The use of time-expanded speech as an aid in the diagnosis of articulation disorders.

The recordings of 20 speech defective speakers' readings of a standard prose passage were presented to 36 student speech clinicians under three listening conditions: (1) unaltered; (2) time expanded to 150% of the original recording time; and (3) time-expanded to 200% of the original recording time. Results of their evaluations indicate that the expanded versions of the recordings elicited more accurate evaluations by the student clinicians. Implications of these findings concerning the possible usefulness of time-expanded speech as an aid in the training of listening skills in student speech clinicians are discussed.

Clinician bias: the effects of pretesting information on the evaluations of speech clinicians.

The purpose of this investigation was to determine if pretesting information has a biasing effect on the speech evaluations of student speech clinicians. A total of three groups of clinicians evaluated 10 speech characteristics of each of 17 recorded speech samples. The clinicians participated in two listening sessions. In the first (control) session, they evaluated the speech samples using a 4-point scale of defectiveness and were given no background information on the speakers. In the second (experimental) session, they were given fabricated case history information prior to their evaluations. Results of their ratings indicate that the pretesting information had a biasing effect on the clinicians' evaluations of the speech samples. Implications of these findings and suggestions for future inquiry are discussed.

Listeners' perceptions of nonspeech characteristics of normal and dysarthric children.

The purpose of this investigation was to compare listeners' evaluations of nonspeech characteristics of dysarthric and normal speakers. A total of 16 speakers, eight cerebral-palsied and eight normal children ranging in age from 6 to 11 years, recorded six three-word phrases. The cerebral-palsied speakers were judged to have disorders in at least two of the three areas of articulation, phonation, and resonance, and all were judged to have an articulation disorder as one of the two disordered areas. All normal speakers were found to have normal speech characteristics. A master tape containing the 16 speakers' recordings was presented to 25 listeners, who were asked to rate each of the speakers on a 24-item semantic differential scale that contained polarized adjective pairs concerned with personality and physical-appearance traits. Results of their ratings indicated that for all but two of the 24 adjective pairs, the normal speakers were judged more favorably than the dysarthric speakers, and for all but one of the 24 adjective pairs there were significant differences in judges' ratings between the normal and dysarthric speakers. It appears, therefore, that dysarthric speech adversely affected listeners' perceptions of the dysarthric speakers' personality and physical-appearance characteristics.

A comparative study of adolescents' perceptions of normal-speaking and dysarthric children.

The purpose of this investigation was to compare adolescents' perceptions of the nonspeech characteristics of dysarthric and normal speakers. Recordings of six three-word phrases produced by 16 speakers, eight cerebral palsied and eight normal-speaking children, were presented to 19 students for judgments of nonspeech characteristics on a semantic differential scale containing 22 bipolar adjective pairs. Results of their ratings indicated that for all 22 adjective pairs the normal speakers were judged more positively than the dysarthric speakers. Moreover, the differences in listeners' ratings between the two speaker groups were statistically significant for all adjective pairs. Implications of these findings and their comparison to the judgments of peer and adult listeners are discussed.

Adolescents' perceptions of normal and voice-disordered children.

The purpose of this investigation was to compare adolescents' perceptions of the nonspeech characteristics of voice-disordered and normal speakers. Recordings of six three-word phrases produced by 16 speakers, eight voice-disordered and eight normal-speaking children, were presented to 19 adolescent students for judgments of nonspeech characteristics on a semantic differential scale containing 22 bipolar adjective pairs. Results of their ratings indicated that for 15 of the 22 pairs (68.2%) the normal speakers were judged more positively than the voice-disordered speakers. Moreover, for 9 of the 22 pairs (40.9%) the differences in listeners' ratings between the two speaker groups were statistically significant. Implications of these findings concerning developmental trends in the perception of voice disorders and their potential impact on mainstreaming in schools are discussed.

The use of time-expanded speech in judgments of hypernasality.

Twenty-five samples of various degrees of hypernasal speech were time altered (expanded) to 0%, 125%, 175% of the original time. Inexperienced and experienced student clinicians rated the magnitude of hypernasality and degree of confidence of ratings for each sample in each condition. The results indicate significant differences in ratings of hypernasality for the expansion conditions for both groups. Experienced clinicians showed less variability and higher mean ratings than inexperienced clinicians.

A normative study of children's performance on the short form of the Token Test.

Spellacy and Spreen's Short Form of the Token Test was administered to 240 normal middle-class children, 20 in each of 12 half-year age groups from 6.0-6.5 through 11.6-11.11. Results of their performance on this test indicate that there are only slight differences between any of the 12 age groups investigated. The findings of this investigation provide normative data on children's performance on the Short Form of the Token Test, information that should be very useful in screening procedures for assessment of receptive language skills in children.

Articulation improvement and oral tactile changes in children.

To determine whether oral tactile changes accompany improvement in articulation, 16 children were administered two tests of oral tactile perception and two tests of articulation competency at three time periods: before, during, and at the end of speech therapy. A positive relationship between articulation proficiency and oral tactile perception was established. Both articulation tests reflected a progressive improvement for the group across the three testing periods. One of the tests of oral tactile perception reflected the same progressive trend and the second test, while not exhibiting the identical trend, did reflect a general improvement from the first to the third testing period. These findings indicate that oral tactile skills appear to be important factors in speech production and that such skills should be assessed in cases of functional articulation disorders,

Effect of prior exposure to stimulus material on identification of speakers' sex, height, and weight.

The purpose of this series of experiments was to determine the effect of prior exposure to taped stimulus material on listeners' judgments of speakers' sex, height, and weight. Two series of listening experiments, one concerned with sex identification and one with height and weight identification, were conducted. Results indicate that prior exposure to stimulus material improved the accuracy of listeners' judgments of speakers' sex, height, and weight. However, they also indicate that it is not necessary to receive prior exposure to the complete set of stimuli to achieve accurate judgments. Prior exposure to the stimulus material improved listeners' confidence in their estimates of height and weight but not in their judgments of speakers' sex.

Correlational analysis of speakers' heights, weights, and temporal speech features.

Correlation coefficients on height, weight, and temporal features were computed separately for 15 female and 15 male speakers. With only one exception, speakers' heights and weights were not significantly correlated with their temporal characteristics. Implications of these findings and suggestions for future research are discussed.

Effect of vocal disguise on estimations of speakers' heights and weights.

The purpose of this investigation was to determine the effect of speakers' attempts to disguise their voices on listeners' accuracy in judgments of height and weight. A total of six sentences were recorded by 26 speakers, 13 females and 13 males, under five conditions: in a normal manner (control condition) and in a manner in which they attempted to sound much taller and much shorter, as well as much heavier and much lighter than they actually were. Five master tapes were constructed, a control tape and four disguise tapes. Each tape contained a total of 156 sentences arranged in random order. A total of 40 judges, 20 in the height-estimation experiment and 20 in the weight-estimation experiment, participated in three sessions, one for each of the three master tapes (one control and two disguise tapes) in each experiment. They were asked to judge the height and weight of the speaker of each sentence and, using a seven-point confidence rating scale, to indicate the over-all confidence in their judgments at the end of each session. Although a majority of speakers yielded height and weight estimates consistent with the intended disguise conditions, the differences in listeners' estimates among all conditions were relatively small. Moreover, differences between speakers' actual heights and weights and listeners' height and weight estimates in the normal (control) condition were also relatively small. Implications of these findings and suggestions for future research are discussed.

Effect of vocal disguise on judgments of speakers' sex and race.

To determine the effect of speakers' attempts to disguise their voices on listeners' accuracy in judgments of speakers' sex and race, 26 speakers, 13 women and 13 men, recorded six sentences under three conditions: (a) in a normal manner, (b) in a manner in which they attempted to sound like a member of the opposite sex, and (c) in a manner in which they attempted to sound like a member of the black race. Three master tapes were constructed, one for each of the three conditions. A total of 40 judges, 20 in an experiment on sex identification and 20 in one on race identification, participated in two sessions, one for each of two tapes (control and disguise) in each experiment. In each session they were asked to judge the sex or race of the speaker of each sentence and, using a seven-point confidence rating scale, to indicate the over-all confidence in their judgments at the end of each session. Analysis indicated that, although listeners' accuracy for sex and race identification was greater under the control than disguised conditions for the majority of speakers, the differences between the two conditions were relatively small. Implications of these findings and suggestions for future research are discussed.

Listeners' identification of environmental sounds.

To determine if listeners can accurately identify various environmental sounds, a total of 40 sounds, including animal, inanimate, musical, and human sounds, were recorded, randomly arranged on a master tape, and presented to 30 judges for identification. Analysis indicated that, in general, listeners can accurately identify environmental sounds. However, their accuracy was not equal for all four classes of sounds investigated: human sounds yielded the highest accuracy, followed by musical and inanimate sounds, and animal sounds were least accurately identified. Implications and suggestions for research are discussed.

Effect of listeners' comprehension on judgments of speakers' height and weight.

To determine the effect of listener's comprehension on speakers' height and weight judgments, a total of 40 speakers, 20 Americans (10 females and 10 males) and 20 non-Americans representing various nationalities (10 females and 10 males), recorded a standard prose passage. A total of three master tapes were constructed to represent the three experimental conditions: (1) American speakers' readings in English, (2) foreign speakers' reading in their native languages, and (3) foreign speakers' readings in English. A total of 30 judges, 15 females and 15 males, participated in three sessions, one for each of the three master tapes, and were asked to provide direct estimations of the height and weight of each speaker from the tapes. Listeners' comprehension does not appear to play a major role in height and weight judgmental tasks.

Listeners' identification of human-imitated animal sounds.

To determine whether listeners can accurately identify human-imitated animal sounds, 20 speakers (10 females and 10 males) recorded their imitations of cows, cats, dogs, pigs, and sheep. These recordings were randomly arranged on a master tape and presented to 30 judges for identification. Analysis indicates that listeners can accurately identify various human-imitated animal sounds. Implications and suggestions for future research are discussed.

Effect of vocal disguise on estimations of speakers' ages.

To determine the effect of speakers' attempts to disguise their voices on listeners' accuracy in age judgments, 26 speakers, 13 females and 13 males, recorded six sentences under three conditions: (a) in a normal manner, actually were, and (c) in a manner in which they attempted to sound much older than they actually were. Three master tapes were constructed, one for each of the three conditions. A total of 20 judges, 10 females and 10 males, participated in three sessions, one for each of the three master tapes. In each session they were asked to judge the age of the speaker of each sentence and, using a seven-point confidence rating scale, to indicate the over-all confidence in their judgments at the end of each session. Although a majority of speakers yielded age estimates consistent with the conditions of intended disguise, the differences in listeners' estimates among all three conditions were relatively small. Moreover, differences between speakers' actual ages and listeners' age estimates in the control condition were also small. Implications of the findings and suggestions for research are discussed.

Listeners' discrimination of real and human-imitated animal sounds.

To determine if listeners can accurately distinguish between real and human-imitated animal sounds, a total of 165 recorded sounds (55 real and 110 human-imitated) of cats, cows, dogs, pigs, and sheep were randomly arranged on a master tape and presented to 30 listeners for discriminative judgments. Results indicate that, in general, listeners can accurately discriminate real from human-imitated animal sounds. Suggestions for future research are discussed.