Description of Research Methodology

    For this investigation, I intend to use an observational, descriptive-analytic approach, with my own role as participant-observer. This method conforms generally to the "qualitative" approach to research described by Goetze and LeCompte (1984). After obtaining permission from the relevant schools, teachers, parents and students, I will observe science classes over extended time periods (i.e., 10-15 weeks), employ audio and/or videotape when allowable, and take extensive field notes of my observations. I will also return to these classrooms after an interval of 2-3 months to document the stability of my observations.

    Such procedures have been previously employed in studies of scientific reasoning in nonhandicapped students. For example, Fleer (1992) investigated the scientific understandings of young children and the conceptual change that occurred during science teaching. In order to accomplish this, Fleer collected one hundred hours of video tape and 190 hours of audio tape over a period of six months, in three science classrooms: a kindergarten class, a transition/year one classroom, and a year 2/3 classroom. Teacher's science lessons were videotaped and students interviewed about their scientific understanding before, during, and after the completion of specific science units.

    Fleer analyzed transcriptions of children's and teacher's discourse while engaged in scientific investigations in order to document successful interactional types that facilitated conceptual development in young children. She reported that the transition/year one teacher engaged in more successful discourse and more competently "scaffolded" conceptual change. Fleer characterized more successful discourse as that which was more "on-task" and which generated more learning outcomes, as revealed through discourse analysis. The results of the investigation were taken to support the importance of the interactive role of the teacher, to structure and promote student reasoning.

    As another example, Rosalind Driver (1989, p. 89, citing Brook, Driver, & Johnson, 1989, p. 70) employs student-student discourse to make the case that observational evidence (e.g., through experiment or activity) may not be sufficient in itself for students to reconstruct their scientific ideas. In one case, students were provided with a balance with a deflated balloon on each end. They were asked to predict what would happen if one of the balloons were inflated:

Daniel: Air's heavy, right. It's heavier, isn't it?

Joanne: No.

Daniel: It is, it is.

Ann: It's the same, air weighs nothing.

Daniel: Look, it'll go down -- air's heavy.

Jaspal: Look, listen! When we blow the balloon up it's gonna come down, isn't it, 'cos the air in the balloon's heavier and gravity pulls it down.

Joanne: Yes, but air's light, so how can it come down?

Ann: It floats so it'll stay the same (Driver, 1989, p. 89).

    One of the balloons is then inflated, but two of the students remain unconvinced that air has weight:

Ann: Hey, it's gone down.

Joanne: But what makes it do down?

Jaspal: Look, we're doing about air, right: It's heavier than normal air outside -- gravity pulls the balloon down.

Ann: Air's light, it'll make the balloon float.

Daniel: How come it came down then?

Ann: I dunno. I thought it'd stay the same.

Joanne: If it were light it would go up wouldn't it.

Jaspal: Look, gravity pulls it down -- it pulls the air down.

Daniel: Only when it is in the balloon (Driver, 1989, p. 89).

    Through analysis of such discourse, from this and other investigations, Driver gains an understanding of how some children conceptualize the concept of weight. That is, that weight is a property of solid objects only, that it is what makes things fall down when released, or press down on surfaces. These ideas may be consistent with a child's everyday observations, but are not consistent with scientific explanations. An understanding of such conceptualizations, Driver argues, is helpful in promoting conceptual change in students, rather than their simple acceptance of empirical evidence.

    As a third example, student-teacher discourse can be used to evaluate whether a particular scientific concept is accepted because it is stated by accepted authorities (e.g., teachers, texts, scientists), or because it makes sense to the individual learner in that it reflects the learner's individual experience and knowledge schemes, that is, that the statement is believed. Driver (1989) presents a particularly informative insight from discourse taken from her Children's Learning in Science Project:

P: I can't really explain, but there's summat where you think, well this table it's made up of particles -- I think it's too, well you can't see any particles or owt, so it's just -- just can't believe it. You know, that this table's made out of particles -- hundreds of millions.

T: You don't believe it?

P: Well, I do in me own way, you know, but well wood's wood, I mean it grows from trees -- you know more or less -- well-- you think -- well fair enough it's made out of particles, but it's you can't really believe that this table's made out of particles.

T: What about the atmosphere in this room? Can you accept that that is made of particles?

P: Not really because -- 'cos you can't really see 'em. I mean, for all we know there could be particles, but in another way, for all we know it could be scientists saying that there's particles in the air and making us believe it. Well it could just be normal sky you know, because there is sky coming all the way down -- it could be sky -- you know in t'buildings and that (p. 103-104).

    Information from all the investigations described above can be employed to draw conclusions about students' understanding of science concepts and the development of this understanding in the presence of classroom science education. Similarly, I wish to collect information (primarily, but not exclusively discourse) that I can then analyze in order to draw conclusions about questions such as those mentioned above, relevant to the development of scientific reasoning in students characterized as learning disabled or mildly mentally handicapped.

. . . . . .

Method

Sample and Settings

    Since I wish to observe and describe the development of scientific reasoning in students with mild mental handicaps and learning disabilities, I will select classrooms which employ inquiry-oriented approaches to science education. I have identified some settings in the Greater Lafayette, Indiana, and Mesa, Arizona schools which use such approaches to science education, and with whom I have previously collaborated. I will primarily examine "self-contained" elementary science classes for students with learning disabilities and mild mental handicaps; however, if circumstances allow, I would also like to observe the science learning of students with mild disabilities in regular science classrooms. Throughout the year, I would like to observe in as many as two learning disabilities classrooms, two classrooms for students with mild mental handicaps, and two regular science classrooms in which students with mild disabilities are included. Since I wish to observe the entire social/academic classroom environment, I shall include in my observations all students in a given classroom who have been designated "learning disabled" or "mildly mentally handicapped" by their schools. I shall also collect as much referral information as possible on each student, so that I might evaluate any observed differences in development of scientific reasoning as a function of identified intellectual, academic, or social characteristics.

Procedure

    Data sources. Multiple data sources are necessary to validate observations and strengthen conclusions (Goetze & LeCompte, 1984). Data sources will include field notes, audiotape and/or videotape recordings of my participant observation; formal and informal interviewing of students, teachers, and other school personnel; and artifacts, such as student journals, written student products, student portfolios, teacher anecdotal records, and test protocols. I will also include, when allowable and relevant to the purposes of this investigation, information from student Individualized Educational Plans.

Data Analysis

    Data analysis in this type of investigation is inductive. Analytic induction "involves scanning the data for categories of phenomena and for relationships among such categories, developing working typologies and hypotheses upon an examination of initial cases, then modifying and refining them on the basis of subsequent cases" (Goetze & LeCompte, 1984, p. 179-180). Data will be assimilated and evaluated in order to develop hypotheses about the development of scientific reasoning in students with mild disabilities. Discrepant-cases and negative-cases will be used to further understanding and refine hypothetical constructs. For example, Driver (1989) employed several examples of student discourse from different activities to demonstrate how students can share ideas and experiences to bring their thinking forward. However, analysis of discrepant events and negative cases supported the necessity of appropriate teacher interaction, in the form of provided experiences, shared conventions and information from the scientific community, and carefully directed questioning.

    Observed events will be subjected to the constant comparative method, in which incidents, categories, and constructs are subjected to overlapping comparisons (Goetze & LeCompte, 1984). Finally, obtained and internally validated results will be compared with qualitative and quantitative descriptions of the development of scientific reasoning and understanding in nonhandicapped populations as well as populations with mild disabilities (e.g., Driver, 1989; French, 1989; Fetherstonhaugh & Treagust, 1992; Mastropieri & Scruggs, 1992).

    Validation. Internal validity of findings is supported by documentation of conclusions supported by multiple data sources, by analysis of discrepant or negative cases, and by validation of event comparisons. External validity of findings with respect to generalizability to other instances of science education for students with mild disabilities is more difficult to establish. However, Yin (1989) makes a convincing distinction between statistical generalization, such as that employed by survey researchers, and analytical generalization, which is appropriate for studies of individual cases (including experimental research when subjects are not drawn at random from a parameter). With analytical generalization, the investigator generalizes the findings of a particular case or set of cases to a theoretical framework, rather than to other, hypothetical cases. In the present instance, I intend to incorporate a "replication logic," in which multiple cases of science learning in different contexts generalize to the same (presently unspecified) theoretical framework. Such validation procedures are presently being undertaken in a descriptive-analytical investigation of the accommodation of a student with emotional handicaps in a mainstream science class (Mastropieri, Scruggs, & Bohs, in preparation).

Results

    Results of this investigation will be incorporated into a research report which will include three important features, as described by Erickson (1986). First, the report must allow readers to appreciate and understand the setting, the participants and the events that have taken place in the course of the investigation. This is accomplished by careful description of all relevant setting variables, taking into account how these may change over the course of instruction. For instance, in a presently ongoing investigation of science instruction of students with disabilities in mainstream settings, I have observed and recorded how a teacher's decision to allow students to rearrange their desks to their own liking provided implications (both positive and negative) for two students with physical disabilities who were attending that class. Such setting variables can interact in unpredictable ways with instructional variables, and must be carefully described for the benefit of readers as well as for the benefit of the investigation. Likewise, careful description of the participants and the classroom events that take place over time is necessary for full understanding to develop.

    The second important feature of the research report involves giving readers access to the entire range of evidence, to allow them to function as coanalysts of the investigation. The obvious dilemma in this feature is providing sufficient evidence for readers to function realistically as coanalysts, without overwhelming readers with a plethora of information. Information from the investigation must be provided in a manner that is thorough, yet succinct. This can be accomplished, I believe, in careful classification of events so that they can be easily combined, objective and thorough listing of negative cases, even (or especially) when they appear to conflict with the investigation's hypothesis, and justification of what is meant by a "typical" case. The latter procedure, I think, is often incompletely qualified in science education research, but can be more successfully achieved by carefully stating how a particular piece of evidence (e.g., from classroom dialogue, interview, or artifact) qualifies as "typical." If the argument is convincing, it will allow for a more parsimonious presentation of evidence. If, on the other hand, the argument is not convincing, it may suggest the author is selective in reporting, a situation that must be avoided.

    As a third important feature, the report must provide the theoretical and personal framework within which the report has been undertaken, so that the author's perspective can be easily understood by readers. In the present instance, I will describe my observations, results and conclusions, with respect to my own theoretical and personal framework, to the extent that I am aware of these influences. Information from my personal experiences, observations, research, and study of science education and the characteristics of students with disabilities has provided a particular framework for understanding students with mild disabilities and how they might develop their awareness of science concepts. This framework can be expected to undergo some change as a result of this anticipated experience. I will note how I perceive my own theoretical and personal framework, how it has interacted with observations, and how it may have changed throughout the course of the investigation. In this way readers as well as myself can gain insight into conclusions drawn from the investigation.

    Overall, readers of the research report will be provided with the purpose of the investigation; a clear description of setting, participants, and events; a clear description of the data collection and data analysis procedures; and a clear description of the conclusions of the investigation and the logical connections between empirical evidence and those conclusions. In addition, the connection between the findings of the investigation and an overarching theoretical framework will be explicated.

Importance

    Mastropieri and Scruggs (1992) recently reported that 66 empirical investigations of science learning of students with disabilities have been undertaken since 1954. However, only half of these investigations included students with learning disabilities or mild mental handicaps. Further, none of those investigations included careful and systematic descriptions of the course of development of scientific reasoning in students with mild disabilities, in inquiry-driven classroom learning contexts. The results of the presently proposed research is intended to add greatly to our present knowledge of how students with mild disabilities acquire important scientific concepts, and should contain important implications regarding the facilitation of scientific reasoning among students with mild disabilities.

    Review of recently conducted research has suggested that, contrary to popular contemporary paradigms in special education, students with mild disabilities can benefit greatly from inquiry-based science instruction, given that such instruction is appropriately sequenced and structured. Further, there is reason to believe that such appropriately sequenced and structured experiences can develop the reasoning abilities of such students when applied in relevant contexts. Such findings have positive implications for the characterization of special learning needs within the context of a scaffolding model of instruction (e.g., Cazden, 1988), in which such learners acquire scientific concepts under specific conditions of teacher support through instruction, modeling, and demonstration; guided practice with gradual release of responsibility; and finally, independent practice or application of skills and concepts.