The ecological relevance of the test of infant motor performance elicited scale items.

Key Words: Ecology; Environment; Motor development; Pediatrics, evaluation; Tests and measurements, functional; Validity.

The Test of Infant Motor Performance (TIMP) is a test currently under development)[1,2] that is designed to assess the postures and movements used by preterm and full-term infants under the age of 4 months as they interact with people, objects, and their environment. The hypothetical construct or latent trait underlying the TIMP is the postural and selective control of movement that is required for functional activities such as changing positions and stabilizing the head to allow the eyes to focus on people or objects. The TIMP (Version 3.0) consists of 28 items scored dichotomously (present, not present) from observations of spontaneous behaviors (Observed Scale) and 25 elicited items scored on 4-, 5-, or 6-point hierarchical scales (Elicited Scale). The Elicited Scale items require handling of the infant according to standardized procedures in order to evoke a postural or motor response (see Fig. 1 for an example). We wanted to determine whether the postural and movement demands placed on infants during the administration of elicited TIMP items were representative of the environmental demands placed on infants during caregiving activities. An environmental demand was defined as moving, stimulating, or positioning an infant in a way that is assumed to require a postural response or movement by the infant.

[Figure 1 ILLUSTRATION OMITTED]

The validity of the TIMP has been studied.[1-3] According to the "Standards for Tests and Measurements in Physical Therapy Practice,"[4] validity is defined as the meaningfulness of interpretations that can be inferred from a measurement. Validity is often divided into three categories: criterion-based validity, content validity, and construct validity.[5-7] Criterion-based validity is defined as the ability of a test to systematically demonstrate a relationship to external criteria.[4,5] The criterion-based validity of the TIMP has not been studied. Content validity is the degree to which test items reflect the elements of its construct as they have been defined.[4,5] Content validity of the TIMP has been investigated in a previous study during the initial development of the test.[3] How representative the content of the TIMP items is to the universe of possibilities was assessed through a review of the test items by 21 expert judges.[3] The results supported the content validity of the TIMP in that the test was found to contain a representative sample of items measuring infant postural and motor responses.[3]

Construct validity reflects the extent to which a test measures the underlying theoretical concepts and premises on which the test is based[6] and provides the basis for making inferences from a measurement.[4,5] Evidence of construct validity is provided by studies of (1) age differentiation (test scores increase with age), (2) correlational evidence (relationship to another test), (3) differences between defined groups, or discriminative validity, (4) factor analysis (a statistical procedure used to assess the underlying dimensions within a set of variables), and (5) internal consistency of the instrument.[7,8] Internal consistency measures reflect how well a group of items assesses the same trait.[3] Discriminative validity and age differentiation have been assessed for the TIMP.[2] The TIMP was administered to 137 infants with a high, medium, or low degree of medical risk (based on medical risk scores for mortality and morbidity) who ranged in age from 32 weeks postconceptional age to 3.5 months postterm age. A multiple regression analysis demonstrated that TIMP scores increased with age and decreased as the number of medical complications increased.[2] The TIMP, therefore, has evidence supporting its discriminative validity for medical complications and maturation. The internal consistency coefficient for the TIMP is .98, indicating that the test items on the TIMP correlate well with each other and thus form a homogeneous scale.[1] Intrarater and interrater reliability of measurements obtained with the TIMP have been assessed,[9] and test-retest reliability evaluation is in progress. Studies are also in progress to examine the predictive validity of the TIMP and to further examine the discriminative validity of the TIMP. Eventually, the developers of the TIMP are hopeful that the test will be responsive to effects of physical therapy provided to infants for the purpose of improving functional motor performance. The test would be more efficient and useful if items unrelated to the functional needs of daily life or the prediction of developmental disability could be eliminated.

It is important that the TIMP, a test designed for use primarily in clinical settings such as the special care nursery or outpatient clinic, contain items that are related to the infant's experiences and behaviors in his or her ecological niche (de, the natural setting of the home). If the demands placed on infants during administration of TIMP items are similar to those imposed on infants during caregiving, the test can be considered to have ecological relevance (de, relevance to daily life experiences).

We examined the relationship between the demands placed on infants during the administration of the TIMP Elicited Scale items and the environmental demands placed on infants by their caregivers during the functional activities of dressing, bathing, and play. We did not evaluate the infants' response to the demands. The purpose of our study was to determine whether the demands placed on infants during the administration of the TIMP Elicited Scale items can be related to the demands of caregiving in natural settings.

Method

Subjects

Twenty-two caregivers and 22 infants participated in this investigation. There were 15 preterm infants (33 weeks 3 days postconceptional age to 12 weeks 5 days postterm age corrected for prematurity) and 7 full-term infants (6 days to 12 weeks 4 days of age). Scores on the TIMP are supposed to reflect an infant's maturation and degree of medical risk[2]; therefore, infants were from several age groups and two risk groups (low and high) (Tab. 1). This method ensured that subjects included infants with varied developmental levels. Within each age group outlined in Table 1, subjects included one infant with low and one with high numbers of medical risk factors. The degree of medical risk was determined by the score on the Newborn Infant Form of the Problem-Oriented Perinatal Risk Assessment System (POPRAS).[10,11] Low medical risk was defined as having a POPRAS score lower than 71. High medical risk was defined as having a POPRAS score greater than 70.

Table 1
Sampling Block Design

Age                                    High      Low
Group   Postconceptional Age           Risk      Risk

  1     32 weeks to 33 weeks 6 days
  2     34 weeks to 35 weeks 6 days   Block 1   Block 2
  3     36 weeks to 37 weeks 6 days   (n=5)      (n=5)
  4     38 weeks to 39 weeks 6 days
  5     40 weeks to 41 weeks 6 days

 Age                                   High      Low
Group   Postterm Age                   Risk      Risk

  6      2 weeks to 3 weeks 6 days
  7      4 weeks to 5 weeks 6 days
  8      6 weeks to 7 weeks 6 days    Block 3   Block 4
  9      8 weeks to 9 weeks 6 days    (n =6)    (n =6)
 10     10 weeks to 11 weeks 6 days
 11     12 weeks to 13 weeks 6 days

The following eligibility criteria were adhered to during subject recruitment:

1. The expected date of delivery of the mother was known.

2. The infant's postconceptional age or postterm age, as calculated from the expected date of delivery, fell within one of the age groups defined in Table 1. We verified whether the infant's age at birth, as calculated from the expected date of delivery, corresponded with the physician's assessment of the infant's gestational age at birth. The two ages had to match within 2 weeks of each other. The physician's assessment of gestational age was allowed to span no more than 2 weeks (eg, between 33 and 34 weeks gestational age). A match, therefore, involved a gestational age based on an expected date of delivery of no more than 1 week less than the lowest or 1 week more than the highest physician's estimate when a 2-week range was given.

3. The infant was the product of a singleton birth.

4. The parents gave informed consent for their participation in the study. If the primary caregiver was not a parent (eg, a nurse), then the primary caregiver also provided signed informed consent.

5. The parents gave informed consent for the videotaping of their child and themselves using standard departmental forms for photographic release.

6. Hospitalized infants were medically stable (not on mechanical ventilation or with intravenous lines; a nasal oxygen cannula was allowed).

7. Hospitalized infants had their physician's approval for participation in the study.

8. The infants in the low-risk group were without congenital deformities, diagnosed syndromes, bronchopulmonary dysplasia, or documented insult to the brain (no intraventricular hemorrhage or periventricular leukomalacia, perinatal asphyxia, seizures, or abnormal electroencephalogram or brain scan). The low-risk group represented the "normal" population (de, infants who did not have medical complications typically associated with poor motor outcome, although their developmental status was not formally documented).

Subjects (caregivers and their infants) were selected as a sample of convenience.[8] Four blocks of infants are outlined in Table 1. The total sample size was 22. The age range of preterm infants (32-42 weeks postconceptional age) was divided into five postconceptional age groups spanning a 2-week interval. The age range of full-term infants or preterm infants who were of postterm age (2-13 weeks postterm age) was divided into six postterm age groups, also spanning a 2-week interval. The sampling block design was divided into two age groups: postconceptional age and postterm age. There were two blocks in each age group (one block for high-risk infants and one block for low-risk infants), for a total of four blocks. Block 1 consisted of 5 infants with a high degree of medical risk, 1 infant from each of the postconceptional age groups. Block 2 consisted of 5 infants with a low degree of medical risk, 1 infant from each of the postconceptional age groups. Block 3 consisted of 6 infants with a high degree of medical risk, 1 infant from each of the postterm age groups. Block 4 consisted of 6 infants with a low degree of medical risk, 1 infant from each of the postterm age groups. Sampling biases were controlled by (1) strictly adhering to the eligibility criteria, (2) including infants of different sex (at least 2 male infants and 2 female infants were represented in each block), and (3) including infants of different ethnic or racial backgrounds (1 infant in each block from the Hispanic-American, African-American, and non-Hispanic European-American groups, for a total of at least 7 infants from each of the three groups represented in the total). Other variables, such as first-time parents versus parents with other children, level of education of the primary caregiver or mother, age of the primary caregiver, marital status of the mother, and level of infant caregiving experience of the primary caregiver were not controlled. Descriptive information was collected from each family on these variables.

A summary of the characteristics of the 22 infants and their caregivers is presented in Table 2. The sampling plan and eligibility criteria for the selection of the infants were adhered to with only one exception. Because of an inaccurate calculation, the infant selected for the low-risk 6- to 7-week age group was actually 8 weeks 0 days postterm age. Seventeen of the infants were seen in their homes, and 5 of the infants were seen in the hospital for the videotaped caregiving observations. Two of the 5 hospitalized infants were in the preterm age, low-risk group, and 3 infants were in the preterm age, high-risk group. Four of the 5 hospitalized infants had nurses as their primary caregivers; all other caregivers were biologic mothers.

[TABULAR DATA 2 NOT REPRODUCIBLE IN ASCII]

Procedure

An observational research design using event sampling was our method. We used a systematic approach to quantifying behavior, with videotaping allowing for a microanalysis of the observations.[12,13] The infants were videotaped by the first author while their primary caregiver interacted with them during typical caregiving activities. Control for possible extraneous factors that could threaten the internal validity of this study was imposed by observing the same caregiving activities for all subjects. Four activities were videotaped: undressing, bathing, dressing, and 5 minutes of play. By controlling the interactive activities, each infant had an equal opportunity to experience the same caregiving demands. The infants who were at home were videotaped at whatever time the primary caregiver reported to be the infant's most content, alert time, which was also convenient for the family. The infants who were in the hospital were videotaped 1 hour before a feeding. The caregivers were instructed to dress and bathe the infant in their normal manner and then play with the infant for 5 minutes however they normally played. All infants were videotaped on one occasion for as long as it took the caregiver to perform bathing and dressing, followed by 5 minutes of timed play. The length of videotaped observations ranged from 7 to 30 minutes ([bar] X = 16.4, SD = 4.8).

In observational research, a particular event or a particular time interval may be the unit categorized.[12,13] We used a continuous recording strategy. We first detected events or demands and then categorized them according to our predefined coding scheme. The infant's response to a demand was not recorded for the purposes of this investigation. Only the occurrences of the demand, as defined by each code, were recorded. The predefined coding scheme of demands placed on the infants represented the 25 TIMP Elicited Scale item administration descriptions and was developed initially using pilot videotapes. There were 25 codes that were numbered to correspond to the 25 TIMP items and were defined by what we believed to be observable, concrete features. For example, code number 13 is "with the infant in a supine position and head in midline, the caregiver flexes both of the infant's hips and knees up toward the abdomen and then releases them." The corresponding TIMP item, number 13, states: "Position: Infant supine; head held in midline and hips and knees flexed onto abdomen, knees together. Procedure: Slowly release the infant's legs while maintaining the head in midline." The codes were not the exact item descriptions for test administration, but rather approximations of the item descriptions. For the purposes of administration of the TIMP, positions and procedures were standardized. For the purposes of coding behaviors, however, item descriptions had to be redefined to reflect characteristics of the observed behavior in a variety of contexts, yet still maintain the test item's identity. The coding scheme was later refined during the process of establishing rater reliability. A code (number 26), known as "other" demands, was added to account for any demand placed on the infant that did not correspond to a TIMP item. The recording forms consisted of a list of all the codes as well as a blank sheet for the observer to describe each of the "other" demands. (See Appendix for a list of brief descriptions of environmental demands that correspond to TIMP Elicited Scale items. This list, however, is not the detailed coding scheme.)

Our data consisted of frequency counts for each predefined code. The first author coded all of the videotapes. A coding session consisted of viewing a videotape from start to finish in regular time (standard speed) of the videotape and pausing each time a demand was placed on the infant (de, the infant was handled or an external stimulus was presented to the child, such as a rattle was shaken or a toy was held in the infant's visual field). The recording could be rewound and replayed if needed in order to verify that the observed behavior met the criteria defined in the code (typically three to four replays were used). After a demand was identified, a check was placed next to the appropriate code on a data recording form. After viewing the entire videotape, the frequency of occurrences for each code was summed. On average, 29 demands were identified and coded from each videotape.

Reliability

To analyze the reliability of coding, percentage of agreement between two observations of the same videotape viewed at different times (intrarater) or between two observers of the same videotape (interrater) was used. Percentage of agreement is a statistic that estimates the ability of raters to agree on category ratings, but it does not correct for chance agreement.[8] A stronger statistical method for assessing reliability, such as the kappa statistic, could not be used in our study because the total number of demands between any two sets of observations of the same videotape were not necessarily the same. In our study, the demands placed on the infant had to be identified first and then coded by the rater, with no time information recorded. Time is often used to identify the events the rater must code in some observational research designs. Bakeman and Gottman[12,13] recommend using percentage of agreement to assess reliability in observational designs where the observer both detects and codes events with no time information recorded on the videotape.

Interrater reliability of four randomly selected videotapes was secured before the remaining videotapes were analyzed by the first author. The second observer of the videotapes was a pediatric physical therapist who had knowledge of the TIMP. She knew the purpose of the study. The raters trained together for approximately 10 to 15 hours per week over 6 weeks and made refinements to the coding scheme. They also developed general rules for coding. When the interrater reliability study began, no further changes in the code definitions were made. The results of the interrater reliability study were 80%, 83%, 95%, and 96% overall agreement on demands identified for each of the four videotapes.

Intrarater reliability was analyzed after coding every sixth videotape to prevent intraobserver reliability decay. After coding the sixth videotape, the first videotape in the series of six videotapes was recoded. The results of the intrarater reliability analysis were 87%, 96%, 96%, and 91% overall agreement.

Data Analysis

To analyze the data collected via video recording, the frequency of occurrence for each demand was calculated. The data were then analyzed from the following three perspectives: (1) the number and percentage of infants in the sample who received environmental demands corresponding to each TIMP item, (2) the number and percentage of all demands corresponding to TIMP items that were imposed on each infant, and (3) the frequency of repetition of the same TIMP item demand within the caregiving sessions of all infants.

The data were explored for relationships between five independent subject variables and two dependent variables. The independent variables were preterm versus postterm age, male versus female, low risk versus high risk based on medical complications, race or ethnicity (Hispanic American versus African American versus non-Hispanic European American), and level of education of the caregiver (less than high school education, high school education, some college, and college degree or more). The dependent variables were the rate of environmental demands per minute and the rate of environmental demands corresponding to TIMP items per minute. The rate of demands in the groups based on age, gender, and degree of medical risk were compared using t tests because these groups contained two categories. The rate of demands in the groups based on ethnicity and level of education were compared using a one-way analysis of variance because these groups contained three and four categories. The level of significance for accepting group differences was set at P [is less than] .05 because there was no critical concern for further decreasing the probability of a Type I error.[8] The SPSS/PC+ (Version 6.1) statistical software program(*) was used for data analysis.

Results

The length of time of the videotaped observations ranged from 7 to 30 minutes and averaged 16.4 minutes (SD = 4.8). The mean total time of the videotaped observations was shorter for infants who were in the preterm age group than those who were in the postterm age group ([bar] X = 13.9 minutes, SD = 3.2, range = 7-18, for preterm age infants and [bar] X = 18.4 minutes, SD = 5.0, range = 12-30, for postterm age infants; t = -2.46, P = .02). The average number of total environmental demands observed across the sample (the average frequency of occurrences of demands placed on infants) was 52.9 (SD = 22.8, range = 22-119). The mean number of demands corresponding to TIMP items observed across the entire sample was 26.4 (SD = 15.8, range = 7-69). Because observations varied in length, demands were prorated for length of observation. The mean rate was 3.23 (SD = 1.07, range = 2.13-7.0) environmental demands per minute of observation time. The mean rate of environmental demands that corresponded to TIMP items was 1.58 (SD = 0.83, range = 0.6-4.38) demands per minute of observation time. For infants in the sample, therefore, approximately 50% of all demands placed on infants during caregiver interaction corresponded to TIMP items.

Table 3 summarizes the total number and percentage of demands reflecting TIMP Elicited Scale items that were observed for each subject. The range of demands corresponding to TIMP items that were observed was from 16% to 68%. The mean number of demands corresponding to different TIMP items observed during a caregiving session for the sample as a whole was 9.2 (SD = 3.7). On average, approximately 37% of the TIMP Elicited Scale items were represented in the observed demands placed on infants during a typical caregiving session. The mode was 36%. Figure 2 shows the number of infants who had the specified percentage of TIMP items represented in the observed demands.

Table 3. Demands Corresponding to Test of Infant Motor Performance (TIMP) Items Observed for Each Subject

                Total No. of        Percentage of
                Demands Related     Demands Related
                to TIMP Items       to TIMP Items
Subject No.     Observed            Observed

1L                    6                 24
2L                    5                 20
3L                   12                 48
4L                    9                 36
5L                    9                 36
6L                    4                 16
7L                   13                 52
8L                    8                 32
9L                   13                 52
10L                  13                 52
11L                   9                 36
1H                    6                 24
2H                    5                 20
3H                    7                 28
4H                   10                 40
5H                    7                 28
6H                   17                 68
7H                   14                 56
8H                    5                 20
9H                    7                 28
10H                  15                 60
11H                   9                 36

[Figure 2 ILLUSTRATION OMITTED]

Table 4 summarizes the variety of positions in which the infants in the sample were placed. The majority of infants in the sample (81%) received demands corresponding to TIMP items in at least three different positions in space. The three positions in which infants were typically placed and received demands corresponding to TIMP items were supine, suspended (either vertically or horizontally), and sitting. Only 9 of the 22 infants (41% of the sample) were placed in the prone position during the caregiving session, and only 2 of the 22 infants (9% of the sample) received a demand corresponding to a TIMP item in a side-lying position.

Table 4. Number of Infants Who Received Demands Related to Test of Infant Motor Performance (TIMP) Items in the Specified Position

Position in Which TIMP Item   No. of Infants   Percentage
Demands Were Received           (N = 22)       of  Sample

Supine                              22            100
Suspended                           20             91
Sitting                             16             73
Prone                                9             41
Side lying                           2              9

Figure 3 illustrates that demands corresponding to 23 of the 25 TIMP items (92%) were observed to occur in the videotape of at least one infant in the sample. Demands corresponding to item numbers 17 and 20 were observed in only one infant, whereas the most frequently occurring demand corresponding to item number 13 was seen in 21 of the 22 infants. Only two TIMP items (3 and 23) were not represented by the observed demands.

[Figure 3 ILLUSTRATION OMITTED]

The majority of demands corresponding to TIMP Elicited Scale items were observed to occur more than one time during a caregiving session. Demands related to TIMP items seen more than four times during one or more caregiving sessions are presented in Table 5. Demands related to TIMP items seen two to four times during one or more caregiving sessions are presented in Table 6. Observed demands related to TIMP item numbers 8, 11, 14, 17, 18, 20, and 22 occurred at a low frequency (in fewer than 30% of the sample). Demands reflecting these TIMP items, however, were often observed more than one time. The exception was the demand corresponding to item number 17, which was observed only one time in one subject.

Table 5. Demands Corresponding to Test of Infant Motor Performance (TIMP) Items Observed Greater Than Four Times in a Subject

                    Demands Corresponding
                    to TIMP Item Numbers
No. of Subjects     Observed [is greater than] 4 Times

8                          24
5                          13
3                           2
2                           1, 9, 10, 15, 16, 25
1                           7, 19, 20, 22

Table 6. Demands Corresponding to Test of Infant Motor Performance (TIMP) Items Observed Two to Four Times in a Subject

                    Demands
                    Corresponding to TIMP
                    Item Numbers
No. of Subjects     Observed 2-4 Times

11                  15
10                  13
 7                  16
 6                   2, 9, 24
 5                   5, 19
 4                   1, 6
 3                   4, 7, 8, 10, 21, 25
 1                  11, 14, 18

The data were examined for trends and relationships regarding demands experienced by infants with different characteristics (Tab. 7). None of the sample characteristic variables studied (age, sex, degree of medical risk, level of education of the caregiver, or ethnicity) were related to significant differences in the average number of demands received per minute. The mean rate of total demands received was 3.0 (SD = 1.4, range = 2.13-7.0) demands per minute for the preterm age group and 3.4 (SD = 0.6, range = 2.47-4.47) demands per minute for the postterm age group, with no statistical difference between the two groups (t = -0.93, P = .36).

The majority of infants in the sample received environmental demands corresponding to seven TIMP items (2, 9, 13, 15, 16, 19, and 24). The two TIMP items not represented in observed demands were numbers 3 and 23. All but two of the TIMP Elicited Scale items were represented in the observed demands during the caregiving routines of preterm infants (32 weeks postconceptional age to 13 weeks postterm age) and full-term infants (birth to 13 weeks of age) during dressing, bathing, and play activities. The typical infant received environmental demands that related to 37% of the TIMP items. No differences were identified based on subject characteristics except for length of the caregiving routine.

Discussion

We investigated the ecological relevance of the TIMP by assessing whether the demands placed on infants during item administration (eg, handling procedures) can be related to the demands of caregiving in natural settings. Bronfenbrenner[14] contended that research is ecologically valid if it is carried out in a natural setting and involves objects and activities of daily life. Furthermore, "ecological validity" has been deemed as "the extent to which the environment experienced by the subjects in a scientific investigation has the properties it is supposed or assumed to have by the investigator."[14(p515)] The following findings support the notion of ecological validity of the TIMP Elicited Scale items: (1) almost all (92%) of the Elicited Scale items were represented at least once in the observed demands placed on the infants in this sample during caregiving, (2) the average infant experienced demands corresponding to 37% of TIMP Elicited Scale items during caregiving, and (3) approximately 50% of all environmental demands placed on the infants corresponded to TIMP items. Many of the other environmental demands not related to the demands of TIMP item administration procedures involved a unilateral movement of an extremity or resembled a TIMP item demand but was not an exact approximation of the TIMP item demand as it had been defined in the coding scheme.

Neonatal and infant tests published in the past 2 decades usually have not been studied for ecological validity. Test items have not been examined for their relevance to function in natural settings. For example, the Movement Assessment of Infants,[15] the Neurological Examination of the Full-Term Newborn Infant,[16] the Neonatal Neurobehavioral Examination,[17] and the Neurological Assessment of the Preterm and Full-Term Newborn Infant[18] primarily have items that require physical handling of the infant to elicit a measured motor response (often a reflex). The relationship between the physical handling of the infant on these tests and the functional or environmental demands experienced by an infant in daily life has not been investigated.

Newer infant motor tests have focused on an observational approach. For example, the developers of the Alberta Infant Motor Scale (AIMS) believe that infants should be observed as they move about freely with minimal handling and that the therapist should respond to the infant's cues rather than expect the infant to respond to demands imposed by the therapist.[19] Functional movement is thought to be assessed by observing the infant move within a task context determined by the infant's own interest in the surroundings.[19] The AIMS has not been studied for ecological validity. Furthermore, using observation of spontaneously generated infant movement as the sole means for evaluating an infant's postural control seems to us to be valid only for infants who are old enough, strong enough, and motivated enough to move independently. A newborn infant may not be able to move freely or change positions in space, but may demonstrate use of postural control strategies in response to challenges or external demands that are encountered in his or her caregiving environment. Ninety-two percent of the TIMP Elicited Scale items corresponded to typical functional demands that infants received during caregiving in the natural setting, demonstrating the ecological relevance of the TIMP Elicited Scale administration procedures.

This investigation is the first step needed to demonstrate that the TIMP is ecologically valid and truly assesses infant postures and movements in response to interactions with the environment during functional activities, which is a question concerning construct validity. Further study is needed to demonstrate whether infants' responses to the administration of the TIMP relate to the responses from naturally occurring demands during daily life to substantiate the ecological validity of the TIMP. By demonstrating that the Elicited Scale administration procedures are related to functional demands encountered in the natural setting, we have provided evidence that the conceptual foundation of the TIMP is sound. Our study, therefore, also lends support to the construct validity of the TIMP.

Conclusion

This investigation is the first to examine the ecological relevance of the TIMP. The majority of TIMP Elicited Scale items have been found to be reflective of environmental demands placed on a sample of 22 infants with diverse backgrounds by their caregivers during the functional activities of dressing, bathing, and play. Our results, therefore, provide evidence supporting the construct validity of the TIMP, as well as the notion that the TIMP is ecologically valid. Although the demands of the TIMP appear to be similar to demands imposed during daily life experiences in natural settings, future studies are needed to examine the relationship between the responses to these environmental demands and responses to administration of the TIMP in order to strengthen the construct validity of the TIMP and substantiate its ecological validity.

Acknowledgments

Appreciation is expressed to Russell E Carter, EdD, PT, Carolyn B Heriza, EdD, PT, and Sandra B Levine, PhD, PT, for assistance in the design and analysis of this investigation; to Kathy Gartlan, PT, for participation in establishing interrater reliability; to Gail Liberg, OTR/L, Manohar Rathi, MD, Dharmapuri Vidyasagar, MD, and the nursing staff of the special care nurseries at Christ and University of Illinois Hospitals for assistance in subject recruitment; to the caregivers and parents of our infant subjects for their willingness to allow themselves and the infants to be videotaped and observed during caregiving activities; and to Robert J Palisano, ScD, PT, and Darl W Vander Linden, PhD, PT, for their review of an earlier version of this manuscript.

Table 7. Mean Rate of Demands Related to Test of Infant Motor Performance (TIMP) and All Demands for Various Groups of Subjects

Variable                    Group                N   [bar] X   SD

Mean rate of
TIMP-related demands

   Age                      Preterm             10     1.4     1.1
                            Postterm            12     1.7     0.5

   Sex                      Male                11     1.6     0.5
                            Female              11     1.6     1.1

   Degree of medical risk   Low risk            11     1.8     1.0
                            High risk           11     1.3     0.6

   Educational level        < High school       2     2.8     0.5
    of caregiver             High school        2     2.9     0.1
                             Some college       8     3.3     0.8
                             College degree   1.0     3.3     1.4

   Ethnicity                Hispanic American    7     1.4     0.7
                            African American     7     1.5     0.7
                            European American    8     1.9     1.1

Mean rate of all demands

   Age                      Preterm             10     3.0     1.4
                            Postterm            12     3.4     0.6

   Sex                      Male                11     3.2     0.7
                            Female              11     3.3     1.4

   Degree of medical risk   Low risk            11     3.4     1.3
                            High risk           11     3.0     0.8

   Educational level        < High School        2     2.8     0.5
    of caregiver            High School          2     2.9     1.1
                            Some college         8     3.3     0.8
                            College degree      10     3.3     1.4

   Ethnicity                Hispanic American    7     3.0     0.6
                            African American     7     3.0     0.6
                            European American    8     3.6     1.6

                                                t or F
Variable                    Group               Ratio    P

Mean rate of
TIMP-related demands

   Age                      Preterm             -0.88    .39
                            Postterm

   Sex                      Male                -0.03    .98
                            Female

   Degree of medical risk   Low risk             1.48    .15
                            High risk

   Educational level        < High school
    of caregiver            High school          0.62    .55
                            Some college
                            College degree

   Ethnicity                Hispanic American    0.74    .49
                            African American
                            European American

Mean rate of all demands

   Age                      Preterm             -0.93    .36
                            Postterm

   Sex                      Male                -0.29    .77
                            Female

   Degree of medical risk   Low risk             0.85    .40
                            High risk

   Educational level        < High School        0.25    .78
    of caregiver            High School
                            Some college
                            College degree

   Ethnicity                Hispanic American    0.73    .49
                            African American
                            European American

(*) SPSS Inc, 444 N Michigan Ave, Chicago, IL 60611.

References

[1] Campbell SK, Osten ET, Kolobe THA, Fisher AG. Development of the Test of Infant Motor Performance. Physical Medicine and Rehabilitation Clinics of North America. 1993;4:541-550.

[2] Campbell SK, Kolobe THA, Osten ET, et al. Construct validity of the Test of Infant Motor Performance. Phys Ther. 1995;75:585-596.

[3] Campbell SK, Kolobe TH, Girolami GL. Development of a Neonatal Motor Scale: The Test of Infant Motor Performance. Final Report to the Foundation for Physical Therapy Inc, June 29, 1993. Alexandria, Va: Foundation for Physical Therapy Inc; 1993.

[4] Task Force on Standards for Measurement in Physical Therapy. Standards for tests and measurements in physical therapy practice. Phys Ther. 1991;71:589-622.

[5] Rothstein JM. Measurement and clinical practice: theory and application. In: Rothstein JM, ed. Measurement in Physical Therapy. New York, NY: Churchill Livingstone Inc; 1985:1-46.

[6] Anastasi A. Psychological Testing. 6th ed. New York, NY: Macmillan Publishing USA; 1988.

[7] Dunn WW. Validity. Physical and Occupational Therapy in Pediatrics. 1989;9(1):149-168.

[8] Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice. East Norwalk, Conn: Appleton & Lange; 1993.

[9] Osten ET. Examination of the Rater Reliability of the Test of Infant Motor Performance. Chicago, Ill: Department of Occupational Therapy, University of Illinois at Chicago; 1993. Master's thesis.

[10] Hobel CJ. Identification of the patient at risk. In: Bolognese RJ, Schwarz RH, Schneider J, eds. Perinatal Medicine: Management of the High-Risk Fetus and Neonate. 2nd ed. Baltimore, Md: Williams & Wilkins; 1982:3-28.

[11] Hobel CJ, Hyvarinen MA, Okada DM, Oh W. Prenatal and intrapartum high-risk screening, I: prediction of the high-risk neonate. Am J Obstet Gynecol. 1973;117:1-9.

[12] Bakeman R, Gottman JM. Observing Interaction: An Introduction to Sequential Analysis. Cambridge, Mass: Cambridge University Press; 1986.

[13] Bakeman R, Gottman JM. Applying observational methods: a systematic view. In: Osofsky JD, ed. Handbook of Infant Development. 2nd ed. New York, NY: John Wiley & Sons Inc; 1987:818-854.

[14] Bronfenbrenner U. Toward an experimental ecology of human development. Am Psychol. 1977;32:513-531.

[15] Chandler LS, Andrew MS, Swanson MW. Movement Assessment of Infants: A Manual. Rolling Bay, Wash: Chandler, Andrew, and Swanson; 1980.

[16] Prechtl HFR. The Neurological Examination of the Full-Term Newborn Infant. Philadelphia, Pa: JB Lippincott Co; 1977.

[17] Morgan AM, Koch V, Lee V, Aldag J. Neonatal neurobehavioral examination: a new instrument for quantitative analysis of neonatal neurological status. Phys Ther. 1988;68:1352-1358.

[18] Dubowitz L, Dubowitz V. The Neurological Assessment of the Preterm and Full-Term Newborn Infant. Philadelphia, Pa: JB Lippincott Co; 1981.

[19] Piper MC, Darrah J. Motor Assessment of the Developing Infant. Philadelphia, Pa: WB Saunders Co; 1994.

Appendix.

Environmental Demands Corresponding to Test of Infant Motor Performance (TIMP) Elicited Scale Item Administration Descriptions

1. While infant is held vertically upright and head in midline, the caretaker makes eye contact with infant (approximately 38 cm [15 in] away from infant's face) and moves her face from side to side.

2. Infant is held in an upright sitting position with no support given to the head.

3. Infant is held in an upright sitting position with head and neck supported in midline, and a gentle upward traction is given to the head.

4. Infant is held in an upright sitting position and head is moved into flexion or falls forward at least 45 degrees.

5. Infant is held in an upright sitting position and head is moved into extension or falls backward at least 45 degrees.

6. Infant is lowered backward from a sitting position with the head unsupported.

7. While infant is in a supine position, the caregiver passively rotates the infant's head so the jaw is nearly directly over the shoulder and maintains the head to the one side for at least 2 seconds.

8. With infant in a supine position, the caregiver moves the infant's head to midline, releases the head in midline, and presents no visual stimulation.

9. With infant in a supine position and head in midline, the caregiver presents a visual stimulus to the infant in midline approximately 20-38 cm (8-15 in) away from face.

10. With infant in a supine position and head turned either side, a visual stimulus is presented within 20-38 cm away from face so that the infant can make eye contact, and then the visual stimulus is moved toward the opposite side.

11. With infant in a supine position, the caregiver covers both of the infant's eyes with a cloth.

12. With infant in a supine position, the infant's entire face is covered with a cloth by the caregiver.

13. With the infant in a supine position, the caregiver flexes both of the infant's hips and knees up toward the abdomen and then releases them.

14. With infant in a supine position, the caregiver adducts one leg across the midline of the infant's body.

15. With infant in a supine position, the caregiver adducts one arm in a pulling manner across the midline of the infant's body.

16. With the infant in a supine position, the infant is pulled to a sitting position by the arms distal to the elbow.

17. The infant is placed in a side-lying position with caregiver's hand on pelvis.

18. The infant is placed in a side-lying position, and the infant's upper body is lowered relative to the legs.

19. The infant is suspended belly down with support given at the ribs or abdomen.

20. With infant in a prone position, the caregiver flexes both of the infant's hips and knees up under the trunk and then releases them.

21. Infant is placed in a prone position with the face down in midline.

22. With infant in a prone position and head turned to one side, an auditory stimulus is given near the back of skull.

23. With infant in a prone position, the caregiver extends both of the infant's arms along sides of the body.

24. The caregiver supports the infant in vertical suspension and then vertically lowers infant to a support surface.

25. Infant is supported in a laterally tilted vertical suspended position for at least 1 second, and head is not supported.

26. The caregiver moves the infant or positions the infant in a manner not previously defined.

ME Murney, PT, PCS, was a student at the University of Illinois at Chicago at the time this research was completed in partial fulfillment of the requirements for her Master of Science degree in physical therapy. She is currently a pediatric clinical specialist in private practice. Address all correspondence to Ms Murney at 8540 S Lamon Ave, Burbank, IL 60459 (USA).

SK Campbell, PhD, PT, FAPTA, is Professor and Director of Graduate Studies, Department of Physical Therapy, University of Illinois at Chicago.

This study was approved by the University of Illinois at Chicago (H-93-599) Institutional Human Subjects Review Board.

During the course of this project, the authors were partially supported by Grant MCJ IL 179590, Maternal and Child Health Bureau, US Public Health Service.