An Empirical Study about the Influence of Word of Mouth on Client Changing Behaviour : The Case of Computer Industry in the State of Kuwait

The Word of Mouth (WOM) was recently considered by marketing management, and adapted by marketers in many industries such as computer manufacturing industry. The study investigated how WOM is presented by sender influences consumer switching behaviour during interaction in the different categories of situations within the computer industry. Five variables were tested to see their correlations with the WOM. These variables are: Switching Cost, Sender expertise, Receiver expertise, Sender Trustworthiness, and Tie Strength. Using convenience-sampling approach, the study obtained 130 usable surveys. Data was randomly selected from four different universities in Kuwait. Data analysis was done through the software SPSS, and the results were obtained through ANOVA. The regression analysis statistical method was implemented to test the influence of the five variables on the variable: Sender influence in Positive WOM. This paper drove its significance in providing empirically evidence that would help marketers to plan strategies or to choose the appropriate way in using the direct influence of WOM in marketing transactions.


Introduction
Word of Mouth (WOM) and customer switching behaviour is an important topic that has been considered in recent years by marketing management field and thus marketers.Previous literatures and studies reveal that WOM increases the loyalty of client and thus reduces damages of the company profit (Hoest, 2004).Marketing departments in our modern organizations become aware of the significance of adapting Word of Mouth to be a primary strategy which impacts their customer-purchasing decision.Studies reported that the sender (the person who recommends products or services to someone else) credibility such as expertise, trustworthiness and opinion leader was considered a significant source that affects receiver (the person who is affected by sender) buying behaviour and thus switching behaviour (Anderson, 1998).
Voluminous studies indicated that WOM did get the greatest consideration of the literature in the marketing markets (Herr et al., 1991;Hogan et al., 2004;Bansal, Taylor, & James, 2005;Dimitriades, 2006).Studies suggested that WOM communication has deemed as valuable mechanism for encouraging the promotion of a company's products and services (Dimitriades, 2006).
Moreover, WOM communication for non-profit organization is regarded as the primary source of the firm-initiated promotional efforts thought opinion leader (Herr et al., 1991).Studies explained that the use of word of mouth is nine times operative when compared to other advertising technique; this is because WOM has more influence on consumer attitude that converts unfavourable or neutral predisposition into positive attitudes (Bansal, Taylor, & James, 2005).In his study, Hogan et al. (2004) suggested that WOM could triple the effectiveness of advertising.WOM is one of the imperative communication channels that modern marketing should pay great consideration to its effect on customer buying and switching behaviour (Anderson, 1998;Gilly et al., 1998).

Literature Review
The introduction has presented an overview of the research part of this work, the problem statement and research overall purpose objectives.This chapter presents studies and reviews literature related to the topic.This chapter primary considers recent finding and researches which have tested the effect of Word of mouth on switching behaviour.

Conceptual Background on Word of Mouth
Prior studies have viewed WOM as a central intermediate for affecting customer behaviour (Arndt, 1967).The word of mouth affects strongly customer decision making and it relies on many factors.Generally, a consumer has propensity to be depicted to marketer dominated bases, while the personal WOM is found to be the most effective (Sandy et al., 2011).
Moreover, word of mouth has been expressed as positive association about a service companies and this communication is viewed as a central relational outcome.Other studies have defined word of mouth as a pattern of social relation between customers (Datta et al., 2003).Kirby and Marsden, (2006) suggested that the applied practice of WOM is implemented in the marketing: "the promotion of a firm services and products via an initiative is conceived to get customers express positively about that company".Without a doubt, the consequences of word of mouth concept are utilized to enhance marketers influence, as indirect form of the perception of "connected marketing" (Kirby and Marsden, 2006).In their study, scientists suggested that the significance of word of mouth for facilities sector is to view it as a main power in the market (Mangold et al. 1999).

The Effect of WOM
Previous studies have suggested that WOM is particularly significant for marketing service.WOM is also effective to influence a consumer to try a particular service provider (Rick Ferguson, 2008).The old literature in marketing research provided by Mangold (1987) has suggested that the effect of word of mouth in the services setting has more influence on acquiring judgement than other bases of inspiration.In their study, Kirby and Marsden, (2006) have suggested that positive communication between customers had a superior power than any media in terms of brand changing for small customers and domestic virtuous.Similarly, prior studies have reported that word of mouth is more significant than promoting in enhancing customer consciousness (Sheth, 1971).
Furthermore, studies reported positive WOM is the information source which has powerful impact on customer decision making.While consumers receive numerous marketing advertisements that are designed to get their attention and buying behaviour, WOM is regarded to be a highly trusted information source (Walsh et al., 2005).For example, WOM provides recommendations with regard to a product and service provider which give positive feedback about particular products or service features and encourage families, relatives and friends to buy from a specific supplier.WOM helps marketers in inviting new consumers who are imperative for a company that has economic success for a long term (Kim, et al., 2006).WOM also helps to decrease negative feedback and disagreement of prevailing consumers (Wangenheim, 2005).
According to Dye (2000), word of mouth was nine times as significant as promoting at transmitting unfavourable or real tendencies into positive defiance.It was predicted as 80% of all purchasing decision were quiet inspired by WOM announcement (Jeffrey Steven, 2008;Mazzarol et al., 2007).WOM has either a positive or negative effect on buying decisions in comparison with merchandises customers of facilities who have a tendency to recognise a higher level of risk during the pre-purchase choice phase (Morone & Taylor, 2004).Studies have explained that the essential scheme is that customer behaviour engages risk in a way that slightly decision taken by a customer leads to significances which might not done ahead with any conviction, and few of which are expected to be objectionable (Brown et al., 2005).
Studies have concluded that to decrease the uncertainty and risk related with consumer purchasing decision, they ask for evidence about the service and/or product (Brown et al., 2005).Word of mouth was initiated to be one of the most effectual info source that would reduce perceived risk in product or service (Brown et al., 2005).Prior studies claims that the higher the awareness related to risk accompanying with product or services, the greater dependence on the informal communication source (Reingen, 1987).
WOM is vital for service providers who offer-many services, and their experience is credibility established.In these services consumers depend greatly on recommendation and proposals from others who have practiced the provision.Previous studies have recognized the limited studies that investigate the customers' previous experiences of word of mouth and mostly the influence of consumer-marketer associations and service quality awareness on word of mouth purposes (Gremler et al., 2001).
Thus, our study attempts to enhance prior studies and literature by investigating factors that affect WOM behaviours of customer.Studies have reported that WOM is an outcome of consumer-firm affiliations (Reichheld, 2003;Brown et al., 2005).Marketers have to be alert about the importance of this association and its benefits to enhance customer WOM tendency.
The significant role of WOM has been acknowledged by dispersal of innovation scientists and has been recognized as one of the greatest vital communication source among customers themselves (Derbaix & Vanhamme, 2003).Word of mouth is found to be a practise of individual stimulus that social communications among senders and a receivers affect the receivers' behaviours and thus buying decisions (Merton, 1968).The role that social communication greatly effects thoughts and is recognized by sociologists, who clarified the value of "opinion leaders" in this practice (Katz & Lazarsfeld, 1955).The capability of one customer to affect other views is of precise awareness to a marketer who is seeking to market their services and products.
In our contemporary organization, WOM plays a major part for service firms, as imperceptibility makes the pre-acquisition provisional of services difficult (Sandy et al., 2011;Pedro Longart, 2010).Studies suggested that word of mouth is significant when services are compound or have high apparent risk (Walsh, 2005a).Moreover, studies suggested that WOM deemed as greatly reliable information source as the sender is generally free of the service provider, in this case (Silverman, 2001).Therefore, WOM has an important role for amenities that possess great qualities, such as specialized and commercial services (Walsh, 2005a).

Factors Enhance WOM Effects on Customers
Previous studies suggest that there are many factors which enhance WOM effects on receiver.In the basic form, two parties construct WOM: Sender and Receiver.Studies suggested that the personal relationship participation plays an important part in selecting the service providers (Walsh et al., 2005b).Source credibility is regarded as expertise and trust (Gianfranco & Wayne, 2011;Pedro Longart, 2010).Moreover, the effects of a message on a customer behaviour related to how that receiver perceives the communication situation.In his study, Ohanian (1990) suggests that source credibility is an important condition to imply affirmative features communication which affects the receiver's reception of a communication.Expertise defined as "the extent to which a communicator is perceived to be a source of valid affirmations", while trustworthiness is defined as "the degree of confidence in the communicator's intent to communicate the assertions" (Ohanian, 1990, p. 41).This paper focuses on both dimensions; source credibility of sender: expertise and trustworthiness of sender.The following sections present studies and literatures about the study variables.

Sender's Expertise
Marketing studies have suggested that expertise, which has been regarded as source credibility, is an important dimension that affects customer-buying decision, (Dholakia & Sternthal, 1977).Studies demarcated expertise as the unit to which the information's source observed is precise and true (Bristor, 1990).Moreover, expertise is predicted to encourage influence since receivers consume slight interest to check the certainty of the source's declarations by recovering and reviewing their own opinions (Sternthal & Craig, 1982).Studies and literatures have found that a sender of a WOM possesses a high degree of expertise, from the receiver's perspective.In their study, researchers have suggested that sender's expertise is a powerful indicator of the influence of word of mouth on the receiver's buying decision (Bansal & Voyer, 2000).

Sender's Trustworthiness
In his study, Bristor (1990) explains that trust is the extent along which the foundation of information is perceived reflecting the actual feelings or opinion of sender.Moreover, studies suggested that it is very important for a receiver to have an interest in the listener and his experience with a service or product.The higher degree of trust would lead to higher influence in WOM exchanging of information (Dichter, 1966).Trust is regarded as the extent of confidence that the listener has had and the level of recognition of the presenter (Rick Ferguson, 2008;Ohanian, 1990).In his study, we examined the influence of source trust on the persuasion of the exchange information, the conversationalist was recognized to be greatly trusted, and the message was further influenced than dogmatic communication in making behaviour alteration (Rick Ferguson, 2008;Ohanian, 1990).Moreover, researches have explained that trust would affect the communicator's persuasiveness (McGinnies & Ward, 1980).This concluded to recognize a source which was supposed to be both an expert and trustworthy yields the best view alteration (Ohanian, 1990).

Receiver's Expertise
Studies have explained that the characteristics that affect the WOM receiver would also affect his expertise (Gilly et al., 1998).Pervious literatures report a negative association between the extent of familiarity of an information pursuer and the extent of external search for information (Sandy et al., 2011).Moreover, studies clarify that skilled persons involve in inadequate pursuits for private source information (Sandy et al., 2011).Studies suggest that those who hold a high level of expertise would obviously give tiny attention to an info search earlier to purchase (Sandy et al., 2011).Furthermore, those who have less service knowledge and experience distrust their aptitude to make suitable and reasonable service or adoptions.This implies that they are expected to pursue the estimations of others for merchandise recommendation.

Interpersonal Influences
In his study, Bristor (1990) uttered that a word of mouth network is a social network that consists of a group of people who are involved in WOM, in addition to the relationships among them.The social network (either electronic or traditional network) is the association between the conversationalists of the info and the receivers.For example in Facebook, people communicate and provide their perception and attitude toward various product and services.In his study, suggested that conversationalists and receivers of WOM are energetic self-governing respondents through communications process, while receivers frequently recruit product conversations by requesting communicators for info (Arndt, 1967).The bond between two persons would influence the WOM process is tie forte, or the force - (Sandy et al., 2011;Bristor, 1990).This tie strength plays a significant part in the flow of info through the networks.

Switching Costs
Researches define Switching Costs as the one-time costs that consumers subordinate with the operation of changing from one supplier to another (White & Schneider, 2000).Switching costs is depended on the extent to which the customer has achieved satisfaction previously (Keaveney & Parthasarathy, 2001).Numerous studies have exposed that dissatisfaction is a foremost aspect affecting switching behaviour (Keaveney & Parthasarathy, 2001).The switching behaviour can damage companies continuously in the market like insurance, telecommunication and computer industry (Keaveney & Parthasarathy, 2001).The reasons of switching behaviour have been articulated in many studies; mostly dissatisfaction is the main factor.Switching costs are the driver of re-buying of service or product from the same provider.Studies have found that the advanced switching costs are recognized to be, the more prospective the customer will repurchase.Consumers who deemed to be low switching costs describe as "Switchers".On the other hand, customers who are high switching costs describe as "Stayers" who miss the practical awareness, they are reluctant to capitalize the money, energy and time needed to alert their service or product provider.In his study, Burnham et al. (2000) stated that there are three types of switching cost.These switching costs are described in Table 1.Furthermore, Burnham et al. (2000) have reinforced that the better practical, monetary and interpersonal switching costs will be related with higher customer aims to stay with a compulsory supplier.

Research Questions
The current study develops many questions that would help us to explore various issues surrounding the background and the problem statement of this study.The study aims to reveals if there are any differences between the finding of previous studies and the effect of WOM in the context of computer industry in Kuwait market.To achieve this purpose, the study has laid out the following research questions: -To what extent WOM (provided by families, friends and exporters) has influence on customer switching behaviour?
-To what extent WOM (provided by families, friends and exporters) has influence on customer buying behaviour in computer manufacturing industry?
-Will the WOM have more impact on receiver's switching behaviour in the computer manufacturing industry if the tie strength and similarity between the receiver and sender are high?
-Will the WOM have more influence if a laptop or computer is perceived as high switching costs?
-Does Age of groups have significant differences in regard to the changing of laptop or computer?

The Framework of the Study
The current study attempts to consider various issues surrounding WOM and categorises the independent variables under four categories: 1-sender characteristics, 2-receiver characteristics, 3-interpersonal relationship, and 4-switching costs.The model in Figure 1 explains the structure of the research work.The variables that are independent in this model are sender expertise, sender trustworthiness, receiver expertise, interpersonal relationship and switching costs.The aim of this model is to test the impact of positive WOM, expressed by the five independent factors, on switching behaviour of receivers.H3: the more the receiver's expertise is high the lower influence of positive WOM on receiver's switching behaviour in computer industry.
H4: The stronger tie strength between the receiver and the sender would influence of positive WOM on receiver's switching behaviour in computer industry.
H5: The higher the switching cost in computer industry, the lower the influence of WOM on receiver to switching the product.

Research Methodology
The methodology presents the research design and purpose, the questions and hypotheses of the study, the data collection procedures, the study population and sampling, the scale validity and reliability and the statistical tools used.The current study adapted quantitative research design; a survey method adapted to collected data.The data obtained are analysed by using SPSS v.

Aim of the Study
Previous studies have suggested three main purposes for any research, these purposes are exploratory, descriptive, or explanatory (Zikmund, 2000).The aim of this research is explanatory, but descriptive measures were applied to support the main research aim.Data collocation tools were developed through prior literature and researches.A pilot study was applied to assure the reliability and content validity of the scale implemented in the questionnaire.A random sample was selected from students who were enrolled in B.A. and MBA programs in public and private universities in Kuwait.The participants were asked to reveal their perceptions towards various factors that affected their buying behaviours in the computer industry.This study uses both explanatory and descriptive purposes.This paper uses a descriptive purpose to describe the study of the five independent variables in the proposed model, and to describe participants' information and demographic factors.The explanatory aim is implemented to clarify the association and correlation between these five independent variables.

Data Collection Methods
To accomplish the above mentioned research objectives, the data for this research work was selected randomly through self-administered questionnaires.Since the reliability and the content validity were significant, the survey of the current research has been developed from previous studies (e.g.Rick Ferguson, 2008).The survey of this study was a self-administered questionnaire that consists of three pages and with two sections.The first section included the variables switching cost, sender's expertise, receiver's expertise, sender's trustworthiness, and the tie strength as well as the influence of sender's Word of Mouth on switching behaviour.Using 5-point Likert-type scale, respondents rated their responses from (1) 'Strongly Disagree', to (5) 'Strongly Agree'.Respondents also indicated whether they were affected by positive WOM in his/her buying behaviour.
The second section of the survey included the characteristics of respondents.Respondents were requested to supply characteristics information such as current position, year of experiences, highest educational qualification, age, nationality and gender.The following table presents the authors from whom we have adapted questions/items and the number of questions/items for each variable.Finally, respondents were requested to reveal any other remarks in the space provided to express their opinion.Table 2 summaries the variables and the number of questions/items of each variable.

Study Population and Sampling
The study has primary considered B.A. and MBA students who study in Kuwait University, Arab Open University, GUST and AUK in Kuwait.In a random manner, respondents were approached by visiting these universities and through using E-mail, wherein the soft copies of the survey were sent to the participants.Moreover, the study sent the electronic copies of the survey by using social networking sites such as Facebook, where students from different universities in Kuwait were easily approached.A total of 160 surveys were distributed among students.The distribution was done through two weeks' period in the early part of April 2011.University students were approached, and the study has explained the nature of this research work and requested respondents if they would be ready to fill in the survey.

Scale Validity and Reliability
Due to availability concern of the preceding study in this area, the researcher has to test the reliability and content validity of the processes.To make sure the used questionnaire is reliable and valid, the study conducted pilot study, by sending the questioners to fifteen respondents randomly, and then evaluated their replies.Also, we will measure the internal consistency method to estimate reliability.To ensure high reliability and better result the study implements Cronbach's alpha test, which is a famous measure implemented to estimate the reliability of constructs.Consequently, Cronbach's Alpha value was computed and the results were matched with the 0.7 level, the cut-off point, which showed that the reliability of the scales were significant (Nunnally & Bernstein, 1994).However as previously detailed, a reliability between 0.5 and 0.6 is considered suffice by Nunnally (1967).

Data Analysis and Tools
The data analysis method has started by coding the collected data.The responses obtained were analysed through the software SPSS V.18.The use of SPSS computer program will assist in the coding and analysis process of the data collected from the survey questionnaires.The study will use the following statistical tools: -Descriptive analysis (the means, frequencies, and standard deviation) and ANOVA test were applied.
Descriptive statistical tools were engaged to designate the personal characteristics of the respondents, while ANOVA test was done to check the differences among demographic variables regarding to the study variables (e.g. the effect of WOM of sender).
-Cronbach's alpha test was conducted to check the reliability and the internal consistency of the survey factors.
Methodologist suggested that a significant value of Alpha would be 0.70; however, an Alpha value of 0.60 is also acceptable as a minimum cut-off point (Nunnaly and Bernstein, 1994;Zikmud, 2000).
-Personal Correlation was implemented to assess the relationship between the variables, (e.g. the relationship between sender expertise and receiver switching behaviour).It's much recommended to test the relationship between the variables before conducting the regression analysis (Zikmund, 2000).
-Linear Regression Analysis was conducted in this paper to find the estimate strength between dependent and independent variables (Zikmund, 2000).

Result and Discussion
This paragraph presents the results of the questionnaires.It begins with describing the personal information of respondents and presents the results of ANOVA test.The second section presents the results of normality assessment, and the third section presents reliability and validity results.After that, the results of Correlation Analysis of the study variables will be presented.Finally, the chapter describes the outputs of the Linear Regression Analysis.The findings are discussed in accordance to the research objectives and hypothesis.

Response Rate
Using convenience-sampling approach, the study selected four universities to distribute the surveys.These universities are; Kuwait University, Arab Open University, GUST and AUK in Kuwait.The study acknowledges the importance of neutrality and accuracy in filling the surveys.Due to high connection and relationship, the study ensures that respondents are aware about the topic of WOM.The target respondents are B.A. and MBA students who study in these universities.
The study distributed 160 questionnaires (40 surveys for each university).From 160 questionnaires distributed, the study obtained 130 surveys.The study found 10 questionnaires were not usable due to the incompletion of the most questionnaire items; these questionnaires were not included in the analysis information and then were eliminated.As such, the study managed to obtain 120 usable surveys and the respondent rate was (75%).Table 3 shows the outputs of response rate for the study.The study took 3 weeks to collect back all the questionnaires.After collecting the questionnaires, responses were coded and gathered in an Excel file, and then the data was transposed to SPSS worksheet.The questionnaire was divided into two parts.The first part was about the demography of respondents and the second part was about the research questions.The following section displays the demographic factors and respondents' profiles.

Respondent's Personal Information
Table 4 describes clearly personal information about respondents.As a result, Table 4 supplies the following section and investigates the effect of demographic factor such as age and its effect on the positive WOM of sender.

ANOVA Test
Analysis of Variance (ANOVA) is a common statistical technique that compares the means of different variables.
The reason for doing the ANOVA Test is to see if there is any difference between the mean age of respondents and the effect of WOM.Table 5 presents the findings of ANOVA Test.The result of ANOVA reveals that the younger age group has stronger effect of WOM of senders.The only exception is with receiver's expertise, because the more receivers have expertise, the less he might be affected by WOM of sender.The old aged people are affected slightly by the WOM compared to younger generation.

Assessing Normality
The second statistical test performed in this study is testing skewness and kurtosis of the variables, which was implemented to assess the data normality.According to Hair et al. (2006) assessing data normality is paramount important since data that are not normally distributed will affect the overall outcome when we conduct hypotheses testing.Therefore, the study includes each item to measure skewness and kurtosis.
Skewness is a measure that study the symmetry or the lack of symmetry of data distribution.Kurtosis is used to measure whether data are peaked or flat with respect to a normal distribution.According to Hair et al. (2006), normality is indicated when skewness value is less than 2 (<2) and kurtosis value is less than 3 (<3).Table 6.4 shows the skewness and kurtosis values of the normality assessment.As shown in Table 6, a skewness and kurtosis of each item is smaller than 2, therefore, all the items are considered normal.Therefore, we can proceed with the other tests and it is rational to employ the estimated mean as centre of location.

Reliability and Validity Results
Reliability test is known as the scale internal consistency that measures the degree to which the factors are homogeneous.After the normality assessment, studies stressed that a study should measure the reliability of a construct before performing statistical analysis such as correlation and regression tests (Nunnally, 1994).
Similarly, Cronbach's and Meehl, (1955) recommended to conduct various tests in a study to guarantee that the developed scales/factors of the questionnaire are reliable in what were proposed to measure before measuring the correlation between variables and regression tests.Hence, to measure the internal consistency of the proposed variables, the test of Cronbach's alpha was employed.Studies suggested that to improve the value of Alpha's coefficients, it is recommended to delete some items when the scored value of a variable is less than 0.7, on the other hand, if the value did not attain the minimum score of the cut-of-point then it is preferable to be omitted (Nunnally, 1994;Hair et al., 2006).Table 7 shows that the lowest coefficients α value in this test is 0.983, which exceed the suggested cut-off point (0.7).Thus we may conclude that the items/variables that explain the constructs of the model were highly reliable and significant.

Descriptive Statistical Analysis
Table 8 shows the perceptions of respondents to variables.As shown, Respondents revealed that sender have great influence WOM on receiver regarding the technology industry in Kuwait (Mean=3.71,SD= 1.21).
Respondents revealed that the second and the third most important factors that lead to affective WOM are Sender expertise (Mean=3.61,SD=0.903) and the tie strength between receiver and sender (Mean= 3.36, SD=1.23) representatively.In generational respondents revealed good perception toward sender trustworthiness (Mean= 2.88, S.D= 0.77).Switching costs found to be relatively weak in its effect on the WOM (Mean=2.62,S.D=1.11).Among the variables, the least important factor is receiver expertise (Mean=249, SD=1.37).The following table presents the average results of descriptive analysis.

Correlation Test
They current study performed correlation test to measure the strength of the linear relationship between the study variables.Previous studies clarified the importance of testing the relationship between variables (dependents and independents variables) before conducting Regression Analysis (Coakes and Steed, 2007).As presented in Table 9, the results shows that the correlation coefficient ranks among the variables are strong enough indicated a significant correlation.Cohen (1988) proposed that the value of Pearson's correlation coefficient r is separated into three groups: Related to Cohen (1988), the results reveals a large correlation between Sender expertise, Sender Trustworthiness, Tie Strength, Receiver expertise (r > 0.5) with WORD OF MOUTH of sender.Similarly, the results indicates a large negative correlation between switching cost and receiver's expertise with WORD OF MOUTH of sender (r < 0.5).From table 4.7 we can conclude the following:  There is a positive and strong association between sender expertise and influence of WOM of sender (r = 0.8, p=0.00 < 0.01).this mean that if sender have good expertise in computer industry and the receiver aware about this issue, he would greatly affected by his advices to buy or avoided buying a specific computer or technologies.


There is a positive and strong association between Sender trustworthiness and influence of WOM of sender (r = 0.773, p=0.00 < 0.01).This means that if receiver trust sender he would greatly affected by his advices to buy or avoided buying a specific computer or technologies.


There is a positive and strong association between Tie strength and influence of WOM of sender (r = 0.753, p=0.00 < 0.01).This means that if receiver and sender have a tie relationship, receiver would greatly affected by his advices to buy or avoided buying a specific computer or technologies.


There is a negative and strong association between Switching Cost and influence of WOM of sender (r = -0.874,p=0.00 < 0.01).This mean that an increase in switching cost would reduce the effect of the WOM of sender.


There is a negative and strong association between Receiver's expertise and influence of WOM of sender (r = -0.857,p=0.00 < 0.01).This imply that in increase in receiver expertise would reduce the effect of the WOM of sender.

Regression
To further investigate and verify our theoretical framework, the study employed ordinary least square (OLS) test.OLS technique as a linear regression model has implemented in this paper to inspect the extent to which the independent variables influence the dependent variable.The regression analysis uses to examine the influence of independents variables (Sender expertise, Sender Trustworthiness, Tie Strength, Receiver expertise and Switching Cost) on the dependent variable (Sender influence in Positive WOM).Tables 10, 11 and 12 summarized the results of the OLS analysis.
The outputs of the Regression study reveals that the model is significant (F= 14.397; p < 0.00) and the correlation between dependent and independent variables is strong (R = 0.928).The coefficient of determination (R 2 ) for the regression is (0.861) indicating that (86.1%) of the variation in the dependent variable (Perceived information Security) was explained by a portion of the variation in the independent variables included in the regression model.This mean that 13.9% (100%-86.1%) of variance were not included in the proposed model.2, the results of the regression line indicated that the variance in the effect of WOM of sender was explained by the five variables.
Figure 2. Theoretical Framework related to the regression analysis Finally, the results of hypotheses tests are described in Table 13.

Findings
This study proved that leaders of companies should make more studies on WOM.Word of Mouth (WOM) and customer switching behaviour is an important topic that has been considered in recent years by marketing management field and thus marketers.WOM is viewed as beyond the control of marketers, thus it's vital for a marketer to comprehend and how WOM affects their customers buying decision as well as switching behaviour.
The main goal of this paper was to explore the impacts of WOM on consumer buying behaviour in the computer manufacturing industry.The result of ANOVA reveals that the younger age group are the stronger effect of WOM of sender.The old aged people are affected slightly in WOM compared to younger generation.The results shows that kurtosis of each item is smaller than 2, therefore, all the items are considered normal.The weakest coefficients α scored is 0.983 in which it exceeds the suggested cut-off point (0. in this research work were regarded as highly reliable and satisfied.Moreover, respondents revealed that sender has great influence WOM on receiver regarding the technology industry in Kuwait.Respondents revealed that the second and the third most important factors that lead to affective WOM are sender expertise and the tie strength between sender and receiver representatively.In general respondents revealed good perception toward sender trustworthiness.Switching costs was found to be relatively weak in its effect on the WOM.Among the variables, the least important factor is receiver expertise.

Conclusion
The results of this study are valuable mainly for two reasons.First, the results of the effect of WOM of sender in the technology industry in Kuwait can be compared to similar research studies conducted abroad.Secondly, this study provides valuable information to IT corporate that provides technological devices such as Laptop, IPod, IPad etc. and products to Kuwait.The main contribution to method of this research work is the investigation of various effects of sender and receiver on WOM in the State of Kuwait.The study has shed light on the importance of understanding of these factors in order to boost retail business and sales technological products, and enhance our perception toward the effect of WOM.The study has proposed a model that would enhance understanding about the effect of WOM of sender.The contribution of this study would be very important to technological industry that focuses on Kuwait as market segment.Finally, the contribution to knowledge of this research work is made to show the effect of WOM theme as it provides empirically evidence about factors that influence receiver in Kuwait.In some instances, the results from this study and other previous studies support each other, as indicated in the discussion chapter.

Recommendation
For future studies, the study recommends that other factors that affect WOM to be considered.For example, future studies may investigate the effects of receiver loyalty, product risk, sender opinion leader etc.moreover the study recommends future studies to consider demographic factor in the analysis such as receiver gender, income and educational level.Moreover, the study suggests that future studies consider larger target of respondents that include other products and services.Finally, due to similarity in culture, language, religion and other personal factors, we recommend future studies to be conducted in other Gulf Region countries such as in KSA and Bahrain.

Table 1 .
Typology of consumer perceptions of switching cost Procedural Switching Costs Financial Switching Cost Relational Switching Costs Source:Burnham et al., 2003.

Table 2 .
Summary of the variables and number of question/items

Table 3 .
Research response rates

Table 4 .
Characteristics of personality's respondent

Table 6 .
The results of normality testing

Table 7 .
Reliability results

Table 9
Correlation is important at the 0.01 level for a two-tailed test.

Table 12 .
Regression 1, coefficientsLinear regression is a useful test to model the value of a dependent scale variable depended on its linear association to one or more predictors.This table shows the coefficients of the regression line.It states that the Sender influence in Positive WOM is equal to: The results of regression indicated that sender influence in Positive WOM is explained by the five variables; Sender expertise, Sender Trustworthiness, Tie Strength, Receiver expertise and Switching Cost.The effect of the five antecedents on Sender influence in Positive WOM was assessed by linear regression analysis by estimating the following equation: Receiver expertise and Switching Cost = β1 Sender expertise + β2 Sender Trustworthiness + β3 Tie Strength + β4 Receiver expertise + β5 Switching Cost + ε

Table 13 .
Summary of hypotheses testing The results of the correlation test shows that the five variables are correlated Positive WOM, these variables are; Tie Strength, Receiver expertise, Sender Trustworthiness, Sender expertise, and Switching Cost.Related toCohen (1988), the results reveal a large positive correlation between Tie Strength, Sender Trustworthiness, Sender expertise, Receiver expertise with WORD OF MOUTH of sender.Similarly, the results indicate a large negative correlation between switching cost and receiver's expertise with WORD OF MOUTH of sender.The ordinary least square (OLS) test technique as a linear regression model has been implemented in this research to test the impact of independent variables on the dependant variable.The results of regression indicated that sender influence in Positive WOM is explained by the five variables; Sender expertise, Sender Trustworthiness, Tie Strength, Receiver expertise and Switching Cost.