Article 3

Microcomputers as Powerful Tools for Naturalistic Inquiry

by Randy Yerrick, 353 Flanagan Hall, East Carolina University, Greenville, NC 27858. (919) 328-6736.

E-mail: ryerrick@curie.uncg.edu

Abstract

There has been much discussion in the camps of science educators on how to create scientific communities in today's school science classrooms. State and national reforms recommend such a view of "science for all Americans [1] [2] [3]." Traditional tools used to examine classrooms are hardly adequate for examining and understanding efforts towards current reform visions. This researcher offers insight of one approach steeped in current technology to aide a teacher/researcher with naturalistic methodological knowledge to begin to make sense of the volumes of videotape he/she has collected. This paper discusses one treatment of such data, its implicit values, and recommendations for technological set up for maximum efficiency and strength for claims made from such a treatment.

The need for alternative and competing frameworks

Educational research organizations are departing from a singular empiricist paradigm as issues surrounding the understanding of teaching and learning become more complex in contemporary classrooms. It is not a choice for bettor or worse but the realization that a monolithic research perspective will not likely lead to a deep understanding of a human endeavor like education. Broad generalizations are often demanded of research for the purposes of sweeping national reforms [1] [3]. However, such claims may tax our understanding of these institutions which are imperfect by their very nature. Obviously these human, social, and cultural meanings are not necessarily objective or rational.

Other disciplines are the focus of such study as well. Science, for example, for centuries was judged solely on the basis of its pristine, rational, and privileged end products and public claims of objective methods [4]. A recognizable limitation to the general public's perception of science is that inferences came as a result of seeing only what scientists wanted the public to see. Recent qualitative studies suggest scientists operate within a context that has been profoundly misrepresented [5] [6]. Within the past two decades researchers studied scientific communities in which scientists were noted to be heavily influenced by irrational human motivations and interests like competition and bias. These less flattering portrayals that starkly contrast an empiricist or objectivist interpretation of science corroborated Kuhn's [7] argument that scientists recognize only data that fits into their traditional paradigm. More damaging was the heavy influence social pressures like competition, greed, status, and skepticism played in the treatment of ideas. Rarely, if ever, does the more rational framework of interpretation get adopted within the scientific community on the sole basis of its objectivity or rationality. This is especially true if it is politically maligned within the ranks of power of scientific communities. Instead, censorship and excommunication prevail. Why else would it have been necessary to rediscover oxygen at least three times [7] before it began breaking down the phlogiston paradigmatic thought ? It has led to some naturalistic researchers to conclude that an idea is considered rational within a particular community and not solely on the basis of its validity, or reliability. Rather, it is based upon whether every community member agrees that it is rational [5] [8].

How interpretive frameworks can aide our understanding of education

Historically, educational studies have represented a narrow paradigm and research focus. For example, the process/product research of the late 60's and 70's presumed that an accurate portrayal of classroom contexts, achievement, time on task, and other important factors influencing teaching and learning could be attained through observational checksheets amd surveys. This paradigm has been found deficient as a tool for understanding many contemporary research foci including classroom values, beliefs, and culture. As a result, education, like science, can be better understood through competing constructions of reality which contrast an empiricists treatment of the world. Standardized testing, one form of an empirical treatment of knowledge, has done little as a framework for interpreting academic failure or prescribing change. Perhaps this is some states' motivation to extract current testing from the ranks of public schools. Why should researchers expect to walk into the context of a classroom in Eastern North Carolina with a ready made framework for interpreting the important events if, earlier frameworks progagated from a traditional paradigm has made little impact or offered little insight? It seems, more than ever, that there is a need to understand what is happening in troubled schools from different perspectives.

Differences in qualitative treatments of the world

How does a study using reported surveys of massive numbers and broad and deep generalizations differ from that of an ethnography which generates a rich understanding of a single context? In naturalistic inquiry an important distinction is how truth or the validity of the report is measured. Contrasting quantitative studies, truth is decided on the basis of internal validity and whether the subjects recognize the claims made. Credibility departs from quantitative validity in that subjects of the study need to recognize the "truth" of the observations and generalizations regardless of the frame of reference used make generalizations or hypotheses [9]. Likewise, dependability of a naturalistic study contrasts quantitative reliability in a test of consistency. The challenge is that evidence must be provided for the reader as to how similar respondents in similar contexts would produce similar results. This is not the same as the reliability or stability of an instrument as defined in quantitative research. In contrast, dependability does not assume that there is minimal variance, but that the variance is attributable to a particular event, revision, or collective shift in the reality of the context [10]. For the naturalistic researcher it is a serious responsibility to record one's own participation within the context of the data as a participant observer. Within this interaction resides the type of beliefs and frameworks applied to the situation, and a record of how the study was influenced chronologically by such influences. Dependability, as such, is not from within the instrument but measured from outside by someone who can audit and demarcate how the researcher arrived at their final analysis of the events studied.

In this way validity and reliability have their distant relatives in credibility and dependability respectively. Similarly, in qualitative research, generalization departs from the empiricist research perspective. Most naturalistic inquirers maintain that no true generalization is really possible as all observations are defined by specific contexts [11]. Transferability of the conclusions, however, lie in contrast to generalizability claims made from vast and varied data sets. The responsibility of generalizability is then left squarely in the hands of the reader. Due to the richness of the context and hypotheses made in naturalistic inquiry it aides researchers to glean tools and insights to generate their own working hypotheses in other contexts.

Some implicit values of naturalistic inquiry

To accomplish these means, a researcher should maintain several key ingredients to assure that the report rings true to the subjects of the study and that it has internal consistency. These include: 1.) broad an purposeful data collection 2.) prolonged engagement with the data, 3.) persistent observation in and out of context, and 4.) triangulation of perspectives. Maintenance of these values helps the researcher to observe and understand routines and predictable daily events; to see through the eyes of natives; to elicit and compare alternative constructions from many sources; and, weigh the most useful framework describing the most data. Often categories which emerge from this process do not fall in line with original expectations or even the same body of research literature. The bonus is that these categories represent constructions that describe a more rich context and mark a path for deeper understanding comparable with the interpretations of the subjects within the context.

Criticisms of naturalistic inquiry

For such a perspective to be practiced, time needs to be focused on getting a grasp of the cultural milieu. Much like the work of Margaret Meade or Dianne Fosse this type of research demands of the observer that they lay aside past beliefs wholly inapplicable to the context observed. Consequently, a blatant limitation of this type of research is the amount of time that must be spent with the data in the process of gathering, revisiting, developing hypotheses, and testing interpretations. Such a process does not necessarily lend itself to institutions of higher education, research organizations, or other venues operating on the basis of rewarding certainty and objectivity of claims, the number of pieces of knowledge produced, or generalizability and usefulness of such products to vastly different knowledge consumers.

Another critique of this type of research is that it is too subjective. Such a perspective take issue through an interpretation that categories simply emerge as if characteristics and influences of the observer are irrelevant. As Kuhn [7] and others have revealed, even the communities which make the broadest empiricist claims are persuaded by the lenses they use to view the world. Claims of inordinate subjectivity is addressed by naturalistic researchers through an audit trail that chronologically and systematically documents when and how recent findings, interpretations, and views have influenced the report. Naturalistic researchers then do not simply look for obvious items that mysteriously emerge from the data but admit their bias and put on the table for all to examine the types of frameworks they have introduced to the data and at what time. These audit trails are to insure that the researcher did not simply find what they set out to look for.

Videotape transcribing devices are tools of the trade

A highly respected educational researcher was once asked what the single greatest contribution to his understanding of teaching and learning. His response after much contemplation was, "The videotape." The videotape presents a systematic way in which a qualitative researcher can collect and analyze data. The problem then arises as to how one decides what to do with video tapes once they are collected.

Digital movies, while easily sequenced and searched, like laser disk programs, have not yet paralleled the ease of gathering data with the 8mm mini camera. In addition, expensive and specialized production hard drives may be able to store vast amounts of digitized data but standard hard drives are not yet cost efficient for storing large volumes of audio-visual (AV) data such as classroom interactions, interviews, or small group settings. In order to use video tapes in research, fundamentally a linear mode of data collection, one always needs to decide on some strategy for videotape management and analysis.

Microcomputer technology can assist with these tasks and allow the researcher to make reasonable progress of knowledge production through the drudgery of observation ad-infinitum. As many naturalistic studies have revealed [12] [13] [14]. Microcomputer technology is no longer a venue for simply compiling numeric representations of the world and analyzing relationships outside of the beliefs that generated them. In the age of AV computers, contexts can be revisited, transcribed, and understood by outsiders through a variety of means. The tools needed by the naturalistic researcher to analyze extensive volumes of videotape libraries may include:

1. an AV desktop PC for viewing and capturing data, complete with video board for importing NTSC (e.g.: MacIntosh Power PC 7200-66av or IBM PS 350 or greater is optimal). While S-video is available on many cameras and players, universal RCA jacks are testament to NTSC's acceptance as the standard format for transfering audio and video signals between VCR's, monitors, laser disk players and other devices. Importing a video signal through a video board and running a video capture program will allow the user to monitor and even perform single and multiframe captures to the hard drive. While these frames and movies do not serve for viewing complete events, they do serve as important pointers, icons, or guides to further inquiry.

2. a video recorder equipped with a Control-L or LANC compatible real time VCR which allows COM port communication of real time information (e.g.: Sony SLV-R1000). This feature allows 2-way communication between the computer and VCR. While Control-S and Control-P enables communication between other VCR's and remotes, this feature is used exclusively in this system to relay real time data that corresponds to the tape location.

3. a software package that allows real time VCR and Camera controlling program, which mediates the 2-way communication using LANC for stamping and recording elapsed tape time (e.g.: C-Video v1.5). Important events can be located and scripted by their real time information on a computer document that can be edited like any word processing program or used for retrieving precise tape locations.

4. a database program that allows the importation of text, numeric, graphical, and real time event tags of a variety of data. (e.g.: Filemaker Pro v2.1)

Another possible addition to this quartet is an organizational software package that allows relationships between data and frameworks to be correlated and shown by graphical representations (e.g. NUD-IST v3.0.5c [15]).

Methodological Considerations

Broad and Purposeful Data Collection

Six guiding words for understanding a contextual setting are, "look carefully," "look carefully," and "look carefully." A serious and confounding error many beginning naturalistic researchers make is presuming that they know what they are looking at or looking for. In contrast, prominent leaders of this field caution against applying one's own set of beliefs to those of the context observed. The first step toward understanding any context is to acquire or generate many handles through which the observer can interpret the data.

One prominent feature of this quartet is the generation of real time transcripts of videotape. This allows the generation of a document that categorizes, times, and collects all observations entered in an overview or detailed transcript ( Figure 1). C-Video [16] is a software program that facilitates the viewing, cataloging and retrieval of information stored on video tape. Using C-Video, segments of the video tape are "tagged" with both beginning and end times creating a clip to which descriptive text can then be added. While some devices and software operate by burning a time or bar tag into the tape, C-Video records the calibrated time displayed on the VCR and numerically enters it into the open C-Video document. Tagging an important event is as simple as a mouse click or keystroke to remind yourself to return to that event later. This electronic document in real-time corresponds directly with the elapsed time of the VCR and replaces the need to store further information on the tape or risk damage beyond typical replay wear. The document, not the tape, is edited in future work and this document can then be used to review the tape either by selecting a tagged time or performing a keyword search.

After stamping the desired portion of the video, text can be added to describe, catalog or code that specific segment or clip. The text can include self-generated categories, names, quotes, descriptions of settings or partial transcriptions. This text becomes a description of the tagged times as well as a record that is searchable using categories or keywords you have selected. Once the C-Video document has been created it can be used to search and review the video-tape to which it corresponds. Stamped events can be viewed, edited, searched by text, time or category, and all VCR controls like fast forward, play, pause and rewind can be operated using the mouse. C-Video documents can also be searched using keywords contained in the descriptive prose and assembled within a new document. In this way viewing can become an act of creating categories that represent recurrent views, characteristics, or attributes of contextual events. These categories can later be used to chunk data and generate grounded theories. Grounded theories, which act as larger interpretive frameworks, do not precede but emerge from the data [17] [18].

Prolonged Engagement with Data and Persistent Observation

Rich descriptions created about a variety of contextual attributes (e.g. transcripts, journals, scanned classroom artifacts, screen captures) can then be used to make generalizations across and within the data set. Glaser and Strauss [17] described one such method for collecting descriptive data and utilizing them to construct working hypotheses and generalized relationships that would be otherwise dismissed as isolate events by the myopic observer. For example 3x5 cards with specific bits of information could be sorted and placed in provisional categories with similar attributes. Once a sizable collection of cards is accrued, the researcher writes a statement that best describes their thinking about the criteria for including or excluding a future event. As Glaser and Strauss described,

"...the constant comparison of the incidents very soon starts to generate theoretical properties of the category. The analyst starts thinking in terms of the full continua of the category and its conditions under which it is pronounced or minimized, its major consequences, its relation to other categories, and its properties." (p. 106) [17].

The creation of cards can be simulated in a more powerful way by using this quartet of technology. Databases can be constructed and linked through C-Video so that searches can be made on videotaped data by time, topic, venue, or even graphic displays of the classroom organization (Figure 2). All this can be performed through normal database functions that otherwise would relegate videotapes of classroom events separate from any other media or data. In the end, all of this can be described and entered into designed attributes for the purpose of checking the internal consistency of grounded theories.

Such a design can also reduce the number of hours a particular lead researcher needs to devote to a project as several different individuals can work in the process of stamping videos and describing classroom artifacts. The more general information can then be viewed more systematically at a later date. It allows workers who are relatively unfamiliar with other portions of the data to log and volume large video with time cues and simple text like "04:31 John talks" or "00:11:35 Classroom disruptions" that can later be used as starting points for future hypothesis generation. An informed participant aware of the treatment of other data collected in the project can begin to ask what attributes surround class disruptions and pose working hypotheses about accepted values within that context.

Part of the process of generating rich descriptions is the creation of multiple ways of observing and documenting views of the same data. A coined phrase, "making the strange familiar and the familiar strange" has been attached to this type of research. A database provides a venue for returning to the same data to check for internal consistency of the categories generated. For example, if a researcher believes that students were collectively acting in a hostile manner towards another student, but cannot understand the slang associated with the event, the researcher can turn down the volume completely which would allow him/her to focus solely on the non-verbal cues being used in establishing the mood of the class and status of the students. Turning off the video device but stamping by purely audible cues would also simulate the compensation of visually impaired observers toward poignant cues distracted by human's reliance upon visual response. In the same way, if an integrated database of stamped video times is rich with a multitude and variety of event characteristics, then searches can be performed using a range of descriptors and artifacts.

Triangulation of Perspectives

There are always events which do not fit the original creation of categories. In qualitative research these are not explained away or covered by a statistical treatment of the norm. Instead, these are often powerful tools for checking the consistency of working hypotheses throughout the data. Categories are expected to evolve and change, and the glitches in expected events often point directly to the limitation and direction or revision for future categories. For example, in a recent study interviews were conducted in which teachers were asked before and after a major reform institute about their curriculum and their interpretations of the two week institute [19]. While initial interpretations of the interviewers focused on curriculum topics and assessment, a subsequent framework was able to make more generalizable claims that teacher beliefs were jading their interpretation of the institutes message and their subsequent plans to follow through on any reform recommendations. An audit trail revealed that this alternative interpretation was adopted when the constructs used were not consistent with teacher beliefs regarding assessment and students.

Diagrams and maps can be produced from the categories assigned to hold and track data that graphically represents their relationships and that also serve to demonstrate how different data sets and findings are compatible within a subset of categories generated (Figure 3). A graphical organizing software package like NUD-IST can accentuate the video management system by allowing relationships between data and frameworks to be correlated and shown through graphical representations. C-video documents can be directly imported into NUD-IST software. Documents can also be checked against other frameworks of interpretation as it is likely to change throughout the research. This additional qualitative tool can display relationships that otherwise remain buried in the complexity of other interpretations. The ability to test hypotheses for their consistency cannot routinely be maintained mentally and hard copy production of these tests on 3x5 cards can be cumbersome. It is a process however expedited by this technology that aides in the checking of hypotheses against potentially divergent perspectives and potentially refuting data subsets.

Implications

The goal of creating emergent categories for grounded research is to make explicit contrasting beliefs and contextual influences that depart from knowledge of the data provided at the onset of a research project. Researchers of this perspective seek to challenge traditional beliefs and views that are often imposed onto data at the cost of understanding thoroughly the phenomena observed. Kuhn and others have reminded us that, though solidly grounded and long accepted, any lens can serve as a blinder to important information. Through the work of naturalistic researchers have unpacked more clearly events that evaded a rich and consistent treatment and have plagued human understanding for centuries. One popular venues of study is that of the difficulties in communication between men and women. Likely observed for millennia but rarely understood, explanations for intergender conversant difficulties have eluded researchers as such differences are too familiar and too recognizable to be seen in ourselves. However, Tannen [20] has enabled millions of readers to understand their own patterns by revealing inherent values in discourse of others and applying new naturalistic frameworks of study and interpretation.

The naturalistic researcher does not claim to be objective or to see things as they really are. It merely adds credence to the knowledge shared within a context and report to outsiders the nature of this insight. In their methods, naturalistic researchers attempt to document how, where, and when the framework used for analysis get recognized. That said, even quantitative research has been argued to find what it sets out initially to look for [7]. The naturalistic researcher differs however in their attempts to document and make clear origins of potential bias to leave the responsibility of generalizability to lie within the reader.

Each of these tasks are made more efficient through recent technological advances in AV processing [21], database management, and even hardware communications. In an era in which videotape is becoming a central medium for qualitative research, it becomes even more important to find ways to expedite arduous analysis tasks complicated by the linear mode of videotape media. Science education as a discipline is progressing towards a less conservative approach to examining learning as evidenced through more and more qualitative studies being produced. This technology quartet and similar setups are designed to forward the values of naturalistic inquiry and will likely change the appearance and substance of future research journals as it becomes more available.

References

1. National Science Education Standards (1996) National Academy Press. R

2. Michigan State Board of Education, (1991). Michigan Goals and Objectives: New Directions for Science Education in Michigan", The Michigan Center for Career and Technical Education. R

3. AAAS (1989) Science for all Americans: Project 2061. American Association for the Advancement of Science Report of Goals in Science Mathematics, and Technology. Washington D.C. R

4. Latour, B., & Woolgar, S. (1979) Laboratory Life: The construction of scientific facts. Princeton, NJ: Princeton University Press. R

5. Traweek, S. (1988) Beamtimes and lifetimes: The world of high energy physicists. Cambridge, MA: Harvard University Press. R

6. Fleck, L. (1979) Genesis and development of scientific fact. Chicago: Chicago University of Chicago. R

7. Kuhn, T. (1970). The structure of scientific revolutions. Chicago, IL: University of Chicago Press. R

8. Lemke, J. (1990). Talking science: Content, conflict, and semantics. New York: Ablex. R

9. Lincoln, Y. S. & Guba, E. G. (1985). Naturalistic inquiry. Beverly Hills, CA: Sage. R

10. Guba, E. G.(1981). Criteria for assessing the trustworthiness of naturalistic inquiries. Educational Communication and Technology Journal, 29, 75-92. R

11. Erlandson, D. A., Harris, E. L., Skipper, B. L., Allen, S. D., (1993). Doing Naturalistic Inquiry: A guide to methods. Newbury Park, CA: Sage. R

12. Allen, S. D. (1990). Principal teacher interactions in urban high schools: Two case studies. Unpublished doctoral dissertation, Texas A&M, College Station, TX. R

13. Dingwall, R., Eekalar, J., and Murray, T. (1983). The protection of children: State intervention and family life. Padstow, UK: T. J. Press. R

14. Jackson, T. F. (1991). The development of a training and support program for older custodial and landscape workers. Unpublished record of study proposal, Texas A&M University, College Station, TX. R

17. Glaser, B. G. & Strauss, A. L. (1967) The discovery of grounded theory. Hawthorne, NY: Aldine. R

18. Guba, E. G., and Lincoln, Y. S. (1989). Fourth generation evaluation. Newbury Park, CA: Sage. R

19. Yerrick, R. & Parke, H. (1996 In press). Struggling to promote deeply rooted change: the filtering effect of teachers' beliefs on understanding transformational views of teaching science. Science Education, New York ,NY: Wiley. R

20. Tannen, D. (1994). Gender & Discourse. Oxford, NY: Oxford University Press. R

Dr. Yerrick is an Assistant Professor at East Carolina University, 353, Flanagan Hall, School of Education, Greenville, NC 27858. His specializations include the study of shifts in classroom discourse, student learning, and microcomputer technology as applied to the learning of science.

Appendices

Figure #1

Figure #2

Figure #3

To get to the top of this page, click here.

To get back to the current issue of the EJSE, click here.

To get back to the EJSE's Archive page, click here.