product returns
Improve the internet order fulfilment process for online fashion retailing
MSc Technology & Operations Management
University of Groningen, Faculty Economics and Business
August 14, 2015
A.H.J.S. Hettinga
Student number: 1989685
Fivelstraat 15A
9715BD Groningen
TEL.:+31(0)615116614
e-mail: a.h.j.s.hettinga@student.rug.nl
First supervisor: Drs. J. Van Leeuwen2
Abstract
A lack of physical, tactile product information and customer behaviour at the online storefront is a drawback of online buying, what, inter alia, results in large percentages of online purchases being returned leading to high costs for the online retailer. Product returns, where this research focuses on, erode 30-35 percent of potential profits. The purpose of this research is to see if the product web technology of product images (model presentation versus torso presentation) influence customer behaviour and subsequently influence the amount of product returns, with the use of a dataset and a survey. So, this research investigates the online storefront in the reverse supply chain to improve the internet order fulfillment process. The used conceptual model is based on the technology acceptance model with their actual determinants for online fashion retailing. The technology acceptance model is a foundation for the examination of customers’ approval of online shopping and can therefore be used to see whether it is influencing customer behaviour. At the end it can be linked with the amount of product returns.
To investigate whether product images influence the amount of product returns, this research used a dataset and a survey. The dataset from the consortium apparel company consists of more than 16000 orders where a sample of 22% of the accompanying style identities is used to investigate whether product images influence customer behaviour in returning more or less products. This research used a within-subjects survey with 75 respondents (70 valid respondents).
The findings show that there was a small difference in product returns between the use of a model presentation and the use of a torso presentation. With a model image presentation there were 0,5% more product returns compared to torso presentation. Contrary, from the survey except one variable (information quality), all variables were positively scored by the respondents for the model image presentation. Due to these outcomes this research stated that the customer is better
informed by the model image presentation and therefore is less likely to return products. So, even though from the dataset the conclusion is that model image presentation will result in more costly product returns, from the survey it is clear that online customers prefer model image presentation and that they are informed better by this technology and are therefore less likely to return products compared to the torso image presentation.
3
Contents
Abstract ... 2
1. Introduction ... 4
2. Theoretical background ... 6
2.1 The online storefront ... 6
2.2 Influence of product images on behaviour of online shopping customers ... 7
2.3 Technology Acceptance Model ... 8
2.4 Reverse supply chain ... 9
2.5 Research model and hypotheses ... 12
3. Methodology ... 17
3.1 Data collection ... 17
3.2 Sample data ... 17
3.3 Data analysis ... 18
3.4 Measurement instruments and scales ... 19
3.5 Reliability Analysis ... 21
4. Results ... 22
4.1 Product returns from the consortium partner ... 22
4.2 Profile of the respondents ... 24
4.3 Hypothesis testing ... 25
4.3.1.Model vs Torso image presentation PIQ, PSvQ, PSQ, trust, enjoyment, PEoU and PU ... 25
4.3.2 Perceived ease of use on online buy intention ... 26
4.3.3 Perceived Usefulness on online buy intention ... 27
4
1. Introduction
Online shopping is increasing over the years (CBS, 2014). Online shopping refers to the shopping behaviour of consumers in an online store or a website used for online purchasing purpose (Monsuwe et al., 2004). However, there is a major drawback of buying online, namely a lack of physical, tactile product information and the knowledge of the customer about the product. The lack of this information results, inter alia, in large percentages of online purchases being returned leading to high costs for the online retailer (Stock et al., 2006). Product returns, where this research focuses on, erode 30-35 percent of potential profits (Rogers and Tibben-Limbke, 2000). Even though the amount of online shoppers is increasing, the financial results of the online shop companies suffer from the costs associated with product returns. (Hunter, 2003; Jeszka, 2014)).
The product returns have a negative impact on the efficiency of the order fulfillment process. Supply chain research typically focuses on institutional-level issues (Mollenkopf et al., 2007) whereas
marketing research, which mainly focuses on the consumer, only leads to small changes in the supply chain. This research advances the existing literature in at least two ways. This research combines both type of researches in the reverse supply chain: , if there is an influence on customer behaviour
(marketing) and the amount of product returns (reverse supply chain). Two technologies of product presentation will be taken into consideration.
This study will focus on an area which is prime for research on product returns: online fashion retailing which has to deal with an increasing number of product returns (CBS, 2014). Companies continue to innovate their website designs to gain a competitive advantage (Limayem et al., 2000; Web, 2010) and to enhance customer experience by creating an enjoyable website atmosphere (Eroglu et al., 2003). The two technologies of product presentation that will be taken into
consideration, model and torso product image presentation, are examples of gaining competitive advantage and enhance customer experience at the online storefront (Visinescu, 2014).
5 This research will investigate whether model and torso images presentation at the online storefront influence the amount of product returns and moreover why customers return more or less products.
In this paper the extended technology acceptance model (TAM) (Davis, 1989) is used to determine the variables effecting online shopping. TAM is a model for the examination of customers approval of online shopping (Stoel and Ha., 2009). Perceived ease of use (PEoU) and perceived usefulness (PU) are the two variables of the existing TAM (Davis, 1989), which will therefore also be used in this research. Moreover in this research, the most actual determinants of online (fashion) retailing of the TAM model resulting from the literature review will be implemented into the TAM model. PEoU, PU and those determinants give a more specific explanation/prediction of consumer acceptance of online shopping with the use of the model or torso product presentation technology. Those variables and determinants will be tested on both the model and torso images. Our objective is to test
whether model or torso presentations will get a higher score and investigate why customers who are being faced with product web technology integrations such product images are returning more or less products. A survey is used to test this. Moreover a set of data from a consortium company is used to determine the amount of product returns on both product image web technologies.
So the research question for this research is: ‘What is the influence of model and torso images on customer behaviour at the online storefront and on the amount of product returns for online fashion retailing?
To answer this main research question, this research divided the research question into sub questions which will be the input for the theoretical background.
- What is the influence of product images on the behaviour of online shopping customers? - How can the Technology Acceptance Model be implemented to test whether product web
technologies have influence on customer behaviour and preferences at the online storefront and on the amount of product returns?
- How are product returns part of to the internet order fulfilment process?
6
2. Theoretical background
In this section, the sub questions as mentioned in the introduction, which give input to the main research question, will be explained. In the last paragraph, the research model and hypotheses will be presented. Research at the interface of logistics and marketing exists in the literature. For
example, it has been suggested that logistics academicians should investigate key questions that face managers and executives acting in distribution and direct marketing to customers (Rao et al., 2011). De et al, (2013) investigated the influence of product-oriented web technologies (such as zoom functions, alternative photos, colors etc.) and conclude that customers who see the product when it is worn by a model and customers using alternative photos (use of a torso) return more products. This research will focus on a new topic to investigate whether it has influence on returning products: product images. Moreover, this research will further investigate why consumers who use such a certain product web technology are returning more or less products.
2.1 The online storefront
In e-commerce, the website is the storefront and interaction between sellers and buyers. This interaction replaces the face-to-face communication between seller and buyers in traditional commerce (Park and Kim, 2003).
The online storefront is the point where the customers buy their products and where for that reason the reverse supply chain (that will be explained in 2.4) starts. When people buy more or less and in which quantities they return products influences the reverse supply chain. “The design of a website is the online consumers’ gateway to the organization” (Bontis and de Castro, 2000: 367) and is
therefore largely on user interface design. So, how does a company present and display their products at the online storefront? Due to the growing e-commerce the design of the online
storefront is changing and developing rapidly. The influence of product images on the design of the online storefront consequently will be explained in the next section.
7 2.5). Chen and Tan (2004) stated that a poorly designed website affects the visitors’ shopping
experience negatively; hence, in today’s competitive environment, a company needs a website that becomes preferred and accepted by the majority of the customers.
2.2 Influence of product images on behaviour of online shopping customers
On an online storefront, consumers can not physically inspect or try on an item prior to purchase. Therefore, an effective visual product presentation is critical for successful online business (Peck and Childers, 2003). Due to the growing importance of e-commerce, the design of the online storefront changes rapidly and consequently the web of presenting the products.
In this case of apparel, it is suggested (Kim and Kim, 2004; Fiore et al., 2005; Kim and Forsythe, 2009) that sensory-enabling presentation strategies that provide an analogous sense of tactile experience is an effective strategy to reduce perceived risk and increase the likelihood of pleasurable shopping experience. One of such a visual presentation method is the use of a model which is wearing the clothing or the use of a torso (with 2D visual enhancer). By providing a simulated product experience online, a retailer can evoke a more positive online shopping experience (Klein, 2003). Through visual presentation with a model presentation, it is possible to provide a sensory experience that enhances both a consumer’s visual perception and mental imagery while evaluating a product (Kim and Forsythe, 2009).
Then and Delong’s (1999) studied preferred apparel presented on a realistic human models online. They suggested that customers are more confident when shopping online for apparel if it is displayed three-dimensional form with the use of a model or mannequin. It has been suggested that by
providing high quality and interactive visual product information, retailers can reduce customers’ evaluation difficulty when shopping for apparel online (Fiore et al., 2005). So, by the presentation with a model the whole image can be determined and for that reason it can reduce customers’ evaluation difficulty.
8
2.3 Technology Acceptance Model
In this section, the model that will be used in this research will be described. In this research of investigating the influence of product-oriented web technologies (product images) online, the technology acceptance model (TAM) (Davis, 1989) is used.
Figure 1. Original Technology Acceptance Model (Davis, 1989)
The TAM model (Davis, 1989) (see figure 1) is used to understand the variables that effect online shopping (figure 1). In the information technology (IT) literature, technology acceptance models are among the most accepted theoretical frameworks for studying individual IT usage intentions (Ajzen et al., 2000; Davis, 1989; Davis et al., 1989; Venkatesh, 2003).Technology acceptance model is a model for examination of customers approval of online shopping (Stoel and Ha, 2009). Perceived ease of use (PEoU) and perceived usefulness (PU) are the two variables of TAM (Davis, 1989). PEoU stands for the degree to which a person believes that using a particular system would be free of effort, according to Visinescu (2014). PU is the degree to which a person believes that using a particular system would enhance his or her performance (Visinescu et al., 2014).
TAM posits that perceived ease of use and perceived usefulness influence individuals’ attitudes toward using a system, which in turn influences intentions to use the system (Davis, 1989; Straub et al., 2007). TAM also proposes that perceived ease of use influences perceived usefulness. Moreover, TAM was successfully applied to the online shopping environment, and perceived ease of use and perceived usefulness have both been found to be significant predictors of intentions to buy online (Klopping et al., 2004; Koufrais et al., 2002; Lee et al., 2005; Vijayasarathy, 2004).
So the extended TAM-model is used to understand the variables effecting online shopping. Besides, due to the growing importance of that online shopping and the rapid way of development in changes of the online storefront, the importance of the mentioned variables ease of use and usefulness have increased on websites (Elliot and Speck, 2005).
9 detail to see which variables will be higher scored for model or torso image presentation and which will more or less influence product returns.
Variable from extended TAM
Actual determinants
Perceived Ease of Use (PEoU)
Perceived info quality (PIQ) Perceived system quality (PSQ) Perceived Service Quality (PSvQ)
Perceived Usefulness (PU) Trust
Enjoyment
Table 1. Variable from extended TAM model with their actual determinants, applicable on this research
For the TAM variable Ease of Use (Davis, 1989) this paper will use three significant determinants of ease of use, namely ‘info quality’, ‘service quality’ and ‘system quality’ (Celik & Yilmaz, 2011; Stoel & Ha, 2009). These are the three factors that evaluate the quality level of a website according to Pavloe (2003). Shih (2004) divided that website quality level into three determinants: first, Shih (2004) considered perceived information quality (PIQ) as the output quality of an information system, then used it to represent information characteristics. We considered information quality to be the output. Thus it assessed using consumer perceptions of the quality of information on the web by online shopping. Second, perceived service quality (PSvQ) is a defining characteristic of
information systems (Shih, 2004). It consists of tangibles, reliability, responsiveness, assurance and empathy. Third, perceived system quality (PSQ) is defined as the transaction characteristic of an information system, supporting functions, which are measured as system quality (Shih, 2004). Stoel and Ha (2009) found that there are two determinants for the primary determinant of usefulness, namely ‘trust’ and ‘enjoyment’ (see table 1). They are predictive towards online shopping. They play considerable roles in consumers’ adoption of e-shopping, what is also found in previous research (Stoel and Ha, 2009). Pavlou (2003) found significance in his hypothesis “trust is in positive relation with e-commerce participation”. So, to test whether online customers are more or less likely to return products, this research takes these variables with their actual determinants.
The complete overview of the operationalised variables are given in the research model (2.5).
2.4 Reverse supply chain
10
reasons according to Jeszka (2014): profit margins are shrinking, sales managers are increasingly
more sensitive to the costs of maintaining inventory (including unsold goods); opportunity costs appear, as the possibilities to recover the value of the product returns. One way to improve the reverse supply chain, with respect to product returns, is to investigate whether product web technologies as product images at the online storefront influence the amount of product returns.
There are several different drivers for reverse logistics, such as customer preferences and shortening of product lifecycle (Carter & Ellram, 1998; Tan, Yu & Arun, 2002). This is especially true for online fashion retailers where apparel is characterized by rapid changes in collections. Therefore an agile and responsive reverse supply chain is needed to respond to that short product lifecycle and low predictability (Fernie et al., 2010; Masson et al., 2007).
Reverse logistics is a part of returns management (manage how the products will be returned) which in turn is a part of supply chain management (Mollenkopf and Closs, 2005). Activities involving reverse logistics are often reactive in nature instead of proactive which means it is often a result of a consumer or downstream channel member action and not a result of a planning decision of an organization (Tibben-Lembke and Rogers, 2002). However, organizations can behave proactively to avoid/handle reverse flow. Proactive management of reverse flow which has to result in a positive impact on the financial position of an organization (Langley et al., 2008). This positive impact on the financial position due to proactive management of the reverse flow will be investigated with the technology of product presentation at the online storefront. To look at the beginning of the reverse supply chain at the online storefront, where people buy the products, is an example of proactive management what has to result in positive impact on the financial position (Langley et al., 2008) due to, inter alia, the amount of product returns. The amount of product returns is the case in this
research, whether model or image product presentation have influence on product returns, and why.
As was mentioned, this research takes the TAM model to investigate how customers react, with respect to product returns, on model and image product web technologies. To be more specific and take into account the actual determinants from literature, this research wants to investigate how the model and image product presentation will be scored by the respondents. These scores, resulting from the survey, do investigate the preferences of the customers and if they meet their
requirements. This marketing part of the research has received increased attention in the marketing and in the supply chain literature as it reflects the ability of a firm within a channel of distribution to positively influence the relationship that it has with its customers (Horvath et al., 2005).
11 as a means of satisfaction and customer delight (Boyer & Hult 2005). How efficient an organization handles their reverse flow in the supply chain will have a powerful impact not just on costs, but also on revenue and customer goodwill. Research by Mermelstein (2006) shows that if the return management of products is convenient for customers, then they are more likely to shop, and if the return management is troublesome, they are not likely to shop there again (cited in Jack, Powers and Skinner, 2010). Organizations today cannot ignore the reverse flow of products and how they handle it because of volumes of returns are increasing world-wide (Stock et al., 2002). Moreover, as
mentioned, online fashion retail has to deal with a short product lifecycle. For that reason it is important that there do not go more than the necessary amount of products out of the company. Or the other way around, that not many products will come back in the company. When they come back, for example one week later, the products could be worthless due to the short product life cycle. And the process of picking, shipping and packing was a costly operation.This research is consistent with that by looking which product web technology (model versus torso image) meets customer requirements, how to deal with that technology and whether they can make a better feeling with the product, make a better choice and therefore return more or less products.
12
2.5 Research model and hypotheses
The research area is the online storefront as starting point for the purchase of products potentially leading to product returns. The data from the online storefront is available to answer if product images influence the amount of product returns. A survey will validate the data and will give input to the behavioural intention ‘why’ of product returns concerning the product web technology of product images (model versus torso images).
Figure 2 presents the complete conceptual model. At the left side is the product web technology presented that will be used for this research, product images with the use of a model and with the use of a torso image. These product images will be tested at the effect on product returns. This association will be mediated by the mentioned variables of the TAM model: ease of use and usefulness with their significant determinants according to the literature (H6 (+) H7 (+)).
Figure 2. Conceptual model RQ
13 that product. With the variable online buy intention from the TAM, this link can be made to conclude in the conclusive part of this research if this not investigated link in the literature is really more or less likely.
Using the determinants of the TAM variables ease of use and usefulness and linking those with the purchase intention linked with product returns, the conceptual model is well stated according to the literature.
The first seven hypotheses are proposed to check whether the use of a model presentation is
significant different than the use of a torso image. To refer to literature, the conceptual model is also drawn as described in literature. For that reason the first seven hypotheses are not drawn among each other but divided into two columns. In the first column the five actual determinants of the variables of the TAM model, namely ease of use and usefulness, are presented. Those determinants are explained in the second column. Moreover, to determine whether there is an influence on the dependent variable buy intention (to predict one variable (PU and PEoU) to another variable (PBI)) a regression analysis has to be made (Field, 2009). This is covered in the conceptual model.
In the next table all variables are operationalised (second column) from the existing literature (third column).
Construct Operational definition References
Perceived Info Quality (PIQ)
Output quality of an information system, represent info characteristics
Shih (2004)
Perceived System Quality (PSQ)
Transaction characteristics of an information system, supporting function of which are measured as system quality
Shih (2004)
Perceived Service Quality (PSvQ)
Characteristics of information system; tangibles, reliability, responsiveness, assurance and empathy
Shih (2004)
Trust The extent to which one believes that the new
technology usage will be reliable and credible
Mc Knight (2001)
Enjoyment The extent to which one believes that shopping will
provide reinforcement in its own right, going beyond performance consequences
Childers et al., (2001)
Perceived Usefulness (PU)
The degree to which an individual believes that using a particular system will enhance his or her behavioural intention
Davis (1989); Van der Heijden (2004) Perceived Ease of Use
(PEoU)
The degree to which an individual believes that using a particular system would be free from effort
Davis (1989); Van der Heijden (2004) Online buy intention
(BI)
Likelihood of return to web store and buy from this store
Van der Heijden (2004)
14 In table 2, the TAM variables ease of use and usefulness are presented, even though they are
operationalised by PIQ, PSQ and PSvQ, because these two variables will be analysed whether they influence buy intention for both the product images of torso and model presentation. In the
theoretical background was stated, that the actual determinants (PIQ, PSvQ, PSQ, trust & enjoyment) make it possible to go in more detail which variables will score higher for model or torso image presentation and which will more or less influence product returns in this research. Therefore the next seven hypotheses are formulated. For the first seven hypotheses we propose hypotheses to check whether the use of a model presentation is significantly different compared to the use of a torso image.
According to Shih (2004), that the PIQ is defined as output quality of an information system the completeness provided by a model presentation is likely to have a higher PIQ. Furthermore, interactive features, such as product images with the use of a model, support the information retrieval process (Roy et al., (2001) and Palmer (2002)).
H1: The use of a model as product image will positively influence perceived information quality compared to the use of a torso image in the product presentation.
Product presentation with a model is perceived to have a higher PSvQ because the presentation by a model, compared to a torso model, gives a more tangible and reliable feeling to the customer (Shih, 2004). It provides the customer with more personalized information.
H2: The use of a model as product image will positively influence perceived service quality compared to the use of a torso image in the product presentation.
The system quality is likely to be a bit higher with the use of a model compare to torso presentation due to the fact that a model presentation will provide more characteristics to the customer and is therefore more reliable (Shih, 2004).
H3: The use of a model as product image will positively influence perceived system quality compared to the use of a torso image in the product presentation.
15 model image) will provide the customer with a more reliable and trustworthy feeling.
H4: The use of a model as product image will positively influence trust compared to the use of a torso image in the product presentation.
Perceived enjoyment is also expected higher with the use of a model compared to the use of a torso image. To see a model wearing the particular jeans will provide a more complete image compared to a torso and so, according to Childers et al., (2001) shopping will provide reinforcement in its own right, going beyond performance consequences.
H5: The use of a model as product image will positively influence perceived enjoyment compared to the use of a torso image in the product presentation.
Due to Fiore et al., (2005), high quality and interactive visual product information such as product images with the use of a model can reduce consumers’ evaluation difficulty when shopping for apparel online. Therefore the ease of use, the degree to which a person believes that using a particular system would be free of effort, so the hypotheses will be:
H6: Websites with model presentation are perceived to have more ease of use than websites with torso images.
Product images with a more, a visualization with more level of undertanding according to Kaplan (1989), will enhance usefulness: the degree to which a customer believes that using a particular system would enhance his or her shop intention (Van der Heijden, 2004). Therefore:
H7: Websites with model product images are perceived to have higher usefulness than websites with torso images.
TAM was successfully applied to the online shopping environment, and perceived ease of use and perceived usefulness have both been found to be significant predictors of intentions to buy online (Klopping et al., 2004; Koufrais et al., 2002; Lee et al., 2005; Vijayasarathy, 2004).
This is also according to the conceptual model that these two variables are related to the
16 Therefore the next two hypotheses are formulated.
H8: The PU with a model presentation will positively influence buy intention online compared to the use of a torso presentation.
As mentioned in the theoretical background, it has been suggested that by providing high quality and interactive visual product information, retailers can reduce customers’ evaluation difficulty when they are shopping for apparel online (Fiore et al., 2005). So, by the presentation with a model the whole image can be determined and for that reason it can reduce customers’ evaluation difficulty and can therefore create a higher buy intention.
17
3. Methodology
This research acquired input to investigate whether the product web technology of product images have influence on the amount of product returns. This input came from the consortium company, a large online fashion retailer. This investigation involved a complemented survey structured to investigate consumer behaviour with respect to product returns and online shopping. This paper wants to investigate the specific product web technology of product images with a model and with the use of a torso. The data are from one product category, for both gender, namely ‘jeans’. The survey explains the ‘why’ why customers using the relevant product web technology (product images), returned more or less products. This is worked out with Excel and SPSS.
3.1 Data collection
The data collection process involved several steps. First, the consortium company was contacted to inform whether they would participate in providing data, helping in the survey and what they would expect. This is done with a face to face interview at the head office of the consortium company.
Second phase was to collect and analyse the data with orders and accompanying amount of product returns from the consortium company. This was done with Excel. The survey, to complement the dataset, was self-administered. According to the research question, this survey has to answer the behavioural aspect of why customers will return more or less products when they use one of the two investigated product web technologies (model and torso product image presentation). After ten days of sending requests, 75 responses were received, five of which were discarded for incompleteness. So 70 respondents completed the survey. The demographic characteristics of the sample are presented in table 7.
The survey was done with a within-subject method. This method let the respondent randomly show one of two investigated product image presentations (Charness, 2012). For each of the two product images the same questions were asked. The questions covered the variables and are from past research. In section 3.4 (table 3) the measurements and scales, who covered the variables and are used in the questionnaire, are presented.
3.2 Sample data
The data for this research came from the consortium company, a large online fashion retailer. On the product page of the investigated product group ‘jeans’ for boys and girls, the customer can use two different product web technologies on the focal product. Besides the zoom function, the customer is provided with the use of a model as product image presentation or the use of a torso as product image presentation.
18 January and 30th April. This dataset contains in total more than 16000 orders and these orders are from 145 unique style identities. In total 32 randomized style identities will be investigated, 16 style identities with model image presentation and 16 with torso product presentation. The number of 16 style identities is based on the fact that there were only 16 from the 145 style identities presented by a model. To keep the amount of investigated style identities the same, this research is using in total 32 different style identities. As mentioned, for the product group ‘jeans’ for boys and girls. For the specific order in that period, the dataset contains information regarding the transaction date, member ID, gender, group description (boys or girls), class identification (jeans), style identity, style description, colour description and whether or not the item was returned.
This dataset will be complemented by a survey. The survey consists of a sample of 75 (70 valid responses) what is sufficient since this research used a within-subjects or repeated measures method to analyse ‘why’ customers return products attached with questions from literature according to the variables from the conceptual model used in this research. The sample as used by similar research with fashion marketing on the web (Griffith et al., 2001) was 150 with a between subject experiment. Moreover in this survey, the majority of the respondents are students (88,4%), who are well-used as respondents in similar research.
3.3 Data analysis
The raw purchase and product returns information from the consortium company is analysed by using Microsoft Excel. This analysis shows if there is an influence of product images (model and torso images) on the amount of product returns, which is part of the research question. This is done by calculating the mean of the product returns, standard deviation and research units (the mentioned 16 for each product image). The mean is calculated for the randomized chosen 16 style identities for each product presentation. This research did not calculate the average of the percentages because in that case the style identities with a low amount of orders have the same impact as those with a high amount of orders. This affects the view wrongly (Field, 2009). A t-test is executed in Excel to
determine if there is a significance difference for the two total averages in the amount of product returns.
19 So, Microsoft Excel and SPSS will provide the raw purchase and product information data (as
explained in the theoretical background) into useful information as the influence of the web technology on the amount of product returns and why the customer returned that product.
The first seven hypotheses, in which will be investigated what the direct influence is of a model or torso image presentation on the behaviour of the online customer, are conducted with a Wilcoxon Signed Rank test. The reasons for this test are the ordinal data due to the Likert-scale and the related groups due to the within-subject method (Field, 2009). The Wilcoxon Signed Ranks test is applied instead of one-sample t-test when the normality assumption is not met, as in this case with ordinal data (Chan, 2003).
The rank table gives some interesting information, to test the rank given by the respondent on the particular product image on the specific variable (Field, 2009) (Appendix B). Due to the hypotheses in this research which expect higher scores on the variables for the product image presentation of a model, this Wilcoxon Signed Rank test gives a negative result when the score on a question with the model image presentation is lower than the score on the torso image presentation. This is due to the Likert-scale of the questionnaire that range from “strongly disagree” to “strongly agree” (Field, 2009). The test provides a positive result when the score of the model image is higher than the score on the torso image presentation (Appendix B).
By examining the final test statistics table we can discover whether these negative or positive
differences in the rank table are significant or not (Field, 2009) (Appendix B). The Z-score in this table, based on the positive ranks of the rank table, corresponds with a p-value. This p-value is used later with the hypothesis testing to determine whether hypothesis proposed are supported or not supported.
The hypotheses 8 and 9 are conducted with a regression analysis as for this part of the conceptual model the research aims to predict one variable (PU and PEoU) to another variable (PBI). Therefore a reliability analysis is conducted (Field, 2009).
3.4 Measurement instruments and scales
Each item of the construct was measured based on a five-point Likert scale from “strongly disagree” to “strongly agree” in the questionnaire.
20
Variables Items References
Perceived Info Quality 1. The information provided about the product is accurate.
2. The product provides me with a complete set of information.
Davis (1989) and Gefen et al., (2003)
Perceived System Quality
1. The website operates reliable.
2. The information is readily accessible to me.
Shih (2004) Perceived Service
Quality
1. The product as displayed provides me with a confident feeling.
2. Product gives personalized information.
Shih(2004)
Trust 1. I trust the way the product is displayed, the
information about the product to be true
2. The way the product is displayed is trustworthy.
Gefen et al., (2003)
Perceived Enjoyment 1. Using the way this product is displayed would provide
me a lot of enjoyment.
2. When products are displayed in this way, purchasing will be more interesting.
Moon and Kim, (2001)
Perceived Usefulness 1. The internet would improve my performance when
searching for and purchasing goods.
2. The internet will makes it easier to search for and purchase goods.
3. The internet will probably enhance my effectiveness in goods searching and purchasing.
4. Using internet to acquire a product would allow me to do my shopping more quickly.
5. The internet will increase my productivity when searching for and purchasing goods.
Davis (1989) and Gefen et al., (2003)
Perceived Ease of Use 1. When products are displayed this way it will make it easy for me to buy a product.
2. This way of displaying products will be clear and understandable for me.
3. This way of displaying products will provide me to do shopping these products more easy
Davis (1989) and Gefen et al., (2003)
Online buy intention 1. When products are displayed in this way I intend to use frequently to shop.
2. If I could, I would like to continue purchasing when products are displayed as here.
Moon and Kim, (2001)
Table 2. Measures and scales
The survey instrument of this research addresses two major purposes: first is to examine the
21 already used questions in particular questionnaires. The scales of the study were adopted from the earlier literature.
3.5 Reliability Analysis
Cronbach’s Alpha was used to calculate the reliability of the constructs that measured the different variables in this research. Cronbach’s alpha is a method to check the degree to which the different underlying items are well enough related to be combined in a single variable (Ponsioen, 2015). According to Ponsioen (2015) and Field (2009), it is generally accepted that a Cronbach’s alpha of 0,7 and higher indicated sufficient internal consistency between the items. And the scale should be at least 3 items. For that reason, some determinants are combined due to their 2 items (table 3).
There was only one variable which does not reach the level of .70 or higher, but when this variable is combined with the first experiment the overall score reached the level of .70. Moreover, the
variables website quality (PIQ, PSvQ and PSQ) and Trust and Enjoyment are not taken into account in the regression analysis. Table 1 shows the reliability scores for all the constructs (α) and the number of items in the construct (N) for the two experiments.
The measured variables perceived ease of use and perceived usefulness, all on a Likert-scale from 1 (strongly disagree) to 5 (strongly agree) are highly reliable.
Due to the fact that the number of items of the variable online buy intention is 2, this variable is neglected in the reliability analysis where the number of items have to be 3 or more (Field, 2009). This minimum level of items was also the reason to combine the first three variables (PIQ, PSvQ and PSQ) into one variable ‘website quality’. This is consistent with literature (Shih, 2004).
Scales Items (N) Cronbach Alpha (α)
Website Quality I* 6 .888
Website Quality II* 6 .834
Trust&Enjoyment I* 4 .851
Trust&Enjoyment II* 4 .652
Perceived Usefulness I* 4 .908
Perceived Usefulness II* 4 .844
Perceived Ease of Use I* 3 .924
Perceived Ease of Use II* 3 .792
Table 4. reliability of measurement instruments separately for presentation with model and without model *I: without use of a model
22
4. Results
In this section the results will be presented.
4.1 Product returns from the consortium partner
The outcomes resulting from the data file with product returns from the consortium partner, are presented.
From the 145 unique style identities, 32 are randomly selected (16 different product images with a model and 16 different product images with a torso). So in total, 22% of the style identities is investigated.
The amount of orders from the 16 different products for each presentation was higher for the product image presentation with a torso (see table 5). However, the amount of orders that came back was also higher. To compare these values, the average product returns percentage is calculated in table 5. From these 16 different style identities for each presentation at the online storefront the average product returns percentage is higher for the model image presentation compare to torso image presentation.
The detailed calculation can be seen in appendix B.
Product image presentation ordered returned Average product returns (%)
Model 1492 226 15,1
Torso 3465 507 14,6
Table 5. Amount of orders with amount of product returns for each product image presentation
According to Field (2009), to calculate whether the means are significant different, a one sample t-test is made. This is done in Excel. The outcomes are presented in table 6.
Torso Model Mean 0,1463 0,1513 Variance 0,0079 0,0082 Observations 16 16 Hypothesized Mean Difference 0 df 30 t Stat -5,000 t Critical one-tail -2,457
24
4.2 Profile of the respondents
Personal and demographic information of the respondents of the survey such as gender, age, hours shopping online on apparel, if they return products and the main reasons to return products when shopping online (table 7).
Variable Category Frequency (N) Percentage (%)
Gender Male Female 37 33 52,9 47,1 Age < 15 Years 15-25 Years 25-35 Years 35-45 Years > 45 Years 0 61 6 0 2 0,0 88,4 8,7 0,0 2,9 Hours shopping online
on apparel
< 1 hour per month 1-2 hours per month 2-5 hours per month 5-10 hours per month > 10 hours per month
31 26 8 5 1 43,7 36,6 11,3 7,0 1,4 How often do you
return products then
Never Sometimes Often Always Not the case
8 44 13 1 5 11,3 62,0 18,3 1,4 7,0 Reasons of product returns
Wrong size, I will order another size
Wrong size, I will not order another size
Other size than I ordered I ordered different sizes and kept the right size
I ordered different sizes and sent them all back
Product did not fit
Other color than I expected Other materials than I expected Product was damaged
42 6 5 23 33 3 10 22 5 0 28,0 4,0 3,3 15,3 22,0 2,0 6,7 14,7 3,3 0,0
25
4.3 Hypothesis testing
In this section the results of the hypotheses from the research model will be presented.
4.3.1. Model vs Torso image presentation PIQ, PSvQ, PSQ, trust, enjoyment, PEoU and
PU
The first seven hypotheses are conducted, as mentioned in the methodology part, with the Wilcoxon signed rank test.
The following table (table 8) shows the mean score on the combined questions from the
questionnaire that covers the particular variable. The last column shows the p-value, so whether there is a significant difference or not, according to a significance level of 0.01.
Variable Model (mean) Torso (mean) Difference (mean) p-value Results
PIQ 3,85 3,52 0,33 0.011 Not supported
PSQ 3,97 3,50 0,47 <0.01 Supported PSvQ 3,69 2,70 0,99 <0.01 Supported Trust 3,94 3,51 0,43 <0.01 Supported Enjoyment 3,62 3,04 0,58 <0.01 Supported PU 3,71 3,05 0,66 <0.01 Supported PEoU 3,94 3,51 0,43 <0.01 Supported
Table 8. mean scores on each of the variables for model and torso presentation
To give insight which mean rank is scored for each of the variables, so the comparison of the different product image presentations, the next table (table 9) shows what the difference in mean rank is. The last column is the same as in table 8, it shows the p-value, so whether there is a significant difference or not, according to a significance level of 0.01. This research used a
significance level of 0.01 because that was also the level used in similar research (De et al., 2013) and this stringent level gives a particular confident result (Field, 2009).
Variable Model (mean rank) Torso (mean rank) Difference (mean rank) p-value Results
PIQ 24,76 25,61 -0,85 0.011 Not supported
PSQ 21,32 14,88 6,44 <0.01 Supported PSvQ 28,62 13,78 14,84 <0.01 Supported Trust 21,53 14,06 7,47 <0.01 Supported Enjoyment 34,67 14,29 20,38 <0.01 Supported PU 27,30 17,28 10,02 <0.01 Supported PEoU 29,94 13,63 16,31 <0.01 Supported
26 With the use of a significance level of 1% the only variable which is not significant in change is the variable of perceived information quality. The Wilcoxon signed rank test showed that the score of PIQ for the model image presentation and the torso image presentation did not elicit a statistically significant change in the score for PIQ (Z= -2,554, p=0,011). Indeed, median score for both model image presentation and torso image presentation was 4.00. The mean rank was moreover quite the same (M=25,61 with model and M=24,76 without model, table 9).
For all the other variables, the Wilcoxon signed rank test showed that the scores for model image presentation compare to the torso image presentation are significant higher (p=0,000). Their p-value is <0.01.
4.3.2 Perceived ease of use on online buy intention
Hypothesis Model Variables
β S.E R square p Results
H8 PEoU BI with torso 0,77 0,09 0,529 0.000 Supported H8 PEoU BI with model 0,65 0,16 0,450 0.000 Supported
Table 10. Regression results perceived ease of use on online buy intention
There was a significant positive relationship between perceived ease of use and online buy intention with the use of a torso for the product presentation, B=.77, p<0.001. This means that an increase in Perceived Ease of Use with the use of a torso image for the product presentation will lead to an increase in buy intention online by 0.77. This increase is significant (table 10).
27
4.3.3 Perceived Usefulness on online buy intention
Hypothesis Model Variables
β S.E R square p Results
H9 PU BI with torso 0,89 0,07 0,714 0.000 Supported H9 PU BI with model 1,08 0,09 0.645 0.000 Supported
Table 11. Regression results perceived usefulness on online buy intention
There was a significant positive relationship between perceived usefulness and online buy intention with the use of a torso for the product presentation, B=.89, p<0.001. This means that an increase in PU with the product presentation with a torso image will lead to an increase in buy intention online by 0.89. This increase is significant (table 11).
28
5. Discussion
This paper investigated the relation between product-oriented web technologies and the costly operation of products returns using the technology acceptance model (TAM) (Davis, 1989), a dataset with more than 16000 orders and a survey. This research takes into account the extended variables of that TAM model, namely ease of use and usefulness. Further, this paper identifies the actual determinants of those variables with respect to online shopping.
Where literature supports that visual web technologies may positively affect consumers’ experience and buy intention (Fiore et al., 2005; De et al., 2013), this research investigated if this really was the case for a large clothing company taking into account the amount of product returns. In the reverse supply chain, this paper gets this subject at the core of the problem: at the online storefront, how does that influence the number of product returns and thus the entire reverse supply chain? The online storefront is the point where people start buying and where, with respect to product returns, the reverse supply chain starts. In this research a survey investigated why customers are more likely to return more or less products.
The effect on buy intention, from the data file of the consortium company with more than 16000 orders, proved to be positive for the torso image presentation on the online storefront. The amount of orders was much higher compared to the products presented by a model. However, the jeans presented by a model were the more exclusive and alternative jeans compared to the other more basic jeans. So, another option could be to test the two technologies within one style identity. However, this was not possible with the submitted data from the consortium company. The reason for the consortium company to use that particular technology for the apparel with or without model presentation is unknown. But the online buy intention in the survey was higher scored by the respondents for the model presentation compared to the torso presentation (M1= 3,58, M2= 2,88). Moreover, from the calculations and t-test with the data file, the amount of product returns were higher for the model image presentations.
29 This research took seven variables which were tested on the way of product presentation at the online storefront, as mentioned model and torso presentation. The biggest difference by far per variable of the mean score is for the variable perceived service quality. So the respondents scored higher at model presentation on confident feeling and the way the product gives them personalized information, compared to torso presentation. This is in line with Lee et al (2005) who investigated that product images presented by a model gave the customer a better feeling and more interactivity. For the mean rank the strongest effect on attitude of customers with a model presentation and a torso presentation is perceived enjoyment. So, the respondents ranked enjoyment and purchase interest higher for a model presentation compared to torso image presentation. This is in line with Childers et al (2001) findings who also stated that enjoyment and purchase interest were higher scored in a hedonic environment, with its more pleasure presentation.
Where perceived usefulness has a higher beta for model presentation, what implicates a higher increase in buy intention online for model presentation, for perceived ease of use the beta is lower for model presentation. In both regressions the R square is higher for torso images which means that the total variability in buy intention online can be explained by larger amounts for the torso image presentation.
5.1 Limitations
There were several limitations in this research. First, The sample of the survey consisted almost completely of students (88,4%). Even though students are well-used in similar research (Griffith et al., 2001; De et al., 2013), in this research for the investigated product group ‘jeans’ for boys and girls it is more likely that these jeans will be bought by older parents (Guldemond; GfK Panel Services, 2011). Besides the sample, the product group ‘jeans’ is also a limitation for the experience. Customers could react different on different product groups (in this case of apparel: jeans, shirts, accessories). Further, the jeans presented by a model and the jeans presented by a torso were not the same jeans. The jeans presented by a model were more exclusive and alternative jeans, as mentioned in the discussion. The use of data from one company is an acknowledged limitation of the present study. Although results at the retailer level of analysis showed no divergence from those at the customer level of analysis, this research should be replicated using a broader sample of Internet retailers to confirm the robustness of the findings of this research.
30 (Charness, 2011) but it can also lead to a change in behaviour by respondents when they see the other product image presentation (Charness, 2011) and therefore affect the results.
Due to the fact that this research wants to investigate a difference in product returns between product web technologies, there are two different technologies used which are used by the consortium partner. Another limitation is therefore that there are various other product presentations available on (apparel) product websites.
Another limitation is the way the survey is conducted. The survey was made by the sample who are not attached with the previous orders from the data file used from the consortium partner. They made it apart from previous orders at home or somewhere else. So, there is a limited experiential and emotional response compare to a survey that was conducted with customer who really bought that product.
31
6. Conclusion
This paper investigated whether product images (model versus torso images) have influence on the online buying behaviour and the amount of product returns for online fashion. There is an influence in product returns when products are presented at the online storefront with product images with the use of a model compared to the use of torso image presentation. According to the data file from the consortium company, customers are less likely to return products when torso product images are used to present a product at the online storefront.
According to the survey, contrarily, customers are less likely to return products when at the online storefront model product images are used. With model product image presentation, they indicated that they are better informed about the product and therefore can make a better choice and are less likely to return products. In this research only one of the investigated seven variables was not significant in difference between a model and torso image presentation, namely perceived info quality. All the other variables scored higher for the model image presentation, and these scores were significant.
Where both the hypotheses 8 and 9 are significantly supported, there is a difference between the ease of use with model and torso presentation. A model presentation is not more easy to use compared to the torso presentation, so we reject H8 in this research in which was stated that a model presentation would have a higher score on ease of use compared to torso presentation. A model presentation provides the customer particularly with more interactivity and a better feeling but it results apart from a deliberate choice (and that they are less likely to return that product) also in a harder choice to make in buying that product.
In order to optimize the internet order fulfilment process, and specifically in this research the reverse supply chain from the online storefront point of view, finding a well-chosen product web technology can be beneficial in costs related to the product returns. Moreover it can be beneficial for the short product life cycle of apparel that could have a lower value during that period.
33
7. References
Ajzen, J. (1991). The theory of planned behaviour, Organ. Behav. Hum. Decis. Process. 50: 179–211. Bontis, N., & De Castro, A. (2000). The 1st World Congress on the Management of Electronic Commerce: review and commentary. Internet Research: Electronic Networking Applications and Policy, 10(5): 365-373.
Boyer, K.K., & Hult, T.G. (2005). Customer Behaviour in Online Ordering: A Decision Scoring Model. Decision Sciences, 36(4): 569–598.
Calisir, F., Bayraktaroglu, E., Gumussoy, C., Topcu, I., & Mutlu, T. (2009). The relative importance of usability and functionality factor for online auction and shopping web sites. Emerald, 420-439.
Carter, R. C., & Ellram, M. L. (1998). Reverse logistics: a review of the literature and framework for future investigation. Journal of Business Logistics, 19(1): 85-102.
CBS (2014).
http://www.cbs.nl/nl-NL/menu/themas/vrije-tijd-cultuur/publicaties/artikelen/archief/2014/2014-4076-wm.htm. Retrieved on December 23, 2014. Çelik, H. E., & Yilmaz, V. (2006). Extending the technology acceptance model for adoption of e-shopping by consumers in Turkey. Journal of Electronic Commerce Research, 12(2): 152-164.
Chan, Y.H. (2003). Biostatistics 10: Quantitative Data – Parametric & Non-parametric Tests, Singapore Med J , 44(8): 391-396.
Charness, G., Gneezy, U., & Kuhn, M.A. (2012). Experimental methods: between-subject and within-subject design. Journal of Economic Behaviour & Organization, 81(1): 1-8.
Chen, L., & Tan, J. (2004). Technology adaptation in e-commerce: key determinants of virtual stores acceptance. European Management Journal, 22(1): 74-86.
Childers, T.L., Carr, C.L., Peck, J., & Carson, S. (2001). Hedonic and utilitarian motivations for online retail shopping behaviour. Journal of Retailing, 77(4): 511-535.
Davis, F.D. Perceived usefulness, perceived ease of use, and user acceptance of information technology, MIS Q. 13, 1989, pp. 319–340.
Davis, F.D., Bagozzi, R.P., & Warshaw, P.R. (1989). User acceptance of computer technology: a comparison of two theoretical models. Management Science, 35: 982–1003.
De, P. (2013). Product-Oriented Web Technologies and Product Returns : An Exploratory Study. Elliott, M. & Speck, P. S. (2005). Factors that affect attitude toward a retail website. Journal of Marketing, 73(1): 40-51.
Eroglu, S., Machleit, K., & Davis, L. (2003). Emprical testing of a model of online store atmospherics and shopper responses. Journal of Psychology, 139-150.
34 Fiore A.M., Kim J., & Lee H.H. (2005). Effect of image interactivity technology on consumer responses toward the online retailer. Journal of Interactive Marketing, 19(3): 38–53.
Gefen, D., Karahanna E., & Straub, D.W. (2003). Trust and TAM in Online Shopping: An integrated model. MIS Quarterly, 27(1): 51-90.
Griffith, D.A., Krampf, R.F., & Palmer, J.W. (2001). The role of interface in electronic commerce: consumer involvement with print versus online catalogs. International Journal of Electronic Commerce, 5(4): 135-153.
Guldemond, E.
(2011).http://www.textilia.nl/nlL/Nieuws/onderzoek_steeds_meer_ouders_kopen_kinderkleding_on line-110630164400#.VbYmovkXxdw
Horvath, P.A., Autry, C.W., & Wilcox, W.E. (2005). Liquidity implications of reverse logistics for retailers. Journal of Retailing, 81(3): 191-205.
Hunter, S.D. (2003) Information technology, organizational learning and the market value of the firm. Journal of Information Technology and Application, 5(1): 1-28
Jack, P.E., Powers, L.T., & Skinner, L. (2010). Reverse logistics capabilities: antecedents and cost savings. International Journal of Physical Distribution & Logistics Management, 40(3): 228-246.
Jayarman, V., Ross, A., & Agarwal, A. (2008). Role of information technology and collaboration in reverse logistics supply chain. International Journal of Logistics Research and Applications, 11(6): 409-425.
Jeszka, A.M. (2014) Product returns management in the clothing industry in Poland. Scientific Journal of Logistics, 10(4): 433-443.
Kaplan, R., Kaplan, S., & Brown, T. (1989). Environmental preference: a comparison of four domains of predictors. Environment Behaviour, (21): 509–530.
Kim J., & Forsythe S. (2009). Adoption of sensory enabling technology for online apparel shopping. European Journal of Marketing, 43(9/10): 1101–1120.
Kim, E.Y., & Kim Y.K. (2004). Predicting online purchase intentions for clothing products. European Journal of Marketing, 38(7): 883–897.
Klein, L.R. (2003). Creating virtual product experiences: the role of telepresence. Journal of Interactive Marketing, 17(1): 41–55.
Klopping, I.M., & McKinney, E. (2004). Extending the technology acceptance model and the task-technology fit model to consumer e-commerce. Journal of Information Technology Learning Performance, 22: 35–48.
Koufrais, M. (2002). Applying the technology acceptance model and flow theory to online consumer behaviour, Journal of Information Systems Res, 13: 205–223.
Langley, C. J., Coyle, J. J., Gibson, J. B., Novack, A. R., & Bardi, J. E. (2008). Managing Supply
Chains - a logistics approach (8th ed.). Kentucky: South Western College.
Limayem, M., Frini, A., & Khalifa, M. (2000). What makes customers buy from internet? A
35 Lee, G.G., & Lin, H.F. (2005). Customer perceptions of e-service quality in online shopping. Journal of Retail and Distribution Management, 33(2): 161–176.
Lee, H.H., Fiore, A.M., & Kim, J. (2005). The role of the technology acceptance model in explaining effects of image interactivity technology on consumer responses. International Journal of Retail Distribution Management, 34: 621–644.
Masson, R., Iosif, L., MacKerron, G., & Fernie, J. (2007). Managing complexity in agile global fashion industry supply chains. The International Journal of Logistics Management, 18(2), 238-254.
Mc Knight, D.H., & Chervany N.L. (2001). What trust means in e-commerce customer relationships: an interdisciplinary conceptual typology. International Journal of Electronic Commerce, 6(2): 35–60.
Mollenkopf, A.D., & Closs, J.D. (2005). The hidden value in reverse logistics. Supply Chain
Management Review, 9(5), 34-43.
Mollenkopf, D. A., Rabinovich, E., Laseter, T. M., & Boyer, K. K. (2007). Managing Internet Product Returns: A Focus on Effective Service Operations, 38(2): 215–250.
Monsuwe, T.P.Y., Dellaert, B.G.C., & Ruyter, K.D. (2004). What derives consumers to shop online? A literature review. International Journal of Service Industry Management, 15(1): 102-121.
Palmer, J.W. (2002). Web site usability, design, and performance metrics. Information Systems Research, 13(2): 151-67.
Park, C. and Kim, Y. (2003). Identifying key factors affecting consumer purchase behaviour in an online shopping context. International Journal of Retail & Distribution Management, 31(1): 16-29. Pavlou, P.A. (2003) Consumer acceptance of electronic commerce: Integrating trust and risk with the Technology Acceptance Model. International Journal of Electronic Commerce, 7(3): 101-134. Peck J., & Childers T.L. (2003). To have and to hold: the influence of haptic information on product judgments. Journal of Marketing, 67(2): 35–48.
Ponsioen, S. (2015). Reliability analysis. EBM766B20.2014-2015.2 University of Groningen, Groningen. 19 January.
Ranganathan, C., & Ganapathy, S. (2002). Key dimensions of business-to-consumer websites. Information and Management, 39:.457–465.
Rao, S., Goldsby, T.J., Griffis, S.E., & Iyengar, D. 2011.Electronic Logistics Service Quality (e-LSQ): Its Impact on the Customer’s Purchase Satisfaction and Retention. Journal of Business Logistics, 30(2):16 7–79.
Roy, M.C., Dewit, O., & Aubert, B.A. (2001). The impact of interface usability on trust in webretailers. Internet Research: Electronic Networking Applications and Policy, 11(5): 388-398.
Shih, H.P. (2004). An empirical study on predicting user acceptance of e-shopping on the Web. Information & Management, 41(3): 351-368.
36 Stock, J., Speh, T., & Shear, H. (2006). Managing Product Returns for Competitive Advantage. Sloan Management Review association, 48(1): 57.
Stoel, L., & Ha, S. (2009). Consumer e-shopping acceptance: antecedents in a TAM. Journal of Business Research, 62(5): 565-571.
Straub, D.W., Burton-Jones, A., & Veni, V. (2007). Breaking the TAM Logjam, Journal of Assoc. Information System, 8: 223–229.
Tan, A.W.K., Yu, W.S., & Arun, K. (2002). Improving the performance of a computer company in supporting its reverse logistics operations in the Asia-Pacific region. International Journal of
Physical Distribution & Logistics Management, 33(1), 59-74.
Then, N. K. and Delong, M. R., “Apparel Shopping on the Web,” Journal of Family and Consumer Sciences, Vol. 91, pp. 65-68, 1999.
Tibben-Lembke, S. R., & Rogers, S. R. (2002). Differences between forward and reverse logistics.
Supply Chain Management: An International Journal, 7(5), 271-282.
Van, der, Heijden, H. (2004). User acceptance of hedonic information system. MIS Quarterly, 28(4): 695-704.
Venkatesh, V., Morris, M.G., Davis, G.B., & F.D. Davis. User acceptance of information technology: toward a unified view, MIS Q. 27(3): 425–478.
Visinescu, L., Sidirova, A., Jones, M., & Prybutok, V. (2014). The influence of website dimensionality on customer experiences, perceptions and behavioural intentions: An exploration of 2D vs. 3D. Information & Management, 52: 1-17.
37
Appendix A
In this appendix, the Dutch questionnaire made in Qualtrics is presented. These questions were asked for both images (model and torso image presentation). Q1-19 are formulated with a vife-point Likert scale (1= strongly disagree, 5=strongly agree).
Online fashion
I1. Wat is uw geslacht? man (1) vrouw (2)
I2. Wat is uw leeftijd in jaren?
(1) 15-25 (2) 25-35 (3) 35-45 (4) >45 (5)
I3. Hoeveel uren shopt u online naar kleding per maand?
minder dan 1 uur per maand (1) 1-2 uur per maand (2) 2-5 uur per maand (3) 5-10 uur per maand (4) 10 uur of meer per maand (5)
I4. Als u kleding koopt online, retourneert u dan wel eens de producten? nooit (1) soms (2) vaak (3) altijd (4) niet van toepassing (5)
Q1 De informatie over dit product is juist gepresenteerd
Q2 Het product zoals hier gepresenteerd geeft mij een complete set informatie Q3 De productpresentatie ziet er betrouwbaar uit om mee te werken
Q4 De informatie is gemakkelijk toegankelijk voor mij met deze productpresentatie Q5 Het product zoals weergegeven geeft mij een zelfverzekerd gevoel
Q6 Deze productpresentatie geeft mij gepersonaliseerde informatie
Q7 Ik vertrouw erop dat, met deze productpresentatie, de informatie waar is Q8 Manier waarop dit product wordt weergegeven is betrouwbaar
Q9 Ik ervaar waargenomen plezier met online shoppen als het product op deze manier wordt gepresenteerd
Q10 Producten op deze manier presenteren maakt online shoppen interessanter Q11 Met deze productpresentatie zal ik vaker gebruik maken van online winkelen
Q12 Als ik kon, zou ik graag blijven inkopen wanneer producten worden weergegeven als hier Q13 Deze productpresentatie zou mijn presentaties in de zoektocht naar deze producten verbeteren Q14 Deze productpresentatie zal de zoektocht naar het juiste product en de aankoop ervan
vergemakkelijken
Q15 Het zal mijn effectiviteit bij het vinden en de aankoop van deze producten te verbeteren Q16 Met deze productpresentatie zal ik sneller shoppen online.
Q17 Deze productpresentatie maakt het mij gemakkelijk om een product te kopen Q18 Deze productpresentatie is begrijpelijk en duidelijk voor mij
38
Appendix B
In this appendix the data file from Excel and the output from SPSS is shown.
Datafile how the amount of product returns is calculated
39 Mean, median, minium and maxium scores per variable
Z-scores and corresponding significance level per variable with and without model presentation
40 one example of regression output (PEoU vs BI with model)
