Deep Dives

The Issue

In the classroom, your students will read that “a pathogen transmitted through fruits and vegetables is making people susceptible to an illness.” You know that having familiarity with all these words would be helpful, but the curriculum only highlights illness as a target vocabulary word to teach. Given the volume of complex academic vocabulary students encounter, an important instructional decision needs to be made. Although some words can be learned implicitly through incidental exposure, others require explicit instruction. However, which words are “important enough” to stop and teach so that there is enough instructional time for students to read the text? What are the chances that your students know what a pathogen is, and is it worth the instructional time needed to teach it? What about transmitted? Could that be a good option since some students may be familiar with the root <trans>? Alternatively, susceptible may be a good word to teach since students may encounter it in other reading passages? You suspect a handful of students may have been exposed to the word illness, but what about those who don’t have that background knowledge? Deciding which words to prioritize for instruction is a challenge many teachers will encounter, but what does the research say about which vocabulary words are the best to teach?

The Research

Many vocabulary studies emphasize the instructional strategies used to teach it, not necessarily the scientific rationale for which words are the best to teach. However, several theoretically grounded approaches and criteria have been generated to support teachers who are frequently bombarded with decisions on which words to prioritize for instruction (Hadley & Mendez, 2021; Nagy & Hiebert, 2011). These approaches are based on our present knowledge of vocabulary acquisition and word exposure. So, although researchers have not directly compared how well the approaches work, they may still provide useful guidance for selecting words. Most popularly, the tiered approach, regarded as one of the most widely implemented approaches to select words for vocabulary instruction, is frequently used amongst teachers (Beck et al., 2002; Hellman, 2018; NCIL, 2023; Steele & Mills, 2011). In this approach, words are separated into three tiers, where tier one words are those that do not necessitate instructional time as they are likely to be learned implicitly in oral language (e.g., cat, run, sick). Tier two words are highlighted as being the best candidates for instruction because they consist of high-frequency academic vocabulary that students will likely encounter during reading, thus making reading more accessible (e.g., obvious, consequence, predict). Lastly, tier three consists of words that are specific to a particular domain or context, often making them inappropriate for general instruction due to limited generalizability (e.g., morphology, pathology, algorithmic). Simultaneously, different curricula have attempted to support teachers’ decision-making by pre-selecting words to explicitly teach. Additionally, resources such as Selecting Words for Instruction (SWIT; Graves et al., 2014), the Academic Vocabulary List (AVL; Gardner & Davies, 2013), and Words Worth Teaching (WWT; Biemiller, 2010) have been generated to help reassure teachers that they are teaching the “right” words.

Though these selection approaches provide a meaningful starting point for teachers, uncertainty about how each approach compares to one another remains. 

As noted earlier, one of the reasons for this uncertainty is that studies that address vocabulary for instruction have emphasized the effects of the instructional strategies, not the word selection approaches themselves. For this reason, we cannot determine whether the effects are the result of the words selected, the instruction provided, or a combination of the two. As such, there is a lack of research comparing the outcomes for children who were taught using one selection approach versus another (e.g., words from the AVL list versus the WWT list). Though each vocabulary selection approach incorporates distinct considerations, they also often share overlapping criteria to drive decisions on which words are selected. For example, multiple approaches prioritize words that are likely to be unfamiliar and useful across contexts; however, different approaches may differ in the weight they place on characteristics such as morphological complexity or using different parts of speech. Consequently, even though researchers may be using the same general criteria, they will likely generate different word lists, thus making a direct comparison difficult. A relatively recent systematic review of word selection approaches found that the tiered approach was the most common method used for vocabulary selection, but around 45% of the expected tier two words actually fell into a classification of tier one (Hadley & Mendez, 2021). This inconsistency suggests that even when researchers use the same selection criteria, a comparison across studies would be difficult.

Because of this subjectivity, teachers may be tempted to rely on the pre-selected words identified in the curriculum to guide decisions on which words to prioritize for explicit instruction. This temptation is further intensified by pressures to adhere to high-quality instructional materials (HQIM), meaning that if the curricula selects certain words, then those are the words that should be explicitly taught. However, this approach is complicated by several realities. Some curricula highlight as many as ten or more words for explicit instruction. And while research does not presently support the idea that there is a target number of words that should be taught per lesson because there is much more nuance to word learning (Beck et al., 2002; Nagy & Hiebert, 2011; Graves et al., 2014), it is unlikely a teacher will be able to teach so many words in one day’s lesson. Critical attention should be given to the reality that curricula may not be reliable in selecting the “best words”. A review of the four most common core reading curricula used in kindergarten classrooms concluded that the majority of curricula not only selected words that would be considered “too easy” and thus not the best use of instructional time, but there was also no discernable pattern for the class (i.e., nouns, verbs, etc.) of word selection so students were limited in the variety of words that they were exposed to (Wright & Neuman, 2013). Based on these realities, though fidelity to HQIM can provide teachers with a starting point for word selection, it may also have the unintended consequence of restricting teachers’ professional judgement and limiting students’ exposure to meaningful words aligned to their unique needs.

More broadly, although curriculum, word lists, and other selection approaches can all help to guide decision-making about which words should be prioritized for instruction, they cannot accurately account for the characteristics of your learner. Student characteristics such as their level of background knowledge or cultural and linguistic experiences largely influence vocabulary acquisition (Fisher & Frey, 2009). Thinking back to our earlier example, if many of the children in your classroom have parents who are medical professionals, illness may not be the best word for explicit instruction since they may likely already be familiar with it given their language environment. Conversely, if your students don’t have as much access to medical care, this same word may be novel and worth the instructional time. Alternatively, you may have a large population of students who speak both Spanish and English in your classroom, so spending time explaining that transmit in English and transmitir in Spanish are cognates could be a valuable use of instructional time. Whatever the characteristics of your learners may be, word selection should acknowledge the influences of age, cultural and linguistic background, and access to language-rich environments as these variables largely relate to vocabulary learning (August, et al., 2005; Beck & McKowen, 2007; White, 1990). At school entry, children with more access to language-rich environments may know twice as many words as their peers (White et al., 1990), children with learning disabilities will likely know fewer words than their peers (Quinn et al., 2020), and multilingual children may present as having a smaller vocabulary compared to their monolingual peers (August et al., 2005). Therefore, approaches that assume that a particular word is the best word to teach should be met with questions on whether that word is appropriate for the students in your classroom.

In the absence of comparative studies to support the efficacy of a particular approach, it is important to reiterate that children do not need explicit instruction on every unfamiliar word that they encounter (Freebody & Anderson, 1983). Estimates from Nagy and Anderson suggest that printed school English has approximately 88,500 distinct words (1984), so the proportion of unfamiliar words that students are likely to encounter far outweighs the availability of time and resources to teach those unknown words. Instead, teachers should be empowered to make strategic instructional decisions based on the research we do have. A summary of work on early childhood vocabulary suggests the following criteria:

1. select words from a variety of parts of speech such as nouns, verbs, adjectives;
2. select words that are conceptually related to one another such as pathogen, disease, and infection;
3. avoid selecting words that are used in everyday speech or can be easily labeled such as cat, dog, or pig, since they may be “too easy”. Likewise, words that children lack an existing conceptual knowledge of or that occur infrequently in academic texts such as ubiquitous or arbitrary, would likely be “too hard”;
4. select words that are useful to the topic being discussed or would help children better communicate their thoughts and ideas; and
5. select words that are sensitive to the unique characteristics of the learners (Hadley & Mendez, 2021).

Similarly, criteria put forth by Nagy and Hiebert (2011) suggests that teachers should select words based on their role in language (i.e., does it appear frequently in text and across different genres?), their role in the lexicon (i.e., does it have a large morphological family?), their role in student’s existing knowledge (i.e., is it familiar or conceptually difficult to understand?), and their role in the lesson (i.e., is this word necessary to understand the text and will it be encountered again?). As research on the criteria to guide word selection continues to evolve, teachers may consider turning to the shared principle that underlies many of these recommendations: select words that will support your students to make meaningful connections across academic concepts and learning experiences.

In Sum

A large and rich vocabulary is a foundational component of reading proficiency (Scarborough, 2001). However, continued research is needed to determine the most effective approaches for selecting the vocabulary words that require explicit instruction. Several sensible approaches grounded in principles of word learning, such as words selected from curriculum, the tiered approach, and word lists, are available to assist teachers; however, these approaches are primarily informed by theory as opposed to head-to-head comparative studies.

However, the absence of a single definitive approach to word selection does not suggest that teachers should dismiss the existing body of research. Instead, it highlights the importance of integrating what is known about these approaches with professional judgement.

Because selection approaches fall short in their ability to account for the unique characteristics of students, teachers must account for factors such as students’ prior knowledge, their cultural and linguistic background, and their language learning environment. Combining theoretically driven approaches and professional judgement may reduce uncertainty in word selection while also promoting more effective use of instructional time.

Questions we just don’t have the answers to:

1. Are vocabulary words from established word lists better matched to students’ instructional needs than words selected by teachers who better know their students?
2. How confidently can the suggestions provided in word learning approaches be extended to word selection in later elementary instruction?
   1. Secondary instruction?
   2. Discipline-specific instruction?
3. When vocabulary interventions are successful, how much of this success comes from the words selected?
   1. How much comes from the instructional method?
4. How do different word selection approaches affect the development of shallow versus deep word knowledge?

Nitty-gritty of some of the research:

Bass & Barron, 2014

Participants

• Four typically developing children (two girls, two boys), attending a preschool primarily serving children from low-income environments, in a metropolitan region of the U.S.

Study Design

• Over a 12-week period, children were assigned to alternating treatment conditions. In one condition (instructionist-led), children participated in an interactive, adult-led read-aloud, and in the other condition (automated), children listened to a pre-recorded narration via headphones. Over the 12 weeks, children were taught 36 new words that were embedded into the stories they heard, and were tested on their depth of vocabulary knowledge using a decontextualized definition task. Target vocabulary words were primarily chosen from Coxhead’s Academic Word List (AWL; 200), however, Tier 2 words based on Beck and colleagues' tiered approach were used when the AWL words did not fit the story context (2002).

Findings

• Overall, children learned more words and demonstrated a greater depth of vocabulary knowledge in the instructionist-led condition. Despite this advantage, the automated condition had a higher preference rating among the children. Children with higher vocabulary scores (based on their performance on popular standardized vocabulary assessments) learned more new words compared to their peers with lower scores; however, unique patterns of vocabulary growth were observed for all participants.

Notes/Limitations

• The generalizability of the findings from this study is limited by its small sample size and use of single-subject design. Also, although this study selected vocabulary words using two selection approaches (i.e., AWL and Tier 2), this study did not compare these approaches or examine whether either word selection method affected student’s ability to learn new words.

Arthur et al., 2016

Participants

• 278, pre-kindergarten to third-grade students in four different states across the U.S. 82% of the children were White, 7% were Black, 5% were Asian, 8% were Hispanic or Latino, 3% were American Indian, and 5% were other races. 11% were reported as having disabilities, and over 54% of the samples had mothers with a bachelor’s degree or higher.

Study Design

• The effects of a single-dose, double-dose, and business-as-usual condition were compared to examine the effects of vocabulary instruction dosage. The instruction was embedded in the Let’s Know! language comprehension instructional supplement over a 21-week period. In the single-dose condition, teachers completed one lesson per week, targeting 4 academically relevant words per unit. In the double-dose condition, students received the information from the single-dose and a supplemental vocabulary support lesson with additional activities meant to deepen children’s vocabulary knowledge. Words were selected based on several criteria including: 1. relevance to the overall unit focus; 2. qualifying as Tier 2 based on Beck and colleagues' tiered approach (2002); and 3. the ability to form words from the target word (i.e., differently, differentiate, differ).

Findings

• The single-dose and double-dose conditions generally produced similar gains in vocabulary learning for the target words and outperformed the control condition when looking at all grade levels. Grade level differences did exist when comparing the pre-test and post-test results of each condition, and effects of the double-dose condition were most beneficial for children in kindergarten. The double-dose condition produced a slightly larger effect size; however, when looking across grade levels, this effect was minimal.

Notes/Limitations

• Though this study used several criteria rooted in theories of vocabulary acquisition to inform the selection of target words, it did not evaluate whether this influenced gains seen in vocabulary knowledge. Furthermore, the results of this study suggest that increased instructional time may not be the most efficient use of instructional resources, as there is a large degree of variability that still must be accounted for.