Children's Relationships Between Repetition and Affect

Theo B.C. Poiesz, Tilburg University
ABSTRACT - In this paper an attempt will be made to gain a better understanding of children's affective reactions to brandname repetition. For this, reference will be made to differences in cognitive-development. These differences are presumed to affect the conditions that are deemed necessary for positive repetition-affect relationships to take place. One of these conditions is specified by the functional exposure hypothesis. Empirical evidence is presented suggesting that age does influence the type of relationship between repetition and affect.
[ to cite ]:
Theo B.C. Poiesz (1986) ,"Children's Relationships Between Repetition and Affect", in NA - Advances in Consumer Research Volume 13, eds. Richard J. Lutz, Provo, UT : Association for Consumer Research, Pages: 633-636.

Advances in Consumer Research Volume 13, 1986      Pages 633-636


Theo B.C. Poiesz, Tilburg University

[The author wants to thank Jan van Gerwen for his valuable contributions.]


In this paper an attempt will be made to gain a better understanding of children's affective reactions to brandname repetition. For this, reference will be made to differences in cognitive-development. These differences are presumed to affect the conditions that are deemed necessary for positive repetition-affect relationships to take place. One of these conditions is specified by the functional exposure hypothesis. Empirical evidence is presented suggesting that age does influence the type of relationship between repetition and affect.


Over the years, repetition-affect phenomena have received a great deal of attention in the psychological and consumer behavior literature. For a recent and comprehensive review on these so-called 'mere exposure phenomena' see Obermiller (1985). Empirical results that were reported in the literature were obtained almost exclusively with adult subjects. It seems justified, however, not to limit research to adult subjects only, as one may assume that also children are frequently and repeatedly exposed to marketing stimuli. For example, when considering the case of brandnames as one type of these stimuli, Atkin and Heald (1977) calculated brandname exposure to be an average of 3.65 times per children's television commercial. In order to get an impression of the total number of children's brandname exposure opportunities, this figure would have to be multiplied by the number of commercials over time and supplemented with brandname exposure through other media. For adults, in (dated) literature some estimates of the number of exposures to marketing stimuli are presented (e.g. Advertising Age, 1970; Bauer and Greyser 1968, Ebel 1957), but for children no such estimates are known to be available. Anyway, apart from the practical significance and methodological problems of such figures, we may safely assume that the number of children's brandname exposures is impressive.

The few studies that were reported on repetition-affect relationships of non-adult subjects produced conflicting results (e.g. Cantor 1968, Heingartner and Hall 1974). Reviewing the available evidence, Rossiter (1980) concludes that a persuasive effect of brandname repetition within a children's commercial is unlikely (p.177). However, one may argue that in its effect upon behavior, the issue of repetition within commercials is less critical than the number of repetitions over commercials. The present paper will address the latter issue, thus not confining itself to the effect of the number of repetitions per advertising message.

The 'mere exposure' hypothesis (Zajonc 1968) states that the mere repeated exposure of an individual to a stimulus object enhances his/her attitude toward it. By 'mere exposure' is meant: a condition which just makes the stimulus accessible to the individual's perception. To put it briefly, empirical studies on exposure effects produced conflicting results. Also, the mere exposure hypothesis directly contradicts the existence of a different phenomenon well established in the literature: the preference for novel as opposed to familiar stimuli (e.g. Berlyne 1980, 1970). Recently, the 'functional exposure' hypothesis was introduced (Poiesz 1983, and elsewhere these Proceedings) which attempts to provide a consistent explanation for the seemingly contradicting empirical results. This hypothesis states that positive repetition affect relationships will occur only and to the extent that repetition, or its psychological counterpart: familiarity, may be considered psychologically functional.

The function of familiarity is to be understood as its possibility to help a person reach a desired psychological state. In an ambiguous experimental situation, for example, this would be the possibility to reduce subjective uncertainty - but only to the extent that uncertainty is associated with apprehension as an undesired psychological state. Whether uncertainty will be associated with apprehension can only be ascertained by the simultaneous consideration of the person, the object(s) and the situation involved. If uncertainly/apprehension is reduced by the more familiar stimuli, these stimuli will be evaluated more positively. As another example take an ambiguous laboratory consumer choice situation in which Ss are requested to make a selection out of a number of alternative brands. Also here, the function of the familiarity of brandnames is dependent upon the specific combination of subject, (choice-)objects and situation. If Ss perceive little risk in the choice situation, familiar stimuli can not be functional by reducing that risk. Therefore, the more familiar stimuli will not be evaluated more positively. If perceived risk increases, so does the function of familiarity. However, if perceived risk is high, also the need for choice-relevant information is high. As mere brandname familiarity itself is not choice relevant, its functionality will be low under conditions of high perceived risk. Therefore, in a functional exposure interpretation it is expected that there will be a preference for familiar brandnames only in a condition of intermediate perceived risk and not/less under conditions of low and high perceived risk. The functional exposure hypothesis differs from other uncertainty reduction hypotheses by its emphasis on the function of familiarity as determined by the interaction of person, object and situational variables. According to the functional exposure hypothesis uncertainty reduction alone is neither a sufficient nor a necessary condition for positive repetition-affect relationships to occur.

The empirical evidence supporting the functional exposure hypothesis is of two types (see Poiesz 1983). The first type of evidence was obtained in the social psychological laboratory. Empirical results supported the hypothesis that positive repetition-affect relationships will be found only in task situations in which subjects are uncertain (and apprehensive) with regard to their personal performance on a task. The second type of evidence concerned laboratory subjects' choice behavior in a 'consumer task'. Some support was found for the hypothesis that under conditions of perceived risk (as the consumer behavior counterpart of the concept of subjective uncertainty/apprehension) more familiar brandnames are preferred to less familiar brandnames, and that this preference will be absent/ smaller under conditions of high and low perceived risk.

The functional exposure hypothesis does imply some (pre-or sub)conscious cognitive activity. First, the task demands must be well perceived and understood. Second, the assessment of functionality requires some cognitive activity by which familiarity, as a relevant stimulus characteristic, is related to the requirements of the task. In the developmental literature, strong evidence is provided for the existence of age differences with respect to the conditions stated above. With regard to the first condition, older children (13, 14 years of age and up) have been reported better capable of selecting useful (i.c. functional) information from stimulus components and of tailoring attention to task demands (e.g. Roedder 1981, Hale 19?9, Flavell and Wellman 1977, Ross 1976, Stevenson 1972).

Regarding the second condition, Piagetian theory and research (see e.g. Chestnut 1979 for a reconsideration of Piagetian theory in children's consumer behavior research) suggest that children of about 14 years of age may be expected to differ from younger children (up to about age 12) in their capacity to cognize about the significance or functionality of events. This should also hold for events taking place in an experimental setting such as the occurrence of exposure frequency differences. Also, the older children are better capable of thinking in terms of self-generated hypothetical propositions, of theorizing about new relationships, and of thinking abstractly (see Roedder 1981). Combining the evidence we may expect the probability of observing positive repetition affect-relationships among experimental subjects in an ambiguous, apprehension-generating situation to increase with age (in the range from 9 to 14 years of age). It is hypothesized, therefore, that children of 13/14 years of age will show more positive repetition-affect relationships in such a situation than children of 9/10 years of age. Parenthetically, reference is made here only to levels in the developmental trend, and not to a strict stage theory of cognitive development. In the developmental trend, levels have been identified that are supported by research and theoretical consensus (Fischer and Silvern 1985). For sake of clarity, the above mentioned hypothesis relates to the difference between age levels on the basis of a global distinction with regard to cognitive development. The hypothesis does not pertain to any particular single aspect of cognitive development as a determinant of the expected effect.


Method. Subjects: 70 children of 9/10 years of age and 76 children of 13/14 years of age (resp. primary and secondary school) participated in the present experiment.

Stimulus material. The stimuli were 6 slides, each showing the set of eyes of 12 year old children. This type of stimulus was chosen for to exclude the possibility that the nature of the stimulus, such as, for example, its subjective complexity, would interact with age level. The 6 stimuli were selected out of a large set of eye pairs on the basis of their relative affective neutrality which was established by a small group of children at the two age levels; these children did not participate in the experiment. Slides had a standard format, eye pairs had been photographed from a standard distance and only eyes were shown (no hair, etc.). The photographed children all looked straight in the camera and had a "neutral" facial expression when being photographed.

Equipment. The equipment consisted of a Kodak carousel slide projector, a projection screen placed in front of the subjects, a white instruction board with text in brown. The instruction could be easily read by all subjects.

Design and procedure. The design involved two experimental groups of subjects distinguished on the basis of age, and a control group for each of the age levels. The experiment was carried out at the schools of the subject groups, in the respective classrooms. Per condition, subjects participated collectively. Each of the two experimental groups was confronted with the procedure typically employed in repetition-affect studies. That is, there was no explicit task related instruction prior to the exposure frequency manipulation. The affect rating of all stimuli immediately followed this manipulation. For the two control groups, there was no exposure frequency manipulation and subjects in these groups rated affect associated with the various stimuli on the basis of a single exposure to each stimulus only.

Instructions. 'In a few moments I will show you a number of slides of the eyes of children. Later-on I will explain you the reason why. The eyes are those of children of the ... school in Eindhoven (a nearby city - note by the author). These children are in the same grade and are of your age. Please look at the screen now'. SubJects in all conditions received these instructions. It was assumed that the combination of the absence of unambiguous instructions and the exposure of unknown stimuli would generate apprehension, an assumption proved valid in comparable experimental situations (Poiesz 1983). A particular stimulus was shown 0, 1, 2, 5, 10 or 20 times. Subjects in the control conditions saw each stimulus once only. Frequencies were intermixed, so that each stimulus did not follow itself. Stimuli were exposed for 2 seconds each. Interstimulus intervals were 4 seconds. Because of practical constraints, the stimulus presentation and the affect rating had to take place for all subjects per condition simultaneously. As a result, the order of the frequency levels in the rating phase was fixed: 5, O, 20, 1, 10 and 2 (exposure time: 2 seconds each). Stimuli were rated on the 7-point "smiling-faces" scale (Jolibert and Baumgarten 1979). Subjects were instructed how to interpret the scale. They were requested to 'indicate how nice you think the child is whose eyes you just have seen'.

Results and discussion. Assuming psychologically equal intervals at the exposure frequency continuum, linear trendscores were calculated per condition (see Winer 1971). This was also done for the two control conditions, although, in fact, the equal exposure frequencies in these conditions can not be subjected to a trend interpretation. However, for reasons of comparison, trendscores could be calculated without violating analytical rules. A linear trend analysis showed an unexpected significant interaction between the base lines of the two control groups (F1,87 = 15.95, p<.01), implying that, apparently, stimuli were not equal in affective neutrality and differing for both groups. Consequently, a direct comparison between the two age groups was not possible. The only prudent comparison that could be made was, per age level, the one between the experimental repetition affect relationship and the respective baseline. Then, for children of both age groups, it was observed that repetition-affect relationships differed significantly from the baselines. However, for the younger children the trend of the experimental group was negative relative to its baseline (F1,68 = 7.04, pc.01), while for the group of older children this trend was positive relative to its baseline (F1,74 = 5.08, p<.05). Even though these results confirmed the hypothesis - the younger children (9/10 years of age) would show less positive repetition-affect relationships than children of 13/14 years of age - the interpretation is hampered by the significant difference between the control groups of both age levels and by limitations in the set-up of the present experiment: it did not allow for a counterbalancing of stimuli over stimulus repetition levels. Nor could the order of the stimuli in the rating phase be varied. Therefore, a second experiment was set up, in which procedural flaws of the first study were eliminated.


Basically, the theoretical background of Experiment II was identical to the one of the Experiment I. Again, two age groups were confronted with a manipulation of exposure frequency, and it was hypothesized that the repetition affect relationship of children 13/14 years of age would be more positive than the relationship of children 9/10 years of age.

Method. Subjects: 60 children participated in this experimental study. 30 Children of about 9/10 years of age, and 30 children of about 13/14 years of age (primary and secondary school, respectively).

Stimulus material. For sake of simplicity, in this experiment a less 'multidimensional' type of stimulus was used than in the previous experiment. Stimuli were 6 nonsense words, formed by four consonants and one vowel. These words has been selected out of a set of 33 words on the basis of their affective neutrality for both age levels. This was done in a study, prior to the actual experiment, with a group of 13 children per age group. These children evaluated stimuli on the "smiling faces" scale referred to earlier. The selected words were DONJW, VINCB, RETBG, SBETG, SGLIB, and FUPKZ. The stimuli were presented on slides, all in the same format.

Equipment. Employed were a KODAK carousel slide projector and a projection screen, placed at a distance of about 15 ft. from the subjects.

Procedure. In this experiment, subjects participated individually. After entering the experimental room, the subject was requested to sit in front of the screen. The standard instruction was: 'In a few moments, I will show you a few slides. Later, I will explain you why. Each slide shows the name of a child. The names are all unknown to you - you probably never saw them before. Now, look at the screen and I will show you the slides. Please watch them closely'. During the stimulus presentation, each stimulus was exposed for two seconds. Interstimulus intervals were 4 seconds. The stimuli were presented in random order, but so that each stimulus did not follow itself. A particular stimulus was shown 0, 1, 3, 5, 10 or 15 tines. The exposure phase was followed immediately by the rating phase, in which stimuli had to be rated one by one on the 7-point "smiling faces" scale. Subjects were instructed how to interpret the scale. In the rating phase, repetition levels were rotated over the exposure positions 1 through 6.

Results. Table 1 and Figure 1 show the average affect score per exposure frequency level and per age group.





The results indicate that there is a difference between the age groups with regard to the slope of the repetition-affect relationship. The tendency for the older children is positive, and for the younger children negative. The difference is significant (F1,58 = 10.28, p<.005), confirming the hypothesis.


The results of the two experiments combined provide support for the argument that cognitive development differences co-determine differences between repetition affect relationships.

The results may be worthwhile in that they suggest that age level (level of cognitive development?) may have an autonomous influence on the repetition-affect relationship.

Several critical questions remain to be answered, however. One of these questions is whether age related differences exist with regard to the perception of the experimental situation. A second question is whether other age related differences than cognitive development differences can be held responsible for the obtained results. For example, a point that received possible too little attention in the reported experiment is the one concerning the subjective uncertainty/apprehension assumed to be generated by the experimental situation. Yet another question concerns the interpretation of the slopes (positiveness/negativeness) of the observed relationships. Note that no explicit hypotheses were formulated on the slope of the relationship of each of the group - only on the nature of the difference between the two. The average relationship of the older children was found to be more positive than that of the younger children, as expected. For the latter group, the relationship was negative, however. A possible explanation for this effect is that if the more familiar stimuli are not psychologically functional for the younger children, the repeated exposure of these stimuli may generate boredom or even irritation. This would have expressed itself in more negative affect towards these more familiar stimuli.

It would be premature to consider the external validity and significance of these findings, considering the artificial nature of the controlled experimental situation in which they were obtained and the importance of yet unanswered questions. However, two points suggest the possible relevance of additional research. First, although the stimuli employed in both studies were not marketing stimuli in the strict sense, there may not be a fundamental difference between the nonsense words as used in the second experiment and real life brandnames. Second, even though the experimental situation with which subjects were confronted here was not a consumer choice situation, the behavior in both situations may often be conceived of as problem-solving behavior involving perceived risk or uncertainty as conceptualized here.

Converging the evidence on adult repetition-affect relationships and the obtained (limited) evidence on children's repetition-affect relationships, three hypotheses may be formulated:

1) Children of age 13/14 on (including adults) will show positive affect towards frequently exposed brandnames in buying situations with a moderate degree of perceived risk(which is to be interpreted as determined by the interaction of consumer, product and situational variables).

2) Children up to age 9/10 will show no positive affect towards frequently exposed brandnames in buying situations with a moderate degree of perceived risk.

3) Children (and adults) of all ages will show no positive affect towards frequently exposed brandnames in buying situations with either a very high or a very low degree of perceived risk.


Advertising Age (1957), May 13.

Atkin, C. and G. Heald (1977), "The Content of Children's Toy and Good Commercials," Journal of Communication, 27, 107-144.

Bauer, R.A. and S.A. Greyser (1968), Advertising in America. The Consumer View. Division of Research. Graduate School of Business Administration. Boston: Harvard University.

Berlyne, D.E. (1970),"Novelty, Complexity, and Hedonic Value," Perception and Psychophysics, 8, 279-286.

Berlyne, D.E. (1980), Conflict, arousal and curiosity, New York: McGraw-Hill.

Cantor, G.N. (1968), "Children's 'like-dislike' Ratings of Familiarized and Nonfamiliarized Visual Stimuli," Journal of Experimental Child Psychology, 6, 651-657.

Chestnut, R.W. (1979), "Television advertising and young children: Piaget reconsidered," in J.W. Leigh and C.R. Martin Jr. (Eds.), Current Issues and Research in Advertising, Ann Arbor: University of Michigan.

Ebel, E.W. (1968) op. cit. Bauer, R.A. and S.A. Greyser.

Fischer, K.W. and L. Silvern (1985), "Stages and Individual Differences in Cognitive Development," Annual Review of Psychology, 36, 613-648.

Flavell, J.E. and H.M. Wellman (1977), "Metamemory," in R.V. Kail and J.W. Hagen (Eds.), Perspectives on the development of memory and cognition, Hillsdale, N.J.: Lawrence Erlbaum Associates.

Hale, G.A. (1979), "Development of children's attention to stimulus components," in G.A. Hale and M. Lewis (Eds.), Attention and cognitive development, New York: Plenum Press.

Heingartner, A. and J.V. Hall (1974), "Affective Consequences in Adults and Children of Repeated Exposure to Auditory Stimuli," Journal of Personality and Social Psychology, 29, 719-723.

Jolibert, A.J.P. and G. Baumgartner (1979), An empirical comparison of the properties of three scaling techniques, Institut des Etudes Commerciales, Universite de Grenoble, 39-57.

Obermiller, C. (1985), "Varieties of Here Exposure: The Effects of Processing Style and Repetition on Affective Response," Journal of Consumer Research, 12, 17-30.

Poiesz, Th.B.C. (1983),-The relationship between exposure frequency and consumer affect: toward a functional interpretation, Dissertation Tilburg University.

Roedder, D.L. (1981), "Age Differences in Children's Responses to Television Advertising: An Information Processing Approach," Journal of Consumer Research, 8, 144-153.

Ross, A.O. (1976), Psychological aspects of learning disabilities and reading disorders, New York: McGraw-Hill.

Rossiter, J.R. (1980), "The effects of volume and repetition of television commercials," in Adler et al. (Eds.), The effects of television advertising on children Lexington, Massachusetts, Lexington Books.

Stevenson, H.W. (1972), Children's learning, New York: Appleton-Century-Crofts.

Winer, B.J. (1971), Statistical principles in experimental design, 2nd Ed., New York: McGraw-Hill.

Zajonc, R.B. (1968), "Attitudinal effects of mere exposure," Journal of Personality and Social Psychology, Monograph Supplement, Vol.2, 2, 1-27.