Poster group
Details of individual items:
poster
Research on the genetic and environmental determinants ofinfant temperament revealed only moderate contribution ofthe genotype to the individual differences. Howeverrecent studies have shown that the psychometric strategyof popular questionnaires such as EASI or IBQ may lead tothe inconsistencies in the genetic data (Goldsmith atal., 1997). This rises the question to which extent theresults from the literature are generalizable totemperament questionnaires that are free from thesedisadvantages?Using sample of 74 healthy twin pairs (corrected age 8-11months) we analyzed heritable influences on a variety oftemperamental dimensions. Assessment of temperament wasbased on 'Baby's Day' Questionnaire (Balleyguier, 1981)that is developed with different conceptual approach andpsychometric strategy than EASI or IBQ. The factoranalyses was applied to reduce 25 highly interrelatedtemperamental scales to several independent factors.As a result of factor analysis we revealed 5 principalcomponents representing the basic temperament dimensions(traits). Three of them - Activity, Sociability andIrritability - are traditionally considered as the traitsof temperament. Two other factors - Aggressiveness andSensitivity to signals of punishment - have never beenassessed in other infant questionnaires.The model fitting analysis provided evidence that duringsecond half of first year of life the individualdifferences in Activity, Irritability and Aggressivenessare predominantly determined by genotype (89%, 85% and92% respectively). The individual differences inSociability depend mainly on shared environment (88%),whereas Sensitivity to signals of punishment isinfluenced by both genetic and shared environment factors(41% versus 52%).Current results in the context of literature arediscussed.
poster
Numerous studies have attempted to explore temperament in infancy, withadditional work quantifying specific dimensions such as activity anddistress. Although reports from various laboratories have described thetrajectory for negative emotionality over the first year (Calkins, Fox &Marshall, 1996; Lewis & Ramsay, 1995), almost nothing is known withrespect to how early feeding, that is, breast- vs. formula-feeding, maymediate the infant’s expression of distress. Except for a fewinvestigations (Bernal & Richards, 1970; DiPietro, Larson & Porges,1987), little published work describes behavioral differencesattributable to feeding, though what does exist suggests that breast-fedinfants are more irritable. The present study was aimed at furtheringthis line of research by supplementing observer ratings withquantifiable estimates of behavioral and physiological distress. Ninety-four infants, enrolled in a longitudinal study of stress andcoping, were seen at 2-, 4-, and 6-months when visiting theirpediatrician’s office for a routine well-baby checkup. While in thewaiting room, mothers completed the Infant Behavior Questionnaire (IBQ;Rothbart, 1981) and a salivary cortisol sample was obtained from theinfants. Following the physical exam, the infants’ response to adiptheria-pertussis-tetanus (DPT) inoculation was video-recorded forlater coding. Approximately 20 minutes later a second saliva sample wasobtained. Following the 2-month visit, it was determined that 35 infantswere still being exclusively breast-fed, and 42 infants were exclusivelyformula-fed, and these are the infants for whom data was analyzed forthe present report. Based on the IBQ ratings, summary scores fordistress to stimuli, distress to limitations, duration of orienting, andsmiling were tabulated. The infants’ maximum level of distress and timeto quiet were calculated after coding the video-recorded DPTinoculation. Finally, the infants’ stress response was determined basedon cortisol output.The results showed that negative emotionality (distress) was ratedhigher for the breast-fed than for the formula-fed infants, withpositive emotionality (smiling and orienting) higher for the formula-fedthan the breast-fed infants. The analysis of observed behavior showedthat breast-fed babies cried for a longer number of intervals at themaximum level of distress. Conversely, formula-fed infants calmed morequickly after the DPT inoculation. Breast-fed infants displayed ahigher baseline and post-inoculation level of cortisol output, relativeto the formula-fed infants, but both groups of infants rose markedlyfrom baseline in response to the inoculation.These results add to a small but consistent literature that suggestsbreastfeeding may be associated with heightened arousal. Whiledifferences in interactive style between breastfeeding and formulafeeding dyads have been widely reported, the discussion will highlightthe differences between the composition of breast-milk and formula as amore suitable explanation for these findings.
poster
Until recently, the examination of helplessness focused on olderchildren and adults, because researchers commonly assumed that youngchildren were invulnerable to helplessness (e.g., Miller, 1985; Rholeset al., 1976). Research now suggests that some children as young asfour years of age succumb to the negative effects of failure (e.g.,Hebert & Dweck, 1985; Heyman et al., 1992; Smiley & Dweck, 1994);however, there has been little work on the origins of helplessness or onthe age at which helplessness first can be seen. The current studyassessed whether the behaviors that characterize helplessness in olderchildren and adults fit together into a meaningful pattern reminiscentof helplessness in toddlers. One hundred twenty-four toddlers (56 female, 68 male) were seenin a laboratory playroom at 25 and 32 months of age. Toddlers' taskavoidance, persistence, negative affect, and shame were measured duringan impossible shape-sorting task and their reluctance to engage with asubsequent similar, but possible, task was also measured. Previous studies examining helplessness in older children andadults have found relations between several behavioral measures ofhelplessness, but no study has used factor analyses to determine whetherthese behaviors fit together into a meaningful pattern. A factoranalysis of toddlers' behavior showed that the individual behaviorscharacteristic of helplessness did not load onto a single factor. Instead, two factors emerged at each age; these two factors were similaracross ages. The two factor structure that emerged at both ages suggeststhat helplessness falls along two distinct dimensions in toddlers: onemore affective and one more behavioral. At both 25 and 32 months, negative affect and shame loadedpositively and persistence loaded negatively on one factor (AffectiveHelplessness) which was relatively stable (r .28, p < .05) across thetwo ages. The Behavioral Helplessness factor consisted of slightlydifferent behaviors at 25 and 32 months and was not stable. At 25months, task avoidance and reluctance loaded positively and persistenceloaded negatively on Behavioral Helplessness, whereas at 32 months, taskavoidance and reluctance loaded positively and shame loaded negativelyon Behavioral Helplessness. It appears that a relatively stable pattern of behaviorreminiscent of helplessness exists by the end of the second year oflife. The rudimentary form of helplessness that is present by the endof the second year continues to develop throughout the third year. Theconstruct of Affective Helplessness changes subtly during the third yearof life despite the fact that it is relatively stable and its factorstructure remains relatively constant. For example, at 25 months, lowpersistence loads on both Affective and Behavioral Helplessness, and infact, loads more heavily on the latter. By 32 months, low persistenceno longer loads on Behavioral Helplessness and becomes an even strongercomponent of Affective Helplessness. As toddlers mature, the affectiveand behavioral components of their actions appear to become moreintertwined and a more cohesive construct emerges. Now that a relatively stable pattern of behavior reminiscent ofhelplessness has been identified in toddlers, future research shouldexamine whether this pattern of behavior relates to measures ofhelplessness in preschoolers and older children.
poster
Maternal reports of infant temperament correspond weakly withobservational measures of temperament, even when mothers both observe andrate infant behavior (Seifer, Sameroff, Barrett, & Krafchuk, 1994). Thislack of correspondence has been attributed to maternal bias in ratinginfant behavior, although other factors, such as the differences in thecontexts in which observations and ratings are made may also affectbetween-measure correspondence. Maternal ratings are typically based oninfant behavior in contexts in which mothers are present and involved withtheir infants to varying degrees, whereas, observations are typically madein contexts in which mothers are not involved. Furthermore, observationsare often made during care-taking tasks that do not elicit distress,whereas, maternal reports focus on how infants respond in potentiallydistressing situations. Additionally, the lack of correspondence betweenratings and observations may reflect on the validity of the observationsrather than the validity of maternal ratings. Testing the predictive andconstruct validity of each in relation to other measures is necessary toaddress this possibility. In the present study, we examine the agreementbetween maternal reports and observations of temperament in a series ofstandardized laboratory settings in which contexts vary by virtue ofmothers' active involvement and the emotion arousing properties of theactivities. We will then compare the predictive validity of each type ofmeasure in relation to other variables. The community sample of 92 primiparous mothers completed threesubscales of the IBQ (distress to limitations, distress to novelty,soothability) when their infants were 5 months old. At 6 months, infantsparticipated in a laboratory assessment of temperament consisting of fouractivities similar to those reflected in the IBQ scales, with the degreeof mother involvement varied across activities. The behavioral data isbeing continuously coded on a 7-point scale ranging from high positiveaffect to high negative affect yielding several observationally basedreactivity scores (e.g., peak intensity of negative affect, ratio ofpositive to negative affect). We will address the following reliability and validity issues: Q1)is the agreement between maternal ratings and independent observationshigher than in previous studies due to the comparability of the contextsin the standardized setting and those reflected in the IBQ; Q2) doesagreement increase when behavioral observations are aggregated acrosscontexts relative to agreement based on observations based on a singlecontext; Q3) is agreement higher based on observations from motherinvolved versus mother uninvolved contexts; Q4) what is the relativepredictive validity of the observational data and maternal ratings inrelation to outcomes (preliminary analyses indicate that maternal ratingsof temperament are associated with maternal self-efficacy and maternalsensitivity); and Q5) does including both measures of temperament increasethe prediction of these outcomes thus demonstrating improved constructvalidity. Question 1 will be examined through simple correlations;Questions 2, 3, and 4 will be examined using simple correlations and bycomparing the relative strength of the correlation coefficients; andQuestion 5 will be addressed using hierarchical multiple regression and/orthe construction of latent variables.
poster
When parents report on behavior of their infants and children, there isoften inaccuracy in those reports. Sometimes the inaccuracy issystematically related to parent characteristics and is referred to asbias. We designed this study to examine sources of inaccuracy and bias inparents' reports of infant temperament. Mothers (n112) of 6-month-olds rated 4 2.5-minute behavior episodes oftheir own children (videotaped immediately before the ratings were made inthe laboratory) and 2.5-minute behavior episodes of 6 standard children.Trained observers also rated each of these video segments. Both mothersand trained observers used 14 bipolar scales defined at each pole byadjective triads. Four summary scales were derived: mood, activity,approach, and intensity. Thus, there was a profile of 16 scores for eachstudy child, and 24 scores for the standard children. In addition, bothparents and trained observers rated the behavior of these infants in theirhomes (aggregated across 6 home visits). To compare mother and observer ratings, q-correlations of the profiles forown children and standard children were computed for each mother. Theaverage of the correlations between mothers and trained observers forstandard children was .83 (with all but one of these 112 correlations above.60). This level of agreement would be consistent with high inter-raterreliability. In contrast, the average of the q-correlations for ownchildren was .32 (19% of these correlations were negative). This isconsistent with unreliability of raters. The mean difference between theq-correlations was .51 (1.75 SD), a very large effect.The discrepancies between mothers and observers ratings of own childrenwere then compared to questionnaire assessments of temperament and ratingsof behavior observed in the home. In those cases where mothers were morenegative in their ratings than observers, the mothers' responses werelikewise more negative on the Carey Infant Temperament QuestionnaireDifficulty Scale (r.24) and on the aggregated adjective triadmother-report scales of mood (r-.46), activity (r.41), approach (r.42),and intensity (r.44). This mother-observer discrepancy was not related tothe observer ratings of infant behavior observed in the home (allcorrelations less than .11). Furthermore, to insure that the discrepancyscore correlations were not driven by the level of negativity in the motherratings in the laboratory, this analysis was repeated covarying the averagenegativity scores. All partial correlations of discrepancy with motherreports remained significant (partial correlations between .20 and .33).These data support the conclusions that (1) mothers can be accuratereporters of behavior when they are observing children unrelated to them,(2) this reporting accuracy breaks down when observing their own children,and (3) negative biases in laboratory reporting are associated withnegative ratings using more standard instruments. We speculate thesefindings reflect the fact that mothers are engaged in an importantrelationship with their young infants, which likely colors their ability tobe objective reporters of their behavior. Even though it may make ourlives as infancy researchers more difficult, we should always keep in mindthat mothers' own children are special to them.
poster
Individual differences in attentiveness have been observed in normalchildren at all stages of development. Ruff and Rothbart (1996) suggestthat the variability in attention among children may be a reflection oftemperamental differences. Temperamental dimensions that may influencethe allocation of attention include reactivity and regulation (Rothbart &Derryberry, 1981). Differences in reactivity, both behaviorally andphysiologically, can be observed from a very early age. However, with theemergence of inhibitory control processes at the end of the first year,reactive infants may become better able to regulate their behavior andattention in challenging situations. This study will examinelongitudinally individual differences in attention and how theseattentional differences may be related to temperamental differences. It ishypothesized that the relation between the behavioral measures ofreactivity and the ability to control attention and behavior duringchallenging situations will be mediated by the child's inhibitory skills.Thus, a highly reactive infant who acquires good inhibitory skills by 12to 18 months should be better able to control attention and behaviorduring challenging situations than a reactive infant who has not acquiredthese skills.A diverse sample of 150 children participating in a longitudinal studyexamining social and emotional development were observed in a laboratorysetting when the children were 6, 12, and 18 months of age. Attention wasassessed at 6 and 12 months of age by coding the intensity of interest,orientation and manipulation exhibited by the infants toward an object .At 18 months, the duration of orienting to a television screen displayingflashing pictures of different objects and scenery was coded.Physiological reactivity was assessed at each age by measuring heartperiod and respiratory sinus arrhythmia throughout the visit. Thesemeasures were used to assess physiological change (from a baseline measureof heart rate) during the challenging tasks. At all ages, emotionalreactivity and early regulatory behaviors were assessed in response tofrustrating situations (arm restraint and prohibition tasks). Also,toddlers' behavioral control is measured during clean-up and a delay taskat the 18-month asssessment. A composite of behaviors will be used toidentify children exhibiting good inhibitory control at 12 and 18 months.In addition, mothers filled out the Infant Behavior Questionnaire(Rothbart, 1981) at 6 and 12 months and the Toddler Behavior AssessmentQuestionnaire (Goldsmith, 1988) at 18 months, therefore, mothers' reportsof their children's attention, reactivity and regulatory skills will beexamined in conjunction with the laboratory measures.Preliminary analyses revealed moderate stability of attention in thelaboratory from 6 to 12 months. In addition, the laboratory measures ofattention at 6 and 12 months were significantly related to the mothers'reports of positive affect and soothability. Moderate stability is alsoexpected in physiological reactivity across the three ages. Additionalrelations between maternal report and laboratory measures will beexplored. Results will be discussed within a larger framework oftemperament and its developing impact on behavioral regulation.
poster
Recent research in the area of infant temperament has focused on three methodologies for measuring child behavior: parent report, home observation, and laboratory observation. One potential difficulty with choosing among these measures is that there is likely to be only modest consistency across measures. At least one recent study (Kochanska, Tjebkes, & Forman, 1998) reports little to no correlation between a standard temperament questionnaire and the laboratory procedures (LAB-TAB, Goldsmith & Rothbart, 1993) that were derived from it. Several factors may affect the concordance between such measures. For example, maternal perception is thought to consist of both an objective component and subjective component (Bates & Bayles, 1984). The subjective component may be influenced by the mothersD5 psychological state, or life experiences. The goal of this study was to examine multiple factors that may affect maternal perception of infant temperament and compare the relations between these factors and both maternal perception and laboratory observations of temperament. A second goal was to compare the concordance between the two types of measures as a function of these subjective factors.Subjects in the study were 326 six-month-old infants (173 boys) who were assessed in a laboratory paradigm designed to elicit temperamental distress to limitations. Using two tasks (a barrier task and an arm restraint task), measures of distress were calculated. Infant behaviors were rated by mothers using the Infant Behavior Questionnaire (Rothbart, 1981). In addition, several self-report measures of maternal functioning were collected, including the Parenting Stress Index (PSI; Abidin, 1983); the Symptom Checklist 90 - Revised (SCL 90-R, Derogatis, 1986) and the Dyadic Adjustment Scale (DAS, Spanier, 1976). Data on marital status, maternal age, socioeconomic status, race, child birth order and child sex were also collected.Two sets of analyses were collected. First, the factors that might affect maternal perception were correlated with both maternal perception and laboratory observations of distress to limitations. These analyses indicated that depression, hostility, parenting stress, socioeconomic status, and maternal age were all correlated with maternal perception of temperament. However, only the measure of stress associated with having a difficult child was correlated with the laboratory measure of distress to limitations. A second set of analyses examined the concordance of laboratory assessment and parent perception of temperament as a function of several factors that affected maternal perception. These analyses indicated greater concordance between the two temperament measures when mothers were married, rating their first born infants, rating girls, experiencing less stress and reporting less hostility. These results suggest that maternal reports of infant temperament should be considered within the context of sociodemographic and psychological factors, and that assessment of such factors may improve both rates of concordance across multiple measures and predictability of infant behavior.
poster
In the infant temperament literature, there is some evidence that two patterns of negative affectivity are observable. Fox and Calkins demonstrated that the tendency to display these temperamental traits (they labelled them fearful reactivity and frustrated reactivity) are linked to different patterns of physiological reactivity and different behavioral outcomes. The pattern of behavior marked by fear is linked to inhibited behavior (crying, clinging, withdrawal in the face of novelty), while frustrated reactivity is linked to uninhibited behavior (Calkins & Fox, 1992; Fox, 1989). These two types of negative behavior clearly differ in origin, and may lead to very different behavioral outcomes. Data from Kagan and colleagues and Calkins and Fox suggest that the outcome for fearful infants is likely to be in the direction of internalizing problems (Calkins, Fox, Rubin, & Coplan, 1995; Kagan et al., 1988). Other researchers have suggested that poorly regulated anger/frustrat!ion plays a role in the development of hostile aggression (Derryberry & Reed; 1994; Eisenberg et al., 1996; Fox & Calkins, 1993; Zahn-Waxler, 1993). The goal of the present study was to describe the physiological, emotional, and attentional characteristics of highly frustrated infants, characteristics that may predispose these infants to later difficulties.A sample of 162 six-month-old children was selected on the basis of parents' report of their infants' temperament and a laboratory assessment of temperament. Infants were classified as frustrated if they were both frustrated in the laboratory and were rated by mothers as high on the distress to limits scale on the Infant Behavior Questionnaire (Rothbart, 1981). Infants were classified as non frustrated if they were both low on frustration in the laboratory assessment and low on the IBQ scale of distress to limitations. These two groups of infants were compared on a number of physiological, attentional and emotional measures. Physiological reactivity and regulation were assessed using a baseline measure of respiratory sinus arrythmia (RSA, measured using Porges' vagal tone algorithm) and suppression of vagal tone during an attention demanding task. Attention was measured using an adaptation of a task by Kochanska and colleagues (Kochanska, Tjebkes, & Forman, 1998). Emotion!al regulation was assessed during the frustration tasks. Emotional responsivity to others was assessed using an empathy task. Results indicated that male and female infants were equally likely to be classified as frustrated and non-frustrated. Infants classified as frustrated were more reactive physiologically (p < .05) and less able to regulate physiological reactivity than their non-frustrated counterparts (p < .05), as evidenced by their high baseline vagal tone and their lack of vagal suppression during the attention-demanding task (see Figure 1). Frustrated infants were less attentive, less responsive to the emotions of others, and used different emotion regulation strategies than non-frustrated infants (p's, < .05). These infants were also classified by their mothers as more active, less attentive, and more distressed to novelty. The discussion of these findings will focus on whether this cluster of characteristics may constitute a unique temperamental type and why these associated characteristics may place the infant at risk for later adjustment difficulties.
poster
Transition to parenthood is a critical life event that increasesemotional distress among parents, which, in turn, may affect thedeveloping relationship with the infant. Increases in perceivedstress, daily hassles, and depressive tendencies have already beenindicated. Adult attachment style has been suggested to be animportant stress precipitating individual characteristic, withinsecure attachment acting as a vulnerability factor while secureattachment as protective. On the other hand, the infants perceivedtemperamental difficulty vs. easiness may also be a source of stressfor the parent. Furthermore, given that adult attachment containinternal models of both self and others, there may be differences inhow secure and insecure parents perceive their infants temperament,this, in turn, affecting emotional distress. Thus, differentialperception of infant temperament may moderate the relation betweenadult attachment and emotional distress. Given the lack of studiesfocusing on the joint relevance of adult attachment and infanttemperament on parental emotional distress, this was addressed in thepresent study. The subjects were 482 parents (307 mothers and 175fathers) of 5-7 mo old infants. The measures comprisedself-questionnaires on attachment (Relationship Questionnaire;Bartholomew & Horowitz, 1991), infant temperament (Infant BehaviorQuestionnaire; Rothbart, 1981), and standardized questionnaires onperceived stress, vital exhaustion, depressive tendencies, and maritalsatisfaction. The results indicated that fearful attachment and perceived infantdistress proneness were related to parental emotional distress.Further, among mothers, anxious-ambivalent attachment was positivelywhile secure attachment negatively related to parental distress. Also,different attachment styles varied according to perceived infanttemperament with fearful mothers reporting highest levels of infantdistress to novelty, while secure mothers and fathers the highestlevels of infant soothability and smiling, respectively. Differentlevels of infant soothability, smiling, activity, orientation anddistress to novelty were also found to moderate the relation betweendifferent adult attachment dimensions and emotional distress.This study demonstrates the importance of considering both theintrapersonal parental attachment styles as well as perceived infanttemperament as related to the level of parental emotional distress.
poster
Despite increasing interest in the development of structuredobservational approaches to the assessment of temperament in infancyand early childhood (Goldsmith & Rothbart, 1991), few studies haveexamined the predictors of dimensions of temperament derived fromobservational assessments. This study was designed to: (1) examinethe stability of dimensions of temperament derived from observationalassessments across the span from 6 months to 3 years, and (2) examinethe associations between observations of temperamental qualities ininfancy and early childhood and measures of parental personality. At 6 months of age, infants were videotaped in episodes ofthe Laboratory Temperament Assessment Battery (Lab-TAB, Goldsmith &Rothbart, 1988) designed to elicit anger proneness, fear, and joy. Inaddition, parents completed the Multidimensional PersonalityQuestionnaire (Tellegen, 1982) and the Infant Behavior Questionnaire(IBQ; Rothbart, 1978). At age 3, 45 of the children seen at 6 monthswere videotaped in episodes of the Preschool Lab-TAB (Goldsmith &Reilly, 1995) designed to elicit fear, distress,interest/persistence, and positive affect. At both time points,videotapes were scored using global rating scales adapted fromGoldsmith & Reilly (1995). Preliminary analyses suggested that observed indicators ofpositive affect, shyness, and negative affect exhibited modest tomoderate levels of stability across the period from 6 months to 3years (r's ranging from .25 to .40, p's < .05). Althoughparent-reports of infant temperament predicted conceptually similardimensions of observed infant temperament, there were fewassociations between reported infant temperament and observedtemperament at age 3. There were relatively few associations between parentalpersonality and dimensions of infant temperament; however, parentalpersonality, particularly maternal personality, assessed whenchildren were infants, emerged as a significant and meaningfulpredictor of observed temperament at age 3. For example, higherlevels of maternal harm avoidance (fearfulness) were associated withhigher levels of fearfulness (r .42, p< .01) and lower levels ofboldness (r .25, p < .05) in children at age 3. Maternalaggression, stress reactivity, and alienation (measures of negativeaffectivity) were positively associated with negative affect/distressat age 3 (r's ranged from .31 to .49, p's ranged from .05 to .001);maternal aggression and alienation also predicted undercontrolledbehavior (i.e., hyperactivity and impulsivity) at age 3 (r's between.35 and .51, p's < .01). Maternal social potency positively predictedchildren's positive affect at age 3 (r .26, p < .05). Fewerassociations were found between dimensions of paternal personalityand child temperament at 3 years. Higher levels of paternal socialcloseness were associated with higher levels of positive affect (r .30, p < .05) and lower levels of child fearfulness and negativeaffect/distress at age 3 (r's -.29 and -.42, respectively, p's <.05 and .01, respectively), and paternal aggression was positivelyassociated with undercontrolled behavior (r .37, p < .01) at age 3.Hierarchical regression analyses suggested that parental personalityrepresents a significant predictor of observed temperament at 3years, even when the effects of infant temperament are controlled. Results are discussed in the context of previous research onthe stability of temperament in early childhood and the role ofparental personality in the expression of temperamental qualities inearly development.
poster
This investigation identified individual differences incontextually-inappropriate emotional responses (namely fearful responses incontexts designed to elicit pleasure). Emotions organize behavioraccording to relevant cues provided in the environment (Frijda, 1984,1996). A dysregulated emotional response pattern associated with thetendency to respond with contextually-inappropriate emotion may be avulnerability for maladaptive functioning and some forms of psychopathology(Cole, Michel, & O'Donnell, 1994). Support for this endeavor may be foundin recent research identifying individual differences in rhesus macaquedefensive responding associated with out-of-context displays of fearfulbehavior (Kalin and Shelton, in press). Research on human individualdifferences in this emotional response style has not been conducted. Thisproject assessed the temperamental domains that might be associated with aresponse style of contextually-inappropriate affect. Participants were140 12 month-old twin pairs from a larger, longitudinal study on emotional,social, and cognitive development. Negative facial affect, body behaviors,and distress vocalizations were rated for three observational measures oftemperament from the Laboratory Temperament Assessment Battery (LabTAB;Goldsmith and Rothbart, 1996) - Stranger Approach, Modified Peek-a-Boogame, and Puppet Game. Maternal report of temperament was assessed withthe Infant Behavior Questionnaire (IBQ; Rothbart, 1981). The resultssupport the hypothesized relationship between expression of fear-relatedbody behaviors during the pleasure-eliciting episodes and higher scores onthe IBQ Fear scale. Infants expressing any bodily fear within the ModifiedPeek-a-boo game scored higher on the IBQ Fear scale than those expressingno bodily fear (p<.05). Infants who showed extreme bodily fear (more than75% of the time) showed a trend (p<.10) to have higher IBQ fear scores thaninfants who showed no bodily fear all during Modified Peek-a-boo. Whenanalyses were conducted to compare individuals who expressed any bodilyfear in both the Modified Peek-a-Boo game and the Puppet Game with thosewho did not express any bodily fear during either episode, the former grouphad higher scores on the Fear scale (p<.05) and showed a trend to scorehigher on the Soothability scale (p.07). The temperamental domains ofactivity level and soothability were also related to this profile. Whencomparing groups according to the presence of bodily fear during the Puppetgame, infants who expressed any bodily fear scored higher on theSoothability scale (p<.05). Planned comparisons of extreme bodily fearversus no bodily fear expressed during the Puppet game showed the extremegroup scoring lower on Activity Level (p<.05). Additional comparisons wereconducted to analyze these 'no fear' and 'extreme fear' groups asqualitatively different types by limiting the comparison to individualswithin these two groups. For the Modified Peek-a-Boo game, the extremebodily fear group scored lower on the Activity Level scale (p<.05) andhigher on the Soothability scale (p<.05). Those who expressed extremelevels of bodily fear during the Puppets game had lower scores on theActivity Level scale as well (p<.05). Findings with a second samplereplicated the relationship between low activity level and expression ofbodily fear during the Puppet game (p<.05). Future analyses will extendinquiry into the physiological and observed behavioral correlates of thisemotional response style.
poster
The purpose of this thesis was to examine the development, structure, andcardiac physiology of behavioral approach and inhibition in children agedsix months to two years. Three goals were addressed to serve this purpose. The first goal was to search for early forms of an adult personalityconstruct, sensation seeking, as it exists in toddlers. A large battery ofbehavioral variables indexing this trait in 2-year-olds proved to beinternally consistent. Furthermore, factor analysis of these variablesdiscerned distinct forms of sensation seeking which corresponded roughly tothree of the four subdimensions of the trait found in adult investigations(thrill and adventure seeking, social disinhibition, and experienceseeking) as well as a factor which indicated a propensity to expresspositive affect. The second aim of this inquiry was to test the proposition that behavioralapproach and behavioral inhibition comprise two separable dimensions. Thecontention that approach tendencies appear prior to 6 months, whereasinhibition shows substantial development over the second half of the firstyear was supported, as latencies to reach for high- and low-intensity toyswere related to one another at 6 months of age, but not at 12 months. Alsosupporting a distinction between approach and withdrawal was factoranalysis of 2-year behaviors resulting in one factor comprised of variablesincluding positive affect scores and exploration of novel surroundings anda second factor which included negative reactions, particularly in responseto highly intense laboratory procedures. Indicating stability ofbehavioral approach, fast latencies to grasp toys at 12 months werestrongly predictive of high approach and low inhibition at 2 years. The third goal was to explore relations between cardiac physiology andsensation seeking/approach behavior. Consistent with expectations, highlevels of approach behavior were related to tendencies to respond to anambiguous auditory stimuli with a cardiac deceleration rather than acardiac acceleration. In addition, low cardiac vagal tone and highbaseline heart rate were associated with high inhibition in social situations.