**3. Hypothesis Development**

Our research hypotheses were developed by:


The research was based on six hypotheses (H1–H6) shown below, "Consumers' willingness to change their shopping habits to reduce environmental impact" ("W", a single exogenous variable) being seen as the input construct (allowing the statement of the hypothesis H1 below), considering our focus on "Retailers' digital transformation to aid consumers to adopt more sustainable lifestyles and to make informed choices in the omnichannel world" ("D"), through the facilitating function of "Retailers' increased concentration on responsibly answering to sustainability as a personal value of consumers changing their behavior" ("R", a mediating variable allowing the statement of the hypothesis H4 below). As shown in Figure 2 below, "Retailers' increased concentration on responsibly answering to sustainability as a personal value of consumers changing their behavior" ("R"):


**Figure 2.** The theoretical research model. (Source: Own research, based on the literature).

The theoretical research model designed on the basis of a valuable quantitative model [175] allowed for the exploration of the causal relationships between the abovementioned main constructs identified (to which was added a fifth construct within the challenging context of the digital disruption impacting consumers' perceptions, attitudes, intentions, behavior, and experience: "Retailers' need to translate consumers' uncertainty into trust, identifying risks associated with disruptive technologies and making them less severe" (T). This above-mentioned research model designed in accordance with the theory of structural equation modeling (SEM) reflects the following hypothesized influences shown by the SEM's path diagram:

**Hypothesis 1** (**H1**)**.** "Consumers' willingness to change their shopping habits to reduce environmental impact" ("W") has a positive influence on "Retailers' increased concentration on responsibly answering to sustainability as a personal value of consumers changing their behavior" ("R").

**Hypothesis 2** (**H2**)**.** "Retailers' sustainability agenda, including by fulfilling consumers' sustainability demands with new products and processes" ("S") has a positive influence on "Retailers' increased concentration on responsibly answering to sustainability as a personal value of consumers changing their behavior" ("R").

**Hypothesis 3** (**H3**)**.** "Retailers' sustainability agenda, including by fulfilling consumers' sustainability demands with new products and processes" ("S") has a positive influence on "Retailers' digital transformation to aid consumers to adopt more sustainable lifestyles and to make informed choices in the omnichannel world" ("D").

**Hypothesis 4** (**H4**)**.** "Retailers' increased concentration on responsibly answering to sustainability as a personal value of consumers changing their behavior" ("R") has a positive influence on "Retailers' digital transformation to aid consumers to adopt more sustainable lifestyles and to make informed choices in the omnichannel world" ("D").

**Hypothesis 5** (**H5**)**.** "Retailers' digital transformation to aid consumers to adopt more sustainable lifestyles and to make informed choices in the omnichannel world" ("D") has a positive influence on "Retailers' need to translate consumers' uncertainty into trust, identifying risks associated with disruptive technologies and making them less severe" ("T").

**Hypothesis 6** (**H6**)**.** Consumers' willingness to change their shopping habits to reduce environmental impact ("W") has a positive influence on Retailers' need to translate consumers' uncertainty into trust, identifying risks associated with disruptive technologies and making them less severe ("T").

### **4. Research Methods**

Based on the above-mentioned comprehensive literature review, a quantitative study was employed to investigate the major shifts in sustainable consumer behavior on the Romanian retail market within the context of the Green European Deal, and retailers' priorities in agilely adapting to these significant evolutions. A structural equation model and an associated "Sustainable Consumer Behaviour on the Romanian Retail Market Questionnaire" were used. The data collection (on the basis of this questionnaire) in the quantitative study was performed via a survey conducted in a supermarket chain (data collected face-to-face inside the supermarket) in Romania from 7 June 2021 to 28 June 2021. In this study, we took into account 1005 respondents, all customers of the supermarket chain, the age structure being presented in Table 1 below. The sampling methodology was a random one, a mixture between the systematic approach (every 20th person) and a stratified approach (gender structures). Thus, the study has a high legality for this welldefined group: "Customers in Romania who do their shopping physically in the stores of the analyzed supermarket chain" (for reasons related to privacy, we cannot disclose the name of the store chain; to ensure the anonymity of the respondents, no personal identification data were collected). The research tool has a high level of validity and was effective in measuring major changes in the sustainable behavior of consumers (the defined group), as well as the way in which retailers with physical store chains adapted. The questionnaire consisted of a set of 38 questions answered by the above-mentioned sample of 1005 respondents. As shown in Table 1 below, the average time spent to answer the questions was 8.30 min. As for the type of questions, closed questions were used, and the type of data sought (to be obtained from these closed questions) was from factual, opinion, and behavioral questions.


**Table 1.** Sample Structure (research).

Participation was voluntary and the confidentiality of the responses was ensured. All respondents answered the 38 questions and went through all the testing phases.

The generation of the items was based on a combination of inductive and deductive strategies, and involved reviewing the literature and our own previous research as well as conducting interviews with stakeholders (experts and clients). We generated an initial set of approximately 100 items, focusing on their environmental impact habits, retail trends and sustainability agenda, their digital transformations, and associated risks, as reflected in the cited studies. In parallel, we conducted focus group interviews with top and middle management (10 people), commercial workers (20 people) from four (4) companies in Romania, and retail clients (20 people). We had an approximately equal distribution by sex, and the respondents were between 27 and 52 years old, with a median of 38 years. Participants were asked about the topics listed above and were asked to come up with a concrete example of daily activities, the purpose being not only to evaluate the applicability of the constructs, but also to identify directions that were not captured in the literature. Following these phases, new elements were identified regarding the risks associated with disruptive technologies, but also regarding the sustainability demands from customers. Thus, an additional number of over 30 items was generated, bringing the total number to 130.

The measuring tool for major changes in sustainable consumer behavior (the defined group), as well as the way in which retailers with physical store chains adapt, was analyzed by five professors from different fields belonging to the departments of managementmarketing, economics, and sociology. The research verification form was analyzed using the "face validity technique", and the feedback obtained was used to improve the tool. Improvements were made to the selection of items (deletion and addition), unclear terms, ambiguous wording, etc. Following this process, we removed about 50 items, reducing the cart to 80. Next, a group of three university assistants from the same departments, together with Master's students, formed a new jury to validate the content for a new selection, but also to assign the items to the relevant variables. In this new phase, another 30 items were removed and another 20 underwent major changes. These 50 items were revised by the authors once again in order to adapt the theoretical content and make sure that it reflects exactly the defined dimensions. All these steps helped us to make sure that the items modified and retained for empirical analysis were clear, respecting of the hypotheses, and consistent. We dropped another number of items, leaving 38. All items were assigned a 5-step Likert scale: W (between 1—strong disagreement, and 5—strong agreement); R, S, D, T (Yes, Partly true, Neutral, Rather not, and No). For the next step, we ran a pilot test involving 30 respondents (these 30 respondents were not included in the final study; they are not among the 1005 respondents on whom the study is based). We thus aimed to increase the level of stability and internal consistency of the instrument. The 30 respondents provided additional feedback on the "face validity" process, marking some errors, elements less known to those in academia, but also other suggestions aimed at increasing the quality of the tool. The response time was also measured, and certain questions were reworded to shorten the time taken to complete the questionnaire. Subsequently, the reliability of the instrument was measured using Alpha Cronbach. The Cronbach Alpha value was greater than 0.6 for each construct and was considered acceptable. The final form is presented in Table 2 below.


**Table 2.** The latent variables and related items.

**Table 2.** *Cont.*



**Table 2.** *Cont.*

