TheCognitive Processing Inventory (CPI)
Designed for:
Developed by
Scott L. Crouse, Ph.D
Copyright © 2007 LDinfo Publishing
Jump to: Section 1 - Overview of the CPI Section 2 - Statistical Properties of the
CPI
Section 3 - The Processing Model of the
CPI
Section 4 - The CPI as Part of a Comprehensive LD Assessment Section 5 - Rating and Scoring the CPI Section 6 - Interpretation of the CPI Section 7 - Installation and Use of the CPI Professional Software
Section 1
Overview of the Cognitive Processing Inventory (CPI)
Development of the CPI: The CPI has been developed over the past twenty years to assist in the process of non-biased differential diagnosis of learning disabled students. It is based upon a thorough review of current neuropsychological research in addition to formal cognitive assessment of over 2000 students, hundreds of parent and student interviews, and direct behavioral observation of the learning disabled population. Ultimately, the intent was to develop a tool which could supplement standard assessment practices and also quantify the information which parents already have about their own child’s learning and cognitive skills. It has been found that the CPI can reliably differentiate specific subtypes of learning disabilities and promote clearer understanding of cognitive processing skills and appropriate educational intervention.
Standardization of the CPI: The initial standardization of the CPI used a two-phase process completed over the course of one year. Phase one of the process involved distribution of CPI rating forms to over 5000 families in a suburban Minnesota area. Approximately 700 of these parent and student/self ratings were completed and returned for data analysis. Teachers were also asked to complete CPI ratings for students who had taken part in the home-rating process. After development of preliminary norms during phase one, phase two was initiated which involved posting an internet web-based CPI rating program. This on-line CPI rating allowed anyone with an internet connection to take part in the standardization process and facilitated the gathering of additional ratings from across the country. The web-based CPI has also enabled ongoing monitoring and revision of the norms.
Age and Gender: The current total standardization sample of 3315 cases were separated into six age/gender groups to be used in the development of the norms. Table 1.1 shows the total number of individuals in each group.
Table 1.1
Standardization Samples by Age and Gender
Age Group Male (N) Female (N)
4.0-6.9 349 302
7.0-10.9 540 490
11.0-14.9 402 499
15.0-18.9 252 509
19.0-22.9 416 1346
23+ 293 1079
Ethnicity: Ethnic distribution within the CPI standardization group has been adjusted to match current U.S. census data.
Because of concern regarding nonbiased assessment practices, further data analysis was conducted to determine the significance of any potential differences found between ratings of "white" and "non-white" individuals. Two-tailed t-tests were performed comparing white and non-white average raw scores obtained in each processing domain for parent, teacher, and student/self ratings. These statistical comparisons found no significant differences across any of the six processing domains or within the Global Processing Index as a function of ethnicity. This suggests that the CPI is able to provide a nonbiased assessment of cognitive processing skills.
Appropriate use of the CPI: The CPI is intended to provide an observable rating of a student's information processing and/or learning style. It primarily represents a convenient means of gathering information from parents about how they view their child's learning and thinking skills. And for older students (age 12 and up) and adults, valuable information can be directly obtained about how they view themselves. Although teacher norms are included, because the CPI is most accurate when completed by someone with long-term familiarity with the student, teacher ratings tend to be somewhat less reliable than parent ratings. For this reason it is suggested that the CPI only be used with teachers who are very familiar with the student in a variety of settings over a relatively long period of time. In general, parent ratings are preferred (as well as self-ratings for individuals over the age of 12).
The CPI is not intended to be used as a sole or even primary assessment of information processing. In general, it should be used in conjunction with objective assessment data and always interpreted with sound professional judgment.
Brief description of the rating scales: The information processing model chosen for the CPI includes the following six general areas of cognitive processing, each of which is based upon well-researched theories of learning and cognition:
Visual processing
Auditory processing
Sequential/rational processing
Conceptual/abstract processing
Processing speed
Executive Functioning
In order to evaluate these processing areas, rating forms are completed by parents, teachers, or the students themselves. These forms consist of 10 questions pertaining to background information followed by 50 rating items. Each item is rated on a scale of 1 to 5 with 1 indicating "obvious difficulty" and 5 indicating "obvious strength" for the person being rated. A rating of 3 is to be used either to indicate "average skill" or when the rater is uncertain about the correct response. This insures that "uncertain" ratings will not significantly skew the results.
Two versions of the rating forms are available in both English and Spanish; the CPI for children ages 4-7, and the CPI for ages 7-adult.
The CPI vs formal cognitive assessment: An individually-administered cognitive assessment instrument is primarily designed and intended to evaluate a student's general intellectual/cognitive abilities. Although subscale or subtest scatter analysis can be an effective means of developing hypotheses regarding an information processing "style", the utility of such approaches has not been clearly substantiated by research. This is probably due to the limited sample of behavior which is available within each subtest along with the inherent error of any "one time only" test of skill.
In contrast, a rating scale such as the CPI is intended to evaluate parents' (and teachers') long-term knowledge and understanding of how a specific child typically performs on everyday activities which have been observed over the course of several months or even years. Each item within the CPI has been carefully chosen because of its direct empirically-based relationship to the given processing area. As such, a certain level of reliability and validity is "built-into" the CPI rating process. Combining the CPI with formal cognitive assessment provides a very solid base of interpretive data for identifying an information processing pattern within a specific individual.
Section 2
Statistical Properties of the CPI
Scores Provided by the CPI:
Presently, the scores provided by the CPI are based upon a normative sample of over 3000 individuals ages 4 through 60, across the United States. Norms are provided for parent ratings, teacher ratings, and self ratings across all age groups. Standardized scores (provided through either hand or computer scoring) include:
• Standard Scores (SS) in each processing domain,
• A Global Processing Index (GPI), and
• Standard Deviation of Difference scores (SDD)
Standard Scores (SS) in each processing domain reflect how a particular student's rating compares to the population "norm" for that rating group (parent, teacher, or self). The mean is set at 100 and the standard deviation set at 15 in order to facilitate interpretation. These standard scores are provided on the Technical Report format of the CPI computer scoring program and can also be plotted on the Graphic Report format to provide a visual display of relative processing strengths and weaknesses.
The Global Processing Index (GPI) is also provided on the Technical Report format as a standard score (with a mean of 100 and standard deviation of 15) and simply represents the average "overall" rating of a particular subject. This is used to determine the relative significance of any differences found across the six processing domains. The GPI provides a standardized comparison between a specific person's rating and the "norms" of the standardization sample who have already been rated. Scores within + or - one standard deviation of the mean (from 85 to 115) represent the "average range" of the standardization group and account for approximately 67% of the total population scores. Although it is on the same scale as most IQ scores, the GPI should not be confused with a measure of general intelligence. The GPI is intended to provide a general measure of overall information processing skill (as viewed by those completing the ratings) which can then be used to determine relative processing strengths and weaknesses in specific areas (see below). Although a low GPI score may suggests rather significant overall information processing difficulty, it would not necessarily suggest low cognitive ability. This is a very important distinction.
The Standard Deviation of Difference (SDD) score is provided to assist with the interpretation of the significance of differences found between standard scores in each processing domain and the Global Processing Index. From a diagnostic standpoint, Standard Deviation of Difference (SDD) scores are considerably more valuable than the GPI because these SDD scores reflect measurable differences within the individual rather than in comparison to other students.
SDD scores provide a measure of the significance of differences found between an individual's GPI and their particular rating in each specific processing area. As with the GPI, the SDD scores relate to the differences found within the standardization group of the "normal population". SDD scores within the -1 to +1 range represent the average or "normal" differences found within 67% of the general population. Based upon the norms, approximately 16% of the general population would obtain an SDD score in any given area lower than -1 and 16% of the general population would be expected to have SDD scores above +1. Only 2% of the general population would be expected to have SDD scores either less than -2 or greater than +2. From a diagnostic perspective, a processing SDD score below -1 could be considered significant enough to negatively impact learning.
For interpretation, "moderate discrepancy" is found in the range of 1 to 2 SDDs (either positive or negative) from the GPI. 2 or more SDDs from the GPI represents a "severe discrepancy" and would indicate either a significant strength or significant weakness in that processing area. Besides the SDD scores in each processing area, the CPI also provides SDD scores reflective of differences between dichotomously paired processing domains. These would included: Auditory Processing vs Visual Processing, and Sequential Processing vs Conceptual processing. Again, 1 or more SDDs between these processing areas would indicate a moderate to significant difference.
SDDs from GPI Interpretation
2 + Significant Relative Strength
1 to 2 Moderate Relative Strength
-1 to +1 Average Range
-1 to -2 Moderate Relative Weakness
-2 + Significant Relative Weakness
The computer-generated report automatically calculates the number of SDDs that a specific processing standard score falls above or below the Global Processing Index. For example, if the GPI is 100 and the Auditory Processing SS is 88, that may translate to a difference of -1.5 SDDs. This would suggest that Auditory Processing is moderately discrepant from the GPI indicating an apparent relative weakness in this area. The actual SDD values found within the various norm groups are listed in the norms tables in Appendix E.
Intra-cognitive vs inter-cognitive interpretation:
Beginning with version 5.0 of the CPI scoring software, examiners are provided the option of either utilizing the default “intra-cognitive” interpretation (evaluating the significance of differences within the individual subject) or switching to a somewhat more traditional “inter-cognitive” interpretation (evaluating the significance of differences between the individual subject and the mean of the norm group). Before deciding which interpretation to use it is very important to fully understand the design and intent of the CPI as well as the pros and cons of each option.
Intra-cognitive comparison represents the true design and intent of the CPI and also follows the philosophical underpinnings of the term “Learning Disability”. In essence, a true learning disability exists when a student’s underachievement is caused by information processing differences within his or her brain. By default, the CPI provides a formal evaluation of intra-cognitive processing differences by comparing the differences noted on the CPI rating forms with “normal” differences found within the norm group. To do this, the CPI scoring program first calculates the subject’s global processing index (GPI) for each rating as well as standard scores in each processing area. These standard scores are then converted to standard deviation of difference (SDD) scores based upon the variance of scores among the norm group. These SDD scores range from -4 to +4 with a mean of 0 (zero). A score of zero simply means that the specific processing area score exactly matches the GPI of that particular rating. Since each specific rating for a particular subject is converted to this scale, all ratings can be directly plotted for visual comparison of relative strengths and/or weaknesses. In other words, even if one rating was rather critical and another rating somewhat lenient, intra-cognitive interpretation gives all ratings a mean of zero so that relative highs and lows can be directly compared.
In conrast, Inter-cognitive comparison represents a means of more directly comparing a given subject’s processing skills with the mean of the norm group. In other words, this interpretation simply compares a given subject’s standard scores in each processing area with the statistical mean of the norm group (100). Standard scores between 70 and 85 are considered to be of “moderate concern” while scores below 70 would be considered “severe concern”. Although this is a traditional interpretive approach (typically used with other types of assessment instruments) it does not provide a means of comparing the significance of differences within the individual subject. The option of inter-cognitive interpretation is offered primarily for situations in which a subject may have generalized information processing issues (across categories) which may not be revealed via intra-cognitive comparison.
Reliability of the CPI:
Test-Retest Stability - An evaluation of test-retest reliability was performed which compared initial and follow-up parent CPI ratings of 150 students at approximate one-year intervals. Of the 150 follow-up sets of ratings, 92% were found to show a pattern of processing which identified identical areas of relatively significant strength and weakness as were noted on the initial rating. This suggests an overall stability correlation of approximately .92.
Internal Consistency - In order to assess the internal consistency and overall reliability of the CPI a split-half method was employed in which the entire CPI item pool and each subscale was randomly divided into 2 similar forms. These split-half correlations were then gathered from the entire normative sample of 3315 cases and are presented in Table 2.1. Correlations across specific processing areas range from .85 to .92 with overall Global Processing Index (GPI) correlations ranging from .94 to .96. This data verifies that the CPI has very strong internal consistency. Table 2.1 Split-Half Reliability Coefficients
Parent Teacher Self Auditory .88 .91 .87 Visual .85 .90 .85 Sequential .87 .89 .89 Conceptual .89 .92 .86 Processing Speed .88 .88 .90 Executive Functioning .89 .92 .89 Global Processing Index .95 .96 .94
Validity of the CPI:
Content Validity - The initial item selection and categorization for the CPI was derived through an empirically-based process which utilized published research findings related to cognitive processing along with hundreds of direct observations and interviews with parents and teachers of learning disabled students. The initial item groupings were then refined through an extensive process of factor analysis to ensure that each item was indeed loading on the specific "processing" factor being rated.
Predictive Validity - In order to evaluate predictive validity of the CPI, parent ratings were obtained on 150 students in grades 3 through 12 who were concurrently being formally evaluated for possible special education services. Using only a pattern analysis* of the general processing clusters of the CPI, correct predictions of learning disability placement were made for 118 of the 150 students (78%) with 12% false-positive and 10% false-negative predictions.
Concurrent Validity - Using data collected during the normative process, concurrent validity was evaluated through parent, student/self, and teacher CPI ratings of students in grades 3 through 12 who had previously been formally identified as having some form of learning disability. For parent ratings of LD students, 86 out of 102 (84%) were found to demonstrate a significant information processing weakness on the CPI at the .15 level. For student/self ratings of LD students, 39 out of 55 (71%) were found to demonstrate a significant information processing weakness on the CPI at the .15 level. For teacher ratings of LD students, 67 out of 86 (78%) were found to demonstrate a significant information processing weakness on the CPI at the .15 level.
Table 2.4 Concurrent Validity Identified LD Students
Ratings Total Significant Correlation Parent 102 86 .84 Student/self (LD) 55 39 .71 Teacher 86 67 .78
Overall, these reliability and validity studies provide very strong support for both the CPI as a screening/assessment instrument and the underlying information processing model chosen.
Section 3
The Processing Model of the CPI
There are many differing and often conflicting theories of cognitive development and information processing. The information processing model chosen for the CPI includes the following six general areas of cognitive processing, each of which is based upon well-researched theories of learning and cognition:
Visual processing
Auditory processing
Sequential/rational processing
Conceptual/holistic processing
Processing speed
Executive Functioning
These broad processing areas have been chosen because of strong historical and empirical support along with the direct educational implications each can provide.
The auditory/visual (sometimes referred to as verbal/non-verbal or linguistic/visuospatial) comparison is the oldest and most extensively researched processing dichotomy. These were the original processing areas described when the concept of a learning disability was first introduced. Various intelligence scales such as Wechsler scales, McCarthy, DAS, Stanford Binet, etc. have continued to emphasize these as primary processing areas. Research has consistently supported the assumption that the brain processes auditory and visual information in very different ways, and when a student demonstrates a significant preference for one over the other, a learning disability can occur.
The sequential/conceptual (sometimes referred to as sequential/simultaneous or successive/holistic) comparison has evolved directly from extensive neuropsychological research conducted over the past several decades. The implications from this research have clearly suggested that each of the two cerebral hemispheres processes information in a rather unique fashion. The left hemisphere appears to process information in a very orderly and detailed manner whereas the right hemisphere takes a more general, holistic processing approach. The Kauffman Assessment Battery for Children (K-ABC) was developed directly from this neuropsychological research base. Although the term "rational thinking" has taken on a somewhat different meaning, taken literally, the word "rational" means "part by part" which is virtually synonymous with our understanding of left-brain processing. However, since the term "sequential processing" is more widely associated with learning disabilities it will be primarily used to describe this processing domain within the CPI.
Processing speed is one of the most recently identified cognitive processing areas, well-supported in research and literature. The actual speed of neural transmission has been measured and found to correlate negatively with some forms of learning difficulties (e.g. low processign speed = high probability of learning difficulty). This has been endorsed as a distinct area of information processing within the latest editions of the WISC, Woodcock-Johnson, and various other cognitive assessment batteries.
The area of Executive Functioning has been added to version 5.0 of the CPI due to the increasing popularity of this cognitive construct within the educational and mental health communities. Executive Functioning refers to the overall ability to manage or regulate all of the various cognitive and emotional processes. This involves initiation, planning, organization, and execution of various tasks as well as the ability to cope with transitions or regulate emotional responses. Subjects with Executive Functioning issues often need externalized structure (i.e. lists, schedules, etc.) and typically respond well to increased structure and predictability in their lives. Weakness in this area is often associated with an attention deficit disorder.
Educational Implications: The intent of any information processing model is to explain the reasons for the educational difficulties experienced by students. The processing model used with the CPI not only helps to explain existing problems but also provides specific guidance for intervention along with the potential to predict other possible areas of difficulty (both educational and non-educational) which typically relate to a particular information processing style. A text is also available entitled "Uncovering the Mysteries of your Learning Disability: Discovery, Self-awareness, Self-advocacy" which is based upon the processing model used within the CPI and is intended to help students understand and cope with their specific learning disabilities. This text provides direct and specific home and classroom interventions for each of the 6 processing subtypes covered by the CPI.
Clarification of the 6 general processing domains:
Visual Processing involves the ability to understand, remember and utilize visual information even when it becomes abstract or complex. When they see something, especially something complex, do they understand it quickly and easily. Can they “visualize” things (like pictures, shapes, words, etc.) in their head? Can they remember information that they see?
Visual Processing involves:
seeing differences between things remembering visual details filling in missing parts in pictures remembering general characteristics visual-motor coordination visualization and imagination organization of their room, desk, etc. artistic skills
Students with a general visual processing disability often experience most learning difficulty in the areas of math and spelling because they have trouble “visualizing” words, letters, symbols, etc.
Specific difficulties may include:
- writing poor handwriting poor spelling (cannot visualize the words) - math difficulty visualizing problems difficulty with cluttered worksheets - reading slow speed poor comprehension - general poor organization/planning/neatness difficulty rechecking work for accuracy difficulty learning by demonstration difficulty learning by video
Auditory Processing involves the general ability to understand, remember, and utilize auditory information.. Can they “keep up” when people talk very fast? Can they tell voices apart easily (even on the phone)? Can they imagine the voices of familiar people in their head? Can they remember information that they hear?
Auditory Processing involves:
hearing differences between sounds/voices remembering specific words or numbers remembering general sound patterns understanding even when they miss some sounds blending parts of words together music
Students with a general auditory processing disability usually have most difficulty with general reading, general writing, and language (understanding and expressing). Specific difficulties may include:
- reading poor decoding of new words poor comprehension - writing poor spelling/mechanics poor sentence structure - communication difficulty with expression poor receptive language - general difficulty following oral directions difficulty learning in lectures
Sequential/Rational processing is generally regarded as the brain's detailed filing system. It involves the ability to learn, memorize, organize, and express detailed or specific information including facts, figures, and formulas.
This is very much like a computer organizes and stores information. How well does a student remember details (like names, addresses, facts, etc.)? How organized are they?
Students with poor sequencing skills may benefit from external structure (such as lists, schedules, reminders, etc.). Such students also sometimes have stronger conceptual processing abilities (reasoning, abstract thinking, creativity) and may learn best when first presented overviews, summaries, and underlying concepts rather than detailed facts.
Sequential/Rational processing involves:
Short-term memory for details long-term retrieval of details fine-motor coordination finding the words you want to say or write organization of your thoughts and materials writing mechanics (spelling, punctuation) reading speed/sounding out new words attention to details putting words and thoughts in order
Students experiencing a general Sequential/Rational processing disability often have most learning difficulties in the areas of basic reading, math computation, expressive language, and writing mechanics. Specific difficulties may include:
- handwriting speed/clarity letter reversals spelling/mechanics letters in wrong sequence (order) - reading decoding (sounding our words) speed/fluency remembering details attention/concentration - math remembering formulas/steps - communication finding words for verbal or written expression - general planning lengthy assignments remembering details paying attention - easily distracted by surroundings remembering names of people or objects following specific directions
Conceptual/Holistic processing involves understanding “the big picture”, overall patterns and underlying concepts for use in higher-order thinking, creating, and reasoning.
Conceptual/holistic filing is like throwing things into boxes with very general labels.
Conceptual/Holistic (right-brain) processing involves:
memory for general themes or ideas reasoning spatial awareness general knowledge inferential thinking estimation/approximation conceptual understanding creativity/inventiveness reading comprehension use of context rhythm music art
Students experiencing a general conceptual/holistic processing disability often perform quite well during early school years but later experience difficulty with reading comprehension, math reasoning, and creative writing. Specific difficulties may include:
- reading understanding irony, inferences, sarcasm general comprehension - math generalizing to new situations story problems - written language creative writing - communication general language comprehension understanding humor - general global/general awareness attention - may focus too much on a specific area
Processing Speed involves how quickly the brain is able to act or react in various situations. Problems can arise when information is either processed too slowly (i.e. the subject can't keep up) or too quickly (i.e. the subject responds impulsively or carelessly).
All LD students experience some processing speed difficulty when required to process information through their weakest processing “channel” or “modality”. But for other LD students, a general weakness in processing speed causes difficulty in all areas.
It is like having the brain work at 40 miles per hour when the rest of the world (and all the information) is going 55 miles per hour. Such students just can’t keep up.
Relatively low Processing Speed is sometimes associated with ADHD - Inattentive type while relatively high Processing Speed is sometimes associated with impulsivity which can be a characteristic of ADHD – Hyperactive type.
Processing Speed involves:
short-term memory (with time pressure) long-term retrieval (with time pressure) talking speed, word-finding writing speed reading speed attention reasoning (with time pressure) general response speed
Students experiencing a general Processing Speed disability often have learning difficulties in all academic areas due to their inability to process all types of information quickly. Specific difficulties may include:
- reading reading speed ability to stay focused while reading - math completing a series of problems - written language writing speed mechanics clarity (with time pressure) - communication delays in responding slow, deliberate speech word-finding difficulties - general coping with implied or expressed time pressures always "a step behind" difficulty maintaining attention to tasks exceeding time limits during tests trouble with social pressures to perform "faster"
Executive Functioning refers to the overall ability to manage or regulate all of the various cognitive and emotional processes. This involves initiation, planning, organization, and execution of various tasks as well as the ability to cope with transitions or regulate emotional responses. Weakness in this area is often associated with an attention deficit disorder.
Executive Functioning skills involve:
ability to stay focused on tasks ability to plan and anticipate organization of thoughts and materials ability to follow-through and complete tasks ability to cope with unstructured situations ability to cope with changes in routine ability to regulate emotions
Students experiencing general Executive Functioning difficulties often struggle academically with work-completion, organization, and motivation for any task which is perceived as difficult, frustrating, or simply unappealing. Specific difficulties may include:
- reading motivation when material is “boring” speed/fluency - skipping words or lines remembering details attention/concentration - math difficulty seeing the “relevance” difficulty maintaining motivation to complete practice worksheets - general planning lengthy assignments remembering details paying attention - easily distracted by surroundings completing assignments following specific directions ability to keep school a “priority”
The CPI as Part of a
Comprehensive LD Assessment
As mentioned previously, the CPI is not intended to be used in isolation to
identify LD students but is ideally suited as one component of a broader
educational assessment. It may be
most appropriate to use the CPI as a pre-referral screening measure to provide
an initial indication of a student's learning/processing style followed by more
comprehensive cognitive and academic assessment if indicated.
Obviously, the goal of any assessment is to obtain data across settings
which consistently points in the same interpretive direction. In order for this to occur, it is very
important that all forms of data be interpreted in terms of the same
information processing model.
The CPI and Achievement: Table 4.1 is provided to
demonstrate the relationship between the processing model of the CPI and
specific areas of academic achievement.
Table 4.1
Correlation Between CPI Processing and
Achievement
|
