacturing focus. These strategies are to prevent and avoid the change before it occurs. The unplanned change control actually acts as a filter to reduce the amount of change that affects the whole system. The changes which pass through the control filter have to be dealt with by the system, through the system's flexibility. Correa and Gianesi (1994) associate this framework with the service operation's flexibility and introduce six dimensions of service operations' flexibility namely design flexibility, package flexibility, delivery time flexibility , delivery location flexibility, volume flexibility, system robustness flexibility and customer recovery flexibility. Silvestro (1993), in turn, proposes three dimensions of service flexibility: volume flexibility, delivery speed flexibility and specification flexibility. Aranda (2003) has even adapted a manufacturing flexibility study and renamed the flexibility dimensions as expansion; distribution of information; routing; equipments and personnel; market; services and servuction, process, programming and volume flexibility.
Overall, previous research in service flexibility had heavily relied on manufacturing studies to conceptualize either the externally-driven dimension, or the internally-driven dimension of service flexibility.
2.4 Flexibility and Firms' performances
What organizational performances could be most affected by the enhancement of operational flexibility? Customer service could be the immediate result. Daugherty and Pittman (1995) imply that flexibility in operations would result in better response to customer needs. Also in the short run, a flexible operation system would also lead to better cost performances in terms of reducing clients' costs, attaining high employee productivity and increasing high capacity utilization as the system will produce consistent results in the forms of less occurrences of glitches and confusions. In the long run, we expect indirect benefits of financial performances such as return on asset and return on investment.
Other authors have discussed the impacts of operations' flexibility on other performance. For example, Selveira (2006) cites studies in manufacturing sectors that relate flexibility and performances such as that of Swamidass and Newell (1987) on flexibility and growth and profitability, Fiegenbaum and Karnani(1991) on extra profit, Narashiman and Das(1999) on flexibility and cost reduction, Jack and Raturi (2002) on financial and delivery performance For service firms, Aranda (2003) who studies Spanish engineering firm and discovers that flexibility moderates the efficiency performance but not customers' satisfaction. Slacks (2005) argues that the issues of flexibility apply both to manufacturing and services firms. Thus, the flexibility in operations impacts firms' performances, be it the manufacturing or service-type of companies.
This study involves the usage of questionnaire as the method for data collection on the causal effect of the service operations' flexibility to firm performance. This questionnaire adopts the 7-point Likert scale ranging from 1 = strongly disagree to 7 = strongly agree for the service operations' factor whereas for performance of the firm, the range is set from 1 = lower to 7 = higher. These 2 different indications for the 7- point Likert scale used, depend on the nature of the items or factors that are studied. The number of items for service operations' flexibility and firm performance were 10 items and 11 items respectively. Questionnaires were distributed to 2 types of services which were the hospital and auto repair services. The questionnaires were distributed to respondents from four different countries i.e. Malaysia, India, South Africa and Indonesia. See Table 1.
Table 1. Country and service categories
We have narrowed down our selection to four economically- developing countries with the assumption that the business environment is more volatile, thus studying service flexibility should become more relevant and important. The two industries are considered as 'service shop' characterized by their high customer contacts but low intensity of labor use. When customers are more involved in acquiring the service delivery, business environments would normally become more difficult to handle, thus some proof of flexibility in their operations may help ease the situation.
212 questionnaires were successfully returned for analyses and this compiled questionnaires is sufficient for the multivariate analysis i.e. Structural Equation Modeling (SEM) as Hair et al. (2006) also commented that the number of samples as small as 100 to 150 observations could be used to start using the SEM depending on conditions of the data. Moreover, the maximum likelihood estimation provides valid and stable results with sample size as small as 50 under ideal conditions (Hair, 2006). The following table displays the respondents' profiles as representatives of the study sample.
Table 2. The profile of respondents (N=173)
Years of operation of Firm
More than 10 years
The 212 dataset was coded and saved into SPSS and during the process of data screening for outliers, many data sets were considered as the missing values and were deleted from the analysis, and consequently only 173 dataset were left to be analyzed. This sample size was still deemed adequate for the application of structural equation modeling (SEM) to address the research objectives. In short, this study has applied a three-stage structural equation modeling, using AMOS version 18, a model-fitting program to assess the validity of the measurement model, the confirmatory factor analysis of the service operations' flexibility towards firm performance. Next, we examined the good-fit of the full-fledged hypothesized model as in Figure 1. Finally, we validated the model to assess the moderating effects of types of services and types of countries on the hypothesized model.
Table 3. Measurement of the variables of the hypothesized model
We have been able to offer new, unique and innovative services to our customer
We have been able to integrate some features of services into an alternative package
We have been able to offer a large number of service features and displays a varietyof services
We have been able to anticipate the service delivery to customers'requirements
We have been able to recover the service to customer after something goes wrong
Our ability to remain operating effectively,despite some shortcomings in elements of service has been enhanced
There seems to be less confusions in procedures for the employees to execute their responsibilities
Managers seem to contradict themselves while making important decisions
There has been fewer termination of activities due to maintenance glitches
Employees know what to do when there is a system failure such as 'blackout' or accident
Attaining high employees'productivity
Maintaining high capacity utilisation
Response to customers'/clients'requests
Response to customers'/clients'complaints
Growth of market share
Return on assets
Return on Investment
3.1 Assessing validity and reliability
Our first objective is to validate the measures of the studied constructs, i.e the SOF and the performances. Hair et al. (2006) define reliability as an assessment of the degree of consistency between multiple measurements of a variable. This study assesses the consistency of the entire scale with Cronbach's alpha and its overall reliability of each factor of service operations' flexibility and firm performance. Both factors have yielded alpha coefficient which exceeds the values of 0.70 suggested by (Hair et al., 2006) and 0.60 as suggested by Nunnally (1978). (See Table 1). From this result of Cronbach's alpha coefficient value, this questionnaire is well accepted and admissible, or in short, proves to be reliable. In order to validate the instrument, this study also considers construct validation using the analysis of moment structures software (AMOS) with maximum likelihood (ML) to analyse the data. This approach is called the confirmatory factor analysis which is more advanced, as the hypotheses are based on the underpinning theory (Norzaidi & Salwani, 2009), as discussed in the next section.
3.3 Inter-item correlations matrix
Table 4 and Table 5 reveal the inter-items' correlations among items in the service operations' flexibility construct and items in the performance construct. The correlations among the bivariate items show that there has been no item that is greater than 0.9. This supports the fact that there is no multicollinearity that exists among the items. It gives initial evidence that the items are distinct to each other to represent the specified construct. However, there is a negative offending correlation that contaminates the estimates using Pearson correlation coefficients. Most of the items are significant at the 0.05 level.
Table 4. Correlation among items for Service Operations' Flexibility (SoF)
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
Table 5. Correlation among items for Firm Performance (Perf.)
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
3.4 Confirmatory Factor Analysis (CFA)
In this study, the confirmatory factor analysis is used to determine the construct validity of the questionnaire items. In simpler terms, the analysis is adopted to know how well can the construct explain the variables under the construct (Siti Aishah & Kaseh, 2008). In this analysis, whenever the correlation of the items within the same construct is relatively high it is said to have the construct validity. Also, when the factor loading or the regression weight and the squared multiple correlations (SMC) of the items are significantly correlated to the specified construct, this would also contribute to the fact that the construct validity is present. In this study, we conducted 2 confirmatory factor analyses on each measurement model i.e. the measurement model of Service Operations Flexibility (SoF) and the measurement model of Firm Performance (Perf.).
Figure 1. Hypothesized model of service operations flexibility on the firm performance
3.5 The hypothesized model and the modeling strategy
The model to be tested postulates a priori that service operation flexibility has, indeed, influenced firm performance. (See Figure 1). As we believe that this is a first attempt to quantify SOF for service industries done in emerging market settings, we prefer to set a fairly simple structural model of a one-to-one relationship of the latent variables; i.e. SOF and company performances. The items to measure SOF are implied and derived from Harvey et al. (1997) and Correa & Gianesi (1994). For the performance measures, they are common measures in many operation-related studies. The items were loaded onto respective factors of service operations flexibility (SOF) and firm performance (Perf.). Each of the factors was measured by 10 observed variables for SOF and 11 items for Perf., the reliability of which is influenced by random measurement error, as indicated by the associated error term. Each of these observed variables is regressed onto its respective factor.
Hair et al. (2010) stress 3 distinct types of modeling strategy i.e. the confirmatory modeling strategy, competing models strategy and model development strategy. Each of these three represents a slightly different approach in modeling. The confirmatory approach is the most straightforward strategy. As the name implies, the confirmatory approach lets a researcher specify a single model composed of a set of relationships and apply SEM to assess the model adequacy. In other words, the process is to find support whether the model fits the data. Secondly, competing model strategy revolves around testing several models i.e. the alternative models through overall model comparisons. The assessment of all models would yield the best model that could represent the data collected which is much stronger than a test of a single model alone. The last one is the model development strategy that begins with the basic model framework and following the adequacy and reasonableness of improving the framework through modifications of the structural or the measurement models. It starts with model that is built based on theoretical judgments that will be empirically tested using SEM. Following this, the model can be modified based on the researcher's judgments or suggestion given by the modeling software used and this re-specification must also be theoretically viable. In this study, the model development strategy was employed.
This section also presents the results of the structural equation modeling that address the objectives of the study.
4.1 Validity of the measurement model of service operations' flexibility and firm performance
The confirmatory factor analysis (CFA) was run for each of the measurement model as in the hypothesized model in Figure 1. Initially, it has been mentioned that each of these factors was measured by the 10 items for service operations' flexibility and 11 items on firm performance. Each item was assumed to load only on its respective dimension. From the initial findings of CFA, the measurement models yielded a non-fit model and therefore several modifications had been conducted until an acceptable fit model based on the suggestions forwarded by the modification indices is reached. (See Figure 2 & 3). In each case, a few items were removed due to violation of estimation. The revised versions of 2 CFAs for each of these latent constructs have demonstrated an adequate fit to the empirical data. [CFA for SOF: Normed Chi-square=2.129, p-value=0.030, CFI=0.986 & RMSEA=0.081 and for Perf.: Normed Chi-square=2.056, p-value=0.128, CFI=0.995 & RMSEA=0.078]. Therefore, the following Figure 2 & 3 are the re-specifications of the measurement model after the estimation using Maximum Likelihood is conducted. From the confirmatory factor analysis results in Table 6, we observe that the factor loadings of all observed variables or items are adequate, ranging from 0.56 to 0.93. The factor loadings or regression weight estimates of latent to observed variables should be above 0.50 (Hair et al, 2006; Byrne, 2001). This indicates that all of the constructs conform to the construct validity test which means that all items have belonged to the specified constructs. The results, therefore, have fulfilled our first objective.
Figure 2. Measurement ModelofService Operations Flexibility (SoF)
Figure 3. Measurement model of Firm Performance (Perf.)
Table 6. Final confirmatory factor analysis results of construct variables
Service Operations Flexibility
We have been able to offer new, unique and innovative services to our customer
We have been able to integrate some features of services into alternative packages that are requested by customers
We have been able to offer a large number of servicesfeatures and variety
We have been able to anticipate the service delivery to customers'requirements
We have been able to recover the service to customer after something goes wrong
Employees know what to do when there is a system failure such as 'blackout' or accidents
Maintaining high capacity utilisation
Response to customers'/clients'requests
4.2 Validity of the structural model of service operations' flexibility on the firm performance
In order to examine the hypothesized model, all the 2 measurement models were then structurally linked. As in Figure 1, it contains the measurement model of the service operations flexibility on the firm performance, both of which comprise of first order factors. In addition to this, the items that best explain the construct is the items that have higher loadings on the same construct and this can be referred to Table 4.
Figure 4. Standardised coefficients of the re-specified model
4.3 Bayesian Confirmatory Factor Analysis (CFA)
Byrne (2010) argues that the Maximum Likelihood (ML) estimation of Likert-scale items produces negligible effects of non-normal and non-continuous data whenever each variable/item has at least 5 categories of response and carries a large sample size. However, severe effects of non-normal non-continuous data occur whenever each variable has 4 or less categories of responses and a small sample size which is less than 200. Under this condition, this study could only use 191 questionnaires for analysis and therefore, the Bayesian estimation is recommended for re-affirming the previously conducted CFA in section 4.1. The Bayesian CFA analysis was conducted in the AMOS software to estimate the unstandardised weights produced by this analysis with the unstandardised loading obtained in the CFA using the Maximum Likelihood procedure. The results of the comparative analysis are shown in the following table.
Table 5. Comparative Analysis (Maximum Likelihood and Bayesian Estimation)
Inf. &St. >StL (Location)
From the results, we can see only a small difference exists between the loadings generated from both the ML estimation and Bayesian estimation. This supplies evidence that the CFA using the ML estimation in this study is acceptable and may proceed with the analysis.
4.4 Adequacy of the causal structure of the service operations flexibility on the firm performance
Figure 2 summarises the results of structural equation modelling of service operations' flexibility on the firm performance. The structural model has yielded consistency of the re-specified causal relationships with the data (relative Chi-square = 1.551; CFI=.985 & RMSEA = 0.057). All of these fit indices have satisfied their critical cut-off points and therefore this re-specified model fits the data following the re-specifications based on modification indices. The parameter estimates of the re-specified model are free from any offending values and the path coefficient of the causal structure is statistically significant at 0.005 levels, and is of practical importance. In sum, the analysis reveals that service operations' flexibility have well-explained almost 64% of the variability of the firm performance. Our second objective, has, therefore, been met.
Next, the final objective of the study was to examine the structural invariance of the re-specified model in Figure 2 across a moderating factor i.e. types of services offered. To test types of services-invariant, a simultaneous analysis on two types of services industry which were the hospital and autorepair samples was conducted. Firstly, the structural paths (service operations flexibility vs firm performance) were not constrained which produced the baseline Chi-square value. Next, the structural path was constrained in order to produce another chi-square value, which was then tested against the baseline value for statistically significant differences. The result of this invariance analysis was presented in Table 5.
Table 5. Results of multiple group modeling
Types of services
The invariance test across the types of services i.e. hospital and auto-repair has resulted in a statistically insignificant change in the Chi-square value, Chi-square (df = 1) = 3.28, p > 0.001. This means that the difference in the Chi-square values between the unrestricted model and the constrained model do not produce a poorer fit model. The path coefficient does not vary significantly across types of services. It is justifiable to conclude that types of services did not interact with the exogenous variable to influence the firm performance and therefore, type of services is not a moderating variable.
5.0 DISCUSSIONS AND CONCLUSIONS
The findings of the study show that 7 identified item that measured service operations flexibility influence 5 types of company performances, and the relationship is not moderated by the types of services, countries, and level of performance towards the firm performance; i.e. that the service operation flexibility model does comply with all countries and also to types of services as far as this study is concerned. From the 7 items; 5 are earlier identified from the literature to measure the external flexibility dimension and 2 items seek to measure the internal robustness. Our model, however, does not differentiate the two dimensions of operational flexibility (see figure 1). From the 11 items that are thought to measure company performance, only 5 items; mostly from customer service performances, have been included in the final model. In addition, the results have achieved the psychometric adequacy of the measure of service operations' flexibility to firm performance as in this research. This can be seen as all fit indices obtained, have exceeded the recommended values of GFI, CFI, TLI > 0.90, RMSEA < 0.08 (Mohamad Sahari, 2001); indicating that the model fits the data following several modifications or adjustments on the hypothesized model as suggested by Modification Indices (MI). The loadings range from 0.34 to 0.93 and succinctly the construct validity for service operations' flexibility on firm performance is supported. The result is achieved after taking into consideration the modification indices (MI) and we allow the residuals or error for items FPF4 and FPF 5 under the factor of service operations' flexibility to correlate as suggested by MI. Besides that, errors of items FPC3 and FPFi1 are also correlated. A close examination of those items in the questionnaire shows that the items were probably phrased in a very similar way according to the respondents' points of view. Errors in measuring the items are therefore hypothesized to correlate with each other as shown in Figure 2.
From the contemplative analysis, the item coded as FPSR 1 seems to be of the same meaning with item FPSR 4, leading to item FPSR 1 being deleted. This is also applied to items coded as FPSR 2 and FPSR 5. These two items also mirror the same meaning and following the recommendation from the modification index, item FPSR 2 was removed from the analysis as the regression coefficient was too low. For item FPSR 3, the reliability index from the beginning for this item is very low and not included in the analysis. We acknowledge that the respondent might get confused with this item as it was negatively worded. Therefore, the remaining items could best explain the service operation flexibility as shown on the re-specified model. For the latent construct of firm performance, items FPC1 and FPC2 were removed from the model. Item FPC3 'on maintaining high capacity utilisation' could be argued to be the result of 'reducing customers/clients costs' and 'attaining high employee productivity'. Therefore the respondents' perception signifies item FPC3 alone as the best indicator for variable of cost. For the variable of customer service variable, item FPC4 which asked the respondents on the response to customers/clients complaints was also taken out from the model. Other accepted 3 items were more appropriate to signify the customer service, rather than the customers/clients just merely being complaints-based. In terms of the financial aspect, items FPFi2, FPFi3 and FPFi4 which enquired the return on assets, return on investment and operating profit could be grouped to one indicator i.e. item FPFi1 which was the growth of market share. Even so, the weak 0.36 factor loading of this item to the company performance may indicate that the SOF has a weak connection with the firms' overall financial performance such as the growth of market share. Thus, the remaining items could best explain the firm performance. All in all, we could explain that the 7 factors will be able to influence most measures of customer service performance. This finding is consistent with Daugherty and Pittman (1995) who imply that flexibility in operations would result in better response to customers' needs. Other related studies (Swamidass and Newell (1987) on flexibility and growth and profitability, Fiegenbaum and Karnani(1991) on extra profit, Narashiman and Das(1999) on flexibility and cost reduction, Jack and Raturi (2002) on financial and delivery performance) are evidently not supported. Part of the reason could be down to the service sector and manufacturing sector's different nature, although Slack (2005) argues that the results could be generalized to both sectors.
It is common to lose items due to the non-fit model after taking into consideration the modification index. The original number of items for service operations' flexibility which included the internal and external flexibility was 10 items. After trimming, as a result of CFA, on the hypothesized model, 7 items were left to be of practical value that clearly defined and explained the service operations' flexibility meanwhile the 3 items were deleted. For the latent construct of firm performance that included the cost, quality and financial 11 items and finally leaving 5 items that were considered of practical importance. This means that the remaining items could clearly define the latent construct of firm performance. The re-specified model is the best model as service operations' flexibility could predict 80% of the firm performance as disclosed in this study. This model could be a good foundation to proceed with the best fit model in the future.
Finally, this research also gives an impact towards managerial implications on firm performance. The service shop organization should develop the seven prescribed flexibility initiatives to enhance their operations' flexibility. For example, the item "We have been able to offer new, unique and innovative services to our customer" reflect the importance for both industries to develop new services that can add value to the customers' money. For example, it may be possible for an autorepair shop to send the mechanic and provide a standard service at the customers' specified locations. A private hospital could offer a new health scanning service using a new equipment. To give another example, as for item no 2 'We have been able to integrate some features of services into alternative packages that are requested by customer'; several private hospitals have started a 'health tourism' package where customers are offered a package price for integrated services. An autorepair shop could offer a combined service for a special package price. All these efforts will enhance the level of customer service as proposed in the model.
In this study, the dimensions of service operations' flexibility and its relationship with service firms' performances were tested using a confirmatory approach. We operationally defined service operations flexibility as the capability of service firms to adapt to internal and external driven changes. We further subdivided the construct into six dimensions of flexibility: design, package, variety, delivery, service recovery, and system robustness. Service firms' performance consists of cost, customer service and finance. Respondents of the larger scale, that is four countries are operations managers or equivalents who are involved in the decision-making process in operations. Two service industries were selected; private hospitals and auto repairs. The two industries are considered as service shops, characterized by a high degree of customer interaction and low labor intensity (Schmenner, 1986). The structural relationship between service operations' flexibility and service firms' performances is tested using the Structural Equation Modeling using AMOS procedures. The final model fits the data with substantial changes adapted from the original hypothesized model. As we have not found any moderation effects of the type of industry, we believe that the model is suitable for both hospitals and auto repairs, and therefore, has contributed to the accomplishment of this study.
The authors gratefully acknowledged the financial support received in the form of a research grant (UKM GUP EP 0718113) from Universiti Kebangsaan Malaysia (UKM).