Concurrent and Predictive Validities of the Bayley Motor Scale and the Peabody Developmental Motor Scales

ROBERT J. PALISANO

Concurrent and predictive validities of the Bayley Motor Scale and the Peabody Developmental Motor Scales were examined by administering both tests to 23 full-term and 21 healthy premature infants at 12, 15, and 18 months of age. For both groups, a correlation analysis of age-equivalent scores indicated Bayley scores had good to high correlation with Peabody gross motor scores (range, r = .78 to r = .96) and unacceptable correlation with Peabody fine motor scores (range, r = .20 to r = .57). When results were reported using standardized quotients, mean Bayley quotients for the full-term infant group were significantly higher than Peabody gross motor quotients. Prediction of motor development at 18 months of age was limited (range, r = .25 to r = .60), with the exception of Peabody fine motor scores for the premature infant group (r = .75). This study provides evidence of concurrent validity between Bayley motor and Peabody gross motor age-equivalent scores, and it suggests the need for further testing, using separate assessments of gross motor and fine motor ability, to determine motor development at later ages.

Key Words: Child development, Growth, Infant, Motor skills.

Many pediatric physical therapists are involved in the evaluation of full-term and premature infants who are at risk because of a delay or disorder in motor development. An important aspect of the physical therapist's evaluation of the high-risk infant is the administration of motor assessments that provide infor­mation pertinent to determining devel­opmental status and making recom­mendations for management. When functional ability is documented, the re­sults of motor development assessments complement findings obtained from the evaluation of muscle tone, reflexes, righting reactions, and equilibrium re­sponses. Knowledge of an infant's func­tional ability is helpful particularly in interpreting the significance of question­able findings for muscle tone and auto­matic motor responses, results that may

be transient or may become less pro­nounced over time.1,2 The reliance on developmental testing to assist in mak­ing clinical decisions, however, entails the responsibility of selecting assessment methods that yield reliable and valid measurements of motor development.

Conclusions about an infant's motor development level or developmental quotient should be based only on the results of an assessment involving 1) a standardized administration procedure and 2) a normative sample that reflects the cultural background of the infants for whom the assessment is being used.3

The Motor Scale of the Bayley Scales of Infant Development4 is, perhaps, the most widely used standardized motor scale that was normalized using Ameri­can infants. Normalized between 1958 and 1962 on 1,262 infants aged 2 to 30 months, the Bayley Motor Scale con­tains 81 items scored on a pass-fail basis, 69 of which measure gross motor be­haviors. Testing takes 15 to 20 minutes to complete, and results are expressed either by a standardized psychomotor developmental index (PDI) or by an age-equivalent score. The Bayley Motor Scale has been used in the re­search of developmental outcomes for high-risk infants, particularly premature infants,5,6 and to measure the effects of treatment in infants with Down syn­

drome.7 Although widely used, the Bay-ley Motor Scale contains a small num­ber of items for each level of develop­ment and omits stages in the motor developmental sequence. For example, it contains no items for running or kick­ing, and a single item incorporates all methods of prewalking progression (eg, creeping, crawling, hitching on but­tocks). The Bayley Motor Scale, there­fore, does not provide an in-depth mo­tor assessment nor does it delineate gross motor and fine motor development.

In contrast to the Bayley Motor Scale, the revised Peabody Developmental Motor Scales8 have been standardized recently on a nationwide sample of 617 infants and children aged 1 through 83 months and consist of separate Gross Motor and Fine Motor Scales, allowing for a more detailed assessment of motor development. The Gross Motor Scale is divided into five skill categories (reflexes, balance, nonlocomotor, loco­motor, and receipt and propulsion of objects), and the Fine Motor Scale is divided into four skill categories (grasp­ing, hand use, eye-hand coordination, and manual dexterity). The Gross Mo­tor Scale contains 10 items at each age level, and the Fine Motor Scale contains either 6 or 8 items at each age level. Test items are scored on a 0- to 2-point scale with a score of 1 indicating partial suc-

Dr. Palisano is Assistant Professor, Program in Physical Therapy, Hahnemann University, Broad and Vine, Philadelphia, PA 19101-1192 (USA). This study was conducted while he was a doctoral student at Sargent College of Allied Health Profes­sions, Boston University, Boston, MA.

This study was supported in part by Project No. 901, Division of Maternal and Child Health Serv­ices, Department of Health and Human Services, and by an educational grant from the Foundation for Physical Therapy.

This article was submitted July 12, 1985; was with the author for revision 16 weeks; and was accepted March 14, 1986. Potential Conflict of In­terest: 4.

1714 PHYSICAL THERAPY

RESEARCH

cess. Both scales can be administered in 45 to 60 minutes, and they present test results in a variety of ways, including a standardized developmental motor quo­tient (DMQ) and an age-equivalent score. Based on the greater number of test items and the partial credit an infant receives for emerging skills, the Peabody scales are more sensitive potentially to change in motor development than is the Bayley Motor Scale. The ability to measure small changes in motor devel­opment is an important quality of an assessment tool that is used to monitor progress in infants demonstrating delays in development.

Although the more recently normal­ized Peabody Developmental Motor Scales offer potential advantages for testing high-risk infants, information on concurrent validity is necessary to deter­mine whether Peabody test results are comparable to results obtained with the more established Bayley Motor Scale. The purpose of this study was to com­pare concurrent and predictive validities of the Bayley Motor Scale and the Pea­body Developmental Motor Scales by administering both tests to groups of full-term and premature infants at 12, 15, and 18 months of age. Concurrent validity is concerned with the corre­spondence between a test's results and the results obtained with a previously available test that is considered reliable and valid. Predictive validity, as it per­tains to this study, indicates the extent to which motor development can be predicted from knowledge of prior test performance.9 Folio and Fewell con­cluded that concurrent validity exists for the Bayley Motor Scale and the Peabody Developmental Motor Scales on the ba­sis of statistically significant (p = .02) correlation coefficients of .36 (Bayley Motor Scale and Peabody Gross Motor Scale) and .37 (Bayley Motor Scale and Peabody Fine Motor Scale).8 When va­lidity is investigated with the correlation analysis, however, the magnitude of the correlation coefficient is the most criti­cal factor; a correlation coefficient below .60 is not considered to be acceptable, regardless of the level of statistical sig­nificance.10 I also investigated the con­current validity of the Bayley Motor Scale and the Peabody Developmental Motor Scales by examining the magni­tude of the correlation coefficients and the mean scores obtained for each as­sessment tool. Additionally, by testing each infant three times, I determined

the ability of each assessment tool to predict motor development over a six-month period and to measure change in motor development. The clinical impli­cations of the results of my study are emphasized in this article.

METHOD

Subjects

The subjects were 23 full-term infants (13 male, 10 female) and 21 premature infants (12 male, 9 female) who were born in 1982 at one of four Boston hospitals. Data were collected as part of a longitudinal study of how to account for gestational age at birth when evalu­ating premature infant motor develop­ment.11 Approval for the study was granted by each hospital's human sub­jects committee, and a consent form was signed by a parent of each infant. The full-term infants had an uncomplicated neonatal history and no medical prob­lems during the first year. Based on neo­natal history and health during the first year, the premature infants were consid­ered to be at low risk for a disorder in motor development. Premature infants selected for the study did not have any of the following medical problems that can contribute to a poor developmental outcome12: neonatal asphyxia, respira­tory complications (infants received a maximum of three days of low-concen­tration, supplemental hood oxygen while in the incubator), seizures, intra­cranial hemorrhage, congenital anoma­lies, or orthopedic deformities. The pre­mature group had a mean gestational age at birth of 31.3 weeks (s = 1.0) and a mean birth weight of 1,602.5 g (s = 270.6).

Procedure

The Bayley Motor Scale and the Pea­body Developmental Motor Scales were administered to all infants within a two-week period after their 12-month, 15-month, and 18-month birthdays. When premature infants are tested, develop­mental quotients commonly are based on adjusted age, which accounts for ges­tational age at birth.13 (An infant born two months prematurely and tested at 12 months' chronological age would have an adjusted age of 10 months.) To compare this practice for the two assess­ment tools examined in this study, I based the developmental quotients on adjusted age for the premature infants.

The mean adjusted ages of the prema­ture group were 10, 13, and 16 months for each successive test session.

Before the testing, I established inter-rater reliability by testing 10 infants not included in the study, and a second therapist independently scored each in­fant's performance. Reliability was de­termined by computing the percentage of agreement on items actually tested. The mean percentages of agreement were 90.0 for the Bayley Motor Scale, 87.5 for the Peabody Gross Motor Scale, and 90.8 for the Peabody Fine Motor Scale.

I administered all tests, which were performed in each infant's home with at least one parent present. A typical test session required one hour to complete. All test items were administered and scored according to the standardized procedures contained in the test man­uals. Neither the Bayley nor the Pea­body scales require that items be ad­ministered in a specific order; therefore, the following procedures were imple­mented to improve data collection. Items from the two assessment tools were interspersed and administered by developmental position (gross motor) or materials presented (fine motor). This procedure minimized the time required for testing and, thus, optimized the in­fants' performances. Several motor be­haviors on the two scales are similar, differing only in criteria for a passing score. For example, item 46 ("Walks alone") on the Bayley Motor Scale re­quires an infant to take 3 independent steps, whereas item 66 ("Walking") on the Peabody Gross Motor Scale requires the infant to take 4 or 5 independent steps. Consequently, motor behaviors common to both scales were scored si­multaneously using the specific criteria for each scale. This procedure elimi­nated unnecessary repetition and also ensured that differences in test results were not attributable to variability in an infant's performance.

Data Analysis

Each infant's age-equivalent scores and standardized motor quotients were calculated for both the Bayley Motor Scale and the Peabody Developmental Motor Scales. The Bayley PDI is deter­mined for age intervals of 1 month. The Peabody DMQ is determined for age intervals of varying length. The age lev­els that pertain to this study are 8

Volume 66 / Number 11, November 1986 1715

through 9 months, 10 through 11 months, 12 through 14 months, 15 through 17 months, and 18 through 23 months.

Concurrent and predictive validities were examined using correlation analy­sis, statistical analysis of mean scores, and change in mean age-equivalent scores. Correlation analysis was per­formed using the Pearson product-mo­ment correlation coefficient. The mag­nitude of each correlation coefficient was interpreted using criteria for exam­ining validity proposed by Meyer10

(Tab. 1). Analysis of mean scores is es­sential for judging concurrent validity. Regardless of the magnitude of the cor­relation coefficients, the clinical inter­pretation of the results would be affected if mean scores were significantly higher or lower for one of the assessment tools. The significance of differences in mean Bayley motor and Peabody gross motor scores was examined for each group by a 2 × 3 (test × age) mixed analysis of variance (ANOVA) with repeated meas­ures on age. For each assessment tool, the significance of the age-related changes in each group's mean age-equivalent scores was tested by a one-factor ANOVA for repeated measures, and significant age effects were analyzed further by the Newman-Keuls method of multiple comparisons. The .05 prob­ability level was used to test for statisti­cal significance.

RESULTS

Concurrent Validity of Age-Equivalent Scores

Correlation analysis indicated good to high correlation between Bayley motor and Peabody gross motor age-equivalent scores for both groups with correlation coefficients ranging from .78 to .96 (Tab. 2). The correlation between Bay-ley motor and Peabody fine motor age-equivalent scores was not acceptable for either group; only the correlation coef­ficient for the premature group at 15 months of age was greater than .50. Mean Bayley motor and Peabody gross motor age-equivalent scores differed for the premature group by no greater than 0.4 months and for the full-time group by 0.2 months, 0.8 months, and 1.5 months, respectively, for the three test ages (Tab. 3). For each group, the ANOVA indicated that mean age-equiv­alent scores for the two assessment tools did not differ significantly (full-term

TABLE 1 Criteria for Assessing Validity Based on Magnitude of Correlation Coefficient10

Range of Correlation Coefficients .85 or above .80-.84 70-.79 .60-.69 .59 or below

Correlation Between Test Scores high very good fair to good poor unacceptable

TABLE 2 Concurrent Validity of Bayley Motor and Peabody Gross Motor (GM) ant Age-Equivalent Scores

12 Months 15 Months

d Fine Motor (FM)

18 Months Group Scales

Full-term (n = 23)

Premature (n = 21)

Bayley, Peabody GM Bayley, Peabody FM Bayley, Peabody GM Bayley, Peabody FM

r .95a

.20

.84a

.49b

r .83a

.31

.96a

.57c

r .78a

.40

.84a

.30

TABLE 3 Mean Bayley Motor and Peabody Gross Motor (GM) and Pine Motor (FM) Age-Equivalent Scores of Full-Term and Premature Infants

Group

Full-term (n = 23)

Premature (n = 21)

Scale

Bayley Peabody GM Peabody FM Bayley Peabody GM Peabody FM

12 Months 15 Months 18 Months

13.0 12.8 13.6 10.0 10.2 10.7

s 1.5 1.7 1.3 1.2 1.3 1.3

16.5 15.7 17.0 13.4 13.0 14.3

s 2.7 1.8 1.5 2.3 1.9 1.9

20.6 19.1 19.1 16.2 16.1 17.2

s 3.3 1.8 1.4 2.2 2.3 1.5

TABLE 4 Concurrent Validity of Bayley Psychomotor Developmental Index (PDI) and Peabody Gross Motor (GM) and Fine Motor (FM) Developmental Motor Quotients (DMQs)

Group Quotients

Full-term (n = 23)

Premature (n = 21)

PDI, DMQ GM PDI, DMQ FM PDI, DMQ GM PDI, DMQ FM

12 Months

r .93a

.17

.63b

.23

15 Months

r .86a

.20

.92a

.59b

18 Months

r .77a

.34

.88a

.32

group: F = 2.21; df= 1,44; p = NS; premature group: F = 0.02; df= 1,40; p = NS).

Concurrent Validity of Developmental Motor Quotients

Similar to the age-equivalent scores, correlation analysis indicated good to high correlation (with one exception) between Bayley PDIs and Peabody gross motor DMQs, whereas correlations be­

tween PDIs and fine motor DMQs were unacceptable (Tab. 4). Correlation coef­ficients between PDIs and gross motor DMQs ranged from .63 to .93 with only two correlation coefficients below .86. Poor correlation between PDIs and gross motor DMQs was found only for the premature group at 12 months of age (r = .63). All correlation coefficients between PDIs and fine motor DMQs were below .50 with the exception of the

aP<.Q01. b p < .05. c p < .01.

a p < .001. b p < .01.

1716 PHYSICAL THERAPY

RESEARCH

TABLE 5 Mean Bayley Psychomotor Developmental Indexes (PDIs) and Peabody Gross Motor (GM) and Fine Motor (FM) Developmental Motor Quotients (DMQs) of Full-Term and Premature Infants

Group Quotienta

Full-term (n = 23)

Premature (n = 21)

PDI DMQ GM DMQ FM PDI DMQ GM DMQ FM

12 Months 15 Months 18 Months

104.4 91.7 96.7 96.9 90.1 95.7

s 12.9 12.9 9.7

14.8 15.5 10.9

111.2 94.0 98.5

102.0 91.8

101.7

s 16.4 11.9 10.6 19.4 16.0 12.4

112.1 96.1 87.4

101.7 97.1 99.7

s 19.9 13.3

8.9 14.4 14.2

9.9

TABLE 6 Prediction of Bayley Motor and Peabody Gross Motor (GM) and Fine Motor (FM) Age-Equivalent Scores of Full-Term and Premature Infants

Bayley Motor Peabody GM Peabody FM Group

Full-term (n = 23)

Premature (n = 21)

Age (mo)

12 15 12 15

15 Months

r .70a

.64b

18 Months

r .54b

.59b

.56b

.66b

15 Months

r .85a

.73a

18 Months

r .60b

.58b

.54b

.72a

15 Months

r .65a

.77a

18 Months

r .25 .35c

.75a

.38c

correlation coefficient for the premature group at 15 months of age (r = .59).

The mean PDIs were higher than the mean gross motor DMQs (Tab. 5); all mean quotients reported in this study, however, were within the normal range. The mean PDIs were 12.7 to 17.2 points higher than the mean gross motor DMQs for the full-term group and 4.6 to 10.2 points higher for the premature group. The ANOVA for repeated meas­ures (age) was highly significant for the full-term group (F = 16.7; df= 1,44; p < .001), but it was not significant for the premature group (F = 2.92; df = 1,40; p = NS). For comparison, the Bay-ley Motor Scale had a normative mean PDI of 100 and a standard deviation of 164; the Peabody Developmental Motor Scales had a normative mean DMQ of 100 and a standard deviation of 15.8

Therefore, when mean quotients of the premature group were based on adjusted age, all mean PDIs and DMQs reported in this article were within the normal range.

Prediction of Age-Equivalent Scores

Predictive ability of 18-month Bayley motor and Peabody gross motor and

fine motor age-equivalent scores from 12-month scores was unacceptable with the exception of the premature group's Peabody fine motor scores (r = .75) and the full-term group's Peabody gross mo­tor scores (r = .60) (Tab. 6). For the 3-month intervals, predictive ability of Peabody gross motor scores ranged from unacceptable to high (range, r = .58 to r = .85), whereas predictive ability of Bayley motor scores ranged from unac­ceptable to fair (range, r = .59 to r = .70). The predictive ability of 15-month Peabody fine motor age-equivalent scores from 12-month scores was good for the premature group (r = .77) and poor for the full-term group (r = .65). The predictive ability of 18-month Pea­body fine motor scores from 15-month scores was unacceptable for both groups.

Change in Mean Age-Equivalent Scores

Between each test age, the full-term and premature infants made from 2.8 to 4.1 months' gain in mean Bayley motor age-equivalent scores, from 2.8 to 3.4 months' gain in mean Peabody gross motor age-equivalent scores, and from 2.1 to 3.6 months' gain in mean Peabody fine motor age-equivalent

scores, gains that were significant for each group (Tab. 3). For the full-term group, the age effect was well above the level required for statistical significance for the Bayley Motor Scale (F = 102.5; df = 2,44; p < .001), the Peabody Gross Motor Scale (F = 232.7; df = 2,44; p < .001), and the Peabody Fine Motor Scale (F = 155.9; df = 2,44; p < .001). Similarly, for the premature group, the age effect was highly significant for the Bayley Motor Scale (F = 121.8; df = 2,40; p < .001), the Peabody Gross Mo­tor Scale (F = 137.7; df = 2,40; p < .001), and the Peabody Fine Motor Scale (F = 214.2; df = 2,40; p < .001). For each assessment tool, multiple com­parison tests revealed that mean age-equivalent scores were significantly higher at each successive test age (p < .01). The mean age-equivalent scores of the full-term group generally were 2 to 3 months' higher than those of the pre­mature group, an anticipated finding considering that the premature infants were born an average of 2 months pre­maturely.

DISCUSSION

The results provide evidence of con­current validity for Bayley motor and

a Quotients of premature group based on adjusted age.

a p < .001. b p < .01. c p < .05.

Volume 66 / Number 11, November 1986 1717

Peabody gross motor age-equivalent scores. Correlation analysis revealed that good to high correlation existed between age-equivalent scores and that mean age-equivalent scores for the two assessments were not significantly dif­ferent. The difference between mean Bayley motor and Peabody gross motor age-equivalent scores for the full-term group, however, increased at each suc­cessive test age. The full-term group's mean age-equivalent score of 20.6 months at 18 months of age suggests that the Bayley Motor Scale may over­estimate motor development at certain age levels. Both groups of infants made significant gains in mean Bayley motor and mean Peabody gross motor and fine motor age-equivalent scores during the 6-month period. The infants' rate of development supports the concepts that motor development is orderly and se­quential and that motor abilities of healthy infants increase with age, imply­ing that both Bayley and Peabody age-equivalent scores provide valid meas­ures of age-related change in motor de­velopment.

The correlation results for gross mo­tor development are more favorable than those reported by Folio and Fewell8; however, the Peabody test man­ual does not describe the subjects tested or present mean scores, preventing a comparison of results. The low correla­tion between Bayley PDIs and Peabody gross motor DMQs (r = .36) reported by Folio and Fewell may reflect, in part, their sample having included infants un­der 1 year of age who were administered both gross motor and fine motor items from the Bayley Motor Scale. Con­versely, because the Bayley Motor Scale contains no fine motor items above the 8.9-month age level, the infants in this study almost exclusively were adminis­tered gross motor items from the Bayley Motor Scale. The Bayley Motor Scale, therefore, served as a measure of gross motor development. Consequently, the correlations between the Bayley motor and Peabody gross motor scores were higher than the correlations between the Bayley motor and Peabody fine motor scores. The unacceptable correlation be­tween the Bayley motor and Peabody fine motor scores reported by Folio and Fewell8 and in this study suggests that performance on gross motor items is not related strongly to performance on fine motor items. Therapists, therefore, are encouraged to perform separate assess­

ments of gross motor and fine motor development.

The significantly higher mean Bayley PDIs achieved by the full-term group pose a threat to concurrent validity when Bayley motor and Peabody gross motor results are expressed using stand­ardized quotients. The difference in mean quotients for the two assessment tools may be attributable to the method used to determine standardized quo­tients for the Peabody Developmental Motor Scales. The Bayley PDI is based on age-level intervals of 1 month, whereas the Peabody DMQ is based on variable and wide-ranging age levels. The DMQ for a 12-month-old infant, for example, is based on the perform­ance of infants in the normative group whose ages ranged from 12 to 14 months. In this study, the ages of the full-term infants corresponded to the youngest month of each age level for each test session and, therefore, their lower mean DMQs and inability to achieve a mean DMQ of 100 are not surprising.

The method of determining standard­ized quotients for the Peabody scales also presents a problem when testing premature infants. This problem was avoided in this study by the selection of chronological test ages that corre­sponded to the youngest month of an age level. Depending on the chronolog­ical age of the premature infant, adjust­ing age to account for gestational age at birth may not change the DMQ. For example, an infant born 2 months pre­maturely and tested at 14 months' chronological age would have an ad­justed age of 12 months. The DMQ would be determined for the 12- through 14-month age level and, therefore, would be identical for both chronologi­cal and adjusted ages. In contrast, the Bayley PDI would be higher when de­termined for adjusted age, as compared with chronological age.

The findings of this study indicate that, when results are expressed as stand­ardized quotients, decisions regarding an infant's motor developmental status could differ depending on whether the Bayley Motor Scale or Peabody Gross Motor Scale is used. Standardized DMQs are advantageous for determin­ing whether an infant's motor develop­ment is within age expectations or for presenting research findings. Alter­nately, age-equivalent scores are more meaningful for infants whose motor de­

velopment is below the normal range of variability. Age-equivalent scores also provide a clearer indication of func­tional ability than do standardized DMQs. Therapists may find that age-equivalent scores are more appropriate than standardized quotients for clinical use and, thus, avoid the problem of interpreting differences between the standardized quotients obtained with the Bayley Motor Scale and the Peabody Gross Motor Scale.

Neither Bayley nor Peabody 12-month age-equivalent scores effectively predicted motor development at 18 months of age in this study. Correlation coefficients were higher between succes­sive 3-month intervals, indicating that the ability to predict motor develop­ment decreases as the timespan involved increases. The variability associated with infant development and differences in the motor competencies required for success at older age levels probably are limiting factors in attempting to make predictions. For the Bayley Motor Scale or the Peabody Developmental Motor Scales to be of value in predicting de­velopment at older ages, correlation coefficients of .75 or greater must be demonstrated between scores obtained over periods of time when major changes in motor ability occur (eg, at 3 and 12 months of age or at 12 and 24 months of age). Until evidence of pre­dictive validity is demonstrated, the Bayley Motor Scale and the Peabody Developmental Motor Scales should be used only to provide current informa­tion on the motor ability of an infant relative to the performance of the nor­mative group. Repeated testing is nec­essary to determine development at older ages.

This study suggests that therapists may choose between the Bayley Motor Scale and the Peabody Developmental Motor Scales when results are reported using age-equivalent scores. The Bayley Motor Scale may be the preferred as­sessment tool when time is restricted, such as in infant follow-up clinics. When evaluation of both gross motor and fine motor development is desired, however, the Peabody Developmental Motor Scales may be more appropriate than the Bayley Motor Scale. Addition­ally, the Peabody Developmental Motor Scales may be selected when a more thorough evaluation of motor develop­ment is indicated. The Peabody scales also potentially are able to document

1718 PHYSICAL THERAPY

RESEARCH

development over relatively short pe­riods of time and measure response to physical therapy intervention.

In this study, changes in mean age-equivalent scores of healthy full-term and premature infants were highly sig­nificant over a six-month period for both the Bayley Motor Scale and the Peabody Developmental Motor Scales. These findings, however, may not be generalizable to infants with identified delays or disorders in motor develop­ment. Research with the experimental edition of the Peabody Developmental Motor Scales14 revealed statistically sig­nificant treatment effects for a group of children with delayed development15

and a group of severely and profoundly retarded nonambulatory children.16 The results of these two studies suggest that the Peabody scales are an appropriate assessment tool for children with delays or disorders in motor development. The latter study especially is important be­cause of the low levels of function in the children tested. A criticism of norm-referenced motor scales is that they are not suitable for evaluating short-term responses to treatment in nonambula­tory children, particularly children with cerebral palsy.17 Further research to ex­amine the validity of using the Bayley Motor Scale and the Peabody Develop­mental Motor Scales to document de­velopmental changes in high-risk infants and infants with identified delays or dis­orders in motor development may aid physical therapists in the proper selec­tion of evaluation tools.

CONCLUSIONS

This study provides evidence of con­current validity for the Bayley Motor Scale and the Peabody Gross Motor Scales on the basis of age-equivalent scores. The results also indicate that the Bayley Motor Scale and the Peabody Developmental Motor Scales effectively measure age-related change in motor development for healthy full-term and premature infants but do not predict motor development at later ages. When the data are based on standardized quo­tients, however, problems may exist in the clinical interpretation of the Pea­body DMQ, presenting a threat to con­current validity. This study suggests that therapists should perform separate as­sessments of gross motor and fine motor development and that repeated testing is necessary to determine motor devel­opment at later ages. Further research is

needed to determine the validity of us­ing the Bayley Motor Scale and the Pea­body Developmental Motor Scales to measure change in motor development for infants with identified delays or dis­orders in motor development.

Acknowledgment. I thank Dr. Ste­phen Haley, Assistant Professor, De­partment of Physical Therapy, Sargent College of Allied Health Professions, Boston University, for his advice in pre­paring this manuscript.

REFERENCES

1. Amiel-Tison C: A method of neurologic evalu­ation within the first year of life. Curr Probl Pediatr 7:1-50, 1976

2. Prechtl HFR: Assessment methods for the newborn infant: A critical evaluation. In Stratton P (ed): Psychobiology of the Human Newborn. New York, NY, John Wiley & Sons Inc, 1982, pp 21-52

3. Standards for Educational and Psychological Tests. Washington, DC, American Psychologi­cal Association, 1974

4. Bayley N: The Bayley Scales of Infant Devel­opment. New York, NY, The Psychological Corporation, 1969

5. Campbell SK, Wilhelm IJ: Development from birth to 3 years of age of 15 children at high risk for central nervous system dysfunction. Phys Ther 65:463-469, 1985

6. Field TM, Dempsey JR, Shuman HH: Devel­opmental follow-up of pre- and postterm in­fants. In Friedman SL, Sigman M (eds): Preterm Birth and Psychological Development. New York, NY, Academic Press Inc, 1981, pp 299-311

7. Harris SR: Effect of neurodevelopmental ther­apy on motor performance of infants with Down's syndrome. Dev Med Child Neurol 23:477-483, 1981

8. Folio MR, Fewell RR: Peabody Developmental Motor Scales and Activity Cards. Allen, TX, Teaching Resources, 1983

9. Anastasi A: Psychological Testing, ed 5. New York, NY, Macmillan Publishing Company, 1982, pp 137-144

10. Meyer CR: Measurement in Physical Educa­tion. New York, NY, The Ronald Press Com­pany, 1974, pp 86-89

11. Palisano RJ: Use of chronological and adjusted age to compare motor development of healthy preterm infants to full-term infants. Dev Med Child Neurol 28:180-187, 1986

12. Koops BL, Harmon RJ: Studies on long-term outcome in newborns with birth weights under 1500 grams. Advances in Behavioral Pediatrics 1:1-28,1980

13. Gesell A, Amatruda CS: Developmental Diag­nosis. New York, NY, Paul B. Haeber, 1947, pp 290-295

14. Folio MR, Dubose RF: Peabody Developmental Motor Scales (Revised Experimental Edition). Nashville, TN, George Peabody College for Teachers, 1974

15. Jenkins JR, Sells CJ, Brady D, et al: Effects of developmental therapy on motor impaired chil­dren. Physical & Occupational Therapy in Pe­diatrics 2(4): 19-28, 1982

16. Ottenbacher K, Short MA, Watson PJ: The effects of a clinically applied program of vestib­ular stimulation on the neuromotor perform­ance of children with severe developmental disability. Physical & Occupational Therapy in Pediatrics 1(3):1-11, 1981

17. Wolf JM: Results of Treatment in Cerebral Palsy. Springfield, IL, Charles C Thomas, Pub­lisher, 1969

Volume 66 / Number 11, November 1986 1719