Action Research on Student-Selected Vocabulary
by Jill Macchiaverna, graduate student
Students come to us with an infinitely varied range of background knowledge and experiences. The vocabulary bank inside each student’s head differs greatly from one to the next. How then are teachers to find an equitable direction for vocabulary instruction in their classroom? This research examines the role self-selection of vocabulary by students could play in vocabulary instruction by teachers.
Theoretical Grounding and Use of Research
This research is based in Unfoldment Theory. During the Romanticism Period in the 1700s, Jean-Jacques Rousseau “postulated that children’s learning would evolve naturally as a result of their innate curiosity” (Tracey & Morrow, 2017). Unfoldment is not about letting children do whatever they want, but about framing education through students’ natural interests. As Rousseau is quoted by Tracey and Morrow, “The goal is not to see the end of suffering but to see the increase of self-mastery, which necessarily includes hardship and difficulty” (Tracey & Morrow, 2017, p24). When we apply this theory to vocabulary, the question becomes: What words would the children like to learn? They can be hard words. Really hard words. “The goal is not to see the end of suffering…” But when students select their own vocabulary to learn, it is based on need and intense interest, which makes it much more likely that they will learn the vocabulary. The student’s self-selection allows their learning to follow individual curiosity and interest.
Pedagogical Research Question:
How can a student's self-selected vocabulary inform instruction related to choosing appropriately challenging vocabulary words?
Teachers consider many complex factors when choosing which vocabulary to include in classroom instruction. Many standards include vocabulary guidelines and many research-supported strategies exist for vocabulary instruction, but very little research exists for guidance on choosing vocabulary for instruction (Graves, et al., 2013).
Dashiell and DeBruin-Parecki outline a model for increasing student vocabulary and understanding to “bridge the vocabulary gap for those students who are most at risk” (Dashiell & DeBruin-Parecki, 2014, p513). Their F.R.I.E.N.D.S. model acronym stands for “Foster conversations, Robust and motivational instruction, Interactive reading, Engaging and literacy rich environment, Numerous opportunities to practice, Direct and explicit instruction, [and] Sophisticated and rare words” (Dashiell & DeBruin-Parecki, 2014, p513). Minus the direct instructional elements, their descriptions of each step are all components that contribute to the ideal home literacy environment that teachers want for their students.
Graves, et al., outline a process -- Selecting Words for Instruction from Texts (SWIT) -- that their subjects acknowledged was helpful but tedious, with steps for:
determining which words students are not likely to know
categorizing the words as essential (students must know them to understand the text), valuable (not essential but widely applicable), accessible (important for English Language Learners and others with limited vocabularies), or imported (words that describe themes or concepts in the text but don’t actually appear in the text)
determine the optimal type of instruction for each word
Powerful Instruction - clear definition, contextual examples, critical thinking questions about the words; conducted separately from the reading
Brief Explanation - stopping during the reading to explain unfamiliar words
Infer Meaning - guiding students to use context clues or background knowledge to reveal the definition
After all that, the teacher can finally teach the vocabulary.
The researchers who have studied the effects of self-selection on student vocabulary learning may not be incredibly numerous, but findings seem to consistently point to self-selection contributing to greater learning (Packer, 1970; Peterson, 1974; Noble, 1981, Fresch, 2005). The incorporation of self-selection leads to increased student engagement, self-motivation, and application of the vocabulary words (Graves, et al., 2014; Packer, 1970). Sylvia Ashton-Warner’s key vocabulary concept was cited in two of the studies I read as being the pioneering model for self-selection of vocabulary (Packer, 1970; Peterson, 1974). The Ashton-Warner approach included but was not limited to: having students choose the word they wanted to learn that day, discuss that word with friends, and do a variety of activities and manipulatives related to vocabulary (Packer, 1970). Even adults learning Chinese learned more vocabulary words when they chose which words to learn (Peterson, 1974).
Some studies also focused on how the self-selected words compared to text-selected words. Kindergarten and first grade readers chose much different vocabulary than what was suggested by basal readers (Packer, 1970). Eighth grade readers also chose different vocabulary words overall than adults when determining relevant vocabulary for a passage, but they still all chose the same few essential words, regardless of their reading ability (Harmon, et al., 2005).
In summary, while there is abundant information on best vocabulary instruction practices, little exists in the form of guiding the choice of vocabulary selection for instruction, but the research that does exist suggests allowing students to choose a portion of the words they learn leads to greater learning. This research investigated a strategy for using students’ self-selected vocabulary to help teachers determine which words to teach in class.
This was a very small study to determine the connection between self-selected vocabulary and words chosen for classroom vocabulary instruction.
Participants – The Participant was a 10-year-old female in the fourth grade at a mid-sized elementary in suburban Oklahoma. She was chosen to be the subject of this study first and foremost for convenience, as she is my daughter. There are additional characteristics that make her a good Participant for a vocabulary study. She is an exceptional reader when compared to her peers in volume of reading but she is typical in comprehension.
Data Sources – The data consists of words and the corresponding average frequency with which those words occur in the American English language. The Participant collected words that were unknown to her as she came across them in any text. I looked up the frequency of the word in the Corpus of Contemporary American English (COCA) housed at https://www.wordandphrase.info/frequencyList.asp, which is free to search for words in the 1-60,000 frequency range. “The corpus contains more than one billion words of text (25+ million words each year 1990-2019) from eight genres: spoken, fiction, popular magazines, newspapers, academic texts, and (with the update in March 2020): TV and Movies subtitles, blogs, and other web pages” (COCA, 2021).
Data Collection – The Participant wrote the words down and the date on which she found them in a small notebook. I conducted daily, ten-minute reading sessions with the Participant to facilitate capturing unknown words, but the Participant was not prevented from capturing words outside of the reading sessions. The texts were of the Participant’s choosing. As words were collected, I facilitated finding appropriate definitions of the words with the Participant. By allowing the Participant to identify and capture unknown words, I was able to determine what level of vocabulary words would be challenging for the Participant. Data collection occurred from February 4, 2021, through March 15, 2021.
Data Analysis – After the student’s word collection, I compiled the words and their corresponding frequencies into a spreadsheet. Words that showed up in the 1-5,000 range were considered Tier 1 (very common). Words that showed up in the 5,001-60,000 were considered Tier 2 (widely applicable across domains but not as common). Words that fell in the 60,001+ range were considered Tier 3 (highly domain-specific).
I did not encounter previous studies that classified Tier 2 and Tier 3 within specific frequency ranges. The distinction made in this study between Tier 2 and Tier 3 is arbitrary, a decision made by me, inspired by evidence that most native American English speakers know the first 5,000 words by fourth or fifth grade and that the average high-schooler knows about 40,000 words (Graves, 2013). I assumed that average high-schoolers do not all know the same 40,000 average words, and therefore the 60,000 cap for Tier 2 felt reasonable. Also, the searchable word frequency list is only free for the first 60,000 words.
The words and their frequencies were plotted over time. By plotting the frequency of unknown words collected by the Participant over time and noting whether any words appeared more than once, I was able to produce a frequency range that indicated the Participant’s current Level of Challenge with respect to vocabulary. The term Level of Challenge was chosen because the unfamiliar words indicate at what level of frequency words became challenging for the participant.
A good process that helps teachers choose which vocabulary to teach is Selecting Words for Instruction from Texts (SWIT) (Graves et al., 2013). The process assumes the teacher will know with which words her students are unfamiliar. By monitoring current levels of vocabulary challenge as demonstrated by unknown words selected by students, the teacher can use the frequency of potential vocabulary words to decide whether the words are likely to fall in an accessible range for her students.
The instructional model consists of students self-selecting, capturing, and researching definitions for unknown words from texts they encounter in school. Once a week, students will be in charge of data entry, adding their (previously unknown) words and frequencies into a cloud-based spreadsheet. The spreadsheet will update automatically with new entries, giving the teacher a way to monitor the average levels of challenge among her students. The teacher will use this information to select vocabulary words for classroom instruction that are appropriately challenging for the ranges of abilities among her students.
Table 1 - Sample Data Set
Table 2 - Sample Chart for Monitoring Levels of Vocabulary Challenge
Findings and Discussion
The analytical tool used to find the frequency of selected words in the American English language was produced alongside the Corpus of Contemporary American English (COCA) and can be found at https://www.wordandphrase.info/frequencyList.asp. As mentioned previously, the more frequent a word is, the lower the ranking. Any words that did not appear in a search of COCA were assumed to be Tier 3 words, and were given the rank of 60,001 (this arbitrary division between Tier 2 and Tier 3 words came about because users can only freely search the first 60,000 words in the COCA database).
The Participant self-selected unfamiliar words over the course of several weeks from books which were being read for pleasure. She wrote the words down in a small notebook. The Participant and I researched the definitions of the words together. Separately, I looked up the frequency of the student’s word selections and plotted this quantitative data over time. No words showed up more than once in self-selection.
I am including some data collected before my research plan was approved because the collection strategy did not change. Other data that I didn’t anticipate using in my original research plan comes from the student’s current curriculum. I looked up text-selected vocabulary words that appeared in the student’s assignments from the time-frame similar to the study, to give a comparison for the level of frequency -- and therefore challenge -- of the vocabulary likely to be included in the student’s classroom instruction. High frequency words are generally considered “easy” and Tier 3 words are considered more challenging due to their specificity and infrequent use.
The student self-selected 61 unfamiliar words to learn over the course of 39 days. The self- and text-selected word frequencies were plotted over time to determine if there were any trends in the level of challenge among the vocabulary. No trends emerged over time. Judging by frequency rankings, however, it was evident that overall the self-selected words were more challenging than the text-selected words (see Tables 3 and 4).
Table 3: Word frequencies for the student’s self-selected vocabulary plotted over time. Some days, the student recorded multiple words. Other days, the student encountered no unfamiliar words.
Table 4: Word frequencies for the curriculum’s text-selected vocabulary plotted over time. One set of vocabulary was presented to students during an unknown time between August 2020 and March 2021. One set was presented during the week of February 22, 2021. A third set was presented during the week of March 22, 2021.
Of the student’s self-selected words, 9% were Tier 1, 76% were Tier 2, and 18% were Tier 3. Of the 24 text-selected vocabulary words examined, 42% were Tier 1, 58% were Tier 2, and none were Tier 3 (see Table 5).
Table 5: Frequency ranges of Self- and Text-Selected Vocabulary Words
How can a student's self-selected vocabulary inform instruction related to choosing appropriately challenging vocabulary words? This data indicates that the student regularly encounters words much more challenging in her self-selected reading materials than in her curriculum reading materials. This is a phenomenon that other researchers have encountered as well (Fresch, 2005). This is helpful information because the teacher may decide the student should spend time developing other strengths instead of vocabulary, as the text-selected vocabulary is likely not new to the student and/or not going to be a challenge to learn.
To keep the student engaged and to continue to monitor comprehension, the teacher could designate this student as the class Word Finder for a period of time, in which the student is tasked with identifying which words she thinks her classmates will need to know to be able to understand assigned reading. The study that looked at eighth graders who were tasked with vocabulary selection found they read more carefully and were highly engaged in the assignment because of their perceived accountability to their classmates (Harmon, et al., 2005).
The results of this study are of interest to the parents as well. They can take pride in the student’s accelerated reading and feel validated about surrounding the child with good books and supporting the child’s self-selection of books. While not indicated in the quantitative data, my experience indicates parents should continue to read with their children beyond early childhood education to check for understanding. As the Researcher and the Parent of the Participant, there were several times when I called attention to a particular word that I thought she might not know, though she didn’t say something right away. I would do this when I was suspicious and there were no context clues to help her figure the word out. Sometimes she really did know the word and sometimes she didn’t. The process of asking helped the Participant get comfortable with looking up unfamiliar words when context clues were unhelpful, eventually making a habit of it on her own.
Limitations of this study include a small sample size (one student) and limited access to curriculum vocabulary. A consideration for future research would be to include more students and also comparisons with more text-selected vocabulary in the analysis. If possible, this research would be furthered with an analysis of large sections of assigned text to determine what percentage of Tier 1, Tier 2, and Tier 3 words are included in entire texts, rather than just in vocabulary lists.
PROFESSIONAL DEVELOPMENT RESOURCES
Using Students’ Self-Selection of Vocabulary to Inform Instruction
Tracey, D. H., Morrow, L. M. (2017). Lenses on reading: An introduction to theories and models. The Guilford Press.
Graves, M. F. et al. (2013). Words, words everywhere, but which ones do we teach? The Reading Teacher, 67(5), 333-346. DOI: 10.1002/TRTR.1228
Dashiell, J., DeBruin-Parecki, A. (2014). Supporting your children’s vocabulary growth using F.R.I.E.N.D.S. model. The Reading Teacher, 67(7), 512-516. DOI: 10.1002/trtr.1250
Packer, A. B. (1970). Ashton-Warner’s key vocabulary for the disadvantaged. The Reading Teacher, 23(6), 559-564. Stable URL: https://www.jstor.org/stable/20196369
Peterson, C. C. (1974). Self-selection of vocabulary in reading instruction. The Journal of Educational Research, 67(6), 253-254. Stable URL: https://www.jstor.org/stable/27536590
Harmon, J., et al. (2005). Vocabulary Self-Selection: A Study of Middle-School Students’ Word Selections from Expository Texts. Reading Psychology, 26(3), 313–333. https://doi-org.argo.library.okstate.edu/10.1080/02702710590969689
Noble, E. F. (1981). Self-selection: a remedial strategy for readers with a limited reading vocabulary. The Reading Teacher, 34(4), 386-388. Stable URL: https://www.jstor.org/stable/20195254
Fresch, M.J. (2005). Observing the self-selection of an emerging literacy learner. Reading & Writing Quarterly, 21(2), 135-149. https://doi-org.argo.library.okstate.edu/10.1080/10573560590915941
Harmon, J.M., et al. (2008). Pick a word -- not just any word: using vocabulary self-selection with expository texts. Middle School Journal, 40(1), 43-52. Stable URL: https://www.jstor.org/stable/23044611
Corpus of Contemporary American English. (2021). [Help page]. Retrieved on February 18, 2021, from COCA: https://www.english-corpora.org/coca/