October 4, 2011

ETEC – 570

0
 October 4, 2011
ETEC – 570

Literature Review
Learning Styles and
Multiple Intelligences
10/1/2011

1
 October 4, 2011
ETEC – 570

Literature Review Learning Styles and Multiple Intelligences

Though researchers create and test learning styles and multiple intelligences measures,

in order to research instructional methods and learning styles, the literature suggests that the

reliability of such instruments remains a consistent problem while studies indicate that there

are no significant differences. During this literature review it became apparent that there was a

dearth of empirical studies on learning styles and multiple intelligences. The thrust of such

scholarly investigation primarily concerned instrument development, as opposed to application.

Perforce, the majority of articles reviewed pertain to the measures themselves, rather than

their utilization.

One of the few articles found that actually tested and applied a measure of learning

styles and multiple intelligences was conducted by Rezaei and Katz (2004), "Evaluation of the

Reliability and Validity of the Cognitive Styles Analysis". Rezaei and Katz applied Riding's

Cognitive Styles Analysis (CSA) to multiple populations in order to determine its reliability as an

instrument (2004). The measure is administered via computer and examines two dimensions

along four scales. The two dimensions consist of the Wholist-Analytic (WA), and the Verbal-

Imagery cognitive styles. Though the authors applied the measure in three separate studies,

under different conditions, they found its reliability so varied that the tool could not be

considered valid. The Verbal-Imagery dimension was particularly problematic based on the low

variance of scores. Rezaei and Katz subsequently made multiple suggestions to improve the

CSA's reliability, including increasing the level of difficulty in the test for the aforementioned

dimension. They noted that the CSA's reliability in test-retest trials was especially problematic,

despite having an underpinning conceptual model that was sufficiently robust.

2
 October 4, 2011
ETEC – 570

By contrast, Furnham (2009) employed his own instrument, Self-Assessed Multiple

Intelligence, a self-reporting measure, to examine eight of Gardner's multiple intelligences. His

study entailed a comparison of the aforementioned measure of concurrent and discriminant

validity testing, employing the following tools: Haselbauer's Multiple Intelligences Test, The NEO

Personality Inventory—Revised, Irwings's General Knowledge Test, and Biggs's Study Process

Questionnaire. He created a theoretical background based on the Big Five Personality test, a

free online self-assessment, to compare and contrast his findings. Though the author notes that

Irwings's General Knowledge Test is a performance test, he categorizes all of the others, save his

own, as being personality-based preference tests. Upon reading this, one must wonder

whether this apples-to-oranges approach to research invalidates any findings, let alone

reliability or validity.

Both the Abstract and the Methods sections state that there were 187 participants.

However, the Results section contradicts this information, indicating that there were a total of

18 participants who completed the multiple intelligence test twice, over a six week period. In

the subsequent paragraph, there is an obvious misprint (test is misspelled as rest). There is no

information in either the text or the tables clarifying the disparity between these numbers. A

further confounding element appears in the Discussion section, wherein the author notes that a

bigger sample size would have yielded better results. The author concludes on a mystifying

note stating that since the quality of empirical evidence for multiple intelligences is so poor that

many researchers believe creating measures to test this theory is counterproductive.

"The Validity and Reliability of the Marmara Learning Styles Scale (MLSS)" authored by

Mustafa Otrar was initially promising, despite the fact that this publication was more devoted to

3
 October 4, 2011
ETEC – 570

developing a tool than researching its application. Otrar used Dunn's learning styles theory to

generate an item bank. The author describes 6 rounds of internal consistency analysis, applied

to a 223 items scale that yielded a 100 item scale with a Cronbach's Alpha value of p=0.9452

rounded to p=0.95. Factor analysis was applied to enhance validity, with one item being

removed for a small load value (0.46), resulting in a 99 item scale. Test-retest reliability was

determined by two analyses: t-test and Pearson product moment correlation. This study stood

out because of the large sample size (N=1643) used in developing the tool, which was reduced

for various undeclared reasons to N=909. Twenty two learning styles were identified.

At this point, the author lapses into pure speculation, positing that differences in

learning styles were attributable to "cultural differences". The author also claims that unknown

external variables affected the internal validity of his findings. Otrar makes bold assumptions in

asserting that "It is presumed that such difference stems from cultural differences" (Otrar, 2007,

p.1417). The author quickly concludes the article by stating "…that cultural diversity may cause

difference [sic] in learning styles" (Otrar, 2007, p.1417).

M. Onur Cesur and Seval Fer translated the English language Learning Style Survey (LSS)

into Turkish, and then proceeded to investigate its validity and reliability. They conducted a

descriptive level quantitative research study, entitled "What Is Validity and Reliability Study of

Learning Style Survey?" that was published in The Journal of Theory and Practice in Education.

Unfortunately, they do not disclose that the English version of the LSS had never been

studied for reliability until their Results and Suggestion section, placing this very important

piece of information at the end of the article. This, in and of itself, is arguably unethical, calling

into question the reliability and validity of their research.

4
 October 4, 2011
ETEC – 570

Other issues stemming from this inquiry involved their small sample population of

768 English Prep students, enrolled in multiple Turkish universities. The authors admit that

other researchers called for a 10:1 participant to item ratio, in order for a measure to provide

reliable results that can also be used for higher order empirical research (Cesur and Fer, 2009,

p.293). Their version of the LSS included 52 items, 12 subscales and had a Cronbach's Alpha

coefficient of 0.88. The measure of inductive style had a Cronbach's Alpha coefficient of 0.42,

which is far below the standards for Social Science research. Overall, the authors were more

concerned with translation issues and validity than reliability. The sheer number of pages

dedicated to translation and validity would make this obvious to novice researchers.

In the final analysis, developing tools to test learning styles and multiple intelligences

comprised the bulk of literature found that had not been presented in class. This was a limiting

factor as the authors of this paper did not want to duplicate previous research. Therefore,

literature related to measurement development necessarily constitutes the bulk of articles

reviewed herein.

This, combined with the limited number of empirical studies suggests that this field is

either developing slowly or that its foundational theories are not conducive to empirical

research.

More empirical studies, such as Rezaei and Katz's, should be undertaken to determine if

the conceptual models of learning styles and multiple intelligences are, themselves, sound.

5
 October 4, 2011
ETEC – 570

References

Cesur, M.O., & Fer, Seval (2009). What is validity and reliability study of learning style survey?.

Journal of Theory and Practice in Education, 5(2), 289-315.

Furnham, A. (2009). The validity of a new, self-report measure of multiple intelligence. Current

Psychology, 28(4), 225-239.

Otrar, M. (2007). The validity and reliability of the Marmara learning styles scale (MLSS).

Educational Sciences: Theory and Practice, 7(3), 1403-1419.

Rezaei, A. R., & Katz, L. (2004). Evaluation of the reliability and validity of the Cognitive Styles

Analysis. Personality and Individual Differences, 36(6), 1317-1327.