Content uploaded by Anne C. Frenzel
Author content
All content in this area was uploaded by Anne C. Frenzel on Oct 15, 2014
Content may be subject to copyright.
UNCORRECTED PROOF
ORIGINAL PAPER
1
2Types of boredom: An experience sampling approach
3Thomas Goetz •Anne C. Frenzel •Nathan C. Hall •
4Ulrike E. Nett •Reinhard Pekrun •Anastasiya A. Lipne-
5vich
6
7Springer Science+Business Media New York 2013
8Abstract The present study investigated different types
9of boredom as proposed in a four-categorical conceptual
10 model by Goetz and Frenzel (Zeitschrift fu
¨r Entwick-
11 lungspsychologie und Pa
¨dagogische Psychologie
12 38(4):149–153, 2006). In this model, four types of bore-
13 dom are differentiated based on degrees of valence and
14 arousal: indifferent, calibrating, searching, and reactant
15 boredom. In two studies (Study 1: university students,
16 N=63, mean age 24.08 years, 66 % female; Study 2: high
17 school students, grade 11, N=80, mean age 17.05 years,
18 58 % female), real-time data were obtained via the expe-
19 rience-sampling method (personal digital assistants, ran-
20 domized signals). Boredom experiences (N=1,103/1,432
21 in Studies 1/2) were analyzed with respect to the dimen-
22 sions of valence and arousal using multilevel latent profile
23 analyses. Supporting the internal validity of the proposed
24
boredom types, our results are in line with the assumed four
25
types of boredom but suggest an additional, fifth type,
26
referred to as ‘‘apathetic boredom.’’ The present findings
27
further support the external validity of the five boredom
28
types in showing differential relations between the bore-
29
dom types and other affective states as well as frequency of
30
situational occurrence (achievement contexts vs. non-
31
achievement contexts). Methodological implications as
32
well as directions for future research are discussed.
33
34
Keywords Boredom Emotions Achievement
35
Experience sampling
36
Introduction
37
…it is probable that the conditions and forms of
38
behavior called ‘boredom’ are psychologically quite
39
heterogeneous (Fenichel 1951, p. 349; see Fenichel
40
1934, for original German quote)
41
Boredom is a frequently experienced emotion
1
(Larson
42
and Richards 1991; Nett et al. 2011) that due to its prev-
43
alence is often described as a plague of modern society
44
(Klapp 1986; Pekrun et al. 2010; Spacks 1995). Despite
45
potential benefits of boredom under specific situational
46
conditions (e.g., in initiating creative processes and greater
47
self-reflection, Seib and Vodanovich 1998; see Vodanovich
48
2003a for an overview), empirical evidence strongly indi-
49
cates that boredom corresponds to a number of detrimental
50
experiences and behaviors. For example, boredom has been
A1 T. Goetz (&)
A2 Department of Empirical Educational Research, University of
A3 Konstanz, Universitaetsstr. 10, 78457 Constance, Germany
A4 e-mail: thomas.goetz@uni-konstanz.de
A5 T. Goetz U. E. Nett
A6 Thurgau University of Teacher Education, Thurgau, Switzerland
A7 A. C. Frenzel R. Pekrun
A8 Department of Psychology, University of Munich, Munich,
A9 Germany
A10 N. C. Hall
A11 Department of Educational and Counselling Psychology, McGill
A12 University, Montreal, Canada
A13 U. E. Nett
A14 Department of Psychology, University of Ulm, Ulm, Germany
A15 A. A. Lipnevich
A16 Division of Education, Queens College, City University of New
A17 York, New York, NY, USA
1FL01
1
As boredom is not a prototypical or basic emotional experience
1FL02(e.g., Ekman 1984; Rosch 1978; Shaver et al. 1987), it has
1FL03alternatively been classified in terms of constructs such as affect or
1FL04mood.
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Motiv Emot
DOI 10.1007/s11031-013-9385-y
Author Proof
UNCORRECTED PROOF
51 found to relate to nicotine and alcohol consumption (Amos
52 et al. 2006), drug use (Anshel 1991), stress and health
53 problems (Thackray 1981), juvenile delinquency (New-
54 berry and Duncan 2001), truancy (Sommer 1985), dropout
55 at school (Bearden et al. 1989), and negative achievement
56 outcomes (Goetz et al. 2006,2007a; Pekrun et al. 2010,
57 2009).
58 Given the high frequency of boredom in various situa-
59 tions encountered in daily life and the variety of detri-
60 mental experiences to which boredom is related, it is rather
61 surprising that to date there has been little research con-
62 ducted on this specific emotion (Pekrun et al. 2002,2010;
63 for reviews see Vodanovich 2003b and Smith 1981). One
64 possible reason for this neglect is that boredom represents
65 an inconspicuous or ‘‘silent’’ emotion as compared to more
66 intensive and consequently more easily observable affec-
67 tive states such as anger or anxiety (for the school context
68 see Goetz et al. 2007b). Further, from the perspective of
69 clinical practice, people’s experiences of boredom are also
70 not typically regarded as relevant to psychopathological
71 diagnoses, in contrast to anxiety or hopelessness/help-
72 lessness (cf., Miller and Seligman 1975; Zeidner 1998,
73 2007). Finally, boredom is not a prototypical emotion
74 (Shaver et al. 1987) and has no prototypical facial
75 expression (Ekman 1984).
76 In addition to research questions that pertain to the
77 prevalence, effects, and detection of boredom experiences,
78 a more fundamental question exists concerning the con-
79 ceptual definition of boredom. From the perspective of
80 component-process definitions of emotions (Kleinginna
81 and Kleinginna 1981; Scherer 2000), boredom consists of
82 affective components (unpleasant, aversive feelings), cog-
83 nitive components (altered perceptions of time), physio-
84 logical components (reduced arousal), expressive
85 components (facial, vocal, and postural expression; for
86 body movements and postures related to boredom see
87 Wallbott 1998), as well as motivational components
88 (motivation to change or leave the situation; see Goetz and
89 Frenzel 2006; Johnstone and Scherer 2000; Pekrun et al.
90 2010). From this perspective, boredom shares certain
91 characteristics with other emotional experiences (e.g.,
92 motivational component—motivation to change the situa-
93 tion is also salient in anxiety) but at the same time it is
94 clearly different from these emotions (e.g., physiological
95 component—reduced arousal is not typical for anxiety).
96 In addition to the need to conceptually define this con-
97 struct, there remains the need to also qualify the potential
98 conceptual dimensions underlying experiences of boredom.
99 As opposed to operational definitions, this approach
100 reflects a more specific way of classifying emotional
101 experiences along multiple dimensions. This dimensional
102 approach is highlighted in the well-known circumplex
103 models of affect (Russell 1980; see also Watson and
104
Tellegen 1985), in which affective states are characterized
105
by two orthogonal dimensions of valence and arousal. In
106
dimensional approaches, boredom has mainly been out-
107
lined as an unpleasant emotional state of relatively low
108
negative valence (slightly unpleasant on average; e.g.,
109
Fisher 1993; Goetz et al. 2007a,b; Perkins and Hill 1985).
110
Whereas the valence assumption is rather consistent in this
111
relatively small research literature, findings concerning the
112
dimension of arousal associated with boredom are mixed.
113
For example, several researchers have classified boredom
114
as a low-arousal emotion (e.g., Hebb 1955; Mikulas and
115
Vodanovich 1993; Titz 2001), whereas others have
116
described it as a high-arousal emotion (e.g., Berlyne 1960;
117
London et al. 1972; Rupp and Vodanovich 1997; Sommers
118
and Vodanovich 2000). Consequently, there exists an
119
ongoing research debate as to whether boredom is best
120
understood as a low- or high-arousal emotional experience
121
(Pekrun et al. 2010; for both low and high arousal, see
122
Harris 2000), and further, an unanswered research question
123
as to why findings concerning the arousal associated with
124
boredom are so varied in nature.
125
The notably heterogeneous theoretical assumptions and
126
empirical results with respect to the arousal dimension of
127
boredom can be traced back to classic statements in psy-
128
choanalytic literature from the 1930s, in which the multi-
129
faceted nature of this emotion was hypothesized (see
130
statement by Austrian psychoanalyst Fenichel 1934,1951
131
above). To our knowledge, this idea did not receive sig-
132
nificant attention in the scientific community until it was
133
echoed over six decades later by Phillips (1993) who
134
suggested that boredom does not appear to represent a
135
single entity, but rather multiple ‘‘boredoms’’ (p. 78; i.e.,
136
‘‘types’’ of boredom).
137
With respect to recent empirical studies concerning
138
potential boredom types, findings from a qualitative study
139
by Goetz and Frenzel (2006) suggest that individuals do
140
indeed report experiencing different types of boredom—
141
not only with respect to arousal but also valence. One
142
methodological shortcoming of the study by Goetz and
143
Frenzel (2006) was that it relied on retrospective self-
144
reports that may have been adversely affected by recall
145
biases or cognitive schemas about emotions rather than
146
actual experiences (Robinson and Barrett 2010; Robinson
147
and Clore 2002; Roseman et al. 1996). In an effort to more
148
comprehensively evaluate the utility of the multiple bore-
149
dom types approach, the present study analyzed empirical
150
data collected in real-life situations. More specifically, our
151
study explored the longstanding assumption that a theo-
152
retical model consisting of multiple boredom types may
153
better reflect individuals’ actual experiences of this emo-
154
tion in everyday life as compared to existing models, from
155
which mixed results are obtained. The first goal of the
156
study was to contribute to the boredom literature by
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
157 addressing the possibility that the boredom arousal con-
158 troversy may be the result of assessing different boredom
159 types associated with differing levels of arousal. The sec-
160 ond goal was to explore potential differences in boredom
161 experiences with respect to varying degrees of valence as
162 suggested by the preliminary findings of Goetz and Frenzel
163 (2006). The latter study explored the extent, to which the
164 slightly negative valence typically associated with bore-
165 dom may simply reflect an average across boredom sub-
166 types, each differing to some extent in how unpleasant they
167 are.
168 A model of boredom experiences
169 Boredom types: Valence and arousal dimensions
170 Goetz and Frenzel (2006) proposed a conceptual model
171 consisting of four different types of boredom that were
172 derived through qualitative interview data concerning the
173 phenomenology of boredom in academic settings. Fifty-
174 ninth-graders (50 % female, M
age
=14.86) were asked to
175 ‘‘Imagine someone who does not know how it feels to be
176 bored. Please try to describe to him or her how it feels.’’
177 After responding to this domain-general question, students
178 were asked to mentally recall a domain-specific boredom
179 experience, namely, a class they perceived as boring in
180 nature. Based on a component model of emotions (Scherer
181 1984), students were then asked the following questions:
182 ‘‘What did you think when you were bored?’’ (cognitive
183 component), ‘‘What would you have liked to have done
184 most when you were bored?’’ (motivational component),
185 and ‘‘How did your body feel when you were bored?’’
186 (physiological component).
187 Consistent with the aforementioned assumptions
188 regarding the existence of types of boredom (e.g., Fenichel
189 1934,1951; Phillips 1993), students’ responses to all
190 interview questions were rather heterogeneous and often
191 contradictory in nature. For example, with respect to the
192 first domain-general question regarding how boredom
193 feels, students’ responses referred to relaxation (36 %) and
194 tiredness/inertness (34 %) as well as need for activity
195 (22 %), aggression (12 %), and unrest (12 %). In an
196 attempt to identify coherent themes across participants’
197 heterogeneous statements, the authors adopted the well-
198 known circumplex model of affect (Russell 1980; see also
199 Watson and Tellegen 1985) in which discrete affective
200 states are characterized by the orthogonal dimensions of
201 valence and arousal. It is important to note that in Goetz
202 and Frenzel (2006), it was the subtypes of boredom that
203 were classified along these two dimensions (within-emo-
204 tion classification) rather than discrete emotional experi-
205 ences (e.g., boredom vs. enjoyment). The authors identified
206
four types of boredom differing in their level of valence as
207
well as arousal (see Fig. 1).
208
The first boredom type was labeled ‘‘indifferent bore-
209
dom’’ and assumed to correspond with low arousal and
210
slightly positive valence. Statements reflecting indifferent
211
boredom included descriptors such as relaxation and
212
cheerful fatigue, and reflected a general indifference to,
213
and withdrawal from, the external world. The second
214
boredom type, referred to as ‘‘calibrating boredom,’’ was
215
associated with higher (but still relatively low) arousal
216
compared to indifferent boredom and slightly negative
217
valence. Statements reflecting calibrating boredom indi-
218
cated wandering thoughts, not knowing what to do, and a
219
general openness to behaviors aimed at changing the sit-
220
uation or cognitions unrelated to the situation. Thus, cal-
221
ibrating boredom represented a slightly unpleasant
222
emotional state associated with receptiveness to boredom-
223
reducing options but not actively searching for alternate
224
behaviors or cognitions.
225
The third boredom type was labeled ‘‘searching bore-
226
dom’’ and was characterized by a more negative valence
227
and higher arousal than calibrating boredom. Students’
228
statements describing searching boredom reflected a sense
229
of restlessness and an active search for alternative actions
230
as evidenced by statements referring to the need for activity
231
and specific thoughts about hobbies, leisure, interests, and
232
school. This type of boredom was experienced as rather
233
unpleasant and was associated with not only being gener-
234
ally open to, but actively seeking out specific ways of
235
minimizing feelings of boredom.
236
The fourth and final boredom type was classified as
237
‘‘ reactant boredom’’ and was characterized by the highest
238
levels of arousal and negative valence. Statements reflect-
239
ing reactant boredom indicated a strong motivation to leave
240
the boredom-inducing situation and avoid those responsible
241
for this situation (e.g., teachers). Students discussed sig-
242
nificant restlessness, aggression, as well as persistent
243
thoughts about specific, more highly valued alternative
244
situations. This final boredom type was thus experienced as
245
very unpleasant in nature and was strongly associated with
246
a need to escape the situation. The four boredom types may
247
develop from one type into another based on situational
248
factors (e.g., searching boredom may develop into reactant
249
boredom in a tedious classroom setting).
250
Boredom types: Relations to other affective states
251
According to the boredom types proposed by Goetz and
252
Frenzel (2006), each boredom type should differentially
253
correspond with other positive and negative affective states
254
in accordance with its degree of valence (from positive to
255
negative; see Fig. 1, x axis). Boredom types that are
256
slightly negative, or even positive in valence, should be
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
257 more strongly associated with positive affective states (e.g.,
258 enjoyment) and less strongly associated with negative
259 affective states (e.g., anger). Conversely, boredom types
260 characterized by high negative valence should correlate
261 more strongly with negative affective states and less
262 strongly with positive affective states. Empirical findings
263 demonstrating the assumed differential relations between
264 boredom types and other affective measures, in a manner
265 consistent with the increasing levels of negative valence
266 from the first to the fourth boredom type, would serve to
267 support the external validity of the four boredom types.
268 Boredom types: Situational prevalence
269 Preliminary findings by Goetz and Frenzel (2006) suggest
270 that the prevalence of each boredom type may differ as a
271 function of contrasting situational characteristics, for
272 example, non-achievement situations as compared to
273 classroom settings. As non-achievement situations typi-
274 cally afford greater degrees of freedom with respect to the
275 choice or termination of activities (e.g., leisure time), types
276 of boredom that are lower in negative valence (e.g.,
277 indifferent boredom) may be more prevalent in non-
278 achievement settings than in situations experienced at
279 school or at work where activity selection and withdrawal
280 opportunities are more limited. Conversely, boredom types
281 that are higher in negative valence (e.g., reactant boredom)
282 are assumed to be more strongly associated with achieve-
283 ment situations (e.g., classroom activities) as compared to
284 non-achievement situations (e.g., shopping).
285 Goetz and Frenzel (2006) further stated the boredom types
286 may change over time. However, due to the availability of
287
withdrawal opportunities in non-achievement situations,
288
boredom types of relatively low negative valence (e.g.,
289
searching boredom) may not often develop into boredom
290
types of high negative valence (e.g., reactant boredom).
291
That is, in non-achievement situation students are likely to
292
change the activity or the situation before boredom types of
293
extreme negative valence can develop. Hence, quantitative
294
findings indicating the anticipated differences in the situ-
295
ational prevalence of the boredom types would provide
296
empirical support for the external validity of the proposed
297
boredom typology.
298
Research hypotheses
299
Based on an in-depth qualitative assessment, Goetz and
300
Frenzel (2006) proposed a conceptual model distinguishing
301
between four types of boredom as a function of differential
302
degrees of valence and arousal. Using real-time data
303
obtained via the experience sampling method, the current
304
quantitative study evaluated the afore-mentioned assump-
305
tions that follow from these preliminary findings. More
306
specifically, we anticipated that the proposed four-part
307
classification of boredom types would be empirically
308
confirmed (Hypothesis 1), that the proposed boredom types
309
would correspond with other affective states in a manner
310
consistent with increasing levels of negative valence
311
(Hypothesis 2), and finally, that these boredom types would
312
differ as anticipated with respect to situational character-
313
istics (Hypothesis 3). Thus, whereas Hypothesis 1 focuses
314
on the internal validity of the proposed four boredom types,
315
Hypotheses 2 and 3 evaluate the external validity of the
Valence
Arousal
Indifferent
boredom
Searching
boredom
Reactant
boredom
Calibrating
boredom
Phenomenologically similar to
An
g
erWell-bein
g
negativepositive
Fig. 1 Localization of boredom
types relative to negative
valence and arousal (Adapted
from Goetz and Frenzel 2006)
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
316 four-part boredom typology with respect to other affective
317 states and across different situations.
318 Hypothesis 1: Boredom types
319 We anticipated that four types of boredom classified based
320 on the dimensions of valence and arousal would be observed
321 in real-life settings. More specifically, we expected to find
322 four boredom types differentiated in a two-dimensional
323 space with the first dimension representing valence (from
324 positive to negative valence) and the second dimension
325 reflecting arousal (from low to high arousal; see Fig. 1).
326 Hypothesis 2: Relations of boredom types with other
327 affective states
328 We further expected that the four boredom types would
329 correspond with other affective states (positive and nega-
330 tive) in a pattern consistent with the increasing levels of
331 negative valence of four boredom types. Indifferent, cali-
332 brating, searching, and reactant boredom were each
333 expected to relate more negatively than the previous type
334 to positive affective states, and each boredom type (in that
335 same order) was also expected to relate more positively
336 than the previous type to negative affective states.
337 Hypothesis 3: Prevalence of boredom types
338 Finally, we hypothesized that the prevalence of the four
339 proposed boredom types would be moderated by the spe-
340 cific characteristics of a given situation. Boredom types of
341 positive or low negative valence and low arousal were
342 expected to be more frequently experienced in non-
343 achievement situations. In contrast, boredom types that are
344 high in negative valence and arousal were expected to be
345 more often experienced in achievement situations.
346 Method
347 Sample and data collection
348 To investigate our research hypotheses, two studies were
349 conducted employing similar data collection protocols but
350 different samples (university students vs. high school stu-
351 dents). The studies thus allow for tests of generalizability
352 across different age groups while controlling for potential
353 method biases.
354 Study 1
355 The first sample consisted of 63 German university students
356 (66 % female) with a mean age of 24.08 years (SD =4.14;
357
range =[19.92; 45.08]). Students participated on a volun-
358
tary basis and were recruited primarily from psychology
359
courses (45 %) and education programs (35 %). The
360
remaining participants (20 %) were enrolled in Sociology,
361
Law, Arts, Physics, Politics, Literature, and Sports programs.
362
Data were collected using the experience sampling
363
method (Csikszentmihalyi and Larson 1987; Hektner et al.
364
2007). Following initial instruction on how to use the
365
personal digital assistant (PDA) devices, students’ PDA
366
responses were assessed over a two-week period (the
367
instructor was contacted in case of technical problems).
368
Consistent with the aim of obtaining representative data on
369
individuals’ experiences throughout an entire day, our
370
assessment employed a time randomizing procedure (sig-
371
nal-contingent sampling; see Hektner et al. 2007), in which
372
the number of signals, the earliest and latest possible sig-
373
nal, as well as the minimal time lag were used as default
374
parameters. More specifically, the PDAs emitted six audi-
375
ble signals per day between 10 a.m. and 10 p.m., after
376
which participants were asked to immediately complete a
377
digital questionnaire on the PDA screen. When this was not
378
feasible (e.g., during exams), participants were instructed
379
to complete the questionnaire as soon as possible thereafter
380
prior to the questionnaire expiring 5 min later. Students
381
who could not complete the questionnaire immediately
382
were instructed to describe their current experiences (i.e., a
383
state assessment) and were explicitly informed not to ret-
384
rospectively describe their experiences when the signal
385
occurred in order to minimize recall bias.
386
Study 2
387
The second sample consisted of 80 German high school
388
students (58 % female) from grade 11 with a mean age of
389
17.05 years (SD =0.54; range =[15.92; 18.58]). Partici-
390
pants were randomly selected from 25 classrooms from 9
391
German high schools (school type: Gymnasium). With few
392
exceptions, German high school classes occur on weekday
393
mornings between approximately 8 a.m. and 1 p.m., with
394
students completing homework or having leisure time in
395
the afternoon hours.
396
As in Study 1, data collection employed the experience
397
sampling method, in which students were provided the PDA
398
devices and instructed to activate their device whenever they
399
attended a class in a core school subject (i.e., mathematics,
400
German, or English). Core subjects were selected as they
401
were required of all students and to restrict the time com-
402
mitment for this study. Upon activating the device, it emitted
403
an audible signal within the next 40 min (combination of
404
event and signal-contingent sampling; see Hektner et al.
405
2007). In addition, students were randomly signaled three
406
times between 2 p.m. and 10 p.m. on weekdays and 6 times
407
between 10 a.m. and 10 p.m. on weekends (signal-
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
408 contingent sampling; focus was homework or leisure activ-
409 ities). Students were requested to complete the digital
410 questionnaire on the PDA screen immediately following
411 each auditory signal, and if responding later (e.g., during an
412 exam), to refer to their current experiences as opposed to
413 their experiences when the signal occurred (the question-
414 naire expired if not completed within 5 min). To minimize
415 classroom disruption, class teachers were informed of the
416 study protocols, all teachers agreed to student participation,
417 and the number of participants did not exceed four students
418 per class (total class size was *30 students).
419 Study measures
420 The same set of variables was assessed in Studies 1 and 2
421 using single items. Each of the study variables was assessed
422 with respect to the specific activity in which students were
423 currently engaged (cf., Clore 1994), thus providing a clearer
424 and more direct assessment of state experiences than would
425 global measures pertaining to more general situations.
426 Current activity
427 To determine the specific nature of the situation in which
428 participants completed the state questionnaires, the item
429 ‘‘What is the main type of activity in which you are cur-
430 rently engaged?’’ was provided with the following two
431 response options: (1) an achievement activity (e.g., classes,
432 lectures, studying in the library, homework, studying at
433 home, job setting), and (2) not an achievement activity (e.g.,
434 sleeping, eating, leisure time). Thus, subjects were prompted
435 to answer whether they regarded their current activity as a
436 situation in which achievement was or was not salient.
437 Based on students’ responses, a dichotomous variable indi-
438 cating the situation type during survey completion was
439 constructed (achievement =1, non-achievement =0).
440 Intensity of boredom and other affective states
441 The intensity of students’ current emotional experiences,
442 subjective well-being, and satisfaction were assessed using
443 multiple single-item measures with each item formulated
444 as follows: ‘‘While engaging in this activity, how strongly
445 do you experience [construct].’’ The specific constructs
446 assessed included (1) boredom, (2) well-being, (3) satis-
447 faction, (4) enjoyment, (5) anger, and (6) anxiety. The
448 response format for each item was a 5-point Likert scale
449 ranging from 1 (not at all)to5(very strongly).
450 Valence and arousal associated with boredom experiences
451 Following responses of 2 or higher on the questionnaire
452 item on boredom intensity—indicating that, at minimum, a
453
small intensity of boredom was being experienced—the
454
questionnaire presented two follow-up items measuring the
455
perceptions of valence and arousal that were associated
456
with the boredom experience (conditional assessment). The
457
items for both valence and arousal were formulated as
458
follows: ‘‘At this moment, how does it feel to be bored?’’
459
By formulating questions this way, students were explicitly
460
instructed to refer to boredom when answering this ques-
461
tion (even if other emotions might have been also experi-
462
enced by students at this moment). Response options for
463
valence consisted of a 5-point Likert scale ranging from 1
464
(positive)to5(negative), with higher scores indicating
465
greater negative valence, and for arousal included a 5-point
466
Likert scale ranging from 1 (calm)to5(fidgety).
467
Data analysis
468
As the focus of the present study was on types of academic
469
boredom, only those assessments in which students repor-
470
ted a minimum of intensity of boredom (at least 2 on the
471
Likert scale) were included in the analyses. Data obtained
472
in both studies reflect a two-level structure, with measures
473
at specific assessment points nested within persons. The
474
two-level structure of the data was taken into account in
475
our analyses through the use of complex commands in
476
Mplus 5.2 software (Muthe
´n and Muthe
´n2007).
477
Hypothesis 1
478
To address Hypothesis 1 (boredom types; H1), latent pro-
479
file analysis (LPA; Muthe
´n and Muthe
´n2000) was con-
480
ducted to identify responses that have similar patterns of
481
valence and arousal [LPA is also known as a latent class
482
analysis (LCA) when conducted with observed continuous
483
variables]. The conceptual goal of an LPA is to detect
484
unobserved heterogeneity in a sample so as to reveal
485
homogenous subsamples of responses that share a similar
486
pattern (in our case, groups of assessments within students;
487
Muthe
´n2001). However, LPA differs from cluster analysis
488
in that it is model-based and probabilistic in nature (Nylund
489
et al. 2007).
490
Latent profile analysis assumes that a categorical latent
491
variable underlies the observed outcome variables and
492
determines the structure of the response pattern, thus
493
defining the class membership. More specifically, LPA
494
seeks to identify the smallest number of latent classes that
495
sufficiently describe the association between observed
496
variables. Latent classes are created in such a way that
497
indicator variables are statistically independent within
498
classes. For this study, the latent classes were expected to
499
represent the boredom types. To determine the optimal
500
number of classes, the Bayesian Information Criterion
501
(BIC; Schwarz 1978) was used. The BIC accounts for the
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
502 log likelihood of a model, the number of model parameters,
503 and the sample size (Nylund et al. 2007). BIC values thus
504 provide relative information about different models, with
505 lower BIC values corresponding to better models. The BIC
506 has been found in Monte Carlo simulations to perform well
507 compared to other fit indices in identifying the correct
508 number of underlying classes (Nylund et al. 2007).
509 Hypothesis 2
510 To address Hypothesis 2 (associations between boredom
511 types and other affective states; H2), the means for the
512 criterion constructs consisting of well-being, satisfaction,
513 enjoyment, anger, and anxiety were assessed for each of
514 the boredom types. The most likely class membership
515 calculated by the LPA is a probability-based score, and
516 analyses based solely on this measure do not account for
517 the possibility that boredom experiences that belong to the
518 same class may markedly differ in their probabilities of
519 class membership (Clark and Muthe
´n2009). To account
520 for such differences, an Mplus feature allowing for mean
521 comparisons on the basis of pseudo-class draws was
522 employed (Wang et al. 2005). In this analysis, several
523 random draws are made from each individual’s posterior
524 probability distribution to determine class membership,
525 resulting in different pseudo-groups between which mean
526 comparisons regarding auxiliary aspects can be computed
527 (Clark and Muthe
´n2009).
528 Hypothesis 3
529 To address Hypothesis 3 (associations between boredom
530 types and situation type) the analytical procedure adopted
531 for Hypothesis 2 was applied. The analysis allows us to
532 compare the means of each assumed boredom type
533 between the levels of the dichotomous variable distin-
534 guishing between non-achievement and achievement
535 situations.
536 Results
537 Descriptive statistics
538 Descriptive analyses of boredom experiences indicated that
539 among participants in Study 1/2 (university/high school
540 samples), 63/80 students reported at least some degree of
541 boredom (at least 2 on the Likert scale) on 1,103/1,432
542 assessments out of a total of 3,945/3,645 assessments. Of
543 the assessments on which boredom was indicated, 587/947
544 were submitted in a reported achievement situation and
545 516/485 were completed in a reported non-achievement
546 situation.
547
Table 1presents means and standard deviations of
548
boredom valence, boredom arousal, well-being, satisfac-
549
tion, enjoyment, anger, and anxiety for participants who
550
reported a minimum of intensity of boredom (at least 2 on
551
the Likert scale; see ‘‘bored’’ column) as well as for all
552
assessments across participants (see ‘‘all assessments’’
553
column).
554
The correlation between the dimensions of valence and
555
arousal was rather small in Study 1 (r=.23; p\.001;
556
university sample) and negligible in Study 2 (r=-.05;
557
p=.270; high school sample). The strength of these cor-
558
relations is consistent with the results of previous real-time
559
assessments (Goetz et al. 2010).
560
Boredom types (H1)
561
Table 2presents the results of the LPA. Due to the scaling
562
of our variables (valence, arousal), the maximum number
563
of classes could be 25 (5 95). For both samples, the BIC
564
was lowest for a solution indicating five latent classes. The
565
entropy of the five-class solution was 0.94/0.89 in Study
566
1/2 indicating a satisfactory certainty for class membership.
567
The probabilities of class membership by latent class are
568
presented in Table 3. In Study 1/2, the probabilities that an
569
assessment classified as belonging to latent class kin fact
570
belongs to class krange from [0.74; 1.00]/[0.79; 1.00] with
571
a mode of each 1.00. These results indicate that
Table 1 Means and standard deviations of scales
Study 1 Study 2
Bored All
assessments
Bored All
assessments
M SD M SD M SD M SD
Valence 3.34 1.24 –
a
–
a
3.46 1.21 –
a
–
a
Arousal 2.50 1.25 –
a
–
a
2.05 1.20 –
a
–
a
Well-being 2.73 1.02 2.94 1.19 2.53 1.22 2.94 1.37
Satisfaction 2.72 1.03 2.91 1.19 2.51 1.20 2.94 1.34
Enjoyment 2.48 1.03 2.70 1.19 2.28 1.20 2.78 1.40
Anger 1.85 1.04 1.62 1.03 1.80 1.14 1.56 1.02
Anxiety 1.56 0.91 1.50 0.91 1.44 0.89 1.34 0.82
Study 1: university student sample. Study 2: high school student
sample. Answer formats were 1 (positive)to5(negative) for negative
valence, 1 (calm)to5(fidgety) for arousal, and 1 (not at all)to5(very
strongly) for well-being, satisfaction, enjoyment, anger, and anxiety.
Bored =assessments indicating a magnitude of at least 2 on the
boredom item. Bored: Study 1 N=1,103 assessments in 63 univer-
sity students; Study 2 N=1,432 assessments in 80 high school stu-
dents. All assessments: Study 1 N=3,945 assessments in 63
university students; Study 2 N=3,645 assessments in 80 high school
students. In these analyses, the two-level structure of the data was
taken into account
a
Valence and arousal were assessed exclusively for situations in
which boredom was experienced
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
572
assessments within students can with satisfactory certainty
573
be assigned to five latent classes as outlined by the LPA.
574
LPA provides probability scores for each boredom
575
assessment concerning latent class membership, resulting
576
in five different scores for each boredom assessment in the
577
present study. The probability score for latent membership
578
in a specific class indicates the probability of belonging to
579
this class. For each boredom assessment, the five proba-
580
bility scores of latent class membership thus correspond to
581
the five classes (that when summed add up to 1). These
582
probability scores of latent class membership were subse-
583
quently logit-transformed for the purposes of linear
584
regression analyses (Clark and Muthe
´n2009).
585
Table 4provides means for arousal and valence, as well
586
as numbers of assessments within classes for the five classes
587
identified by the LPA. The results are notably similar for
588
Studies 1 and 2 (university/high school samples). Class 1 is
589
characterized by slightly positive valence (2.13/2.30 for
590
Studies 1 and 2, respectively; values below the midpoint of
591
the 5-point scale ranging from positive [1] to negative [5])
592
and very low arousal (1.00/1.00; also clearly below the
593
midpoint of the scale ranging from calm [1] to fidgety [5]).
594
Classes 2 and 3 have relatively similar valence levels (scale
595
ranging from positive to negative; 3.20/3.31 for Class 2;
596
3.33/3.41 for Class 3) but differ with Class 2 reflecting lower
597
levels of arousal (2.00/2.00) than Class 3 (3.00/3.00). Class 4
598
is characterized by high levels of both negative valence
599
(3.81/3.67) and arousal (4.41/4.26). Class 5 is characterized
600
by high levels of negative valence (4.07/4.16) combined
601
with low levels of arousal (1.00/1.00). It is important to note
602
that the integers in the class means concerning the arousal
603
variable are a result of the LPA algorithm that aims to obtain
604
maximum homogeneity within classes. As the two variables
605
(valence, arousal) are each measured on an ordinal, 5-point
606
Likert scale, and the variances of arousal were relatively
607
small due to maximal homogeneity within classes, integer
608
values were produced. However, beyond the integer values
609
observed for the arousal dimension, there is considerable
610
variance in the valence variable for each class (standard
611
deviations in Study 1 for Classes 1–5: 0.85/1.14/0.94/1.30/
612
0.48; standard deviations in Study 2 for Classes 1–5: 0.49/
613
0.96/1.04/1.32/0.89).
614
With respect to Hypothesis 1, mean levels in valence
615
and arousal related to Classes 1–4 reflect the proposed
616
boredom types, with Class 1 representing ‘‘indifferent
617
boredom,’’ Class 2 representing ‘‘calibrating boredom,’’
618
Class 3 representing ‘‘searching boredom,’’ and Class 4
619
representing ‘‘reactant boredom.’’ However, Class 5 was
620
not anticipated. According to its levels of valence and
621
arousal, we labeled this type of boredom ‘‘apathetic
622
boredom.’’
623
To investigate how strongly boredom intensity was
624
related to latent boredom classes, means of boredom
Table 2 Information criteria values of class solutions for boredom classes
Number of classes
Study 1 Study 2
123456123456
No. of free param. 4 7 10 13 16 19 4 7 10 13 16 19
Loglikelihood -3,616.51 -3,520.04 -3,468.26 -3,439.08 -3,230.15 -3,221.12 -4,577.79 -4,351.04 -4,247.61 -3,605.30 -3,578.19 -3,578.19
BIC 7,261.05 7,089.13 7,006.57 6,969.24 6,572.40 6,575.34 9,184.64 8,752.92 8,567.85 7,305.02 7,272.60 7,294.39
Entropy 0.74 0.83 0.81 0.94 0.98 0.92 0.89 1.00 0.89 0.84
BIC Bayesian Information Criterion. The two-level structure of the data was taken into account. Study 1—university student sample: N
Level 1
=1,103, N
Level 2
=63; Study 2—high school
student sample: N
Level 1
=1,432, N
Level 2
=80
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
625 intensity of the latent classes as indicated by the LPA were
626 calculated. Further, we calculated correlations between the
627 probability scores for latent class membership and boredom
628 intensity. For both analyses, logit-transformed probabilities
629 of class membership were used. Means and standard errors
630 (in parentheses) of boredom strength (probabilities of class
631 membership) for latent Classes 1–5 in Study 1 (university
632 student sample) were: 2.53 (0.07)/2.61 (0.04)/2.64 (0.05)/
633 3.27 (0.07)/2.85 (0.09); and in Study 2 (high school student
634 sample) were: 2.80 (0.07)/3.06 (0.06)/3.17 (0.06)/3.48
635 (0.08)/3.33 (0.06). Thus, in both studies an increasing
636 intensity from Class 1 to Class 4 was found, and Class 5
637 showed a level of boredom intensity between those found
638 for Classes 3 and 4. However, the range in boredom inten-
639 sity across Classes 1–5 was rather small in both studies (the
640 difference between the highest and the lowest value was
641 0.74/0.68 in Study 1/2, with possible difference of 4.00 in
642 both studies). This finding indicates that the probability
643 scores for latent class membership were not strongly related
644 to boredom intensity. Correlations between the logit-trans-
645 formed probabilities of boredom class membership and the
646 variable of boredom intensity for Classes 1–5 were for Study
647 1: -0.08 (p=.030)/-0.12 (p=.011)/-0.09 (p=.027)/
648 0.30 (p\.001)/-0.04 (p=.386); and for Study 2: -0.09
649 (p=.011)/-0.06 (p=.048)/0.00 (p=.902)/0.13 (p=
650 .002)/0.01 (p=.746). In Study 1, one correlation of med-
651 ium size (0.30) was found for Class 4 (reactant boredom).
652 As the correlation was also positive and significant in Study
653 2 (0.13), the observed tendency was for the logit-trans-
654 formed probabilities of being a member of Class 4 to
655 coincide with higher levels of boredom intensity. However,
656 it is important to note that the effect sizes of the correlations
657 observed were negligible or small in magnitude.
658 The relative frequencies of Classes 1–4 were similar for
659 Studies 1 and 2 (see Table 4). However, they were dif-
660 ferent for Class 5 (apathetic boredom), with 10 % in Study
661 1 (university students) and 36 % in Study 2 (high-school
662 students).
663
In order to explore whether the different boredom types
664
occur randomly within individuals over time, or whether
665
there was a higher probability for some individuals to
666
experience specific types of boredom relative to others, we
667
obtained intraclass correlations (ICCs) of the probability
668
scores of the boredom class membership across assessment
669
points. ICCs of the probability scores of the boredom class
670
membership for Classes 1/2/3/4/5 were as follows: Study
671
1=0.22 (p\.001)/0.10 (p\.001)/0.09 (p\.001)/0.18
672
(p\.001)/0.24 (p\.001); (median: 0.18); Study
673
2=0.19 (p\.001)/0.06 (p=.002)/0.19 (p\.001)/0.11
674
(p\.001)/0.08 (p\.001); (median: 0.11; for both studies,
675
each ICC value was significantly different from zero).
676
These scores indicate that a large proportion of the variance
677
(cf., Papaioannou et al. 2004) of the probability of expe-
678
riencing a specific boredom type was within-person vari-
679
ance. However, as these scores were significantly different
680
from zero, and 7 of the 10 scores were above or equal to
681
0.10 (a threshold suggested by Papaioannou et al. 2004),
682
this also suggests that a meaningful, albeit relatively small,
683
proportion of variance was between-person variance indi-
684
cating a higher probability for some individuals to expe-
685
rience specific types of boredom relative to others.
686
Results indicating the five-class solutions observed in
687
both Studies 1 and 2 are presented graphically in Fig. 2.
688
Dots within the circles (dark-colored: Study 1; light-col-
689
ored: Study 2) indicate mean levels of negative valence and
690
arousal for the five classes in each study (see Table 4). For
691
both studies, the size of the circles (circular areas) repre-
692
sents the number of assessments (across all subjects) within
693
each class relative to the total number of assessments (also
694
across all subjects). Thus, the sizes of circular areas for the
695
two studies are directly comparable. This figure illustrates
696
that aside from differences in the relative frequency of
697
boredom classes, virtually identical results were observed
698
in both studies with respect to the number of classes
699
indicated by the LPA and their locations on the dimensions
700
of negative valence and arousal.
Table 3 Probabilities of latent class membership by latent class
Latent class Probability of latent class membership
Study 1 Study 2
1234512345
1 0.74 0.00 0.00 0.00 0.26 0.89 0.00 0.00 0.00 0.11
2 0.00 1.00 0.00 0.00 0.00 0.00 1.00 0.00 0.00 0.00
3 0.00 0.00 1.00 0.00 0.00 0.00 0.00 1.00 0.00 0.00
4 0.00 0.00 0.00 1.00 0.00 0.00 0.00 0.00 1.00 0.00
5 0.10 0.00 0.00 0.00 0.90 0.21 0.00 0.00 0.00 0.79
Study 1—university student sample: N=1,103; Study 2—high school student sample: N=1,432. The sum of probability values of membership
for a given latent class (row) is equal to 1
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
701 Relations between boredom types and other affective
702 states: H2
703 Figure 3depicts mean levels of well-being, satisfaction,
704 enjoyment, anger, and anxiety associated with each of the
705 five boredom classes for both the university student sample
706 (Study 1) and the high school student sample (Study 2). For
707 each of the five constructs in both studies, boredom classes
708 were generally found to differ with respect to mean levels
709 (main effect p\.001; based on an Mplus feature allowing
710 pseudo-group comparisons by accounting for the proba-
711 bility of class membership; Clark and Muthe
´n2009).
712 However, these differences were not significant for some
713 between-class contrasts on specific constructs (e.g., Classes
714 4 and 5 did not differ on well-being in the high school
715 sample).
2
Nevertheless, a clear overall picture emerged for
716 both samples. Class 1 (indifferent boredom) corresponded
717 with the most positive profile of emotions and well-being
718 (relatively high means for positive affect and well-being,
719 relatively low means for negative affect). Conversely,
720 Class 4 (reactant boredom) was characterized by the most
721 negative profile (relatively low means for positive affect
722 and well-being, relatively high means for negative affect,
723 especially for anger). The valences of the profiles on the
724 criterion measures for Classes 2 (calibrating boredom) and
725 3(searching boredom) were between the more extreme
726 profiles observed for Classes 1 and 4. Finally, the profile
727 found for Class 5 (apathetic boredom, not hypothesized)
728 indicated relatively low levels of positive affect,
729
satisfaction, and well-being, as well as relatively low levels
730
of negative affective states.
731
Relations between boredom types and situation type:
732
H3
733
For each of the five boredom classes in the two study
734
samples, the percentage of students’ boredom experiences
735
that were reported in an achievement situation, as com-
736
pared to situations not related to achievement, was calcu-
737
lated.
3
The mean percentages of boredom reports occurring
738
in achievement situations corresponding to latent Classes 1
739
through 5 were for Study 1: 37/55/53/57/59; and for Study
740
2: 55/65/68/75/72. For both samples, the classes signifi-
741
cantly differed with respect to the percentage of boredom
742
reports occurring in achievement situations (p\.001;
743
based on an Mplus feature allowing pseudo-group com-
744
parisons by accounting for the probability of class mem-
745
bership; Clark and Muthe
´n2009). Beyond this main effect,
746
however, not all classes differed from each other with
747
respect to situation type (e.g., Classes 4 and 5 did not
748
significantly differ for the high school sample; see footnote
749
2).
750
Concerning differences between the two studies, a
751
greater proportion of boredom reports in achievement sit-
752
uations was found in Study 2 (high school students) rela-
753
tive to Study 1 (university students; Study 1/2: 53 %/
754
66 %). Figure 4shows the ratio of boredom reports
755
occurring in achievement situations in relation to boredom
756
reports occurring in non-achievement situations in one
757
latent class. The reported values were calculated in two
Table 4 Means and standard deviations (arousal, valence) for the five boredom classes and number of measures within latent classes
Latent class Boredom type Study 1 Study 2
MNo. measures within class MNo. measures within class
Valence Arousal Valence Arousal
1 Indifferent 2.13 1.00 178 2.30 1.00 153
2 Calibrating 3.20 2.00 331 3.31 2.00 297
3 Searching 3.33 3.00 244 3.41 3.00 253
4 Reactant 3.81 4.41 244 3.67 4.26 214
5 Apathetic 4.07 1.00 106 4.16 1.00 510
Study 1: university student sample. Study 2: high school student sample. Response formats were 1 (positive)to5(negative) for valence and 1
(calm)to5(fidgety) for arousal. The variances are fixed in the analysis using the Mplus standard procedure. Study 1: SD
arousal
=0.23; SD
negative
valence
=1.11. Study 2: SD
arousal
=0.17; SD
negative valence
=1.02
2FL01
2
With respect to Hypotheses 2 and 3, a total of 120 pairwise
2FL02 comparisons were calculated [for each construct 10 comparisons
2FL03 between the 5 classes 96 constructs (well-being, satisfaction,
2FL04 enjoyment, anger, anxiety, dichotomous variable achievement vs.
2FL05 non-achievement situation) 92 samples (university vs. high school
2FL06 student sample)]. Of the 120 comparisons, 18 were not significant in
2FL07 the university student sample and 22 were not significant in the high
2FL08 school student sample [in the case of significance: all ps\.017 for
2FL09 the university student sample; with one exception (p=.037) all
2FL10 ps\.019 for the high-school student sample].
3FL01
3
Correlations (calculated across all assessments) between the
3FL02affective state measures and the situation variable (coded as follows:
3FL030 =non-achievement situation, 1 =achievement situation) were as
3FL04follows for Studies 1/2: -.30/-.32 (ps\.001) for well-being; -
3FL05.28/-.14 (ps\.001) for satisfaction; -.28/-.31 (ps\.001) for
3FL06enjoyment; .14/.06 (p=.011/.056) for anger; .16/.08 (p=.006/.026)
3FL07for anxiety.
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
758 steps: First, the percentages of boredom reports occurring
759 in achievement situations relative to all boredom reports
760 were calculated separately for each latent class. Second, we
761 divided these values by the overall percentage of boredom
762 reports in achievement situations collapsed across boredom
763 classes.
764 A relatively clear picture emerged for both studies: The
765 values below 1 found for Class 1 (indifferent boredom)
766 indicated that if boredom was reported among assessments
767 in this latent class, there was a greater than chance prob-
768 ability that it was reported in a non-achievement context. In
769 contrast, the values above 1 observed for Classes 4 (reac-
770 tant boredom) and 5 (apathetic boredom) indicated an
771 above-chance probability that boredom was instead repor-
772 ted in achievement settings in these classes. The values for
773 Classes 2 (calibrating boredom) and 3 (searching bore-
774 dom) fell between those for Classes 1 and 4/5, and were
775 very close to 1, suggesting that the percentage of boredom
776 reports occurring in achievement situations were similar to
777 the percentages of boredom reports across situations.
778 Discussion
779 The present research aimed to empirically investigate dif-
780 ferent types of boredom as experienced in real-life settings
781 based on hypotheses derived from preliminary qualitative
782 research by Goetz and Frenzel (2006). The study hypoth-
783 eses proposed that individuals’ boredom experiences may
784 be differentiated with respect to valence (positive to neg-
785 ative) and arousal (Hypothesis 1), that boredom types
786 should differ in relation to other affective states (Hypoth-
787 esis 2), and that differential prevalence should be found for
788 boredom experiences in achievement versus non-achieve-
789 ment situations (Hypothesis 3). Hypothesis 1 concerned the
790
internal validity of the four-part boredom typology,
791
whereas Hypotheses 2 and 3 addressed the external validity
792
of the boredom types in relation to conceptually relevant
793
constructs and situation types. Data from two experience
794
sampling studies evaluated these hypotheses with samples
795
of university students (Study 1) and high school students
796
(Study 2).
797
Boredom type classification
798
The primary aim of the present study was to classify
799
individuals’ boredom experiences along the dimensions of
800
valence (positive to negative) and arousal. Thus, our study
801
is in line with previous studies on discrete emotions that
802
classify types of affective experiences according to the
803
dimensions of valence and arousal as opposed to a single
804
dimension (see Russell 1980; Watson and Tellegen 1985).
805
In other words, this research aligns with previous studies in
806
adopting a dimensional approach that takes the combina-
807
tion of two dimensions into account when classifying
808
affective experiences. As an extension of previous studies,
809
our focus is thus not on discrete emotional experiences
810
(e.g., enjoyment, pride, relief, anger, anxiety, boredom),
811
but rather on subtypes of one specific emotion, in this case
812
subtypes of boredom.
813
In both studies, LPA results suggested five boredom
814
classes—one more latent class than was initially observed
815
by Goetz and Frenzel (2006). The first four classes
816
observed in both studies were directly consistent with those
817
found in Goetz and Frenzel’s (2006) typology, and thus
818
support the internal validity of the proposed four boredom
819
types. Class 1 showed slightly positive valence and very
820
low arousal, most closely reflecting ‘‘indifferent boredom’’
821
(relaxed, withdrawn, indifferent). Classes 2 and 3 had
822
slightly negative valence and higher arousal than Class 1,
Class 1
Class 2
Class 3
Class 4
Class 5
Valence
negativepositive
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
5.00
1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
Study 1
University Student Sample
Study 2
High School Student Sample
yt
eg
difmla
c
Arousal
Indifferent
boredom
Calibrating
boredom
Searching
boredom
Reactant
boredom
Apathetic
boredom
Fig. 2 Mean values of negative
valence and arousal for latent
boredom classes. Note Dots
within circles indicate the mean
levels of valence and arousal for
each class in each study. Circle
size represents the relative
number of measures for each
class
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
823 with Class 3 having higher arousal than Class 2. Accord-
824 ingly, whereas Class 2 appeared to represent ‘‘calibrating
825 boredom’’ (uncertain, receptive to change/distraction),
826 Class 3 more closely resembled ‘‘searching boredom’’
827 (restless, active pursuit of change/distraction). The fourth
828 boredom type proposed by Goetz and Frenzel (2006),
829 namely ‘‘reactant boredom,’’ was represented by Class 4 in
830 each study—a class having high levels of negative valence
831 and relatively high levels of arousal (highly reactant,
832 motivated to leave the situation for specific alternatives).
833 Finally, the fifth and unanticipated boredom type was
834 characterized by a high level of negative valence and very
835 low arousal. In contrast to the hypothesized boredom
836 typology described above, this finding suggests an addi-
837 tional class of boredom experiences that are especially
838 unpleasant and associated with very low arousal levels.
839 Given the similarity of this construct description to learned
840 helplessness or depression, this boredom type was referred
841 to as ‘‘apathetic boredom.’’
842 In interpreting our findings concerning the different
843 boredom types, an intuitive assumption may be that the
844 specific boredom types in fact are determined by the overall
845
intensity of the boredom experience (e.g., with indifferent
846
boredom being milder and less intense relative to reactant
847
boredom). However, despite a consistent tendency across
848
both studies for reactant boredom to correspond with higher
849
levels of boredom intensity, the five types of boredom were
850
rather weakly related to boredom strength in terms of effect
851
size. As such, the present findings provide empirical support
852
for different boredom types, but not simply as a function of
853
the intensity of the boredom experience.
854
To summarize, the present LPA results provide empir-
855
ical support for the four hypothesized boredom types as
856
differentiated based on the dimensions of valence and
857
arousal. Further, these findings suggest that the constructs
858
of calibrating and searching boredom (Classes 2 and 3)
859
may be more similar than initially assumed with respect to
860
valence and that an especially unpleasant and debilitating
861
type of boredom similar to apathy may also be experienced
862
in real-life situations.
863
More generally, the present research indicates that there
864
exists a considerable degree of variance within the con-
865
struct of boredom. This ‘‘within-boredom-variance’’ was
866
observed not only with respect to the arousal dimension,
University student sample
C1
C2
C3
C4
C5
Latent
class
1.00
1.50
2.00
2.50
3.00
3.50
4.00
Indifferent
Boredom
type
Calibrating
Searching
Reactant
Apathetic
Well-being Satisfaction Enjoyment Anger Anxiety
Well-being Satisfaction Enjoyment Anger Anxiety
High school student sample
C1
C2
C3
C4
C5
Latent
class
1.00
1.50
2.00
2.50
3.00
3.50
4.00
Indifferent
Boredom
type
Calibrating
Searching
Reactant
Apathetic
Fig. 3 Affective experiences in boredom classes. Note Means of
Likert scale values from 1 to 5 of well-being, satisfaction, enjoyment,
anger, and anxiety associated with each of the five latent boredom
classes both for the university student sample (Study 1) and the high
school student sample (Study 2)
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
867 consistent with prior mixed findings in the boredom liter-
868 ature, but also for the valence dimension, which contradicts
869 prior limited findings that define experiences of boredom as
870 having slightly low negative valence. Thus, the present
871 research evaluated the relatively unexplored emotion of
872 boredom and provided support for preliminary qualitative
873 findings (Goetz and Frenzel 2006) as well as long-standing
874 theoretical assertions that boredom may be best understood
875 as multiple ‘‘boredoms’’ that differ based on valence and
876 arousal.
877 Concerning the five boredom types observed in this
878 research, an additional finding unrelated to the initial study
879 hypotheses is also of interest: Individuals do not randomly
880 experience the different boredom types over time but rather
881 tend to experience specific types of boredom. This asser-
882 tion is supported by a meaningful amount of between-
883 person variance (up to 24 % of the total variance) found for
884 the probability scores of experiencing the five boredom
885 types (for an interpretation of the extent of between-person
886 variance, see Papaioannou et al. 2004). Thus, we can
887 speculate that experiencing specific boredom types might,
888 to some degree, be interpreted as being due to personality-
889 specific dispositions. Alternatively, the between-person
890
variance in experiencing specific types of boredom might
891
be due to differences between individuals in the boredom-
892
arousing situations they encounter (e.g., some students
893
taking a specific class that promotes reactant boredom).
894
Relations between boredom types and other affective
895
states
896
To evaluate the external validity of the observed boredom
897
types, relationships between boredom classes and other
898
affective states were assessed (as hypothesized based on
899
boredom class valence; see Fig. 1, x axis). The present
900
findings provide clear empirical support for the anticipated
901
relations of boredom and a number of positive affective
902
states (enjoyment, well-being, satisfaction) as well as
903
several negative affective states (anger, anxiety).
904
In both studies, experiences of Class 1 boredom (indif-
905
ferent) corresponded with a generally more positive profile
906
than the other boredom types, a finding consistent with the
907
assumptions of Goetz and Frenzel (2006). Further, Class 4
908
boredom (reactant) was instead associated with a signifi-
909
cantly more negative profile. Class 5 boredom (apathetic), a
910
subtype not initially proposed by Goetz and Frenzel (2006),
Achievement settings
High school student sample
C1
C2
C3
C4
C5
Achievement settings
University student sample
C1
C2
C3
C4
C5
Latent
class
Latent
class
0.50
0.60
0.70
0.80
0.90
1.00
1.10
1.20
0.50
0.60
0.70
0.80
0.90
1.00
1.10
1.20
Indifferent
Boredom
type
Calibrating
Searching
Reactant
Apathetic
Indifferent
Boredom
type
Calibrating
Searching
Reactant
Apathetic
Fig. 4 Relative portion of
achievement situations across
boredom types. Note The values
were calculated in two steps. In
the first step, the percentage of
boredom reports occurring in
achievement situations in
relation to all boredom reports
was calculated separately for
each latent class. In the second
step, these values were divided
by the overall percentage of
boredom reports in achievement
situations aggregated across
boredom classes. Thus, a value
above 1 means that this type of
boredom was experienced with
the above-chance probability in
learning and achievement
situations whereas a value
below 1 means that this type of
boredom was experienced with
the above-chance probability in
non-achievement situations
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
911 was similar to Class 4 boredom with respect to the negative
912 direction and the magnitude of relations with positively-
913 valenced measures of enjoyment, well-being, and satisfac-
914 tion. However, these more aversive boredom types were
915 shown to differ significantly in their relations with the neg-
916 ative affective states of anger and anxiety with much lower
917 levels of negative emotions being observed for Class 5 as
918 compared to Class 4 boredom. To summarize, our findings
919 provide empirical support for the external validity of the
920 proposed boredom types and suggest that indifferent and
921 reactant boredom represent the least and most aversive
922 boredom experiences, respectively. Further, we demonstrated
923 that calibrating and searching boredom types fell between
924 these extremes with respect to valence. Finally, the fifth
925 boredom type was shown to be associated with low levels of
926 both positive and negative affective states (e.g., anger).
927 Prevalence of boredom types across situation types
928 In line with Hypothesis 3, the present data revealed that the
929 relative frequency of boredom types that are positive or
930 low in negative valence are more commonly experienced in
931 non-achievement settings as compared to situations related
932 to learning and achievement. Conversely, we found that the
933 relative frequency of boredom types of high negative
934 valence were lower in non-achievement settings as com-
935 pared to achievement situations. Additionally, the relative
936 frequency of Class 1 boredom (indifferent) was shown to
937 be higher in non-achievement related activities as com-
938 pared to achievement contexts. This finding is consistent
939 with the preceding results showing this first type of bore-
940 dom to be the least unpleasant, even slightly pleasant, and
941 thus, more likely to be tolerated in non-achievement set-
942 tings than more aversive types of boredom. Further, this
943 finding may be explained by non-achievement situations
944 typically allowing for greater freedom to modify or escape
945 boring activities than achievement settings (for boredom-
946 specific coping strategies, see Nett et al. 2010,2011). In
947 addition to situational factors, there might also be an
948 interaction of situation and person variables leading to the
949 experience of specific boredom types. For example,
950 extroverts might be more prone than introverts to experi-
951 ence reactant boredom in situations that are hard to modify
952 or to leave (cf., Hill and Perkins 1985; Smith 1981).
953 With respect to the possible positive aspects of boredom
954 experiences (Seib and Vodanovich 1998; see Vodanovich
955 2003a for a review), it may be assumed that different types
956 of boredom can differ with respect to their potential to
957 initiate positive thoughts and actions. For instance, indif-
958 ferent boredom experienced mainly in non-achievement
959 settings may be related to constructive behaviors such as
960 stimulating greater self-reflection and creativity (cf., Baird
961 et al. 2012; Harrison 1984; Sio and Ormerod 2009). At the
962
same time, the potential benefits of boredom in more
963
restrictive achievement situations may be more limited.
964
Further, our studies revealed that indifferent boredom was
965
the least commonly experienced boredom type (16 % in
966
university students, 11 % in high school students). Hence,
967
the potential benefits associated with this type of boredom
968
are likely to be outnumbered by the negative consequences
969
of more aversive boredom types (cf., Pekrun et al. 2010).
970
A fifth boredom type: The case for apathetic boredom
971
The results of our analyses suggest that the preliminary
972
model of boredom experiences proposed by Goetz and
973
Frenzel (2006) should be expanded to include the fifth
974
boredom type: apathetic boredom. This boredom experi-
975
ence appears to be especially unpleasant, but differs from
976
the other highly aversive boredom type—reactive bore-
977
dom—in corresponding with low arousal and the absence
978
of both positive and negative affective states. Thus,
979
whereas reactive boredom is highly aversive and is asso-
980
ciated with high arousal, apathetic boredom is equally
981
aversive yet lacking in arousal—an emotion type more
982
similar to learned helplessness or depression (cf., Fenichel
983
1934,1951 for an early statement on this relationship).
984
This pattern is consistent with empirical findings showing
985
positive relations between boredom and depression
986
(Farmer and Sundberg 1986; Vodanovich 2003b). Of par-
987
ticular concern is the relative frequency of apathetic
988
boredom observed in the present research, namely with
989
respect to the university student sample in which it com-
990
prised 36 % of boredom experiences.
991
A five-class boredom typology
992
Expanding upon the preliminary model suggested by
993
Goetz and Frenzel (2006), findings from the present
994
research based on quantitative data obtained from real-
995
time assessments provide empirical support for the five-
996
class model of boredom experiences. As outlined in Fig. 5,
997
four boredom types are distinguished based on valence
998
(positive to negative) and arousal, with the fifth boredom
999
type (apathetic boredom) not falling in sequence with the
1000
others due to having very high negative valence combined
1001
with very low arousal. Our data further revealed that
1002
calibrating and searching boredom were more similar than
1003
the other boredom types with respect to the valence
1004
dimension. It is also important to note that the empirically
1005
derived model differs from the initially hypothesized
1006
model with respect to the relations between boredom types
1007
and phenomenologically similar constructs. Although the
1008
assumed relations to relevant constructs were found for the
1009
four hypothesized boredom types, the assumption that
1010
more negative types of boredom would coincide with other
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
1011 negative affective states was not fully supported by our
1012 data. More specifically, apathetic boredom was found to
1013 be highly aversive yet corresponded with low levels of
1014 both positive and negative affective states. However, it is
1015 important to note that as only the negative activating
1016 emotions of anger and anxiety were assessed in this study
1017 (Pekrun et al. 2002), it is possible that apathetic boredom
1018 might correspond with high levels of deactivating negative
1019 emotions such as sadness. Finally, unlike the hypothesized
1020 model, our model does not assume directional relations
1021 among boredom types as very little is known about the
1022 temporal transition from one boredom type to another.
1023 In Fig. 5, the average level of valence and arousal when
1024 experiencing boredom is plotted in relation to the five
1025 observed boredom types. As indicated by the proximity of
1026 averaged boredom experiences to the calibrating and
1027 searching boredom types, the present findings suggest that
1028 these two specific classes of boredom experiences are most
1029 likely to represent the ‘‘typical’’ boredom experience from
1030 the perspective of emotion prototypes (e.g., Armstrong
1031 et al. 1983; Clore and Ortony 1991; Johnson-Laird and
1032 Oatley 1989; Ortony et al. 1987; Pekrun et al. 2010; Russel
1033 1991). Moreover, our results significantly qualify this
1034 assertion in showing three types of boredom to be notably
1035 distant from this averaged boredom measure with respect
1036 to valence and arousal (indifferent,reactant, and apathetic
1037 boredom). The averaged boredom experience is located in
1038 the lower right quadrant of the figure—a classification
1039 based on valence and arousal that is in line with previous
1040 approaches to locating boredom according to its underlying
1041 dimensions (e.g., Russell 1980).
1042
It is important to note that the response options for our
1043
arousal dimension ranged from calm to fidgety, with fidgety
1044
likely not reflecting maximum arousal as would more
1045
extreme anchors, such as ‘‘highly agitated’’ or ‘‘panicked’’
1046
(cf., arousal scales ranging from ‘‘as calm as one can feel’’
1047
to ‘‘as aroused as one can feel’’ in Reisenzein 1994). Thus,
1048
when comparing the present five-part boredom typology
1049
and averaged boredom against classical circumplex mod-
1050
els, it is possible that all of the observed boredom types and
1051
averaged boredom may be located near the lower end of the
1052
arousal scale. In sum, our results (see Fig. 5) indicate that
1053
when plotted according to the classical dimensions of the
1054
circumplex model, the identified boredom types are pri-
1055
marily located in the quadrant reflecting negative valence
1056
and low arousal. At the same time, they seem to also reach
1057
or even extend beyond these borders into other quadrants
1058
(e.g., indifferent boredom as a low-arousal/pleasant expe-
1059
rience, reactant boredom as a high-arousal/unpleasant
1060
experience).
1061
Our findings do not contradict but rather expand the
1062
assumptions underlying circumplex models of affect in
1063
showing substantial within-boredom variance with respect
1064
to valence and arousal. As a consequence, from the per-
1065
spective of circumplex models, a specific subtype of a
1066
discrete emotion might be rather similar to a specific sub-
1067
type of another discrete emotion. For example, boredom
1068
that is negative in valence and high in arousal (i.e., reactant
1069
boredom) might be similar in valence and arousal to the
1070
‘‘typical’’ experience of anger. However, although there
1071
might be an overlap in emotions with respect to their levels
1072
of valence and arousal, they may nonetheless differ in other
1073
ways not captured by two-dimensional circumplex models.
1074
For example, it is possible that further investigation of
1075
dimensionally similar emotion types based on component
1076
definitions of emotions (e.g., Kleinginna and Kleinginna
1077
1981; Scherer 2000) could reveal differing components for
1078
emotions that are similar in terms of valence and arousal.
1079
In sum, although our approach is in line with circumplex
1080
models of emotions we do emphasize that the levels of
1081
valence and arousal previously assigned in this model to
1082
boredom represent averaged values that do not exclude
1083
within-boredom variance. Moreover, we believe that sim-
1084
ilar degrees of within-emotion variance found for boredom
1085
may also be characteristic for other discrete emotions.
1086
Study limitations
1087
One potential shortcoming of the present research concerns
1088
the samples assessed in that the two samples consisted of
1089
older students in higher levels of the educational system
1090
(university students and 11th grade students). Future
1091
research that includes samples of younger and/or older
Vale nce
Arous al
Indifferent
boredom
Searching
boredom
Reactant
boredom
Calibrating
boredom
Apathetic
boredom
negativepositive
ytegdifmlac
BOREDOM
Fig. 5 Five boredom types and mean boredom along valence and
arousal dimensions
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
1092 students (e.g., primary school students, mature university
1093 students) is needed to evaluate how strongly our results can
1094 be generalized across educational levels. Second, whereas
1095 the classroom represents a prototypical achievement setting
1096 in which boredom experiences can be analyzed, other rel-
1097 evant achievement settings such as workplace should also
1098 be investigated (e.g., office vs. manufacturing domains, full
1099 vs. part-time employment, etc.; cf., Grubb 1975; Pekrun
1100 and Frese 1992; Smith 1981; for a job-related boredom
1101 scale, see Lee 1986).
1102 Third, due the present study aim to collect a broad
1103 sample of students’ real-time experiences of boredom
1104 (randomized, no equidistant or continuous assessments), it
1105 was not possible to analyze the temporal ordering of dif-
1106 ferent boredom types. For example, it is possible that
1107 searching boredom may precede reactant boredom in
1108 classroom settings as was initially hypothesized by Goetz
1109 and Frenzel (2006). Experimental studies in which bore-
1110 dom experiences are manipulated could help to inform the
1111 research literature on this emotion and shed light on the
1112 possible temporal ordering and development of the pro-
1113 posed boredom types.
1114 Fourth, and related to the previous point, the assumption
1115 that subtypes of boredom may be rapidly induced or
1116 manipulated by environmental factors, as opposed to
1117 gradually developing from one subtype into another over
1118 time with respect to changes in levels of valence and
1119 arousal might be investigated in future research (e.g.,
1120 experimental studies). Future studies focusing on this
1121 aspect are further recommended to explore the use of more
1122 specific response formats (e.g., 11-point format or use of a
1123 slider) or multi-item scales assessing valence and arousal in
1124 order to have more continuous values on both constructs.
1125 Finally, due to the restricted number of items in the state
1126 assessment (to not compromise the validity of the real-time
1127 assessment), the number of external variables used for
1128 validating qualitative differences between the boredom
1129 types was limited. Further, we did not gauge the extent to
1130 which the types of boredom differed with respect to the
1131 component processes involved beyond the dimensions of
1132 valence and arousal. Future studies that investigate the
1133 degree to which the suggested boredom types share simi-
1134 larities with respect to specific components as outlined in
1135 component-process definitions of emotions are encouraged
1136 (e.g., with respect to the cognitive component: whether
1137 altered perceptions of time are found for all boredom
1138 types). Related to this point, future studies might also be
1139 designed to allow for an empirical investigation of whether
1140 students referred in their description of boredom to ‘‘pure’’
1141 experiences of boredom or whether other emotions expe-
1142 rienced at that time had an impact on those descriptions
1143 (cf., Van Tilburg and Igou 2012). In our study we cannot
1144 exclude the possibility that, to a certain degree, boredom
1145
types reflect mixed emotions (e.g., Larsen and McGraw
1146
2011) despite our best efforts to ensure that students
1147
referred exclusively to boredom when answering questions
1148
about valence and arousal.
1149
Conclusion and implications
1150
The results of this study suggest that individuals indeed
1151
experience different types of boredom that may be quali-
1152
fied along the dimensions of valence and arousal. Future
1153
research is warranted in which these boredom types are
1154
further explored with respect to potentially differential
1155
relations with antecedent variables (cf., Daschmann et al.
1156
2011), boredom-related coping strategies (e.g., Nett et al.
1157
2011), or their effects on critical outcome variables (e.g.,
1158
achievement; cf., Pekrun et al. 2010).
1159
In addition, an intriguing avenue for future research
1160
concerns the development of scales to better evaluate dif-
1161
ferential, in vivo experiences of boredom (state assessments
1162
of boredom types). For example, reactant boredom could be
1163
assessed as follows: ‘‘How strongly do you experience
1164
boredom at this moment’’ (from 1 [not at all]to5[very
1165
strongly]). Given a minimum level of boredom reported,
1166
reactant boredom could be assessed with items like ‘‘I feel
1167
restless at this moment’’, ‘‘I feel good at this moment’’
1168
[inverted], ‘‘I wish to leave this situation.’’ Finally, research
1169
evaluating various experiences of boredom with respect to
1170
different contexts (e.g., leisure time, achievement domain)
1171
may provide intriguing results concerning the prevalence of
1172
the boredom types in different settings.
1173
Our results may significantly contribute to previous
1174
research in the field of psychometric studies on boredom.
1175
They do not contradict previous studies that investigated
1176
different dimensions of boredom (e.g., disengagement,
1177
high arousal, low arousal, inattention, time perception;
1178
Fahlman et al. 2013). Rather, they suggest that those
1179
dimensions might be seen from the perspective of different
1180
types of boredom and not from a prototype perspective.
1181
Further, our findings do not contradict results of studies
1182
focusing on antecedents of boredom (e.g., internal vs.
1183
external stimulation, lack of meaning, monotony, being
1184
over- or under-challenged; Dahlen et al. 2004; Daschmann
1185
et al. 2011), but instead expand upon these studies by
1186
suggesting that careful examination of antecedents of
1187
boredom types could help identify possible reasons for why
1188
boredom types develop (e.g., dispositional factors) or are
1189
observed in specific situations (e.g., environmental factors).
1190
It is anticipated that future research in which boredom
1191
subtypes are considered will shed light on the ongoing
1192
discussion concerning the positive as well as negative
1193
effects of boredom on critical outcomes in achievement
1194
settings and everyday life (see Vodanovich 2003a). More
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
1195 specifically, given the present findings indicating that dif-
1196 ferent types of boredom may differentially correspond to
1197 criteria variables, it is likely that varied relations with other
1198 outcomes of interest (e.g., health, persistence, learning,
1199 creativity, decision making, etc.) will be observed. Further,
1200 considering that the prevalence of boredom types was
1201 found to differ depending on a situation in which boredom
1202 was experienced (achievement vs. non-achievement set-
1203 tings), future studies in which additional and more specific
1204 differentiations or manipulations of situation type are
1205 employed (e.g., sports as an achievement-oriented leisure
1206 setting, shopping for leisure vs. necessity, academic
1207 achievement vs. non-academic achievement) may also
1208 yield intriguing moderating effects on the prevalence as
1209 well as consequences of different boredom types.
1210 Finally, research efforts exploring the temporal devel-
1211 opment of the different types of boredom as experienced in
1212 achievement as well as non-achievement settings may also
1213 help to address unexplored hypotheses concerning transi-
1214 tions between boredom types over time. Similarly, future
1215 studies may also inform the development of intervention
1216 programs aimed at reducing individuals’ experiences of
1217 boredom, particularly in situations that afford fewer
1218 opportunities for changing or withdrawing from boring
1219 activities (e.g., the secondary school classroom, older
1220 adults in assisted care facilities).
1221 In closing, the present findings are consistent with the
1222 assertions of early boredom theorists in suggesting that
1223 ‘‘ …it is probable that the conditions and forms of behavior
1224 called ‘boredom’ are psychologically quite heteroge-
1225 neous…’’ (Fenichel 1951, p. 349; see also Fenichel 1934,
1226 p. 270 for the original German quote). To our knowledge,
1227 this study represents the first quantitative investigation into
1228 the internal and external validity of this assumption, and
1229 provides encouraging empirical support for a more differ-
1230 entiated perspective on how individuals experience bore-
1231 dom in everyday life. Moreover, whereas our results could
1232 help to shed light on a number of ongoing debates con-
1233 cerning the phenomenology, antecedents, and effects of
1234 this ubiquitous emotion, they also highlight the potential
1235 utility of exploring within-construct variability in other
1236 emotions along dimensions that may otherwise be assumed
1237 to represent the entire construct. With respect to enjoy-
1238 ment, for example, it is possible that a subtype of enjoy-
1239 ment characterized by relatively low levels of valence and
1240 arousal (‘‘quiet joy’’) may be differentiated from one high
1241 in both dimensions (‘‘excitement’’), with other types of
1242 enjoyment being observed that fall in between. Similarly,
1243 for anger, ‘‘silent anger’’ as compared to ‘‘rage’’ might be
1244 the extreme poles for this emotion. Given our results, it is
1245 possible that the presumed homogeneity of specific discrete
1246 emotions with respect to the dimensions of valence and
1247 arousal may be overestimated. Therefore, in contrast to
1248
assuming, evaluating, and/or classifying discrete emotional
1249
experiences with respect to mean levels of valence and
1250
arousal, the present research suggests that discrete emo-
1251
tions may be best explored by examining the existence,
1252
prevalence, development, antecedents, and consequences
1253
of within-emotion variability along these dimensions,
1254
thereby acknowledging the complex nature of individuals’
1255
discrete emotions that are experienced in real-life settings. 1256
1257
References
1258
Amos, A., Wiltshire, S., Haw, S., & McNeill, A. (2006). Ambivalence
1259
and uncertainty: Experiences of and attitudes towards addiction
1260
and smoking cessation in the mid-to-late teens. Health Education
1261
Research, 21(2), 181–191. doi:10.1093/her/cyh054.
1262
Anshel, M. H. (1991). A survey of elite athletes on the perceived
1263
causes of using banned drugs in sport. Journal of Sport Behavior,
1264
14(4), 283–310.
1265
Armstrong, S. L., Gleitman, L. R., & Gleitman, H. (1983). What some
1266
concepts might not be. Cognition, 13(3), 263–308. doi:10.1016/
1267
0010-0277(83)90012-4.
1268
Baird, B., Smallwood, J., Mrazek, M. D., Kam, J. W. Y., Franklin, M.
1269
S., & Schooler, J. W. (2012). Inspired by distraction: Mind
1270
wandering facilitates creative incubation. Psychological Science,
1271
23(10), 1117–1122. doi:10.1177/0956797612446024.
1272
Bearden, L. J., Spencer, W. A., & Moracco, J. C. (1989). A study of
1273
high school dropouts. School Counselor, 37(2), 113–120.
1274
Berlyne, D. E. (1960). Conflict, arousal, and curiosity. New York,
1275
NY: McGraw-Hill Book Company. doi:10.1037/11164-000.
1276
Clark, S. L., & Muthe
´n, B. O. (2009). Relating latent class analysis
1277
results to variables not included in the analysis. Available from
1278
http://www.statmodel.com/download/relatinglca.pdf.
1279
Clore, G. L. (1994). Why emotions require cognition. In P. Ekman &
1280
R. J. Davidson (Eds.), The nature of emotion (pp. 181–191). New
1281
York: Oxford University Press.
1282
Clore, G. L., & Ortony, A. (1991). What more is there to emotion
1283
concepts than prototypes? Journal of Personality and Social
1284
Psychology, 60(1), 48–50. doi:10.1037/0022-3514.60.1.48.
1285
Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality
1286
Inventory (NEO PI-R) and NEO Five Factor Inventory.Profes-
1287
sional Manual. Odessa, FL: Psychological Assessment Resources.
1288
Csikszentmihalyi, M., & Larson, R. (1987). Validity and reliability of
1289
the experience-sampling method. Journal of Nervous and Mental
1290
Disease, 175(9), 526–536. doi:10.1097/00005053-198709000-
1291
00004.
1292
Dahlen, E. R., Martin, R. C., Ragan, K., & Kuhlman, M. M. (2004).
1293
Boredom proneness in anger and aggression: Effects of impul-
1294
siveness and sensation seeking. Personality and Individual
1295
Differences, 37, 1615–1627.
1296
Daschmann, E. C., Goetz, T., & Stupnisky, R. H. (2011). Testing the
1297
predictors of boredom at school. Development and validation of
1298
the Precursors to Boredom Scales. British Journal of Educational
1299
Psychology, 81, 421–440. doi:10.1348/000709910X526038.
1300
Ekman, P. (1984). Expression and the nature of emotion. In K.
1301
S. Scherer & P. Ekman (Eds.), Approaches to emotion (pp.
1302
319–343). Hillsdale, NJ: Erlbaum.
1303
Fahlman, S. A., Mercer-Lynn, K. B., Flora, D. B., & Eastwood, J. D.
1304
(2013). Development and validation of the multidimensional
1305
state boredom scale. Assessment, 20(1), 68–85.
1306
Farmer, R., & Sundberg, N. D. (1986). Boredom proneness: The
1307
development and correlates of a new scale. Journal of Personality
1308
Assessment, 50(1), 4–17. doi:10.1207/s15327752jpa5001_2.
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
1309 Fenichel, O. O. (1934). Zur Psychologie der Langeweile. Imago
1310 (Leipzig), 20270-281.
1311 Fenichel, O. (1951). On the psychology of boredom. In D. Rapaport
1312 (Ed.), Organization and pathology of thought: Selected sources
1313 (pp. 349–361). New York, NY: Columbia University Press.
1314 doi:10.1037/10584-018.
1315 Fisher, C. D. (1993). Boredom at work: A neglected concept. Human
1316 Relations, 46(3), 395–417. doi:10.1177/001872679304600305.
1317 Goetz, T., & Frenzel, A. C. (2006). Pha
¨nomenologie schulischer
1318 Langeweile [Phenomenology of boredom at school]. Zeitschrift
1319 fu¨ r Entwicklungspsychologie und Pa¨ dagogische Psychologie,
1320 38(4), 149–153. doi:10.1026/0049-8637.38.4.149.
1321 Goetz, T., Frenzel, A. C., & Pekrun, R. (2007a). Regulation von
1322 Langeweile im Unterricht. Was Schu
¨lerinnen und Schu
¨ler bei
1323 der ‘‘Windstille der Seele’’ (nicht) tun [Regulation of boredom in
1324 class. What students do (not) when experiencing the, windless
1325 calm of the soul’]. Unterrichtswissenschaft, 35(4), 312–333.
1326 Goetz, T., Frenzel, A. C., Pekrun, R., & Hall, N. C. (2006). The domain
1327 specificity of academic emotional experiences. Journal of Exper-
1328 imental Education, 75(1), 5–29. doi:10.3200/JEXE.75.1.5-29.
1329 Goetz, T., Frenzel, A. C., Pekrun, R., Hall, N. C., & Lu
¨dtke, O.
1330 (2007b). Between- and within-domain relations of students’
1331 academic emotions. Journal of Educational Psychology, 99(4),
1332 715–733. doi:10.1037/0022-0663.99.4.715.
1333 Goetz, T., Frenzel, A. C., Stoeger, H., & Hall, N. C. (2010).
1334 Antecedents of everyday positive emotions: An experience
1335 sampling analysis. Motivation and Emotion, 34(1), 49–62.
1336 doi:10.1007/s11031-009-9152-2.
1337 Grubb, E. A. (1975). Assembly line boredom and individual
1338 differences in recreation participation. Journal of Leisure
1339 Research, 7, 256–269.
1340 Harris, M. B. (2000). Correlates and characteristics of boredom
1341 proneness and boredom. Journal of Applied Social Psychology,
1342 30(3), 576–598. doi:10.1111/j.1559-1816.2000.tb02497.x.
1343 Harrison, A. (1984). Creativity, class, and boredom: Cognitive
1344 models for intelligent activities. Journal of Education, Boston,
1345 166(2), 150–169.
1346 Hebb, D. O. (1955). Drives and the C. N. S. (conceptual nervous
1347 system). Psychological Review, 62(4), 243–254. doi:10.1037/
1348 h0041823.
1349 Hektner, J. M., Schmidt, J. A., & Csikszentmihalyi, M. (2007).
1350 Experience sampling method: Measuring the quality of everyday
1351 life. Thousand Oaks, CA: Sage Publications, Inc.
1352 Hill, A. B., & Perkins, R. E. (1985). Towards a model of boredom.
1353 British Journal of Psychology, 76, 235–240.
1354 Johnson-Laird, P. N., & Oatley, K. (1989). The language of emotions:
1355 An analysis of a semantic field. Cognition and Emotion, 3(2),
1356 81–123. doi:10.1080/02699938908408075.
1357 Johnstone, T., & Scherer, K. R. (2000). Vocal communication of
1358 emotion. In M. Lewis & J. M. Haviland-Jones (Eds.), Handbook
1359 of emotions (2nd ed., pp. 220–235). New York: Guilford Press.
1360 Klapp, O. (1986). Overload and boredom. New York: Greenwood
1361 Press.
1362 Kleinginna, P. R., & Kleinginna, A. M. (1981). A categorized list of
1363 emotion definitions, with suggestions for a consensual definition.
1364 Motivation and Emotion, 5, 345–379.
1365 Larsen, J. T., & McGraw, P. (2011). Further evidence for mixed
1366 emotions. Journal of Personality and Social Psychology, 100(6),
1367 1095–1110. doi:10.1037/a0021846.
1368 Larson, R. W., & Richards, M. H. (1991). Boredom in the middle
1369 school years: Blaming schools versus blaming students. Amer-
1370 ican Journal of Education, 99(4), 418–443. doi:10.1086/443992.
1371 Lee, T. W. (1986). Toward the development and validation of a
1372 measure of job boredom. Manhattan College Journal of
1373 Business, 15, 22–28.
1374
London, H., Schubert, D. S., & Washburn, D. (1972). Increase of
1375
autonomic arousal by boredom. Journal of Abnormal Psychol-
1376
ogy, 80(1), 29–36. doi:10.1037/h0033311.
1377
Mikulas, W. L., & Vodanovich, S. J. (1993). The essence of boredom.
1378
The Psychological Record, 43(1), 3–12.
1379
Miller, W. R., & Seligman, M. E. (1975). Depression and learned
1380
helplessness in man. Journal of Abnormal Psychology, 84(3),
1381
228–238. doi:10.1037/h0076720.
1382
Muthe
´n, B. O. (2001). Latent variable mixture modeling. In G.
1383
A. Marcoulides & R. E. Schumacker (Eds.), New developments
1384
and techniques in structural equation modeling (pp. 1–33).
1385
London: Lawrence Erlbaum Associates.
1386
Muthe
´n, B. O., & Muthe
´n, L. K. (2000). Integrating person-centered
1387
and variable-centered analysis: Growth mixture modeling with
1388
latent trajectory classes. Alcoholism: Clinical and Experimental
1389
Research, 24(6), 882–891. doi:10.1111/j.1530-0277.2000.
1390
tb02070.x.
1391
Muthe
´n, L. K., & Muthe
´n, B. O. (2007). Mplus user’s guide (5th ed.).
1392
Los Angeles, CA: Muthe
´n & Muthe
´n.
1393
Nett, U. E., Goetz, T., & Daniels, L. M. (2010). What to do when
1394
feeling bored? Students’ strategies for coping with boredom.
1395
Learning and Individual Differences, 20(6), 626–638. doi:10.
1396
1016/j.lindif.2010.09.004.
1397
Nett, U. E., Goetz, T., & Hall, N. C. (2011). Coping with boredom in
1398
school: An experience sampling perspective. Contemporary
1399
Educational Psychology, 36(1), 49–59. doi:10.1016/j.cedpsych.
1400
2010.10.003.
1401
Newberry, A. L., & Duncan, R. D. (2001). Roles of boredom and life
1402
goals in juvenile delinquency. Journal of Applied Social
1403
Psychology, 31(3), 527–541. doi:10.1111/j.1559-1816.2001.
1404
tb02054.x.
1405
Nylund, K. L., Asparouhov, T., & Muthe
´n, B. O. (2007a). Deciding
1406
on the number of classes in latent class analysis and growth
1407
mixture modeling: A Monte Carlo simulation study. Structural
1408
Equation Modeling, 14(4), 535–569.
1409
Nylund, K., Nishina, A., Bellmore, A., & Graham, S. (2007b).
1410
Subtypes, severity, and structural stability of peer victimization:
1411
What does latent class analysis say? Child Development, 78(6),
1412
1706–1722.
1413
Ortony, A., Clore, G. L., & Foss, M. A. (1987). The referential
1414
structure of the affective lexicon. Cognitive Science: A Multi-
1415
disciplinary Journal, 11(3), 341–364.
1416
Papaioannou, A., Marsh, H. W., & Theodorakis, Y. (2004). A
1417
multilevel approach to motivational climate in physical educa-
1418
tion and sport settings: An individual or a group level construct?
1419
Journal of Sport & Exercise Psychology, 26(1), 90–118.
1420
Pekrun, R., Elliot, A. J., & Maier, M. A. (2009). Achievement goals
1421
and achievement emotions: Testing a model of their joint
1422
relations with academic performance. Journal of Educational
1423
Psychology, 101(1), 115–135. doi:10.1037/a0013383.
1424
Pekrun, R., & Frese, M. (1992). Emotions in work and achievement.
1425
In C. L. Cooper & I. T. Robertson (Eds.), International review of
1426
industrial and organizational psychology (Vol. 7, pp. 153–200).
1427
Chichester: Wiley.
1428
Pekrun, R., Goetz, T., Daniels, L. M., Stupnisky, R. H., & Perry, R. P.
1429
(2010). Boredom in achievement settings: Exploring control-
1430
value antecedents and performance outcomes of a neglected
1431
emotion. Journal of Educational Psychology, 102(3), 531–549.
1432
doi:10.1037/a0019243.
1433
Pekrun, R., Goetz, T., Titz, W., & Perry, R. P. (2002). Academic
1434
emotions in students’ self-regulated learning and achievement: A
1435
program of qualitative and quantitative research. Educational
1436
Psychologist, 37(2), 91–106. doi:10.1207/S15326985EP3702_4.
1437
Perkins, R. E., & Hill, A. B. (1985). Cognitive and affective aspects
1438
of boredom. British Journal of Psychology, 76(2), 221–234.
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
UNCORRECTED PROOF
1439 Phillips, A. (1993). On kissing, tickling, and being bored: Psycho-
1440 analytic essays on the unexamined life. Cambridge, MA:
1441 Harvard University Press.
1442 Reisenzein, R. (1994). Pleasure-arousal theory and the intensity of
1443 emotions. Journal of Personality and Social Psychology, 67,
1444 525–539.
1445 Robinson, M. D., & Barrett, L. F. (2010). Belief and feeling in self-
1446 reports of emotion: Evidence for semantic infusion based on
1447 self-esteem. Self and Identity, 9, 87–111. doi:10.1080/
1448 15298860902728274.
1449 Robinson, M. D., & Clore, G. L. (2002). Belief and feeling: Evidence
1450 for an accessibility model of emotional self-report. Psycholog-
1451 ical Bulletin, 128(6), 934–960. doi:10.1037/0033-2909.128.6.
1452 934.
1453 Rosch, E. (1978). Principles of categorization. In E. Rosch & B.
1454 B. Lloyd (Eds.), Cognition and categorization (pp. 27–48).
1455 Hillsdale, NJ: Erlbaum.
1456 Roseman, I. J., Antoniou, A., & Jose, P. E. (1996). Appraisal
1457 determinants of emotions: Constructing a more accurate and
1458 comprehensive theory. Cognition and Emotion, 10(3), 241–277.
1459 doi:10.1080/026999396380240.
1460 Rupp, D. E., & Vodanovich, S. J. (1997). The role of boredom
1461 proneness in self-reported anger and aggression. Journal of
1462 Social Behavior & Personality, 12(4), 925–936.
1463 Russel, J. A. (1991). In defense of a prototype approach to emotion
1464 concepts. Journal of Personality and Social Psychology, 60(1),
1465 37–47.
1466 Russell, J. A. (1980). A circumplex model of affect. Journal of
1467 Personality and Social Psychology, 39(6), 1161–1178. doi:10.
1468 1037/h0077714.
1469 Scherer, K. R. (1984). On the nature and function of emotion: A
1470 component process approach. In K. R. Scherer & P. Ekman
1471 (Eds.), Approaches to emotion (pp. 293–317). Hillsdale: Law-
1472 rence Erlbaum.
1473 Scherer, K. R. (2000). Emotions as episodes of subsystems synchro-
1474 nization driven by nonlinear appraisal processes. In M. D. Lewis
1475 & I. Granic (Eds.), Emotion, development, and self-organization
1476 (pp. 70–99). Cambridge: Cambridge University Press.
1477 Schwarz, G. (1978). Estimating the dimension of a model. The Annals
1478 of Statistics, 6(2), 461–464. doi:10.1214/aos/1176344136.
1479 Seib, H. M., & Vodanovich, S. J. (1998). Boredom proneness and
1480 psychosocial development. The Journal of Psychology, 132,
1481 642–652.
1482 Shaver, P., Schwartz, J., Kirson, D., & O’Connor, C. (1987). Emotion
1483 knowledge: Further exploration of a prototype approach. Journal
1484 of Personality and Social Psychology, 52(6), 1061–1086.
1485
Sio, U. N., & Ormerod, T. C. (2009). Does incubation enhance
1486
problem solving? A meta-analytic review. Psychological Bulle-
1487
tin, 135, 94–120.
1488
Smith, R. P. (1981). Boredom: A review. Human Factors, 23(3),
1489
329–340.
1490
Sommer, B. (1985). What’s different about truants? A comparison
1491
study of eighth-graders. Journal of Youth and Adolescence,
1492
14(5), 411–422. doi:10.1007/BF02138836.
1493
Sommers, J., & Vodanovich, S. J. (2000). Boredom proneness: Its
1494
relationship to psychological- and physical-health symptoms.
1495
Journal of Clinical Psychology, 56(1), 149–155. doi:10.1002/
1496
(SICI)1097-4679(200001)56:1\149:AID-JCLP14[3.0.CO;2-Y.
1497
Spacks, P. M. (1995). Boredom: The literary history of a state of
1498
mind. Chicago, IL: University of Chicago Press.
1499
Thackray, R. I. (1981). The stress of boredom and monotony: A
1500
consideration of the evidence. Psychosomatic Medicine, 43(2),
1501
165–176.
1502
Titz, W. (2001). Emotionen von Studierenden in Lernsituationen.
1503
Explorative Analysen und Entwicklung von Selbstberichtskalen
1504
[Students’ emotions in learning situations]. Mu
¨nster: Waxmann.
1505
Van Tilburg, W. A. P., & Igou, E. R. (2012). On boredom: Lack of
1506
challenge and meaning as distinct boredom experiences. Moti-
1507
vation and Emotion, 36, 181–194.
1508
Vodanovich, S. J. (2003a). On the possible benefits of boredom: A
1509
neglected area in personality research. Psychology and Educa-
1510
tion: An Interdisciplinary Journal, 40(3–4), 28–33.
1511
Vodanovich, S. J. (2003b). Psychometric measures of boredom: A
1512
review of the literature. Journal of Psychology: Interdisciplinary
1513
and Applied, 137(6), 569–595. doi:10.1080/00223980309600636.
1514
Wallbott, H. G. (1998). Bodily expression of emotion. European
1515
Journal of Social Psychology, 28, 879–896.
1516
Wang, C. P., Brown, C. H., & Bandeen-Roche, K. (2005). Residual
1517
diagnostics for growth mixture models: Examining the impact of
1518
preventive intervention on multiple trajectories of aggressive
1519
behavior. Journal of the American Statistical Association,
1520
100(3), 1054–1076. doi:10.1198/016214505000000501.
1521
Wasson, A. S. (1981). Susceptibility to boredom and deviant behavior
1522
at school. Psychological Reports, 48(3), 901–902.
1523
Watson, D., & Tellegen, A. (1985). Toward a consensual structure of
1524
mood. Psychological Bulletin, 98(2), 219–235. doi:10.1037/
1525
0033-2909.98.2.219.
1526
Zeidner, M. (1998). Test anxiety: The state of the art. New York, NY:
1527
Plenum Press.
1528
Zeidner, M. (2007). Test anxiety: Conceptions, findings, conclusions.
1529
In P. A. Schutz & R. Pekrun (Eds.), Emotion in education (pp.
1530
165–184). San Diego: Academic Press.
1531
Motiv Emot
123
Journal : Large 11031 Dispatch : 8-10-2013 Pages : 19
Article No. : 9385 hLE hTYPESET
MS Code : MOEM1118 hCP hDISK
44
Author Proof
