Anxiety contribution on students’ text comprehension in various test types

Received Oct 2, 2019 Revised Nov 7, 2019 Accepted Mar 19, 2020 This study aims to understand how anxiety as a variable contributes to affect the ability in comprehending English texts evaluated on students’ performance on various test types, namely 1) multiple-choice tests (MC), 2) fill-in-the-blank tests (FTB), and 3) true/false tests (TF). This research used ex post facto method, with responses collected from eleventh grade high school students in Jakarta. Respondents worked on tests with similar material, although at different times. Students’ anxiety was measured based on behavioral and cognitive characteristics listed on the questionnaire given to respondents along with the English test. The results pointed out that there was indeed a difference of students’ anxiety levels in working on the various English text comprehension tests given. The results of the English text comprehension tests were generally influenced by the anxiety variable, although the effect was found to be small. Teachers are recommended to apply multiple variations of ability evaluation in measuring the students’ English text comprehension by using multiple test modes.


INTRODUCTION
Language is a tool to deliver one's mind, ideas or feelings. As mentioned by Setiadi, language is a system to express meaning. One may deliver what he feels to others through the use of language. For instance, anger, joy, happiness, sorrow, and so on, can be expressed with language so that others understand what one feels [1].
English plays the role of a pivotal international communication medium and enables someone to create proper and conducive atmosphere for discussion amongst those who learn the language. English is the official language of many commonwealth countries and is understood and widely used around the world. Richards stated that, excepting its role as the mother tongue for 320 million people countries like America, England, Canada, and Australia, English is also used as a second language or is the accepted international language in many countries around the world, including in Asia, with a user base of approximately 390 million people. English is thus one of the most commonly used languages in the world [2].
Hence, English education becomes one of the mandatory curricula for students in Indonesia, especially for high schools in Jakarta. Scrinever believes that English education should be designed to give the chance for the students to improve their skills in using the language, dubbed language skills [3]. These language skills are described as what we do with the language. One's ability in comprehending English texts is among the important factors in learning a language, because realistically, daily activities tend to require ISSN: 2089-9823  Anxiety contribution on students' text comprehension in various test types (Supeno) 249 one to tune in and to understand the point of what another tries to communicate through the extensive use of said language. Thus, English education requires understanding in line with the language skills, especially in the context of learning activity. Comprehending English texts can be learned through learning to read properly and carefully grabbing ideas, points, and general feelings that the writer intends to convey. Understanding the contents of a sentence is not simply knowing the meaning of each and every word in the text, but also understanding the general content of said English text. The students' ability in comprehending English texts can be properly detected through continuous evaluation of their studies. Somadoyo stated that there are factors that influence students' potential text comprehension capability and their interest on reading, such as (a) their predispositions, (b) their family, (c) their cultural background, and (d) their school situations. The inability to properly concentrate may be caused by the aspect of anxiety, which may cause the students to fail to understand the contents of a message or news in an English text [4].
Alexander in Zuchdi mentioned that the factors influencing a student's text comprehension includes (a) reading lesson programs, (b) the student's personality, (c) motivations, and (d) the habits and socioeconomic background of the student. The factors influencing one's reading ability may be internal (e.g. intelligence, age, sex, attitudes, psychological needs, etc.) or external (e.g. adequate fitting reading material, social and economic status, ethnic group, peer influence, parental influence, teachers' influence, television, films, and other media, etc) [5].
A student's true ability can be seen through their performance in a test administered to them. The results, in the form of scores, can be then converted into their grades, which can be then fitted into criteria ranging from excellent to very bad. However, these scores do not necessarily reflect the students' true abilities. The results can also be a product of the students (a) cheating, (b) working in a hurry, (c) making careless or carefree mistakes, (d) underestimating the test's difficulty level, (e) feeling anxious about a particularly difficult item, and so forth.
A study examination as an evaluation platform is known to be diverse. Multiple-choice test, for example, is known to be among the most commonly used type of tests, given that it has certain characteristics: (a) it is quick to correct and to evaluate, (b) it can be used for large amount of students to measure their abilities at once, and (c) it can be processed quickly and also can identify the students' ability levels. Gronlund and Linn stated that multiple-choice tests can be used to measure a student's study progress or to measure a student's ability in critical thinking, from simple to complex thinking as it is seen to be fit with respect to the study materials. Furthermore, Gronlund and Linn also mentioned that there are various types of multiple-choice tests aside from the typical one, such as (a) short answer type, (b) true-false type, and (c) pairing type [6]. The items of multiple-choice tests have certain advantages over other tests. For example, this type of test allows for measuring different variance of learning outputs among students, from the simple to the complex, and its form and function are adaptable to suit the contents of the evaluated learning material. This involves a wide range of applications and special uses, to the point that some standardized tests use the multiple-choice format.
Meanwhile, Sudijono mentioned that multiple-choice test can be classified based on the type of items it uses, such as (a) typical multiple-choice, (b) option-question association, (c) identification of exception, (d) causality analysis, (e) case analysis, (f) quantitative comparisons, (g) dynamic relations, (h) complex-option multiple-choice, and (i) items based on the use of diagrams, pictures, or graphs [7].
The measurement of foreign language anxiety is based on the development and the integration of the idea of psychological anxiety as formulated by Stuart and developed by Horwitz [8]. There are classifications of foreign language anxiety into four scales of Likert measurement, namely Not Anxious, Typical, Anxious, and Panicking. In the other hand, Horwitz classified it into five Likert scales in 33 items, aiming to investigate the students' experience of anxiety in regards to foreign language anxiety in the classroom. He developed this into the Foreign Language Classroom Anxiety Scale (FLCAS). The measurement of foreign language anxiety is based on the students' ability of speech and their attitude, in line with Horwitz's ideas as follows: Foreign language anxiety concerns performance evaluation within an academic and social context. It is useful to draw parallels between the anxieties and three related performance anxieties, they are: 1) communication apprehension, 2) test anxiety, and 3) fear of negative evaluation. Due to its emphasis on interpersonal interactions, the construct of communication apprehension is quite relevant to the conceptualization of foreign language anxiety.
Anindyarini and Supahar mentioned that The anxiety faced by students often relates to their learning problems [9]. Yang affirmed that Findings of the study indicated that students did not perceive or show much more anxiety about spoken English test based on Computer-based Spoken English Test (CBSET). During the test, students perceived less anxiety towards the adoption of computers in spoken English test than that generated from the self-efficacy of their speaking abilities and apprehension of failure in the test [10]. Burgucu affects L2 learning is a barrier that stops learners from performing well on tests and whether this anxiety in turn is associated with test takers' educational background, gender, age and English proficiency or class levels [11].
Stroud and Wee, Looking at anxiety from an identity-based, rather than competence-based, perspective sheds a different light on student behavior in the classroom [12]. Meanwhile, Sokolova and Suplatova analyzed the relationship of socio-biographical variables (gender, language proficiency) and generalized anxiety with the foreign-language learning anxiety of adolescents and young adults [13]. Moreover, this anxiety may affect not only the learners' performance in the foreign-language classroom but also their foreign-language competence in various real-life situations. These results might be useful for language teachers, who are expected to deal with emotions in the classroom and to reduce the negative impact of language learning anxiety upon students' learning and performance.
Noormohamadi, mentioned learners' levels of anxiety and strategy use were analyzed, simply providing a general idea of the negative impact on learners' levels between the two at one moment in time. However, the effect of one (the extent of use of leaning strategies) on the other (language anxiety) was not measured [14]. Feryal affirmed that the application of the Foreign Language Learning Anxiety Scale show that teacher trainees feel more tense and nervous in language classes than in any other classes, it embarrasses them to talk in the class, they feel that their friends speak English better than the others, and while speaking they feel tense and nervous [15].
Macayan, et al. mentioned language learning anxiety negatively predicted task performance in a group speaking task. It was found that higher levels of anxiety translated to poorer performance, while those who had lower anxiety levels performed better in speaking. Thus anxiety positively predicted writing. The more anxious students were about writing the task, the better they performed. Students who exhibited lower levels of anxiety performed poorer in writing [16]. Supeno and Suseno stated English text comprehension tests are generally influenced by the anxiety variable, although the effect is found to be small. Teachers are recommended to use multiple teaching methods, and to teach and to motivate the students to read more. Teachers are also recommended to apply multiple variations of evaluation abilities in measuring the students' English text comprehension by using multiple test modes [17].
This research aims to compare the levels of student's anxiety in working on various English text comprehension tests, and to identify the degree of contribution of student's anxiety to the normalcy of their performance in English text comprehension ability through the results of multiple-choice (MC), fill-in-theblanks (FTB), and true/false (TF) tests.

RESEARCH METHOD
The research employed ex post facto method, it was taking the data of the students' test results while also giving them relevant questionnaires with the research instruments. The collected data will be analyzed to discover the contribution of anxiety in affecting the students' performance in various multiple-choice English text comprehension tests on eleventh grade high school students in East Jakarta.
The research was carried out for five months, started from October 2018 and ended in February 2019. The research objects are the national high schools SMAN 93, SMAN 50, and SMA Pusaka in East Jakarta. The sampling method used in this research was cluster random sampling, resulting in the said three high schools. In total, 125 students were chosen as the representative samples used in this research.
The data was quantitative in nature. The data was obtained from the questionnaire that students filled in about their anxiety in working on English text comprehension in several test varieties, namely multiple-choice (MC), fill-in-the-blank (FTB), and true/false (TF) tests. Meanwhile, the anxiety data was obtained from the questionnaire distributed in the end of the English text comprehension tests given out by the class teacher.
The quantitative data was gathered from the students' objective tests which was then interpreted with descriptive analysis to obtain information of intergroup variance from the students from different schools for testing the learning model. The results were then sorted as such that we can infer all information in corresponds to the influence of student's anxiety in working on various English text comprehension tests.

RESULTS AND ANALYSIS
The results of the descriptive data analysis show that there is indeed variance in students' anxiety in working on English text comprehension tests in MC, FTB, and TF tests, shown in Table 1 The summary above points out that there is a variance of students' anxiety score means in working on the various English text comprehension tests. However, the differences are not significant. All three test results (MC, FTB, and TF) lie within the range of 78-79, which can signify the high levels of anxiety. This indicator is seen through analyzing the students' answers of how they respond towards items, such as a) feeling uncomfortable about working on English tests, b) feeling nervous, c) feeling uneasy after working on a test, d) fearing bad grades or failing the test, e) perspiring more than usual, and f) not entirely focusing in working on the English test.
These results show consistency with the proposition expressed by Budiman, who stated that anxiety is a normal reaction to particularly pressing situations in one's life. Aside from being a normal reaction, anxiety is also a psychological or mental symptom in preventive measures against unexpected or undesired things or events, and the discomfort in one's emotional condition is marked by other psychological signs, such as fear, worry, anxiety, nervousness, lack of self-esteem, all the way to the higher levels, which are stress, depression, and panic [18].
Along the same lines, Ramsay stated that anxiety is a feeling of fear, dread, or uneasiness, and some anxious people suffer from conditions, phobias, stress, sadness, panic, and depression [19]. Anxiety is the feeling of fear and nervousness, and people with anxiety also tend to suffer from conditions like phobia, stress, sorrow, panic, and depression. Mardhatillah used the foreign language anxiety scale with indicators for cognitive, affective, and psychomotor manifestations; mental stress, difficulties on concentrating, confusion in answering test items, and excessive fear and nervousness; and fear of negative evaluation with indicators for anxiousness due to assumptions regarding the social situation and being judged by others, not gaining the approval of others, doing something embarrassing in public, being criticized, and being unsupported or rejected [20].
The data distribution of students' anxiety in working on English text comprehension tests can be seen on the figures below.
Meanwhile  Among the three tests, the TF variant scores the highest mean score at 71.85 and a standard deviation of 8.677, which means that students performed better on TF compared to on MC and on FTB tests. The contribution of anxiety variable in influencing the results of text comprehension TF is known to be 9.4%. This means that the remaining of 94.6% is influenced by other variables than anxiety.
This result is in line with the type of test that TF is, which restricts the number of options to whether the item's statement is true or false. The two options leave no space for other items as potential distractions. Students are demanded to be able to carefully reading the sentence and to comprehend the meaning of each word and of the entire sentences. Distraction factor in TF tests can only be placed on the choice of words that may be ambiguous in nature, but when the students understand the context the language conveys, it becomes easier to decide the appropriate answer. Students' mean score on the FTB test is 69.72 with a standard deviation of 19.09, which means that FTB results have the most variance of score distribution among the three test types. The amount of influence that anxiety has over English text comprehension FTB test can be considered little, by only as much as 5.6%. In other words, approximately 94.4% is influenced by other variables than anxiety. The FTB results indicate adequate scores as expected by English teachers, where the minimum requirement to pass is 70. The characteristics of an FTB test make the students read the English text and sequence of sentences more carefully, because the answer to the question is provided in the text. This may contribute to the lesser anxiety detected in this test type.
In a research, Mardhatillah stated that foreign language anxiety arises because of psychological factor, namely lower self-efficacy in comparison to one's actual ability. Furthermore, it is mentioned that there is a negative relationship between self-efficacy and foreign language anxiety. In other words, the higher the self-efficacy, the lower a student's foreign language anxiety is; vice versa, the lower the self-efficacy, the higher and the foreign language anxiety. Mardhatillah concluded that self-efficacy is a factor that influences foreign language anxiety [20].
On the MC type test, the mean score is 62.56 with a standard deviation of 16.75, meaning that the MC type test has a variance in students' score distribution. The influence of anxiety to the MC type test results is found to be 10.6%, or that 89.4% is influenced by other variables aside from anxiety.
The characteristics of the MC test include several options to select as the answer and that there is also the element of distraction options, which may influence students to lose their concentration due to anxiety over their own abilities, which may hinder them from performing optimally and obtaining proper grades. Suseno explained that the characteristics of the MC test allow for multiple different options that have a certain degree of similarity to be fitted into the item, making it harder to find the proper answer to the question because some options may seem 'almost right' [21]. Although the influence of anxiety to the test results is relatively small, this presses the role of the teachers to give the students more confidence in working on the MC tests.
In an attempt to reduce nervousness and fear when students face a classroom language situation, the technique commonly used to relax is to take one or two deep breaths, which then leads to the students collecting themselves and staying calm. Sembodo explained by being relaxed, it is likelier for the students to be able to function cognitively, and therefore to perform in a more controlled and structured way [22].
Said anxiety can cause the students to go through a particularly bad experience in a classroom foreign language use, or to cause them to worry about their seeming lack of knowledge and fluency in line with the language, which means that language anxiety itself can influence achievements in learning, and inadequate ability can cause anxiety, worry, fear, and loss of concentration. An individual will fear making mistakes and fear becoming the butt of a joke to the audience who see them using a second language. This echoes the statement by Spielberger that foreign language anxiety is a feeling that arises within an individual in the form of tautness, fear, nervousness, and worry, which were all connected to the autonomic nervous system. One of the factors influencing anxiety is, indeed, self-efficacy [23].
Clarke explained that reading is connected to lexical understanding and content understanding. Lexical understanding refers to the process of how one recognizes the written symbols and translates them to the spoken language. Content understanding defines as the understanding of how the words and sentences are connected to become a readable text [24].
Citravelu explained that there are several things to note when reading, such as a) reading requires a set of understanding about the rules and conventions of reading the language, b) reading requires understanding of meaning and messages contained within the text, c) understanding of the text requires understanding of the language used in writing the text, d) reading is an active cognitive process, because while reading, one also guesses, predicts, and makes appropriate decisions, e) reading is an interactive process, f) reading is a system directly related to life needs, g) reading is not a single skill, but a set of separate multiple skills used on different types of texts with different purposes, h) a wide array of experience on different types of texts will help someone easily understand the text he reads [25].
Sembodo explained arguments about why anxiety cases tend to happen in the classroom of second or foreign language is put forth by Guiora, who made it clear that language learning is a worrying psychological problem because it directly threatens one's concept of self and worldviews [22]. The anxious students tend to believe that learning a language is always difficult, so they tend to feel down and are incapable when learning a second or foreign language. This assumption is partly influenced by a history of frustration in language learning, which in turn would drive the students into difficulty in learning language. Apart from the possibility that anxiety in testing also contributes to when they take the test, their low grades show that the materials taught are not absorbed and processed.
Sparks and Ganschow affirmed that foreign language educators have to address practical implications of anxiety in the classroom instruction. If anxiety is a consequence of weak language skill rather ISSN: 2089-9823  Anxiety contribution on students' text comprehension in various test types (Supeno) 253 than a cause of poor performance in foreign language classes, then classroom teachers will need to address these language issues as a primary focus of instruction prior to or simultaneously with efforts to address students' anxiety [26]. The most rational alternative solution to minimize the amount of anxiety the students go through can be solved by creating a fun learning method, by using more varieties to evaluate their English text comprehension ability, and by encouraging the teachers to teach the students to read more as a hobby to pique interest, as indicated by the activity framework of AIDA (Attention, Interest, Desire, Action). When the AIDA principle can be sustained and continuously used, then the students' results will be optimal. Supeno, Suseno, & Alhamidi mentioned in terms of increasing students' English text comprehension ability, alternatives such as reading strategies can be used in order to stimulate cognitive growth through six steps, which are fully understanding a text, understanding the meaning within context, to think and to deduce the meaning, to contextually seek the meaning, to further investigate the meanings, and to return to thoughts of lexical meaning) [27].

CONCLUSION
There is a difference of anxiety levels on English text comprehension test results in the MC, FTB, and TF variants. The mean influence of anxiety towards the comprehension test results is generally small. Analysis of the results of the three test types find that the influence is the smallest on the TF variant, followed by the FTB and then the MC type.
Student's anxieties in the studies are marked by nervousness, worry, and baseless fear that undesired or unexpected things would happen. The anxiety levels among students can be slowly minimized through fun learning, increasing the number of variants for English ability evaluations by utilizing multiple test types, and by increasing teacher's performance quality so that they can fulfill the demand to incite interest among students to read, as is dictated by the AIDA framework of Attention, Interest, Desire, and Action.