One major problem in conducting research on the effects of a cognitive strategy designed to teach learners to derive the meanings of words from context is how to judge its efficacy. Researchers have generally compared learning-from-context treatment to so-called “control” participants who learned a list of specific words, or to “control” participants who received an unrelated treatment.
There are problems with both types of groups against which treatments are compared. Teaching specific word meanings versus learning from context. Comparing the teaching of a general strategy for learning from context to teaching specific words does not represent a fair comparison. Contextual analysis is not meant to teach specific words. Instead it is meant to be general strategy aimed at helping students contend with unfamiliar words in a wide variety of texts. The goal of teaching strategies to improve the learning of word meanings from context is to help students learn more words incidentally as they are encountered in everyday reading. Doing so should, in turn, lead to a larger vocabulary over time, as students read texts containing unknown words. If one reads a given number of words and encounters 50,000 unknown words over the course of the year, the difference between learning the meanings of, say, 5% of them and 7% of them will be the difference between 2.500 new words learned and 3,500 new words learned. If an instructor can teach roughly 300 to 400 words per year using direct instruction (1989) the 1.0000 extra words is 2 times as many words as could be taught directly. These estimates recommend that the effects of context-clue strategy training might be powerful, but only over relatively extended periods of reading.
On the other hand, if one teaches 10 words in a lesson and 8 of these are learned, then a student will know those 8 betters if we measured their profit in reading from a single passage. So, for a single passage or a series of passages, it is likely that direct teaching of a set of words will be more efficient than a generalizable set of strategies for learning from context. These were the findings of Jenkins et al. (1989). However, over time the 2% gain in learning efficiency, which might not be visible in a single passage, should lead to larger long- term gains than teaching specific words. Direct instruction of vocabulary does have demonstrable effects on students’ vocabulary learning and comprehension. Stahl and Fairbanks (1986) in their meta-analysis of vocabulary studies, found that vocabulary instruction did significantly improve not only understanding of passages containing the words taught, but also comprehension in general.
“No treatment controls?” Another problem identified in much of the research is that groups receiving instruction in the use of context clues were often paired with a control group that underwent no instruction or practice or were not paired with a control group of any sort. Any learning from context strategy has at least two components to it: learners must be made aware of the importance of focusing on unknown words, and they must be taught specific strategies to learn the meaning of those words from context. A “no-treatment control” (actually, elementary learners are doing some unrelated treatment) would not include the reminders to examine context, nor would in entail the specific strategies taught. Determine whether the effects of the treatment are due to advisement about the
Importance of learning words from context, to a greater sensitivity to unfamiliar words in context, or to the effects of the strategy taught.
To check the difference between teaching specific strategies or sets of context clues and clearly making learners more metacognitively aware of the need to pay attention to unknown words, one needs to compare an active-strategy treatment with a practice-only treatment, in which students are simply asked to derive word meanings from context without any specific instruction in the process. As we noted antecedently in this article, when studies contained such a contrast researches failed to find a difference between the strategy group and the practice-only groups. Sternberg found that both the strategy training and the practice-only controls did significantly better than the controls who memorized definitions. This finding strongly suggests that the effects of context-clue training may largely be due to making children more aware of unknown words in their reading, their reading, rather than as the result of the specific strategies being taught. Elementary learner’s language researchers know the difficulty in separating learners’ knowledge or language, including vocabulary,
from their metalinguistic knowledge or the ability to talk about what they know. The goal of teaching learners more efficient means of learning from context is to increase their store of vocabulary words. This knowledge is often implicit and subconscious. When we ask learners to verbalize how they learn from context or attempt to judge the quality of their learning by their production of a conventionalized from, such as a definition, we may be assessing students on how well they can use language ( either the verbalization of processes involved in learning words or the use of definitions or other school-based ways of talking about meanings) rather than on how they have actually learned about word meaning. This is a problem in virtually all vocabulary research (see Anderson Freebody, 1981), but is especially apparent concerning the use of context.
Deriving words versus incidental learning. Nearly all of the studies have also examined the effect of various contextual strategies on deriving word meanings. On such a task, a student is given a passage and asked to derive the meanings of a word or words identified. Being asked to derive word meanings is a different task from using context to acquire the meaning of unfamiliar words during free reading. In a derivation task, the participants are told that the word is unknown or a nonsense word is substituted. In the more incidental learning of word meanings during free reading, readers are often unaware that they do not know certain words in the text, and thus they do not apply special attention to those words. Instead, Stahl (1991) speculated that incidental learning may be explained using a connectionist metaphor. In this metaphor, a newly encountered word is weakly connected to both an orthographic representation and other concepts which occurred with it in the passage in which it was encountered. As the word is encountered repeatedly, some links to related concepts are strengthened and others weakened, until the person gains a functional knowledge of the meaning of the word. This meaning may not be a dictionary definition, not does a person who knows a word need to produce such a definition. Intentionally deriving word meanings from context requires attentional processes which may or may not distort this natural learning. In other words, people who are able to deliberately derive the meaning of a word from a specific context may not actually do so when reading outside of the experimental context. The difficulty with using a derivation task is that such a task may simply be tapping an ability that students already have. Because learning from context is a natural process, as well as the way in which we have learned most of what we know (1989), teaching learners to derive words from context may be superfluous that is, teaching then something they already know how to do. Because children already do have some prowess in learning words from context and must have implicit strategies to do so, it is possible that this learning from context strategies may merely allow learners to verbalize better what it is that they are already doing. If so, these learners may do better on a measure that requires verbalization of this ability, but may not show any overall improvement in the skill. In other words, they may be better able to talk about what they are doing but may not be able to do it better.