running head: multiple intelligences assessment … · web viewdevelopment and validation of a...
TRANSCRIPT
Multiple Intelligences Assessment 1
Running head: MULTIPLE INTELLIGENCES ASSESSMENT
Development and Validation of a Multiple Intelligences
Assessment for Children
C. Branton Shearer, Ph.D.
Multiple Intelligences Research and Consulting
1316 S. Lincoln St.
Kent, Ohio 44240
330-677-8534
Multiple Intelligences Assessment 2
Development and Validation of a Multiple Intelligences Assessment for Children
Abstract
Three studies describe the development and validation of a self- and parent-report measure of
children's multiple intelligences disposition, The Multiple Intelligences Developmental
Assessment Scales for Children (MIDAS-KIDS). Two initial studies describe the development
and preliminary validation of the instrument. The third validation study involved over 2240
children in grades K - 8 from five states and across a range of socioeconomic levels. Factor
analyses identified an appropriate seven-factor solution. High internal consistency estimates (.82
- .91), inter-rater ratings (.15 - .61), test-retest reliability (.68 - .90), and criterion-related validity
statistics were obtained. Results indicated that MIDAS-KIDS possesses adequate validity and
reliability for classroom use, but further studies will help to substantiate criterion-related validity
and strategies for interpretation. Subscale cluster analysis is recommended for further scale
development.
______ total words
Multiple Intelligences Assessment 3
DEVELOPMENT AND VALIDATION OF A MULTIPLE INTELLIGENCES ASSESSMENT
FOR CHILDREN
The nature of human intelligence has fascinated people around the world for as long as
history has been recorded. Every culture possesses a vocabulary to describe what abilities and
characteristics are responsible for a person to navigate successfully the challenges of everyday
life and make vital contributions to the social welfare. In our Western culture the measurement
of intelligence has been a major focus for psychologists ever since Alfred Binet developed the
first test of cognitive abilities for French school children in 1905. In spite of Binet's refusal to
claim that his test actually measured intelligence his invention became the basis for one of
psychology's most enduring contributions to education and society of the 20th century.
Concurrent with the worldwide growth of the testing industry over the course of nearly
100 years, many theorists have continued to debate the validity of both unitary and multifaceted
models of human intelligence (Block & Dworkin, 1976; Thurstone, 1938; Sternberg, 1988).
Multifaceted theories and tests have not gained much acceptance nor use. The unitary definition
of intelligence -- as exemplified by the intellectual quotient (I.Q.) and assessed by the Stanford-
Binet (2004) and the Weschler (1974) tests -- has become a ubiquitous feature of contemporary
life and every school child's educational experience.
Although prominent theorists such as J.P. Guilford (1967), L.L. Thurstone (1938) and
R.J. Sternberg (1988) have offered multi-factored alternatives to the unitary I.Q. model, none
have had the impact on the worldwide educational community as that of the theory of multiple
intelligences (MI). In 1983, Howard Gardner proposed that human intelligence is best defined as
a "bio-social potential" that allows a person to solve problems, create products or provide
services that are valued within a community. Using this definition and eight criteria, Gardner has
Multiple Intelligences Assessment 4
identified at least eight distinct yet complementary constructs: linguistic, logical-mathematical,
musical, spatial, kinesthetic, interpersonal and intrapersonal and naturalist (1993) (see Appendix
1 for details).
Gardner's (1983, 1993) contextual and culturally embedded definition of intelligence
challenges the assumption that intellectual prowess can be reduced to a single score that can be
measured via paper-and-pencil and decontextualized tests. In the MI perspective, the vast
majority of "real world" problem solving occurs within value-laden, stimulus-rich situations
requiring something more than answering short questions with short answers as quickly as
possible. This contextually-based conception of intellectual ability requires a new approach to
assessment in order to describe the person's intellectual and creative life.
Psychologists and educators both criticize MI theory because there is no data-based,
empirically validated test for each of the various intelligences. Psychologists condemn MI as a
mere “literary” rather than “scientific” theory because it lacks large-scale test-based validation.
Educators require a valid MI assessment upon which to build MI-inspired curriculum and
instruction.
The lack of a research-based MI assessment has lead to two undesirable results. First, to
aid classroom instruction teachers have resorted to the widespread use of brief checklists that are
included in many popular books (and websites) describing the multiple intelligences (Armstrong,
1994; Campbell, Campbell & Dickinson,1992; Chapman, 1993; Kagan & Kagan, 1998). The use
of informal MI checklists has been criticized by Gardner and others (Gardner, 1999; Chen, et al,
1998; Stefanakis, 2002), as misrepresenting MI theory by being overly simplistic, deceptive and
potentially harmful. Second, and perhaps of greater consequence, because there is no recognized
data-based MI assessment the acceptance (or rejection) of MI theory as a valid scientific
Multiple Intelligences Assessment 5
description of the intellectual potential of human beings has stalemated into an argument of
“believers” vs. “non-believers.” This stalemate has placed MI theory and its implementation in a
marginalized position because it is viewed as “not being research-based.”
Gardner’s meta-empirical / trans-disciplinary approach to MI theory building for the
past 25 years has appealed to large numbers of educators around the world, but has failed to
convince data-oriented, experimental research psychologists. Nearly one hundred years of data
collection supporting the validity of the unitary construct of general intelligence (g) (and
embraced by psychologists and the general public in the IQ score) cannot be undermined by
what is referred to as “merely a literary theory” rather than a scientific verity by its critics
(personal communication). Likewise, public policy makers require efficient and statistically
valid measures of student and school performance in order to support the use of any theory or
assessment that will alter the structure or function of schools. U.S. Department of Education
guidelines require that school innovations be "research-based" in order to qualify for federal
funding (Linn, et al, 2002; No Child Left Behind Act, 2002).
Multiple Intelligences Developmental Assessment Scales (MIDAS)
The Multiple Intelligences Developmental Assessment Scales (MIDAS) is a self or other
completed questionnaire that can be administered and interpreted by psychologists, counselors
and teachers. There are four versions of the assessment for various age groups, ranging from
four years through adulthood. The MIDAS inquires about developed skill, levels of
participation, and enthusiasm for a wide variety of activities that are naturally encountered as a
part of daily life. The MIDAS was initially developed in 1987 as a structured interview format to
assess the multiple intelligences for adolescents and adults undergoing cognitive rehabilitation
(Way and Shearer, 1990). A summary of research results concluded that the MIDAS provides a
Multiple Intelligences Assessment 6
“reasonable estimate” of a person’s intellectual disposition in the eight designated areas
(Shearer, 1996; Buros, 1999, 2007).
The MIDAS is 119-item questionnaire with Likert-style individually written response
choices. It was originally developed as a structured interview format to gather information about
a brain injured person's pre-morbid MI profile from someone who knew the person well prior to
injury. At this time, for educational purposes, it is most often group or individually administered
as a self-report description of intellectual activities, skills, and involvements.
Research with the MIDAS has indicated that this instrument possesses acceptable
psychometric properties including factor structure, item consistency, test-retest reliability and
appropriate discrimination with various criterion groups and measures (Buros, 1999, 2007;
Shearer, 1996). Recent research found appropriate patterns and correlational values among
MIDAS scales and matched reading, math and IQ scores (Shearer, 2006). These findings support
the assumption that an adolescent's or adult's perceived multiple intelligences disposition can be
reasonably described by way of self-report through the careful use and interpretation of the
MIDAS Profile.
Objectives
The goals of the following research were to adapt the teen version of the MIDAS for pre-
adolescent children in grades Kindergarten through eighth, that (a) provide a profile of a child's
developed skill levels in seven identified constructs; (b) provide an estimate of propensity for
two intellectual style scales, Innovation and Technical; (c) demonstrate acceptable psychometric
properties (e.g., scale reliability); (d) possess a sound factor structure and discriminate
adequately among appropriate criterion groups; and (e) can be used effectively with children
across a range of cultures, ages, abilities and socioeconomic status.
Multiple Intelligences Assessment 7
Initial Scale Construction
The MIDAS-KIDS questionnaire was developed over a period of two years following the
format and style of questions employed with the teen MIDAS questionnaire. An initial group of
items was generated after a careful review of the behavioral characteristics associated with each
intelligence appropriate for children, as described by Gardner (1983, 1993), in Frames of Mind.
The goal was to make it easy for a child or parent to respond with a minimum of guesswork or
generalization. The items inquire about observable behaviors representing specific skill domains
within each construct as described by MI theory, e.g., writing for Linguistic and calculations for
Math-logic. To enhance reliability, three to four questions regarding activities representing each
designated skill domain within the seven constructs were written.
The items are written in one of three basic forms. Some items ask the informant to
assess the frequency or duration of time the child spends engaged in an activity that characterizes
the construct. Other questions ask the informant to provide a realistic evaluation of the child's
performance on that activity. Still other questions ask the informant to provide an assessment of
the child's enthusiasm for the activity. Response choices are uniquely written to fit the particular
content of each question (see sample item in Appendix 2).
Each item has an "I don't know or Does not apply" choice so respondents are not forced
to guess or assess an activity inappropriate to the child's age or experience. These responses are
counted as missing and not figured into the scale scores. Scores for responses range from zero to
four. On the questionnaire itself, response choices are marked by letter rather than by number to
encourage respondents to respond to the content of descriptive response choice rather than to a
numerical label. Percentage scores for each scale are calculated only from the total number of
responses.
Multiple Intelligences Assessment 8
Study 1.
Six elementary teachers trained in MI theory and two school psychologists reviewed the
initial body of 120 items. Two forms of the resulting 113-item questionnaire were devised by the
author, one for children (third grade and above) and a second for parents (K - second grade). The
reading difficulty of the children's questionnaire was estimated as being Easy at about the fifth
grade level with a Reading Ease score of 86.7 (Flesch Reading Ease, 2001). Within each scale,
items were grouped according to specific domains creating tentative subscales (e.g., Instrument
for Musical and Working with Numbers for Mathematical-Logical) logically consistent with MI
theory.
This 113-item questionnaire was then administered in an in-depth pilot study to a group
of 49 children in grades K through eighth and 74 parents. Some participants were interviewed by
a research assistant and others completed the questionnaire independently. Teachers at an MI-
designed school also administered the questionnaire to individual students and requested that
parents complete it. All participants were given the opportunity to comment on the questionnaire
when finished. General impressions and problem items were thus noted for future consideration.
Seventeen questions were eliminated from the item set because of low reliabilities and
participants' suggestions. For example, "Did you have a hard time learning how to tie your
shoes?" was eliminated because children could not remember this information. The question "Is
it easy for you to keep the beat when you're clapping your hands or tapping your feet?" was
simplified to "How well do you keep the beat when you clap your hands or tap your feet?"
Study 2.
The 96 remaining questions were submitted for content review by subject area classroom
teachers (art, physical education, music, reading, and math). Two school counselors and two
Multiple Intelligences Assessment 9
school psychologists reviewed the questionnaire for sex bias and age appropriateness. A trained
research assistant also conducted in-depth interviews with seven children and parents. During
these interviews the researcher inquired about the meaning of each question and the clarity of
wording. The 96-item questionnaire was then revised by the author so that item content
adequately covered each domain within the designated MI construct. Problematic questions were
also rewritten to be clear in their meaning and appropriate for the range of targeted age groups.
The revised 96-item questionnaire was then administered by classroom teachers to 145
children and 25 parents for further item and scale analysis. Scale means, standard deviations and
coefficient alphas were calculated. Items were subsequently modified or dropped from the scale
if item-total correlations were low (i.e., r < .30) or the response pattern was skewed. Additional
responses and suggestions by interviewees regarding item construction were considered. As a
result of these analyses, 16 items were eliminated so that the final 80-item instrument consisted
of seven main intelligence scales and two intellectual style scales (Technical and Innovative).
A content analysis of the finalized item set was conducted by an experienced MI teacher
to ensure that domain categories were adequately represented. It was found that 45 questions
(56%) inquire about skills and 24 (30%) ask about amount of participation and 11 (14%) ask
about expressed enthusiasm. The reading difficulty of the questionnaire was estimated as being
Very Easy at about the 4th grade level with a Reading Ease score of 92.5 (Flesch Reading Ease,
2001). The parent's version of the instrument was also adjusted accordingly as is appropriate for
an outside informant. All items could be translated into observable terms so that an informant
might reasonably be able to rate particular activities in question.
Multiple Intelligences Assessment 10
Study 3.
The goals of the third study were to (a) replicate the item analysis; (b) examine internal
consistency estimates and inter-rater reliability, (b) explore the factor structure of the instrument;
and (c) initial investigation of its concurrent and discriminant validity (i.e., correspondence
between ratings on the MIDAS scales and appropriate criterion groups and measures).
Method
Participants
Parent and child permission were again obtained for children to participate in this
project. The sample consisted of 106 kindergarten parents, 155 first grade parents, 148 second
grade parents, 152 third grade students, 184 fourth grade students, 288 fifth grade students, 442
sixth grade students, 423 seventh grade students and 245 eighth graders (N=2,241) from six
middle schools and seven elementary schools in five states on the east coast, west coast, and
midwest. The schools were selected to represent a socio-economically diverse array of student
backgrounds, ranging from large inner city neighborhoods, small towns, suburban and rural
school districts. The sample included 1,142 (51%) girls and 1,099 boys. The mean age of the
total sample was 10.8 and 561 (25%) were African-American, 1,525 (68%) were White, and the
remainder (n=67 or 3%) were unclassified.
Measures
Teachers in two schools administered The MIDAS to students and provided their Verbal,
Performance and Full Scale I.Q. scores from the Wechsler Intelligence Scale for Children-R
(WISC-R; Wechsler, 1974). These scores were used in several preliminary validity
investigations described below.
Multiple Intelligences Assessment 11
Procedures
The teachers administered the 80-item questionnaire to intact third through eighth grade
classes using a standard set of instructions. Third grade teachers read the questionnaire aloud
while students followed along. The parents of K-2 students completed the questionnaire at home
and returned it a week later. Teachers provided Verbal, Performance and Full Scale I.Q. scores
from student records. Students from three schools self-reported participation in one or more
activities of interest, curricular and extracurricular programs for validity comparisons.
A group of 57 parents of children in an Academically Talented program at a middle
school were asked to complete the instrument in addition to their child's self- rating. The teacher
of this program also provided the children's WISC-R scores.
Results
Results are presented in two sections: (a) reliability statistics: Alpha, inter-rater
comparisons and temporal stability; (b) construct and criterion group validity: factor analyses,
concurrent measures and contrasted groups.
Reliability
Intrascale reliabilities were calculated (Table 1) and coefficient alpha's ranged from a low
of .83 for Kinesthetic and Linguistic to a high of .91 for Intrapersonal. The number of items per
scale range from 10 to 22. These results indicated strong internal consistency for the seven
scales. Similar results were obtained for the two intellectual style scales, Innovation (.82) and
Technical (.83).
-- Insert Table 1 About Here --
Comparative statistics between the parents' scale scores and the children's self-
assessment are provided in Table 2. Correlation coefficients between these two samples range
Multiple Intelligences Assessment 12
from a low of .15 for Interpersonal to a high of .61 for Musical. All correlations are significant
at least at the .05-level except for Interpersonal.
-- Insert Table 2 About Here --
In a separate study, an entire 8th grade class of students (N=93) at a suburban middle
school completed the questionnaire two times with a one-week separation. Correlations for the
seven main scales ranged from a low of .68 to .82 with most scales near to .80 (see Table 1).
Validity
Factor Analyses. Exploratory factor analysis was conducted using a random sample
from the database of approximately 850 participants nine years old and above. Maximum
Likelihood analysis using Promax rotation (Kappa 4) revealed sixteen factors with eigen values
greater than one that accounted for 56% of the total variance. An examination of the scree plot
showed possible break points between factors six and eight.
A review of the items comprising each of the initial sixteen factors found that the first
four factors matched with theoretical expectations (Intrapersonal, Spatial, Kinesthetic and
Musical). The next three predicted factors—Interpersonal, Linguistic and Logical-mathematical
– were split between two factors each. An additional five factors consisted of only one or two
items each and were unable to be interpreted. It is instructive to note that the split factors for the
three scales are consistent with MI theory and correspond well with proposed content subscales.
For example, the first Logical-mathematical factor consists of questions pertaining to
mathematics and numbers while the second factor contains only questions related to logical
thinking and problem solving.
The scale’s factor structure was then explored using Maximum Likelihood analysis with
Promax rotation (Kappa 4) and a seven factor specification. This analysis accounted for 42% of
Multiple Intelligences Assessment 13
the total variance with a great majority of the items loading primarily on their designated scale
(see Table 3).
-- Insert Table 3 and Figure 1 About Here --
The inter-factor correlations range from .24 (Logical and Linguistic) to .60 (Logical and
Intrapersonal) with a majority in the .45 range (Table 4)
-- Insert Table 4 About Here --
The seven-factor structure did not, however, display a perfectly clean pattern of items
loading on a single factor only. Approximately 30% of the items were found to correlate on one
factor (or in a few instances two) in addition to their designated scale. Nearly all of these co-
loadings are theoretically consistent with multiple intelligences theory and so the most
significant correlations were incorporated into a complex scoring matrix. For example, item #34,
"How good are you at taking things apart and then putting them back together again?"
correlates at .38 with Spatial and at .33 with Kinesthetic. Item #61, “Is it easy for you to get
people to do things your way?” correlates at .63 with Interpersonal and at .36 with Intrapersonal.
As Howard Gardner has noted (1983) successful performance on complex everyday tasks often
requires a combination of intelligences and these co-loading items are theoretically consistent
examples of this idea. A drawback to this scoring procedure is that it produces moderately high
inter-scale correlations ranging from .43 to .79 with a mean of .62 (Table 1).
Subsequent confirmatory factor analyses and item-scale total score correlational studies
were conducted with new samples in order to clarify the optimal scale item composition. These
studies provided confirmation of the seven-factor structure, but also indicated the need to co-
score some items to account for the complex nature of real world performances1.
1 This research was conducted with a seven intelligences version of the questionnaire developed
Multiple Intelligences Assessment 14
-- Insert Tables 5 & 6 About Here --
Criterion Group Validity
Tables 7, 8 and 9 display the mean scale scores of a representative student group, means
by grade level and 12 criterion groups. Scale score distributions approximated normal, and
scores demonstrated good dispersion.
-- Insert Table 7 and 8 About Here --
Some of the criterion groups were determined by teacher selection using various
appropriate criteria (focus talent area groups, LD, Academically Talented) and other groups
were identified by way of students' self-report of voluntary participation (orchestra, social
groups, gymnastics, sports). The middle school for the Visual and Performing Arts is a selective,
citywide magnet school where students must pass an audition and have teacher
recommendations. Most students have several years of instruction and training in their focus
talent areas. The Dance, Vocal, Art, Drama and Music groups in the focus talent areas. All
students have one focus area and some have two.
The Learning Disability (LD) group were students (identified by teachers and test results)
who participate in resource room activities and are provided with individualized instruction to
remediate academic or learning difficulties. The mean I.Q. scores for a group of 12 of these
students are: Full Scale= 92, Verbal= 91 and Performance=96. The Academically Talented
(AT) group consists of students in three enrichment programs selected primarily on the basis of
prior to the identification of the Naturalist intelligence by Howard Gardner in 1996. Subsequent
factor analytic studies confirmed the original seven-factor structure as well as the additional
eighth for the item created to assess Naturalist abilities.
Multiple Intelligences Assessment 15
high I.Q. scores. The I.Q. test scores available for 50 students in this sample result in a mean
Full Scale score of 136, Verbal score of 136 and Performance score of 130.
The voluntary extra-curricular groups (e.g., Orchestra, Social, Gymnastics and Sports)
were all middle school students in public schools. The Orchestra group consisted of students
who generally had several years of experience with their instruments. The Social group consists
of students who indicated that they were members of any of three school organizations: student
counsel, Peer Mediators and Circle of Friends. The Gymnastics group consists of students who
participated for at least one year in gymnastics instruction. The Sports group consists of students
who indicated participation in any of four sports activities: basketball, track, badminton and
soccer.
The representative student group is presented for comparative purposes. It is composed
of undifferentiated public school students in grades 4 through 8th (Table 7). Mean scores per
grade level are displayed in Table 8.
-- Insert Table 9 About Here --
Concurrent Comparisons With Standard Tests
The teachers for two student groups (Academically Talented and LD, N=62) provided
each students' Verbal, Performance and Full Scale scores from the Wechsler Intelligence Scale-
Revised (WISC-R). The correlations between the 10 MIDAS scales and these scores are
presented in Table 10. The highest correlation is between the Linguistic scale and Verbal I.Q.
(r=.60 p<.001). The next strongest and significant correlations also with the Verbal scale are for
Logical (.57) and Intrapersonal (.58). Low and insignificant correlations with the Verbal scale
are observed for Kinesthetic (-.04), Spatial (-.01) and Interpersonal (.14).
-- Insert Table 10 About Here --
The strongest correlations with Full Scale I.Q. are with Logical and Intrapersonal both
at .54. Again, correlations with Kinesthetic, Spatial and Interpersonal are low and insignificant.
Logical and Intrapersonal are also the MI scales most strongly (p<.01) correlated with the
Performance IQ at .39 and .36, respectively.
It is of interest to note that the Innovation scale fails to be correlated with any of the I.Q.
scores, but the Technical scale shows significant correlations across the board. Thus, supporting
their differential validity. Overall, the strength and pattern of these correlations are statistically
significant, theoretically meaningful and in the expected directions. Of course, this small sample
provides only an indication of the scale’s concurrent validity and is not definitive.
The only deviation from the expected pattern is for the Spatial scale and its low
correlation with Performance I.Q. This is a puzzle, but other data appear to indicate that the
Spatial scale examines creative activities and divergent thinking skills rather than logical,
convergent problem-solving. This is supported by the negative correlation between Performance
and Innovation and the significant, positive correlation with the Logical scale.
Summary and Discussion
These results support the hypothesis that a child's perceived multiple intelligences profile
may be obtained with adequate reliability and validity when appropriate verification strategies
are followed to account for sources of respondent bias. The instrument evidenced high internal
consistency across all studies and strong test-retest reliability after a one-week delay.
Correlations between a child's self-report and the parent’s assessment were generally acceptable
in the moderate range except for the Interpersonal scale.
Small scale studies of the MIDAS-KIDS convergent and discriminant validity was
supported by appropriate patterns of correlations with WISC-R scores as well as a contrasted
groups study with criterion groups. Additionally, construct validity for the seven scales was
supported by exploratory factor analysis.
Recommendations resulting from the interpretations of these research results are
possible. Additional scale development is indicated based on the factor analyses. An undesirable
number of items correlate with more than one scale, which results in moderate inter-scale
correlations. Scale simplification may be obtained through the judicious elimination of a number
of complex items that are weakly correlated with a designated factor.
Further scale development should occur which will examine content and criterion related
validity in light of both the factor structure and theoretical predictions. Concurrent validity
comparisons should be replicated with larger sample sizes. Cluster analysis should also be
conducted to investigate subscale patterns within each main scale. This will enhance the
effective interpretation of results for educational uses and substantiate construct validity. It is
recommended that an additional shorter MIDAS-KIDS version be created for parent completion
regarding children between the ages of five and nine.
It must be noted that the MIDAS-KIDS is a measure of perceived intellectual disposition
and as such the results need to be carefully reviewed for validity with the input of the person in
light of other life experiences including grades, hobbies, test results and feedback from teachers.
The MIDAS Profile is a product of the child's Intrapersonal awareness and psychological factors
can influence the results such as depression, ego inflation and self-criticalness. When completed
by a parent, it is important to consider influences that might distort the profile such as limited
knowledge, emotional distress and attitudes toward school.
It has been found that people generally know themselves well and they are able to
provide useful information about their intellectual disposition, which is defined as thinking
performance in everyday life in terms of skill, behavior and preference. A person's MI Profile
may differ from actual test scores that are obtained from contrived testing situations that are very
different from the tasks of daily thinking and problem solving. Lastly, researchers now have a
psychometrically-based tool that produces quantitative information that can be used to further
explore the multi-faceted view of human intelligence in children.
References
Anastasi, A. (1982). Psychological testing: 5th ed.. New York: MacMillan.
Armstrong, T. (1994). Multiple intelligences in the classroom. Alexandria, VA: ASCD.
Binet, A., & Simon, T. (1916). The development of intelligence in children. Baltimore:
Williams & Wilkens.
Block, N.J., & Dworkin, G. (1976). The IQ controversy. New York: Pantheon.
Buros, O. (1999). The thirteenth mental measurements yearbook: Supplement. Highland
Park, NJ: Gyphon.
Campbell, L., Campbell, B. & Dickinson, D. (1992). Teaching and learning through
multiple intelligences. Stanwood, WA: New Horizons for Learning.
Chapman, C. (1993). If the shoe fits…how to develop multiple intelligences in the
classroom. Palatine, Illinois: IRI/Skylight Publishing Inc.
Chen, J., Krechevsky, M. and Viens, J. (1998). Building on children's strengths: The
experience of project spectrum. New York: Teachers College Press.
Cronbach, L. (1951). Coefficient alpha and the internal consistency of tests.
Psychometrika, 16, 297-334.
Flesch Reading Ease. (2001). Office XP Word. Microsoft Corp. Word.
Gardner, H. (1998, 1993). Frames of mind. New York: Basic Books.
Gardner, H. (1999). Intelligence reframed: Multiple intelligences for the 21st century.
New York: Basic Books.
Guilford, J.P. (1967). The nature of human intelligence. New York: McGraw Hill.
Kagan, S. & Kagan, M. (1998). Multiple intelligences: The complete MI book.
San Clemente, CA: Kagan Cooperative Learning.
Linn, R.L., Baker, E. & Betebenner, D. (2002). Accountablility systems: implications of
the No Child Left Behind Act of 2001. Educational Researcher, 31, (6) 3 - 16.
No Child Left Behind Act of 2001, Pub. L No. 107-110, 115 Stat. 1425 (2002).
Shearer, C. B., & Jones, J. A. (1994, April). The validation of the Hillside Assessment of
Perceived Intelligences: A measure of Howard Gardner’s theory of
multiple intelligences. Paper presented at the annual meeting of the
American Educational Research Association, New Orleans, LA.
Mercer, J.R. (1984). What is a racially nondiscriminatory test? In Reynolds, C.R.
& Brown, R.T. (Eds). Perspectives on bias in mental testing. New York: Plenum Press.
Shearer, C. B. (1994). The MIDAS: A professional manual Kent, OH: M.I. Research and
Consulting, Inc.
Shearer, C.B. (2006). Criterion related validity of the MIDAS. Retrieved 6-6-06 from
www.miresearch.org/research.php.
Solomon, G. (Ed.). (1993). Distributed cognitions: Psychological and educational
considerations. Cambridge, UK: Cambridge University Press.
Stanford-Binet Intelligence Scales, Fifth Edition. (2004). Retrieved July 22, 2004, from
http://www.riverpub.com/products/sb5/index.html.
Stefanakis, E. (2002). Multiple intelligences and portfolios. Portsmouth, NH: Heinemann.
Sternberg, R.J. (1988). Beyond I.Q.: A triarchic theory of human intelligence.
Cambridge, UK: Cambridge University Press.
Thurstone, L.L. (1938). Primary mental abilities. Chicago: University of Chicago Press.
Wechsler, D. (1974). Manual for the Wechsler Intelligence Scale for Children-Revised.
New York: Psychological Corporation.
Author Note
Correspondence concerning this article should be addressed to C. Branton Shearer, 1316
S. Lincoln St., Kent, Ohio 44240, [email protected]
______________________________________________________________________
Appendix 1
Multiple Intelligences Definitions and Subscales
______________________________________________________________________
Musical
To think in sounds, rhythms, melodies and rhymes. To be sensitive to pitch, rhythm, timbre and
tone. To be able to recognize, create and reproduce music by using an instrument or voice.
Active listening and a strong connection between music and emotions.
> Musical Ability: awareness of and sensitivity to music, rhythms, tunes and melody
> Instrument: skill and experience in playing a musical instrument
> Vocal: a good voice for singing in tune and along with other people
> Appreciation: actively enjoys listening to music
Kinesthetic
To think in movements and to use the body in skilled and complicated ways for expressive as well
as goal-directed activities. It involves a sense of timing and coordination for whole body
movement and the use of hands for manipulating objects.
> Physical Skill: ability to move the whole body for physical activities such as balancing,
coordination and sports
> Dancing, Acting: to use the body in expressive, rhythmic and imitative ways
> Working with Hands: to use the hands with dexterity and skill for detailed activities and small
work
Mathematical-Logical
To think of cause and effect connections and to understand relationships among actions, objects
or ideas. To be able to calculate, quantify, consider propositions and perform complex
mathematical or logical operations. It involves inductive and deductive reasoning skills as well
as critical and creative problem-solving.
> Problem Solving: skill in organization, problem solving and logical reasoning; curiosity and
investigation
> Calculations: ability to work with numbers for mathematical operations such as addition and
division
Spatial
To think in pictures and to perceive the visual world accurately. To be able to think in three-
dimensions and to transform one's perceptions and re-create aspects of one's visual experience
via imagination. To work with objects effectively.
> Imagery: use of mental imagery for observation, artistic, creative, and other visual activities
> Artistic Design: to create artistic designs, drawings, paintings or other crafts
> Constructions: to be able to make, build or assemble things
Linguistic
To think in words and to use language to express and understand complex meanings. Sensitivity
to the meaning of words as well as the order among words, their sounds, rhythms, inflections.
To reflect on the use of language in everyday life.
> Linguistic Sensitivity: skill in the use of words for expressive and practical purposes
> Reading: skill in reading
> Writing: ability and interest in writing projects such as poems, stories, books or letters
> Speaking: skill in oral communication for persuasion, memorization and description
Interpersonal
To think about and understand another person. To have empathy and recognize
distinctions among people and to appreciate their perspectives with sensitivity to their
motives, moods and intentions. It involves interacting effectively with one or more
people in familiar, casual or working circumstances.
> Understanding People: sensitivity to and understanding of other people's moods,
feelings and point of view
> Getting Along With Others: able to maintain good relationships with other people
especially friends and siblings
> Leadership: to take a leadership role among people through problem solving and
influence
Intrapersonal
To think about and understand one's self. To be aware of one's strengths and weaknesses
and to plan effectively to achieve personal goals. It involves reflecting on and monitoring
one's r thoughts and feelings and regulating them effectively. The ability to monitor
one's self in interpersonal relationships and to act with personal efficacy.
> Knowing Myself: awareness of one's own ideas, abilities; personal decision making
skill
> Goal Awareness: awareness of goals and self correction and monitoring in light of a
goal
> Managing Feelings: ability to regulate one's feelings, moods and emotional responses
> Managing Behavior: ability to regulate one's mental activities and behavior
Innovative: To work in artistic, divergent and imaginative ways. To improvise and
create unique answers, arguments or solutions.
Technical: To work accurately, carefully. To strive for just the right answer and perform
activities in the exact way they are shown.
______________________________________________________________________
Appendix 2
Representative MIDAS-KIDS Item
______________________________________________________________________
Musical:
How well do you keep the beat when you clap your hands or tap your feet?
A= Not very well
B= Well
C= Very well
D= Excellent
E= The best
F= I don't know
Note. The wording of response choices varies according to the content of the question.
______________________________________________________________________
Table 1
Interscale Correlations, Scale Coefficient Alphas and Test-Retest Reliabilities
________________________________________________________________________
MIDAS Scales
________________________________________________________________________
Mus Kin Log Spat Ling Inter Intra Inno Tech
________________________________________________________________________
Musical .85*
Kinesthetic .54 .83
Logical .39 .46 .90
Spatial .45 .56 .42 .86
Linguistic .59 .56 .60 .58 .83
Interpersonal .48 .54 .51 .48 .69 .88
Intrapersonal .44 .50 .63 .47 .68 .72 .91
Innovation .63 .66 .52 .83 .73 .59 .53 .82
Technical .67 .75 .72 .59 .77 .63 .67 .63 .87
Test-retest .79 .79 .90 .68 .80 .81 .82 .80 .81
_______________________________________________________
Note. * Alpha reliabilities are found on the diagonal.
All Pearson correlation coefficients p<.001 for test-retest.
Scale abbreviations: Mus = Musical; Kin= Kinesthetic; Log=Math-logic; Spat=Spatial;
Ling= Linguistic; Inter= Interpersonal; Intra=Intrapersona;
Inn=Innovative;Tech=Technical.
_______________________________________________________________________
Table 2
Parent - Child Inter-rater Statistics
________________________________________________________________________
Sample Equality Equality
Self Parent of Variances of Means
________________________________________________________________________
Scales mean sd r mean sd F Value
______________________________________________________________
Music 66% 13 .61*** 62% 18 8.40** 1.38
Kinesthetic 56 19 49*** 47 17 .31 2.71**
Spatial 56 19 .45*** 52 20 .00 .92
Logical 72 15 .60*** 73 14 1.11 .92
Linguistic 65 13 .37** 69 15 2.46 -1.54
Interpersonal 63 13 .15 59 16 3.54* 1.30
Intrapersonal 65 13 .34* 64 14 .69 .41
Innovation 57 16 .41** 54 15 .12 1.06
Technical 65 12 .46*** 63 12 .55 1.10
__________________________________________________________________
Note. N= 53
* p< .05 ** p< .01 *** p< .001.
______________________________________________________________________
______________________________________________________________________
Table 3
Variance Accounted for by Each Factor
________________________________________________________________________
Factor Initial Eigenvalues
Total
% of
Variance
Cumulative
%
1 17.374 23.165 23.165
2 3.709 4.945 28.110
3 2.656 3.541 31.652
4 2.385 3.180 34.832
5 2.133 2.845 37.677
6 1.840 2.454 40.130
7 1.493 1.991 42.121
8 1.473 1.963 44.085
9 1.331 1.775 45.859
10 1.242 1.655 47.515
________________________________________________________________________
________________________________________________________________________
Figure 1
Factor Scree Plot
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75
Factor Number
0
5
10
15
20
Eige
nval
ueScree Plot
________________________________________________________________________
______________________________________________________________________
Table 4
Factor Correlation Matrix
Factor 1 2 3 4 5 62 .4393 .608 .4194 .484 .488 .4005 .464 .476 .459 .4596 .331 .391 .293 .482 .4087 .365 .456 .249 .584 .401 .568
________________________________________________________________________
________________________________________________________________________
Table 5
Factor Pattern Matrix using Maximum Likelihood with Promax Rotation
Factor1 2 3 4 5 6 7
Ita77 .713Ita78 .590Ita73 .516Ita79 .516 -.223Ita70 .497Ite64 .487 .167Ita76 .478Ita72 .469 .244Ita75 .462 .178 .295Ita71 .457 .215Ita69 .435Ite66 .413 .318Ita80 .405 .155 .153Ite60 .399 .377Ita67 .371Ite58 .371 -.215 .266Log29 .357Ite62 .350 .257 .245Ite59 .334Ita74 .305 .177Lin47 .303 .189 .229Ite63 .296 .260Lin52 .287 -.211 .286 .251Spa33 .662Spa36 .651Spa35 .624Spa37 .619Spa40 .534Spa32 .520 .218Spa34 .388 .333 -.180Lin45 -.197 .380 .287 .240Log28 .368Spa31 .311 .237Log27 .177 .302 .179Spa39 .298 .228Kin17 .292 .211Kin16 .183 .272Mus5 -.232 .249 .226 .167Log24 .807Log26 .740
Log30 .724Log22 .703 .163Log23 .682Lin41 .557 .184Lin44 .290Kin20 .788Kin12 .686Kin13 -.162 .660 .161Spa38 .603Kin19 .462Kin14 .399Kin15 .373Kin21 .268 .239Mus8 .698 -.223Mus2 .670Mus3 .644Mus9 .569Kin18 .260 .547Mus6 .162 .514Mus11 .167 .417Lin46 .639Ite61 .366 .602 -.290Ite65 .405 .428Ite56 .265 .386Lin43 .309 .384Ite57 .230 .323Log25 .214 .273 -.318Lin42 .223 .151 .278 .170Lin50 -.156 -.161 .606Lin53 .195 .533Lin48 -.219 .508Ite54 .289 .409Lin51 .255 .326Ite55 .278 .194 .228 .311Lin49 .188 .271
Note. Values less than .20 not printed.
Abbreviations. Predicted item factors. Lin = Linguistic; Mus= Musical; Kin=
Kinesthetic; Log= Logical-mathematical; Ite= Interpersonal; Ita= Intrapersonal; Spa=
Spatial.
________________________________________________________________________
________________________________________________________________________
Table 6
Factor Structure Matrix
Factor1 2 3 4 5 6 7
Ita77 .628 .271 .303 .259 .252 .313 .289Ita72 .603 .275 .518 .319 .246 .238 .264Ita73 .595 .374 .394 .300 .247 .304 .334Ita80 .588 .338 .470 .447 .303 .368 .271Ita70 .561 .237 .401 .260 .267 .297 .301Ita78 .559 .312 .308 .249 .248 .229 .238Ita71 .555 .244 .458 .231 .245 .248 .265Ita67 .544 .365 .430 .406 .214 .292 .278Ita79 .541 .251 .338 .324 .295 .308Lin47 .538 .330 .465 .415 .291 .272 .397Ite64 .537 .182 .398 .224 .228 .220 .302Ita69 .520 .289 .375 .266 .224 .288 .266Ita76 .510 .291 .282 .272 .309 .334 .301Ita75 .498 .359 .288 .213 .214 .397 .225Ite63 .487 .367 .339 .347 .303 .371 .453Ite66 .470 .313 .278 .261 .228 .409 .219Log29 .463 .226 .339 .321 .278 .242 .269Ite62 .452 .247 .244 .418 .264 .275 .394Ita74 .405 .340 .232 .200 .285 .366 .353Ite59 .364 .159 .193 .260 .170 .218 .259Spa35 .333 .659 .258 .316 .321 .290 .360Spa33 .257 .656 .193 .311 .286 .222 .371Spa40 .385 .646 .335 .450 .366 .277 .353Spa36 .273 .640 .187 .304 .264 .242 .317Spa37 .279 .599 .189 .224 .240 .280 .283Lin45 .275 .567 .309 .330 .434 .543 .540Spa32 .192 .518 .199 .227 .362 .311Spa34 .341 .499 .375 .480 .196 .174Log28 .336 .453 .303 .346 .152 .153 .196Log27 .423 .441 .414 .360 .154 .194 .191Kin16 .392 .434 .310 .374 .314 .229 .246Spa31 .182 .408 .250 .265 .385 .343Kin17 .216 .402 .233 .326 .284 .338 .210Log30 .473 .283 .738 .345 .238 .189 .159Log24 .389 .169 .722 .299Log23 .453 .290 .704 .357 .255 .284 .208Log22 .432 .263 .697 .344 .258 .332 .253Log26 .408 .202 .696 .304 .170 .160 .153Lin41 .416 .315 .601 .311 .338 .387 .342Lin44 .427 .320 .456 .310 .339 .350 .337
Log25 .366 .337 .412 .280 .179Kin20 .329 .293 .355 .749 .278 .320 .282Kin12 .332 .280 .313 .676 .343 .234 .252Spa38 .355 .376 .363 .638 .178 .297 .201Kin13 .159 .227 .590 .349 .242 .347Kin19 .414 .345 .388 .573 .289 .254 .206Kin15 .334 .335 .334 .484 .271 .186 .203Kin14 .231 .278 .217 .425 .190 .165Kin18 .224 .270 .435 .661 .438 .464Mus2 .214 .238 .174 .270 .656 .260 .392Mus8 .302 .249 .226 .198 .652 .166 .384Mus3 .363 .313 .317 .290 .633 .297 .308Mus9 .262 .203 .598 .360 .395Mus6 .389 .324 .398 .395 .579 .332 .301Mus11 .317 .193 .303 .219 .409 .160 .182Kin21 .182 .340 .396 .408 .314 .334Mus5 .356 .228 .381 .270 .360Lin46 .419 .342 .352 .356 .276 .655 .306Ite61 .450 .279 .279 .255 .245 .573 .185Lin42 .462 .448 .458 .411 .503 .560 .529Lin43 .330 .484 .316 .307 .286 .520 .378Ite65 .497 .286 .300 .311 .219 .507 .293Lin52 .500 .222 .375 .371 .363 .506 .485Ite60 .484 .235 .265 .362 .324 .500 .321Ite56 .440 .232 .361 .318 .237 .458 .244Ite57 .305 .211 .157 .231 .192 .378 .218Lin53 .296 .425 .221 .271 .425 .334 .617Lin50 .315 .350 .237 .173 .311 .221 .569Lin51 .334 .288 .248 .304 .433 .501 .542Ite55 .459 .313 .242 .430 .333 .487 .524Ite54 .443 .336 .249 .370 .325 .270 .509Ite58 .453 .383 .174 .302 .359 .390 .492Lin48 .271 .259 .177 .238 .245 .458Lin49 .364 .396 .280 .247 .297 .373 .445Spa39 .211 .413 .215 .332 .343 .422
Note. Values less than .20 not printed.
_______________________________________________________________________
_______________________________________________________________________
Table 7
Representative Sample Scale Statistics
_______________________________________________________________________
Grades
K - 2* 3 - 8** all***
______________________________________________________________________
Scales m sd m sd m sd
Musical 50% 16 56% 18 54 18
Kinesthetic 50 17 54 18 53 18
Logical 44 18 53 20 51 19
Spatial 55 19 57 20 57 20
Linguistic 45 17 54 18 52 18
Interpersonal 49 16 54 18 53 18
Intrapersonal 46 17 55 18 52 18
_____________________________________________________________________
Note. N= 1,679,
* n=407, ** n=1,272. ***n=1,679
______________________________________________________________________
______________________________________________________________________
Table 8
Scale Means and SD Per Grades K - 8*
______________________________________________________________________
Grade Mean and SD
______________________________________________________________________
Scale K sd 1 sd 2 sd 3 sd 4 sd 5 sd 6 sd 7 sd 8 sd ____
Musical 48 18 51 17 51 18 57 19 59 18 58 18 56 18 53 19 55 19
Kinesthetic 47 15 50 18 53 19 61 20 61 19 58 18 53 17 50 18 50 17
Spatial 52 19 57 19 56 20 62 21 63 21 59 21 57 19 54 22 52 17
Logical 42 18 44 18 45 19 56 20 55 21 54 20 52 19 53 19 52 18
Linguistic 43 16 44 17 47 18 56 20 57 21 56 18 54 17 52 19 50 16
Interpersonal 47 17 49 14 50 19 53 21 56 21 56 20 53 17 54 18 54 16
Intrapersonal 46 17 46 16 46 18 53 19 53 23 57 19 54 17 54 18 55 15
Innovation 46 16 50 18 49 16 54 21 55 21 55 20 51 17 49 18 48 16
Technical 43 15 45 14 50 19 62 18 62 18 59 26 56 15 54 16 53 15
______________________________________________________________________
Note. N= 1,679
*Grade n: K=106, 1=155, 2=146, 3=152, 4=141, 5=187, 6=312, 7=320, 8=160.
______________________________________________________________________
______________________________________________________________________
Table 9
High and Low Mean Group Scores for Scales
_______________________________________________________________________
Scales
m Groups m Groups
High Low
_______________________________________________________________________
Kinesthetic
67 Dancers 55 Vocal
61 Gymnasts 55 LD
60 Drama 55 Gifted
Musical
72 Musicians 54 Sports
69 Vocalists 53 Artists
67 Orchestra 49 LD
Spatial
70 Artists 58 Sports
67 Gymnasts 56 Gifted
63 Musicians 52 LD
Logical-Math
71 Gifted 52 Sports
64 Orchestra 51 Gymnasts
61 Musicians 48 LD
Linguistic
66 Drama 56 Social
64 Vocalists 55 Sports
64 Orchestra 46 LD
Interpersonal
64 Dancers 55 Gymnasts
63 Drama 55 Sports
63 Orchestra 49 LD
Intrapersonal
68 Orchestra 57 Vocal
65 Dancers 57 Gymnasts
63 Gifted 47 LD
Innovative
60 Drama 53 Sports
59 Orchestra 53 Social
58 Artists, Musicians, Dancers 48 LD
Technical
65 Musicians 58 Artists, Gymnasts
64 Gifted, Dancers 56 Sports
62 Drama, Orchestra 50 LD
______________________________________________________________________
Note. Groups:
Visual and Performing Arts:
Dancers, n = 56;Vocal, n= 55;Drama, n= 55; Musicians, n= 115; Orchestra, n=28; Artists, n=
114.
Other Groups: LD, n= 44; Gifted, n= 100; Sports, n= 63; Gymnasts, n= 16; Social, n= 34.
_______________________________________________________________________
Table 10
Correlations Between MIDAS Scales and IQ Scores
________________________________________________________________________
I.Q. Scores
Verbal Perform Full Scale
MI _______________________________________________
Scales
Musical .41*** .13 .32**
Kinesthetic -.04 -.06 -.05
Logical .57*** .39** .54***
Spatial -.01 .09 .05
Linguistic .60*** .22 .48***
Interpersonal .14 .28*** .05
Intrapersonal .58*** .36** .54***
Innovation .23 -.04 .13
Technical .46*** .31** .44***
________________________________________________________________________
Note. N=62
AT= Academically Talented, n=53; LD= Learning Disabled, n=9.
Pearson correlation coefficients * p< .05 ** p< .01 *** p< .001
______________________________________________________________________