2.1. Indirect Questioning Technique and Socially Desirable Responding
Socially desirable responding refers to response tendencies that occur primarily in self-assessment procedures, such as those typically used in personality or attitude measurements [
20,
21]. In such cases, the respondents do not give an answer that really applies to them, but answer in a way that is socially desirable from their point of view [
16]. A major reason for this is to leave a positive impression by giving answers that conform to expectations and to receive social recognition or at least to avoid social rejection [
22]. This impression management is often motivated by maintaining or actually increasing self-esteem and is to be distinguished from self-deception, in which a person assumes that the evaluation actually applies to him or her [
23]. Furthermore, socially desirable responding is seen to emerge either as a more general personality trait, which can be further distinguished upon fundamental value orientations and the level of its consciousness [
24], or from context dependent judgements about the perceived desirability a respondent attributes to a specific item [
25,
26]. Ethical issues such as the consideration of social-ecological aspects within consumption decisions are generally perceived as being socially desirable traits [
14] and respective traditional self-report measures are therefore likely to be affected by item desirability and a general need for approval [
27].
The indirect questioning technique is supposed to deprive participants from the motivation to respond in a socially desirable fashion by embarking on the human tendency to project own psychological states onto other entities [
18,
28]. Psychoanalytic theorists discuss projection as a self-serving defense mechanism, which enables humans to outsource owned “shortcomings” in order to avoid psychological threat [
29]. The concept of projection was already applied by Sigmund Freud in various contexts, for example, to explain different psychological phenomena, such as mental illnesses, e.g., paranoia [
30] and neuroses [
31], as well as to explicate human coping with fear [
32] and jealousy [
30] or to comprehend the meaning of dreams [
33]. Freud understood projection as a simple primitive process of the externalisation of internal mental states and emotions in order to obtain a picture of the outside world [
31]. The concept of projection manifests through an attribution of internal drives, properties, and thoughts to an external object [
34]. Thus, when projection processes occur, there is a partial transfer of self-presentation to a presentation of objects [
35], whereby the objects can also be other people [
36]. In this case, people attribute certain characteristics and behaviours to other individuals, which they cannot or do not want in themselves. It is this mechanism that is utilised by the concept of indirect questioning and it has gained great importance not only in psychoanalysis. Indirect questioning techniques have been used for a long time in clinical psychology, motive research, and market research.
In recent decades, various measurement approaches for the indirect determination of needs, attitudes, and values have been developed in order to identify hidden motivations and uncensored attitudes in consumers, which are withheld by the test persons in direct questioning due to feelings of shame or embarrassment [
37]. These methods include, e.g., Thematic Apperception Tests (TAT), sentence completion tests, picture storytelling tests, and cartoon tests e.g., [
38,
39,
40,
41]. However, all these unstructured measurement methods are qualitative-oriented and often require elaborating materials applicable to the respective survey, e.g., drawing appropriate cartoons or selecting and designing adequate images [
42]. In addition, analysing the obtained responses is time-consuming and may introduce distortions due to subjective interpretations of the answers [
42]. Accordingly, the reliability and validity of unstructured projective surveys are deemed as problematic [
43]. To this end, employing structured approaches with closed questions seems to be helpful since they permit less room for interpretation and for a better comparability of the answers [
44].
The indirect questioning technique that is utilised in the current study is one such structured approach. In particular, it enables respondents to project potentially undesirable psychological states (e.g., low levels of caring for ethical issues within a private consumption decision) by a mere alteration of the point of reference for responding to an item. Instead of instructing participants to respond on behalf of themselves (e.g., “How much do
you agree with the following statements?”), which is the common standard in traditional survey methods, they are instructed to respond on behalf of an abstract referent group (e.g., “How much would a
typical consumer agree with the following statements?”). This is supposed to provide respondents with a feeling of impersonality, allowing for more “realistic” answers even if perceived as socially undesirable [
16].
Therefore, we suggest that participants who are approached with indirect questioning to measure consumer social responsibility (e.g., feelings of moral obligation to refrain from consumption options with poor social-ecological value) will omit SDR-related biases within their responses reducing those to lower (i.e., more realistic) levels of consumer responsibility expressions. However, we suggest responses elicited with direct questioning will introduce such biases and therefore indicate higher levels of consumer responsibility expressions. In particular, we expect when consumers judge their perceived social responsibility in regard to consumption options that are of poor social-ecological value, responses on behalf of the self (i.e., direct questioning) will indicate high levels of consumer responsibility perceptions (i.e., problem awareness, self-ascribed accountability, feelings of moral obligation, and pro-sustainable intentions) that are in line with results generally witnessed by traditional surveys. However, we expect for the exact same constructs, when measured with indirect questioning, judgements will indicate much lower levels of social responsibility perceptions, which are more in line with the low real world market shares of ethical products. Hence, we propose the following hypothesis:
Hypothesis 1 (H1). Judgements of social responsibility perceptions regarding consumption options with poor social-ecological value will be lower when elicited with indirect questioning than when elicited with direct questioning.
Furthermore, since we expect that the hypothesised differences between social responsibility perceptions in both questioning types are driven by SDR-tendencies, we expect measures that are not prone to social desirability tensions (e.g., functional benefits of products with poor social-ecological value) to be unaffected by the type of questioning. This approach of introducing constructs that are expected to be rather unsusceptible to SDR as baseline measures has been utilised in former studies on SDR with indirect questioning [
16]. Thus, we propose the following hypothesis:
Hypothesis 2 (H2). Judgements of functional benefits regarding consumption options with poor social-ecological value will be at same levels when elicited with indirect (vs. direct) questioning.
2.2. Response Mode in Indirect Questioning: Prediction or Projection?
In regard to the second objective of this research—evaluating the validity of the indirect questioning technique in more detail—we have developed and applied a tailor-made approach for this purpose. Since participants are asked to indicate responses on behalf of a referent group, a common validity concern is that they might generate mere predictions about the responses of specific referent group representatives [
19]. To provide distinct evidence for reconciling this ongoing concern with regard to the indirect questioning technique, we embark on a theory-driven test that utilises individual subjective norms as a yardstick. Since subjective norms are expectations of individuals about what the immediate personal environment judges to be right or wrong [
45], norm activation theory and the theory of reasoned action predict them to be particularly related to feelings of moral obligation and behavioural intentions [
46,
47]. However, the propositions for a relationship of subjective norms with feelings of moral obligation, and with behavioural intentions, respectively, predicate on the premise that relating constructs pertain to the same individual. It is not proposed by those theories that one’s own immediate environment affects moral feelings and behavioural intentions of an abstract other entity. We utilise this logic to clarify the prediction vs. projection issue.
Specifically, we expect when measuring the constructs subjective norms, feelings of moral obligation, and behavioural intentions on behalf of the respondent (i.e., direct questioning), the proposed systematic relationships between subjective norms and both other constructs should be evidenced since the relating constructs explicitly pertain to the same individual. However, maintaining subjective norms being measured on behalf of the respondent (i.e., direct questioning), while feelings of moral obligation and behavioural intentions are measured on behalf of a referent group (i.e., indirect questioning), a systematic relationship between those constructs would not be expected if respondents simply predict the responses of the referent group, since they would not pertain to the same individual as the subjective norms anymore. However, if participants actually project their own feelings of moral obligation and behavioural intentions onto the referent group, we would expect the relationships to remain at the same levels as when all three constructs are measured with direct questioning, indicating projection as the underlying response mode to indirect questioning. We hypothesise the latter and propose the following hypotheses:
Hypothesis 3a (H3a). Judgements of feelings of moral obligation and behavioural intentions to forego products with poor social-ecological value, measured with direct questioning (i.e., on behalf of the self), are positively correlated with the subjective norms of one’s own social environment for considering social-ecological aspects within consumption decisions.
Hypothesis 3b (H3b). Judgements of feelings of moral obligation and behavioural intentions to forego products with poor social-ecological value, measured with indirect questioning (i.e., on behalf of a referent group), are positively correlated with the subjective norms of one’s own social environment for considering social-ecological aspects within consumption decisions.