Ebjournal Revisited.

Mock Proposal for Thesis

NO ABSTRACT JUST YET… possibly coming soon!

Inference Generation in Primary Classroom Readers

Introduction
Background Information
Children come to classrooms with preconceived notions that they will be instant readers. They are excited to see reading in action, to practice it, and become experts. For all but the most gifted of readers, time is necessary to learn the processes involved in decoding. As students become familiar with sounds, the letter-sound relationship, and fluency, they begin not only to read the text, but take meaning from it. When readers derive meaning from a text, it is commonly referred to as comprehension. Comprehension is often regarded as multi-level constructs of text, and is aided by the activation of background knowledge. (Pearson, P.D., & Anderson, R., 1984; Graesser, A.C., Singer, M., & Trabasso, T., 1994) In the lower levels of comprehension, students first decode words, and then understand their representations into sentences. This generates local meaning. Children may also find explicit answers in the text, and focusing on the local information. The higher order of comprehension skills occur during inferencing. An inference is widely defined as any information gained or processed that is not explicitly defined in the text. (McMackin, M., & Lawrence, S., 2001; Cain, K., Bryant, P., & Oakhill, J., 2004; Graesser, A.C., Singer, M., & Trabasso, T., 1994; McKoon, G., & Ratliff, R. 1992). Inferences fall into several categories, including predictive, explanatory, and causal.
Though all inferences require higher order thinking, there is debate over how students access and use inferences. In the minimal hypothesis, students must have explicit strategies, and inferences are not automatically processed. Students must have strategies to access inferences. There is debate over what memory processes are working when using and accessing inferences, especially in the different levels of comprehension. The minimal hypothesis states that local inferences (also known as local coherence) are processed automatically, while global coherence has to be taught. (McKoon, G., & Ratcliff, R., 1992) In contrast to that is the constructionist theory, which states that inferences are built from the text up to higher order thinking, and that some inferences are generated on-line (during the course of comprehension) and some are generated off-line (retrieving it later from another task, but not during comprehension.) (Graesser, A.C, Singer, M., & Trabasso, T., 1994.) All reference generating deals with short term memory, working memory, and long term memory. The relationship between these, and the processing skills needed for inference is a current debate. (Graesser, A.C, Singer, M., & Trabasso, T., 1994.)
Additionally, an initial review of literature demonstrates that the foci of studies of inference generation lean toward students in the intermediate grades, who are nine years old or older. Several studies have been conducted to ascertain the rate of inference generation in older students, and adults, but the natural occurrence of inferences has not been fully explored.
This study attempts to explore how younger students generate inferences during the course of learning to read. Are they natural, via the constructivist theory, or is explict teaching of strategies that allow them to produce inferences? The author will work with students ranging in ages six years, one month, to eight years, ten months to ascertain what types of inferences are taking place during reading of appropriate level narrative texts. Before students are formally introduced to inference generation via explicit instruction, it is possible that inferences are being naturally generated. Using previous research on upper elementary students and adults, the author will observe when and where natural inferences are made, without prompting or modeling from an instructor. Discussions and interviews will be transcribed, and follow standard protocols.
Statement of Problem
Students working in upper elementary classrooms (grades 3-8) are taught comprehension skills that include explicit instruction. Though there is much debate as to the cerebral resources used in the processing of inferences, students regularly generate them in a variety of settings. As mentioned in the review of literature, extensive research will be performed to establish how and when inferences are generated in the course of group discussion and independent reading. With the exception of initial inferential research provided by Pearson and Anderson (1984) the author has been unable to find a wide discourse as to the natural inference generation of primary students. The review of literature will continue to explore what research exists, and determine the framework for the current study. With a framework established, it will then be possible to regressively observe students in a different population; that is, younger chronologically, and still in the emergent stages of reading to discern whether they are naturally generating inferences.
Research Questions
Identification of inferences becomes the focal point of this study. Do students generate inferences? The study will attempt to establish if inferences are being generated naturally by second grade students within the discussion of text. Additionally, it must be investigated that if inferences are being generated, in what categories can they be indexed? Are there trends in the data obtained from observation of students’ discussion of text?

Operational Definition of Terms

Inference: information gained that is not explicitly defined in the text. (Graesser, A.C., Singer, M., & Trabasso, T., 1994.)

Global Inference: According to McKoon and Ratliff (1992), a global inference “connects widely separate pieces of textual information and that they do so automatically as a necessary part of comprehension.” (p. 445)

Natural Inference Generation: For the purposes of this study, natural inference generation will refer to any inferences made by a student without being directed by a peer or instructor.

Comprehension Age: the age designated to a subject as determined by the Neale Analysis of Reading Ability. (see NARA) Comprehension age convert the corresponding chronological age of a reader based on a student’s ability to process and comprehend text.

NWEA-MAP: Northwest Evaluation Association- Measure of Academic Progress. A computerized reading assessment tool used to determine a student’s level of comprehension by finding the range at which students perform at a consistent level.

NARA: the Neale Analysis of Reading Ability, a reading test that scores comprehension, and demonstrates corresponding reading age. (McGee A., & Johnson, H., 2003)

Assumptions and Limitations

It is assumed that students will have rudimentary decoding skills. For the purposes of this study, they will defined as having the ability to that place them with a reading and comprehension age of no lower than 6 years, 10 months, and no higher than 10 years, 0 months. Students participating in the study will be native English speakers with no learning disabilities or affects that may inhibit the natural generation of inferences. The sample of students will be limited to a convenience sample, and will also be limited to a second grade classroom. Because of the nature of the study, final results may be in case study form. Also, teacher interactions with students, and the selection of texts by teachers will not be assessed. The goal of this study is to determine if and how students are naturally generating inferences.

Review of Literature

Three key spheres of influence will determine the research of the literature for this study. The nature of the study dictates that definitions be clearly defined so that the author can illustrate what inferences are being generated by students, and which ones will occur because of teacher influence. The works of Pearson and Anderson (1983, 1984), Trabasso and Graesser (1994), Trabasso and Suh (1993), McKoon and Ratliff (1992), and others will be studied to establish current theory in terms of the cerebral processes of generating inferences. Additionally, research trends in the identification and use of inferences in upper elementary students, secondary students, and collegiate students will be used to establish how inferences are generated in different populations. The author will determine what role text plays in inference generation in students of all ages. Finally, the literature will be surveyed to discover what studies have been done to assess primary students and establish possible gaps in research.

Methodology

Purpose of Study
The purpose of this phenomenological study will be to discover what inferences are being generated by primary students who have had no formal inference training, within their general reading comprehension discussions. The study looks to establish whether students automatically generate inferences, and at what level (global versus local, or both). This will be examined from an observer’s point of view, in that students should not be influenced in their response. To this point, inferences refer to any information gained while reading that is not explicated defined in the text.

Research Questions

Within a second grade classroom, and using observations of student interaction of text, this study will explore the nature of inference generation in primary (pre-K – grade two students). Do students generate inferences without explicit instruction? What inferences are generated by students? Into what categories of inferences (local/global, explanatory/causal/predictive) do they fall? This study will notate all inferences generated, but will focus on global inferences, and identify inferences generated within this category.

Population-Sample

The author will work with students from a non-denominational, upper middle-class public school in northern Illinois. Students participating in the study will be determined to be of various reading abilities, based on standardized test scores. Ages of the student population will approximately range from 7 years, 5 months, to 8 years, 10 months. The study will focus on students working above their chronological age in terms of comprehension age, students that having corroborating chronological and comprehension ages, and students that are significantly (near one year) below their chronological ages as related to their determined comprehension age.

Instruments:

To determine the comprehension/reading age student, the NARA field test will be administered. This test relates age to comprehension. To supplement this data, students information from local standardized tests, including their Northwest Evaluation Association- Measure of Academic Progress (NWEA-MAP). Following the identification of students’ ability levels, students will be observed and recorded in the course of discussion of texts that correspond with their comprehension age and ability level.

Procedure/Time Table

Students’ results on the NWEA-MAP (winter 2006 results) will be collected. Students will then be given the NARA. Upon scoring of the NARA, and in combination with NWEA-MAP results, students will be placed into two three categories—high achieving readers (those students whose comprehension age is significantly above their chronological age), average achieving readers (students whose comprehension age is near their chronological age), and low achieving readers (students whose comprehension age is significantly below their chronological age.) Once established, students will be observed over the course of three weeks, in both small group and whole group settings. Students will be recorded in their discussions of text, and this discussion will be subsequently transcribed. From the parameters established, information on inferences will be organized into subcategories of types of inferences. At this point, the will continue to remain on if students are generating inferences, if there is consistency, and what, if any trends can be derived from the observations. Overall, the length of data collection is expected to last no longer than five weeks, including assessment of comprehension age.

Analysis

Once the information is collected via recorded observation, it will be organized and analyzed to discover what types of inferences are naturally generated. Specific focus will be placed on identification of global inferences. Are they more prevalent than local inferences? Within the context of global inferences, attempts will be made to identify trends in the data that suggests what specific types of inferences are being constructed; predictive, causal, or explanatory. The analysis will ultimately focus on not how students are generating inferences, but specifically what is being generated.

*****

Well, that’s the craptacular start. I’m guessing you didn’t read all the way down to here.