Poster group
Details of individual items:
poster
The distinctive response patterns of Short and Long Lookers inhabituation have been shown to predict distraction task performance (Fricket al., 1999). In the present study, Short and Long Looking infants wereidentified from a habituation task on the basis of the total duration oflooking at the repeating stimulus and they were then compared in acontingency learning task. Based on evidence that Short Lookers have beenshown to encode information more rapidly (Colombo & Mitchell, 1990), it waspredicted that the Short/Long Looker distinction would be reflected in thecontingency task performance. Method: 53 infants ranging in age from 6 to 13 months performed in visualhabituation and contingency learning tasks. The stimuli in theinfant-controlled habituation procedure were checkerboards (4 x 4, 12 x 12).One checkerboard was presented until the infant decreased looking at it by50%, and then the other was presented as a novel stimulus. Looking wasmeasured on-line by corneal reflection and was found to be reliable for twoindependent observers for a subsample of infants. In the contingency task, a 1.5-minute baseline was followed by 3minutes when touches on a canister were reinforced by brief presentations ofa light/sound stimulus. This was followed by 3 minutes of extinction. Acaregiver held the infant throughout the session.Results & Discussion: In habituation, the Short and Long Looker groupsdiffered on several visual fixation measures. Specifically, the ShortLookers had longer first looks, had earlier maximal fixation, and reachedthe decrement criterion sooner than Long Lookers (Table 1). These findingsconfirm that the Short Lookers reached the decrement criterion sooner andlooked less overall but showed comparable responses to the novel stimulus.In effect, the Short Looker infants were more rapid encoders since theyprocessed sufficient stimulus information in a shorter period of timecompared to Long Lookers while demonstrating the same level ofdiscrimination on the novelty trial. On the contingency task, the Long Looker group increasedresponding above baseline during the contingency period consistent with theoccurrence of learning, t(26) 4.8, p < .03. In contrast, the Short Lookersreached their maximum level of responding earlier than Long Lookerssuggesting more immediate responsiveness to the reinforcer. However, therewas no evidence that the Short Lookers demonstrated learning becauseresponding at the end of contingency did not differ significantly frombaseline. Instead, the response profile was consistent with rapidhabituation to the reinforcer (Figure 1). Taken together, these findingssuggest that individual differences in encoding between the Short/LongLooker groups based on an habituation task were represented in the infants'performance in the contingency task. The linkage of individual differences in the two tasks herecontrasted with evidence that habituation and contingency learning aredistinct information processes (Millar & Weir, 1997). These findings pointto the importance of taking account of encoding differences (viz., ShortLooker vs. Long Lookers) when examining the performance of infants incontingency learning and possibly other information processing tasks whichinvolve the presentation of discrete stimuli. References Colombo, J. & Mitchell, D.W. (1990). Individual differencein early visual attention: Fixation time and cognitive processing. In J.Colombo & J. Fagen (Eds.). Individual differences in infancy (pp.193-228),Hillsdale, NJ: Erlbaum. Frick, J.E., Colombo, J. & Saxon, T.F. (1999). Individualand development differences in disengagement of fixation in early infancy.Child Development, 70, 537-548. Millar, W.S. & Weir, C.G. (1997). Relations betweencontingency learning and habituation in 6- to 12-month-old infants: Furtherevidence. European Journal of Cognitive Psychology, 16, 537-554.Table 1. Means for Short and Long Looker groups on habituationmeasures with the F-values from analyses of variance and significancelevels. Measure Short Looker Mean Long Looker Mean Analysis ofvariance F(1, 49) Signifi-cance level HABITUATION TASK SCORES Look 1 duration 3.9 s 6.0 s 4.88 < .04 Longest Look 2.3rd look 3.7th look 4.97 < .03 Looks to decrement criterion 6.5 looks 9 looks 10.14<.01 Slope -.68 s/look -.73 s/look < 1 ns Increase looking at novel stimulus 2.37 s 2.47 s < 1 ns CONTINGENCY LEARNING SCORES Response change from baseline to contingency -4.3 responses +4.7responses 8.00 < .01 Peak response segment 2nd out of 6 3.3 rd out of 6 6.85 < .02
poster
Individual differences in visual attention and their implications for recognition memory were studied at 5, 7, and 12 months in a longitudinal sample of full-term and preterm (<1750 g) infants. Participants were part of a larger ongoing longitudinal study of the development of specific cognitive abilities and processing speed in full-terms and preterms from infancy through age 3 years. The sample at 5, 7, and 12 months consisted of 153, 144, and 126 full-terms, respectively, and 50, 50, and 56 preterms (some preterms were newly enrolled at 7 months).Infants were given 9 paired-comparison problems, 5 using Face stimuli and 4 using Abstract Patterns. Each problem consisted of a familiarization exposure to two identical stimuli followed immediately by a test in which one of the familiar targets was displayed along with the novel member of the pair (with left-right position reversed in the second part of each test). During the familiarization phase of each problem, the two identical stimuli remained on display until the infant had accumulated the pre-determined study time: 20 s for Faces and 5 s for Patterns at 5 and 7 months; 10 s for Faces and 3 s for Patterns at 12 months. Test pairs were presented for two 5-s periods at all ages, with a minimum of 1 s of looking required during each period. Measures of look duration (mean and peak look), shifting of gaze between two targets (shift rate), and time off-target (exposure time and mean pause) were drawn from the familiarization phase. Mean look and shift rate were also measured during the test phase. Developmental changes during the first year of life were pronounced. Over age, mean looks, peak looks, and pauses became shorter and shift rates got faster. Shift rates increased by as much as 90% from 5 to 12 months, whereas mean look duration fell around 40%. Preterms showed more immature patterns of attention at all ages. They had longer peak looks, longer mean looks, longer pauses, fewer shifts, and lower novelty scores than their full-term counterparts. Infants were remarkably consistent within each age in the various measures of attention. Across problems of the same type (Faces or Patterns), alpha coefficients were in the .60's and .70's; across problems of different types, correlations were in the .30's and .40's; and across phases (familiarization and test), correlations were in the .60's. There was also modest across age stability for adjacent ages (with a number of correlations in the .20's and .30's). Attention measures from the familiarization phase, aside from showing developmental change and cross-age consistency, were related to subsequent recognition memory. At each age, faster shift rates and shorter look duration during familiarization were associated with better recognition, with shift rates relating to novelty scores independently of look duration.
poster
Development and individual differences in sensitivity to distractors are explored in 3- to 5-month-old infants. A peripheric distractor was presented when infants sustained attention on a central stimulus for 4 seconds, and both targets remained present for 10 seconds. A session involved 5 trials, with the central stimulus changing from trial to trial, while the distractor was constant. Three questions were investigated: how do infants organize their attention between stimulus and distractor ? May habituation to distractor be observed ? Is there any intraindividual stability between ages ? When infants are 3-month-old, three strategies are observed. Some infants shift quickly to the distractor and are very active visually; others remain on the centre; infants in the third group had no systematic strategies. However, those differences disappeared two months later, and no individual stability may be observed. Habituation to the distractor appeared only in 5-month-olds. Results are discussed in relation to Richards' and Ruff's data, stressing the role of obligatory attention.
poster
The processes involved in the infant visual habituation-recovery paradigmhave been well studied. However, one aspect of the process, namely, therelationship between latency to fixation and length of looking has beengenerally overlooked. Few studies have focused on whether latency times tofixate a stimulus during an infant-controlled procedure are related tosubsequent looking lengths. The goal of the present study was to rectifythis lapse by examining whether a relationship exist between the latency tofixation and length of look during four distinct trial blocks of aninfant-controlled habituation-recovery-dishabituation paradigm.Specifically, the study was designed to determine whether the latency timesduring baseline, criterion, recovery, and dishabituation trials were relatedto looking times during these trial blocks. Moreover, the sample was dividedby gender to determine whether the relationship (if observed) was genderspecific. The sample consisted of 190 same-sex 5-month-old twins recruitedfrom the Quebec Newborn Twins Study; a prospective longitudinal study aimedat assessing the cognitive, social, and emotional development of twins frominfancy to adolescence. Using an infant-controlledhabituation-recovery-dishabituation procedure infants were presentedrepeatedly with a computer generated multicolored, rotating ball until theirlooking dropped below 50% (criterion looking) of initial baseline looking.Following habituation, the infants were presented with a novel translating,black-and-white checkerboard pattern for two trials. This was followed bytwo presentation of the original stimulus. During the habituation phase thelocation of the visual stimulus shifted between the right and left half ofthe screen for each trial. Prior to each trial, a small red-square flashedin the middle of the screen for two seconds in an attempt to maintain theinfants' attention toward the screen. Latency times were computed bycalculating the time required by the infants to shift their gaze from theprevious location of the stimulus to the subsequent location. In this mannerit was possible that negative latency times could be observed (i.e., theinfants looked toward the appropriate location of the upcoming stimulusprior to its appearance). Looking times were calculated from the time theinfants first looked at the stimulus until they looked away from thestimulus for a period of 1.5 seconds. Table 1 presents the intercorrelationsfor the trial blocks as a function of gender. With the exception of thebaseline trials, latency times were negatively correlated with looking timesfor both males and females. The correlations indicate that infants withshorter latency times looked longer at the stimulus during these trials. Theresults of the present analyses suggest that the time required for theinfants to fixate a stimulus may be related to their level of interest inthe stimulus, particularly during the criterion, recovery, anddishabituation trial blocks. The results will be further discussed in termsof Cohen's attention-getting and attention-maintaining model of infantattention.
poster
The look time habituation paradigm is widely used in research on infants'perceptual and cognitive capacities, but with few exceptions littleprogress has been made in understanding what information processingmechanisms are measured by habituation. Moreover, while there are avariety of criteria in use for determining whether an individual infant hashabituated to a particular display, few of these criteria have beensubmitted to a formal analysis which evaluates either their relation to aspecific information processing model or their statistical properties.In this presentation, we will describe a new approach to the problem ofdescribing and understanding habituation and show how this approach may beapplied at the level of the individual infant. First, we show thathabituation criteria that average infants' responses have an undesirablestatistical property: the expected value on a given trial is dependent oneach infant's mean level of response. Next, we present a regressionapproach (c.f. Ashmead & Davis, 1996) which overcomes this limitation byattempting to fit a 'true' habituation function to the observed look timedata. In doing so, we evaluate several alternative parametric andnon-parametric models for their goodness of fit and ease of interpretationin an information processing framework. The model we most strongly favor isan exponential function with parameters that describe meaningfulinformation processing variables such as the rate of habituation, theinitial response, and the baseline or minimum response level. Finally, wedemonstrate how the use of classical and empirical Bayes methods allows oneto estimate the parameters of the model for individual infants. EmpiricalBayes methods produce improved parameter estimates for an individual infantby 'borrowing strength' from data supplied by others. These methods may beespecially attractive to researchers who work with infants because there isoften little data for individual participants and the collected data arehighly variable.We illustrate the use of these procedures in real data and make suggestionsfor how they might be implemented in on-line settings.
poster
This study was designed to examine attention in infants as theymanipulated objects. In previous research, Ruff (Child Development, 1986)has suggested that examination of an object by fingering and turningreflects focused attention and the intake of information. Likewise, thelatency to examine the object may represent the time that it takes theinfant to activate the attention system and prepare to gather information.Mouthing, however, does not represent focused attention. The present studywas particularly interested in several variables that were not studied byRuff (1986). Particularly, the present study was interested in examiningsex differences, and differences that may occur as the result of beingseated either in an infant chair or their mother's lap. Furthermore, thepresent study was designed to study latency and duration of bothexamination and mouthing behaviors. Fifty-nine healthy, full-term 8-month-old infants participated in thisstudy (32 boys, 29 girls). The infants were randomly assigned to eithersit in an infant chair or in their mother's lap. Infants were observedwhile interacting with four different toys, one at a time. To ensure thateach toy was novel, parents were asked if they had each particular toy athome. In the case of a 'yes' answer, that particular toy was not used.All toys were counter-balanced in the order in which they were presented,and were chosen from a total of six different toys. Each infant was presented a novel, attractive toy and allowed tomanipulate the toy for 60 seconds. Parents were asked to sit quietly andlet their infant play with the object freely. If the toy was dropped itwas retrieved by the experimenter. If the toy was dropped more than threetimes the trial was ended and the next trial was started. Therefore,trials lasted for 60 seconds or until the infant dropped the toy threetimes, whichever occurred first. The trials were videotaped for latercoding, and were coded by an undergraduate who was unaware of anyperformance expectations. On the first trial, there was an interaction such that males spent anoverall greater percentage of time mouthing the toy, while females spent agreater percentage of the time examining the toy [F(1,56)10.27, p<.01].This difference could be attributed to faster brain maturation of thefemale infants, who spent more time examining the toy and focusing theirattention for the intake of information about the toy itself. Furthermore,results indicated that infants sitting on their mother's lap spent anoverall higher percentage of time mouthing the toy than infants who weresitting in an infant chair [F(1,58)4.65, p<.05]. The results suggest that differences in focused attention while examiningtoys may occur as the result of the sex of the infant or the context ofwhere the infant is seated. Ruff has suggested that developmentalprogressions in attention occur between 7 and 12 months of age. Therefore,the present study may represent only a snapshot of information from our8-month-old sample. These findings would be particularly interesting tofollow longitudinally, in order to see if these differences are eliminatedor exacerbated with development.
poster
In the human being habituation has mainly been studied with 4 to 5 month-old subjects even if it can be evidenced in the new-born, the fetus, the adult and even at a another level, the neuron (Kandel, 1979).The models found for habituation are quite different one from the others although most of them find their origin in Sokolov's conception (Sokolov, 1963) and very few contradictory tests have been attempted since. Each model presented until now seems to be relevant enough to avoid any controversy. Our aim is certainly not to question such models well rooted in the context of visual habituation. we rather aim at proposing yet another interpretation of visual habituation in the baby which would be based on Pearce et Hall model (Pearce & Hall, 1980). Once applied to the 4 month-old baby, this model, tested on animals can open the way to an interpretation of visual habituation in terms of stimuli associability hence avoiding the classical conception of a baby that stares at a target in order to create an inner model of the stimulus. Pearce et Hall model suggests that a conditional stimulus is not processed if it is an accurate prediction of what follows. However, an event E1 will draw attention if its associative strength is weak. It seemed to us relevant to adapt this model to the human baby in as much as authors like Haith et col. (1988) noticed an anticipation of visual fixation in the direction of a coming target, which shows that even at a young age, the subjects are able to predict the occurrence and localization of an unconditional stimulus when it appears alternatively either in the baby's right or left visual field.The results we obtained do not conform with the classical sokolovinan interpretation but can accurately be interpreted in terms of associative learning. However, our point of view is in accordance with Sokolov's original conception of habituation and the orienting response. Hence, a close examination of Sokolov's writings and a careful analysis of our results allowed us to establish that Pearce et Hall's model and Sokolov's model are compatible and complementary in order to improve understanding of habituation in the baby.
poster
A series of previous studies conducted in our laboratory investigated 6-month-old infants attention, encoding, and recognition processes using visual stimuli that included a central focal cue surrounded by background context cues (e.g., a picture of a snowman surrounded by a pink plaid pattern). Results indicate that if initial encoding takes place during a relatively brief familiarization period, then recognition of the central target is successful when the background pattern that was present during initial encoding is also present during recognition-test, but not when the test background pattern is different during test. Furthermore, a recent experiment demonstrated that context-dependent recognition can also be observed following a longer encoding period (habituation rather than familiarization), but only when a brief delay (2.5 min) occurs between encoding and recognition-test. Results such as these, demonstrating context-dependent retrieval of focal information, reflect what has been described as a gating function of context.The design of the habituation experiment included two context conditions (same context during habituation and retrieval; different context during retrieval) and three delay intervals between habituation and recognition-test (0, 2.5, or 5 minutes). Because of certain results we decided to add a third context condition in which the area surrounding the focal target was left blank (a no-context-added condition); although the same background context was present during habituation and retrieval, this background was uninteresting and uniformly gray. The focus of the present report is on an unanticipated difference among the three context conditions in the length of time required to meet the habituation criterion and subsequent test performance.Data were provided by a total of 108 6-month-old infants when the habituation experiment was combined with the follow-up to produce a 3 context condition by 3 delay interval design. Three relevant findings emerged. First, significantly longer times to reach habituation criterion were found in the no-context-added condition (64.7 sec) than in the same- or the different-context conditions (39.45 and 44.96 sec, respectively). Second, the within-conditions correlation between duration of looking during habituation and recognition of the focal cue on test trials (novelty preference) was extremely low (r -.09). Third, the between-conditions correlation between the same two measures was extremely high (r +.95).Habituation to focal cue occurred rapidly when that cue was surrounded by a salient context pattern, but occurred much more slowly in the no-context-added condition. Considered in light of previous findings indicating that background context cues receive attentional priority over focal cues, these findings suggest that attention to a salient background context facilitates attention to and encoding of focal cues. Furthermore, differences in habituation looking time between context conditions were also associated with differences between those conditions in recognition of the focal cue. Our interpretation of these findings is that stimulus context serves not only a gating function in focal-cue retrieval; it also serves an attention-getting function during initial encoding of focal information.
poster
Previous studies in our laboratory have investigated 6-month-old infants' attention, encoding, and recognition processes using visual stimuli containing a central focal cue surrounded by a background context (e.g., a cat surrounded by a pink plaid pattern. In a recent experiment, we investigated the influence of context on focal-cue recognition at varying test intervals in an habituation procedure. Infants' recognition of the focal cue was found to vary as a function of the test context as well as the delay between habituation and test. These results suggest that attention to contextual and focal information may vary over time. This issue was the focus of the present experiment, in this case using a familiarization procedure. Sixty-four 6-month-old infants were familiarized to one of two visual compound stimuli (i.e., a cat surrounded by a pink plaid pattern or a window surrounded by a brown starred pattern). Familiarization continued until infants accumulated 10 s of looking on each of two trials. Two recognition test trials were presented 0-, 2.5-, or 5-min after completion of the familiarization phase. Test stimuli included (a) a novel focal cue surrounded by the familiar context pattern, and (b) a familiar focal cue surrounded by a novel context pattern. Total durations of looking on test trials were converted into a novelty preference score, reflecting infants' preference for the novel focal cue (the complement of this score reflects preference for the novel background). These scores were entered into a 3 (Delay) x 2 (Stimulus Set) ANOVA. Results revealed a significant main effect for delay, with higher novelty preference scores at the 2.5-min delay than at the 0-min delay, and a significant stimulus set effect, with higher novelty preference scores for the Window-Brown Stars stimulus set than the Cat-Pink Plaid stimulus set. Although the interaction did not reach significance, two-tail t tests comparing novelty preference scores in each cell of the design against chance (.50) revealed that novelty preference scores were (a) significantly below chance at the 0-min delay for the Cat-Pink Plaid stimulus set and (b) significantly above chance at the 2.5-min delay for the Window-Brown Stars stimulus set. No other comparisons reached significance. Results indicate that attentional deployment varies as a function of (a) the interval between learning and test as well as (b) the saliency of stimulus context. Immediately following learning, attentional priority is given to (novel) context, but only when that context is highly salient.After 2.5-min, infants can ignore context and redeploy attention to the (novel) focal cue. However, because context is still competing for attention, attention is only directed to focal information when the (novel) context is less salient. Although the same general trend is still present after 5-min (greater attention to context when context is salient but greater attention to target when context is not salient), novelty preference scores no longer differ from chance as both target and context begin to fade from memory.
poster
This study examined the development of infant shifts of covert attention toperipheral stimuli when attending to a focal visual stimulus. One goal ofthe study was to determine if there were developmental changes in covertorienting from the ages of 3 to 6 months. A second goal of the study was toexamine the effect of attending to a central visual stimulus when the cuewas presented in the spatial cueing paradigm.Infants at age 14, 20, and 26 weeks (45 per age, 135 total) were tested in acovert attention shift procedure. A central stimulus was presented that wasknown to elicit attention, and following a delay, a 'cue' stimulus waspresented in the periphery for 300 ms, and both stimuli were turned off.Infants typically will not look toward this competing peripheral stimulusbut fixation will remain on the central stimulus. Followingstimulus-onset-asychronies (SOA) of 450, 875, or 1300 ms, a 'target'peripheral stimulus was presented ipsilateral (valid trials) orcontralateral (invalid trials) to the cue stimulus. The 'target' and 'cue'stimuli were a 2 deg by 6 deg rectangle with changing sinusoidal patterns.A no-cue control condition was done in which just the target appeared. Thedelay between the onset of the central stimulus and cue was eitherimmediate, two seconds, or until a significant change in heart rateoccurred. The immediate delay represents a lack of focal stimulus attention,the two second delay represents a period of 'stimulus orienting', whereasthe heart rate deceleration delay represents 'sustained attention' to thecentral stimulus. The time from the onset of the target to the onset of thefirst saccadic eye movement to the target was analyzed as the reaction time.The results show that attention to the central stimulus is important incovert orienting effects in young infants. First, at the onset ofattention, 'stimulus orienting', both facilitation and inhibition of returnoccurred. The two-second condition resulted in expected 'spatial cueing'effects such as facilitation and inhibition of return. All three ages had ashorter saccade reaction time on the valid trials at short SOA's,facilitation. The 20- and 26-week-old infants also had lengthened saccadereaction times on the valid trials for the longest SOA, inhibition ofreturn. Second, when attention was unengaged there were no differences inreaction time for the valid and invalid trials, suggesting that the infantsdid not covertly orient toward the cue when it was presented. The immediatecondition did not result in any effects of the valid-invalid target locationbut only in a decrease in reaction time for both conditions as a function ofSOA. Third, when attention was fully engaged, 'sustained attention' (heartrate deceleration condition), facilitation of reaction time and inhibitionof return occurred at all three ages. The inhibition of return effectgradually increased in magnitude over the three testing ages. Thus, thisprocedure used to examine infant covert orienting shows developmentalchanges in covert attention when focal stimulus attention is engaged.
poster
Recently there has been a great deal of interest in the development ofattention and distractibility in infancy (Lansink & Richards, 1997; Oakes&Tellinghuisen, 1994; Richards, 1997; Ruff, Capozzoli, & Saltarelli,1996). In several studies, researchers have investigated distractibilityby presenting infants with an initial stimulus (e.g., a toy) and thenpresenting a distracting event (e.g., an image on a tv monitor or a slide)while infants are attending to that stimulus (e.g., Oakes & Tellinghuisen,1994; Ruff et al., 1996). Distraction latencies (i.e., the amount of timeit takes infants to turn from the initial stimulus to the distractingevent) are measured. Using this paradigm, research has shown that wheninfants are engaged in focused attention (i.e., looking at the initialstimulus with signs of concentration), they exhibit longer distractionlatencies (i.e., orient more slowly to the distractor) than when theylooking at the stimulus without signs of concentration (e.g., Lansink &Richards, 1997; Oakes & Tellinghuisen, 1994; Ruff et al., 1996). Inaddition, the characteristics of the distractor influence distractionlatencies. For example, infants exhibited shorter latencies to bimodaldistractors than to unimodal ones (Tellinghuisen & Oakes, 1997). One important question, therefore, is do the characteristics of theinitial stimulus also influence infants' distraction latencies? Morespecifically, how do the information processing demands of the initialstimulus affect infants' attention allocation? Novel and familiar objectsplace different processing demands on the child: there is moreinformation to process in novel toys than in familiar toys. The presentstudies investigated how toy familiarity influenced 9- and 10-month-oldinfants' distraction latencies. In two Experiments, infants participatedin a toy familiarization phase first and then a distractibility phase.The distraction paradigm described above was used, and distractionlatencies were measured when infants were in focused and casual attentionat distractor onset. In Experiment 1, 9-month-old infants first werefamiliarized with two initial stimuli and then were tested in thedistractibility task with those initial stimuli and with two novel initialstimuli. In Experiment 2, 10-month-old infants first were familiarizedwith two potential initial stimuli and then were tested in thedistractibility task with either those initial stimuli or with two novelinitial stimuli. In both Experiments, infants had shorter latencies(i.e., oriented more quickly to the distractor) when investigating thefamiliar toys than when investigating the novel ones, and infants hadlonger latencies when they were engaged in focused attention than whenthey were engaged in casual attention. Thus, the information processingdemands of the initial stimulus and attentional state influenceddistraction latencies. Infants were more easily distracted when theinformation processing demands were low (i.e., the toy was familiar) thanwhen the information processing demands were high (i.e., the toy wasnovel).
poster
We are interested in the organization of infants' looking duringpaired-comparison tests. Many researchers studying perceptual-cognitivedevelopment rely on infants to look longer at novel than familiarstimuli. Complex stimuli are hypothesized to be progressively encodedwith repeated presentations leading to an initial familiarity preference(exhibited while the memory or schema is incrementally formed) that isreplaced with a preference for novel items (Bahrick & Pickens, 1995;Hunter & Ames, 1988; Wagner & Sakovits, 1986). We began using acontinuous or serial paired-comparison task hoping to identify whenindividual infants made the transition from a familiarity to noveltypreference, and to determine whether the transition was gradual orabrupt for individual infants.In our first study, we video taped 15 infants between 3 and 6 months ofage as they looked at pairs of faces. One face, the familiar stimulus,was paired with a series of 24 novel faces. Because we repeated thepresentation order when infants remained content, 48 trials werepossible during a 15-minute testing session. Although only 2 infants hadmean novelty preferences across all trials above 60%, all 15 infants hadat least 1 novelty run (defined as an average novelty preference acrossa minimum of 4 trials of at least 60%). Only 4 infants displayed afamiliarity run before displaying a novelty preference. We alsodiscovered that 11 infants displayed an overall side preference duringthe session. The results of this first study failed to confirm theexistence of initial familiarity preferences as suggested in previousliterature, suggesting instead that many infants waiver between noveltyand familiarity preferences.We felt that further investigation of the stability of these sidepreferences was necessary to determine whether the side preferencesmasked infant familiarity-novelty preferences and contributed to therelatively brief novelty runs. We designed our current study toinvestigate developmental changes during the continuouspaired-comparison task. We tested 9 infants using a sequential design.Each infant's looking was video taped 3 times at 3-week intervals.Infants were either 10, 13, or 16 weeks at their first visit. Wemodified the procedure from our previous study by shortening the overalllength of the task, and by presenting animated stimuli rather thanphotographs of faces. We developed 41 brief movies (one designated asthe familiar stimulus) for each session, and simultaneously displayedtwo movies per trial. Each movie consisted of 2 colorful shapes(sphere, cube, cylinder, star, etc.) that moved for 7 seconds once theinfant looked toward either monitor. We are currently coding eachchange in looking direction throughout the task and will presenttime-series data for the 9 infants. Graphic displays of looking times,novelty preferences, and the distribution of right and left looks willbe presented for each visit. Data analyses will focus on whether sidepreferences and/or familiarity preferences are consistently displayedfrom one visit to the next or whether they are more likely to bedisplayed at one age than another. Our aim is to identify individualdifferences in the organization of looking when infants aresimultaneously presented with two interesting events.
poster
The purpose of this study was to test the proposition that, compared tofull term infants, preterm infants may engage in information processingstrategies characterised by a more limited tendency to inhibit attention tonon-salient stimuli. This aim was achieved by examining the responses of32 full term (FT) and 32 preterm (PT) infants (mean gestational age 28wks)to a visual pop-out display. Infants were assessed at 16-18 wks afterexpected date of delivery. They were initially familiarised over a seriesof 20 one-second trials to a visual array consisting of five identicaldistracter items of one colour and one target item differing only incolour. Recognition of both target and distracter colours was then assessedon two, five second test trials in which a new colour (novel stimulus) waspaired with either the target or distracter colour. Novelty preference wasmeasured in terms of time spent looking at the new colour. Order ofpresentation of target/novel and distracter/novel pairs and position of thestimuli were counterbalanced. ResultsTable 1 Percentage of time spent looking at the novel stimulus when pairedwith the target and the distracter stimulus Full Term (FT) Preterm (PT)Novel v Target colour 72.78 52.26Novel v Distracter colour 57.07 52.49A 2 (Group: Full term vs Preterm) x 2 (Test Trial: Target vs Distracter)mixed design ANOVA was carried out. A significant main effect for Group (F 11.64, p< .001) was found. Preference for the novel item during testtrials was greater in the FT group than the PT group. A significant maineffect for Test Trial (F 5.33, p < .05) was also found, with preferencefor the novel stimulus being greater on the novel v target test trail (mean 62.52) than the novel v distracter test trial (mean 54.79). There wasalso a significant Group x Test Trial interaction (F 5.66, p < .02),indicating a different pattern of novelty preference across target anddistracter test trials for the FT and PT groups. During the test phase the FT group demonstrated a significant preferencefor the novel colour relative to the target colour, but not the distractercolour. This indicates that the target had popped out of the visual arrayand been processed during the familiarisation trials. In contrast the PTgroup spent almost equal amounts of time looking at the novel, target anddistracter colours during the test phase. This pattern of looking suggeststhat during familiarisation the target had not been processed morethoroughly than the distracter. Further analysis of the PT group indicatedthat many infants responded in a similar manner to FT infants but data fromthese infants were counterbalanced by data from a group who demonstrated asignificant preference for the familiar target during the test phase.