How do people interpret the value concept? Development and evaluation of the value conceptualisation scale using a mixed method approach

ABSTRACT

Value research has a long and extensive history of theoretical definitions and empirical investigations using large scale quantitative surveys. However, the way the general population understands, defines, and relates to the concept of values, and how these views vary across individuals is seldom addressed. The present study examined subjective interpretations of the term through focus group interviews, and reports on the development of a Value Conceptualisation Scale (VCS) that distinguishes six dimensions of different views on values: normativity, relevance, validity, stability, consistency, and awareness. Focus group interviews (n = 38) as well as several surveys (n = 100, n = 1519, n = 903, n = 94) were used to develop, refine, and test the scale in terms of response variety, temporal stability, as well as convergent and discriminant validity. These systematic results show that views on values do indeed vary significantly between participants. Correlations with dogmatism, preference for consistency, and metacognition were found for corresponding dimensions. The VCS provides an original measure, which enables future research to explore this variation on the conceptualisation of values.

KEYWORDS:

• Human values
• scale development
• mixed methods
• focus group

Introduction

‘What even are values? I think it’s a beautiful word, a winged word that we all use, but what does it mean? I think a lot of people are not aware of their values (…) and I think values aren’t rigid, but something that can change in the course of our lifetime, and does change. (…) Respect, acceptance, fairness, who would say ‘no, I don’t like that’? But the question is, do you live it?’

Focus group participant on values

The social phenomenon called ‘values’ is complex and often hard to grasp. In sociology, values are most commonly defined after Kluckhohn as ‘a conceptualisation, explicit or implicit, distinctive of an individual or characteristic of a group, of the desirable which influence the selection from available modes, means, and ends of action’ (Kluckhohn 1951, 395). Theoretical descriptions such as this one, try to tackle the elusiveness of values by describing stable characteristics of a universal ‘nature’ of human values (cf. Rokeach 1973). They can be important for distinguishing values theoretically from similar concepts such as social norms, attitudes and opinions, but researchers seldom explore these theoretical characteristics in empirical research (Klages, Hippler, and Herbert 1992).

We believe such an empirical perspective could reveal diversity in this ‘nature of values’, meaning differences in the ways people define values, think or feel about the concept. Previous exploratory research has shown that the perception of the value concept does indeed vary greatly, between parents (Essen and Giese 2010), school principles (Branson 2004), and educational administrators (Zupan 2012). However, these previous studies are limited to small social or occupations groups, and a more systematic study is needed to understand the logic behind these differences. Even in Kluckhohn’s definition, this empirical spectrum is somewhat embedded. Values can – in his understanding – be implicit or explicit, subjective or collective, in conflict or complimenting each other, stable or prone to change, guides for action or not relevant in everyday life. But, as Spates summarises in a pointed way, ‘ … until we can be sure that the values reputedly descriptive of a population are their values and not merely some figment of a theoretician’s imagination – we simply cannot know whether the description is correct or not’ (Spates 1983, 35f).

Against this background, the present study asks: What are the ways in which people in contemporary societies understand and relate to the concept of values? We propose to address this question, using the term value conceptualisation to collect and describe diverging views of and relationships with the concept of values using a brand new instrument developed within an exploratory mixed method study.

Thus, we want to go beyond the priority with which people hold some values above others. The study of people’s conceptualisation of values is an important addition to current value research, which is currently dominated by internationally comparative quantitative research (e.g. EVS, WVS, ESS) that primarily focus on competing assumptions about ‘a finite number of universally relevant value types on which people place relative importance’ (Rohan 2000, 259). This discussion presumes that the crucial difference lies in the priorities by which people hold some values above other values, and does not take into account that the concept of values can itself be of different importance and appeal to different social groups. We take a critical standpoint by questioning values as a concept that is universally equally conceived across societies, and propose value conceptualisation as a term to address potential differences.

Materials and methods

Research design

To tackle the challenging task to gain knowledge about what people understand as values, we propose the use of an exploratory mixed-method approach. Qualitative and interpretative methods are in our opinion better suited to inform new survey measures at this current state, than the reliance on theoretical thought alone. Exploratory mixed-method designs are beneficial, when the important variables of quantitative surveys are yet unknown, and few measures and instruments already exist (Creswell and Clark 2011, 75f). They are particularly useful for developing new survey instruments, because the qualitative methods help to open the researchers’ eyes for new dimensions of relevance, assists the wording and comprehensiveness of the items (Bryman 2006), and phenomena can be explored in depth as well as in their prevalence (Creswell and Clark 2011, 75f).

The Value Conceptualisation Scale (VCS) was developed in the course of a research project at the University of Vienna. The empirical work in the project was undertaken in three phases. First, the exploration phase (1) consisted of piloting and conducting focus group interviews, which explored subjective views and negotiating processes around values. These served as a basis for the item development phase (2), in which the VCS was constructed, pretested and implemented into a nationwide survey. The evaluation phase (3) aimed to evaluate the scale using the nationwide data, as well as a repeat survey and validation survey conducted in the following year.

Phase 1 – qualitative exploration

Phase I started with the development of a focus group guide through a small pilot study(Braun and Clarke 2013, 84f). The focus group interviews aimed to collect different experiences and attitudes of a diverse group of citizens on the value concept and self-proclaimed values. Focus group interviews were chosen, because they allow researchers to be open to the system of relevance of the participants and observe negotiating processes between them (Morgan 1997; Bryman and Bell 2015).

Previous qualitative studies (e.g. Essen and Giese 2010) reported problems of asking people directly about their ‘values’, as the term evoked negative associations, such as being too academic or outdated. To prevent the potential exclusion of certain social groups, the understandability and relatability of the interview questions were tested through pilot narrative interviews (Bryman and Bell 2015, 488; Schütze 1983). In light of these pilot interviews, the focus group guide was refined, and it was decided that explicit questions on value conceptualisation should be reserved for later stages of the interviews. This would give the participants time to ease into talking about their personal values, which is also in line with theoretical considerations on meta-cognition by Rohan (2000, 267).

Recruitment of participants was initialised through national and local newspapers, magazines, flyers, social media, and bulletin boards (e.g. inside super markets). Because focus groups benefit from homogenous composition within the groups (Morgan 1997; Bryman and Bell 2015), maximum variation strategy (Suri 2011) was aspired between groups with regard to three characteristics which empirical research ties to differences in values: age, urbanisation and socio-economic position (e.g. Klein 1991; Polak 2011; Spieß 2000). In total seven focus group discussions with 38 people were conducted, including 23 women and 15 men, between 17 and 67 years old. Homogeneity within groups (as well as variation between groups) was established during the research process with the use of online tools and was monitored via a small field survey.

The focus group discussions began with an introductory round, after which a narrative question asked participants to describe important life events that made them who they are today. At least a half hour of open discussion was granted, after which specific questions such as the participants understanding of the term ‘value’ were asked. The group discussion closed with participants being asked to agree on three changes they would like to see implemented in the world.

The focus group material was analysed following Bohnsack’s documentary method (Bohnsack 2010), and served as a basis for item generation. The major advantage in using focus group interviews for item generation, is the richness of the material, the relevance to the participants perspective and the accurate language (Nassar-McMillan et al. 2010). Although still rarely addressed in methodological literature (O’Brien 1993), many researchers have shown the benefits of focus groups for item development (Fuller et al. 1993; Nassar-McMillan et al. 2010; Wolff, Knodel, and Sittitrai 1993; Luyt 2012). In our case, item construction benefited especially from formulating and reflecting interpretation that revealed major similarities and differences in the conceptualisation of values, formulated in the groups.

Phase 2 – item development

The aim of the second phase was the construction of a quantitative survey, including the development of the VCS. Initial items of the scale were drafted in light of the focus group results (phase 1). Different opinions and counter-arguments were identified and collected as direct interview quotes, to look at the exact wording the participants used. Core items of the new scale were formulated in guidance of these quotes (see examples in Table 1).

Table 1. Selected examples of interview quotes and resulting survey items.

Participant quote	Item formulation
‘Well, loyalty is my first priority. That’s really what I try to live.’	My values give me clear guidelines of how to live my life
‘There are often problems, because you don’t know (pause) there are values that oppose each other. (…) They are contradictory.’	My values are often contradictory and not easy to implement
‘If you always have to discuss and defend your values (…) they become pretty aware to you in this domain.’	I am clearly aware of my values
‘What even are values? (…) I think I don’t really have something like that, at least I am not aware of it.’	I find it difficult to describe my values concretely

Based on focus groups (n = 38). Interview sequences were translated into English and adjusted to written language.

Because the quotes collected in the focus group discussions showed many severely oppositional views on value conceptualisations, the items were formulated on a bipolar scale. This form of a scale was developed with specific challenges in mind (e.g. Menold and Bogner 2015), 1) to ensure that items are homogenous and uni-dimensional, 2) to combat the effect of central tendency and ensuring variation of answers, 3) to represent as many manifestations of Value Conceptualisation as possible within the limits of the survey, 4) to distinguish clearly between people with different attribute levels, and 5) to balance the two poles and centre the scale.

To guarantee the quality of the preliminary items, the scale was repeatedly reviewed by a wider team of interdisciplinary experts on values, including sociologists, political scientists and theologians (DeVellis 2017, 102). During this phase a pilot study was conducted with the help of 100 students of the University of Vienna. The participants were asked to fill out the questionnaire, and were explicitly encouraged to write comments about unclear question texts, scale construction, or any unsureness which answer should be picked.

The pilot survey was analysed using basic statistical measures such as frequencies and descriptive statistics. All comments were collected and reviewed by the research team. This test survey revealed some problems with the first draft of the scale. Firstly, the bipolar mode of the scale seemed confusing to some students. Value conceptualisation was the only scale measured with bipolar rating, so it broke with the familiar Likert-scale routine. To counteract this, the scale description and question text were improved, explaining in more detail what the goal was and how the possible answers should be interpreted. Secondly, the position of the scale within the questionnaire was rearranged (from the third to 17^th position).

Following this step the refined VCS was implemented into a nationwide survey (the Austrian Value Formation Study, abbr. ÖWBS). Data collection was commissioned to an external market research firm and carried out via an online panel representative for the Austrian population between 15 and 69 years of age (with special regard to age, gender, federal states and education). Participants were incentivised to take part in the survey by an internal reward system of the market research firm. Only fully filled out surveys were included in the final data set, which consisted of 1,519 participants. The sample showed representative distributions for the country under study (Verwiebe, Seewann, and Wolf 2018).

The final VCS (used in the national survey) consists of six items, each measuring a different dimension of the way people relate to and understand the term values (see Table 2). Each item is rated on a 6-point bipolar scale, and each dimension of value conceptualisation is represented by one item in the final scale.¹

Table 2. Final survey items.

In the following, you see a variety of contrary attitudes towards values. Please locate your views between these opposing poles.
A	Which values are right, differs from person to person	1	2	3	4	5	6	There are values that are objectively right and apply to everyone
B	Values have little meaning in my everyday life	1	2	3	4	5	6	Values are very important in my everyday life
C	You have to guide people towards the right way	1	2	3	4	5	6	Nobody is allowed to dictate values to others
D	In my life, my values have changed greatly	1	2	3	4	5	6	In my life, the same values have always applied
E	My values give me clear guidelines of how to live my life	1	2	3	4	5	6	My values are often contradictory and not easy to implement
F	I am clearly aware of my values	1	2	3	4	5	6	I find it difficult to describe my values concretely

Item A conceptually corresponds to value normativity, item B to value relevance, item C to value validity, item D to value stability, item E to value consistency, and item F to value awareness.

Phase 3 – evaluation

The evaluation phase started with a repeat survey conducted within the national sample. All participants were contacted again one year after the initial testing. 935 people (500 men and 435 women between the ages of 15 and 68) agreed to fill out the original questionnaire again, which amounts to 62 percent of the former sample. The long time period between the two waves minimises the chance of carryover and deliberative effects, but as changes in the value conceptualisation of participants might have occurred (DeVellis 2017, 68f), the results have to be interpreted with care. We evaluated each item’s test-retest reliability through the use of Pearson correlations, as well as comparing the means, standard deviations and percentage of scores within a ± 1 threshold of the original and retest survey.

In a second step, discriminant and concurrent validity of the VCS were determined through a student sample of 94 students at the University of Vienna. Apart from the VCS, the survey included six theoretically related constructs – one for each dimension of the VCS. It also contained two divergent measures. This validity study with Viennese students was analysed by evaluating the VCS Pearson correlation with concurrent and discriminant measures.

Because to our knowledge no empirical or theoretical study on value conceptualisation exists, formulating hypotheses on concurrent and well tested concepts proved challenging. Hence, our goal was to select items that seemed most plausible in light of our qualitative results and the existing literature. Therefore, some theoreticised relationships presented in this paper are rather exploratory attempts to test the boundaries of our concepts in relation to similar theories, and need to be interpreted with caution.

The first concurrent concept was Altmeyers dogmatism scale. Dogmatism can be defined as unjustified and unchangeable certainty in one’s belief. The scale includes 20 items such as ‘The things I believe in are so completely true. I could never doubt them.’ and is rated on a 9-point scale (Altemeyer 2002, 720; translation by Rangel 2009). The second concurrent concept was conventional political participation developed by Barnes and Kaase (2014). In this 4-point scale with 7 items political participation is measured as an indicator for action taken to follow one’s own attitudes and values. The third concurrent measure was the short form authoritarianism scale (KSA-3; Beierlein et al. 2014). It measures three subdimensions of authoritarian attitudes and uses 9 items, such as ‘Proven modes of behaviour should not be questioned’. Items are rated on a 5-point scale. Furthermore, attitude stability (Petrocelli et al. 2010) was included as the fourth concurrent measure. The scale evaluates people’s certainty in their attitudes. A scale with 8 items such as ‘People can always change their attitudes’ is rated on a 6-point scale. Participants were also asked about their preference for consistency (Collani and Blank 2013). This scale consists of 7 items that measure the motive of keeping stable individual traits. Items are rated on a 5-point scale, and include answers such as ‘I feel uncomfortable, if I have to opinions that don’t go together’. The final concurrent concept was metacognition (McMahon and Good 2016). This term is used to describe awareness and knowledge of a thought process especially in the context of moral reason, as well as the choice to regulate cognitive strategies. A recently developed instrument has attempted to measure it with the use of 20 items such as ‘I know which factors are important to consider when making an ethnical decision’ (McMahon and Good 2016, 394) on a 6-point scale and was included into validity survey. On the basis of existing research, a correlation between value normativity and dogmatism, value relevance and political participation, value validity and authoritarianism, value stability and attitude stability, value consistency and preference for consistency, as well as value awareness and moral metacognition is theoretically plausible.

The two divergent measures include a shortform of the big five inventory (Rammstedt et al. 2013), which measures respondents personality. This measure was included to ensure, that respondents answered in regard to their value concept and not themselves or their own identity. A social desirability scale (Lück and Timaeus 2014) was also included, to see if participants answered the questions, with other peoples preferences in mind. The instrument consists of 4 items which are rated either ‘wrong’ or ‘right’.

Results

Qualitative results – the six dimensions of value conceptualisation

To assess, how people understand and relate to the value concept focus group discussions were conducted. These discussions contained more than 40 different statements on participants’ conceptualisation of values. The respondents described their values, how they perceive them and the role of values in their life. The descriptions differed greatly between individuals and revealed a variance between different characteristics ascribed to values. Through the documentary analysis six dimensions of value conceptualisation could be distinguished, which we will shortly summarise and provide exemplary interview quotes:

Firstly, values were described with differing normative standing. Some participants saw values as universal and eternal guidelines, while others described them as highly subjective. Therefore, participants seemed to differ in the general relevance their values hold, (in)dependently from themselves as individuals. We called this dimension value normativity.

P: The values we are talking about, I think, are always subjective. (…) and that’s what makes [this discussion] so difficult.

The participants also varied in describing the relevance values held in their everyday life. Interestingly, people who had difficulty with the term values, often explained that values were largely irrelevant to their day-to-day life. In contrast, other participants described in detail how their everyday experiences were related to their values (for example relationships with family or community members, sex life, work, eating habits, media consumption).

P: Coming here I thought, what an interesting discussion. Because I barely use the term values. (…) I never say, this is a value of mine, as you always hear from politicians on TV nowadays. (…) Our values and stuff like that. I actually never use it.

As a third dimension, value validity emerged. Here interviewees varied greatly in the extent to which they thought, their own values also apply to other people. While some participants employed the idea that you have to guide people towards the ‘right’ values, others were strongly opposed and stressed that everybody has to find their own subjective way, and values should not be dictated.

P: Everyone has their own values. That is a good thing, because often they are controversial. But when the EU tries to dictate what values are, every citizen’s eyebrow should raise. (…) There is a tendency that values are dictated. But then they stop being values, of course.

Fourthly, some participants perceived their values as stable in their life while others experienced changes in their own views during their lifetime. We further call this dimension value stability. Surprisingly many people revealed life-changing events that had a large impact on their values, such as the sudden death of a family member, severe accidents or illness, breaking apart of important relationships, abuse, war or similar experiences. Interestingly, in spite of the negative connotation of these events, the effect on values was seen as mostly positive, in the sense that participants realised what was important to them. Participants also mentioned that sometimes one becomes aware of his or her own values only in hindsight.

P: I had a sport accident at 14 years old, and was totally dependable on other people for half a year. Afterwards the feeling grew, that I never want to be so dependent on other people again. (…) It even inspired my occupational career, now that I think about it.

Fifthly, participants varied in their perception of value consistency, meaning that one’s values can manifest themselves as clear guidelines or are experienced as contradictory to each. One interviewee who was asked to participate in corruption at his work place, explained that this this situation put him in a dilemma, as his loyalty towards his colleagues suddenly contradicted his values of honesty. Other participants prided themselves for clear and unambiguous values.

P1: You don’t always have to be consistent, and tolerate everything. (…) you can live healthily [as a vegan] and drink alcohol, and be not consistent at all. Why should you be?

P2: Yes, that’s a good point (…)

P3: There is this proverb: you either live or you are consistent.

P: …An unbelievable value, an incomprehensible value. For every person, at every point in history, and at every place it is the same. It provides orientation, and strengthens us.

Finally, people expressed their values in differing degrees of explicitness, a phenomenon we called value awareness. Regarding the question what they understood as values, most participants struggled to answer, while some listed examples, and few provided complex abstract descriptions of the value concept.

I: …I would ask you to talk about, what this term ‘value’ is, how you understand it?

P: (6 second pause) … difficult question (laughter).

I: …we want to study values, and wonder how do people comprehend the meaning of this term?

P: Freedom … I asked myself, what is my biggest value? Freedom.

P: This image of energy comes to mind. So there is always energy, it never fully goes away, but is transformed into different forms of energy. I have a feeling that values never vanish, they are transformable into different states.

In summary, the focus groups revealed that the participant’s relation to the term value depended greatly on the characteristics described to their own values and values in general. Six dimensions of value conceptualisation could be distinguished:

1. Value normativity – seeing values as applicable only to oneself, or all humans

2. Value relevance in everyday life – relating to values in everyday situations

3. Value validity – acceptance of values being dictated to others

4. Value stability – experiencing changes in values throughout the life course

5. Value consistency – perceiving values as coherent vs. contradictory to each other

6. Value awareness – thinking consciously about values and explicitly expressing them

Quantitative results – prevalence and evaluation of the VCS

Turning to the quantitative results, the national survey aimed to see how prevalent these views were within the Austrian population and how well the scale performed psychometrically. Table 3 presents the summary statistics for each item and correspondingly each dimension of value conceptualisation. The descriptive results show (see Table 3), that a majority of Austrians disapprove of the idea, that some values are objectively right (57%). They nonetheless ascribe high importance to their values in everyday life (81%) even if they believe values should not be imposed onto others (61%). About half of the participants feel that their values have changes greatly over time (56%), while the other half thinks the same values applied over their life course. Many Austrians also perceive their values as clear guidelines in their life (66%), that they are clearly aware of (74%).

Table 3. Descriptive statistics.

		Answer 1–3	Answer 4–6	M	SD	Var.	Skew.	Kurt.	N
A	Normativity	43%	57%	3.71	1.55	2.41	−0.20	−1.05	1519
B	Relevance	19%	81%	4.51	1.16	1.35	−0.67	0.12	1519
C	Validity	39%	61%	3.91	1.34	1.80	−0.13	−0.73	1519
D	Stability	56%	44%	3.30	1.38	1.91	0.11	−0.78	1519
E	Consistency	66%	34%	3.00	1.26	1.58	0.28	−0.55	1519
F	Awareness	74%	26%	2.68	1.31	1.73	0.52	−0.50	1519

Austrian Value Formation Study (ÖWBS) 2016 (n = 1519). The frequencies displayed represent the cumulative shares of people leaning either towards statement A (answers 1–3) or statement B (answers 4–6) on each bipolar item.

Table 3 shows further descriptive analysis of the final scale. The mean of each of the six items is relatively close to the centre, and variance is above 1.3 in every item. Value relevance in everyday life (Item B) has the highest mean with 4.51 (leaning strongly towards ‘values being very important in everyday life’) and with the lowest variance.²

Table 4 presents the test-retest reliability or temporal stability of the scale³, by comparing data from the national survey and repeat survey. The table shows that Pearson product-moment correlations after one year are ranging between 0.42 and 0.47. Means of both waves are fairly similar, and around three quarters of participants’ answers were within the range of ± 1 score point. Value awareness shows the highest temporal stability overall.

Table 4. Test-retest reliabilities, means, and percent of stable answers (1 year re-test interval).

	R	M Test	SD Test	M Retest	SD Retest	% within ±1
A Normativity	.42***	3.82	1.54	3.79	1.63	68.5
B Relevance	.43***	4.55	1.10	4.66	1.12	84.4
C Validity	.42***	3.90	1.32	4.09	1.29	73.5
D Stability	.42***	3.35	1.35	3.36	1.35	74.6
E Consistency	.42***	2.97	1.22	2.91	1.23	77.7
F Awareness	.47***	2.61	1.25	2.53	1.24	81.4

Austrian Value Formation Study (ÖWBS) 2016 and retest survey (n = 1850). Only people participating in both waves are included in this comparison. 20 outlier cases were excluded.

Regarding the convergent validity of the scale in terms of theoretically related constructs: three of the six dimensions showed correlations in line with expected relationships, while the other three items did not correlate with concurrent concepts: Value normativity correlated positively (r = .257*) with the dogmatism scale, which indicates that people who show great certainty in their beliefs tend to conceptualise their values as objectively right and applying to everyone. Value consistency positively correlated with preference for consistency (r = .354***), indicating that that people who strive to be consistent also view their values as consistent rather than contradictory. Value awareness positively correlated with moral metacognition (r = .384***), thus confirming that people who are aware and knowledgeable about moral decisions are also more aware of their values. No significant correlations was found between value relevance and the political participation scale (r = .146). The interplay of values, attitudes and action is well known to be very complex. Our item might therefore accurately portray the relevance of values in everyday life, even if it does not directly correlate with attitudes towards political action. On a similar note, value validity and authoritarianism also did not correlate (r = .135) as one could have hypothesised. However, this discrepancy in our opinion does not definitively rule out the validity of our scale. It only shows that the preference for authoritarianism is not correlated with the individual expectation that one’s own values also apply to other people. There is also no correlation between value stability and attitude stability (r = −.005). As these items are quite similar on a conceptual level, the outcome should be carefully reflected and possibly examined again within a larger survey.

A comparison of the VCS with discriminant validity measures showed no or comparatively small correlations. Social desirability showed slight correlation only with value awareness (r = .250*), but all five other correlations were insignificant (r ranging between .023 and .136). Each of the big five inventory items was also tested against all six VC dimensions. Of these possible 30 correlations 27 were insignificant and small (r ranging between .17 and .202). Three significant relations were revealed, namely between value normativity and neurocentrism (r = −0.211*), value validity and openness (r = 0.303**), as well as value consistency and conscientiousness (r = 0.240*). These results suggest that discriminant validity exists between social desirability, personality and value conceptualisation.

Discussion

The study at hand utilised a mixed methods approach, to gain a nuanced understanding of the ways in which people understand and relate to the concept of values. Conducting seven focus group interviews, we  were able to explore a large variety of value conceptualisations. Using the transcripts of these focus group discussions, through documentary analysis (Bohnsack 2010) six underlying dimensions of value conceptualisation were distinguished. Each dimension represents a spectrum of views on the nature of values. The dimensions are value normativity (values are seen as applying only to oneself or all humans), value relevance (how often values are referred to in everyday situations), value validity (degree of acceptance for dictating values to others), value stability (previously experienced value changes), value consistency (perceived contradiction or coherence of one’s values), and value awareness (ability to express or explicitly think about values).

The study then developed a brief tool to systematically assess each of these dimensions, the Value Conceptualisation Scale. Applied to a large and representative sample for Austria, the results showed considerable variation of each dimension of value conceptualisation among the population. Findings indicate that some people viewed their values as explicit, stable, and universally applicable guidelines in their life, while others saw them as contradictory, changing, and highly subjective. In contrast to many theoretical contributions, that ascribe values relatively stable characteristics (e.g. Joas 2013; Rohan 2000; Rokeach 1973; Thome 2003) this large variation on the meaning and conceptualisation of values by everyday citizens is surprising. These findings contradict the notion that values are a homogenous or universally perceived phenomenon, that only vary in what they prioritise as desirable. Rather, values seem to also differ in the relevance, stability, normativity, consistency, validity and awareness that individuals experience.

Additionally to these empirical findings, the paper addressed the evaluation of the VCS. Each item showed considerable response variety, while the means remained close to the centre. This indicates that the items response options (6 answer bipolar scale) are covering all important answers (DeVellis 2017, 122f). The item value relevance in everyday life was performing a bit worse than the other items, with its median leaning strongly towards ‘values being very important in everyday life’ and with the lowest variance. This finding might be the results of a sample bias since the survey was mainly concerned with values, and could be further investigated in a broader population survey.

The test-retest reliability, which was conducted with 62 percent of the original sample, was satisfactory taking into regard the one-year interval (compare Braithwaite and Law 1985; Deth 1983; Evans, Heath, and Lalljee 1996; Locke 2000; Park and Peterson 2006; Schwartz 2003). Here, our findings suggest that value conceptualisations stay relatively stable even over longer periods of time.

Concerning validity, an effort was made to explore the boundaries of value conceptualisation with the use of discriminant and concurrent validity measures. The scale reported good discriminant validity, when compared to the big five inventory and social desirability measures. The results of convergent validity testing were consistent with our initial hypotheses in three of the six dimensions, although more theoretical work and comparable measures are clearly needed. Despite that, the results point towards interesting relationships between VCS and dogmatism (Altemeyer 2002), preference for consistency (Collani and Blank 2013), and moral metacognition (McMahon and Good 2016) which should be further explored. Our findings suggest that people who show great certainty in their beliefs tend to conceptualise their values as objectively right and applying to everyone. People who strive to be consistent also view their values as consistent rather than contradictory. And finally, that people who are aware and knowledgeable about moral decisions are also more aware of their values.

The present study showed some limitations that can be addressed by future research. Preliminary tests could not include a larger number of redundant items, which later prevented internal consistency evaluation. Reliability testing would have benefitted from multiple retest waves to further isolate the scales reliability from its temporal stability (Heise 1969). The biggest limitation in terms of validity investigation is without a doubt the novelty of the scale. Previous theoretical as well as empirical work on value conceptualisation is extremely sparse, and we therefore lack resilient reference points that lie beyond our own research programme. Another general critique could refer to the high level of abstraction demanded by the VCS. Our focus on value characteristic in general (e.g. stability of values) increase the scales usefulness in comparative studies, but the application of the scale to specific values (e.g. stability of importance of security) could provide important insights into value-specific differences within value conceptualisation. The VCS development also provides opportunities for future research to study sociodemographic differences in the value conceptualisation of social groups (e.g. do religious vs. nonreligious people have more stable values?), national or longitudinal comparison.

In conclusion, the VCS takes a new perspective by looking at the way people conceive and relate to their own values. The scale can be applied to a variety of different topics within value research, and on the grounds of initial evaluation, we recommend it for future research. Continued use of this scale will help to assess the relevance of value conceptualisation for the beliefs and actions that shape everyday life.

Acknowledgments

The authors thank Julian Aichholzer (University of Vienna), Johannes Gieseke (HU Berlin) and Martin Kroh (DIW Berlin) for advising the research process with their expertise. We also thank Johann Bacher (JKU Linz) and Susanne Vogl (University of Vienna) for comments that greatly improved the manuscript. Lastly, we thank the anonymous reviewers of this journal for their valuable insights that helped improve this article.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Lena Seewann

Lena Seewann is a research assistant at the University of Potsdam. Her PhD theses is part of the research project ‘Forming Values: Contents – Places – Processes’. Her research interests include methods of the social sciences, religion and human values, migration and social inequality.

Roland Verwiebe

Roland Verwiebe is Professor for Social Stratification and Inequality Research at the University of Potsdam. His research interests include social inequality, migration, labour market, poverty, values and attitudes.

Notes

1. This was due to limitations within the research project, wherein the first surveys only provided limited space to test this newly developed scale. The small size of the item pool meant that later tests in item covariance were not possible, which hinders the selection of best performing items. The mixed methods approach counteracted this problem to a degree, as focus group material helped identify, which formulations more commonly and densely described certain views. Future projects should nevertheless explore more variations of items to possibly further improve the wording of the scale.

2. Results of this preliminary test may be biased by a intense public debate on values during this time in Austria, due to the influx of Syrian refugees in the country, and political events connected to it. Further elaboration on the intended meaning of this item should be considered for future research.

3. General reliability testing with cronbach alpha yielded an overall value of 0.23, as is expected, because the items measure different dimensions of value conceptualisation. Only Items E and F showed a higher correlation (0.59) indicating a possible relationship between value awareness and value consistency.