Impacts of Perceived Risk and Attitude on Internet-Purchase Intention

This research paper aims to explore the correlations among perceived risk, attitude and intention of purchase through internet. Also, the research investigates the significant differences among the customers’ perceptions of the perceived risk, attitude about internet-purchase according to; gender, education, age and use experience. The hypotheses are tested with e-survey data collected from 412 participants from Saudi Arabia. According to the best of author’s knowledge, this is the first research that investigates the correlations of the variables structure according to Saudi people. The author recommends future researchers examine the same relations in researches focusing on other markets.


Introduction
The world total number of internet active users represented 3.29 billion users by the end of March 2013 (IWS, 2013).The rapid growth of the internet population has changed business and social interactions.The survival of e-commerce depends on web site recognition and steady traffic flow (Chen et al., 2013).Around 15% of corporations have presence on internet (DDA, 2013).Global e-B2C reached 1043 Billion USD in the year 2012, with increase rate 21.7% compared with the year 2011, the estimated value in 2013 is 1221 Billion USD with increase rate 17.1% (eMarketer, 2013).The online shopping continues to grow every year (Desai et al., 2012).The use of internet as a mean of shopping goods and services has been growing over the past few years.The number of internet stores, which through e-commerce are able to conduct transactions over the internet, has been steadily increasing (Liang & Lai, 2002;Abdelkader, 2013), while revenues from online shopping expose considerable growth (Rohm & Swaminathan, 2004;Santouridis et al., 2012).Companies increasingly rely more on internet services because they are more convenient, interactive, have lower costs and offer a high degree of customization and personalization to their customers (Park & Baek, 2007;Santouridis et al., 2012).
The number of customers who choose to purchase goods and services online is continuously rising, as they are usually able to find better prices in a more convenient manner (Chen & Chang, 2003;Santouridis et al., 2012).A lot of consumers may use the internet only as an information-search tool (Seock & Norton, 2007).Seock and Norton, (2007) found that consumers' information search on the internet was the key predictor of their intentions to purchase through internet.Online retail spending is billions of USDs through these days increasingly and rapidly more than any time before (Forbes, 2009).About the B2B e-commerce usage patterns, there are approximately 27 million Small to Medium-sized Enterprises (SMEs) in the USA, which account for 99.9 percent of all US firms and generate almost half the Gross Domestic Product (GDP) of the country (Sila & Dobni, 2012).In Canada, there are around 1.4 million SMEs (Riding & Orser, 2007).Small businesses (those that have fewer than 100 employees) constitute 98 percent of all business in Canada and generate over 30 percent of Canada's GDP.Overall, Canadian SMEs account for 60 percent of Canada's economic output and 80 percent of national employment and 85 percent of new jobs (Sila & Dobni, 2012).E-commerce of the Middle East reached 9 billion USD through 2012 with an estimation to reach 15 billion in 2015, includes business through mobile around 4.9 billion USD.68% of users are males and 32% are females.Around 70-80% of them prefer cash on delivery (DDA, 2013).The literature confirms that, e-services quality has been in the focus of researchers for the past decade or so.However, it is generally acknowledged that research is still at a relatively early stage (Santouridis et al., 2012).The number of scientific publications on logistics has been growing steadily in recent years, reflecting the significance of this aspect.However, most reported works on distribution and network area are limited to traditional business.Therefore, they are no longer applicable to the background of e-business (Yan et al., 2012).Research is still in need to special topics; such as Information and Communication Technology (Asare et al., 2012), The key Success Factors (Sila & Dobni, 2012) and the Model of Golden Square of Marketing "IPSL", which is a new proposed framework to manage the audience-mind and to handle the CRM, it includes Image, Positioning, Satisfaction and Loyalty (Abdelkader, 2014).Turban ( 2002) defines e-commerce as the process of buying, selling, or transacting business products, services and information using computer networks including the internet (Asare et al., 2012).There are Interaction between persons, computers and networks (ISO, 1998;Hasan et al., 2013).
The literature confirms that, the research may add a lot of contributions in the following important areas:

•
Apply on Internet-Purchase (IP) as one of the most important activities and industries.
• Study the relations among the research variables; Internet-Purchase Perceived Risk (IPPR), Internet-Purchase Attitude (IPA) and Internet-Purchase Intention (IPI).

•
Explore the IP in Saudi market, as one of the biggest markets in all over the world.
• Investigate the differences among perceptions of Saudi customers according to their demographic differences, on their IPPR about IP.

Internet-Purchase Perceived Risk (IPPR)
Over the last years, considerable researches have focused on the concept of Perceived Risk (PR) and its impact on customer behavior (Page & Luding, 2003).Marketers should distinguish between PR and Real Risk.PR is an important construct in marketing, especially in buying behavior (Tuu & Olsen, 2012).PR expresses the risk according to the standpoint of customers and their perceptions.PR is a strong determinant of customer attitude towards internet-purchase (Page & Luding, 2003).The study of Brengman and Karimo, (2012) confirms that trust is a significant factor in internet-purchase more than non-internet-purchase, because of the absence of physical contact.A part of customers view purchase from internet-purchase as a risky business customers, they believe that the internet is not secure place to purchase and they need to examine the product they buy like the ordinary purchase (Page & Luding, 2003).On the other hand, there is a rise in the volume of the internet-transactions as a commercial tool whether B2B or B2C (Alharbi et al., 2013).Coker et al., (2011) define Internet-Purchase Perceived Risk (IPPR) as the degree of uncertainty an individual perceives towards purchasing a product on the internet (Coker et al., 2011).Risk towards internet-purchase varies greatly not just among individuals, but also depending on the product (Coker et al., 2011).IPPR could be classified according to the time factor in two classifications; pre-purchase and post-purchase, they refer to the probability of suffering before/after purchase according to the buyers (Tuu & Olsen, 2012).Generally, IPPR is considered as a construct comprising of a set of interrelated multi-dimensional component losses, due to the uncertainty of achieving goals set in product purchase (Tuu & Olsen, 2012).Perceived risk is an important construct in marketing, particularly in studies of online consumer behavior.Several studies suggested that, IP have two dimensions comprised of product evaluation and information security (Coker et al., 2011).These dimensions refer to the degrees of uncertainty and mistrust about product quality or information security (Coker et al., 2011).PR was assessed by asking respondents to indicate their evaluations on two important aspect; quality and safety (Tuu & Olsen, 2012;Kim & Lennon, 2013).PR includes: Financial, Performance, Physical, Psychological, Social and Time factors (Page & Luding, 2003;Cunningham et al., 2005).Initial internet trust is important in the context of IP because it decreases PR and positively affects purchase intentions (Brengman & Karimo, 2012).In the context of IP, the information privacy is essential in measuring PR (Alharbi et al., 2013).Internet privacy was defined as an "individual's right to access and control their personal information with respect to its collection, use and transfer over the Internet" (Brengman & Karimo, 2012).PR are measured by three variables; trust in security, difficult to judge quality and trust in keeping personal information privacy (Tuu & Olsen, 2012;Kim & Lennon, 2013).The following hypotheses cover significance testing of differences among the customers' perceptions according to their demographic differences: H 1 .There are significant differences among customers' IPPR according to Gender.
H 2 .There are significant differences among customers' IPPR according to Education.
H 3 .There are significant differences among customers' IPPR according to Age.
H 4 .There are significant differences among customers' IPPR according to Use Experience.

Internet-Purchase Attitude (IPA)
Over the last years, considerable researches have focused on the concept attitude.Chen et al., (2013) assumes that an individual's attitudes and subsequent behaviors are affected by his belief about the system's overall quality and the experience of using the system.Attitude is an immediate determinant of intention to perform a behavior (Seock & Norton, 2007).One's attitude may capture the totality of the expectation level and it may provide the baseline for other cognitions of an overall nature, particularly satisfaction (Ha et al., 2010).Researchers agree that attitudes play a major role in shaping behavioral intentions toward brand, store, shopping, purchase and retailer (Seock & Norton, 2007).Literature of Wu and Lo, (2009) confirms that, attitude is the important factor of consumers' purchase intention on extended products.Consumers' use of the internet for information search was positively affected by their attitudes toward the processing speed and transaction security of web sites where they shopped (Seock & Norton, 2007).The internet's information-rich, interactive nature can increase shopping efficiency by improving the availability of product information, enabling direct multi-attribute comparisons, and reducing information-search costs (Seock & Norton, 2007).Literature of Seock and Norton, (2007) has found positive relationships between consumers' attitudes toward online shopping and their purchasing through the internet.Literature shows that researchers measure "Attitude" as a variable by the following items: Bad/Good (Coker et al., 2011;Park et al., 2007;Ha et al., 2010), Pleasant/Unpleasant (Ha et al., 2010;Park et al., 2007), Like/Dislike (Teng et al., 2007;Ying & Chung, 2007;Wu & Lo, 2009), Useless/Beneficial (Coker et al., 2011;Park et al., 2007;Ying & Chung, 2007), Prefer (Wu & Lo, 2009), Attractiveness (Ying & Chung, 2007), Undesirable/Desirable (Coker et al., 2011), Negative/Positive (Coker et al., 2011), Rewarding/Punishing (Park et al., 2007), Enjoyable (Ha et al., 2010), Satisfactory/Unsatisfactory, and Appealing/ Unappealing (Teng et al., 2007).
H 5 .There are significant differences among Consumers' IPA according to Gender.H 6 .There are significant differences among Consumers' IPA according to Education.
According to Seock and Norton, (2007), young consumers, particularly college students aged 18-22 years, are becoming the internet's "hottest" market and a prime source of future growth in online sales in USA.
H 7 .There are significant differences among Consumers' IPA according to Age.
H 8 .There are significant differences among Consumers' IPA according to Use Experience.
H 9 .IPPR has a significant impact on IPA.
H 10 .IPPR has a significant impact on IPI.
H 11 .IPA has a significant impact on IPI.

Items Generating & Validity
The questionnaire was built on "Google Drive", and it was disseminated through many e-means of communications; included: e-Mail, SMS, Tweeter, Facebook, Blog and WhatsApp.These ways of distribution were driven by a lot of earlier studies (Libai et al., 2010;Verhoef et al., 2010;Ojiako et al., 2012;Abdelkader, 2015).There were three main resources to generate the items of the questionnaire.First resource was the literature review.Second one is the semi-structured interviews, that the research conducted 35 in-depth interviews with some participants from Saudi Arabia.Finally, nine experts and academicians from related fields reviewed the primary contents of the questionnaire and suggested some changes to be more valid.

Questionnaire Structure
Appendix A shows the significant items of the constructs.Multi Regression Analysis MRA was used to analyze the related items of each construct.The final structure of the questionnaire includes three constructs.First construct includes 3 items for measuring IPPR.Second one includes 3 items for measuring IPA.Third construct includes 3 items for measuring IPI.An overall evaluation item was added to each construct in order to test the significance of items.The questionnaire based on Likert scale on a five-point scale (1 representing "Strongly Disagree" to 5 representing "Strongly Agree") their level of agreement for each statement of the three constructs.

Sample
Participants of the sample received the URL of the Google-Drive' questionnaire electronically.The URL of the e-questionnaire was available for 45 days started on four th December 2014.Selection conditions are 18+ years old and from Saudi Arabia.Considering the period and selection conditions, only 412 respondents were finally selected from participants.The sample is sufficiently large, over the recommended size of 200 cases (Medsker, 1994;Jalilvand & Others, 2012), or the recommended size of 384 cases (Sekaran, 1992).

Reliability
Statistical Package for Social Sciences (SPSS, V. 20) was used in this research to test the reliability by measuring Cronbach's α.The overall values of Cronbach's α for all items together were (α = 0.884).All of them are more than 0.60, so the reliability of the scale are acceptable (Sekaran, 1992).Also, all sub-scales display acceptable reliabilities, these being of the order above the generally accepted value of 0.70 (Hair, 1998).The sub-scales values of Cronbach's α for each dimension of questionnaire were (α = 0.832 for Perceived Risk), (α = 0.932 for Attitude), (α = 0.975 for Intention).

Means of Items
Most of items' means are high according to the following measure; less than 3.00 (Low level "L"), between 3.00 and 3.50 (Moderate level "M") and above 3.5 (High level "H").

Hypothesis Analysis
The following are the results of hypothesis analysis, from H 1 to H 11 through (SPSS, V 20 ): • • Accept H 9 , because there is a significant impact of IPPR on IPA on all significance level; 0.05, 0.01 and 0.001; (Sig.=0.009,R=0.148, R 2 =0.022 and R Adjust =0.019).

Conclusions
According to the analysis of the collected data, some of the conclusions and suggestions that can be drawn from the research are as follows: • There are no significant differences among Participants' IPPR according to Gender, Education, Age and Use Experience.In general, the levels of sample' IPPR are high.Marketers are in-need to give a special attention in their plans to show their concern of reducing the IPPR.Also, they are not in-need to differ among targeted users according to Gender, Education, Age and Experience of use.
• There are significant differences among Participants' IPA according to Gender and Use Experience.In general, the levels of samples' IPA are high.Marketers are in-need to keep these levels high and increase them.Also, they are in-need to differ among targeted users according to Gender and Experience of use.
• There are no significant differences among Participants' IPA according to Age and Education.In general, the levels of sample' IPA are high.Marketers are in-need to keep these levels high and increase them.Also, they are not in-need to differ among targeted users according to Age and Education.
• There is a significant impact of IPPR on IPA.Marketers are in-need to give a special intension in their plans to care of reduction of the IPPR to increase IPA and IPI.
• There is a significant impact of IPPR on IPI.Marketers are in-need to give a special attention in their plans to show their concern of reducing the IPPR to increase IPI.
Future research should consider bigger sample size.Ideally a larger sample size would provide a clearer understanding of the relationships between the variables.Also, future researches should include sample from more than one country.

Figure 1 .
Figure 1.The structure of the research hypotheses Note.(*) Significant on all significance levels 0.05, 0.01 and 0.001.

Table 1 .
Sample structure

Table 2 .
Sample structure