Evaluating lists of high frequency words

Evaluating lists of high-frequency words Thi Ngoc Yen Dang and Stuart Webb Vietnam National University, Hanoi University of Languages & International Studies / University of Western Ontario This study compared the lexical coverage provided by four wordlists [West’s (1953) General Service List (GSL), Nation’s (2006) most frequent 2,000 British National Corpus word families (BNC2000), Nation’s (2012) most frequent 2,000 British National Corpus and Corpus of Contemporary American-English word families (BNC/COCA2000), and Brezina and Gablasova’s (2015) New-GSL list] in 18 corpora The comparison revealed that the headwords in the BNC/ COCA2000 tended to provide the greatest average coverage However, when the coverage of the most frequent 1,000, 1,500, and 1,996 headwords in the lists was compared, the New-GSL provided the highest coverage The GSL had the worst performance using both criteria Pedagogical and methodological implications related to second language (L2) vocabulary learning and teaching are discussed in detail Keywords: high frequency word lists, corpus study, lexical coverage, headwords Introduction If 95% coverage of a text is considered as reasonable comprehension (Laufer, 1989), L2 learners need to have a vocabulary size of 2,000–3,000 word families (Nation, 2006; Van Zeeland & Schmitt, 2013; Webb & Rodgers, 2009a, 2009b) for comprehension of spoken text, and a vocabulary size of 4,000–5,000 words for written text (Laufer & Ravenhorst-Kalovski, 2010; Nation, 2006) Despite these lexical demands, a reasonable proportion of L2 learners in different contexts are failing to learn the most frequent 2,000 and even the most frequent 1,000 words families after many years of formal English language instruction (e.g., Henriksen & Danelund, 2015; Nurweni & Read, 1999; Webb & Chang, 2012) Generally, L2 learners have less exposure to the target language than L1 children (Ellis, 2002; Nation, 2001) Therefore, learning general service words; that is, the words that occur frequently in a wide range of text types (Nation, 2001; Nation & Hwang, 1995) ITL – International Journal of Applied Linguistics 167:2 (2016), 132–158.  doi 10.1075/itl.167.2.02dan issn 0019–0829 / e-issn 1783–1490 © John Benjamins Publishing Company Evaluating lists of high-frequency words 133 offers theses learners a good return for their learning effort The size of this group of words is relatively small, but they cover a large number of words in different kinds of texts (Zipf, 1949) Focusing on these words ensures that the words that are most likely to be encountered and needed for communication will be learned (Nation & Waring, 1997) Also, knowledge of general service vocabulary will provide a firm foundation for further vocabulary learning Because of their pedagogical value, it has been suggested that general service words should be the initial vocabulary learned by L2 learners (Nation, 2013; Schmitt, 2010) West’s (1953) GSL is the oldest and most influential high-frequency word list However, researchers have questioned the suitability of the GSL for L2 learning purposes due to its age, and have suggested that it be replaced by a more current list (Richards, 1974; Schmitt, 2010) Several lists have recently been developed and might serve as a general service list Nevertheless, it is not clear which list is the best because there has been no research explicitly comparing these lists To fill this gap, this study aims to compare the lexical coverage provided by the GSL and three current wordlists [Nation’s (2006) BNC2000; Nation’s (2012) BNC/COCA2000, and Brezina and Gablasova’s (2015) New-GSL] in a wide range of corpora Lexical coverage is the percentage of words in a text covered by items from a particular word list (Nation & Waring, 1997) It is an important indicator of comprehension (Laufer, 1989; Schmitt, Jiang, & Grabe, 2011; Van Zeeland & Schmitt, 2013) and can reveal the proportion of vocabulary that would be known in a text if a word list is learned Therefore, lexical coverage is the primary criterion for evaluating wordlists By comparing the lexical coverage provided by each list, this study might indicate which list is best suited for L2 learning purposes 1.1 Existing high-frequency word lists There are a number of available high-frequency wordlists Comparing every wordlist is beyond the scope of a single study Analysis of established lists that were developed from large corpora using precise and valid methodologies may provide a reliable list that will serve as the vocabulary foundation for L2 learners West’s (1953) GSL was chosen because it is the oldest and most influential highfrequency wordlist Nation’s (2006) BNC2000, Nation’s (2012) BNC/COCA2000, and Brezina and Gablasova’s (2015) New-GSL were chosen because these lists have been created recently Earlier studies have shown that the BNC2000 and New-GSL provided higher lexical coverage than the GSL (Brezina & Gablasova, 2015; Gilner & Morales, 2008; Nation, 2004) The BNC/COCA2000 was chosen because it is the updated version of the BNC2000 and is expected to provide higher lexical coverage than the BNC2000 Although there are no studies explicitly compared the BNC/COCA2000 with the GSL, as the updated version of the 134 Thi Ngoc Yen Dang and Stuart Webb BNC2000, the BNC/COCA2000 is expected to provide higher lexical coverage than the GSL Apart from the four high-frequency word lists, there is another high-frequency word list that was recently created, Browne’s (2013) New General Service List (NGSL) It was not used in the present study for two reasons First, no precise description has been provided about the cut-off points of the statistical criteria (frequency and dispersion) that were used to select the NGSL words Second, preliminary analysis with nine spoken and nine written corpora in this study shows that the average coverage per item (multiplied by 1,000) provided by the NGSL headwords ranged from 19.68% to 25.76% These coverage figures were much lower than those provided by any of the four lists in the present study (24.11%–34.55%) West’s (1953) GSL contains 2,000 word families The list was developed from a five million running-word written corpus.1 Although frequency was an important criterion in selecting the GSL words, five other criteria (ease of learning, necessity, cover, stylistic level, and emotional neutrality) were also used in selection so that the GSL would consist of words suitable for L2 learning purposes Research has shown that the GSL provided lexical coverage of 72%–90% in a wide range of text types For example, the GSL covers 76.1% of the words in academic writing (Coxhead, 2000), 85.49% of academic speech (Dang & Webb, 2014), 87.1% of fiction (Nation, 2004), and 89.6% of general conversation (Nation, 2004) Because of its impressive lexical coverage, the status of the GSL has been long-established, and it has had a huge influence on L2 learning and teaching practice as well as vocabulary research The GSL has been suggested as the starting point for L2 vocabulary learning and has been widely used as the basis for early graded reader schemes (Nation, 2004) It has been used in the construction of Nation’s (1983) and Schmitt, Schmitt and Clapham’s (2001) Vocabulary Levels Tests and specialized wordlists (e.g., Coxhead, 2000; Coxhead & Hirsh, 2007) as well as numerous L2 vocabulary studies The GSL has three limitations First, the list might not accurately reflect current vocabulary because it is based on texts collected from the 1930s (Carter & McCarthy, 1988; Richards, 1974) Second, it is biased towards written English (Carter & McCarthy, 1988) Third, it is criticized for the low coverage provided by items beyond the 1,000 word level (Engels, 1968) Therefore, researchers have suggested that the GSL should be either revised or replaced by lists which represent more current vocabulary (Richards, 1974; Schmitt, 2010) One candidate to replace the GSL is Nation’s (2006) BNC2000 The BNC2000 words are the most frequent 2,000 word families from Nation’s (2006) 14 BNC 1.  The frequency of some GSL words was estimated frequency, which was calculated by doubling the actual frequency of the word in a 2.5 million running word corpus Evaluating lists of high-frequency words 135 lists The BNC2000 was derived from the 100 million running-word BNC corpus 10% of the BNC corpus was from spoken sources and 90% was from written sources Three criteria were used to select the BNC2000 words: frequency, range and dispersion However, some subjective judgments were also made to minimize the bias caused by the formal, written and adult nature of the BNC For example, common spoken words (e.g., goodbye, ok, and oh), weekdays, months, numbers, letters, and names of countries were also included in the BNC2000 although they not have high frequency in the corpus The strength of the BNC2000 lies in the fact that it provides better coverage than the GSL The BNC2000 had higher coverage than the GSL in research comparing the two lists (Gilner & Morales, 2008; Nation, 2004) Research on vocabulary load and opportunities for learning (e.g., Nation, 2006) also indicates that the BNC2000 provides relatively high coverage (81.03%–96.73%) in different corpora This is much higher than the coverage achieved by the GSL (71.52%–89.6%) However, although attempts were made to include words in the BNC2000 that are common in general spoken English, the list was developed solely from the BNC and is inevitably affected by the British, adult, formal, written nature of this corpus (Nation, 2004; 2012) A second list which might serve as a more current GSL is Nation’s (2012) BNC/COCA2000 Nation created the BNC/COCA2000 from a corpus consisting of six million running-words from spoken sources and four million from written sources To ensure that the list was suitable for young L2 learners, the spoken samples were taken from spoken English, movies, and TV programs while the written samples were taken from texts for young children and fiction To avoid the bias toward British-English, Nation also included materials from American-English and New Zealand-English in his corpus Although frequency and range were important criteria in selecting the BNC/COCA2000 words, Nation also included in the BNC/COCA2000 words that have lower frequency but may be useful in L2 learning context For example, very common spoken words (e.g alright, pardon and hello), numbers, weekdays and months were added to the list although their frequency in the corpus was not very high Because the BNC/COCA2000 is quite new, it has not been explicitly compared with other lists However, its teachingoriented purpose and its derivation from a corpus with a balance between spoken and written texts from different sources, and different varieties of English suggest that the BNC/COCA2000 may be a useful list for L2 learning purposes A third list that was recently developed with an aim to replace the GSL is Brezina and Gablasova’s (2015) New-GSL The New-GSL has 2,494 lemmas It was created from four corpora (LOB, BNC, BE06, and EnTenTen12), which comprise a total size of around 12 billion running-words There are two main differences between the New-GSL and the other three lists First, unlike the GSL, BNC2000 and 136 Thi Ngoc Yen Dang and Stuart Webb BNC/COCA2000, the New-GSL was created from a purely quantitative approach Three criteria (frequency, dispersion, and distribution across language corpora) were used to select the New-GSL words Another difference between the NewGSL and the other three lists is the unit of counting The New-GSL used lemmas as the unit of counting whereas the GSL, BNC2000, BNC/COCA2000 used word families A lemma includes a headword (allow) and its inflections (allowed, allowing, allows) A word family consists of a headword (allow), its inflections (allowed, allowing, allows), and closely related derivations (allowance, allowances, allowable) While the lemma distinguishes between word classes, the word family does not The New-GSL has two strengths First, the total size of the corpora used to create the New-GSL is larger than the corpora used to create the other lists Second, it is divided into two main parts: core vocabulary and current vocabulary, which enables teachers and learners to see the change of general vocabulary over time However, the New-GSL also has two limitations First, it may be biased towards British, written English Three out of the four corpora (LOB, BNC and BE06), on which the New-GSL was based, represented British-English, and three out of the four corpora (LOB, BE06, and EnTenTen12) were made up of written discourse In the only corpus which included spoken English (BNC), spoken samples accounted for only 10% Second, the New-GSL was developed from a purely quantitative approach and may not include words that are not very high in frequency in written language but seem to be useful for L2 learning purposes such as hey, hi, and ok 1.2 Previous research on comparing high-frequency word lists To the best of our knowledge, there are four studies that explicitly evaluate highfrequency word lists: Nation and Hwang (1995), Nation (2004), Gilner and Morales (2008), and Brezina and Gablasova (2015) All of the studies used lexical coverage as the criterion to compare the GSL with other high-frequency wordlists Nation and Hwang (1995), Nation (2004), and Gilner and Morales (2008) found that the GSL did not provide as much coverage as the other lists In contrast, Brezina and Gablasova (2015) found that the GSL provided higher coverage (84.1%; 82%; 80.6%) than the New-GSL (81.7%; 80.3%; 80.1%) in the LOB, BNC, and BE06 corpora, but lower coverage (80.1%) than the New-GSL (80.4%) in the EnTenTen12 corpus However, this may be because the New-GSL had far fewer lemmas (2,494 lemmas) than the lemmatized version of the GSL (4,114 lemmas) (Brezina & Gablasova, 2015) Despite their valuable findings, these studies have a number of limitations First, they compared the GSL with only one or two general high-frequency word lists No studies explicitly compared a larger number of lists Second, except for Gilner and Morales (2008), all studies used the corpora from which the wordlists Evaluating lists of high-frequency words 137 were developed to evaluate the lists Nation and Hwang (1995) used the LOB corpus, the corpus from which the LOB high-frequency wordlist was derived Nation (2004) used Coxhead’s (2000) academic corpus, the corpus from which the AWL was developed, to compare the GSL plus AWL and the BNC3000 Brezina and Gablasova (2015) used the LOB, BNC, BE06, and EnTenTen12, the four corpora from which the New-GSL was created This is problematic, because for a valid comparison, the corpora used to examine the lexical coverage provided by the lists must be different from the corpora from which the lists were developed (Coxhead, 2000) Only Nation (2004) avoided this limitation by using three other corpora apart from the corpus on which the list was based Third, the number of corpora used in these studies was very small: one corpus (Gilner & Morales, 2008; Nation & Hwang, 1995), three corpora (Nation, 2004), and four corpora (Brezina & Gablasova, 2015) Fourth, most of these corpora were written Of those including spoken materials in the comparison (Brezina & Gablasova, 2015; Gilner & Morales, 2008; Nation, 2004), spoken materials accounted for much smaller proportion than written materials The present study follows Nation’s (2004) approach by using other corpora together with the corpora from which the GSL, BNC2000, BNC/COCA2000, and New-GSL were developed in the comparison However, unlike Nation (2004), this study is based on a larger number of corpora (18 corpora), which vary in terms of discourse type, size, and variety of English (Tables 1–2) This is because a general service list is expected to have consistently large lexical coverage in a wide range of texts Using a large number of corpora with a great degree of variety should provide a more accurate picture of the relative value of different lists 1.3 Dealing with the unit of counting One issue when using lexical coverage in comparisons between wordlists is the unit of counting Ideally, the same unit of counting should be used by all researchers to make it possible to compare the results of different studies (Schmitt, 2010) Most earlier studies (e.g., Gilner & Morales, 2008; Nation, 2004; Nation & Hwang, 1995) compared lists using word families as the unit of counting Thus, they did not have to deal with a difference between units of counting However, the unit of counting has varied among a number of recent studies that involved creating different types of wordlists (e.g., Brezina & Gablasova, 2015; Gardner & Davies, 2014) This makes valid comparisons between lists with different units of counting a real challenge, because different definitions of a word may influence the results of corpus-based vocabulary studies (Gardner, 2007) There are four ways to solve this problem 138 Thi Ngoc Yen Dang and Stuart Webb The first option is to compare lists in their original format That is, no changes are made in terms of unit of counting However, this option is not satisfactory because it disadvantages lists that use a smaller unit of counting For example, lemmas have fewer members than word families Thus, lemma lists should have lower coverage than word family lists This is supported by the results of an analysis of the overall coverage provided by one lemma list (New-GSL) and three word family lists (GSL, BNC2000, BNC/COCA2000) (see supplementary information) Although the New-GSL had more headwords (2,228)2 than the three word family lists (2,168; 1,996; 2,000),3 it provided lower overall coverage than the GSL in 13/18 corpora, the BNC2000 in 15/18 corpora, and the BNC/COCA2000 in 16/18 corpora The second option is to convert lemmas from lemma lists into word families so that word families will be the unit of counting in the comparison However, converting lemmas into word families will overestimate the benefit of the lists For example, from the lemma approach, if learners know allow, they may recognize allowed, allowing and allows, and may not recognize derived forms such as allowance, allowances and allowable Hence, when calculating the coverage provided by the word family allow, coverage provided by allowance, allowances and allowable should not be counted because they not belong to learners’ vocabulary repertoire However, converting lemmas in the lemma lists into word families means that allowance, allowances and allowable will all be counted This then conflicts with the principle that guides the lemma approach; that is, learners may not recognize derived forms Option 2, therefore, is not reasonable The third option is to convert word families from the word family lists into lemmas so that lemmas will be the unit of counting in the comparison However, converting word families into lemmas should favor lemma lists over word family lists A word family is made up of a number of lemmas Some of them are frequent while some are infrequent For example, the word family allow is made up of three lemmas: allow, allowance and allowable In the Wellington Corpus of Spoken New Zealand-English (WSC), the lemma allow occurred very frequently (freq = 165) while allowance (freq = 32) and allowable (freq = 4) occurred very infrequently The rationale behind including infrequent lemmas in word family lists is that if learners know one member of the word family, they may recognize other members even if these members not occur very frequently Therefore, if we convert word families into lemmas, a lemma list, which has frequent lemmas only (allow), will have advantages over a lemma list converted from word families, which includes both 2, 3.  Explanation of the number of items in each list is presented in the Methodology Evaluating lists of high-frequency words 139 frequent (allow) and infrequent (allowance, allowable) lemmas Hence, Option may not provide a valid comparison A final option is to compare headwords from different lists That is, inflected forms and derived forms are not counted unless they are headwords Using headwords has four advantages First, it minimizes the difference between the numbers of items in each list because members of the word families/ lemmas will not be included in the comparison For example, in this study the GSL had 2,168 headwords with 11,283 family members, while the New-GSL had 2,228 headwords with only 3,214 lemma members If only headwords are used for the comparison, it will be fairer because the number of items in each list will be around 2,000 Second, using headwords still ensures that the nature of the lists does not change significantly because headwords are usually the most frequent members in word families or lemmas However, it should be noted that because family and lemma members are not included, coverage will always be less than 100% This might be seen as a disadvantage of this approach Third, using headwords also reflects the nature of L2 teaching and learning That is, L2 teachers and learners usually receive lists of headwords without their inflections and derivations, and thus choose headwords to teach and learn first Moreover, they may never focus at all on lemmas and family members Fourth, evaluating headwords also reflects the approach of wordlist creators That is, from the lemma approach, if learners know one lemma member (usually the lemma headword), they may recognize other members while from the word family approach, if learners know one word family member (usually the word family headword), they may recognize other members There is no perfect way of comparison; however, the advantages of using headwords for comparisons between lists outweigh the disadvantages Therefore, it may be the most valid approach to evaluate different wordlists and is the approach used in this study 1.4 Comparing lexical coverage There have been three ways used to compare the lexical coverage provided by different wordlists: overall coverage, average coverage, and the coverage provided by the most frequent items Most earlier studies (e.g., Brezina & Gablasova, 2015; Gilner & Morales, 2008) used overall coverage as the criterion for comparison However, this may not provide a valid comparison for two reasons First, using overall coverage will favor lists with larger units of counting (e.g., word family) over lists that use a smaller unit of counting (e.g., lemmas) Second, even if the same unit of counting is used in the comparison, using overall coverage will favor longer lists For example, the 2,168 GSL word families provide overall coverage of 88.33% in the WSC, whereas the 1,996 word families from the BNC2000 provided overall coverage of 87.62% 140 Thi Ngoc Yen Dang and Stuart Webb Two other ways to compare wordlists are to use average coverage and coverage provided by the most frequent items Nation and Hwang (1995) used average coverage provided by each 100 word families to compare the GSL with other high-frequency word lists Nation (2004) and Gardner and Davies (2014) used coverage provided by the most frequent items (i.e excluding the lowest frequency items from the lists so that each list had the same number of items) to compare wordlists in their studies Each method has its strengths and weaknesses Average coverage is a useful way to compare lists having different number of items Thus, it is able to evaluate lists as a whole However, average coverage does not provide information about the relative value of one item in comparison with other items in a list Moreover, it favors shorter lists because the extra items in longer lists are likely to be the least frequent items In contrast, coverage provided by the most frequent items in lists indicates the relative value of the words in the lists This is useful because lists may be made of very good items and relatively weak items in terms of lexical coverage, and so looking at the most frequent items provides a picture of how the best (and worst) items compare between lists However, looking at the most frequent items favors lists with larger numbers of items because a larger number of infrequent items are excluded Therefore, it may not provide a valid comparison of the lists as a whole The strengths and weaknesses of the two approaches can be illustrated by the performance of the GSL and BNC2000 headwords in the WSC When the average coverage provided by each headword is compared, the GSL provides lower average coverage (.02979%) than the 1,996 headword BNC2000 (.03201%) This indicates that, as a whole, the BNC2000 is superior to the GSL in terms of lexical coverage However, when the coverage provided by the most frequent 1,996 headwords is compared, the GSL provides better coverage (64.59%) than the BNC2000 (63.90%) This is because the GSL had more infrequent headwords excluded (172) than the BNC2000 (0) This suggests that average coverage and coverage provided by the most frequent items on their own may not provide a thorough evaluation of the lists; however, together they may provide a robust assessment of wordlists In this study, both average coverage and coverage provided by the most frequent items are used to evaluate the relative value of wordlists 1.5 Research questions This study will address the following two research questions: Which list of headwords [West’s (1953) GSL, Nation’s (2006) BNC2000, Nation’s (2012) BNC/COCA2000, Brezina and Gablasova’s (2015) New-GSL] provides the highest average coverage in spoken and written discourse? Evaluating lists of high-frequency words 141 Which list has the highest coverage provided by the most frequent 1,000, 1,500 and 1,996 headwords in spoken and written discourse? Methodology 2.1 The wordlists Four wordlists were used in this study: the GSL headwords, the BNC2000 headwords, the BNC/COCA2000 headwords, and the New-GSL headwords The GSL, BNC2000, and BNC/COCA2000 were downloaded from Paul Nation’s website The New-GSL was downloaded from the online Supplementary Data of Applied Linguistics Journal The headwords in the GSL, BNC2000, and BNC/COCA2000 were checked for consistency Four items in the BNC2000 were deleted because they appeared as members of headwords in the other lists As a result, the number of headwords in the GSL, BNC2000, BNC/COCA2000 was 2,168, 1,996, and 2,000, respectively Unlike the traditional definition, in the present study, lemma is defined as a word form (headword) and its inflections without word class distinction For example, from the traditional approach, form (verb) and form (noun) were counted as two separate lemmas, but in the present study they will be counted as one lemma (form) This is because pedagogically, word forms are more important for beginners than word classes (Nation, 2013) Therefore, 266 out of 2,494 lemma headwords in the original version of the New-GSL were excluded because they shared the same forms with other headwords in the list As a result, the New-GSL version used in this study had 2,228 headwords 2.2 The corpora Eighteen corpora were used in this study (Tables and 2) These corpora were in the form of untagged text files The number of tokens in each corpus ranged from 320,496 to 10,484,320 in the spoken corpora and from 1,011,760 to 87,602,389 in the written corpora These corpora represented a wide range of spoken and written discourse and 10 different varieties of English (American-English, British-English, Canadian-English, Hong Kong-English, Indian-English, Irish-English, JamaicanEnglish, New Zealand-English, Filipino-English and Singapore-English) Moreover, there is a good balance between the number of spoken and written corpora The purpose of high-frequency word lists is to provide L2 learners with a solid foundation of lexical knowledge so that they can effectively communicate in English in diverse spoken and written contexts where English is used as an L1, L2 144 Thi Ngoc Yen Dang and Stuart Webb To determine the coverage provided by the most frequent 1,000, 1,500, 1,996 headwords in the GSL, BNC2000, BNC/COCA2000, and New-GSL, headwords in each list were first ranked according to their frequency in different corpora Then, the coverage provided by each set of the most frequent 1,000, 1,500, 1,996 items from each list in each corpus was calculated by adding the coverage of each headword in the set together For example, to determine the coverage provided by the most frequent 1,000 GSL headwords in the WSC, all 2,168 GSL headwords were ranked according to their frequency in the WSC The coverage provided by the most frequent 1,000 GSL headwords in the WSC (63.54%) was the sum of the coverage of each item in the set of the top 1,000 GSL headwords in the WSC Results Table 3 presents the average coverage provided by each headword in the GSL, BNC2000, BNC/COCA2000, and New-GSL To clarify the significance of the lexical coverage per headword, the figures were multiplied by 1,000 This provided figures that are more in line with studies of lexical coverage For example, the average coverage provided by each GSL headword in the WSC was 0.02979% However, when reporting the result, this figure was multiplied by 1,000 (29.79%) to enable readers to see the differences more clearly The ranking of the lists in terms of average coverage is quite consistent for both written and spoken discourse The BNC/ COCA2000 and BNC2000 always ranked first or second The BNC/COCA2000 ranked first in 11 out of 18 corpora The average coverage provided by the BNC/ COCA2000 and BNC2000 was 26.40%–34.55% and 27.03%–33.81%, respectively Using average coverage as the criterion for evaluation, the New-GSL always ranked third (24.39%–32.03%) while the GSL always ranked last (24.11%–31.46%) Table 4 presents the coverage provided by the most frequent 1,000, 1,500 and 1,996 headwords in the four lists No matter which cut-off point was chosen, the New-GSL consistently provided the highest coverage It ranked first in 17 corpora and second in the remaining corpus (SBCSAE) after the BNC2000 The coverage provided by the most frequent 1,000, 1,500 and 1,996 New-GSL headwords was 53.22%–69.51%, 54.01%–70.47% and 54.24%–70.82%, respectively The BNC/COCA2000 usually ranked second in terms of the coverage provided by the most frequent headwords The coverage provided by the most frequent 1,000, 1,500 and 1,996 BNC/COCA2000 headwords was 51.95%–68.09%, 52.68%– 68.88% and 52.80%–69.09%, respectively The BNC/COCA2000 provided higher coverage than the GSL in all 18 corpora no matter what cut-off point was chosen Similarly, the BNC/COCA2000 provided better coverage than the BNC2000 in six out of nine spoken corpora at all three cut-off points As the number of the most   26,078,219   10,484,320    5,641,642    3,243,449    2,841,573    1,112,905      977,923      943,110      512,801      320,496 Spoken corpora BNC (spoken) ICE (spoken) OANC (spoken) Movies WSC HKCSE TV programs LUND SBCSAE 110,022,403   87,602,389   12,839,527    3,467,451    1,024,320    1,021,357    1,019,642 Written corpora BNC (written) OANC (written) ICE (written) FROWN FLOB WWC Mean Token Corpus 59.84 59.26 57.71 57.34 52.87 59.45 64.00 52.27 66.35 64.79 63.71 64.59 65.31 68.20 64.23 66.59 GSL 60.20 59.80 58.06 58.29 53.96 60.21 64.54 56.52 67.49 64.69 66.37 63.90 65.16 66.48 63.98 66.26 BNC 2000 Overall coverage 60.56 59.85 58.44 58.18 53.79 60.17 64.76 52.80 66.70 65.75 64.63 65.39 66.17 69.09 64.89 67.40 BNC/ COCA2000 65.86 65.44 63.93 64.20 59.67 65.74 67.67 54.34 69.98 68.82 68.55 67.96 68.76 71.37 68.83 70.38 New-GSL 27.60 27.33 26.62 26.45 24.39 27.42 29.52 24.11 30.60 29.89 29.39 29.79 30.12 31.46 29.63 30.71 GSL 30.16 29.96 29.09 29.20 27.03 30.17 32.34 28.32 33.81 32.41 33.25 32.01 32.65 33.31 32.06 33.20 BNC 2000 Average coverage 30.28 29.93 29.22 29.09 26.90 30.09 32.38 26.40 33.35 32.88 32.32 32.70 33.09 34.55 32.45 33.70 BNC/ COCA2000 Table 3.  Overall and average coverage of headwords in the GSL, BNC2000, BNC/COCA2000, and New-GSL in 18 corpora 29.56 29.37 28.69 28.82 26.78 29.51 30.37 24.39 31.41 30.89 30.77 30.50 30.86 32.03 30.89 31.59 New-GSL Evaluating lists of high-frequency words 145    1,018,455    1,017,502    1,011,760 LOB Brown KOHAPUR 58.69 59.64 60.39 61.67 GSL 59.11 59.62 60.24 61.63 BNC 2000 Overall coverage 59.27 59.76 60.78 61.87 BNC/ COCA2000 64.79 65.50 65.84 66.94 New-GSL 27.07 27.51 27.86 28.45 GSL 29.62 29.87 30.18 30.88 BNC 2000 Average coverage 29.64 29.88 30.39 30.94 BNC/ COCA2000 29.08 29.40 29.55 30.04 New-GSL Token   26,078,219   10,484,320    5,641,642    3,243,449    2,841,573    1,112,905      977,923      943,110 Corpus Spoken corpora BNC (spoken) ICE (spoken) OANC (spoken) Movies WSC HKCSE TV programs BNC 2000 BNC/ COCA2000 63.54 63.30 64.17 62.82 64.92 63.57 63.54 62.47 64.02 63.94 63.75 64.48 67.49 65.55 68.09 62.77 61.94 63.18 65.29 64.25 65.80 GSL 1,000 headwords 66.92 65.93 65.96 67.02 69.51 65.62 67.81 NewGSL BNC 2000 64.51 64.40 65.37 63.57 66.10 64.48 64.37 63.56 65.08 64.96 64.87 65.74 68.06 66.30 68.88 63.87 63.47 64.50 66.26 65.68 67.01 GSL BNC/ COCA 2000 1,500 headwords 68.19 67.67 67.23 68.22 70.47 67.59 69.39 NewGSL BNC 2000 BNC/ COCA2000 64.79 64.69 65.75 63.71 66.37 64.63 64.59 63.90 65.39 65.30 65.16 66.17 68.20 66.48 69.09 64.21 63.98 64.89 66.57 66.26 67.40 GSL 1,996 headwords 68.77 68.40 67.74 68.63 70.82 68.57 70.12 NewGSL Table 4.  Coverage provided by the most frequent 1,000, 1,500 and 1,996 GSL, BNC2000, BNC/COCA2000, and New-GSL headwords in 18 corpora Mean Token Corpus Table 3.  (continued) 146 Thi Ngoc Yen Dang and Stuart Webb      512,801      320,496 LUND SBCSAE   87,602,389   12,839,527    3,467,451    1,024,320    1,021,357    1,019,642    1,018,455    1,017,502    1,011,760 BNC (written) OANC (written) ICE (written) FROWN FLOB WWC LOB Brown KOHAPUR Mean 110,022,403 Written corpora Mean Token Corpus Table 4.  (continued) BNC/ COCA2000 56.42 56.38 56.93 57.36 57.20 57.68 57.95 57.54 58.30 59.19 58.90 59.45 57.49 57.33 58.00 56.92 56.88 57.42 55.46 55.25 55.99 55.19 55.50 55.90 51.17 51.70 51.98 57.05 57.09 57.61 62.91 63.11 63.41 51.60 55.74 51.95 65.23 66.04 65.46 GSL BNC 2000 1,000 headwords 60.12 61.25 61.34 62.66 61.11 60.62 59.10 59.16 55.26 60.61 65.54 53.22 67.89 NewGSL BNC/ COCA 2000 58.07 58.48 58.73 59.02 59.09 59.31 59.73 59.65 60.22 60.97 61.00 61.31 59.22 59.52 59.97 58.62 59.10 59.29 57.12 57.43 57.88 56.76 57.62 57.65 52.44 53.46 53.40 58.76 59.42 59.55 63.76 64.22 64.47 52.18 56.42 52.68 66.10 67.18 66.47 GSL BNC 2000 1,500 headwords 62.98 63.91 64.17 65.36 64.01 63.52 62.06 62.15 57.97 63.63 66.89 54.01 69.27 NewGSL BNC 2000 BNC/ COCA2000 58.65 59.11 59.27 59.60 59.62 59.76 60.36 60.24 60.78 61.63 61.63 61.87 59.81 60.20 60.56 59.22 59.80 59.85 57.68 58.06 58.44 57.30 58.29 58.18 52.84 53.96 53.79 59.39 60.21 60.17 64.00 64.54 64.76 52.27 56.52 52.80 66.34 67.49 66.70 GSL 1,996 headwords 64.47 65.22 65.61 66.69 65.56 65.09 63.61 63.79 59.35 65.33 67.46 54.24 69.81 NewGSL Evaluating lists of high-frequency words 147 148 Thi Ngoc Yen Dang and Stuart Webb frequent headwords decreased, the number of written corpora in which the BNC/ COCA2000 provided better coverage than the BNC2000 increased The number of written corpora in which the most frequent 1,996, 1,500 and 1,000 BNC/ COCA2000 headwords provided higher coverage than those from the BNC2000 was six, eight, and nine corpora, respectively The GSL and BNC2000 took turns ranking last in most cases The coverage provided by the most frequent 1,000, 1,500 and 1996 GSL headwords was 51.17%–67.49%, 52.18%–68.06%, and 52.27%–68.20%, respectively Coverage of 51.70%–66.04%, 53.46%–67.18%, and 53.96%–67.49% was provided by the most frequent 1,000, 1,500 and 1,996 BNC2000 headwords, respectively The BNC2000 had lower coverage than the GSL in the spoken corpora It ranked last in six out of nine spoken corpora at all three cut-off points In contrast, the GSL had worse performance than the BNC2000 in written corpora It provided the lowest coverage in 3/9, 8/9 and 8/9 written corpora when the coverage provided by the most frequent 1,000, 1,500 and 1,996 GSL headwords was compared It can be inferred from Table 4 that the difference in the amount of coverage between the cut-off points is quite small The most frequent 1,000 headwords from the GSL, BNC2000, BNC/COCA2000, and New-GSL provided coverage ranging from 51.17%–67.49%, 51.7%–66.04%, 51.95%–68.09% and 53.22%–69.51%, respectively In contrast, the next 500 headwords after the most frequent 1,000 headwords only provided additional coverage of 0.57%–1.78% (GSL), 0.68%–2.33% (BNC2000), 0.73%–1.97% (BNC/COCA2000), and 0.79%–3.02% (New-GSL) Similarly, the next 496 headwords beyond the most frequent 1,500 headwords added very little coverage: 0.09%–0.66% (GSL), 0.1%–0.79% (BNC2000), 0.12%– 0.62% (BNC/COCA2000), and 0.23%–1.7% (New-GSL) This clearly shows a large difference in the relative values of the most frequent 1,000 headwords and the subsequent 996 headwords Discussion In answer to the first research question, the BNC/COCA2000 provided the highest average coverage in most spoken and written corpora It provided higher average coverage than the BNC2000 in five out of nine written corpora This is surprising because the BNC/COCA2000 was developed from a corpus that consists of only 40% written texts In contrast, the BNC2000 was derived from a corpus that was made up of 90% written texts Another interesting point relates to the comparison between the average coverage provided by the BNC/COCA2000 and the NewGSL In a corpus-based study, which considers frequency, range and coverage as the criteria for comparison, the BNC/COCA2000 was not expected to perform Evaluating lists of high-frequency words 149 as well as the New-GSL This is because unlike the New-GSL words, which were chosen purely based on frequency and range, the BNC/COCA2000 word selection was based on frequency, range, and intuition Therefore, the BNC/COCA2000 includes a number of items that might be useful in general conversation but may not have high frequency (e.g., mama, cop, buck and darling) Despite this fact, the BNC/COCA2000 provided higher average coverage than the New-GSL in all 18 corpora The excellent performance of the BNC/COCA2000 over the other lists may be because of the corpus used to create it It has been suggested that the ideal corpus to create a general service list should consist of an equal proportion of spoken and written texts (Nation & Waring, 1997; Nation, 2004), and texts from different sources and different varieties of English should be included to reflect L2 learning purposes (Nation, 2004) The corpora from which the GSL, BNC2000 and New-GSL were derived not meet this guideline and consist of mainly written texts In contrast, the corpus from which the BNC/COCA2000 was developed had a balance of spoken and written samples from different sources and different varieties of English As a result, the corpus used to create the BNC/COCA2000 may provide a more accurate picture of general service vocabulary than corpora used to develop the other three lists In answer to the second research question, the New-GSL had the highest coverage provided by the most frequent 1,000, 1,500, and 1,996 headwords in spoken and written discourse This indicates that, despite not performing well as a whole list, the New-GSL might have a number of very good items as well as a number of poor items in terms of lexical coverage This is supported by the fact that a number of headwords unique to the New-GSL were among the most frequent 1,000 NewGSL headwords in a wide range of corpora (e.g., majority, primary, weekend).4 In contrast, a number of headwords unique to the New-GSL occurred fewer than 10 times in multiple corpora (e.g., celebrity, venue, storey) The results of the present study confirm the fact that there is no perfect single way of comparing word lists, and suggest that if we look at lists from different perspectives, we can gain different information that is useful for evaluation If average coverage is used as the criterion to evaluate lists, the BNC/COCA2000 might be the best list However, if coverage provided by the most frequent headwords is used, then a reduced version of the New-GSL might be best In other words, the BNC/COCA2000 may be the best list as a whole, but the New-GSL may consist of a very large number of good items in terms of lexical coverage This raises the question of how many items should be included in a general service list Although there are no fixed upper limits to the number of general service words (Nation, 4.  The BNC/COCA lists include a separate list of transparent compounds so words such as weekend not appear in BNC/COCA2000 150 Thi Ngoc Yen Dang and Stuart Webb 2013; Schmitt, 2010), 2,000 is widely accepted as the cut-off point of general service vocabulary This figure came from the influence of the GSL, which has around 2,000 headwords (Schmitt, 2010), and was further reinforced as the cut-off point of high-frequency words by Nation and Hwang (1995) and Nation (2001) However, a 2,000 word GSL has been questioned by a number of researchers Engels (1968), when examining the distribution of the GSL in ten 1,000 running-word texts, confirmed that the first 1,000 high-frequency words were very good items in terms of frequency and range However, he also questioned the necessity of words beyond the first 1,000 word level because their frequency and range were too low to be included in the GSL This has been strongly supported by subsequent corpus-driven vocabulary research which found a huge difference between the amount of coverage provided by the first and the second 1,000 word levels in different discourse types (e.g., Nation, 2006) Moreover, research with L2 learners from a range of EFL context (e.g., Henriksen & Danelund, 2015; Nurweni & Read, 1999; Quinn, 1968) has shown that a number of L2 learners failed to learn the most frequent 1,000 words after a lengthy period of formal instruction but learn a proportion of items from lower frequency levels Thus, a smaller list of GSL items may be more effective in focusing learners on the most important items Schmitt and Schmitt (2014) suggested a different approach arguing that the 2,000 figure was too small Investigating this issue from the perspective of frequency, coverage, acquisition and use, they suggested that 3,000 might be a more suitable upper limit of highfrequency vocabulary Their argument is supported by Sorell’s (2013) study which found that lists of the most frequent 3,000 words were sometimes superior to shorter lists The present study provides strong support for Engels’ argument The most frequent 1,000 headwords in all lists provided a relatively large proportion of coverage in all 18 corpora while headwords beyond this level provided a relatively small proportion of coverage An analysis of the distribution of the four lists in the 18 corpora also revealed that a number of headwords did not appear in multiple corpora such as hooray (GSL), nought (BNC2000), canoe (BNC/COCA2000), and storey (New-GSL) Taken together, the present research highlights the value of learning the most frequent 1,000 headwords but also questions the need to include a number of items beyond the first 1,000 headwords in a general service list The present study revealed that the ranking of the four lists depended to a large degree on the criterion used for assessment However, perhaps the most clearly apparent finding was that the GSL is inferior to the three newer lists in terms of lexical coverage The GSL provided the lowest average coverage in all 18 corpora Moreover, when the coverage provided by the most frequent 1,000, 1,500 and 1,996 headwords was compared, the GSL also had very poor performance This is surprising because, as mentioned, using coverage provided by the most frequent headwords as the criterion should favor lists with a larger number Evaluating lists of high-frequency words 151 of headwords because a greater number of weak headwords are excluded In this study, the GSL had the second largest number of headwords (2,168 headwords) Therefore, it could be expected to provide better performance when the coverage provided by the most frequent headwords was compared Despite this advantage, the GSL ranked last in six out of 18 corpora and third in 12 out of 18 corpora when the coverage provided by the most frequent 1,000 headwords was compared It ranked last in 11 out of 18 corpora and third in the remaining seven corpora at both the 1,500 and 1,996 cut-off points The low lexical coverage provided by the GSL in this study is consistent with the results of earlier studies by Nation and Hwang (1995), Nation (2004), and Gilner and Morales (2008), which found that the GSL had lower coverage than the other lists The results of this study also support Richards’s (1974) and Schmitt’s (2010) recommendations that the GSL should be replaced by a list that represents current vocabulary An examination of the items in the four lists indicated that a number of words appearing in only the GSL (e.g., haste, inn, shilling, telegram) are infrequent and should be excluded from a current general service list In contrast, a number of words appearing in the other three lists which refer to common technology (e.g., computer, television, technology) or current issues (e.g., diet, energy, environment) did not appear in the GSL Moreover, words for everyday conversation (e.g., beer, chocolate, job) which appeared in the other three lists were also absent from the GSL Taken together, the findings suggest that the GSL should be replaced by a list that reflects current language Pedagogical implications This study suggests that learning the first 1,000 high-frequency items should be prioritized in L2 vocabulary programs because it offers L2 learners the best return for their learning effort It draws their attention to the most important lexical items of English, which provide a much larger amount of lexical coverage than the subsequent 1,000 items The strong correlation between lexical coverage and comprehension (Laufer, 1989; Schmitt et al., 2011; Van Zeeland & Schmitt, 2013) means knowledge of these words may allow learners to improve their comprehension quickly Second, a list of 1,000 items is a more attainable objective in the EFL/ESL context As mentioned, a reasonable number of learners in various EFL contexts failed to master the most frequent 1,000 word families after many years studying English (Henriksen & Danelund, 2015; Nurweni & Read, 1999; Quinn, 1968; Webb & Chang, 2012) Research on vocabulary growth rates has found that, L2 learners can acquire around 400 word families (Webb & Chang, 2012), or 500 lemmas (Milton, 2009) per year Together, these findings suggested that learning 152 Thi Ngoc Yen Dang and Stuart Webb 2,000 or 3,000 item lists may be a rather daunting task for L2 learners Learning 1,000 items, however, may be a feasible goal It greatly improves proficiency, and motivates further learning Moreover, a 1,000 item list is not too long and might get sufficient attention from teachers and learners within the limited time of an English course (Nation, 2001) whereas a 2,000 or 3,000 item list may not Therefore, perhaps there should be two lists that are oriented towards developing vocabulary size: a 1,000 item essential vocabulary list for beginners and a 2,000 item intermediate level list Learners should master the essential vocabulary list before moving to the intermediate level list Sequencing vocabulary learning in this way provides better scaffolding for learners’ vocabulary development It ensures that students will learn the most useful words first This will then facilitate the acquisition of less frequent items in the intermediate level list In return, knowledge of items in the intermediate level list allows learners to consolidate and expand their knowledge of essential vocabulary Once both lists have been mastered, a vocabulary size of 3,000 items might be more easily reached, which may enable learners to approach closer to the 95% coverage, which indicates reasonable comprehension For learners who have not mastered the most frequent 1,000 word families, there are a number of options for them to choose One option is Nation’s (2012) 1st 1,000 BNC/COCA word families If the learners are at a beginner level, they may find Dang and Webb’s (2016) Essential Word List (EWL) a more attainable goal The EWL is made up of the 800 items (headwords and lemmas) from the GSL, BNC2000, BNC/COCA2000, and New-GSL that provided the greatest coverage in a wide range of spoken and written corpora Compared with the four source lists, the EWL is more pedagogically friendly EWL words were classified into lists of lexical words and a list of function words The lexical words were further divided into sub-lists of 50 items according to their frequencies Distinguishing between lexical words and function words takes into account the difference in the way learners deal with different kinds of vocabulary (Carter & McCarthy, 1988) Breaking the EWL into sub-lists of a manageable size makes it easier for teachers to incorporate the teaching of the EWL words into a language learning program Once the learning goal has been set, it is important for teachers to organize opportunities for learners to acquire, consolidate, and expand on their knowledge of the most frequent 1,000 word families Knowledge of form and meaning should be the initial focus because it acts as the foundation for the development of other aspects (Schmitt, 2010; Webb & Chang, 2012) Once learners have mastered this aspect, teachers can help students to gradually develop other aspects of knowledge of the words by including the items in materials and activities Nation’s (2007) Four Strands provides a useful framework for organizing learning opportunities Evaluating lists of high-frequency words 153 From a pedagogical standpoint, it may be useful for list developers to present different versions of their lists that use different units of counting This offers teachers and learners more flexibility in choosing lists that satisfy their needs and contexts This idea is supported by Nation (2016) who points out that one problem of existing lists is that they only focus on a particular level of Bauer and Nation’s (1993) scale (either lemmas or Level word families) Creating different versions of a list with different units of counting allows the list to meet the needs of a wider range of users Headword lists may be the most user-friendly form for teachers and learners Lemma lists may be the most valuable for beginners, and word family lists may be more suitable for intermediate and advanced students Limitations One limitation of the present research is that it is based solely on frequency data Although frequency has always been the primary factor used to evaluate wordlists and is an important factor for selecting vocabulary (Ellis, 2002), it may not determine items that are most useful for L2 learners (Nation & Waring, 1997; Richards, 1974) General service lists which are based only on frequency will be affected by the limitations of the corpora from which they were developed (Nation, 2001; Schmitt, 2010) and lack a number of items which may be infrequent but seem to be useful for L2 learning and teaching purposes (Nation, 2013) A general service list lacking items that learners and teachers feel are useful may discourage them from using it (Sorell, 2013) Thus, apart from objective judgment, subjective judgment should be involved in the vocabulary selection (Richards, 1974; Sorell, 2013) That may be the reason why West (1953), Nation (2006), Nation (2012), and Bardel, Gudmundson, and Lindqvist (2012) used intuition as well as frequency as criteria to exclude and include certain words in their lists to match L2 learning purposes Follow-up research should use other criteria together with frequency to validate general service vocabulary One way to this is to use L2 learners’ and teachers’ intuition to evaluate the usefulness of the items which are unique to different lists L2 learners’ intuition of frequency may reflect the degree to which these learners have been exposed to the language (Schmitt, 2010) Similarly, the intuition of those who have experience teaching L2 learners may provide some insight into the lexical items that might be useful for L2 learners (Wang & Koda, 2005) This study is a starting point for validation of a new general service list It only looks at single words because the four lists used in the comparison are lists of single words To provide a more thorough picture about the performance of the four lists, research that examines the value of these lists from the multiword unit perspectives is needed Also, the validation of specific items that vary between lists 154 Thi Ngoc Yen Dang and Stuart Webb was beyond the scope of the study However, further research is needed to look at the four word lists and the items that are unique to each list using frequency as a criterion for inclusion as well as other criteria such as learners’ needs Conclusion This study has shown that in terms of lexical coverage, the GSL is outdated and should be replaced by a wordlist that reflects current vocabulary The results suggest that there is no perfect way to evaluate lists using a single approach; therefore, different methods of evaluation are necessary to provide an accurate assessment of wordlists If average coverage provided by each headword was chosen as the criterion, the BNC/COCA2000 might be the best list However, if the coverage provided by the most frequent headwords was used as the criterion, items from the New-GSL might form the best list This study highlights the importance of the first 1,000 high-frequency words and suggests that perhaps 1,000 words is a sufficient number of items for a general service list By using headwords as the unit of comparison, this study indicates an approach to evaluating wordlists that use different units of counting Acknowledgements We would like to express our sincerest gratitude to Professor Paul Nation for his valuable advice and to the ITL anonymous reviewers and the editor for their useful feedback We would also like to thank Professor Gerald Nelson for granting us access to the International Corpus of English References Bardel, C., Gudmundson, A., & Lindqvist, C (2012) Aspects of lexical sophistication in advanced learners’ oral production Studies in Second Language Acquisition, 34(2), 269–290. ​ doi: 10.1017/S0272263112000058 Bauer, L., & Nation, P (1993) Word families International Journal of Lexicography, 6(4), 253– 279. ​doi: 10.1093/ijl/6.4.253 Brezina, V., & Gablasova, D (2015) Is there a core general vocabulary? Introducing the new general service List Applied Linguistics, 36(1), 1–22. ​doi: 10.1093/applin/amt018 Browne, C (2013) The new general service list: Celebrating 60 years of vocabulary learning The Language Teacher, 4(37), 13–16 Carter, R., & McCarthy, M (1988) Vocabulary and language teaching London: Longman Evaluating lists of high-frequency words 155 Coxhead, A (2000) A new academic word list TESOL Quarterly, 34(2), 213–238. ​ doi: 10.2307/3587951 Coxhead, A., & Hirsh, D (2007) A pilot science word list for EAP Revue Franỗaise de Linguistique Appliqueộ, XII(2), 65–78 Dang, T N Y., & Webb, S (2014) The lexical profile of academic spoken English English for Specific Purposes, 33, 66–76. ​doi: 10.1016/j.esp.2013.08.001 Dang, T N Y., & Webb, S (2016) Making an essential word list In I S P Nation (Ed.), Making and using word lists for language learning and testing Amsterdam: John Benjamins Ellis, N C (2002) Frequency effects in language processing Studies in Second Language Acquisition, 24(02), 143–188 Engels, L K (1968) The fallacy of word-counts International Review of Applied Linguistics in Language Teaching, 6(3), 213–231. ​doi: 10.1515/iral.1968.6.1-4.213 Gardner, D (2007) Validating the construct of word in applied corpus-based vocabulary research: A critical survey Applied Linguistics, 28(2), 241–265. ​doi: 10.1093/applin/amm010 Gardner, D., & Davies, M (2014) A new academic vocabulary list Applied Linguistics, 35(3), 305–327. ​doi: 10.1093/applin/amt015 Gilner, L., & Morales, F (2008) Corpus-based frequency profiling: Migration to a word list based on the British National Corpus The Buckingham Journal of Language and Linguistics, 1, 41–58 Henriksen, B., & Danelund, L (2015) Studies of Danish L2 learners’ vocabulary knowledge and the lexical richness of their written production in English In P Pietilä, K Doró, & R Pipalová (Eds.), Lexical issues in L2 writing (pp 1–27) Newcastle upon Tyne: Cambridge Scholars Jenkins, J (2009) English as a lingua franca: Interpretations and attitudes World Englishes, 28(2), 200–207. ​doi: 10.1111/j.1467-971X.2009.01582.x Laufer, B (1989) What percentage of text lexis is essential for comprehension? In C Laurén & M Nordman (Eds.), Special language: From humans thinking to thinking machines (pp 316–323) Clevedon: Multilingual Matters Laufer, B., & Ravenhorst-Kalovski, G C (2010) Lexical threshold revisited: Lexical text coverage, learners’ vocabulary size and reading comprehension Reading in a Foreign Language, 22(1), 15–30 Milton, J (2009) Measuring second language vocabulary acquisition Bristol: Multilingual Matters. ​doi: 10.1057/9780230242258 Nation, I S P (2001) How many high frequency words are there in English Language, Learning and Literature: Studies Presented to Hakan Ringbom English Department Publications, 4, 167–181 Nation, I S P (2006) How large a vocabulary is needed for reading and listening? Canadian Modern Language Review, 63(1), 59–82. ​doi: 10.3138/cmlr.63.1.59 Nation, I S P (2007) The four strands Innovation in Language Learning and Teaching, 1(1), 1–12. ​doi: 10.2167/illt039.0 Nation, I S P (2012) The BNC/COCA word family lists Retrieved from Nation, I S P (2013) Learning vocabulary in another language (2nd ed.) Cambridge: Cambridge University Press Nation, I S P (2016) Making and using word lists for language learning and testing Amsterdam: John Benjamins. ​doi: 10.1075/z.208 156 Thi Ngoc Yen Dang and Stuart Webb Nation, I S P., Heatley, A., & Coxhead, A (2002) Range: A program for the analysis of vocabulary in texts Retrieved from Nation, I S P., & Waring, R (1997) Vocabulary size, text coverage, and word lists In N Schmitt & McCarthy, Michael (Eds.), Vocabulary: Description, acquisition and pedagogy (pp 6–19) Cambridge: Cambridge University Press Nation, P (1983) Testing and teaching vocabulary Guidelines, 5(1), 12–25 Nation, P (2004) A study of the most frequent word families in the British National Corpus In P Bogaards & B Laufer (Eds.), Vocabulary in a second language: Selection, acquisition, and testing (pp 3–14) Amsterdam: John Benjamins. ​doi: 10.1075/lllt.10.03nat Nation, P., & Hwang, K (1995) Where would general service vocabulary stop and special purposes vocabulary begin? System, 23(1), 35–41. ​doi: 10.1016/0346-251X(94)00050-G Nurweni, A., & Read, J (1999) The English vocabulary knowledge of Indonesian university students English for Specific Purposes, 18(2), 161–175. ​doi: 10.1016/S0889-4906(98)00005-2 Quinn, G (1968) The English vocabulary of some Indonesian university entrants Salatiga: English Department Monograph IKIP Kristen Satya Watjana Richards, J C (1974) Word lists: Problems and prospects RELC Journal, 5(2), 69–84. ​ doi: 10.1177/003368827400500207 Rodgers, M P H., & Webb, S (2011) Narrow viewing: The vocabulary in related television programs TESOL Quarterly, 45(4), 689–717. ​doi: 10.5054/tq.2011.268062 Schmitt, N (2010) Researching vocabulary: A vocabulary research manual New York NY: Palgrave Macmillan. ​doi: 10.1057/9780230293977 Schmitt, N., Jiang, X., & Grabe, W (2011) The percentage of words known in a text and reading comprehension The Modern Language Journal, 95(1), 26–43. ​ doi: 10.1111/j.1540-4781.2011.01146.x Schmitt, N., & Schmitt, D (2014) A reassessment of frequency and vocabulary size in L2 vocabulary teaching Language Teaching, 47(4), 484–503. ​doi: 10.1017/S0261444812000018 Schmitt, N., Schmitt, D., & Clapham, C (2001) Developing and exploring the behaviour of two new versions of the vocabulary levels test Language Testing, 18(1), 55–88 Sorell, C J (2013) A study of issues and techniques for creating core ocabulary lists for English as an international language Unpublished PhD thesis, Victoria University of Wellington, Wellington, New Zealand Van Zeeland, H., & Schmitt, N (2013) Lexical coverage in L1 and L2 listening comprehension: The same or different from reading comprehension? Applied Linguistics, 34(4), 457–479. ​ doi: 10.1093/applin/ams074 Wang, M., & Koda, K (2005) Commonalities and differences in word identification skills among learners of English as a second language Language Learning, 55(1), 71–98. ​ doi: 10.1111/j.0023-8333.2005.00290.x Webb, S., & Chang, A C.-S (2012) Second language vocabulary growth RELC Journal, 43(1), 113–126. ​doi: 10.1177/0033688212439367 Webb, S., & Rodgers, M P H (2009a) The lexical coverage of movies Applied Linguistics, 30(3), 407–427. ​doi: 10.1093/applin/amp010 Webb, S., & Rodgers, M P H (2009b) Vocabulary demands of television programs Language Learning, 59(2), 335–366. ​doi: 10.1111/j.1467-9922.2009.00509.x West, M (1953) A general service list of English words London: Longman, Green Zipf, G (1949) Human behavior and the principle of least effort: An introduction to human ecology New York NY: Hafner    1,011,760 KOHAPUR 84.42    1,018,455    1,017,502 LOB Brown 82.82    1,019,642 WWC 80.39 81.82 82.73 81.87    1,024,320    1,021,357 78.89 FROWN    3,467,451 ICE (written) 73.18 81.77 71.13 81.33 FLOB   87,602,389   12,839,527 BNC (written) OANC (written) 110,022,403 Written corpora 89.98      512,801      320,496 LUND      943,110 TV programs SBCSAE 89.14      977,923 HKCSE 89.33 88.33    2,841,573    1,112,905 89.89 Movies    3,243,449 OANC (spoken) 89.61 85.09 41.33 37.74 38.16 38.94 38.20 37.76 37.08 36.39 33.75 37.72 32.81 41.50 41.12 37.51 40.74 41.20 41.46 39.25 89.47 82.61 83.42 84.80 84.00 83.37 81.89 81.28 76.26 83.64 75.11 89.38 88.58 85.11 87.62 88.58 88.15 85.49 41.39 41.79 42.48 42.08 41.77 41.03 40.72 38.21 41.90 37.63 44.78 44.38 42.64 43.90 44.38 44.16 42.83 44.82 Average Overall Overall Average BNC2000 1,996 word families GSL 2,168 word families WSC   10,484,320    5,641,642 BNC (spoken)   26,078,219 Spoken corpus ICE (spoken) Tokens Corpus 82.52 83.77 85.00 84.28 83.20 82.02 80.70 75.15 83.28 72.39 90.70 90.85 83.27 89.79 90.80 91.60 86.54 91.01 Overall 41.26 41.89 42.50 42.14 41.60 41.01 40.35 37.58 41.64 36.20 45.35 45.43 41.64 44.90 45.40 45.80 43.27 45.51 Average BNC/COCA2000 2,000 word families 81.26 81.72 82.97 82.56 81.79 80.54 80.12 75.58 82.06 69.80 88.67 87.86 84.60 87.94 87.37 89.45 84.94 89.58 Overall 36.47 36.68 37.24 37.06 36.71 36.15 35.96 33.92 36.83 31.33 39.80 39.43 37.97 39.47 39.21 40.15 38.12 40.21 Average New-GSL 2,228 lemmas Supplementary data Overall and average coverage provided by the GSL, BNC2000, BNC/COCA2000 and New-GSL in 18 corpora Evaluating lists of high-frequency words 157 158 Thi Ngoc Yen Dang and Stuart Webb Author’s addresses Thi Ngoc Yen Dang Vietnam National University, Hanoi University of Languages & International Studies Victoria University of Wellington School of Linguistics & Applied Language Studies PO Box 600, Wellington 6140, New Zealand Stuart Webb University of Western Ontario Faculty of Education John George Althouse Building 1137 Western Road London, Ontario Canada N6G 1G7 Canada ngocyen1011@gmail.com swebb27@uwo.ca ... most frequent headwords as the criterion should favor lists with a larger number Evaluating lists of high- frequency words 151 of headwords because a greater number of weak headwords are excluded... The frequency of some GSL words was estimated frequency, which was calculated by doubling the actual frequency of the word in a 2.5 million running word corpus Evaluating lists of high- frequency. .. Evaluating lists of high- frequency words 133 offers theses learners a good return for their learning effort The size of this group of words is relatively small, but