Identification of Common EFL Errors in Writings of Kosovar Freshmen

This descriptive study aims to identify the most common English language errors committed by EFL freshmen of the second largest university in Kosova. Data was collected from a corpus of hand-written essays from 210 freshmen in English Language and Literature Department at University of Prizren during the winter semester of 2016/2017 academic year. Data was analyzed using frequency count and percentage distribution. Findings of the study revealed that faulty use of prepositions and present simple tense made up almost half of the total number of errors committed.


Introduction
Identification of learners' errors in English language learning is not a new topic, but one that was studied at length by numerous researchers since 1960's (Corder, 1967;Richards, 1970;Corder, 1981;Dulay, Burt, & Krashen, 1982;Taylor, 1986;Ellis, 1989).Language errors are seen as valuable information not only for researchers seeking information on how second language is acquired, but also for teachers looking into how to improve their language teaching, and students probing on which language areas they need to strengthen (Corder, 1967;Corder, 1981;Dulay, Burt, & Krashen, 1982).Consequently, the wave of error research has spread fast to other countries trying to identify and analyze specific errors committed by learners learning English as a foreign or a second language (Royster, 1913;Earhart & Small, 1915;Politzer & Ramirez, 1973;Lu, 2010;Nezami & Najafi, 2012;Novita, 2014;Prvulović, 2014;Kotsyuk, 2015).However, there is scarce research into errors committed by Albanians (Kaçani, 2013) or Kosovars who learn English as a foreign language (Gërmizaj, 2005) and hardly any specific statistical research.The present study aims to bridge this gap by identifying the most common linguistic errors that Kosovar learners tend to make in written English, without inquiring into what "causes of the deviation might be" (Dulay, Burt, & Krashen, 1982, p.139).In addition, it hopes to provide specific data that can be used to improve teaching practices in the short term and raise the awareness of all the parties involved in the learning process in the long term.

Error definitions
The relevant literature reveals a number of error definitions.Dulay, Burt, and Krashen (1982) defined errors as deviations from the Standard English, while James (1998) defined them as faulty language uses that do not match language that native speakers would use.In addition, researchers have tried to clarify the distinction between errors and mistakes in terms of learners' language abilities.Thus, if students make linguistic faults and can self-correct without external assistance this is considered a mistake and as such they show unsystematic errors in performance; in contrast, if students make language faults but cannot self-correct without additional studying this is considered an error and as such they make systematic errors in competence (Corder, 1967;James, 1998).Sometimes performance errors appeare as a result of carelessness or fatigue, whereas competence errors are always a sign of learners' insufficient knowledge of language rules (Dulay, Burt, & Krashen, 1982).Subsequent studies in second language acquisition influenced how errors were viewed by linguists and researchers and showed that learner errors shouldn't be overlooked and ignored, but valued as factual indicators of an ongoing learning process (Corder, 1967).

Error causes and classification
Consequently, the importance given to errors has propelled research alongside two routes: explaining causes and classifying taxonomies.Beginning 1960s, causes of errors have been analyzed and explained using different theories of which contrastive analysis and error analysis are the most prominent (Dulay, Burt, & Krashen, 1982).The former emphasized the influence of the mother tongue in second language acquisition, wherein the interference of first language is seen as impeding the learning process and not assisting it (Dulay, Burt, & Krashen, 1982).The latter emerged as a reaction to error causes that could not be explained by contrastive approach, and in turn included the aspect of the learner's own idiosyncratic language which is used during the process of learning itself (Corder, 1981;Dulay, Burt, & Krashen, 1982).
In the second route of research, language theoreticians have restlessly tried to cover and explain all the errors committed by learners of different backgrounds (Dulay, Burt, & Krashen, 1982).This route generated the four principal error taxonomies: 1. Linguistic category classified errors alongside the language segments where the errors occurred including phonology, syntax, morphology, lexicon, and discourse components; 2. Surface strategy identified errors based on cognitive processes that learners use to alter the new language namely omitting, adding, misforming, or misordering parts of language; 3. Comparative taxonomy compared second language error structures between second language learners and first language learners and aligned errors in two groups: developmental and interlingual, and 4. Communicative effect taxonomy classified errors based on their effect on communication grouping them into global and local errors (Dulay, Burt, & Krashen, 1982).Similarly, Richards (1970) explains error occurrence on developmental basis which he further groups under four groups: 1. over-generalization where learners merge two grammar structures into one faulty structure, for example, the third person singular ending and irregular past forms of verbs; 2. ignorance of rule restrictions where learners are ignorant of exceptions to rules or limitations to general rules with examples of errors of preposition and article use; 3. incomplete application of rules which include incorrect use of question forms and statement forms, and 4. false concepts hypothesized which includes the "occurrence of structures whose deviancy represents the degree of development of the rules required to produce acceptable utterances" (p.12) which covers incorrect use of question forms.
However, defining and classifying errors properly showed to be an arduous activity.For instance, some errors which are affected by the differences between first and second languages systems are also found in languages that have similar grammar structures such as the case of verb inflection in English, Albanian and Spanish (Corder, 1967;Dulay, Burt, & Krashen, 1982).Furthermore, some errors tend to fall into both over-generalization and ignorance of rule restrictions, while others cannot be precisely grouped under certain categories such as intralingual or developmental (Richards, 1970;Dulay, Burt, & Krashen, 1982).These observed difficulties in describing errors and classifying their causes show that knowing only the differences in learners' first language and second language learning systems might not be enough for errorless acquisition (Corder, 1967), nor might one single theory account for all possible errors that come out of the learning process (Dulay, Burt, & Krashen, 1982).

Error significance
Identifying errors is especially significant in teaching.Results from these studies create a strong foundation for making improvements in teaching, showing gaps in student knowledge that need to be worked on further, and creating a favorable environment for successful acquisition (Corder, 1967).In addition, studies show that language learners make use of conscious and subconscious processes in language learning.These practices affect the quality of learning because they "filter" what new information can enter the mind, then organize and monitor the overall classroom instruction (Dulay, Burt, & Krashen, 1982, p.45).The conscious and subconscious processes have more significance in adult learning than it was previously thought; in fact, research shows that adults make use of inner processes when communicating in a foreign language more than they rely on conscious use of grammar rules in speech (Dulay, Burt, & Krashen, 1982).It is this "linguistic ignorance" of errors that shows how students handle gaps in their knowledge to come with a solution in their language use (James, 1998) and are in control of their own learning process governed also by other factors including personal needs and desires (Ellis, 1989).

Errors across nations
Research shows that errors are widespread and attributed not only to non-natives but also native speakers of English Language.Far from being exhaustive, we are enlisting only a few samples to give a simple outline of errors committed across nations.The most frequently mentioned five error types are verb tenses, subject-verb agreement, prepositions, articles and singular/plural forms of nouns.While natives struggle with errors in subject-verb agreement and adjectiveadverb use (Royster, 1913;Earhart & Small, 1915), foreign language learners, as reported by Poulisse in 1997 (as cited in James, 1998), keep on making the error in the third person singular (s) ending, which is one of the most difficult errors to overcome (Dulay, Burt, & Krashen, 1982); subject-verb agreement is a common error among Filipino learners (Gustilo & Magno, 2012); Ukrainian university student's essays are marked by errors in article use, wrong use of verb forms, incorrect use of singular/plural nouns, possessive's, and errors in use of modifiers and quantifiers (Kotsyuk, 2015); Indonesian freshmen at Andalas University struggle with similar errors in their written essays with the five most frequent ones in verb tenses, subject-verb agreement, article, and prepositions (Novita, 2014); the most common error types in Pakistani learners are verb tenses mainly past and present (Zafar, 2016); Iranian learners struggle with articles, verb forms, singular/plural, prepositions, and tenses (Nezami & Najafi,2012); Albanians make a range of errors in articles and plural forms (Kaçani, 2013); bilingual and monolingual Mexican school children make errors in simple past tense and prepositions (Politzer & Ramirez, 1973); Serbian students have difficulties with determiners, especially articles and pronouns (Prvulović, 2014); Malaysian secondary school children have their share of errors with the five most frequent in singular/plural form, verb tense, word choice, prepositions and subject-verb agreement (Darus, & Subramaniam, 2009), and Turkish learners make errors in preposition use (Tunaz, Muyan, & Muratoğlu, 2016).

Methods
In this study, data was collected from a total of 210 written essays from EFL Kosovar freshmen of Department of English Language and Literature at Prizren University in the academic year 2016/2017.The students were asked to write a oneparagraph essay on the topic of student motivation causes within a time-length of 60 minutes as part of their composition course evaluation process.Additionally, they were explicitly asked to avoid grammar and writing mistakes so that the aspect of performance errors is eliminated (Richards, 1970;Dulay, Burt, & Krashen, 1982).The collected errors are contextualized and semantically related because they are written about a particular topic within a given situation and not just random sentences which would make identification difficult (Corder, 1967).
The methodology in this study followed the same complementary procedures used by other researchers including description, categorization, and descriptive statistics (Politzer & Ramirez, 1973).Within the framework of descriptive taxonomy, data was classified using the linguistic category scheme with an added focus on morphological component (Dulay, Burt, & Krashen, 1982).This scheme prevailed over the other taxonomies because of three reasons: 1. they are an easy reference to both teachers and students, 2. most of textbooks and lessons are developed based on linguistic features of language although in recent years this has been bolstered by context (Dulay, Burt, & Krashen, 1982), so teachers can adapt them easily into their classroom teaching, and 3. in order to avoid challenges that other scholars have faced in attempts to define categories (Politzer & Ramirez, 1973;Dulay, Burt, & Krashen, 1982).Accordingly, errors identified in this study adhere to basic morphological structures (noun, verb, adj…) because of their convenience and userfriendly format, while resort to syntactic terms (subject-verb agreement, modifier…) only when the former is inappropriate.
Data was identified and coded using the comprehensive first and second cycle coding method (Saldana, 2009) focusing only on mostly grammar-based errors including erroneous capitalization and the empty subject (it), and not on acceptability, correctness, strangeness, or infelicity of errors themselves (James, 1998).Errors were extracted from participant's writings as were put on paper, without taking into consideration writer's intentions of what they wanted to write (Taylor, 1986).Therefore, the process of error identification included several considerations: The participants' first choice of words determined correction of the rest of the sentence, for example "..our family spend money.." was classified under incorrect use of present simple tense affirmative third person (our family spends money, rather than our families spend money), "if a student feel good", "we don't have anyone to ask….so we failed sometimes" which was grouped with incorrect present simple negative form rather than incorrect past use, "why people don't learning" was aligned under present simple affirmative and not present continuous tense, or "may will happen yet" was grouped under modals instead of future will.However, when the first word was insufficient to determine the category, meaning of the first part of the sentence was considered as a guiding point such as "I remember my dad always says to me…", or "that's what happening to me lately".Finally, when neither of the two was helpful, the arbitrary decision (Dulay, Burt, & Krashen, 1982) based on logic and reason was followed, for example "after they are getting graduated …won't stay" was placed under present tense rather than present continuous.
Although erroneous sentences were chopped for convenience purposes, researchers gave a careful consideration to chopping process by focusing on the overall meaning of respective sentences/text such as "students leave the city …and going to another one", "..better to work …than going at..".
Participants' personal choice of words and sentence structure was given priority over the researchers, experts' choice of sentence construction, or choice of other linguistics forms, so the researcher focused on "observable [rather than inferred] characteristics" of errors (Dulay, Burt, & Krashen, 1982, p. 146).In this view, we have not included sentences that used a personal choice of words and sentence structure such as "and thinking the fact that they have…" , "as a quote say..", "they are going to have mistakes", "the fault goes to someone else's", "students that are interested to study", "choose a place that they don't belong there", "we can do our future better", "they're poor and they should work instead of study", "they think that could not find …", ", to have lack of knowledge, it's a disaster", "they less their motivation", "we are lack of motivation", "to start tell students,", "start stopping this", and "became who they wanted to be".sentences with ambiguous meaning were not included such as "we cannot find our jobs in a near future", "the reason why is this problem going far is the internet", "in the future".
Once errors were identified and selected, an additional coder1 was asked to code in order to address the validity of the coding process whereby the undecided sentences were disregarded from the research data set.In the first cycle, a preliminary linguistic code was assigned to 19 items, then continued into the second cycle in order to further manage and filter the data so that more manageable categories are formed (Saldana, 2009) and it was decided on the final seven.Finally, descriptive statistics was performed on coded structures for frequency counts of incorrect uses and percentage distribution using Excel.

Results
From the total of 210 student-produced papers, 148 (70.5%) papers contained errors and 62 (29.5%) did not.Total number of errors identified is 406.The highest number of errors was seen in prepositions (21%), and the lowest was recorded in wrong use of present simple questions and past simple negative structure (0%) with only 2 errors, respectively (Figure 3.).The most frequent errors registered were wrong use of prepositions and present simple tense structures, which comprised almost half of all registered errors, 171 (42%) (Figure 1,2.).
Within the category of prepositions, the highest number of errors was registered in wrong use of to 20% and the lowest, while 1%.Similarly, learners made use of the five most frequent prepositions including to, for, in, on, Ø (no need for a preposition) which covered the total of 70% of all prepositions (Figure 4.).

Discussion
A prevalent occurrence of language errors in EFL learners shows that a plethora of studies on causes is slow to show significant changes in its overall goal to improve the present teaching practice.The results of this and other studies confirm three significant points in English language teaching: 1.Studies on error analysis offer a comprehensive, if not exhaustive, tool for narrowing done the list of errors that deserve more attention, 2. there are other factors that influence learning, and 3. errors are inevitable signposts of learning.
Studies have shown that errors are "indicators of progress rather than failure" (Dulay, Burt, & Krashen, 1982, p. 136) in the learning process.In order to help learners in their progress towards successful learning there are several steps that teachers can follow.Knowing that it is learners who control the classroom input (Corder, 1967;Dulay, Burt, & Krashen, 1982) and that learning depends on other factors such as environmental elements and personality traits of the learners (Dulay, Burt, & Krashen, 1982;Corder, 1967;Richards, 1970), teachers can still accomplish their tasks if an extra attention is given to the most frequent errors (Royster, 1913), if there is more contextual and purposeful "input", and if "the idea of individualized approach to language pedagogy" is standardized (Ellis,1989, p.91).On the contrary, learners will make use of other devices to accomplish the given language tasks and end up making errors on the way.we don't have anyone to ask…so we failed sometimes some students is have teacher appreciate them the situation are Kosova have a lot of all of us knows I'm agree with some students likes teachers needs to students who studies one of the students who study here the student have to like as a quote say… even they "graduted" university studies seems to be students leave the city… and going to another one using to much social..they lost a lot of time everyone have everybody dream our teacher tell to us every student have a plan somebody ask a student sacrifice all his life everybody have reasons students needs better if I finish university and I couldn't find a job a lot of students wants one student need to be a good person not reading enough books have it's own reason help others who needs your help information that …help us teacher which explain my friends works in a different jobs this strategy increase my motivation smartphones and technology ..is the main reason not just to hear the lessons that teachers says students who works people who wants to be if anyone have their reasons every students have that this mean everything …the student are not maybe this help and motivate us a student that … want to graduate

Table 1 .
Errors related to prepositions a primary reason of [the] lack of motivation a primary reason of lack motivation

Table 2 .
Errors related to present simple tense of verbs