Posted: October 20th, 2022

The Questionnaire Development And Scaling Technique

Study Bay Coursework Assignment Writing Help

This chapter explains the methodology use to collect, examine and interpret the data obtained in order to better understand the relationship between product attributes and customer loyalty. Section 3.2 consists of the research framework, the explanation of research framework and the discussion of each independent variable and dependent variable, in the meantime hypotheses will be developed in section 3.3. Data collection method is discussed in section 3.4. Then, data source will be discussed inside section 3.5 Follow by section 3.6 and 3.7 which talk about the sampling method and sampling size of the research. Last but not least, the questionnaire development and scaling techniques is discussed in section 3.8, the data analysis and measurement in section 3.9 and the data analysis method in section 3.10.

A research framework is a theoretical model of how researcher theorizes or constructs logical sense of the relationships between the determinants and the problem (Sekaran, 2003). This theory is developed through the literature review of previous research in the similar problem area. Thus, put together the logical theory that publish by previous researcher, and develop a methodical basis in order to examine the research problem. In other word, the research framework discusses the interrelationships between the variables that are believed to be important to the area being investigated.

Moreover, a research framework Helps us to develop, hypothesize and examine the relationship between the variables and also helps us to better understand the cases of the problem ( Sekaran, 2003). Through the research framework, hypotheses can be developed to investigate whether there is a relationship between the variables which is related or associated to the research. The truthfulness of the relationship can be tested through various statistical analyses. Therefore, the research framework plays a crucial role in the entire research.

Since the research framework works as the mechanism to investigate the relationship between the variables, it is important to identify the meaning of the variables (Sekaran 2003). A variable is anything that affects the outcome or varying values. The outcome at different time can be different even from the same people or object. For example, the winter jacket, Ali uses to buy the winter jacket when it is winter time for daily use but yet Ali won’t buy a winter jacket at summer time. This is because different season have different weather or different temperature.

In this research, it consists of two variable which is dependent variable (customer loyalty) and independent variables which consists of (product’s price, store location, product’s quality and brand image). Dependent variable (customer loyalty) is the variable that serve as the primarily curiosity in this research (Sekaran, 2003). The mission of this research is to understand the customer loyalty, to forecast its outcome and to make clear its variability. Throughout the analysis of customer loyalty, the solutions or ways to solve the problem being investigated can be found.

On the other hand, product’s price, store location, product’s quality and brand image) are the variables that affect the customer loyalty in a either positive or negative way (Sekaran, 2003). In other word, each unit increase in the product’s price, store location, product’s quality and brand image will affect increase or decrease in customer loyalty. Therefore, the discrepancy in the customer loyalty is determined by the product’s price, store location, product’s quality and brand image.

In conclusion, this particular section is to investigate whether there is a relationship between the independent variable (product’s price, store location, product’s quality and brand image) and the dependent variable (customer loyalty).

3.2.2 Operational Definition

Customer Loyalty

Customer loyalty is defined as the customer intention to repurchase a particular product or service in future time; resist other better alternatives; ignore price differences and as an interaction between customers’ relative attitude towards a brand or store.

Dick and Basu (1994); Wong and Sohal (2003)

Product’s Price

Price is the gaining value that accepted by a customer, whereby engaging both ‘give’ and ‘get’ elements.

Padula and Busacca (2005)

Store Location

Convenience-seeker has no special commitment toward the store or company but yet, just because the store location is nearby and convenience then the customer makes repetitive purchase

Rowley (2005)

Product’s Quality

Product quality means that the characteristics of product that fulfill the needs and wants of a customer.

Russell and Taylor (2006)

Brand Image

The amount of beliefs, understanding and feelings that customers have toward a product or service.

Crompton (1979)

3.3 Hypothesis Development

A reasonably conjectured relationship between two or more variables conveyed into a testable statement (Sekaran, 2003). Hypothesis can be defined as a tentative explanation for an observation, phenomenon, or scientific problem that can be tested by further investigation (The American Heritage, 2000). In other words, it means that something taken to be true for the motive of argument or investigation.

In this research paper, four hypotheses are tested in order to improve the understanding on the effect of product attributes (product’s price, store location, product’s quality and brand image) toward customer loyalty. Null (Ho) and alternate hypotheses (H1) will be used in this research paper. Null hypothesis means that the exact relationship between two variables. It states that the population correlation between two variables is equal to zero which explain there is no significant relationship between the two variables. However, the alternative hypothesis is a statement that shows there is a significant relationship between two variables (Sekaran, 2003).

3.3.1 Hypothesis І: Product’s price and Customer loyalty

According to Romaniuk and Dawes (2005), product’s price is one of the critical product attribute that affect customer loyalty and repurchase behavior. Price can be considered as the important factor that influences buying behavior among the various offerings in the different category of products (Romaniuk and Dawes, 2005). Furthermore, Lichtenstein et al. (1993) stated that product’s price plays a vital role that may influence consumer buying in all purchasing situations. Customers are sensitive to product’s price and often make buying decision based on product’s price (Romaniuk and Dawes, 2005). Price of a product directly affects the customer’s buying intention and customer’s value awareness (Biswas et al., 2002). Therefore, the hypotheses below are developed:

Ho: There is no significant relationship between product’s price and customer loyalty.

H1: There is a significant relationship between product’s price and customer loyalty.

3.2.2 Hypothesis ІІ: Store location and Customer loyalty

Store environments include store location, store design, and in-store stimuli, will affect customer loyalty toward a particular store (Yee and Sidek, 2008). However, there are four types of loyal customer such as contented, convenience-seeker, captive and committed Rowley (2005). Convenience-seeker means that a customer has no special commitment toward the store or company but yet, just because the store location is nearby and convenience then the customer makes repetitive purchase and becomes loyal. Lastly, location of a store is the most important factor that affects customer to make purchase in a particular store and customer loyalty (Kim & Jin, 2001; Craig et al., 1984; Bruner & Masson, 1968; Nevin & Houston, 1980). Hence, hypotheses are constructed as below:

Ho: There is no significant relationship between store location and customer loyalty.

H1: There is a significant relationship between store location and customer loyalty.

3.2.3 Hypothesis ІІІ: Product’s quality and Customer loyalty

According to Yee and Sidek (2008), customers may repurchase the particular brand of product due to the tangible quality of the product sold. Customer’s repurchase decision depends greatly on the quality performance of products or services (Chaudhuri and Holbrook, 2001). Quality is the factors the customer use to make purchase decision or to consume the products (Fandos and Flavian, 2006). The perceived quality of a product is frequently used as a primarily determinant of loyalty and it can also contributes positively increase to loyalty (Ruyter and Wetzels, 1997). In previous research (Boulding et al. 1993), the result shows that there is a positive relationship between product’s quality and customer loyalty or intention to repurchase again. As a result, the hypotheses are developed as below:

Ho: There is no significant relationship between product’s quality and customer loyalty.

H1: There is a significant relationship between product’s quality and customer loyalty.

3.3.4 Hypothesis ІV: Brand image and Customer loyalty

Brand image helps a company to preserve it’s the existing customers and to build loyalty among them (Bloemer and de Ruyter, 1998; Nguyen and Leblanc, 2001; Pan and Zinkhan, 2006). Johnson et al., (2001); Helgesen and Nesset, (2007) declared that brand image has positive relationship with customer loyalty. In other words, the better the brand image perceived, the more loyal a customer will be. Barich and Kotler, (1991); Zeithaml, (1981) affirmed that brand image plays an important role in customer repurchase buying behavior. Brand image helps to retain and to develop loyalty on customers (Dick and Basu, 1994; Porter, 1985; Raj, 1985; Reynolds et al., 1974-1975). Consequently, hypotheses are constructed as below:

Ho: There is no significant relationship between brand image and customer loyalty.

H1: There is a significant relationship between brand image and customer loyalty.

3.4 Data Collection Method

To accomplish the objectives of this research, questionnaire method is used to collect the data from the target respondents. A questionnaire is a set of questions that develop by a researcher for the target respondents to answer (Sekaran, 2003). Questionnaire is an effective and efficient method to collect data because the questions are developed based on the requirement of the researcher and the researcher for sure will know how to measure their own intended variable of interest.

Questionnaire method allows respondents to choose their favorable answers based on their own perceptions, and knowledge (Sekaran, 2003). In this research paper, researcher will provide few rating scale inside one question so that respondent can rate according to their perception. Rating scale is the scale that respondent use to rate behavior need in terms to the variables questions (Eg. Strongly Disagree; Disagree; Somewhat Disagree; Neutral; Somewhat Agree; Agree; Strongly Agree).

On the other hand, there are several types of questionnaire which are mail questionnaire, online questionnaire and personally-administered questionnaire. In this research paper, personally-administered questionnaire will be used to obtain needed data. Personally-administered questionnaire means that a survey is conducted in a specific area, but yet the researcher or member of the research team is personally administer the questionnaire answering process of the respondents (Sekaran, 2003). The major advantage of personally-administered questionnaire is easier for the researcher or member of the research team to collect back all the completed responses in a short time.

Lastly, the researcher is able to explain the questionnaire if the respondents do not understand the question. Through this action, responses from the respondents will be more accurate and reliable because uncertainties of the respondents can be make clear on the spot without any delay of time (Sekaran, 2003). Besides that, administered questionnaire is less expensive, affordable and consumes less time if compare to other types of data collection method such as interview or observation.

3.4.1 Pilot Study

Pilot study is referring a small scale of questionnaire distribute to individuals and gather the information from this limited number of respondents (Sekaran, 2003). In sequence, potential problem areas and deficiencies in the research instruments and protocol prior to implementation before the distribution of real questionnaire can be identify (Zailinawati, Schattner & Mazza, 2006). It can help to decide between two competing study methods, such as using a personally-administered questionnaire rather than interviews.

As questionnaire had been choose as data collection method in this research paper, before to distribute the actual questionnaire to the respondents, pre-test questionnaire is needed. Sekaran (2003) stated that pre-test questionnaire is needed in order to test the understandability of the questions that included in the actual questionnaire paper. Thus, pre-test questionnaire is used to prevent the conflict and contradiction between the variables and questions. In this research, the estimated target respondents for this pilot study are 30 individuals randomly selected respondents from different shopping center.

3.5 Data Source

To conduct this research, data are taken from various different sources to form a sample to test the hypotheses developed. Data is diversified into primary and secondary (Sekaran, 2003).

3.5.1 Primary Data

Primary data means that the direct or first-hand information that can be obtained by a researcher through interview, observation or personally-administered questionnaires (Sekaran, 2003). Interview means that the researcher or member of research team personally questions the respondents to collect information needed (Sekaran, 2003). It can be structured or unstructured and thus it can be carry out through in person, online or telephone.

On the other hands, Sekaran (2003) had stated that observation refers to a researcher or member of research team observes the target respondent’s daily activities, behaviors, attitudes, body language and facial expression to obtain the intended information. All the interest phenomena should be recorded so that biases can be diminished.

Personally-administered questionnaires imply that a written set of questionnaire is prepared for the target respondent to answer and researcher or member of research team is personally administer the answering process (Sekaran, 2003). In addition, it allows the researcher or member of research team to collect back the questionnaire right after the respondents finish the questionnaire and any doubts can be clarify by researcher or member of research team instantaneously.

3.5.2 Secondary Data

According to Sekaran (2003), secondary data refers to information acquired from existing sources such as books, magazine, journals, media, web sites, interview, annual reports of company, financial statement and publication of government.

In conclusion, due to the differences of cultural, education, and income level, the secondary data obtained from other researcher in different countries cannot be 100% represent to the situation in Malaysia. Therefore, this research study is focused on primary data. It means that all the intended data is obtained through personally-administered questionnaire.

3.6 Sampling Method

Sampling method can be diversified into two major types: probability and non-probability sampling (Sekaran, 2003). Probability sampling means that the elements of the population will have an acknowledged chance of being selected as subjects in a sample. Meanwhile, non- probability sampling refers to the elements of the population do not have pre-established chance of being selected as sample subjects. If the subject’s response of a sample can be generalize, then probability sampling will be used. In contrast, if the different response from different subject is needed, then non-probability sampling is more suitable to use than probability sampling.

According to Sekaran (2003), probability sampling can be diversified into two different groups which is complex probability sampling (restricted) and simple random sampling (unrestricted). There are five types of complex probability sampling: stratified random sampling, systematic sampling, area sampling, cluster sampling, and double sampling (Sekaran, 2003). Stratified random sampling involves a separation or stratification process, and then each subject is randomly selects from each stratum. However, systematic sampling means that every nth element in the entire population will be drowned out. Besides that, area sampling refers to geographical clusters. It means that the subjects of area sampling are selected from or within an area. Hence, cluster sampling can be explained as subjects of a sample are selected based on groups of element which have different characteristics among the elements within each group. Last but not least, double sampling involves of a small sample is used to collect the primary data and then a subsample of this small sample will used to investigate the problem more detail.

Conversely, simple random sampling means that each of the elements has equal chance of being selected (Sekaran, 2003). For instances, there are 1000 residents in Bukit Beruang, in order to study their characteristics or behaviors, then 200 residents are randomly selected from the whole population. Anyone of the resident can be selected as the subject to a sample. This sampling method provides least bias and the responses from this sample can represent the characteristic of the population.

Sekaran (2003) declared that non-probability sampling can be diversified into purposive sampling and convenience-sampling. Purposive sampling means that the researcher collects the intended information from a specific group of people. In short, it means that the desired information only can obtain from those who met the criteria set by the researcher. For examples, to understand the behavior of a director, then the respondent who is qualify to answer the questionnaire will be those is in the position as a director in a company.

Last but not least, convenience-sampling refers to the target respondents of a research paper will be those who are available or convenient to provide the responses (Sekaran, 2003). For instances, a cosmetic company, in order to understand the consumer’s choice of product, the company will held such a research at the shopping mall that full of shoppers. Thus, the shoppers will form a sample for the research and the shoppers will be those who are convenient to answer the questionnaire. Therefore, in this research paper, convenience-sampling will use as the sampling method in this research paper. This is because convenience-sampling is speedy, convenient and less expensive if compare to other sampling method. However, simple-random sampling is not suitable to use because it is impossible to get a hold of the entire up to date listing of the population of a place.

3.7 Sampling Size

Sampling means that the process of choosing an adequate amount of elements from the population and this sample of elements can represent the characteristics of the whole population (Sekaran, 2003). The reason why a sample is needed rather than the whole population is because a population consists of thousand of elements and it is impossible for a researcher to collect data from each of the element. Even if possible, it will consume a lot time, expenses and human resources.

Thus, the study of a sample can reduce the probability of fatigue and errors. This is because the researcher can just focus on a sample rather than the entire population which has thousand of elements. For examples, Nokia mobile phones, each day the company produces thousands or millions of mobile phones, it is impossible to test the functionality of each mobile phones because it will waste a lot of time and cost.

According to Sekaran (2003), the sample size can determine the level of precision and how confidence that the result of the research paper can be generalizability. As a result, the sample size for a research paper needs to be emphasized. For examples, if the sample size of a research is less that 30, the result of the research will be not accurate and precise or not normal. This is because such a small sample can’t represent the characteristics of the entire population of 10000 peoples. Hence, if the sample size is too large, even the weak relationships can reach the significance level and the researcher might assume that the result can represent the population but yet, it can’t apply in reality (Sekaran, 2003).

Therefore, in this research paper, a total of 150-200 questionnaires will be distributed to respondents so that the result of the research paper can be more precise and accurate. Thus, the projected questionnaires that can collect back can’t be determined accurately due to the probability of lost questionnaire or uncompleted questionnaire (blank responses).

3.8 Questionnaire Development and Scaling Technique

This questionnaire is designed to study the effect of product attributes on customer loyalty from retail industry. The questionnaire requires customers who can be anyone that visit the retail company such as Tesco, Carrefour, Giant, Mydin and The Store. Researcher or member of research team will Help the respondents in case they have problem in understanding the questions.

The questionnaire consists of two parts. The first part dealt with gathering the background information about the respondent by giving few choices for the respondent to choose, such as age of the respondent, gender of the respondent, marital status of the respondent, cultural background of the respondent, education level of the respondent, monthly income level of the respondent, store that preferable by the respondent and respondent’s frequency of visitation to retail firms.

However, the second part of the questionnaire consists of all the questions about each variable in the research framework. These questions, measured on a five-point Likert scale rating, with 1 being “strongly disagree”, 2 being “disagree”, 3 being “neither agree or disagree”, 4 being “agree” and 5 being “strongly agree”. Likert scale is designed to examine how strongly subjects agree or disagree with the statements on a five-point scale (Sekaran, 2003). Lastly, there are five questions set on each variable.

Table 3.2: Questionnaire Items and Sources

Variables

Questionnaire Items

Sources

Dependent Variable: Customer Loyalty

1) Loyal customers can bring a lot of benefits to a company by repurchase in future.

Ganesh, 2000.

2) In future years, I would still often purchase from this store.

Grace T.R.Lin and Chia-Chi Sun, 2009.

3) I will encourage my good friends or relatives to purchase this company’s products.

An-Tien Hsieh and Chung Kai Li, 2007.

4) This company’s products will be my first choice when I need to buy any retail products.

An-Tien Hsieh and Chung Kai Li, 2007.

5) I can hardly consider changing to other retail store.

Grace T.R.Lin and Chia-Chi Sun, 2009.

Independent Variable: Product’s Price

1) Product’s price is an important factor that affects store choice.

Yavas, 2003

2) Product’s price affects customer loyalty and repurchases behavior.

Romaniuk and Dawes, 2005

3) Increase of product’s price will not affect my purchase decision.

Wong Foong Yee and Yahyah Sidek, 2008

4) The lower product’s price is usually my choice.

T.Pysarchik, 2001

5) I consider product’s price first in every purchase.

T.Pysarchik, 2001

Store Location

1) If the store is easily to reached, then I will visit often.

Evan, Moutinho and Raaij, 1996

2) Store location makes the customers to make repetitive purchase.

Rowley, 2005

3) One reason for visiting the store is convenient access

Guijun Zhuang, Alex S.L.Tsang, Nan Zhou,Fuan Li, J.A.F. Nichollas, 2005

4) Store location affects my frequency of visitation to the particular store.

Yee and Sidek, 2008

5) Store location affects my choice.

Kim & Jin, 2001; Craig et al, 1984; bruner & Mason, A68; Nevin & Houston, 1980

Product’s Quality

1) Product’s quality is the factor that consumer uses to make purchase decision or to consume the products.

Fandos and Flavian, 2006

2) I feel safe in buying product from retail store.

Grace T.R.Lin and Chia-Chi Sun, 2009

3) I am convinces in buying products from this particular store.

Grace T.R.Lin and Chia-Chi Sun, 2009

4) Quality of the products was one reason for me to visit the store.

Guijun Zhuang, Alex S.L.Tsang, Nan Zhou,Fuan Li, J.A.F. Nichollas, 2005

5) Product’s quality is important to affect a customer buying decision.

Matthew A.Waller, Sanjay Ahire, 1996

Brand Image

1) In my opinion, this particular product has a good image in the mind of customers.

Nha Nguyen, Gaston Leblanc, 2001

2) Brand image will influence a customer to repurchase.

Wong Foong Yee,Yahyah Sidek, 2008

3) Products offer by this retail store have a positive image.

Serkan Aydin and Gakhan Ozer, 2004

4) Highly advertised brands are usually very good.

Venessa P.Wickliffe and Dawn

5) I usually buy well known, recognized or designer brands.

T.Pysarchik, 2001

3.9 Data Analysis and Measurement

There are many different types of analysis method can be used to test or examine the truthfulness of the data obtained such as descriptive analysis, frequency analysis, correlation analysis, reliability analysis, and linear regression (Sekaran, 2003). All of these analysis methods are available in the Statistical Package of the Social Sciences (SPSS) software system. However, in this research paper, there is few analysis methods had been used as below:

Reliability Analysis

Reliability analysis is applied to measure the extent to which the responses are error-free and the consistencies of the responses across time and across the different variables in a research paper (Sekaran, 2003). In short, it is an indicator to measure the stability and consistency of the responses and also helps the ‘truthfulness’ of the responses.

Thus, the reliability analysis is determined by using Cronbach’s alpha (Sekaran, 2003). Nunnally (1978) suggests that the minimum acceptable alpha for scale reliability is 0.60 but not much higher than 0.9 suggested by Streiner and Norman (1989), and the results are within range.

Descriptive Analysis

Descriptive analysis means that the description of the variables (Sekaran, 2003). In the beginning stage, the descriptive analysis is used as the foundation of the entire data analysis process. Thus, it acts to show the general pattern of the responses, measures such as central tendency (mode, mean and median) and variability (standard deviation and variances). Mode means the highest occurrence or the most favorable answers.

Additionally, Sekaran (2003) had declared that mean refer to the average value of the set of values summed up. Hence, median is the measurement of the mid-point value or the values that lies in the middle of a set of value lined up in a proper order. When the entire calculation is completed, it shows the central tendency of the values used.

Meanwhile, standard deviation use to show the degree of variation in the values by converted it into a normal or bell-shaped curve distribution (Sekaran, 2003). However, variances mean that the dispersion of a variable in a set of data obtained. It can be calculated by squaring the results, adding up all the values and dividing the total with the number of respondents.

Frequency Analysis

According to Sekaran (2003), frequency analysis means that the examination on the number of occurrence of a phenomenon. It can be calculated in the ways of percentage and cumulative percentage. For examples, demographic of the respondents such as gender, how many are male respondents and how many are female respondents. If one of the gender’s respondents is higher than the other, the percentage in total will be higher too if compare to other gender.

Pearson Correlation Analysis: Hypothesis Testing

Pearson correlation analysis is used in this research study to measure the relationship between two variables and the strength of relationship between the two variables. Thus, it used to test the hypotheses developed whether the hypotheses are accepted, not substantiated, substantiated or rejected (Sekaran, 2003). In other words, it acts as a method to trace the effect or influence of one variable to another variable. Thus, different number interprets different strength (Hair et al,. 2003). The range between correlation coefficient is -1 to 1. The closer the value of 1; the stronger the relationship between the two variables is.

Besides that, the two variables can be related in three different types, whether positive, negative or zero (Vernoy & Kyle, 2007). Negative correlation coefficient shows that there is a negative relationship between two variables. However, positive coefficient indicates that there is a positive relationship between the two variables. Last but not least, zero coefficient means that there is no relationship between the two variables where else each of it won’t affects another.

Table 3.3: Rules of Thumb about the Strength of Correlation Coefficient

Range of Coefficient

Description of Strength

± 0.81 to ± 1.00

Very strong

± 0.61 to ± 0.80

Strong

± 0.41 to ± 0.60

Moderate

± 0.21 to ± 0.40

Weak

± 0.00 to ± 0.20

None

(Source: Hair, Bush, Ortinau. Marketing Research With A Changing Information Environment. 2003. pp 568 – 569)

Multiple Linear Regression Analysis

According to Sekaran (2003), the multiple regression analysis is a group of techniques, procedures and methods which allow for measurement of the degree of relationship between dependent variable and one or more independent variables. It can be use to determine whether the independent variables explain a significant variation in the dependent variable. In other words, multiple regression analysis can also use to check whether there is a relationship and to predict the values of the dependent variable (Malhotra, 1999). In this research, multiple linear regression is used to determine which product’s attributes (product’s price, store’s location, product’s quality and brand image) are the most significant and contributing factors that affect the customer loyalty toward retail firms.

3.10 Data Analysis Method

There are four hypotheses developed in research. The testing of these hypotheses require using different data analysis method. The table below shows that Multiple Linear Regression is suitable to be applied in the data analysis part of this research.

Hypotheses

Data An