A Survey of Indonesian Students' Use of Dictionaries

This article aims to define the profile of Indonesian speakers as dictionary users by utilizing a questionnaire-based research. The survey was distributed to university students who were studying English as a foreign language. The students fall into two groups with pre-intermediate and intermediate levels of English proficiency. They were instructed to complete the survey. The question items encourage students to report their consultation frequency of dictionary use, dictionary choices and ratings, and dictionary information needs. The findings confirmed that both pre-intermediate and intermediate students use bilingual dictionaries more often than monolingual dictionaries. The two groups show indifference towards the evaluation of different dictionary types, since learners rate bilingual dictionaries as good as monolingual dictionaries. The collected data offers insights into students' awareness and knowledge of various dictionary types in the Indonesian lexicographic context.


Introduction
From all the specialisations in dictionary research, the so-called user-perspectives have made a significant contribution to facts about user awareness of the availability and usability of various kinds of dictionaries and other reference works (Müller-Spitzer, Domínguez Vázquez, Nied Curcio, Silva Dias and Wolfer 2018). This allows studies to assess dictionary culture in a particular language community to discover whether users are aware of the value and limitations of the dictionaries they use. There has been increasingly important dictionary research which focused on users' needs for and skills in acquiring dictionary information (Atkins and Varantola 1998;Béjoint 1981;Hatherall 1984;Lew 2004;Tomaszczyk 1979). Unfortunately, the majority of the research has been conducted in English-speaking countries. Against this background, research needs to be done into an Indonesian lexicographic context in which Indonesian speakers might have different ideas on dictionary use and its values (Kwary 2018). However, there has only been little empirical data on dictionary user research by Indonesian speakers. With this in mind, we collected information about the Indonesian lexicographic tradition. Accordingly, we found that the government considers dictionaries as systems of information storage (Moeljadi, Kamajaya and Amalia 2017) while few researchers took an interest in users' considerations for using a specific type of dictionary (Amalia 2014; Yanto 2016). The reason for the lack of empirical data to define the profile of Indonesian dictionary users is probably that Indonesian speakers have not prioritized dictionaries as one of the necessary reference tools (Badan Bahasa Upayakan Leksikografi Berkembang di Indonesia 2017). Methodological issues emerge for collecting information about dictionary culture in a community where dictionary use must occur. Therefore, researchers can solve the problems by carefully choosing representatives to conduct the survey, such as university students or learners of the English language. In addition, as suggested by Hartmann (2000: 390), researchers should consider including all lexicographic information to raise dictionary awareness. The collected data will be valuable information since the results could contribute to enhance a dictionary culture among particular language communities.

The study
Several studies examining users' perspectives have explored specific issues, such as user attitudes, behaviours, preferences, difficulties, dictionary misuse, reference needs and reference skills (Nesi 2014). Examples of important studies include a few empirical studies before the 1980s, such as Tomaszczyk's (1979) and Baxter's (1980) research. They reported preferences for bilingual dictionaries by EFL learners and difficulty in understanding monolingual dictionaries. On the contrary, Béjoint (1981) found that most French students of English (96% of them) possessed monolingual dictionaries as recommended by their tutors. There was also a massive project, the EURALEX and AILA Research Project into Dictionary Use (Atkins and Varantola 1998). The study presented how users from four linguistic communities regard different types of dictionaries; thus it was found that the users prefer bilingual dictionaries to help them with particular language tasks. The superiority of the bilingual type was also reported by Polish students in Lew's study (2004). The study found a very low consultation frequency of monolingual dictionaries across levels, except the advanced level. It was not surprising, therefore, to find that the students mostly consulted their dictionaries for meaning, English equivalents and Polish equivalents. Although the study of users' perspectives has made impressive progress, the majority of research tended to be conducted in English-speaking countries. None of the above-mentioned studies contributed to show dictionary culture among Indonesian speakers, especially in the higher education context in which dictionaries become one of the important reference tools to support their learning process.

Methodology
A questionnaire survey was distributed to eighty-eight students at Airlangga University, Indonesia. The participants were in their first year studying English language and literature. The university students were chosen as representatives of the study because dictionaries are seen to be one of the important reference materials to support their learning process. With this sampling method, the results of the study would provide a better understanding on how Indonesian students acknowledge the value and limitations of dictionaries. The students' levels were equivalent to the A2 and B1 levels of the CEFR (Common Framework of Reference for Languages). The two levels are commonly referred to as pre-intermediate and intermediate levels.

The user questionnaire
The students were instructed to complete the user questionnaire that was distributed in a paper format. For reasons of different lexicographic traditions, language skills, and exposure to dictionary types, two researchers might have had different research questions and this might have been reflected in question items and their formats. The questionnaire form (see Appendix 1) is almost self-explanatory. On the top right-hand corner, there was a small space for the student identification number which would provide information about the students' English proficiency levels. The students had been told that their participation in the study would not affect their grades and they are expected to answer all the questions truthfully. There were three sections which consisted of several details.
Section A identifies how often the students use certain dictionary categories (Indonesian-English, English-Indonesian, and English-English). The question item gives a range of frequency options from 6-7 days a week, 5-4 days a week, 3-2 days a week, and 1-0 days a week).
Section B asks the students to rank the first, second, and third dictionaries they often use, then to give ratings (excellent, good, OK, poor, and awful) for each dictionary mentioned. For this item, the survey also provides information about dictionaries that are readily available in Indonesia (see Appendix 2). It was possible for the students to mention other kinds of dictionaries and rejected dictionaries they did not actually use. The question item is in the form of openended questions. The instruction has made it clear that the participants are encouraged to give as many details as they could (e.g. details about type, title, publisher, author, edition, or medium).
Section C asks the participants to report how often they look for certain information in a dictionary. In this question item, the students were given frequency options of always, often, sometimes, seldom and never. The typical presentation and symbol for each lexicographic information type are given in case the participants cannot relate to different kinds of information given in dictionary entries.

Frequency of dictionary use
In item A1 of the user questionnaire, the students were asked to identify how often they consult bilingual dictionaries in the direction of Indonesian-English. A detailed summary of the students' responses is presented in    Table 2, even though there is a tendency of intermediate students to use English-Indonesian dictionaries more often than pre-intermediate students, there is no significant effect of different levels on the frequency of English-Indonesian dictionary consultation. From this fact, it is suggested that the groups show similar consultation frequency when using English-Indonesian dictionaries. In item A3, the students responded to a question on how often they used monolingual English dictionaries.   Table 3, most of pre-intermediate and intermediate students reported using monolingual English dictionaries 0-1 day a week. This means that the two groups rarely use monolingual dictionaries. An interesting finding appears as the use of monolingual English dictionaries did not seem to increase across levels, indeed the consultation frequency seemed lower among intermediate students. The result of the statistical test, however, shows no significant difference between the students' levels. Therefore, this means that none of the groups used monolingual English dictionaries more often than the other. It was expected that the students with a higher level would use monolingual dictionaries more often than those with a lower level of English proficiency. Let us now examine the effect of levels on the consultation frequency. The tendency that emerges from the test results has shown no significant effect (R 2 =4%, p=0.29) of levels on frequency of dictionary use. We can conclude that the two groups of students have similar habits in using dictionaries, in which monolingual dictionaries are rarely used and English-Indonesian dictionaries are consulted as often as Indonesian-English dictionaries.

Dictionary preference
In section B of the user questionnaire, the students were asked to name three dictionaries they used most frequently. We had set the item by providing three possibilities consisting of three different kinds of dictionaries from which to choose to complete it. Moreover, the three possibilities in the questionnaire are provided to emphasize the students' first choice (item B1), second choice (item B2), and third choice (item B3). A detailed number and percentage of students' responses are separately presented in tables for each level of English proficiency. From 73 valid responses collected from pre-intermediate students, 42 preintermediate students reported using at least one dictionary. Half of pre-intermediate students reported using two dictionaries, while 10 students were identified using three dictionaries to help them with their day-to-day learning process. In Table 4, we provide readers with details of dictionaries mentioned by preintermediate students.      Table 7 above, intermediate students prefer to use bilingual dictionaries to monolingual ones. However, the statistical test has shown that there is no significant difference in the consultation frequency of consulting bilingual and monolingual dictionaries. In other words, we can say that among intermediate students they did not seem to prefer one type to the other. An analysis of choices for individual dictionary types above suggests that bilingual dictionaries are consulted more often and are preferred by all of the students as their first choice. However, the effect of students' levels and dictionary choice need to be examined further. From the choices made in the test result, it was found that the effect of dictionary types was not significant, so the effect holds irrespective of the students' levels. From this fact, it was suggested the two groups of students did not favor one type of dictionary over the other. There also appears to be a tendency for the students to use one specific dictionary, that is Oxford.

Dictionary evaluation
In section B of the user questionnaire, the students were instructed to rate the dictionaries they have named before. The ratings were given in ranking order. The ratings are classified as 'excellent', 'good', 'OK', 'poor', and 'awful'. A detailed number of responses are separately presented in tables for each level of pre-intermediate (see Table 8) and intermediate levels (see Table 10). In Table 8 Table 8, the five top rows from the collected data give information about the most consulted dictionary. The mean rating of each dictionary is also given in the table. The smaller the number of dictionaries rated by pre-intermediate students, the less reliable the ratings were. However, the identification of the least consulted dictionary is still given for completeness. From the facts collected and presented in Table 8 above, 17 pre-intermediate students rated the OALD dictionary as 'excellent' and 'good'. The average rating of the OALD is 4.24. This fact complements our previous finding that the OALD is not only preferred, but also highly valued by pre-intermediate students. From the total of 13 pre-intermediate students, nine of them rated Google Translate as 'good'. Google Translate obtained a mean rating of 3.54. The mean ratings received by Kamusku and Kamus Indonesia-Inggris/Inggris-Indonesia are 3.56 http://lexikos.journals.ac.za; https://doi.org/10.5788/30-1-1607 and 4.00 respectively. Kamus Indonesia-Inggris/Kamus Inggris-Indonesia (Echols and Shadily 2014/2017) could only be identified by fewer pre-intermediate students and received an average score of 4.33, which is surprisingly higher than the OALD. However, the OALD's rating is considered to be more reliable than that of Kamus Indonesia-Inggris/Kamus Inggris-Indonesia (Echols and Shadily 2014/2017). The reason for this is that the less variation a dictionary has, the lower the mean error value of it will be.
There is one interesting case in which the semi-bilingual dictionary (i.e. Kamus Password English Learner's Dictionary for Speakers of Bahasa Indonesia) (Kwary 2011) was identified by one of the pre-intermediate students. This finding is quite surprising, as we can see in Yanto's (2016) study that the dictionary was hardly mentioned by Indonesian lecturers in the English department. As there is only one student who could identify the dictionary, this finding is unreliable in which the mean error value could be the highest.
All the dictionaries rated in Table 8 above can be classified into either bilingual or monolingual dictionaries. A summary of the results is given in Table 9 below. The table presents information about the mean ratings, standard errors, and 95% confidence intervals for both bilingual and monolingual dictionaries.  As shown in Table 9 above, although pre-intermediate students claimed to use bilingual dictionaries more than monolingual ones (see Section 4.2), it emerged that monolingual dictionaries received higher ratings than bilingual ones. The statistical result suggests that there was no significant effect of dictionary types on their ratings. The data somewhat confirms the fact that pre-intermediate students not only use both bilingual and monolingual dictionaries but also value both dictionaries. Let us examine the ratings given by intermediate students. There were 67 valid responses in the students' evaluation.   Table 10, the OALD is mostly consulted and highly rated by intermediate students. The dictionary was also rated as 'excellent' and 'good' and receives an average rating of 4.31. Google Translate, which is in the second place as the most consulted dictionary, received a mean score of 3.20.

Kamus Indonesia-Inggris/Kamus Inggris-Indonesia (Echols and Shadily 2014/2017) and
Kamusku obtained average rating scores of 4.30 and 2.70 respectively. There is an interesting finding that can be seen in the breakdown of responses given in Table 10. We can see that the five top lists start with a monolingual dictionary by a specific publisher (i.e. Oxford Advanced Learner's Dictionary) and end with a smaller dictionary by the same publisher (i.e. Oxford Mini Dictionary & Thesaurus). Four intermediate students rated Oxford Mini Dictionary & Thesaurus quite highly, with an average rating of 4.25. This dictionary was not even mentioned by the pre-intermediate students. Although the number is small, this provides more information about the students' awareness of different kinds of dictionaries.
A further analysis has to be done to see whether there is a significant effect of dictionary types on the ratings. Table 11 gives detailed information about mean ratings, standard errors, and 95% confidence intervals for each dictionary type.  Table 11: Intermediate students' ratings of bilingual and monolingual dictionaries As shown in Table 11, intermediate students valued monolingual dictionaries more than bilingual dictionaries. However, there is no significant effect of dictionary types on the evaluation. We can assume that the two groups evaluate both dictionaries similarly.

Reference needs
In section C of the user questionnaire, the students were asked to report how often they consult their dictionaries to look for particular information. We considered including seven main types of dictionary information in the questionnaire. They are meaning, pronunciation, word class, examples, phrases, context, and synonyms. The options of consultation frequency were made to elicit relative frequency.   Table 12, the most consulted information when the students use dictionaries is word meaning. The concept of meaning in the questionnaire referred to English and Indonesian equivalents, as well as an English definition. For this item, we considered that it was not necessary to mention the concept of word meanings separately. Indeed, more than 80% of the students reported to 'always' or 'often' look for meaning. A breakdown of responses for each level is also It is interesting to see that students with a higher level of English proficiency rely on the consultation of meaning in a dictionary, while pre-intermediate students seemed to be confident enough of their vocabulary knowledge to consult dictionaries for meaning. In item C2 of the questionnaire, around 40% of both pre-intermediate and intermediate students indicated that they need to consult dictionaries to pronounce words correctly. This also means that more than 40% of both preintermediate and intermediate students might be confident enough not to consult their dictionaries for the way to pronounce English words. More reasons possibly emerge from the students. In item C3, the students were asked to report how often they consult their dictionaries to find word class information. It is suggested that there was a tendency for pre-intermediate students to choose 'always' and 'often' options. In contrast, intermediate students tend to choose the two lower frequency options, 'seldom' and 'never'. We can thus assume that the frequency of word-class consultation in dictionaries may decrease as students' proficiency levels rise.
In item C4, pre-intermediate students reported lower consultation (>50% of all responses) of example information in dictionary entries. In contrast, more intermediate students reported consulting examples more often. In item C5 and item C6, the students were instructed to report their needs for phrases and context information in dictionaries. The responses from both pre-intermediate and intermediate students have shown that only a few students indicated that they 'always' or 'often' look for phrases and context in their dictionaries. For the last item C7, the students would indicate their consultation frequency of synonym information. It was found that more students in the pre-intermediate level chose the lower frequency options, while more students in the intermediate level chose the higher frequency options. The higher consultation for synonyms can be seen as additional information on word meaning, in addition to the fact that information on meaning is already given in the English definition or in the translation equivalents.
The comparison in the consultation frequency of dictionary information between the two groups of students is presented in Table 13 below. The responses of the students' reference needs are then ranked by mean scores of the consultation frequency.  We further examine the relationship between information types and learners' levels, whether there was a significant effect on the frequency in which information types were consulted. Indeed, the type of information most frequently searched by pre-intermediate and intermediate students is meaning, with mean scores of 4.35 and 4.19 respectively for the two groups. This is followed by pronunciation (3.28) and word class (3.07) for the pre-intermediate students.
In contrast, synonyms are the second most consulted information by intermediate students (3.62), and pronunciation follows with a mean score of 3.17. The test results have shown that there was an increasing need for synonyms, but not for the remaining types of information.

Conclusion
User studies, which are based on the user questionnaire survey, can answer a number of questions that are relevant to the greatest part of practical lexicography. However, for a survey to be maximally useful, researchers need to be really careful about the exact form of the question. They need to tackle the specific questions so that they can present useful information to readers. The results of this study confirm that both pre-intermediate and intermediate students use bilingual dictionaries more often than monolingual dictionaries. The two groups further show indifference when asked to identify dictionary choices and ratings. Moreover, the attempt to bring the lexicographic information to the students proved useful to some extent. It is interesting to see that the students are able to identify language difficulties and use dictionaries to look for particular information.
Although this study is capable of providing the necessary information about the profile of Indonesian speakers as dictionary users, a questionnaire-based study has its own shortcomings for not being able to explain the reasons for users' choices. For this reason, adopting a multi-method or mixed approach would be useful in order to better explain the manner of consultation behaviour in this Indonesian context for the future.