Infants: Research Techniques

Research with infants requires different techniques than research with older children and adults. Explanations of some research methods and terminology used in infant research follow.

• Infant head-turning technique. Infants tend to look toward a sound source, especially when that sound is novel or new. This research technique capitalizes on that natural tendency. A repetitive sound stimulus is presented from one side, and the infant is trained through reinforcement to turn the head toward the sound when a change in the stimulus occurs. The stimulus change utilized for training is usually a decibel level change. The reinforcement is usually the illumination and activation of animated toys. The animated toys are illuminated and activated when the head is turned in response to changes in decibel level, but are not activated when the head is turned in absence of this change. Once the infants are trained, this method is used to determine whether infants recognize other changes in the sound source.

• Visual attending technique. This technique was developed at the Kansas Infant Research Laboratory (Tims, 1977). A visual stimulus known to be favored by infants (e. g., a black-and-white checkerboard pattern) is projected in front of the infant. The infant controls the presence of the stimulus by looking at it. When the infant looks away, the visual stimulus is removed. Once this behavior has been learned (habituated), a musical stimulus is added to the visual one. When the infant looks away, both the visual and musical stimuli are removed. This allows the infant to control how long the music plays. The researcher records the amount of time the infant continues to look at the visual stimulus.
  
  Research at the laboratory indicates that changes in a musical stimulus cause the child to look at a visual stimulus again once this behavior has been habituated. This allows the researcher to determine whether the infant perceives changes in musical stimuli.
   
• Behavioral state. The behavioral state of infants is usually described in terms of a four-point scale:
  —Quiet sleep (eyes closed, body relaxed, movement generally absent)
  —Active sleep (eyes closed, frequent arm and leg movements, smiles and sucking may be observed)
  —Alert (eyes open, some activity is observed)
  —Crying (eyes open or closed, gross activity or little movement with overall tension, irregular breathing, marked color change, startles)
  
  Early scales split "alert" into two categories labeled "inactive alert" and "active alert." However, the scale was revised to the four-point version because agreement among observers is strongest with only one "alert" category. Children are least responsive to sound during "quiet sleep" and most responsive when "alert."
   

Infants: Vocalization and Singing

The Pitch Range and Contour of Infant Vocalizations

Fox, D. B. (1982). [CD-ROM]. Abstract from: ProQuest File: Dissertation Abstracts Item: 8300247

Research Design:
Qualitative Study
N = 12

Purpose:
The purpose of this study was to determine pitch range and contour characteristics of infant vocalizations from three months to nine months of age.

Procedure:
The infants were individually observed, and their vocalizations were recorded in their homes during seven sessions. Three representative samples from each session were analyzed using analog-to-digital conversion and computer graphics display procedures.

Results:
The composite range was G³ to C⁶. Eighty-two percent of the contours were descending.

An Analysis of the Characteristics of Infant-Child Singing Expressions

Ries, N. L. L. (1982). [CD-ROM]. Abstract from ProQuest File: Dissertation Abstracts Item: 8223568. A review by Rudolf E. Radocy appears in the Bulletin of the Council for Research in Music Education, 80, (Fall 1994), 83–90.

Research Design:
Qualitative
N = 51

Purpose:
The purpose of this study was to analyze the singing of very young children (aged seven to thirty-two months) for the following:

• length and style of vocalization
• range and frequency of pitches
• ascending and descending intervals
• use of rhythm and meter
• tonality.
   

Procedure:
Fifty-one children ranging in age from six months twenty-eight days to thirty-two months, were videotaped for two days in their homes while parents and siblings encouraged the children to attempt three types of singing: 1) familiar "standard" songs, 2) spontaneous songs, and 3) matching isolated pitches or intervals. The findings were grouped into four general age categories: seven months, eleven months, nineteen months, and thirty months.

Results:
Spontaneous songs such as sleep-songs and game-songs, snatches of standard songs, and occasional matched pitches were sung by a few seven-month-olds. By thirty months, 75% sang spontaneous songs, 83% sang standard songs with varying levels of success, and 17% sang whole songs (with elisions) accurately and unassisted.

Overall, the average range was two octaves that began at A below middle C and up. The most prevalent notes were between middle C and up a fifth (C–G) for the two younger groups and D above middle C and up a fifth (D–A) for the two older groups.

Children under thirty months sang mostly descending intervals. At thirty months, unisons and ascending and descending intervals were used about equally. At every age level the most used interval was the major second (descending and ascending).

The younger children used simple rhythms, often with a two to one relationship. Brief segments of pulse and meter were present at eleven months, with pulse and meter becoming common (though the meter sometimes changed) by thirty months.

At all ages, children sang notes harmonically related to the given pitch, either in addition to or instead of the given pitch. In the pitch-matching tests, 91% of the seven-month-old children either matched pitches or sang notes related to the pitch harmonically at least 50% of the time. Ninety-two percent of the eleven-month-old children either matched pitches or sang notes related to the pitch harmonically at least 55% of the time. Seventy-five percent of the nineteen-month-old children either matched pitches or sang notes related to the pitch harmonically at least 70% of the time. The thirty-month-old children matched or sang notes that related harmonically to 83% of all the pitches sung in the pitch-matching tasks.

Infants: Listening, Perception, and Response

Prototypical Representations of Music Structure in Infancy: Theoretical Exploration and a Pilot Study

Lynch, M. P. (1993). Psychomusicology: A Journal of Research in Music Cognition, 12, (1), 31–40.

Research Design:
Descriptive
N = 12

Purpose:
The purpose of this study was:

• to determine whether Western six-month-old infants can discriminate harmonic music intervals
• to determine whether infants have formed prototypical representations of the perfect fifth.
   

Procedure:
Twelve infants participated in the study. None of their parents were musicians, and only Western music was listened to in their homes.

The infant head-turning technique was used. (To determine whether the children would turn their heads in response to a perceived change in the sound, the children had previously been tested for their response to changes in tuning accompanied by changes in decibel levels. All the infants performed head turns in this preliminary phase.)

During the testing procedure, the infant sat on the parent's lap playing with quiet, handheld toys. Equal-tempered harmonic perfect fifths (prototype) and intervals with the same lower note but the upper note raised in frequency by 2.5%, 3.5%, 4.5%, or 5%. Five percent were played through a loud speaker for thirty trials. One half of these trials included a change in the interval to a mistuned interval (raising of the higher note by 4.5%).

Results:
The infants' response to mistuned intervals was not significant, indicating that six-month-old infants cannot discriminate harmonic music intervals. In examining false negatives (lack of head turn in response to a change), it was observed that the tendency for 4.5% and 5.5% mistuned intervals to be perceived as equivalent to the prototype perfect fifth more often than to the mistuned perfect fifth approached significance. The author speculated that this may be due to the existence of prototypical representations of the perfect fifth in the infants.
 

Categorization and Conservation of Melody in Infants

Summers, E. K. (1984). [CD-ROM]. Abstract from: ProQuest File: Dissertation Abstracts Item: 8501103. Reviewed by Marvin Greenberg in the Bulletin of the Council for Research in Music Education, 94, Fall 1987, 69–71.

Research Design:
Descriptive
N = 38

Purpose:
The purpose of this study was to find behavioral evidence that six-month-old infants could categorize melodies by remembering (conserving) interval sequences when the melodies were transposed.

Procedure: Summers used the infant head-turning technique. Correct responses were a head turn to a change in the melody with no turn if there was no change in the melody. The author noted difficulties in the use of formal testing situations with very young children.

Results:
Ten of twenty-three infants discriminated a melody from its major-seventh transposition. Eleven of thirty-eight discriminated a melody from a variant that had different intervals but preserved the contour. The infants had difficulty discriminating two melodies with a contrasting contour that had the same pitch set. There was no difference in discrimination between close and distant transpositions of a learned melody. The author concluded that categorization and conservation can occur far earlier than previous much research suggests.
 

Infants' Perception of Melodies: The Role of Melodic Contour

Trehub, S., Bull, D., & Thorpe, L. A. (1984). Child Development, 55, 821–30.

Research Design:
Descriptive
Experiment 1: N = 92 (average age: nine months, fifteen days)
Experiment 2: N = 85 (average age: nine months, twelve days)

Purpose:
The overall purpose of this study was to determine what changes (transformations) can be made to a melody that is familiar to infants without resulting in the infants perceiving it as a new melody. The study consisted of two experiments.

Experiment 1. The purpose of the first experiment was:

• to determine whether the infant head-turn technique is viable in studies of melody discrimination in infants
• to replicate previous research on melody perception by infants
• to further explore the basis for the perception and recognition of melodies by infants.
   

Experiment 2. The purpose of the second experiment was to provide more detailed information regarding how infants process melodies by increasing the difficulty of the tasks in the first experiment.

Procedure:
Experiment 1. The infant head-turning technique was used (see description). The infant sat on the parent's lap during the testing procedure. One of three six-tone melodies (sinusoidal waveforms) were presented to the infant, paired with one of the transformations below. Except for the first one, the transformed melodies all began and ended on the same note as the original. The sequences were separated by 800 msec of silence. The transformations consisted of:

1. Transposition of the sequence up three semitones, from the key of C to the key of E-flat
2. Change of intervals in the melody while retaining its original shape (contour preserving)
3. Change of octave of some of the notes while retaining the note names and contour of the melody (octave change/contour preserving)
4. Change of octave and contour (octave change/contour violating). These experimental trials were alternated with no-change or control trials to measure spontaneous head turning. Twenty trials were completed, ten experimental and ten control, in random order.
    

Experiment 2. The tasks in Experiment 1 were made more difficult for the infants to perform:

• A distracter sequence of three tones (262–Hz / C₄)was inserted between the standard melody and the transformed melody.
• A fifth transformation was added consisting of change of intervals in the melody, which violated its contour (contour violating). The first and last notes of the melody remained unchanged from the original (as in the other transformations).
• All other procedures remained the same as in Experiment 1.
   

Results:
Experiment 1. There were two major results from Experiment 1:

• The head-turning technique was found to be effective in studying melody discrimination.
• The infants perceived the changes in every transformation of the melody; all of the transformations were seen as different from the original melody. However, recognition of the transposed melody and the contour-preserving melody was significantly more difficult for the infants than recognizing the other transformations was. Their recognition of the two octave-change transformations did not differ significantly.
   

Experiment 2. Infants were unable to perceive the transposition and contour-preserving transformations (these were equally difficult). The contour-violating transformation was significantly easier for them to perceive than the transposition and contour-preserving transformations, but significantly more difficult than the two octave-change transformations. The ability of the infants to recognize the two octave-change transformations was not significantly different. The authors concluded that when infants hear a new melody, they do not remember absolute pitches. They perceive a melody as familiar if the contour and range is the same as a melody heard previously. They perceive it as a different melody if the contour or the range is different. Contour and range are remembered and perceived, even when the absolute frequency or the sizes of the intervals is changed.

Infants' Perception of Melodies: Changes in a Single Tone 

Trehub, S., Thorpe, L. A., & Morrongiello, B. A. (1985). Infant Behavior and Development, 8, 213–23.

Research Design:
Descriptive
Experiment 1: N = 30 (average age: seven months, three days)
Experiment 2: N = 30 (average age: seven months, one day)

Purpose:
The purpose of these experiments was to determine whether infants can recognize the change in the frequency of a single tone in a short tonal melody.

Procedure:
Experiment 1. The infant head-turning technique was used. The infant sat on the parent's lap during the testing procedure. One of three six-tone melodies (sinusoidal waveforms) were presented to the infant, paired with one of the transformations listed below. The melodies were separated by 1500 ms of silence. .

• The frequency of the first note is changed.
• The frequency of the second note is changed.
• The frequency of the third note is changed.
• The frequency of the fourth note is changed.
• The frequency of the fifth note is changed.
• The frequency of the sixth note is changed.
   

These experimental trials were alternated with no-change or control trials to measure spontaneous head-turning. Thirty-six trials were completed, including eighteen experimental trials (one for each transformation) and eighteen control trials.

Experiment 2. A distracter sequence of three tones (262-Hz /C₄) with 900 ms of silence before and after was inserted between the standard melody and the transformed melody. Other aspects of Experiment 2 remained the same as in Experiment 1.

Results:
Experiment 1. Infants were able to recognize the change in the melody when a single tone was changed in any position—they turned their heads significantly more often in response to the change trials than they did during the control trials for all transformations. Transformations that extended the range (the new note was outside the range of the original melody) were significantly easier for the infants to identify as different.

Experiment 2. Infants were able to recognize the change in the melody when a single tone was changed in any position—they turned their heads significantly more often in response to the change trials than they did during the control trials for all transformations. Unlike Experiment 1, transformations that extended the range were not significantly easier for the infants to identify than other transformations.

Infants had significantly more difficulty recognizing the transformation in the first position compared to the third, fourth, fifth, and sixth positions. The difficulty of transformation in the first position compared to that in the second position approached significance.
 

Infant's Perception of Musical Patterns

Trehub, S. (1987). Perception and Psychophysics, 41, (6), 635–41.

This article reviews research about infants' perception of musical patterns, including sections on "melodic processing: contour," "melodic processing: diatonic structure," "temporal processing: rhythm," and "temporal processing: stream segregation." In summary, research reviewed in this article indicates that:

• Infants are able to analyze complex combinations of sounds.
• Infants can recognize when melodies are different from one another, particularly when they are presented immediately one after the other.
• Infants can accurately perceive pitches, implied by their ability to imitate sung pitches after limited training.
• Infants respond to some aspects of underlying musical structure—they more readily detect changes in diatonic melodies than in nondiatonic ones, and they more easily recognize sets of pitches that include relationships embodied in the major triad. (It is not known whether this is because of the ratio relationships between the tones, their relationship to the natural overtone structure, or some degree of familiarity with Western music and its diatonic structure.)
• Infants appear to group patterns of notes on the basis of similar frequency or similar overtone structure.
• Infants recognize differences in rhythm, even when tempos vary.
   

Infants: Effects of Home Musical Environments

A Study of Infant Musical Productivity 

Kelly, L. & Smith, B. S. (1987). In J. C. Peery, I. W. Peery & T. W. Draper (Eds.), Music and child development (pp. 35–53). New York: Springer-Verlag.

Research Design:
Qualitative
Case Study
N = 3

Purpose:
The purpose of this study was to observe three children in three different family situations—one with parents who were professional musicians, one with parents who were musically oriented but not musicians by profession, and one with parents who were not musically oriented. The study sought to:

• determine how musicality was differently expressed in these families
• identify the distinctive features of the children's early musical response and performance
• determine the immediate context signaling the beginning and end of musical behavior.
   

Procedure:
Each of the three children, in different family situations as noted above, was observed weekly for varying lengths of time ranging from fifteen minute to half-day sessions. Each family was also interviewed about the family's musical history, ways that musicality was transferred from one generation to the next, and favorite songs, records, instruments, and occasions for musical behavior in the child. This article reports one of three case studies.

Results:
The differences among the three children suggest a strong relationship between musical experiences at home and the speed and path of the child's musical development. The children from the two musical homes began their musical expression much earlier (six months and ten months), starting with melody, then proceeding to rhythm and words; the child from the nonmusical home began with words (twenty to twenty-one months) and then incorporated rhythm, never singing very much. The children from the two musical homes produced clearly distinguished pitches prior to age two, whereas the child from the nonmusical home did not—tones were more spoken than sung.

The musical environment observed in the family with parents who are professional musicians included the mother singing to the infant beginning at one week of age (including lullabies), rocking the baby, asking the baby to sing, placing the child near the piano so child can play, the father playing piano while the child listens or plays one finger at a time, and the parents placing a guitar or Autoharp on the floor for the child to play with (at nine months old). These parents also played records and the radio as well as rehearsed and performed in proximity to the child.

The musical environment of the family with parents who had an avocational interest in music included the mother singing to infant (including lullabies) at one month, playing a music box, playing a piano while the child listened and sang along, and infrequently offering a toy xylophone. These parents played records and the radio; the child also listened and sang along in church.

The "nonmusical" environment included no parental musical behaviors except for infrequent tambourine, the introduction of a guitar by the observer at twenty-four months, the playing of records, and watching TV (Mickey Mouse and Sesame Street).

Infants: Responses to Different Types of Music

Contrasting Music Conditions, Visual Attending Behavior, and State in Eight-Week-Old Infants

Tims, F. C. (1981). Bulletin of the Council for Research in Music Education, Nos. 66–67 (Spring/Summer), 164–68.

Research Design:
Descriptive
N = 36 (average age eight weeks and two days)
Six experimental groups

Purpose:
The purpose of this study was to examine how different types of music could cause infants to exhibit visual attending behavior, to determine whether infants are able to discriminate among types of music, and to observe the effects different types of music have on the infants' behavioral state.

Procedure:
Thirty-six infants participated in the study—twenty-four experimental subjects and twelve control subjects. Three types of music were played successively for the children as indicated below, and data was collected on visual attending and behavioral state (crying and sleeping). The types of music were:

• sedative (traditional lullabies from a variety of cultures performed on the piano)
• traditional stimulative (the Vivace of Copland's "Piano Sonata")
• intermittent/nontraditional stimulative ("Agony," an electronic piece by Mimaroglu).
   

Results:
The children looked more often, demonstrated more discrimination, and cried and slept more in response to the sedative music. The author attributed this to the possibility that sedative music was easier for the infants to process. There was no observed difference in infants' response to stimulative and intermittent music types.