1

English Language Proficiency Level of BangladeshiStudents:

A Social Psychological Study

Md. Kamrul Hasan

Assistant Professor, English English Language Institute

United International UniversityDhanmondi, Dhaka-1209

Bangladesh

2

3

Content

page

Chapter 1: Introduction 5

Chapter 2: Cloze procedure and

Theoretical Background 12

Chapter 3: Methodology 25

Chapter 4: Result and Analysis 36

Chapter 5: Conclusion 66

Appendices

Appendix A: Cloze Test

Appendix B: The Questionnaire

Appendix C: Variable, Mean and Coding

Appendix D: Mean, Standard Deviation and

Correlation Matrix

Bibliography

4

5

Introduction

In the last three decades or so, the research work in second language

learning has brought out the significance of socio psychological aspects of second

language learning. In the context of successful foreign/second language learning,

several researchers have stressed on the significance of learner variables such as

attitudes, aptitude, motivation (motivational intensity/motivational orientation),

authoritarianism, ethnocentrism, etc. This work shows that achievement in a

second language is related to measures of attitudes and motivation. Two questions

come to the fore in spite of these relations. One of them concerns the validity of

the measurement of attitudes and motivation. The other concerns the precise

meaning of the correlation between measures of attitudes and motivation on the

one hand and second language achievement on the other. Though the nature of

analysis has generally been correlation implication which prompts the obtained

relationships to indicate that attitudes and motivation contribute to determine

achievement in the second language or that proficiency in the second language

influences attitudes and motivation.

The study of attitudinal and motivational variables is gaining in significance

as can be seen from some recent publications in the field. The pioneering research

work in this field was established by the contributions of Canadian scholars such

as W.E Lambert, R.C. Gardner and their associates. Lambert and their associates

conducted a series of studies over a period of twelve years in the U.S.A, Canada

and Philippines. Several of their studies were in Canada (e.g. Gardner and Lambert

1959, Gardner 1966, Anisfeld and Lambert 1961, Lambert, Gardner, Barik and

Tunstall 1962 and Feenstra and Gardner1968). Their findings show that along with

verbal intelligence and language aptitude, attitudes and motivation play

considerable importance as variables in second language proficiency. Gardner and

Lambert tested their initial hypothesis outside Canada in three American settings

(Louisiana, Maine and Connecticut). They were quite successful in establishing

6

that a friendly outlook towards the other group whose language is being learned

can ‘differentially sensitize the learner to the audio-lingual features of the

language, making him more perceptive to pronunciation and accent than is the

case for a learner without this open and friendly disposition’. (Gardner and

Lambert 1972: 134). According to Gardner and Lambert (1972) and others,

motivational orientation can be divided into four different types: integrative,

instrumental, resentment and manipulative. They made a distinction between

integrative and instrumental motivation. A motivational orientation is said to be

integrative if a person learns a foreign /second language in order to enter into an

active interaction with the target language speakers, that is, socially and culturally

he wishes to become a part of the target language community. It is instrumental if

one learns the target language to get a job, etc., that is a person learns a second

/foreign language with utilitarian objectives, such as obtaining admission in a

particular course or for trade purpose. They do not explore the last two

motivations in detail. Gardner and Lambert (1972) emphasize the role of

integrative motivation in second language learning situations. Their French-

American studies show that positive attitudes towards French-American culture

along with favorable stereotypes of the European French are associated with

certain expressive skills in French. Lambert and Gardner made a very significant

distinction between integrative and instrumental motivation and in most of their

studies conducted in Canada and America claimed that higher achievement in

foreign language learning correlated significantly with the motivation to integrate

with the target language. Gardner’s (1985) socio-educational model envisaged

integrative motive as a hypothetical construct comprising motivation, positive

attitudes to the learning situation (i.e. attitudes to the language teacher and the

course) and integrative (i.e. attitudes to the other community or communities,

integrative orientation and interest in foreign languages). They conclude that

success in foreign language learning would be less if the underlying motivation is

instrumental rather than integrative. Many studies provided considerable support

7

for Gardner and Lambert’s hypothesis that language proficiency will increase

directly in proportion to the intensity of integrative motivation (see Gardner 1985,

Gardner and ment 1990 for reviews). It is interesting to note from the

Philippines study that (reported in Gardner and Lambert 1972) they considerably

distanced themselves from their original position and thought that in settings

where there was an urgency about mastering a second language for utilitarian

ends, the integrative orientation won’t be prevailing. But, they found that the

instrumental approach to language study was extremely effective in the context of

Philippines.

Inspired by the work of Gardner, Lambert and their associates, research in

the social psychological aspects of second language acquisition gets intensified in

several other parts of the world (e.g. Lukmani 1972, Oller and Hudson and

Liu1977). These studies demonstrate that the measures of proficiency in second

language learning are significantly related to measures of attitudes and motivation.

However, several other research studies failed to find support for their hypothesis

and raised several questions. For instance, Burstall (1975) working with the

teaching of French in primary schools finds that although pupil’s attitude and

achievement prove to be closely associated, motivational characteristics of

individual pupil’s appear to be neither exclusively integrative now instrumental.

Lyczak, Fu and Ho (1976), Khanna and Agnihotri (1982,1984); Khanna (1983),

Au (1988), Oller, Hudson and Liu (1977) and Young and Gardner (1990) showed

that the theoretical claims of Gardner and Lambert lacked generality as the

variance in second language proficiency was not explained not so much by

motivational or attitudinal variables as by some social variables. Lyczak, Fu and

Ho (1976) found no significant correlation between achievement and attitudinal

variables. Wong (1982) obtained motivational orientation of Chinese students

learning English did not pose any correlation with their achievement. Au (1988)

found desire to learn English correlate negatively with English proficiency

measures. Au (1988) provided substantial evidence against the hypothesis that

8

integrative motive was positively related to second language achievement.

Agnihotri, Khanna and Mukherjee (1982,1988) measured proficiency in the use of

English tenses of 356 undergraduates of University of Delhi and correlated it with

19 sociopsychological variables such as sex, socioeconomic status, schooling,

exposure to English, pattern of language use, attitudes, claimed control on English

etc. They found that the proficiency of the undergraduate students in their use of

English tenses correlates significantly with their schooling, patterns of language

use and stereotypes of English. From their study, schooling came up as the most

significant predictor of achievement in the use of tenses in English. Khanna,

Verma, Agnihotri and Sinha (1990) found that proficiency in English of the adult

immigrants learning English as a second language in Britain correlates

significantly with claimed control, patterns of language use, exposure to English,

use of English in the family, motivational orientation and desire to learn English.

Mathur (1991) worked with twenty-nine students learning German in Delhi and

found that educational background; self-image and exposure to German were most

closely associated with proficiency in German.

From the above, it is easy to draw the conclusion that second/foreign

language learning is an extremely complex process including the contributions of

the learner and his environment. Gardner's (1985) socio-educational model is more

comprehensive as along with individual differences, it also includes the social

milieu, the formal and informal language acquisition contexts and the linguistic

and non-linguistic outcomes as important ingredients of the model. Although

Gardner’s socio-educational model includes ‘social milieu’ as an important

component, his firm belief is that second language learning is essentially a social

psychological phenomenon. Gardner (1988) says that social variables which

reflect the complexity and dynamism of a given social situation can in fact prove

to be confounding variables.

Earlier, we have mentioned that the hypothesis put forward by Gardner and

Lambert (1972) is not beyond question. Several studies like Khanna and Agnihotri

9

(1982) and Khanna (1983) have provided substantial empirical evidence against

Gardner and Lambert’s theory. Khanna (1983) found that achievement in English

was influenced more by schooling, claimed control of English, exposure to

English, use of English among friends, family, etc. than by attitudes and

motivation. Attitudes were found to have significant correlations only with

variable of exposure and language use suggesting thereby that attitudinal variables

have only indirect bearing on achievement in English in India. Similar result was

reported in Khanna and Agnihotri (1982). Sahgal (1983) working with educated

Indians found that the prestigious variants of different phonological variables

correlated far more strongly with schooling, socio-economic status than with such

social psychological variables as integrative motivation, linguistic stereotypes and

ethnocentrism. According to them, the people of India learn English for a variety

of reasons but all of them have an unmistakable instrumental colour: some are

integratively instrumental, some instrumentally instrumental, some manipulatively

instrumental and some instrumental despite resentment. They conclude that the

different types of motivational orientation show a high degree of overlap. It is

clear from their study that there is a substantial overlap between different types of

motivational orientation and that different types of informants are characterized by

different elements of motivations than exclusively by a given individual

motivational orientation. It is possible that the same informant may be influenced

equally by instrumental or resentment motivation. The main point which Khanna

and Agnihotri attempt to make is that the fundamental reasons for learning English

in India are instrumental rather than integrative if we regard manipulative and

resentment motivation as essentially instrumental in nature. Some work has

already been initiated in the attitude towards English in Bangladesh. In this

connection, the work of Shahed (2001) can be mentioned. He has concentrated for

his data in some urban areas only leaving the vast countryside out of his view. In

this context, it seems that his study is a restricted one based solely on a

questionnaire. His work reveals a largely negative attitude towards English though

10

its instrumental function is also well recognized. Since changes due to

globalization are occurring at a rapid pace in Bangladesh, it is felt that the attitudes

towards English in Bangladesh may have become more positive. In the present

work, an attempt will be made to study the attitudinal and motivational variables

in the context of English in Bangladesh. The model proposed by Gardner and

Lambert will be employed and the findings will be compared with those of

Agnihotri and Khanna in India and those of Shahed in Bangladesh.

English is taught as a compulsory foreign language in Bangladesh right

through school up to the end of pre-university level. The need for the knowledge

of English becomes apparent among Bangladeshis as it can give access to the

advancing knowledge of the world. Apart from this, there is need for many to

interact academically and socially for purposes of politics, economics, etc. with

people beyond the national borders. As competition in the employment sector

increases the urge of students to learn English to give them an edge over others

understandably increases too. The need becomes most real in the context of the job

market. At this point, many employers of organizations with international contacts

often see proficiency in spoken and written English as a desired requirement.

Thus, proficiency in English has become more sought-after for Bangladeshis.

However, a large number of people feel that Bangla should be given more

prominence over English at all levels of education, administration and other

sectors. The people of Bangladesh have fought and laid down their lives for their

mother tongue. This carries a unique chapter in the history of the country. After

the inception of independence in the country, the demand for Bangla at all levels

intensified. As a result, the spread and establishment of Bangla in education,

administration and in other sphere of lives got approved by the government as well

as by the common people of our country. Now, within thirty years of its

independence the growing demand for learning English and dissemination of

English in all other institutions, organizations, administrations, etc. is also coming

to the fore. This whole spectrum can have the potential outlook of highlighting a

11

change in the attitude towards English in Bangladesh. The general observations

which can be seen in Bangladesh are the following. English is not the language of

home. The number of programmes available in English has multiplied manifold

with the advent of cable TV. In the recent years, there has been a greater interest

in western music and plays in English among the young generation, particularly, in

the capital city of Bangladesh. The new youth wants to come out of the restrictions

imposed by their previous generations. The more ‘forward looking’ parents want

some amount of English to be used in teaching at all levels of education and

preferably want to send their children to English medium schools. On the other

hand, we also notice a concern for the greater use of mother tongue at the primary

to higher secondary level of education among academicians, researchers and the

politicians engaged in the race to increase their vote banks. It is often seen that the

children of the politicians who demand mother tongue education at all levels of

instruction are sent to study abroad or in English medium schools. English is

primarily learnt for higher education, career enhancement and to upgrade one’s

prestige and personality. It is seen that most educated people in the capital of

Bangladesh are fully aware of the importance of English in education, science and

technology. The general Bangladeshis are positively oriented towards English-

speaking Bangladeshis. In the context of motivational orientation, the relevance of

resentment motivation can be linked in the context of learning English in

Bangladesh. Not everyone is willing to learn English. For them, it is a compulsion

imposed by the system of education; others learn it because they are forced to do

so by their parents/guardians as a means of survival or for some other equally

compelling reasons. The stereotype which can be seen among educated

Bangladeshis is that scientific knowledge is not complete without learning English

as well as without education through the English medium. It would not be far from

truth to say that the ruling class in Bangladesh wants to continue to have English

in administrations and other important sectors because it is an important tool for

keeping the underprivileged marginalized.

12

Cloze Procedure and Theoretical Background

Cloze procedure was initiated by Ebbinghaus in 1897. The origin of the

word ‘cloze’ goes back to 1953 when Taylor was trying to design appropriate

reading materials for native students. Cloze procedure prominently came into light

when Taylor investigated its effectiveness as a tool for determining the readability

of passages of prose in the reader’s native language. Deleting words from a

selection and requiring the examinee to fill in the blanks constructed the test. The

term ‘cloze’ was used with the notion of Gestalt ‘closure’ in mind, referring to the

natural human psychological tendency to fill in gaps in pattern. He coined the

word as a reminder of the concept of ‘closure’ which was well known in Gestalt

psychology. According to Gestalt psychology, the whole is considered different

from the sum of its parts. The development of Gestalt psychology has influenced

the way the process of language and language learning are described. Cloze

procedure is based on the Gestalt psychology. It refers to the ability of individuals

to complete a pattern once they have grasped the structure of the pattern. As for

example, if one sees a picture of a man’s face without one ear and one eyebrow,

provided he grasps what it is meant to be, he will be able to complete the picture.

Gestalt psychology was developed in Austria and Germany toward the end

of the 19th century. Gestalt is a German word which does not have an exact

translation in English. The word puts stress on the notion that properties of the

whole influence the way in which the parts are perceived; perception acts to draw

the sensory data together into a holistic pattern or Geslaltien. For that reason, the

theory is sometimes referred as the theory of the ‘whole’. The proponents of this

theory refute the idea that percepts are built up from distinct atoms of sensation.

The theory leads to the conclusion that human beings possess a natural ability to

perceive unfinished or incomplete Figures as complete entities. A typical example

of visual perception is completing an incomplete geometrical pattern–a pattern

with a closure–such as the ones illustrated below:

13

Thus, if a person has developed a perception of an object, the object will be

perceived completely even if it contains some incomplete parts or closures. The

same approach, applied to language processing, claims that if one has developed

an ability to use language, he will be able to complete his perception of a piece of

a language even if the piece of language contains some closures or blanks. Taylor

(1953) suggested that the closure procedure could be treated as an alternative to

traditional reading comprehension texts. He believed that when certain closures,

i.e., blanks were made in a passage, it would result in an incomplete whole. The

words around the blank would function as the stimuli to activate reader’s mental

mechanism. Thus, the reader would try to find clues from the passage that might

help him trace the missing words. In this case, the reader’s ability to complete the

incomplete passage would be an indication of his reading comprehension ability in

general and that shows, in particular, his language processing ability. In more

technical terms, Cloze procedure is a means of assessing learner’s internalized

grammatical knowledge. In fact, the Cloze procedure indicates the efficiency of

the learner’s utilization of grammatical processes to retrieve the missing or

mutilated parts of the text. Cloze tests are deceptively simple devices that have

been constructed in so many ways for so many purposes that an overview of the

entire scope of literature on the subject is challenging.

Various research projects were conducted to investigate the efficiency of

Cloze procedure as text for non-native learners’ language proficiency. The

possibility of using Cloze procedure for testing foreign–language proficiency has

14

been known since a study by Carroll, Caron and Wilds (1959) employed the Cloze

method with high school foreign–language students. In spite of the utility of their

study, the Cloze procedure is not even mentioned in many standard test books on

language testing (Lado 1961) nor is it discussed in the most widely used language-

teaching manuals (Lado 1964). The definition of Cloze procedure is provided by

Taylor as ‘any passage of appropriate length and difficulty with every nth word

deleted.’ There are several terms of the definition need to be clarified. The first

line of research focuses on determining the value of ‘n’. It is clear that the

minimum value that ‘n’ can take is ‘2’. That is, every other word should be

deleted. Though the minimum is easy to determine, it is not so with the maximum.

To determine the most appropriate value of ‘n’, Cloze tests with every 3rd, 4th,

5th, 6th, 7th, 8th, 9th and 10th word deleted are developed. In practice, tests with

every 5th or 7th word deletion have become more popular. More specially, Cloze

tests with every 7th word deleted tests considered as standard Cloze tests though

they may not possess significant advantages over other types of Cloze with

different rates. MacGinite (1961) has found that words are equally restorable

under sixth, twelfth and twenty-fourth deletion systems but when every third word

is deleted, restoration is more difficult. Fillenbaum, Jones and Rapport (1963)

have shown that deleting words more frequently e.g. one out of five creates a test

of such difficulty that much discriminatory power is lost. On the other hand, to

delete words less frequently than one out of twelve does not substantially change

the quality of the test. We can see that a very low value of ‘n’ e.g. 3 or 4 will make

the test extremely difficult and a very high value e.g. 10 or 11 will make it very

easy.

The second line of research centres around determining the appropriate

number of deletions ranging from 20 up to 100. Since the number of deletion

determines the length of the passages, generally, more care is taken in this regard.

Following the tradition of developing reading comprehension tests, most scholars

believe that the length of the passage should not exceed 300-400 words. These

15

numbers of words would accommodate approximate 40-50 deletions. The number

of items should be determined at a level which would produce both high reliability

and high validity. Experiments show that Cloze tests with 25-30 deleted items

have as high a validity index as those with 40 to 50 deleted items. Thus, the

common sense criterion for the number of deletions in a Cloze test is set to be 25-

30. It means that considering the 7th word deletion rate and 25 to 30 blanks, the

length of the Cloze passages would be somewhere between 175 to 210 words.

The most common form of a paragraph is the so-called ‘I’ shaped one. That

is, the topic sentence which carries the heaviest load of information, usually

appears at the beginning of the paragraph and the concluding sentence which

summarizes the context of the paragraph, comes at the end of the paragraph. In

other words, these two sentences provide a great amount of information which

may have a key role in understanding the total passage. Therefore, it is felt that

leaving the first and the last sentences intact in the Cloze passage will help readers

to comprehend the passage easily. It eventually becomes an accepted norm to

leave the first and the last sentences intact. In fact, later research also supports this

hypothesis. Adding the number of words in these two sentences would lengthen

the Cloze passage to 220-250 words.

Literature on judging the passage difficulty is old and rich. Within the last

few decades many readability formulas have been developed. Readability

formulas, though variant on the surface, have some common characteristics. Most

of them utilize factors, such as, number of syllables or words, length of sentences

or the syntactic complexity of sentences. Research on foreign or second language

indicates that a Cloze score of 53% or above corresponds to ‘what has traditionally

been called an independent level of reading.’ A score of 44% to 53% is in the

appropriate range for instrumental materials, the so-called ‘instrumental level’ and

the score below 44% falls into the ‘frustration level of reading.’ Oller (1972). The

same phenomenon has been also explicitly supported by Haskell (1976). In using

Cloze test procedure as a readability measure for foreign language texts, it is

16

mentioned that if the mean score of a group of foreign students is over 53% correct

on a passage, then they will have the ‘independent’ level which means that the

passage is easy enough for the students and they can read and comprehend the

passage without getting help from a teacher. If the mean score of the students is

between 44% and 53%, then they are reading at the ‘instructional’ level, i.e., the

students will require help from a teacher. If their mean score is less than 44%, then

they will have the ‘frustration’ level, that means, the passage is too difficult for the

students even with the help of a teacher. Other researchers, like Anderson (1976)

suggested that the actual percentage might vary a little.

Considering the complexion and controversies related to readability

formulas, and also taking account of the differences between a native and non-

native speakers of a language, it is difficult to advocate strongly a particular

formula or criterion for selecting a Cloze passage for non-native speakers. For

selecting a passage for non-native speakers, it is suggested first to determine the

content of the materials studied or expected to be studied by the target population

for whom the test is intended to be given. Then, a random sample of passages from

the content should be selected. A particular readability formula should then be

applied to these passages and the average index should be taken as the criterion for

selecting an unseen passage for the group. Among the methodological questions

concerning Cloze tests which have been investigated systematically with non-

native speakers are scoring methods, difficulty levels, grammatical categories of

deletions and the performance of native and non-native speakers on the same texts.

Taking the difficulty level of a passage into consideration, the definition of the

term ‘cloze’ can be redefined as a passage of appropriate difficulty (determined by

readability formulas) and of appropriate length (220-250 words) with every

seventh word deleted. This definition though not a standard one nor

experimentally superior to other definitions can serve as a reasonable working

definition for those who are interested in working with Cloze procedure.

17

In comparison with other earlier tests that generally measured discrete skills,

Cloze test is considered to be an integrative test. After the investigation of Taylor

(1953), a number of studies have been conducted to investigate the

appropriateness of the use of Cloze procedure as a measure of readability of L1

and L2 materials (e.g., Knight 1966, Bormuth 1962, Brual 1962, Anderson 1967,

McLeod 1962, Miller and Coleman 1967 and Keer 1968) measure of reading

comprehension in L1 and L2 (e.g., Bormuth 1969, Weaver and Kingston 1963,

Tuinman, Blaton and Gray 1975 and Alderson 1978) and the measure of overall

L2 proficiency (e.g., Carroll, Carton and wilds 1959, Conrad 1970; Darnell 1970,

Oller and Inal 1971, Oller 1972, Pike 1973, Stubbs and Tucker 1974, Jonz 1976,

Alderson 1979a, Hinofotis 1980, Brown 1980, 1983; Bachman 1985,etc.). Cloze

procedure involves a cognitive task ‘based on understanding and reasoning’ (Rye

1982). It is termed as a ‘constructive language process’ by Ryan and Semmel

(1969). Rye later on pointed out that a more appropriate term for it would be

‘Construction Procedure’ as the subjects were asked to build up or construct the

deleted words with the help of grammatical, syntactical, semantic and stylistic

information.

There can be two bases on which the deletions are generally made. The

deletions are either on the basis of ‘fixed-ratio deletion cloze’ where every nth

word (5th to 10th word) is deleted and replaced by a blank of standard length or

‘variable-ratio’/‘rational’ deletion where specific items, such as, articles, nouns,

prepositions, verbs, etc. are deleted. The latter suits grammatical testing/teaching

.The scoring is generally worked out either on the basis of ‘exact-word’

replacement requiring retrieval of the original word or on the basis of ‘acceptable-

substitute’ that allows acceptance of any alternative word that is contextually

appropriate. Bachman (1985) made two tests, in one of them different types of

deletions were made according to the range of context required for closure, while

in the other a fixed-ratio deletion procedure was followed. Bachman found that

while both tests (rational deletion procedure and fixed-ratio deletion procedure)

18

were equally reliable and had equal criterion validity, the fixed-ratio test was

significantly more difficult.

Oller (1973) has termed the Cloze procedure as a ‘stroke of raw genius’.

The Cloze test is easy to construct, score and also it shows very significant

correlations with other measures of language proficiency. Taylor (1953) attempted

to show that the Cloze test procedure ranked passages as much as the two most

prominent readability formulae, Dale-Chall and Flesch. Cloze researchers have

focused on its utility as measure of overall second language proficiency (e.g.,

Carroll, Carton and Wilds 1959, Oller 1972, Stubbs and Tucker 1974, Mullen

1979, Khanna 1983 and Sahgal 1992). Not only for measuring readability and

assessing language proficiency but also for educational, medicinal, social,

translation and other purposes Cloze tests have played significant roles.

Fillenbaum and Jones (1962) were successful in making distinction between

transcripts of the speech of aphasic patients and control speakers. Klare, Sinaiko

and Stolurow (1972) showed that Cloze procedure possesses valid technique for

evaluating the quality of translations of technical training manuals used by the

military.

One of the major issues related to Cloze tests ‘what does it test/measure?’

Oller (1973) refers to the measurement of Cloze test as ‘grammar of expectancy’.

He feels that the underlying process of taking a Cloze test shows more than

‘passive reading’. On the other hand, Alderson (1983) and Foley (1983) are

sceptical of the validity of Cloze procedure as a measure of overall language

proficiency. Alderson (1983) reports:

‘The Cloze procedure is not a unitary technique, since it results in tests

which are markedly different, different tests give unpredictably different

measures’. (Alderson 1983:211)

Researchers have tried to explore the influence of different cognitive abilities and

styles on Cloze test performance. There are many studies which show contrary

19

views about the existence of the influence of different cognitive abilities and styles

on Cloze test performance. Bialystok and Howard (1979) investigated the

influence of inferencing on Cloze test performance. They have defined inferencing

as the ability to exploit maximally all the available information sources in order to

arrive at new insights into the unknown aspects of the second language. They

found that inferencing is an integral component in performance on Cloze tests. A

lot of research has also been conducted to explore the influence of field

dependent/independent cognitive style (FD/I) on Cloze test performance. FD/I

refers to individual differences in performed ways of perceiving, organizing,

analyzing or recalling information and experience. Field dependence indicates a

tendency to rely on external frames of reference in cognitive activities whereas

filed independence suggest reliance on internal rules and strategies for processing

information and the existence of mental restructuring abilities. The results of the

studies exploring the relationship between FD/I and Cloze test performance are far

from consistent. Hanseen (1984) came to the conclusion that the relationship

between field sensitivity and Cloze test performance might not be very significant

for all cultures.

The next issue related to Cloze test is whether Cloze items are sensitive to

cross-sentential ties. Many researchers, such as, Carroll (1972), Alderson (1983),

Porter (1983) and Markham (1985) found that Cloze items were not sensitive to

long-range constraints. Carroll (1972) emphasized that Cloze test is based on

working with the ‘local redundancy’ of a passage i.e., the missing elements can be

supplied by the linguistic clues present in the immediate environment-generally in

the same sentence. Alderson (1983:216) writes in this regard:

‘My research, and particularly that reported in the Journal of Research in

Reading in 1979, shows that increasing the amount of context on either side of a

Cloze gap has no effect on the case on average. No increase in predictability is

gained by a bilateral context of eleven words rather than five words. If amount of

context has any effect, the critical amount is less than five words’.

20

Porter (1983) also reached the same conclusion. He writes:

‘There was no indication that quantity of context beyond five or six words

bilaterally is a factor affecting predictability’.

On the other hand, there are many researchers, such as, Ramananskas

(1972), Oller (1975), Chihara, Oller, Weaver and Chavez-Oller (1977), Cziko

(1978), Brown (1983) and Jonz (1990) are of the opinion that Cloze items are

sensitive to long-range constraints. Ramananskas (1972) reached the conclusion

that at least some Cloze items are sensitive to cross-sentential constraints and also

said that the affect of such constraints tend to increase as the examinees become

more proficient in the language concerned. Chihara, Oller, Weaver and Chavez-

Oller (1977) found that both native and non-native speakers of English performed

much better on Cloze items in intact texts than did on the same items in the

scrambled texts. Other studies, such as, Shanahan, Kamil and Tobin (1982),

Alderson (1979) and Porter (1978) have shown that Cloze texts are not sensitive to

contexts beyond the sentence. Alderson (1979) found no evidence that increases in

context affected his subjects’ ability to complete items successfully. The current

study of Chavez-Oller, Chihara, Weaver and Oller, Jr. (1985) concluded that some

Cloze items are sensitive to constraints that reach even beyond 50 words on either

side of a blank. Their research also showed that this sort of context sensitivity was

inaccessible to rank beginners but became increasingly accessible to more

proficient language users. With the reviewing of the experimental data from

Chihara et al. (1977), they found about 10% of the items in each of the two texts

examined were highly sensitive to constraints ranging across sentence boundaries.

Researchers have also come up with alternative suggestions of modifying the basic

Cloze test so that it becomes a better measure of cross-sentential cohesive ties.

Those researchers who gave alternative suggestions are Clarke (1979), Bachman

(1982), Bensoussan and Ramraz (1984), Deyes (1984) and Mauranen (1987).

Clarke (1979) and Bachman (1982) suggest that content words tend to reflect

grammatical process at the sentential level. According to these researchers a

21

content-word deletion procedure is superior to every nth word procedure since it is

capable of measuring both syntactic and discourse level relationship in a text.

Bensoussan and Ramraz (1984) worked with a multiple-choice rational Cloze

which they named as the ‘fill-in-test’ is not random; they are chosen because of

their key function in a rational argument which the candidate has to reconstruct.

They found that this ‘fill-in-test’ is a reliable measure of EFL reading

comprehension. According to Deyes (1984), a more discourse-oriented approach

in Cloze testing is needed. Units based on communicative value are more

appropriate deletion items than single words if the aim is to test the comprehension

of a test.

The other important issue related to Cloze procedure is the frequency of

word deletion of a Cloze passage. Words of a Cloze test are generally deleted on

the basis of either rationally or semi-randomly or randomly. The rational Cloze

deletion process includes selection of words for deletion on the basis of linguistics

principle. For example, it can be only nouns or verbs deletion. Most researchers

have preferred to use the more simple every nth deletion procedure. A number of

these have employed an every 5th deletion system. A number of studies have been

conducted to investigate the effect of the rate of deletion on Cloze test

performance and to answer the question as how much context should be left

between Cloze items. MacGinite (1961) compared the effect of deleting every

third, sixth, twelfth and twenty-fourth word. He took into account the fact that

different deletion rates took out different sets of words. To control this factor, only

results on blanks common to all deletion rates were analyzed. Since different

deletion rates were multiple of 3, the possibility of having the same word deleted

in different Cloze tests was fairly high. MacGinite found that words were equally

restorable under sixth, twelfth and twenty-fourth deletion systems but restoration

was more difficult when every third word was deleted. He concluded that context

more than about five words distant exercised relatively little constraint. Anderson

(1976) points out MacGinite employed passages of text that would have been very

22

easy reading for the sample he took and that the effect of employing too easy

reading materials for subjects is to run the risk of masking differences which might

exist. Fillenbaum, Jones and Rapport (1963) compared the effect of deleting every

second, third, fourth, fifth and sixth word and found that Cloze scores increased

moderately as frequently of deletion decreased from every second to every sixth

word. Most of the researchers, such as, Fillenbaum, Jones and Rapport (1963),

Oller (1973), Potter (1968) and Alderson (1980) have preferred to employ the

every nth word deletion procedure. To Cohen (1980), the rational deletion process

shows more of a discrete point test. The semi-random deletion process involves

deletion of every nth word (N can be any number) neglecting its linguistic

characteristics. Random cloze involves the deletion of a percentage of words using

a statistical random sampling technique.

Another issue related to the Cloze procedure is the scoring procedure. The

most common way of scoring responses is to give credit only for the restoration of

the exact word deleted. However, it is not uncommon to take as correct retrieval of

either synonyms of the deleted words or semantically acceptable words. There are

other alternative scoring procedures, e.g. Darnell (1986). Darnell used a scoring

procedure based on a prior determined distribution of response frequencies to each

item as answered by the native speakers.

Many studies, such as, Stubbs and Tucker (1974), Alderson (1980) and

Khanna (1983) showed that there existed high correlations between the exact word

scoring method and the contextually acceptable word among non-native speakers.

Alderson (1983) concluded that the method which gave credit for any semantically

word appears to be the most valid procedure for the purpose of English as a

Second Language (ESL) testing. Thus, it appears that for native speakers the exact

word scoring procedure is the best and for non-native speakers acceptable word

scoring is the best. Several researchers attempted to modify various aspects of the

basic Cloze test. Instead of the usual written form, Peisach (1965) and Potter

(1968) presented the Cloze passage in auditory form. Ozete (1977) used

23

‘attended–choice’ approach in which the students will have choice to select one

out of two words rather a blank at random intervals. Jonz (1976) presented the

‘multiple-choice’ approach where the students had to choose one out of four

choices. On the other hand, Allen (1968) proposed using fill-in-blank tests instead

of the multiple-choice types. Clarke (1979) and Bachman (1982) suggested that a

rational-deletion Cloze test based on content-word deletion procedure was superior

to every nth word procedure.

The following issue related to Cloze test concerns the nature of the text. It is

well known now that if a student is familiar with the text of a Cloze passage, his

performance rises significantly than others who are not familiar with the text of the

Cloze passage. Darnell (1968) showed that engineering majors outperformed non-

majors on passage that was taken from engineering texts. Many studies have

investigated the kind of relationship that exists between the level of difficulty of

the text and the Cloze scores. Agnihotri and Khanna (1991) found that the difficult

text appeared as more difficult Cloze test rather than Cloze test based on easy text.

The most important issue related to Cloze procedure concerns the

performance of native versus non-native speakers’ differences on Cloze tests.

Alderson (1980) found that if Cloze tests were claimed to be texts of linguistic

proficiency then native speakers would perform invariably well on Cloze tests than

non-native speakers. He found that the effect of the changes in deletion frequency

was also similar for native and non-native speakers. It was shown by Alderson

(1980) that native and non-native speakers perform similarly on Cloze tests and

non-native speakers achieve scores as high as the highest native speaker. Although

native speakers are somewhat better at the task they are still far from perfect and

are not clearly distinguished from non-native speakers. Carroll et al. (1959) found

no differences in mean scores between native speakers and non-native speakers.

On the other hand, Oller, Jr. and Inal (1971) found that there were significantly

differences between the mean scores of native and non-native speakers though

Whiteson (1979) found somewhat different results.

24

The last issue concerning Cloze procedure is about the number of deletions

a Cloze test can have. The recommendation comes from the research literature is

that a Cloze test should have 50 deletions and a minimum length of 250 words

passage with 5th word deletion. To test this recommendation empirically, Rand

(1978) designed an item-analysis programme which artificially lengthens Cloze

tests one item at a time. The programme provided standard statistics for a two-

items, three-items, four-items and so forth, up to a fifty-item Cloze test for four

different scoring procedure (exact- word, acceptable word, Cloze-tropy, multiple

choice). Twenty-five items had achieved the maximum reliability achieved across

scoring methods. Rand concluded that little precision is gained by making a Cloze

test longer than 25 items. Sciarone and Schoorl (1989) tried to determine the

number of blanks minimally required to ensure parallelism for Cloze tests

differing only in the point at which deletion starts. They found that the required

minimum number of blanks for Cloze tests depended on the scoring method used.

25

Methodology

Introduction

This chapter discusses the details of the sample, tools and procedure employed to

examine the relationship between social psychological variables on the one hand

and proficiency in English on the other. The proficiency in English is measured

through the Cloze procedure.

Sample

The sample consisted of students from class eight (30), class ten (30) and twelve

(30) from different institutions in Bangladesh. Figure 1 provides the details of the

sample design.

Figure 1: Sample Design N = 90

class 8 class 10 class 12 (30) (30) (30)

school type

I II III I II III I II III (10) (10) (10) (10) (10) (10) (10) (10) (10)

We had 30 students from each class, that is, a total of 90 students. Three different

kinds of schools (an English medium public school, a government aided public

school and an ordinary school) were selected to take care of the socio-economic

variables. English medium children were taken from a school of Dhaka,

government-public school students from Jessore town and ordinary students from

26

Navaran, Jessore, Bangladesh. We preferred the co-educational schools so that we

could collect data from both male and female students of the same class. Random

sample selection was used in each class to select ten students. The information

about the students and their socio-economic background were obtained from their

admission forms in the institutions. All of them were Bengali speakers though

some of them learn Arabic in their earlier years and were in the age group of 12-19

years. The students were learning English as a second language. It is to be noted

that English in most of the institutions in Bangladesh is taught from class one till

class twelve as a compulsory subject. We can also mention that the learning of

English does not take place in a natural setting as most of the teachers of English

are not native speakers of English and the target community is also absent in our

country. The existing scenario shows that the students of ordinary schools in

Bangladesh generally don’t use English with their family members, among

classmates and friends. On the other hand, those who are from English medium

schools use English as a language of peer-group communication particularly with

the opposite sex. The reason for the greater use of English with the opposite sex

may be to impress them. There has been in recent years a greater interest in

western music and plays in English among the youth in Bangladesh. In spite of all

this, English is not the language of home and family for the students in

Bangladesh.

3.3 Tools

We used the following tools to collect data from our informants.

I. A Cloze test (Appendix A).

II. A detailed social psychological questionnaire (Appendix B).

The tools employed in this study were initially designed and tested in a pilot study.

Broadly speaking, we isolated three sets of measures, proficiency in English,

personal and family information and attitudes and motivations.

27

The Cloze Passage

Proficiency in English was measured through the Cloze test procedure. This

procedure was initially designed and developed by Wilson Taylor (1953) in the

early 1950s to measure the readability of texts in the reader’s native language.

This technique is also found to be very useful for testing overall language

proficiency of native speakers. Recently, it has been used as a measure of

proficiency in the case of non-native speakers in foreign and second language

situation. This section consists of a Cloze test designed to measure the proficiency

levels of the students in English. Following the principles of ‘Cloze procedure’,

we used an extract from Jim Corbett’s ‘Lalajee’ where every sixth word was

deleted after leaving the first and the last sentence unmutilated. It was a very

simple and an interesting passage. It was about 242 words in length. We had a

total of 30 blanks. The blank size was kept uniform throughout the entire passage.

Procedure

Examinees were given a copy of the test containing the instructions which

stresses that only one word was to be used in each blank which seemed the most

appropriate within the context of the passage. Students were advised to read the

passage twice before starting to fill in the blanks. Students were also told that

spelling would not count against them as long as the scorer could read the intended

word. The test administrator read through the complete instructions with the

candidates. He then asked whether there were any questions. The time allowed for

this test was thirty minutes. Students were told to consider the test as a challenge

to their English language ability.

Method of Scoring

The test was scored twice for each candidate: once by the exact word

method whereby only the words which were originally deleted from the text were

considered correct and second by the acceptable scoring method whereby any

word which was considered contextually and grammatically correct was also

28

counted as correct. Each blank when filled correctly was assigned one point;

hence, the score range of the Cloze test was 0-30. There were 30 blanks in the

Cloze test passage. Following ‘exact retrieval’ scoring procedure, every ‘exact’

recall was given a score of 1. Thus, a subject could get a maximum score of 30.

We adopted the following criteria for a contextually-acceptable response (i) any

non-grammatical form (e.g., he say for he says) would be automatically excluded

even though the meaning happened to be exact and (ii) any blank which contained

two words was excluded even though the result may have been acceptable

semantically. All other contextually-or semantically-acceptable possibilities were

accepted.

The Social psychological Questionnaire

This section consists of a detailed social psychological questionnaire

designed for the informants (students) to elicit the following kinds of information:

(a) Personal background, age, sex, socio-economic status, types of education

received

(b) Medium of instruction

(c) Attitude towards English

(d) Attitude towards English-speaking Bangladeshis

(e) Motivational orientation

(f) Parental encouragement to learn English

(g) Classroom anxiety

A total of 25 variables (v/vs) were created to examine the patterns of correlation

obtaining in our data. A complete list of these variables is provided in Appendix

C.

(a) Personal background and family information

Age (v. 1) was quantified as a running variable.

29

Sex (v. 2) was quantified in terms of ‘1’ for males and ‘2’ for females.

Exposure to English (v. 3) was quantified as a running variable ranging from 0 to

14.

Schooling (vs 4-6) was seen in terms of the nature of schooling at two different

levels, primary and secondary (post-primary). The score for total schooling (v. 6),

as may be seen from Question 11 (in Appendix B), could vary from a minimum of

6 to a maximum of 18. It was measured in terms of prestige of the institution,

medium of instruction used there and facilities available for learning English. That

is, the medium of instruction and other facilities were seen as indication of the

type of school that the subject attended. Thus, there are English medium schools

where not only the medium of instruction is English but also other facilities

required for an optimal control over English are also available. Similarly, there are

other schools (mostly run by the government) where the medium of instruction is

Bengali and facilities for learning English are minimal. A high school score is

suggestive of a strong English bias in the total education of the informants.

Residence/Address (v. 7) was quantified on a 5-point scale ranging from 1 (lower

class residential area) to 5 (posh residential area). This information was collected

from the address given by the subjects. The score was calculated as follows

Type of Locality Score

Lower Class 1

Lower Middle Class 2

Middle Class 3

Upper Middle Class 4

Posh 5

Socio-economic status or SES (v. 14):

30

The socio-economic status of the informants was elicited from Question 12

(Appendix B). Socio-economic status was measured in terms of the educational

qualification of the subject’s parents (vs 8-9), their monthly income (vs 10-11) and

occupation (vs 12-13). A modified version of Kuppuswamy’s (1962) SES-scale

was used to quantify father and mother’s education, occupation and income. Total

socio-economic status was seen as an aggregate of the above variables (vs 7-13).

The score for total SES (v. 14) could vary from a minimum of 7 to a maximum of

35.

(i) Educational qualifications:

Education is very important in uplifting a family socially and financially.

The following scale was used for the calculation of scores for educational

qualification.

Educational Qualification Score Ph D 5Professional Degrees or Masters 4Bachelor 310+2 2 Below Class12 1

(ii) Occupation: the following scale was used for the calculation of scores foroccupation.

Occupation Score Professionals 6Teacher/Government Servant 5Businessman/Private Company 4Farmer 3Labourer/Clark 2Housewife 1

(iii) Monthly Income: The scale used for the calculation of scores for monthlyincome is as follows:

31

Income (TK. Per month) Scores Below 2000 12000-3499 23500-4999 35000- and above 4

(b) The desired medium of instruction in different domains of education

The medium of instruction (v. 15): In question 13 (Appendix B), we wanted

to provide what our informants thought should be the medium of instruction in

different level/fields of education, such as, school, college, medicine,

management, engineering, etc. All the responses to question 13 were scored as

follows:

Medium of Instruction ScoreBengali 1Both Bengali and English 2Only English 3

The score for total medium of instruction (v. 15) could vary from a minimum of

12 to a maximum of 36.

(c) Attitudes to English-speaking Bangladeshis

An attitude scale was designed (Question 14: Statements i-iii) to measure

informant’s attitudes towards English-speaking Bangladeshis. There is a

westernized elite group in urban Bangladesh; one of their defining features is the

knowledge of English. A 5-point scale was given after every statement which

ranged from ‘agree’ to ‘disagree’.

There were 2 positively (Statements i and ii) worded and 1 negatively

(Statement iii) worded statement(s). The score for total attitudes to English-

speaking Bangladeshis (v. 15) could vary from a minimum of 3 to a maximum of

32

15. Scoring the positive and the negative statement(s) was done as follows:

Positive Negative Agree 5 1Slightly Agree 4 2Neither Agree nor Disagree 3 3Slightly Disagree 2 4Disagree 1 5

(d) Attitude to English (v. 17)

An attitude scale (Question 14: Statements iv-v and xxv-xxx) was designed

to measure the attitudes of the informants towards English. A 5–point scale

ranging from ‘agree’ to ‘disagree’ was given after each statement. There were only

one negatively (Statement xxx) worded statement and the rest were 7 positively

(Statements iv-v and xxv-xxix) worded. The score for total attitude to English

could vary from a minimum of 8 to a maximum of 40. Scoring for the positive

and the negative statements was done as follows:

Positive Negative Agree 5 1Slightly Agree 4 2Neither Agree nor Disagree 3 3Slightly disagree 2 4Disagree 1 5

(e) Motivation to learn English

All the items in these attitudinal scales were followed by a 5-point scale

ranging from ‘agree’ to ‘disagree’. Items were modified to suit the Bangladeshi

situation. The subjects were asked to circle the alternative which best indicated the

extent to which he/she agreed or disagreed with the statements.

Integrative motivation (v. 18) consists of 4 statements (Question 14: Statements

vi-ix) and the score for total integrative motivation could vary from a minimum of

4 to a maximum of 20. Primarily, it focuses on assimilation with the target

33

language community (here, the personality, traits and social roles of the English-

speaking Bangladeshis).

On the other hand, Instrumental motivation (v. 19) (Question 14: Statements

x-xi and xiv and xv) concerned with learning English for social and occupational

mobility. The score total for instrumental motivation could vary from a minimum

of 4 to a maximum of 20.

Not every one in Bangladesh learns English willingly. For some, it is a

compulsion imposed by the system of education; others learn it because it is a part

of their course requirement. The score for total Resentment motivation (v. 20)

(Question 14: Statements xii-xiii) could vary from a minimum of 2 to a maximum

of 10.

Manipulative motivation (v. 21) has been measured through statements xvi and

xvii in Question 14 and the score for total manipulative motivation could vary

from a minimum of 2 to a maximum of 10. Scoring for the motivation to learn

English was done as follows:

Agree 5

Slightly Agree 4

Neither Agree or Disagree 3

Slightly Disagree 2

Disagree 1

(f) Class-room anxiety (v.22) was elicited through 3 statements (Question 14:

Statements xxii- xxiv) and a 5-point scale ranging from ‘agree’ to ‘disagree’ was

given after each statement. The score for total class-room anxiety (v. 22) could

vary from a minimum of 3 to a maximum of 15. It was meant to be an index of

how comfortable a subject feels in the English class. Scoring was done as follows:

34

Agree 5

Slightly Agree 4

Neither Agree or Disagree 3

Slightly Disagree 2

Disagree 1

(g) Parental support in learning English (v. 23) was measured through 4

statements (Question 14: Statements xviii-xxi) and was quantified as in the other

(above) cases using a 5-point scale. The score for total parental support could vary

from a minimum of 4 to a maximum of 20. The students were asked to rate the

extent to which their parents actively encouraged them to study English.

The remaining two variables deal with the Cloze test; one of them is Cloze

test exact (v. 24) and the other is Cloze test acceptable included (v. 25). The

discussion of the two variables can be seen in the previous section 1 which deals

with Cloze procedure. The questionnaire was administered to the students only

after assuring them that the data to be collected from them would be strictly

confidential and it has nothing to do with their college evaluation. The

questionnaire was provided followed by the Cloze test.

We give below a list of variables we examined:

A. Individual Variables

(a) Biological Variables

(i) Age (v. 1)

(ii) Sex (v. 2)

(b) Attitudinal Variables

(i) Attitude to English (v. 17)

35

(ii) Attitude to English-speaking Bangladeshis (v. 16)

(iii) Motivational orientation (vs 18-21)

(c) Class-room Variables

(i) Class-room anxiety (v. 22)

B. Social Variables

(a) Socio-economic Variables

(i) Residence (v. 7)

(ii) Occupation (vs12-13)

(iii) Income (vs 10-11)

(iv) Education (vs 8-9)

(b) Socio milieu Variables

(i) Parental support (v. 23)

(ii) Medium of instruction (v. 15)

C. Language Proficiency

(a) Cloze test variables

(i) Cloze exact (v. 24)

(ii) Cloze acceptable included (v. 25)

36

Result and Analysis

Data was collected from schools; three types of school were identified for

collecting data; an English medium school, a government-aided school and an

ordinary school in Bangladesh. We preferred the co-educational schools so that we

could get data from both male and female students in the same class. As pointed

out in Chapter 3, we collected data from class eight (30), class ten (30), class

twelve (30) students and tried to make equal numbers of male and female students.

We took 30 students from each class, that is, a total of 90 students.

The average age was 15.33 with a range from 12 to 19. We used random

sample procedure for each class to select ten students. We always sought

permission from the head of each institute for the purpose of collecting data. After

getting the required permission, we first got the questionnaire (Appendix B) filled

in and then administered the Cloze test (Appendix A). In addition to it, we

informed the students about the objective of our study and assured them that this

study had nothing to do with the evaluation of their class-performances. Though

they were told to take the Cloze test as a challenge for their knowledge of English.

The administration of the tests and questionnaire was supervised at all times

by the test administrator. This was to ensure consistency in the collection of the

data. The total time spent on administering the questionnaire and the Cloze test

ranged from one to one-half hour. Allowance was made for five minutes for the

students to rest between the administrations of test and questionnaire. Sometimes,

along with me the class teacher invariably gave the instructions and also guided

the students. The teacher read each item of the questionnaire before it was

answered. This was to make sure that the students understood the questions and

responded appropriately. Due consideration was given to the fact that the medium

of questions was not the mother tongue of the students. The data collected was

codified and tabulated for every student. A large matrix was prepared on a

checkered sheet. The matrix consisted of twenty-five variables on the horizontal

37

axis and ninety informants on the vertical axis. Each student had been assigned a

number.

The analysis was done by using a software package SPSS/WIN. The data

were analyzed for means, standard deviation and correlation of 25 variables. The

total number of students in our study is 90. We did not face any problems in the

process of data collection. There was no question left unanswered by our

informants. They were very cooperative, patient and sincere in filling the

questionnaire (Appendix B). In our following tables, number before the issues

listed in Column I refer to the statements numbers in Question 14 in Appendix B.

Table 4.0: Cloze Retrievals (%) of the Students

Class

School 8 10 12

English Medium (n=30) 33.00 47.66 48.33

Government-aided (n=30) 31.66 34.33 41.66

Ordinary (n=30) 16.00 20.66 29.66

Table 4.0 shows vertical progress of performance among schools. The students of

government-run school have performed better than students of ordinary schools.

The students of English medium school have also performed better than students

of government-run school. For example, the students of class eight of government-

run school have better performance (31.66%) than the students of class eight of

ordinary school (16.00%) and the students of class eight of English medium school

have better performance (33.00%) than the students of class eight of government-

run school. The same can be said for other two classes (class ten and twelve) of

students as well.

We also see horizontal progress of performance within schools. The

students of class ten of any schools (English medium or government-run or

ordinary school) have performed better than students of class eight and class

38

twelve students of any of the three schools have performed better than the students

of class ten. Thus, we can say that the students of class twelve of English medium

school are the best performer in the Cloze test. Whereas the students of class eight

of ordinary school have performed very poorly in the Cloze test. Here, we can

also mention that as the students move up to higher classes, their English

proficiency level also increases.

The interpretation of the variables based on Mean score, Standard Deviation (S .D)

and percentage is shown in the following tables.

Age: The mean age of the sample is 15.33 (S.D 2.02). On an average, the students

involved in this study are found to be half a year above the standard age of their

respective classes. The students of the ordinary schools were perhaps sent late for

primary education by the parents.

Sex: The mean score of sex for the sample is 1.42 (S.D .49). This score shows that

on an average, the numbers of female students are more than male students. The

reason behind this can be the policy of our government which has started granting

many schemes, scholarships from class 6 to 12 to encourage female students to

study.

Exposure to English: English is taught from class one as a subject throughout

Bangladesh. The mean score of exposure to English for the sample is 9.43 (S.D

2.27) on a scale ranging from 0 to 14. The reason for this healthy figure is because

of their learning English from class one onwards.

The following Table 4.0 shows the mean and standard deviation of the variables of

our study.

Table 4.1: Variable, Mean and Standard Deviation

Variable No Variable Mean Standard Deviation1 Age 15.33 2.022 Sex 1.41 .4963 Exposure to

English9.43 2.27

39

4 Schooling Primary 5.14 1.725 Schooling

Secondary5.65 1.49

6 Schooling Total 8.92 3.6214 SES 18.50 4.3615 Medium of

Instruction32.27 2.31

16 Attitude toEnglish-SpeakingBangladeshis

11.46 2.35

17 Attitude to English 31.49 4.7918 Integrative

Motivation16.74 3.21

19 InstrumentalMotivation

17.87 2.40

20 ResentmentMotivation

7.93 2.40

21 ManipulativeMotivation

7.23 2.49

22 Class-roomAnxiety

8.53 3.94

23 Parental Support 16.47 3.8424 Coze Test Exact 10.11 3.9425 Cloze Test

AcceptableIncluded

16.73 5.65

Schooling: The mean scores of primary schooling, secondary schooling and total

schooling are 5.14 (S.D 1.72), 5.65 (S.D 1.49) and 8.92 (S.D 3.62) respectively.

The score for primary schooling (v. 4) and secondary schooling (v. 5) could vary

from a minimum of 3 to a maximum of 9 whereas the score for total schooling (v.

6) ranges from 6 to 18. The mean of total schooling shows that the students are

from quite prestigious schools and also there are some facilities for learning

English.

Socio-Economic Status (SES): The mean SES of the sample is 18.50 (S.D 4.36) on

a scale ranging from 7 to 35. The mean score suggests that most of the students

belong to middle class and lower middle class.

40

Medium of instruction: The mean score of medium of instruction for the sample is

32.27 (S.D 2.31) on a scale ranging from 12 to 36. The mean score indicates that

most of the students want greater amount of English to be used in teaching at all

levels of education. The figures tend to suggest that there should be very little use

of English at the primary level and more in professional levels of education. It is

also noticed that if a student prefers English medium at one level of education, he

prefers English medium for other levels of education as well.

Cloze test

The mean score exact, exact and acceptable included of the Cloze test are 10.11

(S.D 3.94) and 16.73 (S.D 5.65) respectively. The score range of the Cloze test

was 0 to 30. The mean scores reflect the poor performance of the students. The

nature of the text coupled with the below standard of our education as well as the

lack of qualified English teachers can be reasons for poor performance of the

students.

Attitude towards English: The mean attitude towards English of the sample is

31.94 (S.D 4.79) on a scale ranging from 8 to 40. On an average, the attitude

towards English of the students is on higher side. The mean score suggests that the

level of attitude towards English of the students falls between slight positive (high)

to total positive orientation. The attitude towards English measured through it’s

suitability for creative writing (see following Table 4.2) is not totally positive

unlike scientific thinking. However, it is considered overwhelmingly (67.7%)

suitable for scientific thinking. Most students agree that the present system of

education has benefited enormously (73.3%) from English and greater efforts

(81.1%) should be made to spread English throughout Bangladesh.

41

Table 4.2: Distribution of the Sample on the Scale Measuring Attitude Towards

English

Agree Agree Slightly

Neither agreenor disagree

Slightlydisagree

DisagreeIssueselicitingattitudetowards English

Noofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

4.Suitableforcreativewriting

23 25.5 24 26.6 9 10.0 9 10.0 25 27.7

5.Suitableforscientificthinking

61 67.7 15 16.6 1 1.1 5 5.5 9 10.0

25.Goodinfluence on thesystem ofeducation

66 73.3 18 20.0 3 3.3 0 0 3 3.3

26.Moreeffort tospreadEnglish

73 81.1 10 11.1 5 5.5 1 1.1 1 1.1

27. Essentialfor aleader

32 35.5 16 17.7 13 14.4 11 12.2 19 21.1

28.Goodforscienceandtechnology

67 74.4 14 15.5 4 4.4 0 0 5 5.5

29.Important forpoliticalunity

22 24.4 16 17.7 15 16.6 5 5.5 32 35.5

30AbolitionofEnglish

5 5.5 5 5.5 7 7.7 1 1.1 72 80.0

The attitude towards English, measured through it’s requirement to become

a leader in Bangladeshi society, is positive. However, about 36 per cent informants

agree that knowledge of English is necessary to become a leader in Bangladesh.

Nearly 75 per cent of the informants agree that progress in science and technology

will be hampered without English. It is expected in the context of Bangladesh that

English will not play an important role in bringing about political unity. Nearly 36

42

per cent of the informants disagree that English is important for achieving political

unity in Bangladesh. The reason behind this goes back to history. The

Bangladeshis have fought for their mother-tongue against the imposition of Urdu

by then West Pakistan government. They have more positive attitude towards

Bengali than any other language. In the present context of Bangladesh, the

pressing need for English in the first globalising society shows that English has

come to stay and will become part of our society. Almost 73 per cent of our

informants disagree that English should be abolished from Bangladesh.

Attitude towards English-speaking Bangladeshis: From Table 4.3, we can see that

most of the informants are positively oriented towards English-speaking

Bangladeshis. The mean score of attitude towards English-speaking Bangladeshis

for the sample is 11.46 (S.D 2.35) on a scale ranging from 3 to 15. The mean score

suggests that most of the students have slight positive attitude (high) orientation

towards English-speaking Bangladeshis.

Table 4.3: Distribution of the Sample on the Scale Measuring Attitude Towards

English-Speaking Bangladeshis.

Agree Agree Slightly

Neither agree nordisagree

Slightly disagree DisagreeIssueselicitingattitudetowards English-speakingBangladeshis

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage No ofpersons

Percentage

1.Warm-hearted

20 22.2 36 40.0 13 14.4 13 14.4 8 8.8

2.Progressive

38 42.2 32 35.5 11 12.2 5 5.5 4 4.4

3.Dishonest

5 5.5 5 5.5 22 24.4 5 5.5 53 58.8

About 42 per cent of the informants agree that the English-speaking Bangladeshis

are progressive. Nearly 59 per cent of our students disagree that the English-

speaking Bangladeshis are generally dishonest. A smaller percentage (22.2%) of

them agree that they are warm-hearted also.

43

Class-room Anxiety: Table 4.4 shows that nearly 47 per cent informants agree that

the other students in their class speak better than them. This shows that they do not

lack confidence in their use of English totally. Nearly 44 per cent of them disagree

that they get nervous and confused when they speak English in their class. On the

other hand, 56 per cent of them agree that they get confused and nervous when

they speak in English in their class. However, interestingly a noticeable percentage

(65%) of them disagree to have a fear of being laughed at when they speak in

English. Whereas about 35 per cent, that is, one third of the students agree to have

fear of being laughed at when they speak in the class. The overall picture one gets

from this spectrum of responses is that most of the students experience class-room

anxiety in a mixed way.

4.4: Distribution of the Sample on the Scale Measuring Attitude Towards Class-

room Anxiety

Agree Agree Slightly

Neither agree nordisagree

Slightly disagree DisagreeIssueselicitingclass-room Anxiety

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

22.OtherstudentsspeakEnglishbetter

22 24.4 20 22.2 12 13.3 12 13.3 24 26.6

23. Getnervousandconfusedwhilespeaking

18 20.0 32 35.5 2 2.2 5 5.5 33 36.6

24.Fear ofbeinglaughed atwhilespeakingEnglish

17 18.8 14 15.5 8 8.8 4 4.4 47 52.2

The mean score of class-room anxiety of the sample is 8.53 (S.D 3.94) on a

scale ranging from 3 to 15. The mean score suggests that the level of anxiety is not

alarmingly high. The reason can be that most of the students claim themselves as

good students in their classes.

44

Parental Support: The mean score of parental support of the sample is 16.47 (S.D

3.85) on a scale ranging from 4 to 20. The mean score suggests that most of the

students have high degree of parental support when it comes to learning English. It

is evident from Table 4.5 that more than 64 per cent informants agree that they

have parental help for their English. Nearly 64 per cent of them agree that their

parents encouraged them to seek help from their English teachers whereas about

61 per cent of them agree that their parents emphasize the importance of English

after leaving school/college. However, about 53 per cent of them agree that their

parents have advised them to spend more time on their English studies. Most

parents understand that the career prospects of their children depend on subjects,

such as, physics, chemistry and mathematics. Therefore, they do not exaggerate

the importance of English as a subject.

Table 4.5: Distribution of the Sample on the Scale Measuring Parental Support

Agree Agree Slightly

Neither agree nordisagree

Slightly disagree DisagreeIssueselicitingparentalsupport

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

18.Help inEnglish

58 64.4 14 15.5 5 5.5 2 2.2 11 12.2

19. More timeto Englishstudies

48 53.3 23 25.5 7 7.7 6 6.6 6 6.6

20.Importance of English

55 63.3 20 22.2 12 13.3 12 13.3 24 26.6

21.Seek helpfrom Englishteacher

57 63.3 20 22.2 12 13.3 12 13.3 24 26.6

45

Table 4.6: Distribution of the Sample on the Scale Measuring Integrative

Motivation

Agree Agree Slightly

Neither agree nordisagree

Slightly disagree DisagreeIssueselicitingintegrativemotivation

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

6.Importantfor interactingEnglish-speakingBangladeshi

53 58.8 20 22.2 4 4.4 3 3.3 10 11.1

7.Importantfor interactingwith variedpeople

77 85.5 7 7.7 1 1.1 3 3.3 2 2.2

8.Helps toappreciateEnglish artand literature

60 66.6 17 18.8 8 8.8 0 0 5 5.5

9.Helps tolive andbehave likeEnglish-speakingBangladeshi

37 41.1 22 24.4 12 13.3 2 2.3 18 20.0

Integrative Motivation: We see from Table 4.6 that out of the four integrative

reasons for learning English in Bangladesh, the largest number (85.5%) of

informants agree to have the desire to ‘interact with more and varied people in the

world’. Nearly 59 per cent of the informants agree that English learning would be

important for them to interact with English-speaking Bangladeshis. About 76 per

cent of them agree that it would help them to appreciate English art and literature.

However, a lesser number of informants (41.1%) agree that English is important

for them because ‘it will help them live and behave like English-speaking

Bangladeshis’. It is obvious from the responses of our informants that most of the

informants have the desire to learn English to interact with people, not necessarily

to assimilate or imitate the way of life of English-speaking Bangladeshis. The

mean score of integrative motivation of the sample is 16.74 (S.D 3.21) on a scale

ranging from 4 to 20. Compared with instrumental motivation, the mean score

46

shows lower level of orientation in integrative motivation. It looks like the

students wish to learn English without surrendering their unique identity as many

students desire to learn English for appreciating arts and literature which are

accessible through the knowledge of English. Very few students feel to imitate the

way of life of English-speaking Bangladeshis

Instrumental Motivation: Most informants, that is nearly 86 out of 90 agree (see

Table 4.7) that English is important for them because ‘it would help them to build

a career’.

Table 4.7: Distribution of the Sample on the Scale Measuring Instrumental

Motivation

Agree Agree Slightly

Neither agree nordisagree

Slightly disagree DisagreeIssueselicitinginstrumentalmotivation

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

10.Helps tobuild career

77 85.5 9 10.0 2 2.2 0 0 2 2.2

11.Useful forhighereducation

70 77.7 14 15.5 4 4.4 1 1.1 1 1.1

14.Makes apersonknowledgeable

68 75.5 10 11.1 5 5.5 2 2.2 4 4.4

15.Adds toprestige andpersonality

48 53.3 19 21.1 6 6.6 5 5.5 15 16.6

Nearly 78 per cent of them agree that English would be useful for them in higher

education in Bangladesh. On the other hand, about 76 per cent of the students

agree that English-learning would make them more knowledgeable. More than

half of them (53.3%) agree that English would add to their prestige and

personality. It looks like that the students wish to use English largely for

instrumental reasons. It is very interesting to note that English is perceived not

only as an academic and economic mobility but also as a means for enhancing

social mobility and individual personality. The mean instrumental motivation of

the sample is 17.87 (S.D 2.40) on a scale ranging from 4 to 20. The mean score

suggests that most of the students are fully aware of the significance of learning

47

English for their future career. They realize well that most white-collar jobs and

social mobility require the knowledge of English.

Resentment motivation: Nearly 63 per cent of the informants agree that they learn

English because it is a part of their course requirement whereas about 52 per cent

of them agree that they learn English as they are forced by their parents/guardians

to learn it.

Table 4.8: Distribution of the Sample on the Scale Measuring Resentment

Motivation

Agree Agree Slightly

Neither agree nordisagree

Slightly disagree DisagreeIssueselicitingresentmentmotivation

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

12.A part ofcourserequirement

56 62.2 20 22.2 0 0 4 4.4 10 11.1

13. Parentswant it to belearnt

47 52.2 8 8.8 15 16.6 2 2.2 18 20.0

The mean score of resentment motivation of the sample is 7.93 (S.D 2.40) on a

scale ranging from 2 to 10. The mean score suggests that most of the students

learn English as more of compulsion than anything else.

Manipulative motivation: Nearly 70 per cent of the informants (see Table 4.9)

agree that English will enable them to get their work done everywhere in

Bangladesh.

Table 4.9: Distribution of the Sample on the Scale Measuring Manipulative

Motivation

Agree Agree Slightly

Neither agree nordisagree

Slightly disagree DisagreeIssueselicitingmanipulativemotivation

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

No ofpersons

Percentage

16.Helps toget work doneeverywhere

45 50.0 19 21.1 6 6.6 7 7.7 13 14.4

17.Helps touse otherpeople forpersonal gains

26 28.8 28 31.1 12 13.3 5 5.5 19 21.1

48

On the other hand, 20 per cent of them disagree on this issue. About 50 per cent of

the students agree that English would help them to use other people for their

personal gains. It shows that manipulative motivation is instrumental in nature.

The mean score of manipulative motivation of the sample is 7.23 (S.D 2.49) on a

scale ranging from 2 to 10. This is not as high as of resentment motivation. The

mean score suggests that most of the students wish to learn English for

instrumental and not manipulative reasons.

Analysis of the Cloze Test

Following Haskell’s (1976) level of judging a Cloze score, we have divided

the sample into three groups in order to measure the overall global proficiency of

the students. This classification is aimed at determining the proficiency level of

second language learners and the difficulty level of the test the students are faced

with.

Table 4.10: Sample Distribution According to Haskell’s Criteria (N=90)

Mean Score Number of Students

Independent Level Over 54 5

Instruction Level 44-53 14

Frustration Level Below 43 71

A close look at the above Table 4.10 tells us that most of the students i.e. 71 out of

90 i.e.78 per cent of the students are at the frustration level. This indicated that the

passage is difficult for majority of the students and even extra help from the

English teacher would have not been useful for them. The average mean score of

Cloze Exact (see Table 4.1: Variables, Mean and Standard Deviation) is only

10.11 which suggests that the majority of the students are not proficient in English.

The poor performance of the students clearly shows that a few students are able to

49

acquire academic/cognitive level of proficiency in English in Bangladesh.

Fourteen students, that is, about 15 percent students are at the instruction level

which means that they are able to read and comprehend the passage and their

performance would have gone up with the help of the English teacher. Only five

students, i.e., nearly 6 per cent of the students are at the independent level which

indicates that they have acquired full grasp over English and have

academic/cognitive proficiency in English. It also shows that without the help

from the English teacher, they are quite capable of comprehending the test. The

average number of correct answers for the Cloze exact scoring is 10.11, with

Cloze acceptable included it rises to 16.73. The mean score is well over 50% now.

In second language learning situations, acceptable Cloze entries should be

considered seriously.

Classifications of the Deleted Words

We have classified the deleted words of the Cloze test (Appendix A) into

content words (Nouns, Verbs, Adjectives and Adverbs) and function/grammatical

words (Articles, Prepositions, Pronouns, Conjunctions, Determiner and Adverbial

particles). This kind of classification helps us to determine which category or

categories are easier or more difficult for students to retrieve. Table 4.11 shows the

exact number of words deleted in each category. Table 4.12 shows the percentage

of retrieval of function and content words in the Cloze test in terms of:

i. Exact retrievals

ii. Semantically and grammatically acceptable (SGA) retrievals

50

Table 4.11: Classification of Deleted Words

Parts of Speech Number of Deletion from Cloze Test

1 Nouns 7

2 Verbs (Main Verb) 2

3 Adjective 1

4 Adverb 1

Content Words

Total Content Words 11

1 Article 1

2 Prepositions 6

3 Pronouns 5

4 Conjunctions 3

5 Determiners 3

6 Adverbial 1

Function

Words/

Grammatical

Words

Total Function Words 19

Total Number of

Deletions

30

51

Table 4.12: Percentage of Retrieval of Content and Function Words (N=90)

Exact (%) Acceptable Included (%)

Content Words 24.94 50.80

Function Words 38.88 58.83

From the above Table, we can say that the students have found the exact function

words easier to retrieve than the content words of the Cloze test.

Table 4.13a: An Overview of Retrievals in Different Categories (Exact and

Acceptable Included)

Open Class Items Exact (%) Acceptable Included (%)

1.Nouns 18.73 50.15

2. Verbs (Main) 34.44 49.44

3. Adjectives 28.88 50.00

4. Adverbs 45.55 58.88

Mean Score 22.45 45.72

Tables 4.13a and 4.13b present an overview of the percentage scores of the

retrievals in different grammatical categories by the 90 students.

52

Table 4.13b: An Overview of Retrievals in Different Categories (Exact and

Acceptable Included)

Close Class Items Exact (%) Acceptable Included (%)

1. Articles 67.77 75.55

2. Prepositions 44.44 64.88

3. Pronouns 48.44 60.44

4. Conjunctions 20.00 45.92

5. Determiner 19.25 56.29

6.Adverbial 44.44 44.44

Mean Score 35.00 52.92

The mean score of the function words is higher than the mean score of the content

words. In the close class items, the percentage of retrievals of the articles based on

acceptable included score is the highest followed by the prepositions. The reason

can be the limited number of items falling in these categories. However, the

percentage of retrievals of the adverbials and conjunction based on acceptable

included score is on lower side among close class items. In the context of open

class items, that is, content words, verbs and adjectives have lower retrievals in

comparison with the rest of the content words. Comparing function words, we can

say from the above Table that function words for the students are easier to retrieve.

In fact, most studies have shown comparable results. Function words tend to be

predictable from the overall structure of the sentence. On this basis, we can draw

the conclusion that functions words are easier to learn.

Correlation Analysis

The level of significance of the values of the correlation coefficient (r-value)

is dependent on the sample size. The sample size of this study is ninety and the

53

loading considered significant is as follows: r ≥ .26 p ≤ .01 and r ≥ .20 p

≤ .05

Table 4.14: Correlation Matrix

1 -.119 -.079 .423 ** .432 ** .870 ** -.024 -.131 .522 ** .006 -.285 ** -.248 * -.109 .057 -.032 -.133 .263 * -.428 **

1 .000 -.282 ** -.243 * -.025 -.006 -.060 -.069 -.025 .195 .251 * .061 -.013 -.061 .146 -.059 .281 **1 .249 * .198 .063 .296 ** .338 ** .122 -.072 -.348 ** -.073 -.319 ** -.326 ** -.343 ** -.345 ** -.325 ** -.142

1 .920 ** .400 ** .504 ** .396 ** .598 ** .138 -.413 ** -.491 ** -.320 ** -.291 ** -.162 -.533 ** .104 -.459 **1 .411 ** .505 ** .430 ** .636 ** .149 -.377 ** -.429 ** -.287 ** -.259 * -.211 * -.499 ** .056 -.469 **

1 .055 -.254 * .463 ** .043 -.290 ** -.178 -.086 .030 -.089 -.164 .120 -.415 **1 .617 ** .589 ** .310 ** -.249 * -.394 ** -.218 * -.313 ** -.266 * -.610 ** -.071 -.271 **

1 .892 ** .395 ** -.325 * -.314 * -.464 ** -.401 ** -.221 -.364 ** -.241 -.2251 .241 * -.358 ** -.390 ** -.258 * -.164 -.251 * -.445 ** .052 -.385 **

1 .021 -.063 -.072 -.089 .078 -.093 .104 -.1501 .405 ** .416 ** .235 * .266 * .354 ** .046 .420 **

1 .326 ** .448 ** .294 ** .549 ** -.128 .574 **1 .536 ** .117 .368 ** .163 .513 **

1 .139 .534 ** .181 .427 **1 .281 ** .206 .208 *

1 .120 .431 **1 -.041

1

SEX 2SES 14CLOZ.EXA 24

COLZ.ACC 25EXP.ENG 3SCH.PRIM 4SCH.SECO 5

SCH.TOTA 6MED.INST 15

AT.EN.BA 16ATT.ENGL 17INTE.MOT 18INSTR.MO19RESE.MO 20MANI.MO 21CL.ANXIE 22PAR.SUP 30

V1 V2 V14 V24 V25 V3 V4 V5 V6 V15 V16 V17 V18 V19 V20 V21 V22 V23

Correlation is signif icant at the 0.01 lev el (2-tailed).**.

Correlation is signif icant at the 0.05 lev el (2-tailed).*.

AGE 1

The correlation matrix of individual and social variables can be seen in Table 4.14

which has been taken from Appendix D.

An Overview

Language Proficiency and Social Psychological Variables:

4.15: Correlation of Social Psychological Variables with Cloze Test

Social Psychological Variable Cloze Test

Age .42**

Sex -.28**

Exposure to English .40**

Total Schooling .59**

SES .24**

Medium of Instruction .13

54

Attitude to English-Speaking Bangladeshis -.41**

Attitude to English -.49

Integrative Motivation -.32**

Instrumental Motivation -.29**

Resentment Motivation -.16

Manipulative Motivation -.53**

Class-room Anxiety .10

Parental Support -.45**

i. A significant correlation (r= .42; p ≤ .01) is found between age (v. 1) and Cloze

exact (v. 24). It suggests that the older students have performed better than the

younger ones in the Cloze test.

ii. A significant inverse correlation (r= -.28; p ≤ .01) exists between sex (v. 2) and

Cloze exact (v. 24). It suggests that boys have performed significantly better than

girls in the Cloze test.

iii. Cloze test exact (v. 24) correlates (r= .40; p ≤ .01) positively in a significant

way with exposure to English (v. 3). This shows that those students who have

more exposure to English perform better in Cloze test.

iv. Cloze test exact (v. 24) shows positive and significant correlation with total

schooling (r= .59; p ≤ .01). These positive correlations show that those students

who have got better primary, secondary and total schooling perform reasonably

well in the cloze test exact.

v. A significant positive correlation (r= .24; p ≤ .05) is found between SES (v. 14)

and exact Cloze (v. 24) retrievals. This suggests that those who have come from

55

high socio-economic status perform better than those of low socio-economic status

in the Cloze test.

vi. Cloze test exact (v. 24) is seen to correlate (r= -.41; p ≤ .01) negatively in a

significant way with attitude to English-speaking Bangladeshis (v 16). This

suggests that those students who have performed better in Cloze test exact do not

have positive attitude towards English-speaking Bangladeshis.

vii. Cloze test exact (v. 24) bears significant negative correlation (r= -.49; p ≤ .01)

with attitude to English (v. 17). This inverse correlation indicates that those

students who have got better score in Cloze test exact do not view English more

favorably.

viii. Cloze test exact (v. 24) holds significant inverse correlation (r= -.32; p ≤ .01)

with integrative motivation (v. 18). This negative correlation shows that the

students who have performed well in the Cloze test exact have less integrative

motivation for learning English.

ix. The correlation (r= -.29; p ≤ .01) of Cloze test exact (v. 24) with instrumental

motivation (v. 19) is also negative. This suggests that those students who have

performed reasonably well in the Cloze test exact do not have high level of

instrumental motivation to study English.

x. Cloze test exact (v. 24) correlates (r= -.53; p ≤ .01) significantly in a negative

way with manipulative motivation (v. 21). This negative correlation suggests that

those students who have got better score in the Cloze test exact do not sustain high

level of manipulative reasons for learning English.

xi. Cloze test exact (v. 24) has inverse correlation (r= -.459; p ≤ .01) with parental

support (v. 23) suggesting that those students who have performed better in Cloze

test exact have received less parental support for learning English.

The correlation matrix of Tables 4.15 shows that proficiency in English is

correlated far more significantly with social variables, such as, schooling, age,

56

parental support, exposure to English, attitude towards English-speaking

Bangladeshis, integrative motivation, etc. Schooling is significantly correlated

with proficiency in English. Variables which have relatively low correlation with

achievement in the Cloze test include sex, socio-economic status and instrumental

motivation. Variables which are slightly more significant include age and

integrative motivation. Exposure to English has a significant correlation with

proficiency in English. It is found among those students who have studied in

English medium school. Attitude to English and manipulative motivation are also

found to have significant correlations with proficiency in English. It appears that

the learners who are negative in their outlook to English and have less

manipulative reasons for learning English tend to be proficient in English. It is

very rare that manipulative motivation turns out to be significant; this area needs

further investigation.

Cloze Test Acceptable Included (v 25):

A highly significant positive correlation (r= .92; p ≤ .01) holds between Cloze test

exact (v. 24) and Cloze test acceptable included (v. 25) (see Table 4.14). This

correlation suggests that the students who have scored well in Cloze test exact

have also got better score in Cloze test acceptable included (in the same Cloze

test). Except for resentment motivation (v 21), all the variables that are either

positively or negatively correlated with Cloze test exact (v. 24) are in the same

way either positively or negatively correlated with Cloze test acceptable included

(v. 25). Cloze test acceptable (v. 25) correlates (r= -.21; p ≤ .01) significantly in a

negative way with resentment motivation (v. 20). This negative correlation

suggests that those students who have scored well in Cloze test acceptable

included do not have high level of resentment motivation for learning English.

Correlation of Individual and Social Psychological Variables:

57

Age (v. 1):

i. A highly significant positive correlation (r= .87; p ≤ .01) exists between

exposure to English (v. 3) and age (v. 1) and it indicates that the older students

have more exposure to English than the younger students.

ii. A significant positive correlation (r= .522; p ≤ .01) exists between age (v. 1) and

total schooling (v. 6) which simply suggests the almost obvious that more the

number of years higher the schooling.

iii. Age (v. 1) has negative correlation (r= -.28; p ≤ .01) with attitude towards

English-speaking Bangladeshis (v. 16) suggesting that younger students have more

positive attitudes towards English-speaking Bangladeshis. It is important to notice

that as children mature, their 'fascination' for the English-speaking adult declines.

iv. A significant negative correlation (r= -.24; p ≤ .05) is seen between age (v. 1)

and attitude to English (v. 17). It suggests that the younger students view English

more favorably than the older students in the sample.

v. A significant positive correlation (r=. 26; p ≤ .05) holds between age (v. 1) and

class-room anxiety (v. 22). This indicates that the older students have experienced

higher levels of anxiety in the class than the younger students.

vi. Age (v. 1) has significant negative correlation (r= -.42; p ≤ .01) with parental

support (v. 23) suggesting that the younger students have received more parental

support for learning English than the older students.

Sex (v. 2):

i. Sex (v. 2) has significant positive correlation (r= .25; p ≤ .05) with attitude to

English (v. 17) suggesting that girls have more positive attitude towards English

than boys.

ii. Sex (v. 2) shows significant positive correlation (r= .28; p ≤ .01) with parental

support (v. 23). This indicates that girls have received more parental support for

studying English than boys.

58

Exposure to English (v. 3):

i. Exposure to English (v. 3) is seen to correlate (r= .46; p ≤ .01) positively in a

significant manner with total schooling (v. 6). This suggests that older students

have more exposure of learning English than the younger students in the sample.

ii. Exposure to English (v.3) correlates (r= -.29; p ≤ .01) negatively with attitude

towards English-speaking Bangladeshis (v. 16). This means that the greater the

exposure to English, the less positive the attitude towards English-speaking

Bangladeshis.

iii. A significant negative correlation (r= -.41; p ≤ .01) exists between exposure to

English (v. 3) and parental support (v. 23). This suggests that more the exposure to

English the less the support, encouragement for learning English from parents. It

suggests that perhaps parents themselves do not have very high level of

proficiency in English. As children grow older and study advance level texts,

parents are probably unable to help them.

Total Schooling (v. 6):

i. Total schooling (v. 6) is seen to correlate (r= .24; p ≤ .05) positively with

medium of instruction (v. 15). This suggests that those who have better schooling

recommend more use of English as a medium of instruction in different levels of

education.

ii. Total schooling (v. 6) has inverse correlation (r= -.35; p ≤ .01) with attitude

towards English-speaking Bangladeshis (v. 16). Surprisingly, it suggests that those

who have better schooling (involving greater exposure to English) don’t have

positive attitude towards English-speaking Bangladeshis.

iii. A significant negative correlation (= -.39; p ≤ .01) is found between total

schooling (v. 6) and attitude to English (v. 17). This suggests that those who have

received better schooling do not view English favorably.

59

iv. Total schooling (v. 6) correlates (r= -.258; p ≤ .05) in a negative way with

integrative motivation (v. 18). It indicates that the students who have got better

schooling have less positive attitude to integrative motivation. This suggests that

the students don’t wish to assimilate to or imitate the way of life of English-

speaking Bangladeshis.

v. Total schooling (v. 6) has inverse correlation (r= -.251; p ≤ .05) with resentment

motivation (v. 20) suggesting that the students of the sample who have better

schooling learn English not out of resentment motivation. They do not learn

English because it is a part of their course or because their parents force them to.

vi. A highly significant negative correlation (r= -.44; p ≤ .01) is seen between total

schooling (v. 6) and manipulative motivation (v. 21). This shows that the students

who have better schooling do sustain less manipulative reasons for learning

English.

vii. Total schooling has negative correlation (r= -.38; p ≤ .01) with parental

support (v. 23). It suggests that the students who have better schooling receive less

parental support to study English. The reason can be that most of the parents

encourage their wards to go for English tuitions.

Socio-Economic Status (SES) (v. 14):

i. Socio-economic status (v. 14) has significant positive correlations (r= .29; p ≤

.01) and (r= .33; p ≤ .01) with primary schooling (v. 4) and secondary schooling

(v. 5) respectively. It indicates that the students those who have come from higher

class of society also have both better primary and secondary schooling.

ii. SES (v. 14) bears significant negative correlation (r= -.34; p ≤ .01) with attitude

towards English-speaking Bangladeshis (v. 16) suggesting that those students who

have come from lower stratum of society hold more positive attitude towards

English-speaking Bangladeshis than those students who have come from high

socio-economic status.

60

iii. A significant negative correlation (r= -.31; p ≤ .01) holds between SES (v. 14)

and integrative motivation (v. 18). This suggests that those who have come from

low socio-economic status tend to less have reservations about assimilating or

imitating the way of life of English-speaking Bangladeshis.

iv. SES (v. 14) is seen to show a significant inverse correlation (r= -.32; p ≤ .01)

with instrumental motivation (v. 19). It indicates that the students who have come

from lower stratum of society wish to learn English for instrumental purposes.

They believe that their knowledge of English will help to get higher education and

good jobs.

v. A significant negative correlation (r= -.343; p ≤ .01) is seen between SES (v.

14) and resentment motivation (v. 20). This suggests that those students who have

come from higher socio-economic status tend to have less resentment motivation

for learning English.

vi. SES (v. 14) correlates (r= -.345; p ≤ .01) negatively in a significant way with

manipulative motivation (v. 21). It suggests that those students who are coming

from higher socio-economic status have less manipulative reasons for learning

English.

vii. We see significant negative correlation (r= -.32; p ≤ .01) of SES (v. 19) with

classroom anxiety (v. 22). This negative correlation indicates that the students who

have come from high socio-economic status do not have any significant amount of

classroom anxiety in English classes because their parents spend substantial

amount of money to encourage their wards to devote more time in acquiring

proficiency in that language.

Attitude to English-Speaking Bangladeshis (v. 16):

i. Attitude to English-speaking Bangladeshis (v. 16) shows positive and significant

correlation (r= .40; p ≤ .01) with attitude towards English (v. 17). This suggests

61

clearly that the students who have a positive attitude towards English have in

general positive attitude towards English-speaking Bangladeshis.

ii. Attitude towards English-speaking Bangladeshis (v. 16) correlates (r= .41; p ≤

.01) positively in a significant way with integrative motivation (v. 18). This

correlation suggests that those students who have positive attitude towards

English-speaking Bangladeshis have wished to integrate with the way of life of

English-speaking Bangladeshis.

iii. Attitude towards English-speaking Bangladeshis (v. 16) bears significant

correlation (r= .23; p ≤ .05) with instrumental motivation (v. 19). This positive

correlation suggests that the students who have more positive attitude towards

English-speaking Bangladeshis have more integrative motivation. This is due to

the fact that they have realized the importance of English for higher education,

better career, etc.

iv. Attitude towards to English-speaking Bangladeshis (v. 16) is seen to correlate

(r= .26; p ≤ .05) in a positive way with resentment motivation (v. 20). This

suggests that the students who desire to integrate with the way of life of English-

speaking Bangladeshis learn English due to the pressure of course requirement or

their parents force them to.

v. Attitude towards English-speaking Bangladeshis (v. 16) is seen to show a

significant correlation (r= .35; p ≤ .01) with manipulative motivation (v 21). This

suggests that the students who have positive attitude towards English-speaking

have also manipulative reasons for learning English.

vi. Attitude towards English-speaking Bangladeshis (v. 16) correlates (r= .42; p ≤

.01) in a significant manner with parental support (v. 23). This correlation

indicates that those students who have more positive attitude towards English-

speaking Bangladeshis have more support for learning English from their parents.

62

Attitude to English (v. 17):

i. Attitude to English (v. 17) correlates (r= .32; p ≤ .01) significantly with

integrative motivation (v. 18). This suggests that the students who have positive

attitude to English have high level of integrative motivation for learning English.

ii. Attitude to English (v. 17) has highly significant correlation (r= .448; p ≤ .01)

with instrumental motivation (v. 19) suggesting that those who have positive

attitude to English have also high level of instrumental motivation for learning

English. This shows that English is perceived by the students not only as an

academic and economic ladder but also as means for enhancing social mobility

and individual personality.

iii. Attitude to English (v. 17) shows positive and significant correlation (r= .294;

p ≤ .01) with resentment motivation (v. 20). This indicates that the students who

have positive attitude towards English have high level of resentment motivation

for learning English.

iv. Attitude to English (v. 17) is seen to correlate (r= .549; p ≤ .01) positively in a

significant manner with manipulative motivation (v. 21). This correlation suggests

that the students who have positive attitude to English have very high level of

manipulative motivation.

v. Attitude to English (v. 17) shows highly positive significant correlation (r= .57;

p ≤ .01) with parental support (v. 23). This suggests that the students who have

positive attitude to English claim to have received more parental support for

learning English.

Integrative Motivation (v. 18):

i. Integrative motivation (v. 18) correlates (r= .536; p ≤ .01) highly significantly

with instrumental motivation (v. 19). This correlation suggests that the students

who have high level of integrative motivation also have high level of instrumental

motivation for learning English.

63

ii. Integrative motivation (v. 18) bears significant correlation (r= .36; p ≤ .01) with

manipulative motivation (v. 21). This indicates that those students who have

higher level of integrative motivation also have high level of manipulative reasons

for learning English.

iii. A highly significant correlation (r= .51; ≤ .01) exits between integrative

motivation (v. 18) and parental support (v. 23). This suggests that the students who

have high level of integrative motivation also have received more parental support

for learning English.

Instrumental Motivation (v. 19):

i. Instrumental motivation (v. 19) shows highly significant positive correlation (r=

.53; p ≤ .01) with manipulative motivation (v 21). This positive correlation

suggests that those students who have higher level of instrumental motivation also

have more manipulative reasons for learning English.

ii. Instrumental manipulative (v. 19) correlates (r= .42; p ≤ .01) significantly with

parental support (v. 23). This suggests that the students who have higher level of

instrumental motivation also have received more parental support for learning

English.

Resentment Motivation (v. 20):

i. A significant correlation (r= .28; p ≤ .01) exits between resentment motivation

(v. 20) and manipulative motivation (v. 21). This correlation indicates that the

students who have higher level of resentment motivation also have more

manipulative reasons for learning English.

ii. Resentment motivation (v. 20) shows significant correlation (r= .208; p ≤ .01)

with parental support (v. 23). This suggests that the students who have more

resentment motivation also receive significant support from their parents for

learning English.

Manipulative Motivation (v. 21):

64

A highly significant correlation (r= .43; p ≤ .01) exits between manipulative

motivation (v. 21) and parental support (v. 23). This correlation shows that the

students who have higher level of manipulative motivation also have received

substantial amount of support from their parents for learning English.

The network of all significant correlations can be seen in Figure 2.

+

+

From Figure 2, we can say that among the socio-biological variables, age,

exposure to English and schooling are not strongly and positively correlated with

Cloze but also among themselves. It is thus possible to say that in general, senior

students with higher exposure and better schooling have higher scores. Though

attitudinal and motivational variables are positively correlated among themselves,

they generally correlated negatively and strongly with Cloze scores. Though those

Age

English LanguageProficiency

Manipulative

Attitude toEnglish

Integrative Motivation

Exposure to English

Schooling

InstrumentalMotivation

ParentalSupport

AttitudeEng.Spe.Ban

. 42

. 40

. 59

- .41

-. 32- . 49

- . 53

- . 29

- . 45

+

+

+

+

+

+

+

+

+

+

+ +

++

-

-

-

-

- -+

+

+

+

-

-

+

65

who are positively inclined towards English are also positively inclined towards

English-speaking Bangladeshis and have strong motivation to learn English. They

do not end up scoring highly on the Cloze. As in several other studies (Agnihotri

et al. 1985, Khanna 1983 and Khanna and Agnihotri 1982,1984), it seems social

variables are more important for proficiency in a second language. This study in

fact has shown that the cluster of attitudinal and motivational variables may in fact

be negatively associated with language proficiency.

66

Conclusion

The main objective of the present study was to examine the relationship

between different social psychological variables on the one hand and proficiency

in English on the other. We had a total 90 students from class eight (30), class ten

(30) and twelve (30) from different institutions in Bangladesh. A set of 25

linguistic and social psychological variables was examined to decide the main

diagnostic determinants of the proficiency levels of the students in English. The

proficiency in English was measured through the Cloze procedure.

We have found a significant correlation between scoring for exact versus

acceptable included responses suggesting the Cloze procedure would well be

appropriate for use with non-native speakers. Except for medium of instruction,

resentment motivation and class-room anxiety, all other variables, namely, age,

sex, socio-economic status, exposure to English, schooling, attitude to English-

speaking Bangladeshis, attitude to English, integrative motivation, instrumental

motivation, manipulative motivation and parental support are significantly

correlated with the Cloze scores. We have observed that age was positively

correlated with the Cloze test, exposure to English and total schooling indicating

that those students who are older have better schooling, more exposure to English

and also have performed better in the Cloze test. Sex has inverse correlation with

Cloze which suggests that male students have done better than female students.

Cloze has shown significant positive correlation with socio-economic status. This

indicates that the learners who are from lower rung of society have not done well

in the Cloze test though they have positive attitude to English-speaking

Bangladeshis and also have realized the importance of English for getting better

education and good jobs. The reason might be that they do not enough access to

English. Those students who have performed better have come from high socio-

economic status and do not sustain positive attitude towards English-speaking

Bangladeshis and view English unfavourably. We have observed that the

maximum numbers of students, that is, over 78 per cent of them are at the

67

frustration level. The other reasons that are responsible for this poor performance

of the students are lack of standard educational system and overall poor teaching

by the unqualified English teachers. It is obvious that those who are from high

socio-economic status can afford to learn English by private tutors and also have

access to other sources. Thus, we see from the result that more positive the

attitudes and more integrative the motivation, the lower the level of proficiency of

English of the learners.

Our study has shown that the variables which appear to be most significant

in their relationship with total achievement includes age, exposure to English and

the kind of school attended. Total achievement in the Cloze test is highly

correlated with primary, secondary and total schooling. Schooling, thus, appears to

be the most significant variable in relation to achievement in English. Schooling

also has been found to be a very significant variable in several other studies.

Williams (1981) concerned with the relative significance of different social and

psychological variables in relation to the ability of Nigerian children in reading

English also found that the ‘type of school’ accounted for no less than 37 per cent

of the variance in the combined scores. Agnihotri et al. (1982) in their socio-

psychological study of achievement in English tenses found that schooling

explained 29 per cent of variance in the total achievement in the use of tenses.

Khanna et al. (1984) pointed out that speech skills were significantly influenced

by the type of school a student had attended. The importance of schooling in

second language learning becomes clear when we realize that it is only good

schools which provide suitable facilities encouraging the conversion of ‘input’ into

‘intake’. Agnihotri et al. (1983) point out that these schools generally have

qualified teaching staff, very good libraries and several recreational facilities

where children can learn while playing. On the other hand, teaching in government

schools generally concentrate on discrete grammatical points and provide very

little interesting exposure. The ordinary schools cater to children coming from

relatively poor socio-economic background. There are minimal opportunities to

68

use English and the library and recreational facilities are limited. Not very

different situation prevails in Bangladesh. It is obligatory for the students to use

English all the time in English medium schools. On the other hand, English is

seldom used outside the class-room in government schools. There are minimal

opportunities to use English for the students in ordinary semi-government schools

and the teaching is done through Bengali even in English classes.

Socio-economic status was found to play a very important role in the

development of proficiency in English. This finding corroborates the results of

many other investigations reported by Burstall (1975). Robinson (1971) noticed

that children with parents in high status occupation tend to receive more parental

support when they approach new learning experiences than do those with parents

in lower status occupation. We have noticed that positive attitude towards the

English-speaking Bangladeshis do not correlate significantly with achievement in

English proficiency. It was expected that the socio-economic status of the learners

would correlate highly with their schooling and with overall achievement in the

Cloze test. Children from higher socio-economic background generally go to

schools and speak English more fluently than others. But, in the test administered

by us, socio-economic status seems to bear very low correlation with achievement.

The present study has shown that age is one of most significant predictors of

proficiency in English. Hamid’s (2001) work based on Syrian college students

acquiring English also found that age was one of the most significant predictors of

achievement and proficiency in English.

We have also found that social psychological variables significantly interact

with social variables and social variable are found to be more important for

language proficiency. Our study has also shown that out of all social psychological

variables only attitude toward English and English-speaking Bangladeshis and

motivation manipulatively instrumental in nature have significant correlations with

proficiency in English. It seems that in foreign language learning situation such as

those examined by Gardner and Lambert (1972), social psychological variables

69

may be more important in determining proficiency/ achievement in a foreign

language whereas in second language learning situations such as that of

Bangladesh, social variables are more important. Khanna (1983) found that the

best predictor of achievement in English were proficiency in English and

schooling. It appears that different sets of social psychological variables are likely

to cluster with different linguistic skills in different social settings. Our study

shows that attitudinal motivations have strong negative correlations. Though the

result is in line with Indian studies where socio-psychological variables, such as,

schooling and exposure to English are found as important predictors for English

proficiency. The high negative correlations of attitudinal motivations of our study

require further investigations.

We have examined the proficiency in English in a comprehensive

sociopsychological framework in our study. We have also noticed how statistical

tools such as correlation can be used to select significant variables from manifold

different social and social psychological variables. The results of our present study

confirm that schooling is the most significant predictor of achievement and

proficiency in second language learning. In addition to schooling, age and

exposure to English also contribute substantially towards language proficiency in

English.

70

APPENDIX A

Please read the whole text once without filling in the blanks. Read it again and fill

in the blanks with the most appropriate word you can think of.

Fill only one word in each blank.

LALAJEE

BY JIM CORBETT

And so Lalajee left me, taking with him the greater part of my savings. That I

would see him 1 I never doubted, for the 2 of India never forget a

3 ; but the promise Lalajee had 4 was, I felt sure, beyond 5

powers of accomplishment. In this 6 was wrong, for returning late 7

evening I saw a man 8 in spotless white standing in 9 veranda. The

light from the 10 behind him was in my 11 , and I did not recognize

12 until he spoke . It was 13 , come a few days before 14 expiry of

the time limit 15 had set himself. That night 16 he sat on the floor

17 my chair he told me 18 his trading transactions, and the 19

that had attended them. Starting 20 a few bags of grain 21 being

content with a profit 22 only four annas per bag 23 had gradually,

and steadily, built 24 his business until he was 25 to deal in

consignments up 26 thirty tons in weight, on 27 he was making a

profit 28 three rupees per ton. His 29 was in a good school, 30

as he could now afford to keep a wife he had married the daughter of a rich

merchant of Patna; all this he had accomplished in a little under twelve months. As

71

the time drew near for his train to leave he laid five one-hundred rupee notes on

my knee.

List of Deleted Words of the Test

1 Again 11 Him 21 And 2 Poor 12 Lalajee 22 Of 3 Kindness 13 The 23 He 4 Made 14 He 24 Up 5 His 15 As 25 Able 6 One 16 Near 26 To 7 Dressed 17 Of 27 Which 8 My 18 Success 28 Of 9 Room 19 With 29 Son 10 Eyes 20 And 30 And

APPENDIX B

INFORMANT NO:QUESTIONNAIRE

1. Name:2. Address:3. Telephone Number:4. Course: Science/Humanities5. Age:6. Sex:7. Date of Birth:8. The first language you learnt to speak as a child:9. At what age did you start learning/using English:10. In which class did you join this school?11. Using the scale below, please tell us about the educational institutions youattended in terms of their prestige, facilities available for learning English andthe medium of instruction:

Prestige Medium of instruction

Ordinary school =1 Bangla =1Quite prestigious =2 Both Bangla and English =2

72

Highly prestigious =3 Only English =3

Facilities for English

Very little =1Some =2A lot =3

(Please put the appropriate number in each box)

Level Prestige Facilities Medium ofInstructions PrimarySecondary

12. Using the Scale below, provide the following information about your familymembers:

Monthly Income Fluency in English(Tk Per Month)Below 2000 =1 Nil =12000- 3499 =2 Very little =23500-4999 =3 Moderate =35000-and above =4 Good =4

Very high =5

Age EducationalQualification

Occupation MonthlyIncome

Fluency inEnglish

Father Mother Brother (s) Sister (s) Any Other

13.What, according to you, should be the medium of instruction at the followinglevels of education?

Level Medium of instruction

Primary schoolSecondary schoolHigher secondary school

73

B.AB.Sc. M.AM.Sc. EngineeringManagementFashion designingTravel and TourismComputer education

14. Please indicate your opinion about each of the following statements by circlingthe alternative which best indicates the extent to which you agree or disagree withthat statement.

i. The English-speaking Bangladeshis are warm-hearted. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

ii. The English-speaking Bangladeshis are very progressive. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

iii. The English-speaking Bangladeshis are generally dishonest. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

iv. English is the most suitable language for creating writing (poetry, drama, etc). (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

v. English is the most suitable language for scientific thinking. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

vi. English is important for me because it will allow me to interact with theEnglish-speaking Bangladeshis. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

vii. English is important for me because it will allow me to interact with moreand varied people in the world. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

viii. English is important for me because it will help me to appreciate Englishart and literature.

74

(a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

ix. English is important for me because it will help me to live and behave likethe English-speaking Bangladeshis. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

x. English will help me to build my career. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xi. English is very important for me because it would be useful for highereducation in Bangladesh. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xii. I learn English because it is a part of my course requirement. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xiii. I learn English because my parents want me to learn it. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xiv. English is important for me because it will help me to become a moreknowledgeable person. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xv. English is important for me because it will add to my prestige andpersonality. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xvi. English is important for me because it will enable me to my get work doneeverywhere in Bangladesh. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xvii. English will help me to use other people for my personal gains. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xviii. My parents try to help me with English. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

75

xix. My parents think that I should devote more time to my English studies. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xx. My parents have stressed the importance on English will have for mewhen I leave school/college. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xxi. My parents encourage me to seek help from my English teacher. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xxii. I always feel that other students speak English better than me. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xxiii. I get nervous and confused when I am speaking in my English class. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xxiv. I am afraid of the other students will laugh at me when I speak English. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xxv. English has exercised a good influence on the system of education inBangladesh. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xxvi. Greater effort should be made to spread English throughout Bangladesh. (a) Agree (b) slightly agree (c) neither agree or nor disagree (d)

slightly disagree (e) disagree

xxvii. Knowledge of English is essential in order to be a leader in theBangladeshi society.

(a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xxviii. If English is abolished, it would hamper the progress of science andtechnology in Bangladesh.

(a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xxix. English has important role than Bangla to play in bringing about politicalunity in Bangladesh.

(a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

xxx. English should be abolished from Bangladesh.

76

(a) Agree (b) slightly agree (c) neither agree or nor disagree (d)slightly disagree (e) disagree

APPENDIX C

Coding Sheet

Variable No. Name Coding 1 Age Running 2 Sex Male=1; Female=23 Exposure to English Running 0-144 Primary Schooling 3-95 Secondary Schooling 3-96 Total Schooling (4-5) 6-187 Residence 1-58 Father’s Education 1-59 Mother’s Education 1-510 Father’s Income 1-411 Mother’s Income 1-412 Father’s Occupation 1-613 Mother’s Occupation 1-614 Total SES(7-13) 7-3515 Medium of Instruction 12-3616 Attitude to English-Speaking

Bangladeshis 3-15

17 Attitude to English 8-4018 Integrative Motivation 4-2019 Instrumental Motivation 4-2020 Resentment Motivation 2-1021 Manipulative Motivation 2-1022 Class-room Anxiety 3-1523 Parental Support 4-2024 Cloze Test Exact 1-3025 Cloze Test Acceptable Included 1-30

77

APPENDIX D

Variable, Mean and Standard Deviation

Variable No Variables Mean Standard Deviation1 Age 15.33 2.022 Sex 1.41 .4963 Exposure to

English 9.43 2.27

4 Schooling Primary 5.14 1.725 Schooling

Secondary 5.65 1.49

6 Schooling Total 8.92 3.6214 SES 18.50 4.3615 Medium of

Instruction 32.27 2.31

16 Attitude toEnglish-SpeakingBangladeshis

11.46 2.35

17 Attitude to English 31.49 4.7918 Integrative

Motivation 16.74 3.21

19 InstrumentalMotivation

17.87 2.40

20 ResentmentMotivation

7.93 2.40

21 ManipulativeMotivation

7.23 2.49

22 Class-roomAnxiety

8.53 3.94

23 Parental Support 16.47 3.8424 Coze Test Exact 10.11 3.9425 Cloze Test

AcceptableIncluded

16.73 5.65

78

Correlation Matrix

1 -.119 -.079 .423 ** .432 ** .870 ** -.024 -.131 .522 ** .006 -.285 ** -.248 * -.109 .057 -.032 -.133 .263 * -.428 **

1 .000 -.282 ** -.243 * -.025 -.006 -.060 -.069 -.025 .195 .251 * .061 -.013 -.061 .146 -.059 .281 **1 .249 * .198 .063 .296 ** .338 ** .122 -.072 -.348 ** -.073 -.319 ** -.326 ** -.343 ** -.345 ** -.325 ** -.142

1 .920 ** .400 ** .504 ** .396 ** .598 ** .138 -.413 ** -.491 ** -.320 ** -.291 ** -.162 -.533 ** .104 -.459 **1 .411 ** .505 ** .430 ** .636 ** .149 -.377 ** -.429 ** -.287 ** -.259 * -.211 * -.499 ** .056 -.469 **

1 .055 -.254 * .463 ** .043 -.290 ** -.178 -.086 .030 -.089 -.164 .120 -.415 **1 .617 ** .589 ** .310 ** -.249 * -.394 ** -.218 * -.313 ** -.266 * -.610 ** -.071 -.271 **

1 .892 ** .395 ** -.325 * -.314 * -.464 ** -.401 ** -.221 -.364 ** -.241 -.2251 .241 * -.358 ** -.390 ** -.258 * -.164 -.251 * -.445 ** .052 -.385 **

1 .021 -.063 -.072 -.089 .078 -.093 .104 -.1501 .405 ** .416 ** .235 * .266 * .354 ** .046 .420 **

1 .326 ** .448 ** .294 ** .549 ** -.128 .574 **1 .536 ** .117 .368 ** .163 .513 **

1 .139 .534 ** .181 .427 **1 .281 ** .206 .208 *

1 .120 .431 **1 -.041

1

SEX 2SES 14CLOZ.EXA 24

COLZ.ACC 25EXP.ENG 3SCH.PRIM 4SCH.SECO 5

SCH.TOTA 6MED.INST 15

AT.EN.BA 16ATT.ENGL 17INTE.MOT 18INSTR.MO19RESE.MO 20MANI.MO 21CL.ANXIE 22PAR.SUP 30

V1 V2 V14 V24 V25 V3 V4 V5 V6 V15 V16 V17 V18 V19 V20 V21 V22 V23

Correlation is signif icant at the 0.01 lev el (2-tailed).**.

Correlation is signif icant at the 0.05 lev el (2-tailed).*.

AGE 1

79

BIBLIOGRAPHY

Agnihotri,R.K. and Khanna,A.L.1991. ‘Evaluating the readability of a schooltextbook: an Indian study’. Journal of Reading 35.4:282-288.

Agnihotri,R.K. and Khanna,A.L.(eds.), 1994. Second Language Acquisition:Socio- cultural and Linguistic Aspects of English in India (RAL 1). New Delhi:Sage Publications.

------(eds.), 1995. English Language Teaching in India: Issues and Innovations(RAL 2). New Delhi: Sage Publications.

------(eds.),1997. Problematizing English in India (RAL 3). New Delhi: SagePublications.

Agnihotri,R.K.,Khanna,A.L and Mukherjee,A.1982. ‘Predictors of achievement inEnglish tenses: a sociopsychological study’. Indian Journal of Applied Linguistics8.3:89-105.

------.1983. Variations in the use of tenses in English: a sociolinguisticsperspective. New Delhi: ICSSR.

------.1988. Tense in Indian English: A Sociolinguistic Study. New Delhi: ICSSRand Bahri Publicatons.

Agnihotri,R.K., Khanna,A.L and Sachdev,I.(eds.),1998. Social PsychologicalPerspectives on Second language Acquisition. New Delhi: Sage Publications.

Alderson,J.C.1978. ‘The use of Cloze Procedure with Native and Non-nativeSpeakers of English’. Ph.D dissertation. University of Edinburgh.

------.1979. ‘The cloze procedure and proficiency in English as a foreignlanguage’. TESOL Quarterly 13.2:219-223.

------.1979a. ‘Scoring procedures for use on cloze tests’. Paper presented at the1979 TESOL Convention, Boston, Massachusetts.

------.1980. ‘ Native and non-native speaker performance on cloze tests’. Language Learning 30.1:59-76.

------.1983. ‘The cloze procedure and proficiency in English as a foreignlanguage’. In Oller,J.W.Jr.1983 (ed.), Issues in language testing research. Rowley, Massachusetts: Newbury House Publishers.

80

Allen,V.F.1968. ‘Towards a thumbnail test of English competence’. TESOLQuarterly 2:241-246.

Anderson,J.1967. ‘A scale to measure the reading difficulty of children’s books’. University of Queensland, Faculty of Education Paper 1.6. St. Lucia: University ofQueensland Press.

------.1976. Psycholinguistic experiments in foreign languages testing. St.Lucia, Queensland, Australia: University of Queensland Press.

Anisfeld,M. and Lambert,W.E.1961. ‘Social and Psychological variables inlearning Hebrew’. Journal of Abnormal and Social Psychology 63:524-529.

Au,S.Y.1984. ‘Social-psychological variable in second language learning’. Master’s thesis. City University.

------.1988. ‘A critical appraisal of Gardner’s socio-psychological theory ofsecond-language (L2) learning’. Language Learning 38.1:75-100.

Bachman,L.F.1982. ‘The trait structure of cloze test score’. TESOL Quarterly16.1:61-69.

------.1985. ‘Performance on cloze tests with fixed-ratio and rational deletions’. TESOL Quarterly 19:535-555.

Bartz,W.H.1974. ‘A study of the relationship of certain learner factors with theability to communicate in a second language (German) for the development ofmeasures of communicative competence’. Ph.D dissertation. The Ohio StateUniversity.

Bensoussan,M.and Ramraz,R.1984. ‘Testing EFL reading comprehension using amultiple-choice rational cloze’. Modern language Journal 68.3:230-239.

Bialystok,E. and Howard,J.1979. ‘Inferencing as an aspect of cloze testperformance’. Working Papers on Bilingualism 17:24-36.

Bormuth,J.R.1962. ‘Cloze Tests as Measures of Readability and ComprehensionAbility’. Ph.D dissertation. Florida State University.

------.1963. ‘Cloze as a measure of readability’. Proceedings of the InternationalReading Association 8:131-134.

81

------.1969. ‘Factor validity of cloze tests as measures of reading comprehensionability’. Reading Research Quarterly 4:358-365.

Brown,H.D.1973. ‘Affective variables in second language acquisition’. LanguageLearning 23: 231-244.

Brown,J.D.1980. ‘Relative merits of four methods for scoring cloze tests’. ModernLanguage Journal 64:311-317.

------.1983. ‘A closer look at cloze: validity and reliability’. In Oller,J.W.Jr. (ed.), Issues in Language testing research. Rowley, Massachusetts: Newbury House.

Brual,T.G.1962. ‘An Experiment Approach to the Determination of theReadability of Selected Children’s Books’. Unpublished Thesis. University ofQueensland.

Burstall,C.1975. ‘Factors affecting foreign language learning: a consideration ofsome recent research findings’. Language Teaching and Linguistics: Abstracts8.1:5-21.

Carrol,J.B.1960. ‘Foreign Languages for Children: What Research says’. NationalElementary Principal 39:12-15.

------.1962. ‘The prediction of success in intensive foreign language training’. InGlaser,R.(ed.), Training research and education. Pittsburgh: University ofPittsburgh Press.

------.1963. ‘Research on teaching foreign languages’. In Gage,N.L. (ed.), Handbook of research on teaching. Chicago: Rand McNally &Co.

------.1972. ‘Defining language comprehension: some speculations’. In Carroll,J.B. and Freedle,R.O.(eds.), Language comprehension and the acquisition ofknowledge 1-29.

Carroll,J.B. and Sapon,S.M.1959. Modern language aptitude test ( MLAT)manual. New York: The Psychological Corporation

Carroll,J.B, Carton,A.S and Wilds,C.P.1959. ‘An investigation of cloze items inthe measurement of achievement in foreign languages’. A report on researchconducted under a grant from the College Entrance Examination Board. Laboratory for Research in Instruction, Graduate School of Education, HarvardUniversity, Cambridge, Massachusetts.

82

Chavez-Oller,M.A., Chihara,T.,Weaver,K.A and Oller,J.W.Jr.1985. ‘When arecloze items sensitive to constraints across sentences?’. Language Learning 35.2:181-205.

Chihara,T.,Oller,J.W.Jr.,Weaver,K.A and Chavez-Oller,M.A.1977. ‘Are clozeitems sensitive to constraints across sentences?’. Language Learning 27.1:63-73.

Chomsky,N.1959. ‘A Review of B.F. Skinner’s Verbal Behaviour’. Language35.1: 26-58.

Clarke,M.1979. ‘Reading in Spanish and English: Evidence from adult ESLstudents’. Language Learning 29:121-145.

ment, R.,Gardner,R.C and Smythe,P.C.1977. ‘Motivational variables in secondlanguage acquisition: A study of francophones learning English’. CanadianJournal of Behavioural Science 9:123-133.

Cohen,A.D.1980. Testing language ability in the classroom. Massachusetts:Newbury House Publishers.

Conrad,C.1970. ‘The Cloze Procedure as a Measure of English Proficiency’. Master’s thesis. University of California.

Cook,V.J.1978. ‘Second language learning: psycholinguistics perspective’. (Survey Article) Language Teaching Abstracts 11.2:73-84.

Cziko,G.1978. ‘Differences in first and second language reading: the use ofsyntactic, semantic and discourse constraints’. The Canadian Modern languageReview 34:437-490.

Darnell,D.K.1968. The Development of an English Language Proficiency Test ofForeign Students Using a Clozentropy Procedure. Boulder,Colorado: Universityof Colorado. (EDRS:ED 024039).

------.1970. ‘Clozentropy: a procedure for testing English language proficiency offoreign students’. Speech Monographs 37:36-46.

Douglas,J.W.B.1964. The home and the school: a study of ability and attainmentin primary school. London: MacGibbon and Kee.

Deyes,T.1984. ‘Towards an authentic discourse cloze’. Applied Linguistics 5.2:128-137.

Feenstra,H.J. and Gardner, R.C.1968. ‘Aptitude and motivation in second

83

language acquisition’. Research Bulletin No.101, University of Western Ontario.

Fillenbaum,S. and Jones,L.V.1962. ‘An application of ‘cloze’ technique to thestudy of aphasic speech’. Journal of Abnormal and Social Psychology 65: 183-201.

Fillenbaum,S.,Jones,L.V., Rapoport,A.1963. ‘The predictability of words and theirgrammatical classes as a function of rate of deletion from a speech transcript’. Journal of Verbal Learning and Verbal Behaviour 2:186-194.

Foley,J.A.1983. ‘More questions on assumptions about cloze testing’. RELCJournal 14.1:57-69.

Gardner,R.C.1965. ‘A language aptitude test for blind students’. Journal ofApplied Psychology 49:135-141.

------.1966.‘Motivational variables in second-language acquisition’. InternationalJournal of American Linguistics 32:24-44.

------.1968. ‘Attitudes and motivation: their role in second language acquisition’. TESOL Quarterly 2:141-150.

------.1980. ‘On the validity of affective variable in second language acquisition:conceptual, contextual and statistical considerations’. Language Learning 30:225-270. ------.1983. ‘Learning another language: A true social psychological experiment’. Journal of Language and Social Psychology 2:219-240.

------.1985. Social Psychology of Second Language Learning: The role of attitudesand motivation. London: Edward Arnold Publishers.

------.1988. ‘The socio-educational model of second language learning:assumptions, findings, and issues’. Language Learning 38.1:101-121.

Gardner,R.C and ment, R.1990. ‘Social psychological perspectives on secondlanguage acquisition’. In Giles, H. and Robinson, P. (eds.), Handbook of Language andSocial Psychology. Chichester: Wiley.

Gardner,R.C., ment, R.,Smythe,P.C. and Smythe,C.L.1979. ‘TheAttitude/Motivation Test Battery-Revised Manual’. Research BulletinNo.15.Language research Group. University of western Ontario, Laondon. Canada.

84

Gardner,R.C and Lambert, W.E.1959. ‘Motivational variables in second languageacquisition’. Canadian Journal of Psychology 13:226-272.

------.1965. ‘Language aptitude, intelligence and second language achievement’. Journal of Educational Psychology 56.4:191-199.

------. 1972. Attitudes and Motivation in Second Language Learning. Rowley, Massachusetts: Newbury House. Gardner,R.C.,Lalonde,R.N. and Moorcroft,R.1985. ‘The role of attitudes andmotivation in second language learning: Correlational and experimentalconsiderations’. Language Learning 35: 207-227.

Gardner,R.C. and Smythe,P.C.1981. ‘On the development of the AttitudeMotivation Test Battery’. Canadian Modern Review 37:510-525.

Gardner,R.C.,Smythe,P.C., ment,R. and Gliksman,L.1976. ‘Second languagelearning: a social psychological perspective’. Canadian Modern Language Review32.3:198-213. Gliksman,I.1981. ‘Improving the prediction of behaviours associated with secondlanguage acquisition’. Ph.D dissertation. University of western Ontario.

Hamid,N.A.2001. ‘The role of anxiety in second language learning: a case studyof Syrian college students’. M Phil thesis. University of Delhi.

Hansen,J.1984. ‘Field dependence-independence and language testing: evidencefrom six pacific island cultures’. TESOL Quarterly 18.2:311-324.

Haskell,J.F.1976. ‘Using cloze to select reading material’. TESOL Newsletter16.11:15-16.

Hatch,E.1976. ‘Conversational analysis: an alternative methodology for secondlanguage acquisition research’. Proceedings of the NWAVE-V Conference. Georgetown.

Hinofotis,F.B.1980. ‘Cloze as an alternative method of ESL placement andproficiency testing’. In Oller,J.W.Jr., Perkins,K. (eds.), Research in languagetesting. Rowley, Massachusetts: Newbury House.

Jokobovits,L.A.1970. Foreign language learning: as psycholinguistic analysis ofthe issues. Rowley, Massachusetts: Newbury House.

85

Jonz,J.1976. ‘Improving on the basic egg: The M-C cloze’. Language learning 26:255-256.

------.1990. ‘Another turn in conversation: what does cloze measure?’ TESOLQuarterly 24.1:61-83.

Kerr,A.H.1968. ‘Cloze Procedure as Measure of the Readability of French as aSecond Language’. Bachelor’s thesis. University of New England.

Khanna,A.L.1983. ‘A Study of Some learner Variables in Learning English as asecond language’. Ph.D dissertation. University of Delhi.

Khanna,A.L. and Agnihotri,R.K.1982. ‘Language achievement and some socialpsychological variables’. Central Institute of English and Foreign LanguagesBulletin 18.2:41-51.

------.1984. ‘Some predictors of speech skills: a sociopsychological study’. International journal of Dravidian Linguistics 13.2:229-251.

Khanna,A.L., Vrema,M.K., Agnohotri, R.K. and Sinha,S.K.1990. ‘Attitudes andmotivation of adult ESOL learners in Great Britain: a cross cultural pilot study’. Language Issues 4.1:4-8.

Klare,G.R., Sinaiko,H.W. and Stolurow,L.M.1972. ‘The cloze procedure: aconvenient readability test for training materials and translations’. InternationalReview of Applied Psychology 21.2:77-105.

Knight,D.W.1966. ‘The reading performance of students with ninth grade readingachievement on occupation information materials written with various levels ofreadability’. Ph.D dissertation. Florida State University.

Krashen,S.D.1978a. ‘Individual variation in the use of the Monitor’. InRitchie,W.C.(ed.), Second language acquisition research: issues and implications. New York: Academic Press.

------.1978b. ‘The monitor model for second language acquisition’. InGingras,R.C.(ed.), Second language acquisition and foreign language teaching. Arlington,va: Centre for Applied Linguistics 1-26.

------.1982. Principles and practice in second language acquisition. Oxford:Pergamon Institute of English.

Kuppuswamy,B.1962. Manual of socio-economic status scale (urban). Delhi:Manasayan.

86

Lado,R.1961. Language Testing. New York: McGraw-Hill.

------.1964. Language Teaching: A Scientific Approach. New York: McGraw-Hill.

Lambert,W.E.1963. ‘Psychological approaches to the study of language. Part II:On second language learning and bilingualism’. Modern Language Journal14:114-121.

Lambert,W.E. and Tucker,G.R.1972. Bilingual education of children: the St. Lambert Experiment. Rowley, Massachusetts: Newbury House.

Lambert,W.E.,Gardner,R.C.,Olton,R. and Tunstall,K.1961. A study of the rules ofattitudes and motivation in second language learning. McGill University Mimeo.

Lambert,W.E., Gardner,R.C., Barik,H.C. and Tunstall,K.1962. ‘Attitudinal andcognitive aspects of intensive study of a second language’. Journal of Abnormaland Social Psychology 66:358-68. Larsen-Freeman,D.E.1976. ‘An explanation for the morpheme acquisition order ofsecond language learners’. Language Learning 26.1:125-134.

Lukmani,Y.M. 1972. ‘Motivation to learn and language proficiency’. LanguageLearning 22.2:261-273.

Lyczak,R., Fu,G.S.and Ho, A.1976. ‘Attitudes of Hong Kong bilinguals towardsEnglish and Chinese speakers’. Journal of Cross-Cultural Psychology 7:425-438.

MacGinite,W.H.1961. ‘Contextual constraints in English prose paragraphs’. Journal of Psychology 51:121-130.

Markham,P.L.1985. ‘The rational deletion cloze and global comprehension inGerman’. Language Learning 35:423-430.

Mathur,C.1991. ‘The Role of Attitudes and Motivation in Foreign LanguageLearning: the Case of German in India’. M Phil thesis. University of Delhi, Delhi.

Mauranen,A.1987. ‘Can gaps measure comprehension? Modifications of cloze astests of reading’. In Lauren,C., Nordam,M.(eds.), Special language: From humanthinking to thinking machines. Philadelphia: Multilingual Matters Ltd.

McLeod,J.1962. ‘The estimation of readability of books of low difficulty’. BritishJournal of Educational Psychology 32:112-118.

87

Miller,G.R.,Cloeman,E.B.1967. ‘A set of Thirty-six prose passages calibrated forcomplexity’. Journal of verbal learning and Verbal behaviour 6:851-854.

Mullen,K.1979. ‘More on cloze tests as tests of proficiency in English’. InBriere,E.J. and Hinofotis,F.b (eds.), Concepts in Language Testing: Some RecentStudies. TESOL 21-34.

Mueller,T.H.1971. ‘Students attitude in the basic French courses at the Universityof Kentucky’. Modern Language Journal 55.5:290-298.

Nida,E.A.1958. ‘Some psychological problems in second language Learning’. Language learning 8.1: 7-15.

Oller,J.1972. ‘Scoring methods and difficulty levels for cloze tests of ESLproficiency’. Modern Language Journal 56:151-158.

Oller,J.W.Jr.1972. ‘Scoring methods and difficulty levels for cloze tests ofproficiency in English as a second language’. Modern Language journal 56:151-158.

------.1973. ‘Cloze tests of second language proficiency and what they measure’. Language learning 23.1:105-118.

------.1975. ‘Cloze, discourse and approximation to English’. In New directions inTESOL, Burt,M.k., and Dulay,H.C.(eds.), 345-355. Washington,D.C.: TESOL.

Oller,J.W., Hudson,A.J. and Liu,P.F.1977. ‘Attitudes and attained proficiency inESL: a socio-linguistic study of native speakers of Chinese in the United States’.Language Learning 27:1-26.

Oller,J.W.and Inal,N.1971. ‘A cloze test of English prepositions’. TESOLQuarterly 5:315-326.

Oller,J.W.Jr., and Perkins,K.1980. Research in language Testing. Rowley, Massachusetts: Newbury House.

Ozete,O.1977. ‘The cloze procedure: a modification’. Foreign Language Annals10.5:565-568.

Peal,E. and Lambert,W.E.1962. ‘The relation of bilingualism to intelligence’. Psychological Monographs LXXVI, No.27(Whole issue No.546).

Peisach,E.C.1965. ‘Children’s comprehension of teacher and peer speech’. ChildDevelopment 36:467-480.

88

Pike,L.W.1973. An evaluation of present and alternative item formats for use inthe test of English as a foreign language. Princeton, New Jersey: EducationalTesting Service.

Pimsleur,P., Mosberg,L. and Morrison,A.L.1962. ‘Students factors in foreignlanguage learning’. Modern language Journal 46:160-170.

Pimsleur,P.,Sundland,D.M and Mcintyr,R.D.1964. ‘Under-achievement in foreignlanguage learning’. International Review of Applied Linguistics 2:113-150.

Porter,D.1978. ‘Cloze procedure and equivalence’. Language Learning 28.2:333-340.

------.1983. ‘The effects of quantity of context on the ability to make linguisticpredictions: a flaw in a measure of general proficiency’. In Hughes.A andPorter,D.(eds.), Current developments in language testing. London: AcademicPress.

Potter,J.C.1968. A taxonomy of cloze research, part 1: readability and readingcomprehension. Southwest Regional Laboratory for Educational Research andDevelopment. California: Englewood.

Ramanauskas,S.1972. ‘The responsiveness of cloze readability measures tolinguistic variable opening over segments of text longer than a sentence’. ReadingResearch Quarterly 8:79-91.

Rand,E.1978. ‘The effects of test length and scoring method on the precision ofCloze test scores’. Workpapers in Teaching English as a Second Language 12:62-71. Richards,J.C.1972. ‘Social factors, interlanguage and language learning’. Language learning 22.2:150-188.

Robinson,W.P.1971. ‘Social factors and language development in primary schoolchildren’. In Huxley,R. and Ingram,E.(eds.), Language Acquisition: models andmethods. New York: Academic Press.

Ryan,E.B. and Semmel,M.I.1969. ‘Reading as a constructive language process’. Reading Research Quarterly 5:59-83.

Rye,J.1982. Cloze Procedure and the Teaching of Reading. London: Heinemann.

Sahgal,A.1983. ‘A sociolinguistic study of the spoken English of Delhi elite’. MPhil thesis. University of Delhi.

89

------.1992. ‘Bilingualism and Scholastic Achievement’. Ph.D dissertation. University of Delhi.

Schumann,J.H.1978. ‘The acculturation model for second language acquisition’. In Gingras,R.C.(ed.), Second language acquisition and foreign language teaching. Virginia: Centre for Applied Linguistics.

Sciarone,A.G. and Schoorl,J.J.1989. ‘The cloze test: or why small isn’t alwaysbeautiful’. Language Learning 39.3:415-438.

Secord,P.F. and Backman,C.W.1964. Social Psychology. New York: McGraw-Hill.

Shahed,F.H.2001. ‘English in Bangladesh: A study of Urban Educated PublicAttitude’. Ph.D dissertation. Jawaharlal Nehru University.

Shanahan,T., Kamil,M and Tobin,A.W.1982. ‘Cloze as a measure of intersentialcomprehension’. Reading Research Quarterly 17.2:229-255.

Smart,J.,Elton,C.F. and Burnett,C.W.1970. ‘Underachievers and over-achievers inintermediate French’. Modern Language Journal 54.6:415-420.

Stevick,E.1976. Memory, meaning and method. Rowley: Massachusetts: NewburyHouse.

Stubbs,J.B and Tucker,G.R.1974. ‘The cloze test as measure of ESL proficiencyfor Arab students’. The Modern language Journal 58: 239-241.

Taylor,W.1953. ‘Cloze procedure: a new tool for measuring readability’. Journalism Quarterly 30:414-38.

Titone,R.1977. ‘Teaching second language in multilingual/multicultural contexts’. Paper presented in UNESCO meeting in Paris.

Tuinman,J.J.,Blanton,W.E and Gray,G.1975. ‘A note on cloze as a measure ofcomprehension’. Journal of Psychology 90.2:159-162.

Weaver,W.W.,Kingston,A.J.1963. ‘A factor analysis of the cloze procedure andother measures of reading and language ability’. Journal of Communication13:252-261.

Williams,D.1981. ‘Factors related to performance in reading English as a secondlanguage’. Language Learning 21.1:31-50

90

Wittenborn,J.R. and Larsen,R.P.1944. ‘A factorial study of achievement in collegeGerman’. Journal of Educational Psychology 35:39-48.

Wittenborn,J.R and Larsen, P.R and Mogil, R.L.1945. ‘An empirical evaluation ofstudy habits for college course in French and Spanish’. Journal of EducationalPsychology 36.8:439-474.

Whiteson,V.1979. ‘A study of proficiency and linguistic variety of non-native EFLteachers’ Ph.D dissertation. Indiana University.

Wong, M.J.1982.‘The Role of Attitudes and Motivation in Second LanguageAcquisition Among Adolescent Native Chinese Speakers in the US’. Ph.Ddissertation. University of San Francisco.

Young, M.Y. and Gardner,R.C.1990.‘Modes of Acculturation and secondlanguage proficiency’. Canadian Journal of Behavioural Sciences 22.1:59-71.

Buy your books fast and straightforward online - at one of world’s

fastest growing online book stores! Environmentally sound due to

Print-on-Demand technologies.

Buy your books online at

www.get-morebooks.com

Kaufen Sie Ihre Bücher schnell und unkompliziert online – auf einer

der am schnellsten wachsenden Buchhandelsplattformen weltweit!

Dank Print-On-Demand umwelt- und ressourcenschonend produzi-

ert.

Bücher schneller online kaufen

www.morebooks.deVDM Verlagsservicegesellschaft mbH

Heinrich-Böcking-Str. 6-8 Telefon: +49 681 3720 174 info@vdm-vsg.deD - 66121 Saarbrücken Telefax: +49 681 3720 1749 www.vdm-vsg.de