The Laboratory Experiment Design

Published Date: 02 Nov 2017

Similarly, online purchase intention was measured from items adopted from Verhagen and Dollen (2009), following the same argumementation of word of mouth intentions. Items and codings are showed in Table 4.13.

In order to investigate whether the influence of store layout on the fourteen dependent variables depends on telepresence, hedonic and utilitarian shopping motivation, a Mixed-design (Split-Plot) Analysis of Variance was adopted.

Hedonic and utilitarian shopping motivation are considered moderators of the research model. In order to measure hedonic motivations, eighteen items were adpopted from the Arnolds and Reynolds (2003) study, and the four items used to measure utilitarian motivations were drawn from Kang and Park-Poaps (2010), as adopted from Babin, Darden, and Griffin (1994).

Telepresence was measured modifying the six relevant items for the 3D online context, that appeared in the Novak, Hoffman and Yang’s (2000) study. They modified the items of telepresence from the study of Kim and Biocca (1997).

In the first part of the questionnaire, participants were asked about their internet experience and the purpose of internet use, in order to provide more insightful results in terms of the sample’s demographic characteristics. They were asked to provide their experience both in 3D online environments and 2D online environments. In addition, they we asked to provide information regarding the kind of products they buy through 3D online stores. The items of purpose of internet use construct were drawn from O’Keefe et al. (2000), and internet experience items from Kevork (2011).

4.3.1.1. Realism check

This study follows an experimental research design approach. Participants of the main study were invited to watch a video of each store layout type and then answer questions about the store. As a result, participants did not really interact with the stores, but were asked to think of an environment where things described in the experiment really happened. To investigate the realism of this experimental design, a realism check with two items was included in the third part of the questionnaire. The items of the realism check were drawn from Wagner et al.’s (2009) study, as they adopted them from Darley and Lim (1993).

4.3.2. Laboratory Experiment Design

This section is dedicated to presenting the phases of the laboratory experiment. The stages of a laboratory experiment design are presented in the following figure (Labwriter 2004).

A within-subjects design was considered most appropriate for five reasons:

An increase in the number of participants leads to an increase of the statistical power; thus, a research design that involves more participants is desirable. At the same time, the experiment is quite large and would require a lot of effort of the respondents’ part. The pilot-test of the experiment showed that participants needed between two and two and a half hours to complete the experiment. In this regard, it is probable that the sample size would not be large enough for a between-subjects design. For example, as there are five treatments in this experiment, if a between-groups design was selected, five times as many participants would be required compared to those needed for a within-groups design.

Due to the long duration of the experiment (i.e., two and a half hours approximately), were unlikely to be recruited within a reasonable time period.

Taking into consideration decrease of variance, as MacKenzie (2012) considers that "the variability in measurements is more likely due to differences among conditions than to behavioural differences between participants".

Avoidance of group differences that take place in a between-groups design.

In order to minimize, or even eliminate, the order or learning effects of the participants, which is a disadvantage of within-group design, the researcher was determining the order of the various treatments that each participant was exposed to. Each participant of the sample has been exposed to a different order of treatments.

The sampling frame of the experiment consisted of undergraduate and postgraduate students of two universities in Greece. More information about the sampling characteristics is provided in the following section (i.e., Sampling Design). A cover letter was developed and was sent to potential respondents in these universities. This cover letter provided information regarding the purpose of the experiment, the stages involved and the processes taking place during each stage, the marketing ethics considered by the research design (e.g., that the information provided will not be related to individual responses), motives for their contribution, and contact information of the researcher. Since all participants were from Greece, the Greek language was used in the experiment. A copy of the cover letter both in English and Greek is included in Appendix C1. The stages of the experiment are presented in the following paragraph.

First Stage of the Experiment

The students that were willing to participate had to send a confirmatory e-mail to the researcher. Then, the researcher contacted each participant in order to arrange a meeting at predefined time intervals. The experiment took place in the researcher’s office in Room 907, 47A Evelpidon & 33 Lefkados Street, Athens. Luck and Rubin (1987) emphasize random participants’ selection to avoid learning effects and increase reliability. In this experiment all participants were randomly selected and at the beginning of the experiment were provided with a list with various combinations of the treatments, sequence in order, to select one. Through this process participants were selecting the order of the layouts that would be presented to them. In this regard, the researcher did not follow a predefined order. Participants were randomly deciding the order as the store layout types were coded (i.e., store#1, store#2 etc.). Also, participants could not select an order that had been chosen by another participant. This served as insurance that a considerable number of participants would not select the same order. For example, the coding of stores could drive the participants for selecting the order: store#1, store#2, store#3, store#4, and store #5. Finally, all participants while they were taking part in the experiment were sitting in the same desk and were provided with the same laptop, in order to eliminate any extraneous factors.

Second Stage of the Experiment

In the second stage, respondents were asked to fill in the first part of the questionnaire. This part included questions such as the purpose of internet use, the purpose of 3D online environments use, the products that they buy from 3D online environments, the shopping motivation, and the degree of telepresence in 3D online environments (Appendix C.2).

Third Stage of the Experiment

In the third stage of the study, the researcher of the study provided a video, and a description of the layout/design of a store to each participant. The participant was invited to review the design of the store and evaluate the characteristics of the store in the appropriate part (i.e., Part B) of the questionnaire. Also, in the instructional leaflet participants were encouraged to concentrate only on the store design/layout and not on any other external factors such as prices and design of the products. There are five stores, thus this process (i.e., Stage#2) followed a five times rotation. The second part of the questionnaire is included in Appendix C.3, both in English and Greek.

Fourth Stage of the Experiment

In the final stage of the Laboratory Experiment the respondents were invited to fill in the final part of the questionnaire. There is evidence that the demographic characteristics of a sample can explain behavioral patterns in marketing research issues (Crisp et al. 2007). The third part of the questionnaire was consisted of questions related to the demographic data of the respondents. The demographic data were descriptively analyzed both to shed light on complicated issues regarding the results of the laboratory experiment, and to provide managerial implications in terms of the sample characteristics. The researcher had been in the same room with the participants throughout the experiment, and was willing to address any issues or difficulties faced by the respondents. The third part of the questionnaire is included in Appendix C.4.

4.3.3. Sampling Design

The sampling design process (Figure 4.8) is of critical importance, since it is integrated in all aspects of problems’ investigation; from definition of the problem to presentation of the results (Malhotra 2007).

The decisions made in each step of this process for this thesis, are:

Step 1: Definition of the target population

Elements or sampling units, extent, and time should be justified in this phase of the process. The elements of this study are individual shoppers and non-shoppers who are acquainted with the internet and 3D online environments and virtual worlds. In order to ensure that all participants would have experience with 3D online environments, the first question of the questionnaire was filtering participants. Participants were asked whether they visit 3D online stores. In case of a negative answer, respondents were asked to stop filling the questionnaire. Two participants had not visited 3D online stores and were excluded from the final sample. All other participants were well experienced users-consumers with at least two years experience in 3D online environments. The study took place between 20/07/2012 and 20/09/2012.

Step 2: Sampling frame determination

The sampling frame is considered as an image of the elements or sampling units of the target population. In this regard, the participants of this study were undergraduate students from the Athens University of Economics and Business, Department of Management Science and Technology, and undergraduate/postgraduate students from Technological Education Institute of Piraeus, Departments of Business Administration, and Accounting.

According to the theory of diffusion of technology, students are considered innovators, and more eager to use and experience new products and services and new environments (e.g., 3D interface) (Kevork 2011), while the internet is recognized as the major communication medium of teenagers (Sorce 2005). The innovative aspect of this style fits with the profile of university students.

Finally, Kearney et al. (2009) report that the sampling of the majority of studies in online environments has been university students (e.g., Liang and Lai 2002; Lee and Benbasat 2003; Clark et al. 2004; Biers and Richards 2005; Lee et al. 2006). Due to the aforementioned reasons, and time and budget restraints, this sampling frame was regarded as appropriate to provide rigorous results and reduce the sampling frame error.

Step 3: Sampling technique selection

Sampling without replacement was selected as the general approach of the sampling technique. Each element of the sampling frame that was included in the sample would not be reselected.

In terms of the available probability or non-probability techniques, a probabilistic simple random sampling, where each element of the sampling frame had the same probability of selection, was selected. The invitation to participate in the study was sent by e-mail to the students of the departments of the aforementioned universities. One of the advantages of this sampling technique is that the sample results can be assigned to the target population.

Step 4: Sample size determination

The extensive duration of the experiment has been one of the main obstacles for achieving a large number of participants. There are very few experiments in marketing research that require two hours of commitment to the survey. Usually, research call for a sample size of about 100-150 respondents in similar kind of research (Malhotra 2007). However there are a several studies with sample size lower than a hundred, such as Donovan et al.’s (1994) where 60 elements took part.

On the other hand, the sample size is strongly associated with the statistic tools that were selected for the data analysis. To test the research hypotheses described in section 4.1, one-way repeated measures Analysis of Variance, mixed/split-plot Analysis of Variance, and Multiple Regression were selected. The 59 respondents of the present study is a quite satisfactory sample for all the assumptions that underpin the use of the aforementioned statistic tools. Finally, taking into consideration that due to budget constraints there was not any participation motive, and in conjunction with the exceptional qualifications of the elements required to participate, the final sample size is quite rare and surprising.

Step 5: Sampling process execution

The last step of the sampling design process deals with the precise execution of the previous stages of the sampling process and consideration of the appropriate approach selection. After the precise execution of all the sampling process’ steps, the data acquired were analyzed to confirm that there are not any significant differences with the target population. It is worthwhile noting that not only the characteristics of the sample are quite similar with those of the target population but the vast majority of the elements are experienced users of 3D online environments.

4.3.4. Data Collection

The questionnaire, which is a popular instrument in marketing studies (Tashakkori and Teddlie 2003), has been the data collection instrument of the quantitative, conclusive research in the main research study. Malhotra (2007, p.300) illustrates the steps of questionnaire design, which are presented in Figure 4.9.

The decisions and corrective actions in each step of this process are:

Specify the information needed

This step is also the first phase of the research design. It is an iterative process in the beginning of the research design that is evaluated and corrective actions are taken where it is considered appropriate, until the final execution of the laboratory experiment. The analysis of the respective items of fourteen dependent variables of the organism dimension and two variables of response dimension will provide the information needed in this study. The respective items of these variables are included in the second part of the questionnaire (see Appendix C). In the first part of the questionnaire there were questions addressed to obtain both behavioral and skill characteristics of the sample, and to validate the appropriateness of the sample’s characteristics with those of the target population. The third part of the questionnaire included questions regarding the demographic characteristics of the sample, and two questions that would ensure the realism of the laboratory experiment.

Specify the type of the interviewing method

The interview took place in the researcher’s office at Athens University of Economics and Business. Each participant visited this office at a predetermined time slot and the questionnaire was administered by the researcher. The researcher was in the same room throughout the interview to guide the respondent in each phase of the interview and to answer in any questions. Also, a cover letter explaining the whole process and the steps of the laboratory experiment was administered to each participant at the beginning of the interview.

Determine the content of individual questions

The topic of the study is not considered sensitive or controversial in order to include neutral or other questions out of the main scope of the research objectives. All the questions of the interview were used to provide insightful results and to answer the research questions of the study.

Design the questions to overcome the respondent’s inability and unwillingness to answer

It was clear before the execution of the experiment that the length of the questionnaire would require considerable effort from the respondents. For that reason, the researcher was encouraging the participants throughout the interview process, emphasizing the value of their help contribution. The only part of the questionnaire that could be considered sensitive by some participants was the third part with the demographic characteristics. While none of the respondents refused to answer this part, the interviewer had already informed participants in the instructional leaflet that all answers would be used anonymously. Also, in order to increase the likelihood of retrieving demographic characteristics, they were placed at the final part of the questionnaire. At this point, participants will have overcome any obstacles and will be more willing to answer (Malhotra, 2007).

Decide on the questionnaire structure

The survey included both structured and unstructured questions. There were some open-ended questions to capture the reasons why users or consumers would buy products in 2D or 3D online environments or not.

In terms of structured questions there were five multiple-choice questions asking issues regarding the frequency of 2D and 3D internet use, the kind of products purchased through 3D online stores, and demographic characteristics. As discussed earlier, a filter dichotomous question was used at the beginning of the survey to ensure that individuals who participate in the study are in some way experienced users of 3D environments. The majority of the questions followed a five-point Likert scale scale format. Five-point Likert scales were used to measure the items of the constructs of the research model.

Determine the question wording

The main concern with respect to the questionnaire wording has been to develop a clear, understandable, well-structured, and theory-driven questionnaire. The guiding principles illustrated by Malhotra (2007, p.311) were followed for the wording of each question:

Definition of the issue

Ordinary words use

Avoidance of ambiguous words

Avoidance of leading questions

Avoidance of implicit alternatives

Avoidance of implicit assumptions

Avoidance of generalizations and estimates

Positive and negative statements use

The translation of the constructs’ items in Greek involved multiple cycles of revision involving an expert with a PhD in sociolinguistics which was also native in the Greek language and graduate in English literature.

Arrange the questions in proper order

The participants received the first part of the questionnaire which was comprised of general and simple questions regarding internet use in 2D and 3D online retail environments. The aim of this arrangement has been to make participants feel at ease with the environment and the researcher, and get interested in the research topic. The instructional leaflet informed respondents about the anonymity of the responses; to that end no identification questions (e.g., name, address, e-mail) were used. Also, the funnel approach was adopted by placing the specific questions of the research model in the second part of the questionnaire. In order to increase reliability, the items of the variables were not placed sequentially, but randomly.

Identify the form and layout

At this phase, the primary goal has been to obtain high quality results by providing a formal, high quality appearance to the questionnaire. This would encourage participants to consider this study as of high value and provide information of high quality.

Reproduce the questionnaire

The questionnaire was reproduced in high quality paper and printing method, to achieve a professional look. None of the questions was split in two pages, to avoid misleading participants. While the questionnaire is large in size, the tendency to crowd questions to achieve a shorter look was avoided, as this could lead to unanswered questions or less enlightening data. Also, results from pretesting indicated that that 9-font size that was originally set, was quite annoying for the participants, and was changed to 11-font size in the final version.

Eliminate bugs by pretesting

Pretesting wad considered essential before executing the final survey, following the recommendations of Luck and Rubin (1987) and Malhotra (2007). The pretesting phase investigated all aspects of the questionnaire design process. Pretests took place in the form of personal interviews in the room that the final laboratory experiment would be executed. The profile of the pretest participants was similar to those of the actual survey and was drawn from the target population of the study. Malhotra (2007) states that a small sample of ten to fifteen respondents is enough for the pretesting scope. Eleven individuals took part in the pretest, while the interviewer followed a debriefing procedure to perceive confusion and other problems related to the questionnaire. Participants were asked to think loudly while answering each question, to explain the meaning of the question and consider whether is confusing and needs rephrasing or not.

4.3.5. Data Analysis and Interpretation

In terms of data preparation, the process proposed by Malhotra (2007, p.429) was followed:

Preparing preliminary plan of data analysis.

Questionnaire checking: The questionnaire was checked in terms of completeness and quality.

Editing: The questionnaire was checked in terms of accuracy and precision.

Coding: All possible answers were assigned to a specific code. Each column of an xls file was used for each question answer.

Transcribing: The presence of the researcher throughout the laboratory experiment and check of the questionnaire avoided the existence of missing values. The researcher with two assistants transferred responses to the xls file.

Data cleaning: Very few missing responses were adjusted and consistency checks were made.

Statistically adjusting the data: Adjusting the data was not necessary according to this design of the research.

Selecting the data analysis strategy: Multivariate techniques were used. The multivariate techniques involved one-way repeated measures analysis of variance, split-plot analysis of variance, and multiple regression.

4.4. SUMMARY

This chapter provides a systematic analysis of the methodology followed for the main research in this thesis (i.e., Stage#1-Delphi method, and Stage#2-Laboratory experiment), placing particular emphasis on the design issues that would lead to rigorous and insightful results. Results from research studies conducted in Information Systems, Human Computer Interaction and Marketing domain areas guided the development of the research hypotheses. Research design and analysis of variables of the research model and their measuring methods, followed the most recent principles of Marketing research orientation. Also, the critical importance of conducting a qualitative study, namely the Delphi method, was introduced. The methodological concerns in terms of the Delphi process, its sample design and analysis of data, were thoroughly analyzed and reasoned defense was presented. Finally, design principles of the laboratory experiment, as well as sampling design, data collection, and analysis of results formed the discrete whole of this methodology.

CHAPTER 5. VISUAL EXEMPLAR OF 3D ONLINE RETAIL STORE LAYOUTS

This chapter describes the design issues involved in the visualization of the five distinct layout types and their underlying implementation techniques. The three-round Delphi’s method outcome supported the existence of five distinct layout types in 3D online retail environments. One of the most important research objectives of this study, as aforementioned, is to investigate the influence of store layout on fourteen variables that the extant literature indicates bear a causal relationship with store layout.

In this chapter, the options and decisions made by the researcher in order to provide the best possible laboratory testing context for the purposes of the main research are explored. The review of these options and decisions has merit in its own right, as it presents a number of design issues and practical concerns for this type of research. Therefore, it is an essential part of methodological design decisions for conducting research in a 3D context.

5.1. DEVELOPMENT OPTIONS

Three development options were considered as the most appropriate to meet the requirements of the distinct 3D store layouts. The characteristics of each option, and the arguments for the most suited for the scope of this study are discussed below.

Development of "real" 3D online retail stores

A possible development option could be the development of the five layout stores within a virtual world or in five different URL links on the internet. While this an attractive option from a research perspective, budget constraints rendered it unrealistic. Specifically, in order to rent virtual land in a virtual world for five stores allocated in different virtual places or in five-storey block, one has to pay at least 500 US$ for a one month period. Also, specialists who have expertise with the specific development tools of the virtual world should be engaged in the development of the stores. Quite the same applies to the development of the stores in URL links.

Visual exemplar of 3D online retail stores

Another development option could be the use of a 3D tool for the development of stores, followed by and a video recording to capture all aspects of the in-store layout patterns. There are freeware programs such as Goggle SketchUp providing the ability to develop a 3D appearance of a building. This option facilitates the development of 3D stores in a laboratory setting and provides a clear view of the interior of a store.

"Paper-and-Pencil" visualization

This option could also meet the requirements of this research and is adopted in many research studies (e.g., Russel and Haney 1997; Pettit 2002; Wright et al. 1998). In this approach, the ground plan of each store layout could be designed on paper in order to visualize the store layout.

The second option is preferable to the third for deriving a clear picture/representation of store layout. Taking into consideration the relevant resources that made the first approach impractical, the second approach was adopted. Google SketchUp v.8 served as the main tool for building and modifying 3D models. This tool offers the additional advantage of import and export capabilities to other design programs.

5.2. RETAIL SECTOR

Retailing activity is taking place in various areas of 3D shopping channel. In the second chapter, several examples of apparel, furniture, real-estate, and tourism industry were presented. Results from the initial exploratory study indicated that the most groeing fields in coming years will be malls (51%of the sample) and retail stores (48.1% of the sample). Similarly, a great proportion of 71.2% of the respondents of the initial research study reported that they usually visit apparel stores. 3D stores selling avatars’ clothing are common in virtual worlds, while there are many companies such as Land’s End, Kenneth Cole, and Levi’s (Faust and Carrier 2009) that facilitate virtual try-on, through technology coming from Bodymetrics (Laird 2012). Following this market’s evolution, the apparel industry was considered a pioneering sector to serve as the experimental part of this study.

5.3. PRODUCT SELECTION AND ALLOCATION OF MERCHANDISE

Second Life is considered one of the leaders in virtual worlds (Shin 2008). To finalize the list of products that would be offered in 3D online stores, it was decided to use the search engine for places in Second Life and visit the ten top listed apparel stores. These places are likely to be the most crowded and "famous" places. An obstacle that this study had to face is that the products offered in the stores visited in Virtual worlds could not be copied and used for the scope of the experiment, because of copyright permissions. Also, design of products from diverse 3D stores is likely to influence the study’s participants in different ways, and bias the results in terms of store layout evaluation. For example the different design of the products displayed is likely to influence respondents’ preferences. To overpass these obstacles, it was decided to use products offered in the Database of Google SketchUp 8. However, the variety of products offered by this program is limited. The use of Adobe Photoshop CS6 was considered appropriate to design clothes that are based on the products offered by Google SketchUp but look quite different. The final database of the products developed was used in its entirety in all other store layout types, except for the "boutique" (i.e., store#4) store layout. The same products were used in all layout types to avoid bias in the evaluation of participants. Also, as is mentioned later in in this section, specifications of the "boutique" layout type does not enable the use of too many designed products. The variety of products designed and used in the stores is presented in the following list and figures:

As far as the allocation of products within each store is concerned, specifications coming from the Delphi method results determined merchandise allocation guidelines in each store separately. Specifically, in the first-round of the Delphi, a panelist introduced the theme-based and similar-based allocation of products; these terms were validated from all panelists in the second-round Delphi. Theme-based display of products refers to a practice where products of the same category (e.g., shirts) are located in the same place in a part of the store, one next to the other. Similar-based display refers to the practice where complete outfits (e.g., shirt, jeans, belt, socks and shoes) are displayed, the one next to the other. The first practice serves the ease of searching among products of the same category, while the second serves matching an outfit’s parts.

5.4. REAL MODELS AS PROMOTERS OF THE VIRTUAL STORE’S COLLECTION

The same brand name was used in each store layout type for internal validity purposes of the forthcoming laboratory experiments. In other words, to avoid bias participants’ evaluation of store layout, a nonexistent brand name was chosen. Clothing displayed in each store was named "Winick’s Collection"; it comes from the first name of the researcher’s avatar in Virtual World Second Life since 2008, which is Winick Ceriano. The first name of the avatar comes from two words, namely "wine" and "Nick", which translated in Greek, implies the real surname of the researcher.

The practice of displaying images of real world models to promote products, in walls of a 3D store, is common in Virtual Worlds retailers’ stores. Following current business practice, it was decided that copying images from models appeared on the Internet or 3D stores was not an option because of copyright permissions. To avoid this difficulty, one of the researchers captured images with an iPhone 4s mobile device, from six individuals who approved for their images to be displayed around each laboratory store layout. The final editing of the images was developed through Adobe Photoshop CS6. The images of the models are presented in figure 5.13.

5.5. CURRENCY SELECTION

A notable amount of virtual worlds such as Whyville and Second life retain their own local currency within the world (Messinger et al. 2009; Balkin 2004-2005; Bainbridge 2007; Stuart 2007). In some virtual worlds, inhabitants can engage their virtual funds to get for example extra privileges, equipment etc, while in others the virtual currency can be exchanged for real Dollars or Euros. For the virtual stores of the laboratory experiment, it was decided that there should be prices presented on the displayed products in order to provide a more realist aspect of the retailing dimension. In case the products were displayed in a currency familiar to participants, it would be likely that prices of the products could influence participants’ evaluation and bias the results. In this regard, an imaginary currency was invented, namely the "$GRT". The "$" is a familiar notation of currency for participants, while the "GRT" symbolism comes from the consonants of the word "Gortyna" which is one of the most historic cities of the island Crete, achieving considerable growth three thousand years ago. A gortynian coin was auctioned at the end of 2011, for about 480.000 US$, which has been a world record for this type of auctions (Archeology Newsroom 2011).

Another attempt to avoid bias in the results has been that there were only four different prices for the products displayed in each store. Two examples of the interactive frames where the products are included are presented in Figure 5.14.

5.6. SCALE OF STORES’ DEVELOPMENT

The scale development of 3D objects is of major concern in 3D environments. A glance at Figure 5.15 may lead an observant to assume an unequal size for these two stores. However, the size of both of these stores presented, is one hundred digital square meters.

The scale of the objects’ development creates a specific illusion in the human eye in terms of the largeness/size of a 3D online store. The specifications of the distinct store layouts required different sizes of store layout. To achieve this, the scale difference between the avatar largeness and all other objects of the store was calculated in each store layout type. This approach becomes clearer in the following sections, where images of each store layout type are illustrated.

5.7. GRAPHICAL INTERFACE OF STORE LAYOUT TYPES

This section is dedicated to the presentation of images of the laboratory store layout types according to the specifications that resulted from the Delphi method. While the analysis of the Delphi study is thoroughly presented in chapter six, the visual representation of specifications for each store layout type is provided in this section. There are six images representing views of each store layout type.

5.7.1. Store #1 (Medium Size Store)

The specifications of this layout type according to the Delphi study results are listed below, while the corresponding visualization is presented in Figures 5.16 to 5.21:

Theme-based/Similar-based display of products.

Demo products or models wearing the products/images posted will help the customer reach a decision.

Posters need to highlight the details of the products.

Insertion of screens in the floor plan to increase the amount of the display space they have.

Requires avatars to move through the store rather than just being able to pan the walls with the camera.

These stores tend to use images on the walls and may also use additional structures, but will leave some room in the middle for a model or two.

5.7.2. Store #2 (Large Warehouse Store)

The specifications of this layout type according to the Delphi study results are listed below, while the corresponding visualization is presented in Figures 5.22 to 5.27:

Helpful display to allow the customer to compare products to each other.

Functionality of comparing similar products.

Theme-based/Similar-based display of products.

Designer should be able to be contacted for further info on the products, because of the way they had the products designed.

Ability to teleport into specific product-related areas.

Easy ability to get into the building through alternative entry points.

A virtual salesmen could guide customers find the products.

5.7.3. Store #3 (Image-reliant Store)

The specifications of this layout type according to the Delphi study results are listed below, while the corresponding visualization is presented in Figures 5.28 to 5.33:

Wall-only-items

Image stores are a great way for the retailer to reduce the lag of the store

Theme-based display of products

Very simple product management for the end-user

Due to simple images, the simulation is much lighter and system requirements can be kept much lower. However, this sacrifices the realism of having a proper 3D model on screen.

Inexpensive Approach: Makes it possible to show a lot of different items in what can be a relatively small space, especially when extra display walls are included

5.7.4. Store #4 (Boutique Store)

The specifications of this layout type according to the Delphi study results are listed below, while the corresponding visualization is presented in Figures 5.34 to 5.39:

They sell small items such as virtual hair for avatars or shoes.

They tend to mimic physical stores with display cabinets and shelves.

Customers browse the store quickly and if they don’t find something they like, they can simply move on to the next one.

The owner may also design note cards that are easy to give away and be shared between avatars/customers.

Demo products also play a major role in this category.

One should be able try on the product before reaching the decision of buying it

Clear display of products.

Limited number of the available products.

Feasible for some products such as artistic items.

Theme-based/Similar-based display of products.

Visual interest: interesting architecture, walls interesting architecture, wall of glass of glass, attractive materials – appeals to residents.

Need to have enough blank space to make it easy for people to see the content of the shelves.

Need to give distinctive names to items for people to be able to differentiate among them.