Schlotz, Wolff; Wallot, Sebastian; Omigie, Diana; Masucci4, Michael D.; Hoelzmann, Sonja C. and
Vessel, Edward A.. 2020. The Aesthetic Responsiveness Assessment (AReA): A screening tool
to assess individual differences in responsiveness to art in English and German. Psychology of
Aesthetics, Creativity, and the Arts, ISSN 1931-3896 [Article] (In Press)
https://research.gold.ac.uk/id/eprint/29310/
The version presented here may differ from the published, performed or presented work. Please
go to the persistent GRO record above for more information.
If you believe that any material held in the repository infringes copyright law, please contact
the Repository Team at Goldsmiths, University of London via the following email address:
gro@gold.ac.uk.
The item will be removed from the repository while any claim is being investigated.
more information, please contact the GRO team: gro@gold.ac.uk
For
AESTHETIC RESPONSIVENESS
1
1
The Aesthetic Responsiveness Assessment (AReA): A screening tool to assess individual
2
differences in responsiveness to art in English and German
3
Wolff Schlotz* 1, 2, Sebastian Wallot* 1, Diana Omigie1, 3, Michael D. Masucci4, Sonja C.
4
Hoelzmann1, 2, and Edward A. Vessel1
5
6
7
8
1
Max Planck Institute for Empirical Aesthetics, Frankfurt am Main, Germany
2
Insitute of Psychology, Gothe-University Frankfurt am Main, Germany
3
Department of Psychology, Goldsmiths, University of London, UK
4
Department of Psychology, New York University
9
Author Note
10
* These authors contributed equally to this article and hence share first authorship
11
Sebastian Wallot,
12
Wolff Schlotz,
https://orcid.org/0000-0003-2356-7766
13
Diana Omigie,
https://orcid.org/0000-0003-1600-0112
14
Michael D. Masucci,
https://orcid.org/0000-0002-9151-7253
15
Sonja C. Hoelzmann,
https://orcid.org/0000-0002-7655-5884
16
Edward A. Vessel,
17
Michael D. Masucci is now at Louisiana State University, Louisiana
18
We have no known conflict of interest to disclose.
19
The authors would like to acknowledge Amy Belfi, Anna Kasdan, Gabrielle Starr and
20
Jonathan Stahl for help with data collection, Christine Knoop for substantially contributing to
21
the translation of the AReA into German, and Kirill Fayn for commenting on an earlier
22
version of the manuscript.
23
https://orcid.org/0000-0002-3626-3940
https://orcid.org/0000-0001-8733-1731
Correspondence concerning this article should be addressed to Sebastian Wallot, Max
24
Planck Institute for Empirical Aesthetics, Grueneburgweg 14, 60322 Frankfurt am Main,
25
Germany. E-mail: sebastian.wallot@ae.mpg.de
AESTHETIC RESPONSIVENESS
26
2
Abstract
27
People differ in how they respond to artworks. Measuring such individual differences is
28
helpful for explaining response variability and selecting particularly responsive sub-samples.
29
On the basis of a sample of items indicating relevant behavior and experience, we
30
exploratively constructed the Aesthetic Responsiveness Assessment (AReA), a screening tool
31
for the assessment of individual differences in responsiveness to art in English and German.
32
Exploratory and confirmatory factor analyses suggested three first-order factors labeled
33
aesthetic appreciation, intense aesthetic experience, and creative behavior, and a second-order
34
factor aesthetic responsiveness. Aesthetic responsiveness was assessed in N = 781 participants
35
from the United States and Germany, and measurement invariance analysis demonstrated full
36
metric and partial scalar invariance across language versions. AReA scale scores yielded good
37
reliability estimates. Validation studies confirmed expected associations between AReA scale
38
scores and measures of related constructs, as well as continuously and retrospectively
39
recorded responses to music, visual art, and poetry. In summary, the AReA is a promising,
40
psychometrically evaluated instrument to assess aesthetic responsiveness built on a mixture of
41
exploratory and confirmatory construction strategies. It can be used as a screening tool both in
42
English and German speaking samples.
43
44
45
Keywords: aesthetic responsiveness, creative behavior, aesthetic experience, screening
scale, validity, measurement invariance
AESTHETIC RESPONSIVENESS
3
46
The Aesthetic Responsiveness Assessment (AReA): A screening tool to assess individual
47
differences in responsiveness to art in English and German
48
There exist individual differences in responsiveness to many different types of
49
information (e.g. to visual brightness, auditory loudness, taste, social or emotional cues), and
50
responsiveness to aesthetic stimuli is no exception. Indeed, aesthetic experiences would
51
appear to be a domain where individual differences in responsiveness are rather large. We
52
may all call to mind individuals whose responsiveness is different than our own: for instance,
53
a colleague may report that they generally don’t get pleasure from visiting museums, or from
54
listening to music. In contrast, we may know other individuals whose level of aesthetic
55
responsiveness to a particular art form is so strong as to be wholly out of our level of
56
understanding.
57
As experimentalists interested in studying the psychological and neural basis of
58
aesthetic experiences, this heterogeneity in aesthetic responsiveness presents a distinct
59
problem. If a large proportion of the potential observers that we sample from the general
60
population do not respond to our stimuli, this may result in inconclusive findings. While at
61
least a portion of variability may reflect individual preferences for specific aesthetic domains
62
or styles, part of this variability likely also reflects trait-level differences in overall aesthetic
63
responsiveness. Here, we present a screening tool developed with the goal of providing a
64
quick assessment of (overall) aesthetic responsiveness.
65
We define aesthetic responsiveness here as the individual capacity to respond to
66
aesthetic stimuli. This definition is mainly based on the notion that aesthetic responses have a
67
common origin in brain areas that mediate responses across different domains, particularly
68
neural systems involved in emotion and reward processing (Berlyne, 1971; Chatterjee &
69
Vartanian, 2016; Vessel et al., 2019). These neural systems can affect peripheral responses
70
via connections with the autonomic nervous and neuroendocrine systems that link central
AESTHETIC RESPONSIVENESS
71
nervous system activity with peripheral physiological responses (Lane et al., 2009). This
72
conceptualization of aesthetic responsiveness implies some sort of generality, such that
73
individual differences in responsiveness may exist across aesthetic domains, response
74
domains (cognitive, emotional, behavioral, and physiological), and time (e.g., repeated
75
exposure). However, this does not rule out stimulus specificity whereby aesthetic stimuli of
76
different domains may result in systematically different aesthetic experiences, for example
77
due to perceptual modality-dependent processing (cf. Jacobsen & Beudt, 2017). In addition,
78
we acknowledge here that some response variance is likely to be due to individual-specific
79
responses, i.e. patterns of responses that differ systematically between individuals (Vessel et
80
al., 2018).
81
We assume that aesthetic responsiveness is a dispositional tendency that generates
82
individual differences in responses to aesthetic stimuli. These individual differences are
83
assumed to be relatively consistent over time and across aesthetic domains, as well as
84
coherent across response domains. It is assumed that individuals with a high aesthetic
85
responsiveness trait level experience aesthetic cognition, emotion and related physiological
86
effects more frequently and more intensively than others, and that they show a greater
87
behavioral propensity towards engagement with art.
88
The construct of aesthetic responsiveness is related to constructs focusing on
89
individual differences in the appreciation of, or engagement with beauty (Diessner et al.,
90
2018; Diessner et al., 2008; Haidt & Keltner, 2004), particularly if appreciation is conceived
91
as a cognitive-emotional, and engagement as an emotional reaction to beauty (Güsewell &
92
Ruch, 2012). However, aesthetic responsiveness differs from these constructs in a number of
93
aspects. First, it focuses on responses to aesthetic stimuli and excludes non-aesthetic stimuli
94
such as talent, virtue, or morality. Second, it explicitely distinguishes between response
95
domains, providing a background for more fine-grained predictions of domain-specific
4
AESTHETIC RESPONSIVENESS
96
responses. Finally, aesthetic responsiveness does not exclusively focus on beauty; it includes
97
responses to aesthetic stimuli that are not necessarily perceived as beautiful.
98
99
5
Regarding associations of aesthetic responsiveness with personality factors, openness
to experience (or open-mindedness) seems to be particularly relevant. Findings from
100
empirical aesthetics studies investigating openness demonstrate that personality is predictive
101
of indicators of aesthetic experience (Fayn et al., 2015; McCrae, 2007; Rawlings et al., 2000;
102
Silvia et al., 2015). Openness has also been linked with aesthetic activities and positive
103
aesthetic attitudes (McManus & Furnham, 2006). Measurements of aesthetic responsiveness
104
should therefore show strong associations with measurements of openness. In comparison to
105
constructs of major taxonomies of personality traits, aesthetic responsiveness is closely
106
linked, conceptually, with a specific facet related to aesthetic experience which is located in
107
the lower level structure of the factor openness. This facet has been labelled aesthetics (Costa
108
& McCrae, 1995), aesthetic sensitivity (Soto & John, 2017), or aesthetic appreciation (Ashton
109
& Lee, 2007). However, openness additionally comprises a number of facets that are not part
110
of the construct of aesthetic responsiveness. For example, a detailed analysis found five facets
111
of openness in addition to the facet aesthetics which have been labeled intellectual efficiency,
112
ingenuity, curiosity, tolerance, and depth (Woo et al., 2014). While these lower level facets
113
can be expected to be empirically related to aesthetic responsiveness, they clearly reflect
114
different constructs. Thus, while aesthetic responsiveness is thought to be similar to the
115
openness facet aesthetics, openness is a much broader construct comprising facets that are
116
clearly distinguishable from aesthetic responsiveness both empirically and with regard to
117
content.
118
As opposed to the concept of aesthetic sensitivity, which has historically been
119
identified as the degree to which an individuals’ aesthetic judgments agree with an externally
120
defined standard (Child, 1964; Eysenck, 1940), aesthetic responsiveness is defined by the
AESTHETIC RESPONSIVENESS
121
strength of the response, regardless of an individual’s subjective sense of taste. Therefore,
122
evaluative constructs as assessed by aesthetic sensitivity tests should be empirically
123
distinguishable from aesthetic responsiveness as well as related constructs such as the
124
personality factor openness. In line with this assumption, individual scores on the Visual
125
Aesthetic Sensitivity Test (Götz et al., 1979), a measure of aesthetic sensitivity, showed only
126
a modest correlation with the openness facet scale Aesthetics (Myszkowski et al., 2014).
127
As a more convenient alternative to a complete assessment of aesthetic responsiveness
128
across all possible aesthetic domains and response domains (e.g. behavioral, physiological,
129
emotional, cognitive), we present a self-resport assessment tool of how individuals have
130
perceived their responses in different stimulus and response domains in their daily life. This
131
approach is particularly useful for screening for individual aesthetic responsiveness in
132
research settings that do not allow for rigorous and comprehensive testing that encompasses
133
all domains.
134
6
Similar scales have been developed for different aesthetic domains, and represent
135
different aspects of aesthetic responsiveness to a greater or lesser degree (Hager et al., 2012;
136
Rowold, 2008; Stamatopoulou, 2004). This includes a recent scale that provides a very fine-
137
grained assessment of aesthetic-emotional responses (Schindler et al., 2017). The measure
138
that reflects a construct most closely related to aesthetic responsiveness is the Engagement
139
with Beauty Scale (EBS; Diessner et al., 2008), which itself is related to the Appreciation of
140
Beauty and Excellence (ABE) subscale of the Values in Action Inventory of Strengths (VIA-
141
IS; Peterson & Seligman, 2004). However, the EBS focuses exclusively on the experience of
142
beauty and is designed to measure engagement with beauty across natural, artistic, and moral
143
domains. This wider scope is not a good match for a more focused assessment of aesthetic
144
responsiveness. Additionally, the EBS does not separate out aesthetic responsiveness to
145
different artistic domains, nor does it assess behavioral indicators of art appreciation. Taken
AESTHETIC RESPONSIVENESS
146
together, none of the existing instruments assesses the breadth of aesthetic responsiveness
147
specific to artworks as defined above with a short scale that can be used for screening
148
purposes.
149
7
We will here present rationale and choices of constructing a scale for the assessment
150
of aesthetics responsiveness that assesses individual responses to aesthetically relevant stimuli
151
from a broad variety of different domains. We present analyses of psychometric properties of
152
two language versions of the scale, English and German. In the subsequent sections, we
153
present results from a number of studies that provided data we used for validation of the scale,
154
namely correlations of scale scores with individual responses to visual art, music, and poetry,
155
as well as with measures of related personality constructs. Finally, a validation study will be
156
presented, where participants filled in the resulting scale together with a measure of the Big
157
Five personality domains and their factes; the analysis focuses on correlations of scale scores
158
with openness and its facets.
159
Scale Construction
160
With a focus on research participant screening for aesthetic responsiveness, an 18 item
161
short scale was developed in the English language, assessing typical responses to and
162
engagement with a variety of aesthetic stimuli, and with an emphasis on visual aesthetic
163
experiences to reflect that a large proportion of art has a visual component (painting,
164
sculpture, dance, film, etc.). Due to the self-report format, the scale assesses perceived (self-
165
evaluated) aesthetic responsiveness, reflecting typical and daily life aesthetic experiences. The
166
items were designed with the aim of assessing general or aggregate experiences, in contrast to
167
focusing on single episodes.
168
One goal of scale construction was to reflect the centrality of “beauty” as a core
169
domain-general aesthetic emotion term (Istok et al., 2009; Jacobsen et al., 2004; Menninghaus
170
et al., 2019) but also to acknowledge that this is not the only path to positive aesthetic
AESTHETIC RESPONSIVENESS
171
experiences, and that research participants often misinterpret “beauty” to refer to objective
172
stimulus traits rather than as an emotional responding arising from the interaction of a
173
perceiver with an object (Reber et al., 2004; Vessel et al., 2012).
174
8
Another key goal of scale construction was to distinguish between those individuals
175
who regularly respond to artworks in an intense way from those who rarely experience more
176
than a commonplace appreciation of aesthetic objects in everyday life. Recent empirical work
177
suggests a potential difference between more everyday positive experiences of beauty and a
178
subset of more intense aesthetic experiences (e.g. “being moved”, “awe”, the “sublime”;
179
(Brielmann & Pelli, 2017; Omigie et al., 2019; Pelowski et al., 2017; Vessel et al., 2012,
180
2013).
181
Such work parallels accounts in the philosophical literature that pit feelings of beauty
182
against those of the sublime (Burke, 1757/2015). In the context of music, for instance, beauty
183
experiences “in which tension and discord have at most a minor place” have been
184
distinguished from other forms of beauty, that may, instead, confront or challenge (Levinson,
185
2012, p. 128). Here, we sought to extend, to the individual differences level, this notion of a
186
distinction in the types of aesthetic states that are possible. We propose that a scale that is able
187
to reveal those individuals that regularly respond to artworks in an intense way would allow
188
experimenters to better account for much variability in responses observable in their data.
189
Another goal of scale construction was to differentiate individuals who actively
190
occupy themselves with the creation of aesthetically relevant products from those who do not.
191
Although creative behavior does not reflect aesthetic responsiveness at the same level as
192
appreciation of aesthetic objects does, we assume that individuals high in aesthetic
193
responsiveness have a higher propensity to actively engage in goal-directed creative processes
194
such as writing, painting, or making music. On the one hand, this is based on well-established
195
associations between openness and creativity (Puryear et al., 2017), suggesting that openness
AESTHETIC RESPONSIVENESS
9
196
contributes substantially to an individual’s creative potential. On the other hand, the link of
197
creative potential with actual creative behavior is assumed to be moderated by a number of
198
factors, suggesting that creative potential can or cannot lead to creative behavior (e.g.
199
Karwowski & Beghetto, 2019). We assume that individuals high on aesthetic responsiveness
200
have a higher creative potential, and that creative behavior is therefore linked with aesthetic
201
responsiveness. However, this link is thought to be moderate, as other factors influence
202
creative potential and its effect on creative behavior. We added items on creative behavior to
203
the scale, thereby broadening the scope of the construct measurement. While emotional,
204
cognitive, and physiological responses to aesthetic stimuli were covered by many items,
205
behavioral indicators of aesthetic responsiveness were represented less well. Therefore,
206
including items assessing creative behavior brings the representation of indicators of different
207
construct-relevant responses to a similar level. While creative behavior seems to be a rather
208
distal indicator of aesthetic responses, it should be kept in mind that it requires continued
209
preoccupation with aesthetically relevant material and therefore reflects an individual’s
210
receptiveness for such material. The inclusion of items related to creative behaviour also
211
aimed to achieve more precise measurements by separating variance components indicating
212
different facets of aesthetic responsiveness. Moreover, adding creative behavior items might
213
be particularly relevant for selecting participants for studies focusing on creative behavior,
214
and therefore potentially increase the utility of the scale.
215
We began by modifying several items from the EBS reflecting experiences with
216
artworks and expanding these into a set of eight questions reflecting either beauty or intense
217
aesthetic experience, across four response domains: cognition (items 3, 16), physiological
218
arousal (items 8, 10), conscious emotion (18, 13) and spirituality/transcendence (items 5, 14).
219
Next, a set of five questions were added to assess aesthetic appreciation of different domains:
220
poetry (item 1), fiction (item 7), music (item 4), architecture (item 11) and nature (item 15).
AESTHETIC RESPONSIVENESS
10
221
Lastly, a set of five items were added to assess behavioral indicators of aesthetic
222
responsiveness; one assessing attendance to museums or performances (item 2) and four
223
probing levels of creative behavior across the domains of writing (item 9), visual arts (item 6),
224
music (item 4) and education (item 12), which we assume to be strongly related to aesthetic
225
responsiveness. To record and score responses, a frequency scale with five categories from
226
“never” to “very often” was implemented. A full list of the 18 items of the original version
227
can be found in the online supplemental material. In sum, aesthetic responsiveness was
228
operationalized as an individual’s perceived frequency of aesthetic experiences as indicated
229
by a variety of cognitive and affective states, responses, and behaviors.
230
This scale construction process emphasizes both, a common origin of aesthetic
231
responses (i.e. aesthetic responsiveness), and multiple facets of aesthetic responsiveness,
232
namely appreciation of aesthetic stimuli, intense aesthetic experiences, and creative behavior.
233
However, it is important to note that the construction of the assessment instrument and its
234
empirical applications were not intended to explore qualitatively different theoretical models
235
of aesthetic experience and its precursors, moderators, mediators, and consequences; or to
236
compare aesthetic responsiveness with aesthetic sensitivity; or to differentiate theoretically
237
refined constructs of the aesthetic process such as aesthetic appreciation, engagement, or taste.
238
The level of detail required for such an investigation and subsequent analysis of the
239
nomological network is beyond the scope of this paper.
240
With the aim of broadening the applicability of this scale, all items were translated to
241
German language by two bilinguals following widely used guidelines (van de Vijver &
242
Hambleton, 1996). Translations were discussed with one of the developers of the English
243
language original scale with regard to differences and similarities in semantic content. The
244
resulting German language version was used in several research projects at the Max Planck
245
Institute for Empirical Aesthetics in Frankfurt am Main, Germany.
AESTHETIC RESPONSIVENESS
246
11
The major aims of this study were (a) to explore and confirm the dimensionality of the
247
scale; (b) to test for measurement invariance of the resulting scale across the English and
248
German language versions; (c) to report scale score descriptive statistics and estimate the
249
reliability of scores of the final scale; and (d) to explore the validity of scale scores using
250
measures of constructs related to aesthetic responsiveness, and investigate associations with
251
responses to specific aesthetic stimuli, namely visual art, poems and music.
252
253
Method
Samples
254
U.S. sample. 285 undergraduate students filled in the scale as part of a battery of tests
255
and questionnaires administered at the beginning of an introductory psychology course at
256
New York University. The battery was completed as an online web survey within the first
257
week of the semester at a time and place of the participants' choosing. Consent was obtained
258
via an online consent form, and all study procedures were approved by the NYU institutional
259
review board. Four cases were excluded as they did not provide any data on the scale. Thus,
260
the final sample comprised 281 participants, 198 (70%) females. The mean age of participants
261
was 18.9 years (SD = 1.1), ranging from 16 to 24 years. One missing item response from one
262
participant was imputed using the item sample mean. All participants had completed high-
263
school.
264
German sample. The German sample consisted of two subsamples. German
265
subsample 1 was a convenience sample of participants from a study on music listening
266
behavior. For this study, 202 participants were recruited, of which 31 did not provide any
267
responses on the aesthetic responsiveness scale, and one had 78% missing responses.
268
Removing these participants resulted in a final sample of 170 participants, 118 females (69%)
269
(7 participants, 4%, did not respond), with a mean age of 31.1 years (SD = 12.5; range: 18 to
270
75 years); 73 (43 %) had completed a university degree.
AESTHETIC RESPONSIVENESS
271
12
German subsample 2 was a convenience sample from a study of poem reading. After
272
the reading study, participants filled in the aesthetic responsiveness scale as part of a larger set
273
of questions. The sample consisted of 123 participants, 92 (75%) females, with a mean age of
274
25.0 years (SD = 5.1; range: 18 to 43 years); 54 (44 %) completed a university degree.
275
German subsamples 1 and 2 were pooled into a German total sample comprising 293
276
participants, 210 (72%) females (7 participants, 2%, did not identify as one of the sexes), with
277
a mean age of 28.3 years (SD = 10.7).
278
In addition, the final version of the AReA was applied in a validation study
279
comprising 207 participants, 124 (60%) females (1 participant, 0.5% did not identify as one of
280
the sexes), with a mean age of 49.9 years (SD = 16.2).
281
282
283
Adding up across countries, the total sample size for this study was N = 781.
Measures
All participants filled in the 18 items of the original version of the aesthetic
284
responsiveness scale, except for validation study 4 where the final 14-item version was filled
285
in. In addition, we used responses on sample-specific scales relevant for validation of the
286
AReA. Measures used for validation studies are described in the respective sections.
287
Data analysis
288
Item development aimed at emphasizing a common factor underlying responses to all
289
items on the one hand, and multifacetedness of responses with regard to general appreciation,
290
intensity, and creativity, on the other hand. We therefore first analyzed heterogeneity of the
291
items using basic item characteristics such as item-rest correlations (IRC) and inter-item
292
correlations to eliminate single items that clearly did not show satisfactory associations with
293
the other items and were therefore not compatible with the assumption of a single common
294
factor. With the aim of identifying items with invariant measurement characteristics in both
295
samples, this was done separately for the US and the German sample. We then split the
AESTHETIC RESPONSIVENESS
13
296
sample randomly by language version into two subsamples, each comprising half of the US
297
and German total sample (random sample 1 and 2; n = 287 each). Using random sample 1, the
298
remaining items were subjected to a parallel analysis based on principal components analysis
299
(PCA) to explore potential dimensional heterogeneity and determine the number of factors to
300
be extracted. We extracted the number of factors estimated ±1 (cf. Lim & Jahng, 2019) and
301
subjected the items to a maximum-likelihood exploratory factor analysis (EFA) with oblique
302
oblimin rotation. We evaluated solutions on the basis of interpretational validity and clarity of
303
the simple structure of rotated factor loadings.
304
To check for stability of the factorial structure across random samples, we tested
305
second-order confirmatory factor analysis (CFA) models in random sample 2. If the EFA
306
suggested a multiple factor solution, these factors were represented in the CFAs as first-order
307
factors which loaded on a common second-order factor Aesthetic Responsiveness. For testing
308
fit of the factorial structure in random sample 2, we ran the following model sequence: First,
309
we tested CFA models separately in the US and German sample to evaluate if the factorial
310
structure showed an acceptable fit in each language version. We used comparative fit index
311
(CFI) and Tucker-Lewis index (TLI) close to .95 or higher, a standardized root-mean-square
312
residual (SRMR) close to .08 or lower, and a root-mean-square error of approximation
313
(RMSEA) close to .06 or lower, as targets for acceptable model fit in accordance with Hu and
314
Bentler (1999). We then proceeded to test for configual, metric, and scalar measurement
315
invariance (Chen et al., 2005; Millsap, 2011) between the English and German language
316
versions of the scale by comparing model fit for the US sample and the pooled German
317
sample from random sample 2. Configural invariance assumes equal factorial structures in
318
both groups. For model identification, the loading of the first measured variable on each latent
319
factor was fixed to one, the latent common first-order factor means fixed to zero, and
320
intercepts, latent factor variances and covariances freely estimated. Metric invariance
AESTHETIC RESPONSIVENESS
14
321
additionally assumes equal factor loadings in both groups. Model specification was the same
322
as for the configural invariance model, except that, first, all first-order factor loadings were
323
constrained to be equal across groups; second, all second-order factor loadings were
324
constrained to be equal. Scalar invariance additionally assumes equal item intercepts. Model
325
specification was the same as for the metric invariance model, except that, first, all item
326
intercepts were constrained to be equal across groups, and the second-order latent factor mean
327
was freely estimated in the German sample, and, second, the second-order factor mean was
328
constrained to be equal between the groups. If one of the invariance assumptions did not
329
hold, we tested for partial invariance by relaxing equality constraints for those parameters that
330
showed substantial modification indices.
331
Although we report chi-square differences (Δc²) for all model comparisons, our
332
decisions on measurement invariance were based on differences in approximate fit indices, as
333
Δc² is highly sensitive to sample size. In particular, differences in CFI (ΔCFI), RMSEA
334
(ΔRMSEA), and SRMR (ΔSRMR) between models with increasing restrictions were used to
335
assess each level of measurement invariance. In the case of metric invariance, changes of
336
ΔCFI ≤ -.010, ΔRMSEA ≥ .015, and ΔSRMR ≥ .015 would indicate non-invariance as
337
suggested by Cheung and Rensvold (2002) and Chen (2007). In the case of scalar invariance,
338
ΔSRMR ≥ .010 would indicate non-invariance, with the other criteria being the same as for
339
metric invariance, as suggested by Chen (2007).
340
We then compared factor scores and scale mean scores between language versions in
341
the combined random samples. Note that factor scores, i.e. latent mean differences, can be
342
meaningfully compared between groups even in the case of partial scalar invariance, whereas
343
composite scores (i.e. differences of mean or sum scores) are biased if full measurement
344
invariance does not hold (Steinmetz, 2013). Nevertheless, studies applying psychometric
345
scales often prefer composite scores over factor scores. Composite reliability was separately
AESTHETIC RESPONSIVENESS
15
346
estimated for the two versions using coefficient omega (McDonald, 1999), which is
347
appropriate for unit-weighted scoring of congeneric scales (McNeish, 2018). Finally, we
348
investigated construct validity of the resulting scale using Pearson correlation coefficients
349
with relevant experimental data and other self-report scales related to the construct of
350
aesthetic responsiveness.
351
All models were based on continuous indicator variables using a maximum likelihood
352
estimator with standard errors and a mean-adjusted c² test statistic (MLM) that are robust to
353
non-normality of indicator variable distributions.1 CFAs and composite reliability calculations
354
were performed using Mplus (Version 7.3); EFAs, parallel analysis, factor extraction and
355
rotation, item, scale and some validity anaylses were performed using Stata (Version 15.1);
356
the remaining validity analyses were performed using R (Version 3.4.0).
357
358
359
Results
Item selection and factor analyses
Although the items were designed to indicate different facets of a disposition to
360
respond to aesthetic stimuli, we assumed that they share variance attributable to a common
361
underlying factor, i.e. aesthetic responsiveness. We therefore expected all items to show
362
relatively high associations with the scale score minus the item itself, i.e. IRC, and at least
363
medium inter-item correlations. Sample-specific IRCs as well as average inter-item
364
correlations were higher in the English language version than in the German language version
365
(see Tables S1 and S2 in the supplementary material for details); three items showed very
366
weak IRCs of less than .30 in the German language version, one of which was also very weak
367
in the English language version. We therefore excluded these items (number 7, 15, and 17 of
368
the original scale, cf. Tables S1/S2) from the scale. This increased the average inter-item
1
We have also tested CFA models for ordered-categorical factor indicators separately for the English
and German language version. As these models yielded similar fit to the data as the models for continuous
indicators, we used the more straightforward continuous indicator CFA models for measurement invariance
analysis.
AESTHETIC RESPONSIVENESS
369
correlations considerably to .46 in the English language and to .35 in the German language
370
version, bringing the whole scale closer towards a more homogenous item sample.
371
16
The resulting 15 items were subjected to a parallel analysis using random sample 1
372
(both language versions together). Parallel analysis suggested extraction of two factors
373
(Eigenvalues PCA: 6.91; 1.37; 1.09; Eigenvalues parallel analysis: 1.41; 1.32; 1.25). We
374
therefore compared rotated factor solutions with one, two, and three factors. Both, the two-
375
and three-factor solutions clearly separated a creative behavior factor. The three-factor
376
solution provided a clearer simple structure and an interpretable third factor, although one
377
item did not fit with the content of the creative behavior factor despite a high factor loading.
378
This was likely due to confounding content (“I enjoy poetry”, while poetry and writing was
379
also prominently represented in two other items loadings on the creative behavior factor). We
380
therefore decided to remove this item and rerun the analysis, resulting in a clear and
381
interpretable simple structure with three factors. Factor 1 represented aesthetic appreciation,
382
factor 2 strong/intense emotional responses to art exposure, and factor 3 different aspects of
383
producing art. One item (“I am deeply moved when I see art”) cross-loaded on the factors
384
representing aesthetic appreciation and intense aesthetic experience. The correlations between
385
the factors were: rf1,f2 = .67, rf1,f3 = .48, rf2,f3 = .46.
386
To check stability of the factorial structure across random samples, we conducted
387
second-order CFAs using random sample 2. CFA models were fitted separately for the
388
English and German language versions. The CFA model showed an acceptable fit to the data
389
in both, the English language (χ² = 112.6; df = 73; p = .002; RMSEA = 0.062, 90% CI: 0.038,
390
0.084; CFI = 0.965; TLI = 0.957; SRMR = 0.050) and German language version (χ² = 119.6;
391
df = 74 (the residual variance of one first-order factor in the German sample had a small
392
negative estimate and was therefore set to zero); p = .001; RMSEA = 0.065, 90% CI: 0.042,
393
0.086; CFI = 0.946; TLI = 0.933; SRMR = 0.050). These results provide support for the
AESTHETIC RESPONSIVENESS
394
17
validity of the factorial structure across different samples.
395
In sum, the 3-factor model provided the best mixture of good model fit,
396
parsimoniousness, and interpretability, and it was confirmed in an independent random
397
sample using second-order CFAs. The final scale was named Aesthetic Responsiveness
398
Assessment (AReA), comprising the sub-scales Aesthetic Appreciation (AA), Intense
399
Aesthetic Experience (IAE), and Creative Behavior (CB), loading on a second-order factor
400
Aesthetic Responsiveness (AReA total). Both language versions of the final scale can be
401
found in the supplementary material to this article.
402
Measurement invariance across language versions
403
We tested the final second-order CFA model for configural, metric, and scalar
404
measurement invariance across the English and German language versions using the US and
405
the pooled German sample. As can be seen from Table 1, the configural invariance model
406
yielded acceptable model fit indices. Comparing fit indices of the model with equal first-order
407
factor loadings to the configural invariance model showed that changes of RMSEA, CFI, and
408
SRMR were minimal and within or close to the pre-defined cut-off values. In addition, all
409
model fit indices suggested a good fit of the metric model. The second-order metric
410
invariance model showed very small deviations from the first-order meric invariance model.
411
We therefore concluded that these results clearly suggest full metric invariance across the
412
English and German language versions of the AReA. In contrast, the test of scalar invariance
413
of observed indicators yielded model fit indices that were clearly beyond pre-defined cut-off
414
values for model fit as well as fit difference to the metric invariance model. Inspection of
415
modification indices suggested that this was due to item intercept equality constraints for few
416
items. Lifting equality constraints for three items (see Table 1 for details) resulted in an
417
acceptable model fit as well as fit-index differences that were within or very close to the pre-
418
defined range for demonstrating scalar invariance of observed indicators. Testing scalar
AESTHETIC RESPONSIVENESS
18
419
invariance of first-order factors showed very small deviations from the observed-indicator
420
scalar invariance model. These results suggest that the English and German language versions
421
of the AReA showed partial scalar invariance.
422
Figure 1 shows structure and coefficients of the final partial scalar measurement
423
invariance model. The good fit of the second-order CFA model supports the assumption of a
424
single higher order factor explaining the covariance between the first-order factors. We
425
therefore suggest that scoring of the AReA should, in addition to computation of scores for
426
the three factors, also include computation of a total score reflecting individual aesthetic
427
responsiveness.
428
Fitting the CFA model shown in Figure 1 to data from another German validation
429
sample of 207 participants resulted in a good model fit (χ² = 110.1; df = 73; p = .003; RMSEA
430
= 0.050, 90% CI: 0.029, 0.068; CFI = 0.958; TLI = 0.948; SRMR = 0.052). Factor loadings
431
and latent factor correlations (not shown here) were similar to the results for random sample 2
432
shown in Figure 1. These results further support the factorial validity of the AReA German
433
language version.
434
Scale scores
435
Table 2 shows average scale mean scores for the US and the German total samples.
436
Although some of the scale score distribution tests indicated slight deviations from normality,
437
the absolute skewness and kurtosis parameters as well as inspection of histograms showed
438
that these deviations were minor. As factor scores from the partial scalar measurement
439
invariance model can be used for unbiased comparison of individual trait standings between
440
language versions, we computed correlations between factor scores and scale mean scores.
441
These correlations were very high (Table 2), supporting the utility of scoring the AReA using
442
sum or mean scale scores.
443
Reliability
AESTHETIC RESPONSIVENESS
444
19
Composite reliability coefficients were all in a satisfactory range of w > .70 for both
445
language versions (cf. Table 2). Coefficients were slightly higher in the US sample, with the
446
exception of the subscale CB. Notably, CB yielded acceptable reliability estimations despite
447
comprising only three items.
448
Results of reliability analysis in the additional German validation sample of 207
449
participants suggested good reliabilities for the AReA total scale (ω = .82) and the subscales
450
AA (ω = .84) and IAE (ω = .80). In contrast, the reliability estimate for the subscale CB was
451
somewhat lower (ω = .63), both in comparison with the other AReA subscales in this sample,
452
and in comparison to other samples (cf. Table 2).
453
Validation study 1: Trait pleasure and responses to visual artworks and music
454
The US validation sample consisted of an independent sample of n = 50 participants
455
(mean age = 27.3 yrs., SD = 6.5; 19 males, 31 females) who participated in either a study with
456
visual artworks (Belfi et al., 2019) or with musical excerpts. In addition to the AReA, all
457
participants completed the Temporal Experience of Pleasure Scale (TEPS; Gard et al., 2006).
458
The TEPS consists of two sub-scales: TEPS-A, which measures anticipatory pleasure (related
459
to reward-sensitivity and imagery), and TEPS-C, which measures consummatory pleasure
460
(related to openness to diverse experiences and appreciation of positive stimuli). Moreover,
461
aesthetic judgement ratings were available for visual artworks (n = 21) and musical excerpts
462
(n = 26).2
463
For the TEPS, we expected both scales to show a positive relationship to the AReA
464
sub-scales AA and IAE. Specifically, the TEPS-C scale should bear a positive relationship
465
with the AReA sub-scales, because openness to experience is conceptually closely linked with
466
aesthetic responsiveness. The results shown in Table 3 largely match these expectations,
2
Note that these two subsamples do not add-up to n = 50, because data of three participants had to be
discarded due to problems with performance and recording of the aesthetic judgements.
AESTHETIC RESPONSIVENESS
20
467
although the TEPS Anticipatory Pleasure scale was only very weakly related to IAE and the
468
AReA total score.
469
For the visual study, a squeeze ball was used to record continuous momentary
470
aesthetic pleasantness of visual artworks presented for either 1 second, 5 seconds, or 15
471
seconds. Artworks consisted of 30 paintings at each duration (90 total), selected to represent a
472
variety of styles, content and periods (15th century to present day, Western and Eastern,
473
representational and abstract). Observers were instructed to squeeze the ball at a level
474
corresponding to their felt pleasure both during the painting presentation and for a "post-
475
stimulus" period after the painting disappeared. In addition, participants provided a
476
retrospective overall rating of how aesthetically appealing each trial was using a trackball in
477
the other hand.
478
For the magnitude of the momentary online and retrospective ratings of visual
479
aesthetic stimuli we expected positive correlations with the AReA sub-scales, again
480
particularly AA and IAE. In this context, associations with online-ratings (i.e., the average
481
and maximum ratings via the squeeze ball during the exposure to the stimuli) should prove
482
more reliable compared to associations with retrospective ratings, as they better reflect the
483
momentary experience, whereas retrospective measures are potentially biased. In addition, the
484
maximum rating might show stronger relations to the AReA sub-scales, because they provide
485
an index of the maximum reactivity of a participant. As we expected that exposure to an
486
artwork for the duration of merely one second is substantially too short to provoke a reliable
487
aesthetic response, we compared associations of AReA subscales with ratings during 1-
488
second exposure separately from ratings during 5- and 15-second exposure.
489
For the sample of participants that received visual stimuli, Table 4 provides
490
correlations between the average and maximum online-ratings, and the retrospective ratings
491
for 1 second duration exposure and 5 and 15 second duration exposure with the AReA sub-
AESTHETIC RESPONSIVENESS
21
492
scales. As can be seen, AReA values were not predictive of aesthetic judgments in the 1-
493
second exposure conditions, but correlated with aesthetic judgments in the longer conditions.
494
However, this was only the case for momentary online ratings, but not for retrospective
495
ratings. Moreover, there was a tendency for stronger relations to the maximum online ratings
496
compared to the average online ratings.
497
For the auditory study, participants listened to 60 s excerpts of music and made
498
continuous ratings of liking on a 0 (Low) to 1 (High) visual slider scale using a trackball.
499
Following each clip, observers gave an overall rating of how aesthetically appealing the clip
500
was. Clips consisted of 16 classical pieces and 16 electronic pieces, blocked by genre in
501
groups of 8 clips. Within these genres, pieces were selected to be stylistically consistent in
502
order to prevent participants from responding purely on the basis of genre. Classical pieces
503
were of 19th century small ensemble music from the Romantic era, which contains a wider
504
range of dynamic and emotional intensity than other periods. Electronic music consisted of
505
dance music with a distinctive beat structure (60-150 bpm), selected to have some degree of
506
change or transition during the clip; songs with a single repetitive motif were avoided.
507
For the sample of participants that received music stimuli, Table 5 provides
508
correlations between the average and maximum online-ratings, and the retrospective ratings
509
for classical or electronic music with the AReA sub-scales. As can be seen, AReA scores
510
were substantially correlated with rating of classical music, even though these correlations
511
were not statistically significant due to the small sample.
512
Validation study 2: Responses to poems
513
The second German validation sample consisted of a sub-set of n = 40 participants of
514
the German subsample 2, where the effects of rhetorical language features on the subjective
515
aesthetic experience of the reader was investigated (Menninghaus, Wagner, Wassiliwizky, et
516
al., 2017). Participants read 10 poems in their original version and 10 poems in a de-
AESTHETIC RESPONSIVENESS
22
517
rhetorized version. Additionally, all participants filled in the AReA and provided ratings of
518
different versions of poems on a 7-point scale for beauty, movingness, melodiousness, joy,
519
and sadness. Previous research on poem and proverb reading has shown that manipulations of
520
rhyme and meter lead to changes in the processing and aesthetic evaluation of language
521
(Menninghaus, Bohrn, et al., 2015; Menninghaus & Wallot, 2020; Wallot & Menninghaus,
522
2018).
523
Because AReA is an instrument designed to assess a person’s responsiveness to
524
aesthetic stimuli, we hypothesized that participants scoring high on the AReA would provide
525
higher ratings on subjective emotional and aesthetic experience for the original poems
526
compared to participants that scored low on AReA. Additionally, we hypothesized that
527
participants scoring high on the AReA would show a greater difference between original
528
poems and their de-rhetorized versions (i.e., without rhyme and meter), indicating greater
529
sensitivity to the absence vs. presence of those poetic language features. The subscales
530
Aesthetic Appreciation and Intense Aesthetic Experience were expected to show stronger
531
associations in contrast to Creative Behavior.
532
Table 7 shows the correlations between the three AReA subscales and the AReA total
533
score with ratings of joy, sadness, beauty, movingness and melodiousness. The average
534
ratings correlated consistently positively with the Intense Aesthetic Experience subscale, and
535
less so with the Creative Behavior subscale. However, in contrast to our hypothesis, only
536
values for beauty ratings correlated positively with the Aesthetic Appreciation subscale. For
537
the difference scores, we found significant positive correlations on three out of the five ratings
538
for the Intense Aesthetic Experience subscale, but none for the other two subscales. While
539
these results support the validity of the AReA, it seems that responses to poetry are more
540
strongly affected by a disposition to intense aesthetic experiences as assessed by the IAE
541
subscale of the AReA.
AESTHETIC RESPONSIVENESS
542
23
Validation study 3: Behavioral activation, music reward, and responses to music
543
The first German validation sample consisted of the whole sample of n = 167
544
participants of the German subsample 1, drawn from a study on evaluating listeners’
545
responses to music in order to identify individuals who show low levels of hedonic pleasure
546
during music listening. In addition to the AReA, participants filled in the German version of
547
the BIS/BAS (Carver & White, 1994; Strobel et al., 2001), and a German ad-hoc translation
548
of the Barcelona Music Reward Questionnaire (BMRQ; Mas-Herrero et al., 2013), and were
549
asked to rate how often they experience chills during music listening in general (possible
550
answers: 1 = “never”, 2 = “rarely”, 3 = “sometimes”, 4 = “often”). In addition, participants
551
were asked to listen to a piece of music that had been selected for reliably eliciting chills
552
across a majority of listeners. Afterwards, participants were asked to rate whether they
553
experienced chills while listening to the given piece of music (possible answers: 1 = “no”, 2 =
554
“yes”, or 3 = “don’t know”). For the latter variable, we removed “don’t know” answers before
555
analysis.
556
The BIS/BAS consists of the following sub-scales: The BIS total score (sensitive to
557
signals of punishment, non-reward and novelty), the BAS total score (sensitive to signals of
558
reward, non-punishment and escape from punishment), as well as three BAS-subscales: BAS-
559
Drive (pursuit of desired goals), BAS-Fun-Seeking (desire for new rewards and willingness to
560
approach), and BAS-Reward (positive responses to occurrence or anticipation of reward).
561
Because AReA was designed to assess a person’s sensitivity to aesthetic stimuli primarily
562
relating to a (positive) emotional response, we hypothesized the following: In relation to the
563
AReA subscales, there should be no particular relation to the BIS total score, as AReA items
564
are not related to negative experiences or their avoidance. In contrast, we expected positive
565
associations with the BAS total score, and particularly with the BAS-Reward subscale, as
566
aesthetic experiences are rewarding. As the BIS/BAS captures strong emotional responses, we
AESTHETIC RESPONSIVENESS
24
567
expected strong positive associations with the AReA subscale Intense Aesthetic Experience,
568
but to a lesser degree to Aesthetic Appreciation.
569
The BMRQ consists of five subscales: BMRQ-Musical-Seeking (e.g. looking out for
570
new music, informing oneself, spending money), BMRQ-Emotional-Evocation (e.g chills,
571
tears, becoming emotional), BMRQ-Mood-Regulation (e.g. keeps me company, helps me
572
relax), BMRQ-Sensory-Motor (e.g. need to dance, tap, sing, hum), BMRQ-Social-Reward
573
(e.g. like to play with others, feeling of connection). In relation to the AReA subscales, we
574
expected positive associations with the BMRQ-Emotion-Evocation subscale, which should
575
tap into the same construct as the AReA Aesthetic Appreciation and Intense Aesthetic
576
Experience subscales. Furthermore, the subscale BMRQ-Sensory-Motor seems to be
577
unrelated to the AReA subscales, because it neither captures any form of evaluation of
578
emotional involvement, nor a productive component in the sense of the Creative Behavior
579
subscale. Associations between the other three subscales of the BMRQ and AReA were
580
difficult to predict, because even though they do emphasize emotional components of music
581
perception, they additionally capture consequences of functions of listening to music that are
582
not specifically addressed in the AReA. Finally, the two chill variables were expected to be
583
positively associated with the AReA subscales Aesthetic Appreciation and particularly
584
Intense Aesthetic Experience, because chills are a bodily response indicative of high
585
physiological arousal (Wassiliwizky et al., 2017) triggered by stimuli with high information
586
content (Omigie et al., 2019)
587
Table 6 shows the correlations between the three AReA scale scores and the subscales
588
of the BIS/BAS, the BMRQ, and ratings of occurrence of chills (trait and state). The
589
hypothesized relations are generally borne out: Specifically, the AReA subscales did not
590
correlate with the BIS total score of the BIS/BAS and the Sensory-Motor score of the BMRQ.
591
Furthermore, the Creative Behavior subscale of the AReA showed the smallest correlations
AESTHETIC RESPONSIVENESS
25
592
with all other measures that were expected to be more strongly associated with the receptive
593
subscales of the AReA. Particularly, the hypothesized positive correlations between the
594
AReA subscales Aesthetic Appreciation and Intense Aesthetic Experience with the BAS
595
Reward subscale, BMRQ Emotional Evocation subscale, and trait and state measures of chills
596
were observed.
597
Validation study 4: Big Five, open-mindedness and its facets
598
In another German validation sample, an online survey presented the final 14-item
599
AReA version as well as a German translation of the BFI-2 (Danner et al., 2019; Soto & John,
600
2017) and was completed by 207 participants (3 participants were excluded due to extremely
601
long response times). We computed Pearson’s correlation coefficients between AReA scale
602
scores and the BFI-2 domain scales as well as the three facet scales constituting Open-
603
Mindedness, i.e. Intellectual Curiosity, Aesthetic Sensitivity, and Creative Imagination. The
604
pattern of correlations will provide additional information on the convergent and discriminant
605
validity of the AReA scales. We expected large correlations between AReA scales and the
606
Open-Mindedness scale, but much smaller correlations with the other domain scales, i.e.
607
Extraversion, Agreeableness, Conscientiousness, and Negative Emotionality. With regard to
608
the facet scales of Open-Mindedness, large correlations with AReA scales were expected for
609
the facet Aesthetic Sensitivity, whereas correlations with the other facet scales were expected
610
to be much smaller. Finally, the correlation between the AReA subscale Creative Behavior
611
and the facet scale Creative Imagination was expected to be higher than with the other facet
612
scales, as an individual disposition to high levels of creative imagination is expected to
613
facilitate creative behavior as assessed by the AReA subscale.
614
Table 8 shows correlations between AReA and BFI-2 scales. As expected, correlations
615
of the AReA with Open-Mindedness were large and highly significant, whereas those with
616
Agreeableness, Conscientiousness, and Negative Emotionality were small and mostly not
AESTHETIC RESPONSIVENESS
26
617
significantly different from zero. Extraversion showed significant positive correlations with
618
the AReA scales, due to a considerable portion of shared variance between Extraverion and
619
Open-Mindedness (r = .36). However, these correlations were significantly smaller than the
620
correlations between AReA scales and Open-Mindedness (difference tests for correlation
621
coefficients: all ps ≤ .001, see supplemental Table S3 for details). Regarding the facets of
622
Open-Mindedness, the AReA subscales correlated significantly higher with the facet
623
Aesthetic Sensitivity than with the other facets (ps < .05, see supplemental Table S3 for
624
details), with the exception of the AReA subscale Creative Behavior. In line with our
625
expectations, CB showed significantly higher correlations with Creative Imagination than
626
with the other facets (all ps ≤ .020, see supplemental Table S3 for details).
627
In summary, results of validation study 4 support factorial, convergent, and
628
discriminant validity of the AReA total and subscale scores in its German version, and
629
therefore further strengthen the evidence for construct validity of the AReA.
630
631
Discussion
We present the Aesthetic Responsiveness Assessment (AReA) which can be used to
632
assess aesthetic responsiveness. The scale is based on an original pool of questionnaire items
633
that was compiled with the goal of identifying potential study participants that are particularly
634
responsive to aesthetic stimuli. The final version comprises three sub-scales: Aesthetic
635
Appreciation (AA), Intense Aesthetic Experience (IAE), and Creative Behavior (CB) of
636
respondents.
637
A main goal of the scale was to allow experimenters to distinguish those individuals
638
who regularly respond to artworks in an intense way from those who rarely experience more
639
than a commonplace appreciation of aesthetic objects in everyday life. In supporting the
640
notion that such a distinction is an important one to make, our scale complements previous
641
scales, such as the EBS (Diessner et al., 2008), which focused on other distinctions (e.g.
AESTHETIC RESPONSIVENESS
642
643
27
between reponses to nature, art and moral beauty).
Indeed, the dissociation of the two reception-oriented sub-scales AA and IAE fits with
644
previous behavioral findings on the special capacity of engagement with art to result in
645
intense aesthetic experiences such as being moved (Menninghaus, Wagner, et al., 2015). This
646
dissociation is in line with neurophysiological findings showing that prefrontal and default
647
mode network brain regions are selectively engaged by strongly moving aesthetic experiences
648
with visual artwork (Belfi et al., 2019; Vessel et al., 2012, 2013). Similarly, it is in line with
649
evidence that experiences of beauty in reponse to music may vary in terms of subjective and
650
physiological arousal (Omigie et al., 2019). The extraction of the CB subscale clearly reflects
651
item content relating to participants’ engagement in the creation of art. We suggest that this
652
makes it highly relevant for occasions when it is important to identify participants that
653
regularly engage in the production of art works. However, in contrast to high reliabilities of
654
the AReA total scale score and scores on AA and IAE, the shortness of the CB scale limits its
655
reliability, which implies a relatively larger measurement error in the assessment of
656
individuals. This should be kept in mind when using the CB scale as a screening tool for
657
selection of individuals.
658
One of the most important findings is the demonstration of measurement invariance
659
for the English and German language versions of AReA. Having established full metric
660
invariance suggests that results of association analyses such as regression using the AReA
661
scales can be meaningfully compared between samples from Germany and the US using the
662
respective language versions. However, one should be cautious when comparing mean levels
663
of responses (i.e. composite scores) across English and German language versions, because
664
full scalar invariance had to be rejected for this instrument. Thus observed differences
665
between the samples cannot be fully attributed to differences in individual latent trait
666
standing. However, partial scalar invariance was found when item intercept equality
AESTHETIC RESPONSIVENESS
28
667
constraints were released for three items from the scales AA and CB. Hence, analyses of
668
composite differences between language versions of the AReA or its subscales AA and CB
669
should use factor scores, i.e. latent mean differences (Steinmetz, 2013), while composite
670
scores can be compared between language versions when analyzing IAE subscale scores only.
671
Using independent samples or sub-samples of participants that took part in different
672
studies on the reception and evaluation of music, visual art, and poetry, we found evidence
673
supporting the validity of scale scores by showing expected correlations with self reported
674
strength of aesthetic responsiveness to visual (validation study 1), musical (validation studies
675
1 and 2) and literary aesthetic stimuli (validation study 3), as well as scales tapping into
676
general (BIS/BAS and TEPS), and more domain-specific hedonic responses (BMRQ).
677
Although due to small sample sizes not all of these correlations were statistically significant,
678
many of them represent rather large effects from a normative perspective (Gignac & Szodorai,
679
2016). These results suggest a broad applicability of AReA as a screening instrument across a
680
variety of domains of art perception.
681
As there is considerable overlap between the construct of aesthetic responsiveness and
682
the personality domain opennenss, relatively high correlations between measures of these
683
constructs should be expected. The pattern of correlations of the AReA with measures of the
684
Big Five personality domains and the facets of Open-Mindedness we found in validation
685
study 4 were in line with these expectations. The large correlations between the Open-
686
Mindedness facet Aesthetic Sensitivity and AReA scales support its convergent validity.
687
However, the size of the correlations clearly suggests that the constructs measured by the
688
AReA are suffiently different to support its utility as an independent measurement instrument.
689
This is further supported by the specific association of CB with Creative Imagination. In
690
contrast, AReA scale scores did not correlate substantially with agreeableness,
691
conscientiousness, and negative emotionality, while the moderate correlations with
AESTHETIC RESPONSIVENESS
29
692
extraversion are likely due to shared variance with opennenss. In total, these results strongly
693
support the construct validity of the AReA in its German language version, and they can be
694
expected to generalize to the English language version, as the measurements are invariant
695
across languages. Nevertheless, future studies should investigate similar correlations using an
696
English speaking sample.
697
We conclude that AReA scores indicate the theoretical construct of aesthetic
698
responsiveness. Our theoretical approach emphasizes the individual subjective experience
699
associated with central processing of aesthetic stimuli. Similar to what has been suggested in
700
the area of stress reactivity (Schlotz, 2013; Schlotz et al., 2011), it implies relatively
701
consistent and coherent responses across time, stimulus domains, and response domains. As
702
this is a rather strong assumption, future studies should systematically assess and compare
703
responses across domains to put these theoretical assumptions to the test. The development of
704
an inventory that systematically assesses responses in different domains would be a valuable
705
contribution.
706
It is not surprising that scores on the AReA subscale Creative Behavior (CB)
707
correlated less often and less strongly with judgments of beauty, pleasantness, or aesthetic
708
appeal in reception-oriented tasks than the other two scales, as creative behavior includes an
709
action-related component beyond simply responding to aesthetic stimuli. It could thus be
710
debated whether CB is part of the construct of aesthetic responsiveness in a strict sense.
711
However, we opted to keep this subscale in the AReA, as it provides useful information at
712
relatively low cost (three items only) on an important aspect of aesthetics; namely a
713
predisposition to engage in art production. Indeed, both, substantial correlations between
714
factors, and good fit of the second-order CFA model provide psychometric evidence that
715
supports keeping CB as a subscale of the AReA.
716
It should be noted that theoretically, aesthetic responsiveness includes both indicators
AESTHETIC RESPONSIVENESS
30
717
of aesthetic appreciation and aesthetic engagement. Both are assumed to be affected by an
718
individual’s trait standing on aesthetic responsiveness. Consequently, the AReA does not
719
separate these constructs systematically (although the subscale Aesthetic Appreciation
720
contains less engagement-relevant items than the other subscales). The relative contribution of
721
aesthetic responsiveness to appreciation and engagement could differ between individuals
722
(individual-specific response patterns), and probably even within individuals across time or
723
stimuli. However, a theoretical conception that separates individual propensities to aesthetic
724
engagement vs. appreciation—as two related but separable facets of aesthetic
725
responsiveness—is not incompatible with our theoretical account of aesthetic responsiveness.
726
Future developments of assessments of aesthetic responsiveness could aim at generating items
727
that more systematically sample specific theoretically defined components of aesthetic
728
responsiveness. One approach could be a systematic separation of aesthetic appreciation and
729
aesthetic engagement. Another one could be a differentiation of response indicators to more
730
specifically reflect emotional, cognitive, behavioral, and physiological domains. Whether
731
such refinements of the operationalization of aesthetic responsiveness have utility and
732
incremental validity compared to the AReA is an empirical question.
733
It is important to note that the construct of aesthetic responsiveness explicitly excludes
734
reference to an external standard and is therefore very different from constructs that refer to
735
quality of judgements of aesthetic stimuli such as aesthetic sensitivity (Child, 1964; Eysenck,
736
1940; Myszkowski & Zenasni, 2016; but see Corradi et al., 2019). This has the great
737
advantage that the AReA can be used in non-experts and experts alike. Our theoretical
738
approach clearly implies that the question of whether these groups differ in their aesthetic
739
responsiveness is not a theoretical but an empirical issue. However, the construct defined here
740
nevertheless refers to responsiveness to aesthetic stimuli, and any measure of the construct
741
has to demonstrate that scores reflect more than just non-specific responsivity. In this sense,
AESTHETIC RESPONSIVENESS
31
742
our finding from validation study 2 that AReA scores correlated more strongly with responses
743
to classical versus electronic music can be seen as a first step towards specificity of
744
responsiveness to aesthetically relevant stimuli.
745
Limitations and outlook
746
There might be certain limitations built into the convenience samples that were used in
747
the current analysis. For example, some studies have found differences in art perception and
748
consumption between experts and laypersons (Elvers et al., 2015; Leder et al., 2014). As our
749
samples comprised laypersons, its properties in a sample of experts might be different. To
750
clarify this point, a future study could investigate measurement invariance of the AReA
751
between laypersons and experts.
752
Also, there is a certain built-in limitation of the scale with regard to the original item
753
pool of the screening instrument: Currently, the items of the scale focus disproportionally on
754
wordings that are suggestive of visual perception of art, especially compared to other domains
755
such as music and literature (or nature). Even though the results of our validation studies
756
suggest that the scale can successfully be applied to those domains, it does not provide a fine-
757
grained distinction between domains. Moreover, the current item pool does not systematically
758
cover response domains. For example, IAE captures emotional and physiological responses,
759
but it does not distinguish between them, and does not comprise items indicating other
760
response domains. Hence, future developments should include a more systematic selection of
761
additional items from different aesthetic and response domains to provide a more fine-grained
762
instrument, potentially also covering negative emotional responses to art (Menninghaus,
763
Wagner, Hanich, et al., 2017). Finally, it might be of interest to explore what background
764
experiences lead to high scores on the AReA. More specifically, it would be interesting to
765
investigate the relative contribution of frequency and intensity of individual aesthetic
766
experiences to scores on the AReA.
AESTHETIC RESPONSIVENESS
767
32
The mixture of exploratory and confirmatory strategies in the construction of the
768
AReA resulted in a stable and meaningful scale structure. However, alternative structures are
769
conceivable that emphasize other aspects of aesthetic responsiveness theory. Such alternative
770
operationalizations could be based on refined theoretical accounts and would provide
771
potentially useful progress in the assessment of aesthetic responsiveness. In addition,
772
multimodal assessments of responses could provide insight into aesthetic responsiveness
773
beyond self-reports.
774
Conclusion
775
Although built on an exploratory scale construction strategy, the AReA is a promising,
776
psychometrically evaluated tool for the assessment of individual differences in aesthetic
777
responsiveness that is particularly suitable for selecting participants for empirical aesthetics
778
studies. It can also be used to study (a) associations of aesthetic responsiveness with other
779
constructs from the area of aesthetic research such as aesthetic sensitivity, (b) associations
780
with constructs from the broader area of personality, such as personality dimensions or
781
ability, and (c) developmental trajectories and factors underlying individual aesthetic
782
responsiveness. As we demonstrated measurement invariance for the AReA, its English and
783
German language versions can be used in parallel to compare samples between these
784
languages.
AESTHETIC RESPONSIVENESS
785
786
33
References
Ashton, M. C., & Lee, K. (2007). Empirical, theoretical, and practical advantages of the
787
HEXACO model of personality structure. Personality and Social Psychology Review,
788
11(2), 150-166. https://doi.org/10.1177/1088868306294907
789
Belfi, A. M., Vessel, E. A., Brielmann, A., Isik, A. I., Chatterjee, A., Leder, H., Pelli, D. G.,
790
& Starr, G. G. (2019). Dynamics of aesthetic experience are reflected in the default-
791
mode network. Neuroimage, 188, 584-597.
792
https://doi.org/10.1016/j.neuroimage.2018.12.017
793
Berlyne, D. E. (1971). Aesthetics and psychobiology. Appleton-Century-Crofts.
794
Brielmann, A. A., & Pelli, D. G. (2017). Beauty requires thought. Current Biology, 27(10),
795
796
797
798
1506-1513.e3. https://doi.org/10.1016/j.cub.2017.04.018
Burke, E. (2015). A philosophical enquiry into the sublime and beautiful. Oxford University
Press. (Original work published 1757)
Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and
799
affective responses to impending reward and punishment: The BIS/BAS scales.
800
Journal of Personality and Social Psychology, 67, 319-333.
801
https://doi.org/10.1037/0022-3514.67.2.319
802
Chatterjee, A., & Vartanian, O. (2016). Neuroscience of aesthetics. Annals of the New York
803
Academy of Sciences, 1369(1), 172-194. https://doi.org/10.1111/nyas.13035
804
Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance.
805
Structural Equation Modeling: A Multidisciplinary Journal, 14(3), 464-504.
806
https://doi.org/10.1080/10705510701301834
807
Chen, F. F., Sousa, K. H., & West, S. G. (2005). Testing measurement invariance of second-
808
order factor models. Structural Equation Modeling: A Multidisciplinary Journal,
809
12(3), 471-492. https://doi.org/10.1207/s15328007sem1203_7
AESTHETIC RESPONSIVENESS
810
34
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing
811
measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal,
812
9(2), 233-255. https://doi.org/10.1207/s15328007sem0902_5
813
814
815
Child, I. L. (1964). Observations on the meaning of some measures of esthetic sensitivity.
Journal of Psychology, 57(1), 49-64. https://doi.org/10.1080/00223980.1964.9916671
Corradi, G., Chuquichambi, E. G., Barrada, J. R., Clemente, A., & Nadal, M. (2019). A new
816
conception of visual aesthetic sensitivity. Advance online publication. British Journal
817
of Psychology. https://doi.org/10.1111/bjop.12427
818
Costa, P. T., & McCrae, R. R. (1995). Domains and facets: Hierarchical personality
819
assessment using the Revised NEO Personality Inventory. Journal of Personality
820
Assessment, 64(1), 21-50. https://doi.org/10.1207/s15327752jpa6401_2
821
Danner, D., Rammstedt, B., Bluemke, M., Lechner, C., Berres, S., Knopf, T., Soto, C. J., &
822
John, O. P. (2019). Das Big Five Inventar 2 [The Big Five Inventory 2]. Diagnostica,
823
65(3), 121-132. https://doi.org/10.1026/0012-1924/a000218
824
Diessner, R., Pohling, R., Stacy, S., & Güsewell, A. (2018). Trait appreciation of beauty: A
825
story of love, transcendence, and inquiry. Review of General Psychology, 22(4), 377-
826
397. https://doi.org/10.1037/gpr0000166
827
Diessner, R., Solom, R. C., Frost, N. K., Parsons, L., & Davidson, J. (2008). Engagement with
828
beauty: appreciating natural, artistic, and moral beauty. Journal of Psychology, 142(3),
829
303-329. https://doi.org/10.3200/jrlp.142.3.303-332
830
Elvers, P., Omigie, D., Fuhrmann, W., & Fischinger, T. (2015). Exploring the musical taste of
831
expert listeners: musicology students reveal tendency toward omnivorous taste.
832
Frontiers in Psychology, 6, Article 1252. https://doi.org/10.3389/fpsyg.2015.01252
833
834
Eysenck, H. J. (1940). The general factor in aesthetic judgement. British Journal of
Psychology, 31(1), 94-102. https://doi.org/10.1111/j.2044-8295.1940.tb00977.x
AESTHETIC RESPONSIVENESS
835
35
Fayn, K., MacCann, C., Tiliopoulos, N., & Silvia, P. J. (2015). Aesthetic emotions and
836
aesthetic people: Openness predicts sensitivity to novelty in the experiences of interest
837
and pleasure. Frontiers in Psychology, 6, Article 1877.
838
https://doi.org/10.3389/fpsyg.2015.01877
839
Gard, D. E., Gard, M. G., Kring, A. M., & John, O. P. (2006). Anticipatory and
840
consummatory components of the experience of pleasure: A scale development study.
841
Journal of Research in Personality, 40(6), 1086-1102.
842
https://doi.org/10.1016/j.jrp.2005.11.001
843
Gignac, G. E., & Szodorai, E. T. (2016). Effect size guidelines for individual differences
844
researchers. Personality and Individual Differences, 102, 74-78.
845
https://doi.org/10.1016/j.paid.2016.06.069
846
Götz, K. O., Borisy, A. R., Lynn, R., & Eysenck, H. J. (1979). A new visual aesthetic
847
sensitivity test I: Construction and psychometric properties. Perceptual and Motor
848
Skills, 49(3), 795-802. https://doi.org/10.2466/pms.1979.49.3.795
849
Güsewell, A., & Ruch, W. (2012). Are there multiple channels through which we connect
850
with beauty and excellence? Journal of Positive Psychology, 7(6), 516-529.
851
https://doi.org/10.1080/17439760.2012.726636
852
Hager, M., Hagemann, D., Danner, D., & Schankin, A. (2012). Assessing aesthetic
853
appreciation of visual artworks—The construction of the Art Reception Survey
854
(ARS). Psychology of Aesthetics, Creativity, and the Arts, 6(4), 320-333.
855
https://doi.org/10.1037/a0028776
856
Haidt, J., & Keltner, D. (2004). Appreciation of beauty and excellence [awe, wonder,
857
elevation]. In C. K. Peterson & M. E. P. Seligman (Eds.), Character strengths and
858
virtues (pp. 537-551). Oxford Universty Press.
859
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
AESTHETIC RESPONSIVENESS
36
860
analysis: Conventional criteria versus new alternatives. Structural Equation Modeling:
861
A Multidisciplinary Journal, 6, 1-55.
862
Istok, E., Brattico, E., Jacobsen, T., Krohn, K., Muller, M., & Tervaniemi, M. (2009).
863
Aesthetic responses to music: A questionnaire study. Musicae Scientiae, 13(2), 183-
864
206. https://doi.org/10.1177/102986490901300201
865
Jacobsen, T., & Beudt, S. (2017). Domain generality and domain specificity in aesthetic
866
appreciation. New Ideas in Psychology, 47, 97-102.
867
https://doi.org/10.1016/j.newideapsych.2017.03.008
868
Jacobsen, T., Buchta, K., Köhler, M., & Schröger, E. (2004). The primacy of beauty in
869
judging the aesthetics of objects. Psychological Reports, 94(3_suppl), 1253-1260.
870
https://doi.org/10.2466/pr0.94.3c.1253-1260
871
Karwowski, M., & Beghetto, R. A. (2019). Creative behavior as agentic action. Psychology of
872
Aesthetics, Creativity, and the Arts, 13(4), 402-415.
873
https://doi.org/10.1037/aca0000190
874
Lane, R. D., Waldstein, S. R., Chesney, M. A., Jennings, J. R., Lovallo, W. R., Kozel, P. J.,
875
Rose, R. M., Drossman, D. A., Schneiderman, N., Thayer, J. F., & Cameron, O. G.
876
(2009). The rebirth of neuroscience in psychosomatic medicine, part I: Historical
877
context, methods, and relevant basic science. Psychosomatic Medicine, 71(2), 117-
878
134. https://doi.org/10.1097/PSY.0b013e31819783be
879
Leder, H., Gerger, G., Brieber, D., & Schwarz, N. (2014). What makes an art expert? Emotion
880
and evaluation in art appreciation. Cognition and Emotion, 28(6), 1137-1147.
881
https://doi.org/10.1080/02699931.2013.870132
882
Levinson, J. (2012). Musical beauty. Teorema, 31(3), 127-135.
883
Lim, S., & Jahng, S. (2019). Determining the number of factors using parallel analysis and its
884
recent variants. Psychological Methods, 24(4), 452-467.
AESTHETIC RESPONSIVENESS
885
886
https://doi.org/10.1037/met0000230
Mas-Herrero, E., Marco-Pallares, J., Lorenzo-Seva, U., Zatorre, R. J., & Rodriguez-Fornells,
887
A. (2013). Individual differences in music reward experiences. Music Perception,
888
31(2), 118-138. https://doi.org/10.1525/mp.2013.31.2.118
889
890
37
McCrae, R. R. (2007). Aesthetic chills as a universal marker of openness to experience.
Motivation and Emotion, 31(1), 5-11. https://doi.org/10.1007/s11031-007-9053-1
891
McDonald, R. P. (1999). Test theory: A unified treatment. Erlbaum.
892
McManus, I. C., & Furnham, A. (2006). Aesthetic activities and aesthetic attitudes: Influences
893
of education, background and personality on interest and involvement in the arts.
894
British Journal of Psychology, 97, 555-587.
895
https://doi.org/10.1348/000712606x101088
896
897
898
McNeish, D. (2018). Thanks coefficient alpha, we'll take it from here. Psychological
Methods, 23(3), 412-433. https://doi.org/10.1037/met0000144
Menninghaus, W., Bohrn, I. C., Knoop, C. A., Kotz, S. A., Schlotz, W., & Jacobs, A. M.
899
(2015). Rhetorical features facilitate prosodic processing while handicapping ease of
900
semantic comprehension. Cognition, 143, 48-60.
901
https://doi.org/10.1016/j.cognition.2015.05.026
902
Menninghaus, W., Wagner, V., Hanich, J., Wassiliwizky, E., Jacobsen, T., & Koelsch, S.
903
(2017). The Distancing-Embracing Model of the enjoyment of negative emotions in
904
art reception. Behavioral and Brain Sciences, 40, Article e347.
905
https://doi.org/10.1017/S0140525X17000309
906
Menninghaus, W., Wagner, V., Hanich, J., Wassiliwizky, E., Kuehnast, M., & Jacobsen, T.
907
(2015). Towards a psychological construct of being moved. PLOS ONE, 10(6), Article
908
e0128451. https://doi.org/10.1371/journal.pone.0128451
909
Menninghaus, W., Wagner, V., Wassiliwizky, E., Jacobsen, T., & Knoop, C. A. (2017). The
AESTHETIC RESPONSIVENESS
910
emotional and aesthetic powers of parallelistic diction. Poetics, 63, 47-59.
911
https://doi.org/10.1016/j.poetic.2016.12.001
912
38
Menninghaus, W., Wagner, V., Wassiliwizky, E., Schindler, I., Hanich, J., Jacobsen, T., &
913
Koelsch, S. (2019). What are aesthetic emotions? Psychological Review, 126(2), 171-
914
195. https://doi.org/10.1037/rev0000135
915
Menninghaus, W., & Wallot, S. (2020). What the eyes reveal about (reading) poetry
916
[Manuscript submitted for publication]. Daprtment of Language and Literature, Max
917
Planck Institute for Empirical Aesthetics, Frankfurt am Main, Germany.
918
Millsap, R. E. (2011). Statistical approaches to measurement invariance. Psychology Press.
919
Myszkowski, N., Storme, M., Zenasni, F., & Lubart, T. (2014). Is visual aesthetic sensitivity
920
independent from intelligence, personality and creativity? Personality and Individual
921
Differences, 59(0), 16-20. https://doi.org/10.1016/j.paid.2013.10.021
922
Myszkowski, N., & Zenasni, F. (2016). Individual Differences in Aesthetic Ability: The Case
923
for an Aesthetic Quotient. Frontiers in Psychology, 7, Article 750.
924
https://doi.org/10.3389/fpsyg.2016.00750
925
Omigie, D., Frieler, K., Bär, C., Muralikrishnan, R., Wald-Fuhrmann, M., & Fischinger, T.
926
(2019). Experiencing musical beauty: Emotional subtypes and their physiological and
927
musico-acoustic correlates. Psychology of Aesthetics, Creativity, and the Arts,
928
Advance online publication. https://doi.org/10.1037/aca0000271
929
Pelowski, M., Markey, P. S., Forster, M., Gerger, G., & Leder, H. (2017). Move me, astonish
930
me… delight my eyes and brain: The Vienna Integrated Model of top-down and
931
bottom-up processes in Art Perception (VIMAP) and corresponding affective,
932
evaluative, and neurophysiological correlates. Physics of Life Reviews, 21, 80-125.
933
https://doi.org/10.1016/j.plrev.2017.02.003
934
Peterson, C. K., & Seligman, M. E. P. (2004). Character strengths and virtues: A handbook
AESTHETIC RESPONSIVENESS
935
936
and classification. American Psychological Association.
Puryear, J. S., Kettler, T., & Rinn, A. N. (2017). Relationships of personality to differential
937
conceptions of creativity: A systematic review. Psychology of Aesthetics, Creativity,
938
and the Arts, 11(1), 59-68. https://doi.org/10.1037/aca0000079
939
39
Rawlings, D., Barrantes i Vidal, N., & Furnham, A. (2000). Personality and aesthetic
940
preference in Spain and England: two studies relating sensation seeking and openness
941
to experience to liking for paintings and music. European Journal of Personality,
942
14(6), 553-576. https://doi.org/10.1002/1099-0984(200011/12)14:6<553::AID-
943
PER384>3.0.CO;2-H
944
Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure:
945
is beauty in the perceiver's processing experience? Personality and Socical
946
Psychology Review, 8(4), 364-382. https://doi.org/10.1207/s15327957pspr0804_3
947
Rowold, J. (2008). Instrument development for esthetic perception assessment. Journal of
948
Media Psychology, 20(1), 35-40. https://doi.org/10.1027/1864-1105.20.1.35
949
Schindler, I., Hosoya, G., Menninghaus, W., Beermann, U., Wagner, V., Eid, M., & Scherer,
950
K. R. (2017). Measuring aesthetic emotions: A review of the literature and a new
951
assessment tool. PLOS ONE, 12(6), Article e0178899.
952
https://doi.org/10.1371/journal.pone.0178899
953
954
955
Schlotz, W. (2013). Stress reactivity. In M. D. Gellman & R. J. Turner (Eds.), Encyclopedia
of Behavioral Medicine (pp. 1891-1894). Springer.
Schlotz, W., Yim, I. S., Zoccola, P. M., Jansen, L., & Schulz, P. (2011). The perceived stress
956
reactivity scale: Measurement invariance, stability, and validity in three countries.
957
Psychological Assessment, 23(1), 80-94. https://doi.org/10.1037/a0021148
958
Silvia, P. J., Fayn, K., Nusbaum, E. C., & Beaty, R. E. (2015). Openness to experience and
959
awe in response to nature and music: Personality and profound aesthetic experiences.
AESTHETIC RESPONSIVENESS
960
Psychology of Aesthetics, Creativity, and the Arts, 9(4), 376-384.
961
https://doi.org/10.1037/aca0000028
962
Soto, C. J., & John, O. P. (2017). The next Big Five Inventory (BFI-2): Developing and
963
assessing a hierarchical model with 15 facets to enhance bandwidth, fidelity, and
964
predictive power. Journal of Personality and Social Psychology, 113(1), 117-143.
965
https://doi.org/10.1037/pspp0000096
966
Stamatopoulou, D. (2004). Integrating the philosophy and psychology of aesthetic experience:
967
Development of the aesthetic experience scale. Psychological Reports, 95(2), 673-
968
695. https://doi.org/10.2466/PR0.95.6.673-695
969
40
Steinmetz, H. (2013). Analyzing observed composite differences across groups: Is partial
970
measurement invariance enough? Methodology, 9(1), 1-12.
971
https://doi.org/10.1027/1614-2241/a000049
972
Strobel, A., Beauducel, A., Debener, S., & Brocke, B. (2001). Eine deutschsprachige Version
973
des BIS/BAS-Fragebogens von Carver und White [A German language version of the
974
BIS/BAS questionnaire by Carver and White]. Zeitschrift für Differentielle und
975
Diagnostische Psychologie, 22(3), 216-227. https://doi.org/10.1024//0170-
976
1789.22.3.216
977
978
van de Vijver, F., & Hambleton, R. K. (1996). Translating tests: Some practical guidelines.
European Psychologist, 1(2), 89-99. https://doi.org/10.1027/1016-9040.1.2.89
979
Vessel, E. A., Isik, A. I., Belfi, A. M., Stahl, J. L., & Starr, G. G. (2019). The default-mode
980
network represents aesthetic appeal that generalizes across visual domains.
981
Proceedings of the National Academy of Sciences, 116(38), 19155-19164.
982
https://doi.org/10.1073/pnas.1902650116
983
984
Vessel, E. A., Maurer, N., Denker, A. H., & Starr, G. G. (2018). Stronger shared taste for
natural aesthetic domains than for artifacts of human culture. Cognition, 179, 121-131.
AESTHETIC RESPONSIVENESS
985
986
41
https://doi.org/10.1016/j.cognition.2018.06.009
Vessel, E. A., Starr, G. G., & Rubin, N. (2012). The brain on art: Intense aesthetic experience
987
activates the default mode network. Frontiers in Human Neuroscience, 6.
988
https://doi.org/10.3389/fnhum.2012.00066
989
Vessel, E. A., Starr, G. G., & Rubin, N. (2013). Art reaches within: Aesthetic experience, the
990
self and the default mode network. Frontiers in Human Neuroscience, 7, Article 258.
991
https://doi.org/10.3389/fnins.2013.00258
992
Wallot, S., & Menninghaus, W. (2018). Ambiguity effects of rhyme and meter. Journal of
993
Experimental Psychology: Learning, Memory, and Cognition., 44(12), 1947-1954.
994
https://doi.org/10.1037/xlm0000557
995
Wassiliwizky, E., Koelsch, S., Wagner, V., Jacobsen, T., & Menninghaus, W. (2017). The
996
emotional power of poetry: neural circuitry, psychophysiology and compositional
997
principles. Social, Cognitive, and Affective Neuroscience, 12(8), 1229-1240.
998
https://doi.org/10.1093/scan/nsx069
999
Woo, S. E., Chernyshenko, O. S., Longley, A., Zhang, Z. X., Chiu, C. Y., & Stark, S. E.
1000
(2014). Openness to experience: Its lower level structure, measurement, and cross-
1001
cultural equivalence. Journal of Personality Assessment, 96(1), 29-45.
1002
https://doi.org/10.1080/00223891.2013.806328
1003
AESTHETIC RESPONSIVENESS
42
1004
Table 1
1005
Fit indices and test statistics for configural, metric and scalar invariance of the second-order
1006
factor model of the AReA between the US (n = 140) and German sample (n = 147) of random
1007
sample 2.
Invariance test
Fit index
Configural Metric
(first order
factors)
Metric
(second order
factor)
Scalar
(observed
indicators)
1008
Partial
Partial
scalar a
Scalar b
(observed (first order
indicators) factors)
232.1
250.9
255.8
356.0
297.7
298.0
c²
df
147
159
161
174
171
172
RMSEA
.064 [.048, .063 [.048, .064 [.049,
.085 [.073, .072 [.058, .071 [.058,
.079]
.078]
.078]
.098]
.085]
.085]
ΔRMSEA
-.001
.001
.021
.008
.007
CFI
.957
.954
.952
.908
.936
.937
ΔCFI
-.003
-.002
-.044
-.016
-.015
SRMR
.050
.069
.073
.089
.079
.080
ΔSRMR
.019
.004
.016
.006
.001
Note. The residual variance of one first-order factor in the German sample had a small
1009
negative estimate and was therefore set to zero in all models.
1010
a
1011
factor) invariance model.
1012
b
1013
test against partial scalar (observed indicators) invariance model. This final model is
1014
presented in Figure 1. See supplemental material for item wording.
Intercept equality constraints lifted for items 5, 11, and 12; test against metric (second order
Equality constraints set for all first-order factor means and the second-order factor mean;
AESTHETIC RESPONSIVENESS
43
1015
Table 2
1016
Mean scale scores, correlations with factor scores, and reliability estimates for AReA
1017
subscales and total score for the US (n = 281) and German sample (n = 293)
AA
US sample
IAE
CB
AReA
AA
German sample
IAE
CB
AReA
1018
Scale mean
scores
Mean
3.5
2.4
2.6
2.8
3.7
2.6
2.3
2.8
SD
0.8
0.9
1.0
0.8
0.7
0.8
1.0
0.7
S
-0.2
0.3
0.4
0.3
-0.5
0.4
0.7
0.3
K
2.6
2.5
2.6
2.6
3.3
2.9
2.6
3.2
p (SK)
.12
.007
.005
.057
.003
.060
.001
.12
r (scores)
.98
.98
.99
.90
.98
.98
.95
.97
.91
.89
.72
.89
.86
.80
.73
.84
Reliability (w)
Note. AA = Aesthetic Appreciation; IAE = Intense Aesthetic Experience; CB = Creative
1019
Behavior; AReA = Aesthetic Responsiveness Assessment total score; SD = Standard
1020
deviation; S = Skewness; K = Kurtosis; p (SK) = Joint skewness/kurtosis test for normality; r
1021
(scores) = Pearson correlations of scale mean scores with factor scores. Tests of average
1022
differences in scale mean scores between the US and German samples showed that the US
1023
sample scored significantly lower on the AReA subscales AA, t(572) = -3.4, p = .001, and
1024
IAE, t(572) = -2.5, p = .013, but higher on CB, t(572) = 3.8, p < .001. In contrast, the AReA
1025
total score did not differ significantly between the samples, t(572) = -0.4, p = .69.
1026
AESTHETIC RESPONSIVENESS
44
1027
Table 3
1028
Correlations between AReA subscales and total score and subscales of the TEPS (n = 50)
1029
TEPS
TEPS-A
TEPS-C
** p < .01
1030
Note. AA = Aesthetic Appreciation; IAE = Intense Aesthetic Experience; CB = Creative
1031
Behavior; AReA = Aesthetic Responsiveness Assessment total score; TEPS-A = Temporal
1032
Expectations of Pleasure Scale, Anticipatory Pleasure; TEPS-C = Temporal Expectations of
1033
Pleasure Scale, Consumatory Pleasure.
1034
AA
.38**
.44**
IAE
.15
.37**
CB
.04
.24
AReA
.18
.38**
AESTHETIC RESPONSIVENESS
45
1035
Table 4
1036
Correlations between AReA subscales and total score and aesthetic judgments of visual
1037
paintings (n = 21).
1038
Aesthetic judgments
AA
1 second exposure
Momentary force rating
Mean
.10
Maximum
.17
Retrospective
-.09
5 and 15 second exposure (combined)
Momentary force rating
Mean
.28
Maximum
.28
Retrospective
.06
* p < .05
1039
Note. AA = Aesthetic Appreciation; IAE = Intense Aesthetic Experience; CB = Creative
1040
Behavior; AReA = Aesthetic Responsiveness Assessment total score. Momentary ratings are
1041
the average of the measured force produced during stimulus exposure. Retrospective ratings
1042
were provided on an analogue scale ranging from 0 to 1.
1043
IAE
CB
AReA
.26
.35
.10
.22
.36
.20
.24
.36
.11
.44*
.43*
.28
.35
.44*
.22
.42*
.45*
.23
AESTHETIC RESPONSIVENESS
46
1044
Table 5
1045
Correlations between AReA subscales and total score and aesthetic judgments of auditory
1046
stimuli (n = 26)
1047
Aesthtic judgments
Classical Music
Momentary force rating
Mean
Maximum
Retrospective
Electronic Music
Momentary force rating
Mean
Maximum
Retrospective
* p < .05
1048
Note. AA = Aesthetic Appreciation; IAE = Intense Aesthetic Experience; CB = Creative
1049
Behavior; AReA = Aesthetic Responsiveness Assessment total score. Online ratings are the
1050
average of the measured force produced during stimulus exposure. Retrospective ratings were
1051
provided on an analogue scale ranging from 0 to 1.
1052
AA
IAE
CB
AReA
.24
.44*
.28
.31
.31
.31
.35
.17
.31
.35
.31
.34
-.15
.23
-.22
-.09
.13
-.19
-.14
-.16
-.25
-.14
.03
-.25
AESTHETIC RESPONSIVENESS
47
1053
Table 6
1054
Correlations between AReA subscales and total score and average ratings of original poems,
1055
as well as differences in ratings for original vs. partly de-rhetorized poems (n = 40)
1056
AA
IAE
CB
AReA
Average ratings for original poems
Beauty
.38*
.58***
.21
.47**
Movingness
.14
.36*
.32*
.34*
Melodiousness
.06
.31*
.16
.23
Joy
.10
.41**
-.001
.21
Sadness
.14
.34*
.32*
.33*
Absolute difference scores of original poems v. poem version without rhyme and meter
Beauty
.24
.38*
.03
.26
Movingness
.24
.33*
.08
.26
Melodiousness
-.01
.22
.11
.14
Joy
.14
.40**
-.07
.19
Sadness
.12
.23
.02
.15
* p < .05, ** p < .01, *** p < .001.
1057
Note. Ratings for beauty, movingness, and melodiousness were averaged across 10 poems,
1058
joy and sadness ratings only across the 5 joyful and sad poems from the same set; AA =
1059
Aesthetic Appreciation; IAE = Intense Aesthetic Experience; CB = Creative Behavior; AReA
1060
= Aesthetic Responsiveness Assessment total score.
1061
AESTHETIC RESPONSIVENESS
48
1062
Table 7
1063
Correlations between AReA scale scores and subscales of BIS/BAS, BMRQ and chills (n =
1064
167)
1065
AA
BIS/BAS
BIS total
-.01
BAS total
.16*
BAS-Drive
.20**
BAS-Fun-Seeking
.25**
BAS-Reward
.25***
BMRQ
Music Seeking
.39***
Emotional Evocation
.36***
Mood Regulation
.32***
Sensory-Motor
.14
Social Reward
.39***
Chills
Trait
.16*
State
.24**
* p < .05, ** p < .01, *** p < .001.
1066
Note. AA = Aesthetic Appreciation; BIS/BAS = Behavioral Inhibition/Activation System;
1067
BMRQ = Barcelona Music Reward Questionnaire; IAE = Intense Aesthetic Experience; CB =
1068
Creative Behavior; AReA = Aesthetic Responsiveness Assessment total score.
1069
IAE
CB
AReA
.09
.20**
.21**
.31***
.29***
.03
.19*
.21**
.14
.22**
.03
.21**
.24**
.27***
.30***
.26***
.25**
.14
.10
.23**
.20*
.11
.08
.03
.15
.35***
.30***
.25**
.12
.33***
.25**
.26**
.18*
.09
.24**
.25**
AESTHETIC RESPONSIVENESS
49
1070
Table 8
1071
Correlations between AReA subscales and total score and Big Five Inventory 2 domain scales
1072
and facet scales of the domain Open-Mindedness (n = 207)
Mean (SD)
Correlations with AReA scales
AA
IAE
CB
AReA
.30***
.18*
.12
.03
.61***
.21**
.13
-.02
.12
.45***
.17*
.07
-.01
.06
.48***
.29***
.16*
.06
.07
.63***
.35***
.71***
.36***
.27***
.42***
.35***
.28***
.26***
.44***
.37***
.64***
.45***
1073
BFI-2 domains
Extraversion
40.5 (7.3)
Agreeableness
45.2 (6.0)
Conscientiousness
43.5 (7.2)
Negative Emotionality
32.1 (7.7)
Open-Mindedness
47.1 (7.0)
BFI-2 facets of Open-Mindedness
Intellectual Curiosity
15.9 (2.8)
Aesthetic Sensitivity
16.5 (2.9)
Creative Imagination
14.7 (3.4)
* p < .05, ** p < .01, *** p < .001.
1074
Note. AA = Aesthetic Appreciation; IAE = Intense Aesthetic Experience; CB = Creative
1075
Behavior; AReA = Aesthetic Responsiveness Assessment total score.
1076
1077
AESTHETIC RESPONSIVENESS
50
1078
Figure 1
1079
Final CFA model for the AReA in the English and German language version including
1080
unstandardized coefficients from the partial scalar invariance model. First- and second-order
1081
factor loading parameters are equal for the two version. Residual variances of first-order
1082
factors and the variance of the second-order factor shown are for the English version in the
1083
first line and for the German version in the second line. Item intercepts and error variances
1084
not shown.
1085