Decibels, Disposition, and Duration: the Impact of Musical Loudness and Internal States on Time Perceptions

ABSTRACT - This study examines the influence of a psychophysical stimulus property of music (loudness) on temporal perceptions, as well as conditions and processes that may govern its effects. Musical volume was manipulated and internal states (affect and arousal) measured in a between-subjects experiment involving 54 female college students. The dependent variables were two dimensions of temporal perception: subjective duration and pace. Results indicate that loudness of music influences both aspects of time perception, that affect moderates the influence of loudness on perceived duration, and that arousal partially mediates the influence of loudness on perceptions of pace. The duration of a time interval seemed shorter to subjects exposed to soft (versus loud) music. This effect was more pronounced among subjects in neutral (versus positive) affective states.

Citation:

James J. Kellaris, Susan Powell Mantel, and Moses B. Altsech (1996) ,"Decibels, Disposition, and Duration: the Impact of Musical Loudness and Internal States on Time Perceptions", in NA - Advances in Consumer Research Volume 23, eds. Kim P. Corfman and John G. Lynch Jr., Provo, UT : Association for Consumer Research, Pages: 498-503.

Advances in Consumer Research Volume 23, 1996      Pages 498-503

DECIBELS, DISPOSITION, AND DURATION: THE IMPACT OF MUSICAL LOUDNESS AND INTERNAL STATES ON TIME PERCEPTIONS

James J. Kellaris, University of Cincinnati

Susan Powell Mantel, University of Toledo

Moses B. Altsech, Penn State University

ABSTRACT -

This study examines the influence of a psychophysical stimulus property of music (loudness) on temporal perceptions, as well as conditions and processes that may govern its effects. Musical volume was manipulated and internal states (affect and arousal) measured in a between-subjects experiment involving 54 female college students. The dependent variables were two dimensions of temporal perception: subjective duration and pace. Results indicate that loudness of music influences both aspects of time perception, that affect moderates the influence of loudness on perceived duration, and that arousal partially mediates the influence of loudness on perceptions of pace. The duration of a time interval seemed shorter to subjects exposed to soft (versus loud) music. This effect was more pronounced among subjects in neutral (versus positive) affective states.

This study seeks to advance psychological time research by exploring the influence of a common environmental stimulus (music) and measured internal states on perceptions of event duration and pace. Music is frequently encountered in commercial environments. It not only shapes the environment, but can influence internal states (Bruner 1990). Thus, music may offer a means by which marketers can influence consumers' temporal perceptions to achieve commercial benefits.

This study examines two key dimensions of temporal perception: duration and pace. Perceived duration refers to how long a time interval seems to last (or, retrospectively, to have lasted). Perceived pace refers to how rapidly the succession of events within a time interval seems to take place. Pace bears a close relation to two other psychophysical variables: informational density and perceived activity (Berlyne 1974). Given a time interval of fixed duration, the sequence of stimulus events (changes) that take place during the interval will seem to happen at a rate which may differ from objective pace.

Perceived duration and perceived pace are conceptually distinct aspects of temporal perception, yet each may play a role in determining the other. Pace perceptions may serve as inferential cues on which time estimates can be based. If changes in the stimulus environment seem to take place in rapid succession, then more stimulus information was encountered. Given that it generally takes longer for more (versus fewer) events to occur, the duration of a time interval filled with more varied stimulation should, in retrospect, seem longer. Conversely, duration estimates may be used to draw inferences about rate of information, i.e., pace. Thus duration estimates and pace perceptions should be positively correlated.

The Influence of Music

Several recent studies have examined effects of music on time perceptions. Kellaris and Kent (1992) found duration estimates to vary as a function of musical modality (key) in a lab study. Keys that facilitated storage and retrieval of information encountered during a time interval produced longer duration estimates than keys that inhibited storage and retrieval. Kellaris and Altsech (1992) found louder (versus softer) music to be judged as longer in duration by female listeners. The authors attributed this finding to the amount of sensory information to which listeners were exposed and to the greater hearing sensitivities of females. Apparently less (more) seemed to happen during a time interval filled with soft (loud) music. Kellaris and Mantel (1994) found the duration estimates of female subjects to vary as a function of their affective states. Female subjects tended to underestimate lapsed time to a greater extent when they were in less (versus more) positive moods. This finding is consistent with the suggestion that females may tend to store and/or retrieve more information about positively valenced events. In a simulation study of waiting in bank teller lines Chebat et al. (1993) found that the effect of visual stimulation on perceived waiting time depended on the tempo of background music. The "wait" seemed longer to subjects exposed to high (versus low) amounts of visual information when slow music was played, and shorter when fast music was played. Yalch and Spangenberg (1990) found foreground versus background music to influence perceived shopping time in a field study. Younger shoppers reported they had spent more time shopping when exposed to background (softer) music, whereas older shoppers reported longer shopping durations when exposed to foreground (louder) music. Clearly, music can play a role in shaping time perceptions.

The present study focuses on the loudness dimension of music. Because loudness refers to an objective, tangible attribute of sound, it may be considered a psychophysical property (Berlyne 1974). Examination of this property is warranted because it is universally characteristic of auditory stimuli and is easily controlled by marketers in applied settings. Moreover, findings pertaining to musical loudness may generalize to other psychophysical variables.

Loudness is expected to contribute positively to both duration and pace perceptions, because loud (versus soft) music confronts sensory receptors with more salient stimulus information (Kellaris and Altsech 1992). Louder, more salient music should evoke higher levels of attention, processing, and recall of the stimulus event (Dowling and Harwood 1986). As more (versus less) stimulus information is encountered and subsequently recalled from a time interval, listeners should attribute longer durations to the interval (Block 1990; Fraisse 1984; Ornstein 1969).

Loudness should have a similar positive effect on perceived pace because the intensity of a psychophysical stimulus should contribute to arousal and perceived activity (Berlyne 1974). The positive association between arousal and perceived activity may create an illusion of faster pace under conditions of more intense stimulation. The loudness of sounds may also serve as an inferential cue on which judgments of pace can be based. As more (less) stimulus information is encountered and remembered from a time interval, listeners may infer that the succession of events or changes that took place during the interval occurred more quickly (slowly).

The Role of Internal States

Internal states should shape the impact of environmental stimuli on the subjective experience of time (Block 1990). Specifically, when the listener is in a positive (versus neutral) affective state, the effect of musical loudness should be less pronounced. This is expected because when the listener's affect is elevated (positive affect), both cognitive organization and motivation to process are enhanced (Isen 1993). Listeners exposed to soft music while in positive (versus neutral) affective states should therefore process and later recall hearing more auditory information during the time period. As a result, their memory-based time estimates are augmented and should more closely resemble those of listeners exposed to more salient information (louder music).

By contrast, time perceptions of listeners in neutral (versus elevated) affective states should be more prone to effects of musical loudness. Whereas less (versus more) salient information is less (versus more) likely to be attended to and later remembered, memory-based time estimates should be shorter among listeners exposed to soft music and longer among listeners exposed to loud music when affective states do not alter processing strategies.

The influence of musical loudness on temporal perceptions may stem (partly) from the arousal it induces. Arousal has been correlated positively with perceived activity (Holbrook and Anand 1990). Hence arousal is expected to mediate the effect of musical loudness on temporal perceptions, particularly perceptions of pace.

Given that studies have reported effects of positive affect that differ from those resulting from arousal (Isen 1993), affect and arousal are expected to operate independently. In this preliminary study, we attempted to capture the influence of subtle variations in affect and arousal by measuring (rather than manipulating) naturally occurring levels of these internal states. In contrast to past studies, we measure internal states after exposure to the stimulus music. This way, the measures in no way interfere with subjects' duration/pace responses.

METHOD

A factorial between-subjects experiment was conducted to investigate the influence of loudness of music (soft=60dB, loud=90dB), affect (neutral versus positive), and arousal (covariate term) on temporal perceptions. The stimuli consisted of musical excerpts played at varied levels of loudness. Affect and arousal states were measured using multi-item scales. Affect groups were created via median split. Dependent variables were retrospective estimates of the stimulus music's duration and perceived pace. The procedure involved exposing subjects to music over a loudspeaker in small groups, and having them complete a brief questionnaire.

Subjects

Fifty-four (N=54) females were recruited from a subject pool at a midwestern university. Subjects were offered course credit in exchange for participation. Ages ranged from 20 to 27 years, with a median age of 21. Previous research has shown females (versus males) to be generally more sensitive to subjective distortions of time (e.g., Krishnan and Saxena 1984). Hence, a female sample should be more efficient for establishing internal validity. It also avoids potential confounds due to gender differences in responses to music (Lacher 1994).

Stimuli

The stimuli were three-minute excerpts of an original, pop style recording of instrumental music. Whereas the style of the music was generally familiar to our subjects, the specific composition was not. Original music was used to avoid confounding stimulus materials with extraneous associations due to prior exposure.

Loudness was manipulated by adjusting volume controls on the amplifier that controlled the loudspeakers in the listening facility. An average sound level of 60dB was produced under the soft condition and 90dB for the loud condition. These levels were chosen to simulate background and foreground listening conditions respectively (Yalch and Spangenberg 1990). Sound levels were checked using a decibel meter placed at room center, at a distance of five feet from the overhead loudspeaker in each listening booth.

Procedure

Subjects were processed in small groups in a behavioral lab. After assignment to treatment groups, each group was directed to one of the sound-proof listening booths adjacent to the main room of the lab. Stimulus music was piped into each booth via identical equipment from a control room. Subjects were instructed to listen to the music without talking and to complete a self-administered questionnaire after the music stopped playing. Unobtrusive video monitoring during the procedure revealed no violations of the instructions.

Measures

The dependent measures included an open-ended, retrospective duration estimate item and a seven-point perceived pace scale. It is important to note that subjects anticipated answering opinion questions about the music, but could not anticipate the duration estimation task or perceived pace measure. Printed instructions told subjects to provide their best estimate of how long the music had played. A prompt ("I estimate the music I heard lasted for about:") preceded blank spaces labeled "minutes" and "seconds." Such open-ended measures are standard in psychological time research (Block 1990; Fraisse 1984).

The perceived pace measure was a single-item, seven-point semantic differential scale with endpoints labeled "slow(1)" and "fast(7)," preceded by the prompt "The music I heard was." A simple, single-item scale was preferred over more elaborate alternatives because a more complex measure would have imposed greater cognitive demands on subjects. A more demanding task could interfere with the feeling states yet to be measured.

Affect was measured via five-point semantic differential items that were summed and averaged to form a composite scale. The individual items were "good(5)/bad(1), pleasant(5)/unpleasant(1), happy(5)/sad(1), positive(5)/ negative(1)." The alpha reliability of the composite scale is .88. To facilitate testing the moderation hypothesis with variance analysis (ANOVA), low and high affect groups were formed via median split (M=3.8). The "low" group (X=3.2), it should be noted, actually represents neutral affect rather than negative or dysphoric mood. The "high" group (X=4.4) represents what Isen would label "mildly positive affect," rather than highly euphoric mood. Because affect was measured rather than induced, the range represents the variation that occurs naturally in daily experience. Affect is expected to be independent of musical loudness because affective response to music stems principally from other dimensions, such as pitch or mode (Bruner 1990; Dowling and Harwood 1986).

Arousal was measured on a multi-item adjective checklist adapted from multiple sources (e.g., Hevner 1935; Kellaris and Rice 1993; Nowlis 1965). Printed instructions asked subjects to "Please circle the words that best describe the music," with supplementary instructions to circle as many or as few as apply. Responses were scored as 1 for circled items and 0 for items not circled. A composite arousal score was formed by summing the following items: "arousing, stimulating, boring(*), exciting, vigorous, calm(*)." Items marked (*) were reverse scored. The alpha reliability is .64 for the summed scale. The checklist technique was used to minimize the impact of the measurement process on the responses and to allow internal states to be recorded quickly and easily before they dissipated.

TABLE 1

OVERVIEW OF VARIANCE ANALYSES

A confirmatory factor analysis (CFA) was performed using LISREL VI (Version 6.6) to further assess the validity of the affect and arousal measures. The correlation matrix constructed from the raw data did not differ statistically from the correlation matrix reproduced by the CFA, c²(34)=32.2, p>.50, thereby indicating that the measurement model fits the data. Consistent with our expectations, affect and arousal are not significantly correlated (Pearson's r=.14).

A seven-point loudness scale (1=soft; 7=loud), preceded by the prompt "The music I heard was..." was included to facilitate a check on the manipulation. A final item asked subjects to record their age.

RESULTS

Independent Variable and Confounding Checks

As a check on the integrity of the independent variables, two variance analyses (ANOVAs) were performed. The first analysis examined the impact of musical loudness and affect group on the soft-loud scale described above. As expected, the loudness treatment produced a significant main effect, F(1,54)=74.34, p<.001, with means in the anticipated direction (X_soft=3.5, X_loud=5.7). There were no significant main or interactive effects of low (neutral) versus high (positive) affect group on the loudness check measure. The second analysis examined the impact of musical loudness and affect group on the four-item affect scale. As expected, a significant difference between low (X=3.2) and high (X=4.4) affect groups was found, F(1,54)=119.67, p<.001, with no main or interactive effect of loudness on affective state. Thus, each independent variable produced a significant main effect on its corresponding measure, with no unanticipated confounds.

MANOVA

The data support the expectation that perceptions of duration and pace should be intercorrelated (Pearson's r=.28, p<.03). Because the dependent variables are intercorrelated, a multivariate variance analysis (MANOVA) was performed to avoid Type I error inflation. Results appear in Table 1.

Consistent with expectations, the MANOVA revealed a significant main effect of musical loudness on the combined dependent variables representing temporal perception, Wilks' D=.72, F(2,49)=9.65, p<.001, and a significant loudness by affect interaction, Wilks' D=.86, F(2,49)=4.08, p<.03.

Effects of Musical Loudness

Given the positive MANOVA findings, individual variance analyses (ANOVAs) were performed on each dependent variable. Results appear in Table 1, descriptive data in Table 2.

The analyses show significant, positive main effects of musical loudness on both perceived duration, F(1,53)=12.34, p<.001, and perceived pace, F(1,53)=6.79, p<.012, with effect magnitudes of w²=.16 and .10 (Keppel 1982) for perceived duration and pace, respectively. Mean duration estimates were shorter (134.3 sec.) under the soft music condition and longer (194.8 sec.) under the loud music condition. Likewise, the pace of the stimulus event seemed slower (X=5.2) under the soft music condition and faster (X=5.8) under the loud music condition.

The Moderating Role of Affect

A significant loudness by affect interaction was found, F(1,54)=8.32, p<.01, supporting our expectation that affect moderates the influence of loudness on perceived duration. The estimated magnitude (w²) of this effect is .10. The shape of the interaction is illustrated in Figure 1.

The positive effect of musical loudness on perceived duration is more pronounced when subjects are in less positive (i.e., neutral) affective states. Under the low affect condition, duration estimates differ significantly between soft (113.5 sec.) and loud (228.7 sec.) music conditions, t=3.57, p<.002. No statistical differences between the duration estimates of soft (169.4 sec.) and loud (151.0 sec.) music groups are observed under the high (positive) affect condition, t=.95, p>.10. The duration estimates of low versus high affect group differ statistically under the loud music condition, t=2.09, p<.05, but not under the soft music condition, t=-1.74, p=.09 (two-tailed test). A similar pattern is seen in the group means for pace; however, the interaction is not significant, F<1.

The Mediating Role of Arousal

If loudness operates through arousal to influence time perception, several conditions will be observed (Baron and Kenny 1986). First, the independent variable (loudness of music) must produce an effect on the dependent variable(s), i.e., duration and pace, a condition already established above. Second, the independent variable (loudness) must produce an effect on the hypothesized mediator (arousal). To establish this relationship, a variance analysis was performed with loudness as the independent variable and arousal as the dependent variable. Arousal was statistically related to loudness, F(1,54)=4.43, p<.04, with loud music predictably producing greater arousal (X=4.3) than soft music (X=3.4).

TABLE 2

MEANS AND STANDARD DEVIATIONS FOR EXPERIMENT

FIGURE 1

INTERACTIVE EFFECT OF LOUDNESS OF MUSIC AND AFFECT ON PERCEIVED DURATION OF A TIME INTERVAL

Third, the hypothesized mediator (arousal) must be related to the dependent variable(s). Arousal was found to be statistically related to perceived pace, r=.23, p<.05, but not to perceived duration, r=.07, n.s. Arousal may thus be a mediator of music's effect on perceived pace; but, having failed to meet the third condition, must be disqualified as a potential mediator of music's effect on perceived duration.

The final condition for mediation is that when variation due to the hypothesized mediator (arousal) is removed from the dependent variable (perceived pace), the statistical relationship between the independent variable (loudness of music) and the dependent variable should become non-significant, or at least significantly weaker. An ANCOVA procedure, with perceived pace as the dependent variable, loudness as the independent variable, and arousal as a covariate term, was used to test for this condition. The impact of loudness on perceived pace became slightly weaker (F=5.1 and w²= .07 with the covariate, versus F=6.8 and W²=.10 without the covariate), but the effect was still statistically significant, p<.03. We interpret this result as evidence of partial mediation.

DISCUSSION & CONCLUSION

This study has examined the influence of music on aspects of temporal perception among young adult females. The loudness of music was found to contribute positively to both retrospective duration estimates and perceived pace of a stimulus event. Affect was found to moderate the influence of loudness on duration estimates, such that louder (softer) music made a stimulus event seem longer (shorter) to subjects in neutral affective states. Conversely, the time estimates of those subjects in mildly elevated affective states do not appear to be influenced by musical volume. Evidence also suggests that arousal may partially mediate the influence of musical loudness on perceived pace. The positive effect of loudness on pace was somewhat less pronounced when the contribution of arousal was taken into account.

The interactive effect of loudness and affect may provide additional insight into previously reported findings. Yalch and Spangenberg (1990) reported a moderating effect of age of shopper on the influence of background versus foreground music (roughly analogous to soft versus loud music, although different styles of music were used across conditions) on perceived shopping time in a field experiment. Whereas younger (versus older) shoppers reported slightly more positive moods across both conditions, the pattern of the interaction is consistent with the present findings. The internal state (mood) of the shoppers, rather than their age (which was incidentally confounded with positive mood), may have been the actual contingency underlying the observed effect of music.

Concerning our mediation hypothesis, the positive effect of loudness on pace was somewhat less pronounced when the contribution of arousal was taken into account; however, given the relatively small difference measured arousal seems to make, some other process(es) may also intervene between loudness and pace. Future research should explore other potential mediators, particularly cognitive process variables such as attention, storage, and retrieval.

As with any study, certain features of the design (e.g., single sex sample, retrospective paradigm, single item measures) impose limitations to generality, each of which suggests an opportunity for further research. Although the measurement of internal states (affect and arousal) in the current study allows for the analysis of naturally occurring fluctuations, our procedure provides evidence of association, but not causality. In addition to manipulating internal states, future research should use both forced and passive exposures to environmental stimuli, explore effects of other traits of music and other environmental stimuli, and, of course, replicate and extend the present study using other samples of listeners.

Although the present findings must be considered preliminary, they suggest the intriguing possibility of influencing consumers' time perceptions in commercial settings by manipulating characteristics of the stimulus environment (such as background music). Perhaps sellers and service providers could effect greater customer satisfaction by diminishing the perceived duration of waiting times, by augmenting the perceived duration of rendering individual service, by slowing or increasing the perceived pace of service encounters as appropriate to the situation, etc. For example, the present findings suggest that in situations where consumers are likely to be in neutral affective states, subjective time can be diminished by playing soft music and augmented by playing loud music. In situations where consumers are likely to be in elevated moods, it may not be possible to manage time perceptions by manipulating the loudness of background music. Future research should continue to explore relationships between external stimuli, internal processes, and time-perceptual outcomes, and well as consumptive consequences of time perception.

REFERENCES

Baron, Reuben M. and David A. Kenny. (1986). "The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations," Journal of Personality and Social Psychology 51, 1173-1182.

Berlyne, D. E. (1974). Studies in the New Experimental Aesthetics: Steps Toward an Objective Psychology of Aesthetic Appreciation. Washington, D.C.: Hemisphere Publishing.

Block, Richard A. (ed.). (1990). Cognitive Models of Psychological Time. Hillsdale, NJ: Erlbaum.

Bruner, Gordon C., II (1990). "Music, Mood, and Marketing," Journal of Marketing 54, 94-104.

Chebat, Jean-Charles, Claire Gelinas-Chebat, and Pierre Filiatrault (1993), "Interactive Effects of Musical and Visual Cues on Time Perception: An Application to Waiting Lines in Banks," Perceptual and Motor Skills, 77, 995-1020.

Dowling, W. Jay and Dane L. Harwood. (1986). Music Cognition. San Diego, CA: Academic Press.

Fraisse, Paul. (1984). "Perception and Estimation of Time," Annual Review of Psychology 35, 1-36.

Hevner, Kate (1935), "Expression in Music: A Discussion of Experimental Studies and Theories," Psychological Review 42, 186-204.

Holbrook, Morris B. and Punam Anand (1990), "Effects of Tempo and Situational Arousal on the Listener's Perceptual and Affective Responses to Music," Psychology of Music 18, 150-162.

Isen, Alice M. (1993). "Positive Affect and Decision Making." In M. Lewis & J. Haviland (eds.), Handbook of Emotion. New York: Guilford, 261-275.

Kellaris, James J. and Moses B. Altsech (1992), "The Experience of Time as a Function of Musical Loudness and Gender of Listener," in J. Sherry and B. Sternthal, eds., Advances in Consumer Research, Vol. 19, Provo, UT: Association for Consumer Research, 725-729.

Kellaris, James J. and Robert J. Kent. (1992). "The Influence of Music on Consumers' Temporal Perceptions: Does Time Fly When You're Having Fun?," Journal of Consumer Psychology 1, 365-376.

Kellaris, James J. and Susan Powell Mantel (1994), "The Influence of Mood and Gender on Consumers' Time Perceptions," in C. T. Allen and D. Roedder John, eds., Advances in Consumer Research, Vol. 21, Provo, UT: Association for Consumer Research, 514-518.

Kellaris, James J. and Ronald C. Rice (1993), "The Influence of Tempo, Loudness, and Gender of Listener on Responses to Music," Psychology and Marketing 10, 15-29.

Krishnan, Lila and N. K. Saxena. (1984). "Perceived Time: Its Relationship with Locus of Control, Filled versus Unfilled Time Intervals, and Perceiver's Sex," Journal of General Psychology 110, 275-281.

Lacher, Kathleen T. (1994), "An Investigation of the Influence of Gender on the Hedonic Responses Created by Listening to Music," in C. T. Allen and D. Roedder John, eds, Advances in Consumer Research, Vol. 21, Provo, UT: Association for Consumer Research, 354-358.

Nowlis, V. (1965), "Research with the Mood Adjective Checklist." In S. S. Tomkins & C. E. Izard, Eds., Affect, Cognition and Personality (pp. 352-389). New York: Springer.

Ornstein, Robert E. (1969). On the Experience of Time. New York: Penguin.

Yalch, Richard F. and Eric Spangenberg. (1990). "Effects of Store Music on Shopping Behavior," Journal of Consumer Marketing 7, 55-63.

----------------------------------------