Abstract
☑ Earn CEs Online
by answering questions on this article.
For more information, visit:
http://my.aerbvi.org/AER-Store/Publications/BKctl/ViewDetails/SKU/AER.
Problem and significance
No tool or assessment protocol exists that can comprehensively measure a young child’s orientation and mobility (O&M) skills. The Preschool Orientation and Mobility Screening was developed in 1987, and it provided the first tool that allowed professionals to determine if further evaluation and services were needed (Hill et al., 1992). However, since this tool was not sufficiently sensitive to developmental changes that would have allowed it to be used for progress monitoring, a variety of checklists have been developed that supplement it (Dodson-Burk & Roman-Lantzy, 2012; Pogrund et al., 2012). Modern testing theory indicates that practitioners cannot uniformly measure the efficacy of their interventions using checklists since checklists do not use precise enough measures (Dale & Salt, 2007). Instead, there is a need for the development of a high-stakes assessment that can serve as a snapshot of a student’s skills to determine if he or she needs intervention or that can be repeated to monitor for progress (Dale & Salt, 2007; Morash & McKerracher, 2017). A variety of tools partially address intervention efficacy, but these are either limited in scope, measure only motor or visual skills, or they are not appropriate for very young children (Leong, 1996).
Under the Individuals with Disabilities Act (IDEA), O&M services are clearly included in education: part B specifies O&M as a “related service,” and part C identifies the need for infants and toddlers to receive interventions (IDEA, 2004). Professionals are required by state departments of education to conduct preliminary and ongoing assessments that are reliable and valid (IDEA, 2004). Because of these requirements, it is necessary that O&M tools be developed that are reliable and valid for children aged birth to 6 years. The Birth to 6 Orientation and Mobility Skills Inventory (B6OMSI) has been drafted and is in use by many O&M specialists but has not yet been demonstrated to be reliable and valid (New Mexico School for the Blind and Visually Impaired, 2020).
Literature review
Visual information stimulates the motor development of young children; it guides visually directed reach, as well as provides the child continuous feedback about their movement in space (Papadopoulos & Koustriava, 2011). Vision plays the primary role as a youngster identifies an object to move toward; it plays a secondary role as the child refines his or her movement patterns through observation of both others and self while the child is engaging in movement (Brambring, 2001). Published reviews of the literature have found that children who are visually impaired (i.e., those who are blind or have low vision) develop motor skills later than do their sighted peers (Houwen et al., 2009; Leong, 1996). Yet, a critical finding from an article by Houwen et al. (2009) is that children’s level of visual impairment, age, and gender are not predictive of their motor scores, meaning the test has no predictive validity and cannot determine the age at which such skills are developed. Other research has demonstrated that children who are visually impaired meet developmental milestones at different ages than do their sighted peers and, in some cases, in different orders (Ferrell et al., 1998).
Research conducted in the area of gross motor skills involving school-age children with visual impairments found that these children lagged significantly behind their same-aged peers without visual impairments prior to intervention (Wagner et al., 2013). It has also been found that the development of locomotor skills in children with visual impairments include shorter strides, slower walking speeds, and other differences in gait parameters (Hallemans et al., 2011). From these studies, it is apparent that children with visual impairments may not, without intervention, develop motor skills in the same ways that their peers with vision do.
In cognitive assessments, the Verbal Scale from the Weschler Intelligence Scale for Children has been used to estimate the IQ of children who are visually impaired because, as a verbal test, it appeared no modification in the protocol was needed for administration (Miller & Skillman, 2003). However, a recent study investigated the reliability of this practice, compared to the use of the Blind Learning Aptitude Test (BLAT), which also tests verbal skills but was designed for children with visual impairment. The results demonstrated that the Vocabulary subtest of the Wechsler Abbreviated Scale of Intelligence (WASI) scores did not have a test–retest reliability that is appropriate for a high-stakes test (Cronbach’s α .85) when administered to people who are visually impaired, while the BLAT scores were reliable (Cronbach’s α .92; Morash & McKerracher, 2017). These findings indicate that the WASI is not yielding accurate scores, despite the common practice of using this type of assessment (Morash & McKerracher, 2017). For these reasons, it is important that assessments used for children who are visually impaired be developed, standardized, and tested on this specific population, rather than adapted from tests that were developed for use with sighted children.
Historically, the practice of O&M was developed by the U.S. governmental entity now known as the Department of Veterans Affairs, its practice later spread to schools (Welsh, 2005). Leong (1996) found a history of articles supporting O&M training for children receiving early intervention services, but noted that, as the youngest branch of the field of O&M, its practice increased primarily during the early 1990s. During that time period, the first screening tool was created (Hill et al., 1992). Recommendations indicated that O&M services should begin before kindergarten, generally at 4–5 years of age, and they called for further research in instructional practices and the development of an O&M curriculum for young learners (Leong, 1996). The curriculum for Teaching Age-appropriate Purposeful Skills was developed and includes a more thorough assessment; however, the tool still lacked a sensitive scoring framework (i.e., one that can detect small changes in performance that are a result of child development, which are appropriate for progress monitoring; Pogrund et al., 2012). More recent studies have further built on those frameworks, demonstrating that early intervention O&M services would be beneficial for the development of very young children, preventing delays rather than providing rehabilitation after a child falls behind (Celano et al., 2016; Dale & Salt, 2007). Ely and Ostrosky (2018) found, through a comprehensive review of the literature, that the key practices of early intervention were demonstrated to be important for intervention with young children who are visually impaired. These practices are family-centered relationship–based interventions, natural environments, child learning, adult learning, and quality team practices (Ely & Ostrosky, 2018).
Sensitivity to change is also an important test characteristic when an assessment will be repeated for the purpose of progress monitoring (Dale & Salt, 2007). A test that is sensitive to child development is expected to yield higher scores in older children, except in cases where something has occurred that would account for regression (Petermann & Macha, 2008). When age is not predictive of motor scores for children who are visually impaired, it is an indicator that the assessment is not valid for this population (Houwen et al., 2010, 2009). The likely reason the outcomes do not match the theoretical expectation of scores increasing throughout a child’s developmental period is that the tests being used are not sensitive to maturation when administered to children with visual impairments (Reimer et al., 2015).
Previously, a variety of assessment tools have been developed that are broadly related to the O&M skills of children with visual impairments (Brown, 1978; Dodson-Burk & Roman-Lantzy, 2012; Pogrund et al., 2012). However, some of these tools assess motor development, other concept development, and many assess only school-age children and are not sensitive enough for tracking the developmental changes that are used for early intervention progress monitoring. The test designed for O&M in preschool children is a “screener” (Dodson-Burk & Roman-Lantzy, 2012), thus it is not appropriate as a high-stakes test (a test in which the score can stand alone to represent a person’s ability on a domain) or as a test for progress monitoring, which would detect change over time (Dale & Salt, 2007; Morash & McKerracher, 2017). A screener is a test that identifies children who need further assessment and can help professionals set the first goals for students. Teachers and specialists find themselves creating their own assessments to conduct the recommended follow-up mandated under IDEA, which, incidentally, means the tools they create are ineffective for comparisons between students, alignments with benchmarks, or evaluations of program efficacy. A research-based assessment created specifically for children who have visual impairments and that is demonstrated to be reliable and valid for this population would fill this gap. The B6OMSI has a precise scale for measuring the abilities of young children, and it can more sensitively detect change over time than any previously available tools (New Mexico School for the Blind and Visually Impaired, 2020). The B6OMSI assessment includes specific skills to elicit from the child, a clear rating scale for each skill, and tabulated summative scores for each of 6 core categories: orientation, mobility, concepts, sensory, cane, and communication. The B6OMSI also meets the 11 recommended practices for assessment identified by the Council for Exceptional Children’s Division for Early Childhood and, while these practices are intended for children with developmental disabilities, they may also be a sign of a high quality in other early childhood assessments (Division for Early Childhood, 2014).
Research question
Previous research (Wall Emerson & Corn, 2006) began to explore O&M for children; however, the scope of this study was focused on what should be taught to school-age children and what preparation is needed for pre-service professionals. Although the study’s results called for assessment under a variety of conditions, it did not provide a comprehensive list of what domains should be measured or how. For very young children, assessment should include representation of the eight domains for O&M instruction identified by Wall Emerson and Corn (2006) with a set of age-appropriate subskills from each, as well as certain developmental milestones that may not have a parallel in instructional goals such as noting the integration of reflexes during assessment as it relates to a child’s overall development, though this is not something that could be taught directly (Brown, 1978). To address the need for an assessment to provide this missing information, the study presented here recruited experts in O&M for young children to identify the key domains that would inform the need for, and progress in, O&M instruction for young children.
The purpose of the current study was to determine the face validity and content validity of the B6OMSI. The results of this study seek to answer the following specific questions: 1. Does the B6OMSI test the relevant areas of O&M for very young children? That is, is the test comprehensive in the domains it measures? 2. Are all items on the B6OMSI contributing to the total picture of a child’s O&M skills? That is, is the test focused on the domains of O&M, without extraneous measures of other areas? 3. Is the validity sufficiently strong for the tool to be used as a high-stakes assessment? That is, is the face and content validity established with >90% consensus, so that the test score can stand alone (Morash & McKerracher, 2017)?
Method
Design
In light of modern test development, a Delphi study is an appropriate way to generate an item pool for this relatively new domain in the field of O&M, and such studies have been found to be an appropriate method in health and human services research (Falzarano & Zipp, 2013). A Delphi study uses consensus among a sample of experts to establish a valid list of answers to the research question, which demonstrates face validity and content validity and is an appropriate research design when there is insufficient literature on the topic to establish validity in alternate ways (Falzarano & Zipp, 2013). Before data were collected, approval was obtained from the Human Subjects Institutional Review Board at Western Michigan University (approval #17-06-02).
Participants
Duration of professional experience of the experts.
Countries of origin, certifications, populations served, and settings for the experts.
Materials and procedures
Each round of the survey was conducted using an online web-based form, which created with Google software suite. Those participants who volunteered at conferences had the option to complete the background and first round on paper if Wi-Fi or computers were not readily available; these individuals participated in follow-up rounds online.
After the participant pool was identified from the experts who volunteered, the first round of the survey was emailed to participants. Researchers who have previously used this study design found that giving participants a list of topics in the first round of a Delphi study was a limitation since it may influence responses toward only the suggested topics (Wall Emerson & Corn, 2006). To avoid swaying participants, the first round of this survey was conducted as an open-ended questionnaire, and participants were asked to generate their own lists of what items they thought needed to be assessed. Participants in the first round were asked to answer the questions from their professional expertise, without looking at existing tools. Items on the previous B6OMSI were used as the a priori codes ensuring their inclusion during the second round, without priming the experts’ memory with the items in that or other existing tools. This process avoided the risk of omission of new items by prompting the experts into presuming that the tools they regularly use to be complete. It also ensured the resulting assessment tool would benefit from the previous work of criterion referencing the B6OMSI to state standards (New Mexico School for the Blind and Visually Impaired, 2020). Participants were asked to complete the first round within two weeks. The survey utilized a variety of vignettes to encourage answers for the diverse population of young children with visual impairments. These descriptions included three ages (prior to crawling, 3 years, and 6 years); two vision levels (blind and a visual acuity of 20/100); three combinations of multiple impairments (visual impairment with a physical disability, visual impairment with a cognitive disability, and deaf blindness); children with cortical visual impairment; and an unrestricted question for anything else the professional would include in an assessment. To reduce the time amount of time required to answer the questions, participants were instructed that they did not need to repeat answers in multiple categories. Answers to the survey that were completed on hardcopy at conferences were entered into the web-based form. A comprehensive list of skills was generated from all of the survey responses, and they were qualitatively coded using the a priori codes from the previous B6OMSI and emergent codes for new themes found in the dataset. Two researchers coded the items before comparing their findings to discuss and agree on how to resolve any differences.
Round two of the survey was created from the list of skills generated through that qualitative process using the responses of the first round and the draft tool. Each potential skill was presented with the question, “Should this item be assessed?” and asked participants to provide their rationale for the answer. The second round of the survey also concluded with an open-ended question asking respondents to include any other items that should be included in the assessment, which allowed participants to add items they thought of while reading the list. Participants were reminded that they were working toward consensus. The strength of findings in a Delphi study is based on the level of consensus rather than the size of the expert pool (Avella, 2016; Falzarano & Zipp, 2013; Powell, 2003).
Subsequent rounds of the survey included the revised list, based on participant consensus. Rationale were included and disseminated along with each question. After each round, the level of agreement was tallied, and items were accepted, rejected, or sent back to the group for another round. A consensus of 90% is appropriate for high-stakes tests (Morash & McKerracher, 2017), and 70% for other tests (Falzarano & Zipp, 2013). Items with less than 30% consensus were considered rejected by consensus in these rounds, with the exception of newly submitted items. Intermediate values were returned to the group for further consideration. Participants were asked to consider rationale and change their answer, if appropriate, based on their expert professional judgments, given the reasons provided by other panelists.
After all rounds of the Delphi study were completed, the final list was compared to items on the B6OMSI so that new items could be added and irrelevant skills could be removed. For the instrument to serve as a high-stakes test, all items that achieved 90% consensus needed to be included.
Results
Item creation and validation.
*Not valid at a high-stakes level.
Discussion
The purpose of this study was to answer three research questions that addressed whether the B6OMSI was comprehensive, without the need to assess extraneous skills, and valid, at a high-stakes level, so the results produced by the instrument could be considered without the need for additional information. The first research question was answered when the Delphi study showed that the existing tool was not comprehensive. The generation of seven new items showed that, although the existing tool comprised 83.8% of the new test, O&M specialists saw benefit in including some additional content. The new research edition of the assessment tool (B6OMSI-RE) will include all of the items that have content validity as established by the panel of experts. Thus, B6OMSI-RE will serve as a comprehensive tool. Regarding the second research question, it was determined that all items on the assessment have face validity, according to the expert panel. Furthermore, because the agreement among experts was above 90%, these items are valid at a high-stakes level. The administration of a high-stakes test will save practitioners time in data collection and progress monitoring because this single test score can represent a meaningful snapshot of student functioning. By using a Delphi study, discussion and consensus were facilitated among a larger pool of experts who could efficiently coauthor a tool.
The new items fit well within the existing test domains. Two new items were generated in the orientation domain: use of landmarks and traffic. These items also increased the test’s representation of the decision-making domain previously identified by Wall Emerson and Corn (2006). Two items were also generated for the B6OMSI-RE’s sensory domain: explores objects and proprioception. The concept domain added ordinal sequencing, and the mobility domain added gross motor skills. Upon review, the B6OMSI-RE represents early childhood components from all eight domains (spatial concepts, spatial skills, perceptual skills, motor skills, mobility skills, environmental knowledge, decision-making, and interpersonal skills) that were identified by Wall Emerson and Corn (2006). Some of the subskills from those domains, such as arranging for rides, hiring and firing drivers, and intersection or traffic flow patterns, are not represented on the test because they are not age appropriate for young children. The B6OMSI-RE is more robust in other skills such as early motor development, early cane skills, and echolocation skills, which are included in addition to what had been previously identified in the literature.
Technology was an assessment item that was agreed upon by most participants, but it did not rise to the high-stakes level. Participants who opposed the inclusion of a technology assessment stated that children needed to be proficient in their own skills before introducing technology; others added that it was culturally inappropriate for young children to be using technology. In response to this division, technology was added to the background information portion of the assessment, rather than to the scored portion. It is the author’s belief that this placement for the technology element will provide the information where it is helpful without creating pressure for an O&M specialist to teach and monitor a skill that may not be relevant in some settings.
The items generated through this consensus process form the content of the B6OMSI-RE. Through the process of generating test items from a pool of experts, with a wide range of stimulus questions using diverse student vignettes, the new tool will provide a comprehensive snapshot of a young child’s O&M skills development. Because the experts come from a wide geographic range and practice in different types of communities, this tool can be appropriately used with young children in any setting where O&M specialists have similar training.
Limitations of the study
Just over 70% of participants in the first round continued through all rounds of the Delphi study. Most of the attrition occurred between the first two rounds, when the qualitative coding was occurring, which resulted in a longer time between rounds.
This study recruited O&M specialists with a wide range of credentials, including Certified Orientation and Mobility Specialist (COMS), National Orientation and Mobility Certification (NOMC), and national certifications according to each participant’s country of origin. Both participants who held NOMC certification stopped responding to follow-up messages before the end of the study. However, since all items that were generated in the first round were included in the research edition of the assessment tool, their contributions were not lost.
Conclusion and directions for future research
Historically, assessments for O&M have been inappropriate for young children or not sensitive enough to use effectively for progress monitoring. Tools have sometimes been developed by a few professionals, creating risks of missing assessment content. This study attempted to create a comprehensive list of skills that should be assessed for young children who are visually impaired who may need or are receiving O&M services. The intent was to build an assessment tool based on the B6OMSI (New Mexico School for the Blind and Visually Impaired, 2020), revising it to include all items where experts identified content validity.
This study has established face validity of all skills included in the B6OMSI-RE through a process of content validity by evaluating the level of agreement among experts. This exercise was a critical first step in the larger process of developing a valid and reliable O&M assessment tool for young children, which laid the foundation for the creation of B6OMSI-RE, a high-stakes assessment tool. Future studies should explore other key parametric properties including reliability and construct validity measures. Field testing to develop procedures that are both standard and efficient for the practitioner will be needed to prepare the assessment for widespread adoption. A longitudinal study would be helpful to the field in identifying the sensitivity to developmental changes, as well as in identifying benchmarks for various ages for subpopulations. This line of research inquiry should be replicated for other O&M assessments used in the field of visual impairment to establish multiple research-based tools. Practitioners can participate in future reliability, validity, or longitudinal studies that arise to contribute to the further refinement of tools they use. However, although research on this tool continues, the working copy is available from the New Mexico School for the Blind website, to ensure O&M specialists will always have access to the most complete version of the B6OMSI-RE. Practitioners should look for published data on the validity and reliability of any assessment tools they are using in their initial and ongoing evaluation procedures.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
