BDI-2: An Inventory of Advantages
The Battelle Developmental Inventory 2nd Edition (BDI-2) is a comprehensive test of early childhood developmental skills. The original version was developed at the Battelle Memorial Institute (Columbus, Ohio laboratory) in 1973 through a contract with the U.S. Department of Education. The tool was designed to evaluate early childhood programs by assessing the developmental progress of children with developmental handicaps.
The BDI-2 is an ambitious test in that it attempts to be a comprehensive assessment of early developmental skills by measuring the following 5 domains and 13 subdomains of children from birth through 7 years, 11 months:
- Self Care
- Personal Responsibility
- Adult Interaction
- Peer Interaction
- Self-Concept and Social Role
- Receptive Communication
- Expressive Communication
- Gross Motor
- Fine Motor
- Perceptual Motor
- Attention and Memory
- Reasoning and Academic Skills
- Perception and Concepts
The BDI-2 also has a separate screening instrument with cut scores for 3 standard deviation options (1.0,1.5, 2.0).
Keep in mind the breadth of the BDI-2 when you compare it to other early childhood tests on the market. Do they measure all of the same domains? Will you need to cobble together other tests (each with different norm samples) to get all of the domains measured? Is your screener linked to test? These are important considerations because scoring additional tests (each with their own scoring rules and norm tables) is costly and time consuming. That said, the BDI-2 communication domain is just that… It measures communication as it’s represented at the different ages, which means it’s a measure of both verbal and nonverbal communication. So an additional test of verbal comprehension may need to be administered in some cases. As an astute and fastidious examiner you need to observe whether there is a pattern of errors on the verbal items that’s being “washed out” by good performance on the nonverbal items yielding a score that’s not a unified representation of the child’s overall communication skills. Or if the parent/caretaker has specific concerns about language development then that’s your cue to make sure it’s being adequately measured. Obviously, articulation would be another specific area that would require a separate tool or clinical judgement.
First of all when evaluating young children the examiner needs two things: Manipulatives to engage the child, and theatrical skills. The BDI-2 will provide you the former but you must bring the latter. The manipulatives come in a “tackle box” that might be a challenge for some people to lug around. However, if you’re testing children of a particular age range like 0-2 years of age (i.e. Part C) then you can remove items intended for the older children to cut down on the mass. Keep in mind that you’ll need items that kids in your age range can’t do in order to get a ceiling. Conversely, if you’re testing older kids, and remove some younger age items, keep some easy (young items) so that you’re able to establish a basal.
Riverside sells the manipulatives but you can also purchase them yourself. If you go that route you must buy items that are functionally identical. For example, the blue blocks are 1.25 inches. One inch blocks would not meet that standard. For this reason Riverside also provides the option of purchasing the “difficult to find” items. I don’t want to spend too much time on purchasing issues but it’s good to know that when you lose the tennis ball that you can just go to your local discount store and buy one. I heard a story from some people at Riverside (possibly apocryphal) where a customer called wanting to purchase a lost penny from the original BDI. Don’t be that person that tries to buy a penny.
Being a flexible person (mentally and physically) is key to evaluating young children. Having a test with flexible administration options means that you can gain standardized information about the child that you couldn’t otherwise obtain with a test that relied solely on structured test items. For example, with infants it would be difficult (or at least time consuming) to structure a situation where you could rate breastfeeding behavior. Plus it would be kind of creepy.
The BDI-2 has three administration options: Structured, Observation, and Interview. Structured is still going to the most preferred because it’s always better to directly document a behavior than to rely on the observation or reporting skills of another person. It’s not so much that people are lying (although that can be the case); instead, it’s that communication is fraught with error which in normal conversation is trivial. But in an assessment context those nuanced misunderstandings can create measurement error. The advantages of having observation and interview far outweigh any disadvantages. In fact, the validity of the test is likely improved because we have access to items that we otherwise wouldn’t be able to measure at all. The point of bringing your attention to a hierarchy of desirability (for lack of a better term) is because, when you have a choice between the options, the structured is going to be most preferred unless you have reason to believe, based on the circumstances of your particular case, that the other options give a better data. Your professional judgement as a flexible examiner will often need to be consulted. Get used to it.
The items on the BDI-2 are “actionable” meaning that you can use the items the child was unable to master as goals or objectives in an IEP/IFSP. Since it’s a skills based test then it’s perfectly reasonable to teach the skills being measured. If the child can’t walk up the stairs then you’re going to need to teach the child to walk up the stairs! You can use the evaluation criteria from the individual items the child didn’t master, but you’re not going to want to use the actual BDI-2 items. For example, if the skill was stacking blocks then you’d want to use conceptually similar items (and likely more than one size/shape). You’re not teaching the test you’re teaching the skills the test items represent.
Scoring and Reporting
The BDI-2 has a straightforward scoring system that, when using the rubric for each item, will yield a 2,1 or zero. A 2 means the child has achieved the milestone, a 1 means the milestone is emerging, and a 0 means the milestone has not yet been achieved. This simple scoring is used on all subdomains which gives you a clear picture of the child’s specific skills with a simple glance at the raw scores. This becomes useful when it comes time to write the IEP/IFSP because as noted in the above section on “actionable items” you can simply scan the record form for items that are a 1 or zero.
Once you score the child’s record form the computer software (of which there are many options that I’ll discuss in a later article) provides numerous reporting options. Besides obtaining the child’s scores you can also get class skills status report that will describe how each class is doing in each domain, numerous government reports that aggregate information for Head Start, IDEA and NCLB, in addition to district level reports that can aggregate scores for each school, classroom or individual student. It will also generate “family reports” that have simple descriptions of the child’s performance complete with easy to read charts of performance, professional narrative reports with more specific descriptions of performance, individual summary reports of all the many scores the BDI-2 provides, and a Child IEP guide to help with the identifying actionable items.
In short, the BDI-2 provides more than just a test to determine special education eligibility. There are Head Start programs that give the test up to three times in a year to show progress; along with using those scores to determine eligibility, and drive goals. If the idea of testing a child three times in a single year seems daunting keep in mind that if the evaluator has worked with the child the test is much quicker to administer because items that require observation would be trivial to complete.
The BDI went through a 5 year development process that included updating the items, refining the scoring criteria, revising the structure (domains/subdomains), and utilizing a nationally representative sample of 2500 children that closely matched the US census to create the norms. The end result is a test with high reliability (domains .90-.96) and well documented validity. In fact the BDI-2, due it’s comprehensive structure, even provides sensitivity and specificity statistics. In brief, sensitivity is how well does the test find kids that meet the criteria while specificity is how well does the test not find kinds that don’t meet the criteria. It’s incredibly rare to find a test that reports this kind of information so if you’ve never heard of it then that is probably why. The recommended level of accuracy for a test should be at least .75 according to Kingslake (1983). With a sensitivity of .83 and specificity of .85 the BDI-2 exceeds the recommended level of accuracy.
The BDI-2 has many advantages that I have attempted to outline in this article. This is by no means a complete listing. The test also has a spanish translation and many scoring options including a web based Data Manager and a recently released Mobile Data Solution (MDS) for Windows. It also provides a Change Sensitive Score that provides an interval metric of student progress. In later articles I’ll highlight some of those additional features and discuss issues of administration, scoring, and interpretation. As well as questions that have come up during BDI-2 Training workshops.