Behavioral thresholds for tones during infancy

An observer-based psychoacoustic procedure for use with young infants.

“…Approximate thresholds for 6-montholds relative to the average thresholds of two aduits tested by Trehub et al were 20 dB at 400 Hz, 23 dB at 1000 Hz, and 15 dB at 4000 Hz. We have not included data from Berg and Smith's (1983) study in which sounds were presented free field because no data were collected from 6-month-olds in these conditions. Although it is difficult to make any firm statement about the sources of this difference, two rather obvious possibilities should be considered.…”

Developmental perspectives on the localization and detection of auditory signals

“…The head-turning or localization response can be elicited by 5 or 6 months of age (Chun, Pawsat, & Forster, 1960), and reinforcement can maintain it at high levels for 30-40 trials within a single session (Moore, Thompson, & Thompson, 1975;Suzuki & Ogiba, 1961;. Indeed, the robust nature of this technique and its applicability in the early years have made it the method of choice for studying absolute thresholds (Berg & Smith, 1983;Olsho, 1985;Schneider, Trehub, & Bull, 1980;Sinnott, Pisoni, & Aslin, 1983;Trehub, Bull, Schneider, & Morrongiello, 1986;Trehub, Schneider, & Endman, 1980;Trehub, Schneider, Morrongiello, & Thorpe, 1989), masked thresholds Nozza & Wilson, 1984;Schneider, Trehub, Morrongiello, & Thorpe, 1989;, critical bands (Olsho, 1985;Schneider, Morrongiello, & Trehub, 1990), and binaural unmasking (Nozza, 1987;Schneider, Bull, & Trehub, 1988) in infants and young children.…”

“…Control or no-signal trials are interspersed with signal trials to ensure that the infant is turning to the signal, not merely to the locus of interesting visual events. Adaptive procedures are typically used, and infants who tum on "too many" no-signal trials are often eliminated, although the number of such false alarms and other exclusion criteria vary considerably across studies (Berg & Smith, 1983: greater than 0.33 false-alarm rate; Sinnott et al, 1983: 2 or more false alarms).…”

“…From this perspective, one might expect all infant data gathered in this manner to be excluded on the basis of false-alarm and related criteria (e.g., not responding on selected probe or suprathreshold trials). Indeed, Berg and Smith (1983) excluded 36% of their lO-month-olds and 54% of their 14-month-olds, whereas Sinnott et al (1983) excluded 41% of all test sessions with 7-to l l-montholds. Perhaps another aspect of the G/NG-D procedure helps keep the false-alarm rate in check-namely, the experimenter's active manipulation of toys in front of the infant.…”

“…Perhaps another aspect of the G/NG-D procedure helps keep the false-alarm rate in check-namely, the experimenter's active manipulation of toys in front of the infant. In fact, Berg and Smith (1983) describe the experimenter as being an entertainer (pp. 412-413).…”