Learning to read requires many building-block skills such as phonological awareness and alphabet understanding. What is not as widely acknowledged is that reading comprehension, an even more complex process, also requires different building-block skills.
One model of reading comprehension proposes that understanding what we read is really the result of 3 levels of skills:
- Literal comprehension
- Inferential comprehension
- Evaluative comprehension
In a recent study in Reading and Writing, An Interdisciplinary Journal, researchers set out to test the validity of this model of reading comprehension. They developed a 20-item multiple-choice test to analyze students’ performance in these 3 areas of reading comprehension and to look for empirical evidence of the theory. Researchers believe that using this model of reading comprehension and assessments that evaluate the different skills could be used to guide instruction.
More than 2,400 5th grade students from 8 school districts in Oregon took an untimed, computer-based reading comprehension assessment in the fall, winter and spring. Students read a fictional narrative passage of about 1500 words and then answered 20 multiple-choice questions, Including 7 that assessed literal comprehension, 7 that assessed inferential comprehension and 6 that assessed evaluative comprehension. Literal questions asked students to identify a specific event from the text, inferential questions asked students to infer implicit meaning from the text and evaluative questions required students to evaluate a situation and make a judgment.
“The idea that there are different levels of reading comprehension, each of which imposes different cognitive demands on the reader and requires varying levels of interaction with the text, is not new,” the researchers write.
“Although literal comprehension is undoubtedly important (without surface-level understanding of a text, deeper interactions with the text are not possible), those designing and providing instruction and developing tests must also recognize that literal understanding is a stepping-stone to more advanced comprehension skills that must also be examined to continue to see growth in student performance.”
Results from this study support the 3-level model of reading comprehension, researchers report. First, as predicted, students had less difficulty with items on literal comprehension than inferential and evaluative comprehension. Secondly, performance in the domain-specific factors—literal, inferential and evaluative—accounted for anywhere from 3.15 % to 4.74% of the variance in students’ scores on multiple-choice comprehension over and above a general reading comprehension.
One criticism of the 3-level model is that it seems to assume a linear progression between the levels. Critics point out that the relationship among the levels may not be hierarchical, but the different skills may work in conjunction with and in support of one another.
In this study, researchers found that although, on average, test items on literal comprehension were easier for students than those on inferential or evaluative comprehension, they can also be more challenging. Items can be easy and challenging within each level of comprehension.
The 3-level model of reading comprehension and the use of test items to evaluate each of the 3 levels could better inform instruction than many current assessments that do not differentiate among types of comprehension skills, the researchers write.
“Examining the structure of reading comprehension: Do literal, inferential, and evaluative comprehension truly exist?” by Deni Basaraba et al., Reading and Writing, An Interdisciplinary Journal, Volume 26, Issue 3, March 2013, pp. 349-379.