1. Introduction TEM-4 is a standardized test for English majors of comprehensive colleges in China. This test is a significant way to measure the English proficiency of English majors. Dictation is an important and difficult part in TEM-4 which combines speech recognition, information understanding, memory and writing (Hu, 2017). Over the years, the average score of the dictation section in TEM-4 is not ideal. There are many factors that affect the accuracy of dictation, such as dictation style, concentration, mental state, etc. 1.1 Need for the study In TEM-4, the score of listening part accounts for 30% of the total score. Dictation, as an important part of listening ability, sets a higher standard for the English majors. Dictation is a comprehensive skill, which can reflect students#8217; comprehensive English level. It is a powerful psychological skill that covers such psychological activities as speech discrimination, comprehension, memory and expression, which can train students#8217; ability of listening, writing and short-term memory of language sequences (Brown Barnard, 1975). Therefore, for a long time, dictation is not only a common training and test of foreign language teaching, but also a compulsory test item of TEM-4. However, the examinees do not perform well in the dictation part. One of the important factors is vocabulary. If the examinees don#8217;t know a word, they can#8217;t identify it, let alone write it down. Also, working memory is another significant factor in the dictation. According to Hu (2017), the problems existing in students#8217; dictation are mainly shown in the following aspects: neglect of prediction before listening, improper grasp of rhythm, unreasonable allocation of attention, insufficient vocabulary, delayed writing speed, improper use of punctuation marks and information omission. Poor instantaneous memory and short-term memory are also the reasons why students#8217; dictation accuracy is not high (Zhang, 2014). In order to explore the influence of both the two factors on dictation, the author will proceed from working memory and vocabulary size to carry out the research of dictation. 1.2 Research purpose The present study intends to investigate the dictation accuracy among English majors with different working memory capacity and different vocabulary size. Specifically speaking, the purpose of this study is three-fold: First, whether the difference in working memory capacity has a significant impact on dictation accuracy; secondly, whether the size of vocabulary has a significant impact on dictation accuracy; thirdly, whether there is a significant correlation between working memory capacity and vocabulary size. By doing so, the author hopes to provide English majors with some practical suggestions to improve their dictation accuracy. 2. Literature Review 2.1 Working memory 2.1.1 Definition of working memory Working memory is a memory system that temporarily processes and stores information with a limited-capacity (Baddeley Hitch, 1974). It plays a key role in the completion of complex cognitive tasks such as language understanding, learning, reasoning and thinking (Wu Jin, 2005). Working memory capacity, also known as working memory span. It is often used as an important indicator of working memory. The working memory span, also known as the complexity span, differs from the simple short-term memory span (digital memory span) in that the working memory span needs to be processed and stored simultaneously, i.e., to memory while completing another task (secondary task). 2.1.2 Models of working memory In 1974, Baddeley and Hitch suggested that working memory was multi-component system which included central executive, phonological#8194;loop and visuospatial#8194;sketch#8194;pad. The phonological loop is responsible for the sound-based information storage and control. It is the basis of memory span. In addition, phonological loop plays an auxiliary role in speech comprehension, vocabulary acquisition and language learning. There is evidence that the capacity of phonological loop is a good predictor of second language acquisition ability in children and adults. The visuospatial#8194;sketch#8194;pad is mainly responsible for storing and processing visual information. The central executive system is the core of working memory, responsible for the relationship between each subsystem and their relationship with long-term memory, coordination of resources and selection and planning of strategies, etc (Baddeley Hitch, 1974). This is the earliest and the most famous working memory model. In 2000, Baddeley et al. revised the classical model of working memory by adding a new component, the episodic buffer. Episodic buffer is a component used to integrate visual, spatial and verbal information. It is a space with limited capacity for integrating information from phonological loops and visuospatial#8194;sketch#8194;pad. Episodic buffers are connected to long - term memory (Baddeley, 2000). Baddeley#8217;s multi-component model has been continuously improved and developed since its introduction, but there are still some assumptions that have not been verified, and some experimental results cannot be explained in this model. However, this model has an important impact on later research, which is undeniable. This study is based on this multi-component model. 2.1.3 Measure of working memory capacity In 1980, Daneman and Carpenter created a method of measuring working memory capacity#8212;reading span test, requiring participants to read a series of sentences, and then recalling the last word of each sentence. The tester divided the cards into five groups, from two sentences to six sentences. Each group consists of three rounds, for a total of sixty sentences. In the three rounds tests for each sentence group, the subjects must answer correctly for at least two rounds before the test can continue. The test ended when the subjects were unable to recall any rounds of words correctly in a group. In this test, the reading span was 2-6. The working memory reading span is measured by the maximum number of sentences in which the subject can correctly read and remember the ending words (Daneman Carpenter, 1980). Unlike the simple digital memory span test and word memory span test, the working memory test requires the participants to correctly understand the sentence and remember the words, that is, to complete both understanding and memory work, which is consistent with the theoretical concept of working memory. Therefore, reading span test is more effective than other tests such as digital span test. The experimental materials for measuring working memory capacity in this paper are arranged according to Daneman and Carpenter#8217;s reading span test. However, there are some shortcomings in this reading span test. First of all, the subjects did not judge the semantics and grammar of the sentences they read, so their degree of processing sentences could not be determined when the subjects read the sentences. Secondly, although the test requires the subject to process and store the sentence, only the stored results are calculated, without considering the sharing of working memory resources by processing and storing functions(Waters Caplan,1996). Waters and Caplan revised the reading span test on the basis of Daneman and Carpenter#8217;s reading span test. Subjects need to judge the acceptability of sentences from the syntax and semantics. Waters and Caplan adopted the new scoring method, which has three criteria: the average response time for subjects to make correct judgments, the average accuracy of acceptable or unacceptable sentences presented, and the number of words they can recall in order (Waters Caplan, 1996). The retest validity of the total working memory score was higher than that of the previous reading span test. The working memory capacity test materials and procedures in this article are adapted from Waters and Caplan#8217;s test. 2.2 Vocabulary size Vocabulary size refers to the number of words or phrase in a language that a person can recognize or use. According to Wikipedia, ”A vocabulary is a set of familiar words within a person's language”. Usually, one#8217;s vocabulary size increases with age. Vocabulary is the foundation of learning any language. There are four types of vocabulary: listening vocabulary, speaking vocabulary, reading vocabulary and writing vocabulary. In English learning, without grammar, one can#8217;t express something well, and without words, nothing can be expressed. Without sufficient vocabulary, it is impossible to listen, speak, read and write, and communicate effectively in English. Actually, to a certain extent, vocabulary size indicates one#8217;s English proficiency level. Sometimes, vocabulary is far more important than structural accuracy. For example, an English beginner can express what he or she wants to express by using some isolated key words instead of sentence in correct grammar. 2.3 Dictation According to Wikipedia, dictation is the transcription of spoken text: one person who is ”dictating” speaks and another who is ”taking dictation” writes down the words as they are spoken. It is a way to strengthen memory in learning human language. The dictation process basically reflects the process of language acquisition: learners first acquire the sound information, then input the information into the brain for processing, and then record it accurately and quickly according to certain requirements. Practice proves that it is an effective way to cultivate and test students#8217; language ability (Qian, 2006). In the process of acquiring the sound information, dictation is good for students to understand the correct pronunciation of vocabulary and reduce pronunciation errors. Certainly, dictation can strengthen students#8217; listening ability. In the process of inputting the information into the brain for processing, dictation can advance students#8217; ability to understand and process information, and fosters unconscious thinking in the new language. In the writing process, dictation can exercise students#8217; short-term memory ability and improve the speed of shorthand and sketch. In addition, it can expose students#8217; spell errors that they may make. 2.4 Related empirical studies on dictation accuracy 2.4.1 Error analysis Since the students didn#8217;t work well in the dictation part, quite a few researchers analyzed the errors in the dictation through investigations, researches and experiments. Yan Minggui (2004) pointed out that the common errors in dictation were reflected in both linguistic factors and non-linguistic factors. He divided the common errors into six parts#8212;phonetic troubles, spelling troubles, grammar errors, changing the original sentence, background knowledge or common sense errors and negative psychology. According to Lv Lan (2011), students also made errors of omission which were caused by insufficient vocabulary and poor short term memory. 2.4.2 Cognitive strategy and dictation Li Zujun (2013) introduced three cognitive strategies --- new word strategy, grammar strategy and predictive reasoning strategy. The study found that the new word strategy is negatively correlated with listening performance, which means that the more vocabulary candidates pay attention to the vocabulary, the lower the dictation score. Contrary to the new word strategy, grammar strategies are positively correlated with dictation scores. Predictive reasoning strategy is another important strategy used in dictation which can improve dictation score. The metacognitive strategy is one level higher than the cognitive strategy. It includes three aspects: planning strategy, monitoring strategy and evaluating learning result strategy. The metacognitive strategy is divided into pre-listening strategy, in-listening strategy and post-listening strategy. In order to improve English scores better, it is necessary to train students in metacognitive strategies (Li Gao, 2014). 2.4.3 Influential factors Xie Wenqian (2017) carried out an empirical study to investigate the relationship between working memory capacity and the comprehension of learners during dictation, and the relationship between working memory capacity and Chinese character writing performance. The results showed that When Chinese is acquired as a second language, there is a significant correlation between working memory capacity and listening comprehension performance, as well as the performance of Chinese characters in dictation. The correlation between working memory capacity and comprehension of learners during dictation is higher than that between working memory capacity and Chinese character writing performance. Liu Ying (2014) conducted an empirical study to explore the effects of working memory and listening style on English learners#8217; dictation accuracy. The results proved that Subjects#8217; working memory and listening style correlates with dictation to a significant level and can explain the variance of dictation score in a proportion of 23.6% and 16.7%. 2.5 Problems in the previous studies Although there are quite a few studies analyzed the mistakes appeared in dictation, some researchers have testified the factors which affect dictation performance by experiments. It is no doubt that both vocabulary and working memory play crucial roles in dictation. However, there is no experiment on the relationship between vocabulary size, working memory and dictation. This study will explore the relationship among working memory capacity, vocabulary size and dictation to find how the former two factors affect the dictation accuracy and how the two factors affect each other. References Baddeley, A. D. (1992). Working memory. Science, 255, 556-559. Baddeley, A. D. (2000). The episodic buffer: a new component of working memory? Trends Cogn Sci, 4,417-423. Baddeley, A. D. Hitch, G. (1974). Working memory. New York: Academic Press. Brown, D. Barnard, H. (1975). Dictation as a learning experience. RELC Journal, 2, 42－62. Chiang, H. H. (2018). English vocabulary size as a predictor of TOEIC listening and reading achievement among EFL students in Taiwan. Theory and Practice in Language Studies, 8, 203. Cowan, N. (2016). Working memory capacity. New York: Psychology Press. Daneman, M. Carpenter, P. A. (1980). Individual differences in working memory and reading. Journal of Verbal Learning and Verbal Behavior, 19, 450－466. Diana, I. S., David, M. O. (2018). Can working memory strategies enhance English vocabulary learning? How, 25, 29－47. Ellis, N. C., Sinclair, S. G. (1996). Working memory in the acquisition of vocabulary and syntax: Putting language in good order. Quarterly Journal of Experimental Psychology, 49, 234－250. Geraldine, G. C. (1986). Dictation improves writing, listening, spelling and vocabulary. The Reading Teacher, 39, 609－610. Jia, M. (2018). On the importance of English vocabulary. English on Campus, 5, 230. Liu, Y. (2014). The effect of students#8217; working memory and listening style on dictation. Unpublished master#8217;s dissertation, Huazhong University of Science and Technology, Wuhan. Londe, Z. C. (2008). Working memory and English as a second language listening comprehension tests: A latent variable approach. Los Angeles: University of California Press. Waters, G. S. Caplan, D. (1996). The measurement of verbal working memory capacity and its relation to reading comprehension. Quarterly Journal of Experimental Psychology, 49, 51－80. Waters, G. S. Caplan, D. (1996). Processing resource capacity and the comprehension of garden path sentences. MEMORY AND COGNITION, 24, 342－356. 段红萍 戈玲玲 (2011)，英语专业四级听写心理机制及策略研究, 外国语文（双月刊）（6）：134－136。

管芳 (2016)，工作记忆的研究综述, 现代交际（12）：134－135。

胡玲玲 (2017)，英语专四听写中的常见问题及应对策略, 英语教师(10)：45－48。

李冰 高翔 (2014)，元认知策略在CET-4听力短文听写中的运用, 外语研究（1）：365－366。

李祖君 (2013)，基于认知策略的英语专业四级听写探究, 学周刊（5）：10－11。

吕兰 (2011)，英语专业学生短文听写错误之分析, 柳州师专学报（1）：111－116 倪锦诚 (2017)，工作记忆容量对二语阅读能力的影响研究, 解放军外国语学院学报（3）：79－85。

钱晓玲 (2006)，听写与大学英语教学, 中国西部科技（36）：88 魏望东 (2003)，中国大学生英语词汇量实际状况分析, 中山大学学报论丛 （2）：225-228 吴文春 金志成 (2005)，工作记忆及其理论模型, 中国临床康复（40）：74-76。

项葆 (2015)，英语短文听写错误分析及其训练策略, 池州学院学报（4）：148－150。

解文倩 (2017)，工作记忆容量对留学生汉语听写成绩影响的研究, 硕士学位论文，南京师范大学。

严明贵 (2004)，英语专业学生短文听写的错误分析, 台州学院学报 （5）: 53－56。