1. Introduction
In this study, the market is divided into segments according to factors influencing the foreign consumers in selecting Chinese restaurants, and the consumers are divided into different groups by the way of market segments, they are “physiological needs”, “common consumption” and “careful consideration”. In the meanwhile, the restaurant selection factors and service quality assessment are analyzed and simplified into several factor combinations, which aims to discuss whether there is differential correlation between factors of them, whether the service quality of Chinese restaurants affects the foreign 留學生dissertation網consumer’s satisfaction, and whether their satisfaction affects their behavioral intention.
In order to study the significance of Chinese restaurant’s service quality to foreign consumers, their satisfaction as well as the influence upon their behavioral intention, in this study, sample restaurants are chosen as non-fast food and middle-high price-level restaurants with a certain level of service. Such restaurants provide seats and table service, as well as an opportunity to observe the interaction between consumers and staffs.
1.1 Philosophy
This study is going to use positivism as its philosophy, because in this study that “the researcher is independent of and neither affects nor is affected by the subject of the research” (Remenyi et al.,1998). It is frequently advocated that the positivist researcher will be likely to use a highly structured methodology in order to facilitate replication (Gill and Johnson, 2002). However, this may not necessarily since it is perfectly possible to adopt some of the characteristic of positivism in the research. As in this study, researcher developed hypotheses which will be tested and confirmed.
1.2 Approach and Data Type
As it is said in the philosophy that this study developed hypotheses, which makes the approach of this study is deduction. It is the dominant research approach in the natural sciences, where laws present http://www.ttc-switch.com/Thesis_Tips/Methodology/the basis of explanation, allow the anticipation of phenomena, predict their occurrence and therefore permit them to be controlled (Collis and Hussey, 2003). In the deduction possesses, the search to explain causal relationships between variables, deduction possesses also dictates that the researcher should be independent of what is being observed.
An important characteristic of deduction is that in deduction possesses concepts need to be operationalised in a way that enables facts to be measured quantitatively. This means that the data type of this study is quantitative. Many business statistics textbooks classify quantitative data into data types using a hierarchy of measurement, often in ascending order of numerical precision.(Diamantopoulos and Schlegelmilch, 1997; Morris, 2003). These different levels of numerical measurement dictate the range of techniques available to you for the presentation, summary and analysis of your data.#p#分頁標題#e#
Quantitative data can be divided into categorical and quantifiable. Categorical data is used in this study which refers to data whose values unlike quantifiable data cannot be measured numerically but can be either classified into categories according to the characteristic that identify or describe the variable or placed in rank order. They can be further subdivided into descriptive and ranked. Both descriptive and ranked data are widely used in the questionnaire design procedure in this study.
2. Hypotheses
Mainly targeting at foreign consumers, the study discusses the correlation between restaurant selection, consumers’ attitude toward service quality and their behavioral intention, with its hypotheses as follows:
Hypotheses 1: consumers in different market segments have significantly different perceptions in selecting restaurants.
Hypotheses 2: consumers in different market segments have significantly different perceptions in their preference toward service quality.
Hypotheses 3: consumers in different market segments have significantly different perceptions in satisfaction and behavioral intention.
3 Questionnaire Design
3.1 Restaurant Selection Factor Items
In the first part of the questionnaire, in order to understand the factors influencing foreign consumer’s selection of Chinese restaurant, this item divides the restaurant selection factors into 9 categories with reference to Yuksel and Yuksel (2002), namely Service quality and staff attitude, Product quality and hygiene, Adventurous menu, Price and value, Atmosphere and activity, Healthy food, Location and appearance, Smoke and Visibility; and divides the consumer’s reselection of restaurants into 4 categories with reference to Soriano (2002), namely Quality of food, Quality of service, Cost/value of the meal and Place. By taking all the above categorization into consideration, this study concludes 4 structural items, namely Service quality, Food Quality, Price and value , Place and location, and designs 22 question items for restaurant selection factors by induction and with reference to Meyer ad Conklin (1998), Soriano and Yuksel.
Among all the factors could influence the foreign customers’ decision making on restaurant, it is believed that service quality would be the most important one. The reason for that is today’s customers are becoming increasingly concerned about the quality of service, plus it has direct effect on restaurant’s profits, it is positively related with customer retention and with customer loyalty(Baker and Crompton, 2000; Zeithaml and Bitner,2000). In order to understand the consumer’s attitude toward the service quality of the restaurant where they have dinner, this part of the questionnaire takes the service quality measuring indicators of SERVQUAL proposed by Parasuraman, etc.(1988) as its main structure, and divides the service quality items into 5 categories, namely Tangibles, Reliability, Responsiveness, Assurance and Empathy, with reference to 29 items and relative studies in the DINESERV measuring indicators proposed by Stevens, etc.(1995) to assess the restaurant service quality. It also refers to Stevens, Alexandris (2002), Yuksel and Yuksel (2002), Park (2004), Tsaur and Lin (2004) to design 18 question items in total. #p#分頁標題#e#
The questionnaire design makes use of Likert scale to assess the selection factors received by the respondent consumers, and divides the answering options into “extremely important”, “important”, “somewhat important”, “commonly important”, “not quite important”, “unimportant” and “extremely unimportant”, with each given a mark of 7, 6, 5, 4, 3, 2, 1 in succession, thereby assessing the consumer’s attitude toward the restaurant selection factors..
In this part, 36 items are used to investigate the foreign consumers’ perceptions of the selection factors of the Chinese restaurant. These items are:
Q1 has a menu that is easily readable
Q2 has comfortable seats in the dining room
Q3 has personnel who are clean, neat and appropriately dressed
Q4 has dining area that is thoroughly clean
Q5 has a visually attractive dining area
Q6 has clean bathroom
Q7 provides a private environment to chat
Q8 ability to perform the promised service dependably and accurately
Q9 Waiting-time before being seated
Q10 Waiting-time before food arriving
Q11 Waiting-time before paying the bill
Q12 serves you in the time promised
Q13 provides prompt and quick service
Q14 has personnel who can answer your questions completely
Q15willing to give you information about menu item, their ingredients and methods of preparation
Q16 has personnel who seem well-trained, competent, and experienced
Q17 sensitive to your individual needs and wants
Q18 has personnel who have passion to provide good service
Q19 make you feel like home
Q20 give extra effort to handle your special requests
Q21 Availability of non-smoking area
Q22 Food variety
Q23 Prices shown clearly
Q24 Innovative food
Q25 Hygienic food preparation
Q26 Fresh ingredients
Q27 Good taste of food
Q28 outstanding characteristic
Q29 Availability of healthy and Nutritious food
Q30 Visibility of food preparation area
Q31 Reasonable food prices
Q32 Promotion activities
Q33 Word of mouth
Q34 Impression from the road
Q35 Convenient location
Q36 Convenient parking

Q1-Q21 is about Service quality:
(Q1-Q6 is about Tangibles, Q7-Q8 is about Reliability, Q9-Q13 is about Responsiveness, Q14-Q16 is about Assurance, and Q17-Q21 is about empathy)
Q22-Q30 is about Food Quality,
Q31-Q33 is about Price and value,
Q34-Q36 is about Place and location

3.2 Consumer Behavioral Intention Items
In order to assess the consumer behavioral intention, this study refers to the consumer behavioral intention measuring indicators developed by Zeithaml, etc.(1996), of which the structural items are Loyalty, Switch, Pay more, External response and Internal response, and also refers to Boulding, etc. (1993) who proposed to measure the consumer behavioral intention in terms of repeated purchase intentions and recommendation intentions. Mittal and Lassar(1996) made use of many indicators, such as the work or service quality, overall satisfaction, recommendation intention and changing intention, to assess if the consumers are willing to continue to receive the company’s follow-up services. The behavioral intention in this study mainly refers to specific behaviors conducted by consumers after their purchase and use of products or services; therefore, the consumer’s behavioral intention is measured in terms of “repeated purchase intentions” and “recommendation intentions” proposed by Boulding, etc. The questionnaire design makes use of Likert scale to assess the service quality received by the respondent consumers, and divides the answering options into “strongly disagree”, “disagree”, “slightly disagree”, “neutral”, “slightly agree”, “agree” and “strongly agree”, with each given a mark of 1,2,3,4,5,6,7 in succession, thereby assessing the consumer’s attitude toward the consumer behavioral intention.#p#分頁標題#e#
In this part, 2 questions are used to investigate the foreign consumers’ perceptions of the consumer behavioral intention of the Chinese restaurant. Please indicate your agreement level to the following question. These questions are:
Will your recommend a restaurant to friends?
Will you go to the restaurant again?

3.3 Consumer Satisfaction
This part of questionnaire is used to measure the satisfaction of the consumers. It enquiries the overall dining satisfaction of the consumers with reference to the measuring mode of “satisfaction” applied by Cronin、Brady and Tomas (2000) in their empirical studies. The questionnaire design makes use of Likert scale to assess the service quality received by the respondent consumers, and divides the answering options into “extremely satisfying”, “satisfying”, “somewhat satisfying”, “commonly satisfying”, “not quite satisfying”, “unsatisfying” and “extremely unsatisfying”, with each given a mark of 7, 6, 5, 4, 3, 2, 1 in succession, thereby assessing the consumer’s overall dining satisfaction.
extremely unsatisfying unsatisfying not quite satisfying commonly satisfying somewhat satisfying satisfying extremely satisfying
Q7 Your overall dining satisfaction of Chinese restaurant
3.4. Demographics
The fourth part of the questionnaire has the aim to gain the distribution status of the respondent consumers’ social and economical backgrounds, six demographic variables are included. This part of questionnaire applies the form of Nominal Scale value in terms of gender, education background, marital status and profession; and it applies the form of Ordinal Scale value in terms of age and individual monthly income of the respondent consumers.
In this part, 6 questions are used to gain the distribution status of the respondent consumers’ social and economical backgrounds. These questions are:
Q1.Your age
A. Less than 20
B. 20-29
C. 30-39
D. 40-49
E. 50-59
F. 60 an over
Q2.Your gender
A. male
B. female
Q3.Your education background
A. high school or below
B. undergraduate
C. postgraduate or above
Q4. marital status
A. single
B. married
C. divorced
D. bereft of one's spouse
Q5. Income per month (RMB)
A. less than 1000
B.1000-2000
C.2000-3000
D.3000-4000
E. more than 4000
Q6. Occupation
A. Student
B. Management
C. Professional
D. Housewife
E. Employee of company
F. Government employee
E. Private business
F. Others

And there are also 3 more questions to help with dividing the consumers into different groups, these questions are:
Q1.How often do you eat in Chinese restaurant?
A. Less than 1 monthly
B. 1-3 per month
C. More than 3 times a month#p#分頁標題#e#
2. How much do you spend per person when you eat in a Chinese restaurant (RMB)?
A. Less than 50
B.50-99
C.100-149
D.150-199
E. More than 200
3. Who do you often go to Chinese restaurant with?
A. friends
B. boyfriend/girlfriend
C. family
D. workmates
E. classmates
F. by myself
4. Sampling
4.1 Sample Size
According to Miller (1991), a sample size is a smaller representation of a large whole. There are usually two ways to determine the sample size. On one hand, the optimal sample size can be calculated, given a desire level of precision and cost, according to the standard level formula. This can only be used in probability sampling. On the other hand, an arbitrary size within the constraint of the research budget, manpower and time available can be set. In this case, sample size is dependent on the research questions and objectives in particular, what you need to find out, what will be useful, what will have credibility and what can be done within your available resources(Patton, 2002). Because of the characteristic of on-line survey, and the time limit, the researcher decided to set the sample size of this study for 100.
4.2 Sampling Technique
According to Saunders (2007), in business research, such as market surveys, probability sampling may either not be possible or appropriate to answering the research questions. This means this study’s sample must be selected some other way. Non-probability sampling provides a range of alternative techniques to select samples based on the subjective judgment, it may be the most practical sampling technique for this study.
To answer the research questions and to meet the objectives with limited resources and the inability to specify a sampling frame, the population is numerous and it would be extremely expensive in terms of time to study all the population elements. Therefore, researcher could use a number of non-probability sampling techniques. These include convenience and self-selection sampling techniques, purposive sampling and snowball sampling techniques.
Self-selection sampling needs individuals to identify their desire to take part in the research, self-select often do so because of their feelings or opinions about research questions or stated objectives. In this case, the researcher has to publish the need for cases such as advertising through certain media to ask the individuals to take part. Base on the situation of this study, it is too expensive for using this sampling technique.
Convenience sampling involves selecting haphazardly those cases that are easiest to obtain for the sample, the sample selection process is continued until the required sample size has been reached. However, this sampling technique has big disadvantage, that is it is prone to bias and influences that are beyond the control, the choice of sample is likely to have biased the sample.
Snowball sampling is used when it is difficult to identify members of the desired population, in this study, the researcher contacted with several cases in the population first, and then ask them to indentify further cases in order to find more cases in limited time.#p#分頁標題#e#
Purposive sampling is preferred by the researcher in this study because it enables the researcher to use judgement to select cases that will best answer the research questions and to meet the objectives. This form of sample is often used when the researcher wishes to select cases that are particularly informative. (Neuman,2000)
As it is said above, considering the advantages and disadvantages of all the four sampling techniques, the researcher decided the sampling techniques used in this study are snowball sampling and purposive sampling.

5. Data Collection
Data collection is conducted completely through internet. That is because of three reasons: (Saunders 2007)
Firstly, in order to maximize the questionnaires converge, the Internet is believed to be most effective way to assure respondents’ variety and quantity. Besides, on-line questionnaire survey can save a lot of money and papers to keep the research more economical and prevent unnecessary waste. Finally, on-line survey generates raw data automatically in database. It saves a lot of time and avoids human mistakes when coding data.
According to Saunders (2007), all questionnaires should be pilot tested prior to collecting data to assess the validity and likely reliability of the questions. The purpose of a pilot test is to refine the questionnaire so that respondents will have no problems in answering the questions and there will be no problems in recording the data. In order to make the questions understood by the respondent in the way intended by researcher and the answer given by the respondent understood by the researcher in the way intended by the respondent.
On-line questionnaires are firstly sent to two foreign classmates and three more foreign friends and relatives of the researcher by E-mails for pilot test, so that it can test if the designed questionnaire is capable of meeting the objectives of this study. The results of these five piloted questionnaires were inputted into a SPSS spreadsheet in order to check the reliability and validity of the scale, and questions as inappropriate, equivocal or inductive are amended.
The content of e-mail also asked them if they could help forward this questionnaires hoping to increase sample size of study. The feedback of these five e-mails involved further information about cases that are particularly informative. One of the feedbacks suggested sending the questionnaires to the students in a Chinese language school for foreigners. And another feedback suggested sending the questionnaires to the foreigners working with him and provided help with forward sending the e-mails to them.
100 e-mails were sent in total and of them answered the questionnaires, all of them are considered effective, that made the sample return of the questionnaire is .
6. Data Analysis Procedures
Several statistical analyses and multivariate techniques are taken for investigating objectives of this study. These analyses are conducted mainly by the software SPSS version 12.0.1#p#分頁標題#e#

6.1 Descriptive Statistics
http://www.ttc-switch.com/Thesis_Tips/Methodology/Characteristics of respondents are firstly summarized after finishing questionnaire survey. In this stage, descriptive statistics are listed in the form of table to show the delineation of the sample.
To narrow down the complexity of data by statistic values such as percentage, mean value and standard deviation and make use of descriptive statistics to explain the social and economical backgrounds of correspondents as well as the distribution status of the variable of consumption levels.
6.2 Inferential Statistics
6.2.1 Validity Analysis
This is the most fundamental verification and analysis process in the questionnaire design, which aims to assess the fitness of consumers toward restaurant selection program, service quality program and their behavioral intention, thereby understanding if there is substantive fitness. If it is below the significant differential level, the amendment and adjustment of description method should continue.
6.2.2 Reliability Analysis
According to McDaniel and Gates (1998), reliability is the degree to which measures are free from random error and thus provide consistent data. “Consistency indicates how well the items measuring a concept hang together as a set” (Sekaran 2000). The most frequently used statistic to measure internal consistency is Cronbach’s coefficient alpha, which provides an indication of the average correlation among all of the items that from the scale. To verify the internal coherence of items in questionnaire measuring indicator by Cronbach’s coefficient alpha, assess the reliability level as well as stabilization level of the whole questionnaire measuring indicator, and adjust the indicator items of lower reliability.
Cuieford (1965) suggested Cronbach’s alpha coefficient value lower < 0.35 as low reliability, 0.35 ≦Cronbach’s alpha coefficien value < 0.7 as just acceptable reliability, Cronbach’s alpha coefficient value ≧ 0.7 as high reliability.

6.2.3 Factor Analysis
To induce common fundamental factors from those of respondent consumers toward the restaurant selection program and service quality assessment program, which aims to deduce many variables into several main categories of factors according to their relevance, so as to narrow down the complexity among variables in hopes of building up an utmost explanation factor for original variables.
6.2.4 Cluster Analysis
To ensure high homogeneity of samples in the same group while high heterogeneity of samples in different group according to specific factors of sample information. Based on the restaurant selection program of respondent consumers, this study makes use of cluster analysis to divide consumers into groups.
6.2.5 One-way ANOVA
It aims to compare the mean value of three or above groups and calculate if the mean value of different groups (matrix) is equal. This method is to verify the differences in cognition of above three different groups toward each question. This study makes use of single factor variation to analyze the discrepancy of different consumer groups toward restaurant selection program.#p#分頁標題#e#
6.2.6 Pearson Product-moment Correlation Analysis
Correlation analysis is used as a solution to verify correlation between two sets of variables which belong to different constructs. This analysis clarifies positive or negative relations between factors and other dimensions.
This is a statistical analysis method applied to verify linear relation of two variables and applicable for depiction of correlation between two consecution variables. However, consecution variables interval scale or ratio scale measuring indicator must be used for consecution variables in questionnaire design. Normality test is used to show the distribution of the score if they are normal. Investigating the result 留學生dissertation網the research can decide which kind of correlation analysis can be used Pearson or Spearman correlation.
This study is applied to analyze the correlation between restaurant service quality assessment program factors and overall dining satisfaction as well as that between overall dining satisfaction and behavioral intention.

7. Access and Research Ethics
The Ethical Issues form was completed and no ethical issues were identified (refer to Appendix where form can be found)