CBMs (Curriculum Based Measures)

--Download file--

CBMs have also been referred to as (CBMs are limited compared to these):

CBA – broader term, less formal, more variety – used in curriculum decision making.
CBE – Evaluating hypotheses from the original curriculum based data to link recommendations for teaching and so forth. (The framework for thinking about CBMs)
DIBELS (Dynamic Indicators of Basic Early Literacy skills) – a type of CBM, for younger

Benefits of CBMs –
• Can be related directly to instructional planning.
• Provides direct assessment of skills (Not indirect – thus avoids the assumption that the score is relevant to the skill you are measuring – because you get the direct assessment).
• We measure fluency when we use CBMs.
• Allows for pre and post test measurement.
• Designed to be cost effective and less expensive and less time consuming.
• Data eliminates ambiguity. Not based on a discrepancy formula, which is a vague concept. You either know it or you don’t.
• Eliminates bias in decision making in the use of local norms.
• Content Validity

Features of CBMs
• Tied to the student’s curriculum.
• Must be a short duration to facilitate frequent administration by teachers.
• It’s capable of having multiple forms – so all the administrations can be different.
• Sensitive to improvement over time (Much more so than Standardized tests)
• The data should allow for goal setting through the use of time series analysis.

CBA (Curriculum Based Assessment)

There are several models of CBAs:

The accuracy based CBA: Measures the percentage correct and provides teachers with information to pinpoint where to begin instructional sequences. Helps eliminate instructional mismatch between the skills of the low achieving students and the curriculum assignments.

In using, convert the responses into a ration of correct to incorrect. Then further evaluate how much time the student has spent academic learning. You want to determine what academic level the student is performing at with high success.

Ex.) Have the child read for three minutes from the basal reader at the 3.3 level (most advanced) and count the errors, and divide by the number of words said correctly. Then do the same thing for 3.2 and 3.1, and keep going down, (2.2, 2.1, etc) until they perform successfully. Then instruct at a level slightly higher than the level at which they are performing well.

Criterion Referenced CBA model: We obtain direct and frequent measures, just as above, but here we measure sequentially arranged objectives in the curriculum. The steps of knowledge in order as needed to complete a task. This is useful at the high school level as you are often beyond the basic skills at this point.

To use: Select the items you want to measure. Select the performance which will show mastery. The types of data vary widely and are not prescribed.

Using CBM for alternative procedure for eligibility determination.

Not based on the old fashioned discrepancy formula. The problem with that traditional model is that it is fully involved within the single child. There is nothing in the environment included in this model.

Single step eligibility: Do Screening – get scores, find kids below average, and target them for further assessment. (very similar to the old IAT)

Multi step model: Compare referred students to normal functioning peers on grade level curricular tasks through repeated samples over time. As the data continues to collect, look for the median score and graph those in relation to the grade level score. Then you can use one of two techniques to determine if further assessment is needed to determine eligibility.

1. Discrepancy Ratio: Used for classroom or school wide norms. Divide the greater academic performance by the lesser performance (normal students median grade by the referred student) = the resulting ratio is greater than one. Has a sign value attached to it. A minus (-) is given if the referred student’s score is lower, which is what you would most likely expect. A plus ratio is assigned if the referred student is greater.
2. Percentile Ratio: Used for district norms. Organize all the scores of all the students in the district by quartiles to aid in determining a cutoff for determining eligibility. The cutoff figure is arbitrary.

Developing Local Norms: Three types: Class wide, school wide, and district wide.

If you were to work to develop school wide norms: You would decide upon three times where you will collect data (Fall, Winter, Spring). Then you will develop key points for each grade level as to what is to be targeted in the data. Pick out the important points in the curriculum that need to be identified. As time progresses, this is being done on a national level. Then develop a sampling plan (get a representative sample of the school’s population). Once you summarize data, you will have an organized listing of the school’s performance with means, SDs, etc.

Curriculum Review and task analysis.

Reading comprehension: An example task analysis.

Know letters
Know sounds
Know to read from left to right
Know blending
Know syllable rules to read words they don’t know
Know decoding of words
Know separation of words (space)
Know context clues
Know definitions of words read
Know sentence structure and separation of sentence
Know paragraph structure
Know how to remember what has been read
Be able to retell what has been read
Be able to locate answers about text within text
Be able to infer meaning from text

BOSS (Direct Observation: Behavioral Observation of Students in Schools)

Used to compare student behavior to that of peers. Compare in the environment in which there is a problem. Compare in another environment area.

Filed under: EDC 514-515 Academic Assessment for Intervention
Copyright: January, 2004 - David Profitt