Turning seven ideas into fifty survey questions.
A model is only as good as its measures. This page shows how the questionnaire was built and fielded, how each idea in the model was turned into survey items, and the exact wording respondents saw, with the reliability of every scale.
- n = 250completed responses
- Online panelprofessional consumer panel provider (Qualtrics), Russia
- Qualified ifowns a mobile device and has used a mobile commerce site or app
- 7-point scale1 = strongly disagree, 7 = strongly agree
- 59 items50 for the model, 9 for cultural control variables
How the questionnaire was built
Every scale was adapted from published research, then checked twice before fieldwork, following the pretest and pilot procedure of Hult, Hurley and Knight (2004).
Adapt established scales
Items were taken from validated scales and reworded for the mobile commerce context, so each construct rests on wording that has worked before.
Expert pretest
5 academics in marketing and information systems judged logical consistency, ease of understanding, item sequence and relevance to the context.
Pilot test
30 mobile commerce users completed the survey to check content quality and the reliability of the measures.
Fieldwork
250 responses collected from qualified panel members through the online panel.
Two kinds of measure, chosen on purpose
Before writing a single item, each construct needs a decision: are its items different windows onto the same thing, or different parts that add up to the thing? The answer changes how the scale is built and how it is judged.
Items are symptoms of the idea
The items should move together, and any one could be dropped without changing the meaning. Judged by internal consistency: loadings, Cronbach’s α, composite reliability and AVE.
Used for: the seven value dimensions, both kinds of exploration, return and recommend, and the two cultural controls.
Items are parts that build the idea
The items cover different facets and need not correlate. Dropping one changes what is being measured. Judged by overlap between items (VIF) and by each item’s weight.
Used for: immersion (eight facets of flow), and the three kinds of value, each built from its dimensions.
The constructs and the exact items
Select a construct to see its items. Loadings above .70 indicate that an item is a reliable reflection of its construct; every reflective item in the study cleared that bar. α is Cronbach’s alpha, CR is composite reliability (both should exceed .70), and AVE is average variance extracted (should exceed .50).
Useful (functional value)
A higher-order construct formed by three dimensions, 14 items in total.
Ubiquity5 items, reflective · adapted from Okazaki and Mendez (2013)α .95CR .96AVE .83
| Survey item | Loading |
|---|---|
| Using Mobile Internet keeps me well informed at all times. | .908 |
| When I cannot wait, and I need a certain type of information immediately, I will use the Mobile Internet. | .907 |
| Mobile Internet provides a quicker way of updating myself. | .880 |
| Mobile Internet is practical because I can use it without difficulty wherever I am. | .929 |
| Mobile Internet gives me the freedom to find the information that I need. | .916 |
Customization5 items, reflective · adapted from Yang and Lin (2017)α .91CR .93AVE .73
| Survey item | Loading |
|---|---|
| I value M-commerce webpages that: are personalized for my usage experience preferences. | .865 |
| I value M-commerce webpages that: offer personalized services and/or products based on my past usage experience. | .886 |
| I value M-commerce webpages that: seem to know what I want according to my location. | .858 |
| I value M-commerce webpages that: seem to know my preferences based on my past search and browsing behaviour. | .835 |
| I value M-commerce webpages that: provide me with information (e.g., deals and ads) tailored to my interests and preferences. | .815 |
Navigation4 items, reflective · adapted from Johnson et al. (2020); Richard and Chandra (2005)α .91CR .94AVE .79
| Survey item | Loading |
|---|---|
| M-commerce webpages provide a clear process for browsing. | .864 |
| Search features (e.g. filters, categories, suggested keywords) on M-commerce webpages help me to locate accurate information. | .892 |
| It is easy to find the necessary information through internal hyperlinks when using M-commerce webpages. | .911 |
| The categorization and layout contents (e.g. site map and menu items) of M-commerce webpages are helpful to obtain what I am looking for. | .897 |
Enjoyable (emotional value)
A higher-order construct formed by two dimensions, 9 items in total.
Aesthetics4 items, reflective · adapted from Li and Yeh (2010); Johnson et al. (2020)α .92CR .95AVE .81
| Survey item | Loading |
|---|---|
| The arrangement of elements (i.e., colours, boxes, fonts, etc.) of M-commerce webpages are attractive. | .884 |
| The M-commerce interface design is professional looking. | .900 |
| The presentation and demonstration of products on M-commerce webpages are aesthetically pleasing. | .912 |
| The overall look and feel of the M-commerce webpages are visually appealing. | .907 |
Enjoyment5 items, reflective · adapted from Nysveen et al. (2005)α .93CR .95AVE .78
| Survey item | Loading |
|---|---|
| Using M-commerce is fun. | .836 |
| I find using M-commerce enjoyable. | .908 |
| I find using M-commerce very entertaining. | .910 |
| I find using M-commerce very exciting. | .909 |
| I feel a sense of adventure while using M-commerce. | .833 |
Social (social value)
A higher-order construct formed by two dimensions, 9 items in total.
Interactivity5 items, reflective · adapted from Fang (2017)α .92CR .94AVE .78
| Survey item | Loading |
|---|---|
| Using M-commerce, I can obtain information without any delay. | .873 |
| When I click on M-commerce webpages’ links, I feel I am getting the instantaneous information I expected. | .835 |
| M-commerce webpages facilitate two-way communication. | .887 |
| M-commerce webpages give consumers the opportunity to talk back. | .879 |
| M-commerce webpages are effective in gathering visitors’ feedback. | .876 |
Social presence4 items, reflective · adapted from Fang (2017)α .93CR .95AVE .82
| Survey item | Loading |
|---|---|
| There is a sense of human contact during experiences with M-commerce webpages. | .931 |
| M-commerce webpages give me a sense of "personalness" as I interface with it. | .909 |
| There is human-like warmth associated with M-commerce webpages. | .889 |
| There is a sense of human sensitivity associated with M-commerce webpages. | .893 |
What the shopper experiences and intends
Immersion, the two kinds of exploration, and the outcome.
Immersion (flow state)8 items, formative · adapted from Engeser and Rheinberg (2008)VIF 2.1 to 3.75 of 8 weights sig.
| Survey item | Weight | Loading |
|---|---|---|
| I feel that I have control over my interaction with M-commerce webpages. | .263 | .858 |
| I am unaware of distractions. | .127 | .673 |
| I am totally absorbed in what I am doing. | .035 n.s. | .788 |
| My attention is focused entirely on what I am doing. | .062 n.s. | .827 |
| Time seems to go very quickly. | .169 | .748 |
| Using M-commerce, I feel just the right amount of challenge. | .234 | .903 |
| Using M-commerce, my thoughts/activities run fluidly and smoothly. | .065 n.s. | .840 |
| Using M-commerce, I know what I have to do each step of the way. | .259 | .842 |
A formative construct is judged differently: by whether its items overlap too much (VIF below 5) and by each item’s weight. Three weights were not significant, but those items were kept because their loadings are above .50, following Hair et al. (2019).
Searching to decide (specific exploration)3 items, reflective · adapted from Litman and Spielberger (2003)α .89CR .93AVE .81
| Survey item | Loading |
|---|---|
| Using M-commerce, I can work hard at a product’s problems that I feel must be solved. | .893 |
| Using M-commerce, I can strive hard to obtain a deeper understanding of a looked-for product. | .919 |
| Using M-commerce, I am interested in discovering how a specific product works. | .898 |
Browsing to discover (diversive exploration)3 items, reflective · adapted from Litman and Spielberger (2003)α .92CR .95AVE .86
| Survey item | Loading |
|---|---|
| Using M-commerce, I enjoy learning about products that I am unfamiliar with and know less about. | .923 |
| Using M-commerce, I find it fascinating to learn new information regarding the general category of products. | .930 |
| Using M-commerce, I enjoy exploring new products. | .932 |
Return and recommend (future patronage intention)4 items, reflective · adapted from Baker et al. (2002)α .88CR .92AVE .74
| Survey item | Loading |
|---|---|
| Revisit likelihood is ... | .803 |
| Using of M-commerce is ... | .853 |
| I say positive things about M-commerce to others. | .895 |
| I would recommend M-commerce to someone who seeks my advice. | .886 |
The first two items are shown as they appear in the thesis table, where the response stems are abbreviated.
Control variables
Two cultural orientations were measured with scales. Age, gender, education and mobile internet plan were single questions.
Uncertainty avoidance (control)3 items, reflective · adapted from Sharma (2010)α .76CR .86AVE .67
| Survey item | Loading |
|---|---|
| I would not describe myself as a risk-taker. | .841 |
| I prefer specific instructions to broad or general guidelines. | .711 |
| I feel safe when I am in my familiar surroundings. | .885 |
Individualism-collectivism (control)6 items, reflective · adapted from Sharma (2010)α .90CR .92AVE .66
| Survey item | Loading |
|---|---|
| I would rather depend on myself than others. | .849 |
| My personal identity, independent of others, is important to me. | .888 |
| I rely on myself most of the time, rarely on others. | .839 |
| It is important that I do my job better than others. | .773 |
| The well-being of my group members is important for me. | .768 |
| I feel good when I cooperate with my group members. | .754 |
How the higher-order constructs were estimated
Useful, enjoyable and social value are not measured directly. They are built from their dimensions using the disjoint two-stage approach recommended for PLS-SEM (Sarstedt et al., 2019).
Estimate the seven dimensions
The model is run with the seven dimensions connected directly to immersion, exploration and the outcome. Their reliability and validity are established here, and a score for each dimension is saved for every respondent.
Build the three kinds of value
The saved scores become the indicators of the three higher-order constructs. The model is run again, and the weights show how much each dimension contributes to its kind of value. Those weights are reported on the results page.
What this design can and cannot say
The sample is an online panel in one country, qualified on mobile commerce use, not a probability sample of all shoppers. All measures come from one self-completed questionnaire at one point in time. That is a sound and common design for testing relationships between perceptions, and the results should be read as evidence about those relationships, not as population estimates or proof of cause and effect. I come back to this on the last page.