**3. Data Collection and Analysis**

To analyze our research questions an online survey was conducted using a questionnaire with statements considering *Environmental awareness aspects*, *Building conditions* as well as *Behavioral*, *Normative* and *Control beliefs* of the TPB model. Finally, this extended form of the TPB served as a guideline for the identification of relevant influencing factors within the existing literature. In the following we introduce the precise statements and questions asked based upon the identified influencing factors. Moreover we provide information on the procedure of data acquisition and the method used to statistically analyze the collected data.

#### *3.1. Survey Content*

The factors and statements used in this study were either drawn directly from the available scientific literature in the context of energy efficiency and residential buildings (such as e.g., [19,20,24]) or were specifically created based on factors identified as relevant. Moreover, we utilized statements from Bearden et al. [40] to assess the potential influence of individuals' *Environmental awareness*. The *Building conditions* [22,32] were examined using self-developed statements related to the structural condition, the energy efficiency as well as a variable representing the comfort in the building and its visual appearance.

The wording of the statements used in our questionnaire can be derived from Table 1. While "Non-Refurbishers" were asked to refer their answers to hypothetical energy-related refurbishment activities on their buildings, "Future-Refurbishers" were asked to refer their answers to those measures stated as intended for the near future.


**Table 1.** Statements used in the questionnaire of this study (wording for "Non-Refurbishers").


**Table 1.** *Cont.*

Answer-scales: I totally agree/I agree/Neither agree nor disagree/I don't agree/I don't agree at all. \* Yes/Not sure/No. \*\* I should take actions ... as soon as possible (asap)/ ... in the next years/there is no need. Source: Content adapted and adopted from Zundel and Stieß [19], Stieß and Dunkelberg [20], Achtnicht and Madlener [24], Bearden et al. [40], Black et al. [22] and Organ et al. [32].

## *3.2. Data Collection*

For the purpose of our study, a Germany-wide online survey was conducted during June and July 2016 using an online panel provided by a market research institute. Our target group were house owners of single- and two-family houses in Germany who lived in these houses at the time of data collection. By asking the house owners whether a refurbishment project was planned or not, the group of "Future-Refurbishers" and "Non-Refurbishers" were identified. Subsequently, "Future-Refurbishers" were asked whether they plan to undertake EERM on the upper or lower building envelope, the façade, windows and/or doors. Only those house owners who stated their intention to realize at least one EERM related to these building components or intended to modernize the heating system (e.g., via solar thermal systems, installation of a ventilation system with heat recovery) were considered as "Future Refurbishers" for the present study. "Future-Refurbishers" without energy-related measures were not considered for this study.

Those individuals who stated to have no refurbishment intentions were considered for this study when a need for EERM was indicated. This need was identified by asking a question considering the perceived energy-related status of those building components.

Only those owner-occupiers (without refurbishment intentions) who stated a "need" or an "immediate need" for improving the energy efficiency of the heating system or of at least one of the mentioned building components were considered for the group of "Non-Refurbishers".

Finally, after data cleaning and sorting out owner-occupiers without energy-related refurbishment intentions (75 respondents) or needs (627 respondents) 351 "Non-Refurbishers" and 734 "Future-Refurbishers" were available for statistical data analysis. The data cleaning procedure followed a combined approach characterized by an analysis of the respondents' answers to the individual question sets as well as the time respondents devoted for answering the questions. After marking questionnaires in which mainly identical answers and/or short processing times were evident, an individual case-by-case examination of suspicious but also incomplete data sets finally led to the exclusion of questionnaires of 320 "Future-Refurbishers" and 345 "Non-Refurbishers".

The characteristics of the respondents of both groups are presented in the Figures 1–4 indicating statistically significant differences between the two groups in terms of age (Figure 1), education (Figure 2), average monthly net household income (Figure 3) and in terms of the construction periods (Figure 4) of the participants' buildings.

**Figure 1.** Age groups of "Future-Refurbishers" and "Non-Refurbishers".

**Figure 2.** Education levels of "Future-Refurbishers" and "Non-Refurbishers".

**Figure 3.** Average monthly net household income (in €) for "Future-Refurbishers" and "Non-Refurbishers".

**Figure 4.** Construction periods of "Future-Refurbishers" and "Non-Refurbishers" buildings.

Figures 2 and 3 show that "Future-Refurbishers" are not only more likely to have a university degree, but also a higher income than "Non-Refurbishers." Since EERM often result in substantial costs, it seems that people with a higher income are more capable of realizing EERM in the future. Moreover, the group of "Non-Refurbishers" is older than the group of "Future-Refurbishers." Further, there are significant differences between the two groups related to the age of their houses with the average construction year of "Future-Refurbishers" buildings being 1968 compared to 1963 for houses of "Non-Refurbishers". This implies that a higher share of "Non-Refurbishers" buildings were built before the 'Thermal Insulation Ordinance' came in place in 1977 in order to enhance the energy efficiency of new buildings in Germany.

The answers of the "Future-Refurbishers" and "Non-Refurbishers" to the statements presented in Section 3.1 were analyzed using binary logistic regression. This method is intended to reveal factors beyond socio-demographic aspects—that allow for a differentiation of both analyzed groups and

potentially for deriving recommendations for overcoming the reluctance concerning residential EERM in Germany.
