A Study Of The Relationship Between Multiple Intelligences Grammatical Errors

A Study of the Relationship between Multiple
Intelligences & Grammatical Errors
Zhuang Wei*
and Dhirawit Pinyonatthagarn
School of English, Suranaree University of Technology, Nakhon Ratchasima 30000, Thailand
Abstract
The purposes of the study were : 1) to make surveys on grammatical errors and multiple intelligences
of Chinese EFL learners; 2) to explore the possible relationship between grammatical errors and multiple
intelligences; 3) to make a comparison between two groups of the learners about the number of errors
after receiving two different kinds of instruction – the control group following undifferentiated instruction
ignoring the subjects’ conditions of multiple intelligences, and the experiment group receiving instruction
catering to each subject’s composition of multiple intelligences. The subjects were 74 first year students
of English major at Guangzhou Automobile College, China, all enrolled in 2009, belonging to two
classes, which was randomly designated as one control group and one experimental group. The method
used in the study was quantitative. The study was conducted through data collection by administering of
questionnaire, writing task, and error correction test to all the subjects. Then it was followed with data
analysis by examination of frequency distributions of multiple intelligences and grammatical errors,
bivariate correlation between the multiple intelligences and grammatical errors, and independent-samples
t test on the two groups of subjects. The study found that among the Chinese EFL learners: 1) the most
frequently made grammatical errors was the incorrect usage of lexical verbs; 2) the obvious intelligence
displayed in strength was intrapersonal intelligence; 3) the most obvious correlation was a negative one
found between visual intelligence and article errors; 4) the experimental group obtaining multiple
intelligences based instruction performed better in error correction than the control group which followed
undifferentiated instruction.
Keywords: Chinese EFL learners; grammatical errors; multiple intelligences
Introduction
Many studies on error analysis and treatment have been conducted since 1960s.
However, it still deserves further investigation in light of various situations and perspectives
such as a study from the perspective of the learners’ multiple intelligences. It is out of
question that a large number of sources have been confirmed to be attributive to EFL
learners’ errors according to the past research. For example, Richards (1971) cites four
* Corresponding author
E-mail address: mrzhuangwei@gmail.com; dhirawit99@yahoo.com
Suranaree J. Soc. Sci. Vol. 5 No. 2; December 2011 (29-50)

30 Relationship between Multiple Intelligences & Grammatical Errors
major causes of errors: (1) overgeneralization, (2) ignorance of rule restrictions, (3)
incomplete application of rules, and (4) false concepts hypothesized. Later in his
1974 paper, he identifies six sources of errors, namely, (1) interference, (2)
overgeneralization, (3) performance errors, (4) markers of transitional competence,
(5) strategies of communication and assimilation and (6) teacher-induced errors.
According to Brown (2000), there are four sources of errors: A. Interlingual transfer;
B. Intralingual transfer; C. Context of learning; D. Communication strategies. However,
few of studies investigated the function of multiple intelligences in error making until
now. Intuitively, we have a faith that smarter people make fewer errors. Extended
from this common sense, the researcher tried to investigate their relationship when
the smartness was broken down in multiple intelligences, and errors were specified
as grammatical errors. Recently, there is an uptrend in the application of multiple
intelligences theory into the study of foreign language learning and teaching. For
example, Mahdavy (2008) finds among the multiple intelligences, only linguistic
intelligence contributes to listening proficiency. Moreover, Loredana and Aneliz (2011)
use interactive multiple intelligences tasks to support the EFL learners, ending with
good pedagogical results.
Put forward by Gardner (1983), that human intelligence is not some static reality
fixed at birth and measured well by standardized testing. Instead, it is a dynamic, ever-
growing reality in one’s life through several different intelligences. The multiple
intelligences (MI) theory is an important contribution to cognitive science and
constitutes a learner-centered philosophy that is “an increasingly popular approach to
characterizing the ways in which learners are unique and to developing instruction to
respond to this uniqueness (Richards & Rodgers, 2001: 123).”
Specifically, in line with Gardner (1999), the definitions of the nine separate
intelligences are as follows:
(1) Verbal (Linguistic) Intelligence: Sensitivity to the meaning of words,
grammar rules and the function of language to persuade, remember, convey information
and reflect upon language.
(2) Logical-Mathematical Intelligence: Ability to see relationships between
objects and solve problems, as in calculus and engineering.
(3) Visual (Spatial) Intelligence: Ability to perceive and mimic objects in
different forms or contexts, as in miming or impressionist painting.

31
Suranaree J. Soc. Sci. Vol. 5 No. 2; December 2011
(4) Musical Intelligence: Ability to hear tones, rhythms and musical patterns,
pitch and timbre.
(5) Bodily-Kinesthetic Intelligence: Using the body, perceptual and motor
systems in the brain to solve a problem, a well - honed sense of timing, an ability to
anticipate what is coming next, an overall smoothness of performance.
(6) Interpersonal Intelligence: Sensitivity to the actions, moods and feelings of
others, the ability to establish person to person relationships and to read the intentions
and desires of others.
(7) Intrapersonal Intelligence: Ability to understand and define inner feelings,
as in poetry and therapy, and the ability to reflect upon one’s own thoughts, feelings
and sense of self.
(8) Naturalist Intelligence: Be attached strongly to the outside world or to
animals, enjoy outdoor activities, and notice patterns and things from nature easily.
(9) Existential Intelligence: Enjoy thinking and questioning, curious about life,
and exhibit the proclivity to pose and ponder questions about life, death and ultimate
realities.
Given such rich connotations contained in the multiple intelligences, it is natural
for the researcher to pose and ponder the corollary that there might be pertinence
between multiple intelligences and grammatical errors. Subjectively, this association
is not a miracle in that we human beings are born with curiosity and impulsion to
explore the unknown world. The eminent physicist Zeilinger (2011:82) says, “We
would not have our civilization if people weren’t curious about things. To me this is
the most important driving force in science.” The present research started as a result
of curiosity over why many EFL learners, in particular, Chinese EFL learners,
following the normal curriculum have little improvement in language accuracy in
writing, as well as an impulsion to solve the problem. Objectively, intelligence and
education existed for a long time in the form of two faces of a coin. Educational
achievement has always been evaluated in terms of personal intelligence. Especially
when Alfred Binet developed a kind to intelligence test called IQ (intelligence
quotient) test to predict academic performance in the early twentieth century, this
evaluation became systematic and scientific. The new developed multiple intelligences
theory and its applications in education including error analysis and error treatment
followed the previous tradition.

Specifically, there are three ways of describing errors as shown in figure 1. The
linguistic way put forward by Chamot (1978) is suitable for this study though it is
considered to be difficult to provide a satisfactory description of learners’ L2
development by quantifying the types of errors they make (Ellis, 1994). An
alternative to linguistic description of errors is to use a surface strategy taxonomy in
light of Dulay, Burt and Karshen (1982) by means of such operations as omissions,
additions, misinformation and misorderings. According to Murrow (2005), the
operations in the surface strategy taxonomy was renamed as missing, superfluous,
incorrect and misplaced, and put to use in the present study. The renaming make the
operations sound more objective in the perspective of the reader. The third way of
describing proposed by Corder (1974), who distinguishes three types of errors
according to their systematicity, concerns more on how learners learn an L2. The
systematicity way did not enter the study as it involves identification of the EFL
learners’ awareness of the existence of a particular rule, which is opaque and tricky
to the researcher except the learners themselves. As a result, this study adopted the
linguistic way combined with surface strategy taxonomy to describe errors made by
the EFL learners, because these two ways are more direct and easier to use in coding
grammatical errors.
Error Description
Linguistic Way Systematicity Way
Omissions, Additions,
Misinformation, Misorderings
Surface Strategy Taxonomy
Presystematic,Systematic
Postsystematic
Quantifying
the types of errors
Figure 1 Error Description
In short, the paper tried to answer four questions: (1) What kinds of errors do
the EFL learners frequently make? (2) What are the characteristics of multiple
intelligences of the EFL learners? (3) Is there any relationship between the EFL
learners’ multiple intelligences and the types of errors they make? (4) Is there any

33
difference in error-correction between the two groups of EFL learners after they have
received two different kinds of instruction?
Methodology
The present research adopted a quantitative approach. Specifically, research
questions one and two involve descriptive operations such as frequency and
percentage. Research question three is engaged with correlation operations. There are
many kinds of correlation, but this study will use Pearson correlation, because
Pearson method should be used only when each variable is quantitative in nature
(Gupta, 1999). In question three, two sets of variables, multiple intelligences and
grammatical errors, were transformed into numbers before they were put into
Pearson correlation analysis. But not all correlations of the two variables will be
taken into account. According to Cohen et al. (2002), correlation may refer to any
statistically significant relationship between two random variables or two sets of
data. Only those that are statically significant, specifically, the sig. value of the
correlation is smaller than 0.05 will be discussed. Research question four entails a
pre-test and post - test process which is a quasi-experimental design applied to the
nonequivalent groups (two intact classes but the subjects being non-randomized in
this case) in essence. The pre-test is a process that measures the difference of the two
groups before intervention, whereas the post-test is the process that identifies the
distinction of the two groups after the intervention. The dependent variable, which is
the grammatical error of the EFL learners, was measured both before and after the
treatment or intervention, which is independent variable, as depicted below:
NR—O—X 1—O
NR—O—X 2—O
NR stands for non-randomization.
O stands for observation.
X1 and X2 stand for different interventions.
The simplest method to compare the mean score of the two groups is independent-
samples t test. Specifically, only when the sig. value of Levene’s test for equal
variances assumed in the output of independent–samples t test is less than 0.05, the
two groups are considered to be significant different (Gupta, 1999).

Subjects
Constrained by the educational setting and with no intention to universalize the
present discoveries, the researcher adopted the non-probable or purposive sampling,
intended to provide reference to and pin hope for future study. The subjects included
in the present study came from Guangzhou Automobile College of China. Seventy-
four first-year English majors were chosen to participate in the study. They were
belonging to 2 classes. In order to keep students’ gender and English level balanced,
the student administration divided them into two classes roughly even but not exactly
equal. By drawing lots, class 1 bearing 38 students was designated as experimental
group, and class 2 having 36 students was specified as control group. They were all
enrolled in 2009, and the researcher was their incumbent English teacher.
Instruments
There were three instruments - writing task, questionnaire, and error correction
test. The data collected through three instruments were finally coded into numbers
for statistical analysis. For example, the writing task was designed to collect numbers
of errors. Though nominal measures were used in marking errors, the quantity of the
errors was expressed in a number. Similarly, the questionnaire measured intelligences in
numbers, and the error correction test measured errors corrected in numbers. So in
essence all the three instruments used a ratio scale of measurement represented by
continuous numbers.
The first instrument, the writing task, employed in the study is a controlled
composition, which has following requirements:
1) Topic: Learn by Yourself or with a Teacher?
2) Tips: Some people think that they can learn better by themselves than with a
teacher. Others think that it is always better to have a teacher. Which do you prefer?
Use specific reasons to develop your essay.
3) Length: Around 120 words.
4) Time: 60 minutes.
In accordance with the actual situation of the EFL learners’ writing, each error
found in the subjects’ writing was dually identified as linguistic errors and surface
errors simultaneously. The linguistic errors are broken down into 10 types, i.e. errors
of auxiliary verb, lexical verb, noun, adjective, adverb, conjunction, article, pronouns
preposition, and punctuation symbol (According to Crystal (2003), a lexical verb is a

35
member of an open class of verbs that includes all verbs except auxiliary verbs.
Lexical verbs typically express action, state, or other predicate meaning. In contrast,
auxiliary verbs express grammatical meaning). The surface errors are categorized
into 4 classes: missing errors, superfluous errors, incorrect errors, and misplaced
errors. The error was marked on the examination paper with the acronyms of
denomination of both the linguistic error (i.e. the error identified in linguistic way)
and the surface error (i.e. the error matched to surface strategy taxonomy) as shown
in the appendix where the subjects’ writing samples are attached. Table 1 below
shows acronyms for the errors.
Table 1: Acronyms for Errors
Linguistic Category
Surface Strategy Taxonomy
Missing Superfluous Incorrect Misplaced
Auxiliary Verbs AVMS AVS AVI AVMP
Lexical Verbs LVMS LVS LVI LVMP
Nouns NMS NS NI NMSP
Adjectives AdjMS AdjS AdjI AdjMSP
Adverbs AdvM AdvS AdvI AdvMSP
Conjunctions ConjMS ConjS ConjI ConjMSP
Articles ArtMS ArtS ArtI ArtMSP
Pronouns PronMS PronS PronI PronMSP
Prepositions PrepMS PrepS PrepI PrepMSP
Punctuation Symbols PSMS PSS PSI PSMSP
The second instrument is a survey questionnaire. The questionnaire introduced
into the study is in English, developed from McKenzie (1999). In order to reduce
ambiguity, the English versioned questionnaire was translated into a Chinese one by
the researcher and double-checked by the peer researchers – Mr. Wang, a Ph.D
student of English of SUT - before it was administrated to the Chinese students. The
questionnaire and its Chinese version are attached in the appendix.
The questionnaire consists of 9 sections, each of which contains 10 pieces of
short statements. Each section corresponds to a particular intelligence. The subjects

were required to complete each section by placing a “1” next to each statement they
felt accurately described them, and leaving the statement they did not identify with
blank. Then they were asked to total the column in each section. Finally the total of
each section was multiplied by 10 to facilitate statistical operation. So each surveyed
subject ended up with a table as follows that characterizes his/her various
intelligences. The higher the score is, the stronger the intelligence is.
Table 2: Table of Multiple Intelligences
Section Intelligences Total Score Total Score×10
1 Naturalist
2 Musical
3 Logical
4 Existential
5 Interpersonal
6 Kinesthetic
7 Verbal
8 Intrapersonal
9 Visual
The third instrument is an error correction test. The error correction test
comprised 50 sentences, all of which were extracted and adapted from the EFL
learners’ writing, and each of which contained at least one error which had been
instructed differently in both control group and experimental group prior to the test.
The score of error correction was decided by the number of the errors corrected. The
learners would gain 0.5 score by successfully correcting one error. The samples of
error correction test are also attached in the appendix for reference.
Procedures
Generally speaking, to draw out errors, the writing task, employed in the study
for both the control group and the experimental group, was a controlled composition.
After the writing task, the subjects were expected to answer the closed-ended
questionnaire to procure numeric data on the distribution of the EFL learners’

37
multiple intelligences. Finally, all the subjects of the experimental and control group
attended the error correction test in the wake of different instruction.
It is worth mentioning that the writing task as an instrument to elicit errors for
investigation was embedded in a mid-term examination among other examination
items. The mid-term exam consists of three parts: reading comprehension, vocabulary
knowledge, and writing task. The total exam time is 180 minutes, within which 60
minutes was specified for the writing of a given topic with required word number.
The intention of the single blinded maneuvering was to prevent the subjects from
knowing of the experiment, and prevent them from putting extra efforts on the task
consciously, and thus keep the data unbiased. Again, provided enough time to write
with the congruous topic and limited with the definite word number, the subjects
were led to production of authentic and valid data. For the same reason, the error
correction test is not an individual test but a component of another examination.
Referring to Puchta and Rinvolucri (2005), the researcher designed two kinds of
instruction - undifferentiated instruction and MI based instruction – for the control
group and the experiment group respectively. Specifically, in class two, all the
grammatical errors found in the compositions were listed, analyzed and explained
one by one, coupled with related grammatical knowledge introduction and followed
by an excise that the student who made a particular error is required to correct it by
him/herself and than cite an example containing the grammatical knowledge he/she
has just learned in written form. In class one, all the instruction process is the same
as in class two except the demand that an error made by a particular student was
required to be corrected and exemplified by him/her in a way that suitable to his/her
strength in some aspect of the multiple intelligences. For example, if a student who
shows the strength in musical intelligence, he/she would be demanded to create
orally, either on his/her own or mimicking a model prepared by the researcher, a
rhythmic or rhymed verse containing the grammar knowledge that he/she was taught
to reinforce the new acquired knowledge. Such personalized instruction for each
student of the experiment group carried out only once. On average each student was
allocated 2 minutes to be instructed plus 3 minutes to rehearse. The experiment
group consisted of 36 students, thus altogether 180 minutes or about four classes
were spent.
Apart from the researcher himself, the research team was composed of another

two English teachers –Vernon and Jenny, who are native speakers of English coming
from Britain and teaching at the same college with the researcher. Each member of
the team played an important role over the study. Before the marking of the EFL
learners’ writing, the team has reached the agreement on the ways of error classification.
During the error marking process, the researcher did the first round, and then Vernon
and Jenny double - checked the marking results, and discussed together whenever
any divergence occurs. The same process was applied to the error correction test. The
following figure illustrated the research process.
Figure 2 Research Process
In figure 2, circles represent instruments, short rectangles stand for groups,
middle rectangles represent intervention, and the biggest rectangles stand for analysis
and comparison. The whole process is composed two sub - processes. The first sub -
process starts from giving writing task to the two groups and ends in analysis of
relationship. The second sub - process begins from two groups receiving different
instructions and stops at comparison of results.
Results
Corresponding to the research questions, the study yielded the following
findings which are described under each sub-title as follows.

39
Findings for Question One
The most frequently made grammatical errors were lexical verb errors in the
linguistic classification, and the incorrect usage errors in surface strategy taxonomy.
The least frequently made errors were adverb errors in the linguistic classification,
and the misplaced usage errors in surface strategy taxonomy. The findings related to
research question one were shown in Table 3 and Table 4 below.
Table 3: Frequency and Percentage of Linguistic Errors
Frequency Percent Valid
Percent
Cumulative
Percent
Valid
1=Auxiliary Verb Errors 86 9.1 9.1 9.1
2=Lexical Verb Errors 188 19.9 19.9 29.1
3=Noun Errors 154 16.3 16.3 45.4
4=Adjective Errors 57 6.0 6.0 51.4
5=Adverb Errors 39 4.1 4.1 55.6
6=Conjunction Errors 54 5.7 5.7 61.3
7=Article Errors 74 7.8 7.8 69.1
8=Pronoun Errors 90 9.5 9.5 78.7
9=Preposition Errors 95 10.1 10.1 88.8
10=Punctuation Symbol Errors 106 11.2 11.2 100.0
Total 943 100.0 100.0
Table 3 summarized the statistical results of frequency and percentage of
linguistic errors made by the EFL learners in their writing. Those linguistic errors
totaling 943 were classified into ten categories, each frequency and percentage of
which was listed ranging from the adverb error number of 39 or 4.1% to the lexical
verb error number 188 or 19.9%.

Table 4: Frequency and Percentage of Surface Errors
Frequency Percent Valid
Percent
Cumulative
Percent
Valid
1=Missing 233 24.7 24.7 24.7
2=Superfluous 147 15.6 15.6 40.3
3=Incorrect 539 57.2 57.2 97.5
4=Misplaced 24 2.5 2.5 100.0
Total 943 100.0 100.0
Similarly, Table 4 numerated frequencies and percentages of surface errors
made by the EFL learners in the same writing, which has the identical error number
in sum but are sorted in another way. Under such way, the most frequent errors made
are incorrect usage, totally 539, occupying more that half of the total which is 57.2%,
while the least frequent ones are misplaced usage counted 24 taking 2.5%. In sum,
the most frequent errors in linguistic category occurred in the use of lexical verbs,
nouns and punctuation marks; the most frequent errors under surface strategy
taxonomy are incorrect usage errors.
The errors of lexical verbs were centered around incorrect uses of verb tenses,
verb voices, non-finite verb forms, subject-verb agreement, and taking nouns for verbs.
Some sentences containing lexical verb errors extracted from the compositions are as
follows.
1) When we were children, we don’t* know many things.
(A tense error made by Lai Fangfang)
2) …and your knowledge is enrich*.
(A voice error made by Lin Yuanling)
3) Learn* with a teacher, we can never feel lonely.
(A non-finite error made by Han Jiaqi)
4) Learning by ourselves mean* we will….
(A subject-verb agreement error made by Peng Shaoting)
5) If we want to success*, we must study hard by ourselves.
(An error of taking noun for verb made by Xu Siming)

41
The noun errors mostly appeared in wrong uses of single and plural nouns, and
putting verbs and adjectives in the place of nouns, etc. Some examples quoted from
the compositions are as follows.
1) Maybe they are good at how to arrange their times*.
(A plural noun error made by Li Yanghua)
2) I have three reason* to stand my opinion.
(A single noun error made by Li Chunrong)
3) I think learn* with a teacher is better than by myself.
(An error of taking verb for noun made by Li Huaxiu)
4) We can learn by ourselves in this small social*.
(An error of taking adverb for noun made by Hao Jing)
Punctuation symbols actually play many grammatical functions that indicate the
structure and organization of written language. The most frequent punctuation
symbol errors are the incorrect uses of comma and period, which may induce run-on
sentences and fragment sentences that are exemplified in the following.
1) We often need learning,* learning make a progress.
(A comma error made by Chen Mengyuan)
2) If we want to get more knowledge. *We must learn by ourselves.
(A period error made by Hao Jing)
3) The most important way of learning something is learning by ourselves, but
we can’t leave the help of teachers.*Especially in the university.
(A period error made by Ke Dixiao)
Findings for Question Two
The findings corresponding to research question two were that the EFL learners
displayed strongest strength in the intrapersonal intelligence, and the weakest in the
interpersonal intelligence, as revealed in the sum scores, means and percentages in
table 5. The sum scores of nine intelligences were produced by adding together the
total scores of each intelligence for each subject in table 2. The means and percentages

of the strength of each intelligence among the nine intelligences were drawn out by
running descriptive statistics in SPSS. When we look at the scores and percentages of
the nine intelligences, we are taking account of all subjects as a whole.
Table 5: Frequency and Percentage of Multiple Intelligences
MI Naturalist Musical Logical Existential Interpersonal Kinesthetic Verbal Intrapersonal Visual
Sum Score 390 368 374 445 351 396 360 500 364
Mean 5.27 4.97 5.05 6.01 4.74 5.35 4.86 6.76 4.92
% 10.99% 10.37% 10.54% 12.54% 9.89% 11.16% 10.15% 14.09% 10.26%
In table 5, the sum score of intrapersonal intelligence is 500, which is the biggest
of all the nine intelligences. The same are true for its mean, which is 6.76, and its
percentage, which is 14.09%. Thus, the strongest intelligence displayed among the
students is intrapersonal intelligence. Contrasted to the intrapersonal intelligence, the
sum score, and its mean and percentage are the smallest comparing with those of
other intelligences. So, the weakest intelligence among the students is interpersonal
intelligences.
Findings for Question Three
The findings related to research question three were that there existed three
kinds of statistically significant correlation between linguistic errors and multiple
intelligences as shown in Tables 6, 7, 8, but there was only one such correlation
found between surface error and one of the multiple intelligences as displayed in
table 9. Those findings come from the analysis of Pearson correlation between MI
and linguistic errors as well as correlation between MI and surface errors. But not
every kind of error correlates to every intelligence significantly, because correlation
would not stand without statistical significance, which is literally decided by the sig.
value 0.05.

43
Table 6: Article Errors’ Correlations with Multiple Intelligences
Multiple Intelligences Article
Visual
Pearson Correlation -.413**
Sig. (2-tailed) .008**
Logic
Kinesthetic
Musical
Verbal
Existential
Naturalist
Table 6 is an excerpt of cells coming from the general table of correlation
between MI and linguistic errors. The cells constituting table 7 are chosen to display
because their sig. values in each pair of correlation are smaller than 0.05. Thus the
corresponding correlations in table 7 are statistically significant. Moreover they
correlate in opposite direction because the Pearson coefficients are negative. Among
the 7 pairs of correlations, visual intelligence correlates with article errors most
eminently, since the absolute value of correlational coefficient is the biggest of all,
though all the correlations are moderate statistically due to the fact that none of the
coefficients is smaller than 0.03.

Table 7: Preposition Errors’ Correlation with Existential Intelligence
Preposition
Existential
Pearson Correlation -.283*
Sig. (2-tailed) .044*
Table 7 is another cell from the general table of the correlation between MI and
linguistic errors, where the existential intelligence has a significant negative
correlation with the preposition error, as the sig value is 0.044.
Table 8: Punctuation Errors’ Correlation with Kinesthetic Intelligence
Punctuation
Kinesthetic
Pearson Correlation .290*
Table 8 is cognate as the above. Also for the same reason, that we found a
significant positive correlation between the kinesthetic intelligence and the
punctuation error.
Table 9: Missing Errors’ Correlations with Multiple Intelligences
Missing
Kinesthetic
Pearson Correlation .309*
Table 9 is a cell that gleaned from the general table of the correlation between
MI and surface errors. The relationship between the kinesthetic intelligence and the
missing error listed in table was the only correlation being detected significant
statistically in view of its sig. value being 0.013. Again, the kinesthetic intelligence
correlated with missing error positively.
All in all, among the above ten pairs of correlation, the most obvious correlation
is the negative one between the visual intelligence and article errors because the

45
absolute value of the correlation coefficient of the pair is 0.413, which surpasses all
the other statistically significant correlation coefficients. The negative correlation
between the visual intelligence and article errors can be interpreted literally as
stronger visual intelligence helps reduce article errors. Moreover, the kinesthetic
intelligence correlating positively with punctuation symbol errors and word missing
errors implies that stronger kinesthetic intelligence contribute to more punctuation
symbol and word missing errors. Thus, in brief, there are some relationship between
some of multiple intelligences and some language errors.
Findings for Question Four
The finding that answered research question four was that the multiple
intelligences based instruction improved the EFL learners’ performance of error-
correction significantly as attested by the pretest-posttest design.
In the process of pretest, both of the two groups received the same writing task.
Through recording of error number of each subject in his/her writing and inputting
the data of the error number of each subject with his/her class affiliation into SPSS to
do the independent – samples t test (The name lists of error number are attached in
the appendix), the researcher found the two groups had no statistically significant
difference in terms of error number as shown in Table 10.
Table 10: Independent-Samples T Test of Error Making
Levene’s Test for
Equality of
Variances
t-test for Equality of Means
F Sig. t df Sig.
(2-tailed)
Mean
Difference
Std. Error
Difference
95% Confidence
Interval of the
Difference
Lower Upper
ERRORS Equal
variances .072 .789 .412 72 .681 .72 1.737 -2.747 4.179
assumed
Equal
variances
not assumed
.411 70.019 .682 .72 1.743 -2.759 4.192

Checking the sig. value of Levene’s test in table 10, the researcher found it was
0.789 which was higher than 0.05, so the equal variances of the two groups can be
assumed. Therefore, we have 95% of confidence to say that the number of errors
made by the two groups is equal in the statistical sense. In other words, the
grammatical level of the two groups was observed to be the same before any
intervention.
In the process of post - test, the two groups were exposed to two different kinds
of error treatment for purpose of comparison. After the two kinds of instruction, an
error correction test was implemented to the two groups, the results of which were
attached in the appendix and summarized in Table 11.
Table 11: Error Correction Score
Class One
Experimental Group
Class Two
Control Group
Value Difference between
Class One and Class Two
Student Number 36 38 -2
Total Score 1040 992 48
Mean Score 28.89 26.11 2.78
The data in Table 11 were computational results from the records of scores
which were collected through an error-correction test conducted to two groups of
subjects. Before the test, class two was taken as a control group in which students
were treated with the same instruction disregarding the variation of multiple
intelligences in each learner, while class one was taken as an experimental group in
which students got tailored instruction based on his or her multiple intelligences. The
score represents competence of error correction. The higher the score is, the more
competent the learner is in dealing the error correction. Table 11 shows class one that
was experimented with MI based instruction got higher scores, which is 1040 in
total, and 28.89 on average. Table 11 also shows class two earned lower scores,
which is 992 in total, and 26.11 on average. Considering that the error number made
by the two classes had no statistically significant difference prior to intervention, and
all the errors appeared in the error correction test were extracted from their previous
writings, the discordance in the test which is 48 in total and 2.78 on average must be

47
the results of different instructional strategies.
If we were still not sure whether the difference had been spawned by
coincidence, then the independent-samples t test would exclude such a possibility to
great extent. Inputting the data of the error correction score of each subject in to
SPSS to do the independent – samples t test, and then checking the sig. value of Levene’s
test in the output table 12, the researcher found it is 0.045 which is lower than 0.05,
so we cannot assume equal variances. Therefore, the error correction difference
between the MI based instruction and the undifferentiated instruction was not out of
coincidence with 95% of confidence. The difference is significant in statistical sense;
in other words, the multiple intelligences based instruction did make a difference on
the students’ performance of error-correction.
Table 12: Independent-Samples T Test of Error Correction
Levene’s Test for
Equality of
Variances
t-test for Equality of Means
F Sig. t df Sig.
(2-tailed)
Mean
Difference
Std. Error
Difference
95% Confidence
Interval of the
Difference
Lower Upper
Score Equal variances
assumed
4.154 .045 1.969 72 .053 2.784 1.413 -.034 5.601
Equal variances
not assumed
1.990 65.283 .051 2.784 1.399 -.010 5.578
Discussion & Conclusion
The findings of the study have multiple implications for language researchers,
teachers, and learners. To language researchers, the finding for research question one
that the incorrect usage of lexical verbs accounts for the most of errors in their writings
confirmed Murrow’s investigation of Japanese EFL learners in Chinese context. The
possible explanation for the phenomenon could be due to the fact that the Chinese
lexical verbs lack variation of tense and aspect like English. Opposite to the lexical
verb, the errors related to incorrect usage of adverbs are least frequently made. The
reasons for it might be that adverbs are easily built and their positions are relatively
fixed in sentences. The finding for research question two was that the strongest

intelligence displayed in the EFL learners was intrapersonal intelligence, but no
significant relationship found between the intrapersonal intelligence and any
language errors. The prevalence of the intrapersonal intelligence among Chinese EFL
learners can be attributed to the influence of cultural heritage and teaching methods.
In Confucianism dominated Asian countries like China, introvert personality and
self-examination behavior are encouraged. For instance, Chinese people were
inculcated from their birth with the famous sayings of Confucius such as “A
gentleman finds faults with himself while a base man finds faults with others”,
“Think twice before acting”. Furthermore, the mainstreamed teacher centered
approach treats students the same way ignoring variation of intelligences among
them. Thus, students who are unduly required to adapt themselves to the instruction
by their teachers become more intrapersonal. In contrast with the strongest
intrapersonal intelligence, the interpersonal was found to the weakest logically. The
finding for research question three was that the most obvious correlation was a
negative correlation found between the visual intelligence and article errors. This
finding can be explained by the fact that while articles function as spatial indicators
in effect, the students who are strong in spatial intelligence must be making less space
related errors like article errors. The visual intelligence, according to its definition, is
space related intelligence. Thus, the stronger the visual intelligence is, the fewer
article errors are. This finding is supplemental to the study by Mahdavy (2008), who
finds the linguistic intelligence contributes to listening proficiency. His study shows
scores of each intelligence positively correlate with listening scores of both TOEFL
and IELTS listening but only linguistic intelligence has a statistically significant
correlation with listening. The finding for question four that the MI based instruction
improved the EFL learners’ performance of error-correction significantly substantiated
the claim by Loredana and Aneliz (2011) that application of multiple intelligences
supports the EFL education.
To language teachers and language learners, the findings that indicated the
distribution of grammatical errors of the EFL learners can be used as a learning guide
to help the language teachers to organize learning materials and the language learners
to pay particular attention to their language deficiency. Thus the teachers teach
purposefully and learners learn efficiently. The findings that characterize the
intelligences composition of the learners enabled the learners to understand

49
themselves and enhance their self-confidence. The findings that revealed some
significant relationships between part of the multiple intelligences and part of
grammatical errors contributed to the improvement of teaching strategy to achieve
better teaching results. The findings that validated the evident effect in the treatment
of grammatical errors offered an alternative to innovate language education.
Further studies are needed to randomize the subjects and expand the selection
participants so that the findings are more persuasive. Also, it is desirable to refine the
instruments to measure data more accurately. Last but not least, it is the researcher’s
hope that the future studies should produce better results if they are done both
quantitatively and qualitatively.
Acknowledgement
The paper is a summary report of my Ph.D. thesis supervised by Dr. Dhirawit
Pinyonatthagarn, School of English, Suranaree University of Technology. I thank my
supervisor for his patience, guidance, and encouragement. I am particularly thankful
to Dr. Sirinthorn Seepho who provided me a lot of constructive suggestions and kind
encouragement in the revision of the paper. Also I would like to take this opportunity
to express my heartfelt gratitude for their help, without which I cannot have achieved
what I have today. Thanks also go to my family, my friends and all the teachers in
the school of English who provided their moral support and various types of
assistance.
References
Brown, H.D. (2000). Principles of Language Learning and Teaching. New Jersey: Prentice-Hall.
Chamot, A. (1978). Grammatical problems in learning English as a third language. In E. Hatch (Ed.), Second
language acquisition. Rowley, MA: Newbury House. pp. 175-189.
Cohen, J., Cohen P., West, S.G., & Aiken, L.S. (2002). Applied multiple regression/correlation analysis for the
behavioral sciences (3rd
ed.). New Jersey: Lawrence Erlbaum Associates, Inc..
Corder, S.P. (1974). Error Analysis in Allen, J.L.P. and Corder, S.P. (eds) Techniques in Applied Linguistics.
Oxford. Oxford University Press. pp. 122-154.
Crystal, D. (2003). A Dictionary of Linguistics & Phonetics (5th
edition). New York: Wiley-Blackwell.
Dulay, H., M. Burt, and S. Krashen. (1982). Language Two. New York: Oxford University Press.
Ellis, R. (1994). The Study of Second Language Acquisition. Oxford: Oxford University Press.

Gardner, H. (1999). Multiple Intelligences: The theory in practice. New York: Basic Books.
Gardner, H., (1983). Frames of mind. The theory of multiple intelligences. New York: Basic Books.
Gupta. V., (1999). SPSS for Beginners. Georgetown: Georgetown University Press
Loredana-Andreea Stancuna & Aneliz-Iulia Craciun. (2011). A multiple intelligences approach: intuitive English
learning – a case study for k – 1 students. Procedia Social and Behavioral Sciences 11, pp. 72–76.
Mahdavy, B. (2008). The Role of Multiple Intelligences (MI) in Listening Proficiency: A Comparison of TOEFL
and IELTS Listening Tests from an MI Perspective. The Asian EFL Journal, September 10, pp. 109-126.
McKenzie, W. (1999). Multiple Intelligences Inventory. Retrievable at the website <http://surfaquarium.com/
Mi/inventory.htm>, 2009.
Murrow, P. (2005). Analysis of Grammatical Errors in Students’ Writing. MCT Journal. No.40, pp.21.
Puchta, H. & M. Rinvolucri. (2005). Multiple Intelligences in EFL. West Sussex: Grafica Veneta.
Richards, J. C. (1971). A Non-Contrastive Approach to Error Analysis. English Language Teaching, 25, No. 3.
Richards, J. C. and G. P. Sampson. (1974). The Study of learner English. Error Analysis: Perspectives on
Second Language Acquisition, ed. J. C. Richards. London: Longman.
Richards, J. C., Rodgers, T. S., (2001). Approaches and Methods in Language Teaching, 2nd
ed. Cambridge:
Cambridge University Press, pp.123.
Zeilinger. A. (2011). Q & A: Anton Zeilinger. Discover Magazine. Waukesha: Kalmbach Publishing Co. Vol. 32,
No.6, pp.82.

A Study Of The Relationship Between Multiple Intelligences Grammatical Errors

More Related Content

Similar to A Study Of The Relationship Between Multiple Intelligences Grammatical Errors (20)

More from Sandra Long (20)

Recently uploaded (20)

A Study Of The Relationship Between Multiple Intelligences Grammatical Errors