English Language – To improve my writing
The link provided from Alan’s blog.
At glance I found it very useful.
Hope could improve my writing.
My aim
Understanding Personal Information Management :
The Comparison of User Strategies – Grounded Theory Issues
Azrina Kamaruddin
Submitted in part fulfillment of the requirements
for the degree of Doctor of Computer Science
December 2009
——————————————————————————————————————–
May my dream this dream will be true. Amiin.
Actions taxonomy
The complete write up refer to keynote Tamodia or blog.
We also perceive patterns of actions in the world, most significantly in the actions of other creatures. In a keynote at Tamodia in Pisa last year, I discussed the production of human action and the rich intertwining of sequential planned or proceduralised activity with more stimulus-driven reactions. Both could be explicit or tacit. The computer aid is trying to ‘read’ these patterns of intentional (and reactive) activity.
| pre-planned | environment-driven | |
|---|---|---|
| explicit | (a) following known plan of action | (b) means-end analysis |
| implicit | (c) proceduralised or routine actions | (d) stimulus-response reaction |
Why I need this:———————————————————————————————————————–
Refer to duming behavior or tacit activities while PIM.
LOW MODELS VER 1.1
These are the models created based on the analysis done for LOW COMPETENCY USER. The partial model is not done yet. There are few things that need to consider. Then the models is merged to have the whole big models for LOW user. This the version 1.1.
There are few things trigger during the model creation. Sometimes it is difficult to exactly make the model and somehow it look similar with another category.
Gt : Glaser and Strauss
This book is about discovering theory from data, what Glaser and Strauss call grounded theory. The major strategy they use is a general method of comparative analysis. They argue that much of current research is primarily the verification of theory or the development of theory through logical deduction rather than from the experimental data itself. Grounded Theory In sociology they propose that the roles of theory is to: * to enable prediction and explanation of behavior * to be useful in theoretical advance * useable in practical applications * guide research on behavior.
“Theory is sociology is a strategy for handling data in research, providing modes of conceptualization for describing and explaining” (p. 3). Basing theory on experimental data ties it to the data and makes it irrefutable. It reduces the opportunitistic use of available theories or the force fit of popular theories to the data. It avoids “exampling”, the often easy task of finding examples to fit any theory.
Glaser and Strauss take the position that the adequacy of a theory can’t be divorced from the process of creating it (p. 5). The purpose of their book is to describe a method and some tips on how to build theory out of qualitative data.
Verification and Grand Theory Most research today is designed to verify existing theories, not generate new ones. Researchers eek out small gains of knowledge from existing “grand theories” rather than explore new areas not covered by existing theories. The existing research culture emphasizes and reveres good scientific, quantitative verification studies and downplays more qualitative studies whose objective is theory generation.
Most theory is thus generated through logical deduction from past studies and knowledge and not from the data itself.
Part I: Generating Theory By Comparative Analysis In comparative analysis different groups or subgroups of people are compared and their differences build into theory. The theory is then tested and refined by considering it with other comparision groups. Looking at the data we begin to ascertain patterns in the data/behavior which can lead to general concepts about it. These concepts can then be built into broader theoretical propositions which can then be evaluated and tested with other comparison groups. “In discovering theory, one generates conceptual categories or their properties from evidence, then the evidence from which the category emerged is used to illustrate the concept” (p. 23). These conceptual categories can then be explored in other comparison groups, which may support the categorical concept or suggest modifications to make it more generalizable. In the pure pursuit of theory generation accurate evidence and verification are not that important, but often researchers get caught up in these two areas (because they are part of the current reward system) and thus limit the possibilities and narrow their range of potential theory generation. Theory generation doesn’t require lots of cases.
One case could be used to generate conceptual categories and a few more cases used to confirm the indication (p. 30). “(The reseacher’s) job is not to provide a perfect description of an area, but to develop a theory that accounts for much of the relevant behavior” (p. 30).
What Theory is Generated Grounded theory can appear in various forms. “Grounded theory can be presented either as a well-codified set of propositions or in a running theoretical discussion, using conceptual categories and their properties” (p. 31). The authors prefer the discussional form, because it is often easier to comprehend and tends not to “freeze” the theory in a set of propositions.
Comparative analysis can generate two types of theory — substantive and formal. Substantitive theory is developed for a specific area of inquiry, such as patient care, professional education, delinquency, etc. Formal theory is for a conceptual area of inquiry such as stigma, deviant behavior, formal organization, socialization, reward systems, etc.
It is often best to begin with generating substantive theory from data and then let formal theory or revisions to existing formal theory emerge from substantive theory instead of using logic to deduce substantive theory from formal theory. More studies generating substantive theory will ultimately generate and improve formal theory. Elements of the Theory The elements of theory generated from comparative analysis are first conceptual categories and their properties and second hypotheses or generalized relations among the categories and their properties.
“The constant comparing of many groups draws the sociologist’s attention to their many similarities and differences. Considering these leads him to generate abstract categories and their properties, which, since they emerge from the data, will clearly be important to a theory explaining the kind of behavior under observation” (p. 36). It’s best to let the categories emerge than to come into the study with pre-set categories based on existing theories. “An effective strategy is, at first, literally to ignore the literature of theory and fact on the area under study, in order to assure that the emergence of categories will not be contaminated by concepts more suited to different areas” (p. 37).
They also advocate that the generation of theory should initially aim at acheiving high diversity in emergent categories, at as many levels of conceptual and hypothetical generalization as possible. The concepts should be analytic and yet sensitizing, giving the ability to grasp references in terms of one’s own experience. Generating hypotheses from the data requires only enough data to suggest the hypothesis, not prove it.
During the research the emergent categories will begin to form patterns and interrelations which will ultimately form the core of the emerging theory (p. 40).
Chapter Three: Theoretical Sampling Building grounded theory requires an interative process of data collection, coding, analysis, and planning what to study next. The researcher needs to by theoretically sensitive as they are collecting and coding data to sense where the data is taking them and what to do next. Coming into a research program with an existing theoretical framework will merely blind them to the richness of the incoming data. As this iterative process continues, the researcher may explore the same group more deeply or in different ways, or may seek out new groups. Comparison groups should be selected based on their theoretical relevance to further the development of emerging categories. It’s best to pick the groups as you go along than choose them all beforehand — let the data be your guide. In theory generation non-comparability of groups is irrelevant, but it can have an effect on the level of substantive theory developed and it’s thus important to pick the right group for the next part of the comparative research. Sometimes it’s best to pick similar groups to gain sensitivity to differences between groups and to estabilish a definite set of conditions when a category will exist. Other times its best to pick very different groups which will maginify the strategic similarities and broaden the scope of the emergent theory. The researcher takes the role of an active sampler, pursuing leads and groups without worrying about being incomplete. The researcher will know when a given search is ended or the appropriate number of groups surveyed when no additional data can be found that develops properties of the conceptual categories. The researcher can saturate on category and then move on to other categories. “In trying to reach saturation he maximizes differences in his groups in order to maximize the varieties of data bearing on the category, and thereby develops as many diverse properties of the category as possible” (p. 62).
Core theoretical categories should be saturated more than peripheral ones. Theoretical sampling is not statistical sampling — it may only require a few groups to exhaust one category, and many groups to exhaust another. The number of groups is based on the extent of saturation and the level of theory one wants to generate. Knowing when to quit is a skill developed through experience.
Theory generation can involve multiple types of data simultaneously, and multiple slices of that data. Often it doesn’t matter what slices the researcher uses — they all are useful in developing categories. In fact , multiple slices helps build better theory by ensuring the generated theory works from multiple perspectives.
IV: From Substantive to Formal Theory One way to generate formal theory from substantive theory is merely to replace specific words with general words and see what happens. Another way is to apply the substantive theory to more diverse groups and situations and interatively expand the substantive theory. Another method is through direct comparisons of data from other substantitive areas in the researchers experience or in the literature, though this is risky and often leads to logical-deductive thinking. Most researchers tend to stay in the substantive theory area and avoid going into formal theory.
V: The Constant Comparative Method of Qualitative Analysis Currently analyzing qualitative data usually involves coding the data to get some quantifiable means to test some hypotheses. Glaser and Strauss advocate combining coding with analysis to help locate and build grounded theory. In this method the data is coded only enough to generate categories and hypotheses.
The authors describe four main stages:
1. Comparing incidents applicable to each category Begin by coding the data into as many categories as possible. Some categories will be generated from himself, some from the language and data of the research situation. As you find more instances of the same category code you will being to refine your ideas about that category. At this point it’s best to stop coding and record a memo of these ideas.
2. Integrating Categories and their Properties The constant comparative method will begin to evolve from comparing incidents to focusing on emergent properties of the category. Diverse properties will start to become integrated. The resulting theory will begin to emerge by itself.
3. Delimiting the Theory Eventually the theory solidifies, and there are fewer changes to the theory as the researcher compares more incidents. Later modifications include taking out irrelevant properties of the categories, integrating details of properties into an outline of interrelated categories. More importantly, the researcher will begin to find ways to delimit the theory with a set of higher level concepts. He/she will find ways to generalize the theory more as they continue to make constant comparisons against it. The number of categories will be reduced. New categories are often created halfway through coding, and it usually isn’t necessary to go back and code for them. You only need to code enough to saturate the properties of the category. Later you can evaluate your categories and emergent theory by moving on to new comparison groups.
4. Writing Theory “When the researcher is convinced that his analytic framework forma a systematic substantive theory, that is it s reasonably accurate statement of the matters studied, and that it is couched in a form that others going into the same field could use — then he can publish his results with confidence” (p. 113). This methodology tends to result in a “developmental” theory, which lends itself to further evolution. It’s an inductive rather than deductive approach.
VI: Clarifying and Assessing Comparative Studies Many types of studies are subsumed under the general name of comparative analysis. Each has differing objectives regarding generation and verification, and you must carefully read them to determine what level of each the study has.
VII: New Sources for Qualitative Data Glaser and Strauss advocate using the library to supplant field data in theory generation. The strategies used in library research can be very similar to those previously discussed for fieldwork. Most researchers underutilize the library and overfocus on their own research data. There are many types of primary and secondary documents that can be interatively explored like fieldnotes and can be coded as well. There may be sets of interviews, surveys, pictures that prove useful in theory generation or in further explaining the fieldnote contexts. The library offers a fantastic range of potential comparison groups to study. Of course, the library strategy has limitations. Existing data may be quite limited in scope. Further clarification may be impossible (especially if the respondents are dead). It can be hard to assess accuracy.
VIII: Theoretical Elaboration of Quantitative Data Even quantitative data, often the exclusive domain of verification studies, can be used to generate grounded theory. The authors propose that this could be a new frontier of theory generation. IX: The Credibility of Grounded Theory Often qualitative research of this type is labeled “sloppy” or “exploratory”. You can improve the credibility by basing your research on multiple comparison groups.
X: Applying Grounded Theory A good pratical ground theory should: * fit the substantive area it will be used in* understandable by laymen in that area * general enough to apply to many situations in the area * allow the user some control over the theory as the daily situations change over time.
Refer back vs retrieval
I am trying to write about the retrieval process. But some how got stuck with the word refer back and retrieval.
Are this 2 concepts the same?
Or does somebody do something , to make retrieval easy?
For example:
Located the information (working in the temporary location) to make it easy accessible for duartion of time?
Hmmmm it sounds similar relation
- in the sense of to be able to access easily without browsing again and again or go to the long way to be able to find some infor again?
Or it need to be retrieve first than refer again (for a duration of time)—hmmm seems logical.
- retrieval:
1) Link 1
C.J. van Rijsbergen
B.Sc., Dip. NAAC, Ph.D., F.B.C.S., F.I.E.E., C.Eng., F.R.S.E.
Information Retrieval Group
Department of Computing Science
University of Glasgow
Glasgow G12 8QQ
SCOTLAND
he Information Retrieval paradigm is about a person (human or physical) having a need for information, and a set of information objects from which this need is to be satisfied. In this chapter we provide general models to formalize the concept of information need.
A searcher has a need for certain information. This need may have both qualitative and quantitative aspects. In the context of an information collection, the searcher tries to formulate this need in terms of this collection, translating the need for information into a need for information objects (documents). The starting point of our discussion will be a searcher having a need for documents in the collection.
The information collection may be a stable (fixed) collection, in which case an Information Retrieval System may preprocess the collection for optimal disclosure. The intention of this preprocessing is that the Information Retrieval System gets an impression of the contents of each document. This calls for a contents description mechanism.
The collection may be a (continuous) stream of documents (like a newsgroup on the Internet). For each next document, the Information Retrieval System has to decide if it will be interesting for the searcher. This is referred to as the Information Filtering paradigm. In this case there can be no preprocessing of all documents. Descibing document contents then is either based on some universal system for contents description, or the system should try to build such a system incrementally when new documents arrive. This is also the situtation on the world wide web, an extremely large, dynamic and volatile collection of documents.
A difference between Information Retrieval and Information Filtering is that a searcher interest is valid during some period of time, while a query only expresses the information need at some point in time. As there may be content fluctuations in the document stream, there may also be fluctuation in the searcher interest.
Information retrieval is a wide, often loosely-defined term but in these pages I shall be concerned only with automatic information retrieval systems. Automatic as opposed to manual and information as opposed to data or fact. Unfortunately the word information can be very misleading. In the context of information retrieval (IR), information, in the technical meaning given in Shannon’s theory of communication, is not readily measured (Shannon and Weaver[1]). In fact, in many cases one can adequately describe the kind of retrieval by simply substituting ‘document’ for ‘information’. Nevertheless, ‘information retrieval’ has become accepted as a description of the kind of work published by Cleverdon, Salton, Sparck Jones, Lancaster and others. A perfectly straightforward definition along these lines is given by Lancaster[2]: ‘Information retrieval is the term conventionally, though somewhat inaccurately, applied to the type of activity discussed in this volume. An information retrieval system does not inform (i.e. change the knowledge of) the user on the subject of his inquiry. It merely informs on the existence (or non-existence) and whereabouts of documents relating to his request.’ This specifically excludes Question-Answering systems as typified by Winograd[3] and those described by Minsky[4]]. It also excludes data retrieval systems such as used by, say, the stock exchange for on-line quotations.
To make clear the difference between data retrieval (DR) and information retrieval (IR), I have listed in Table 1.1 some of the distinguishing properties of data and information retrieval. One may want to criticise this dichotomy on the grounds that the boundary
that in IR we are searching for relevant documents as opposed to exactly matching items. The extent of the match in IR is assumed to indicate the likelihood of the relevance of that item. One simple consequence of this difference is that DR is more sensitive to error in the sense that, an error in matching will not retrieve the wanted item which implies a total failure of the system. In IR small errors in matching generally do not affect performance of the system significantly.
In dilema – have several ideas – to proceed
However, manage to get all the LOW COMPETENCY user open categories.
Having said that, I am in the dilema and have several solutions or testing for it:
1) either start with LC users with axial coding, applied paradigm model and propose it (theoretical sampling].
2) proceed with coding (recording dialectic) with basic competency user and high level user – meaning still in the open coding process.
3) or it seems could be done parallel.
4) How to generate theory/proposition base on the data? A problem for me….
Links to theory creation
1) Alan Dix
2) Research methodolog in this book – bought it.
The process of doing coding – will keep on updates
having problem in converting (message from scribd).
may be the files too big or too many links in the file.
This is an Excel files where the coding process happend.
It was in many sheets because of the recycling and redo in the codes (
mainly because of lack of confidence —bad azrina…bad girl).
Open categories for LC Users
Finished for (I do not know) how many times I redo this Open categories.
Now, I have 60 Open categories for Low competency users. Next steps is:
- comparing with each categories
- draw in visio
- extend with basic and high competency to acquire/amend other categories if exists.
- axial categories
- in between trying to amend the analysis writing.
In visio version:
Task -Open categories for Low ver 1.
-
Archives
- July 2009 (2)
- May 2009 (2)
- April 2009 (2)
- March 2009 (10)
- February 2009 (18)
- January 2009 (8)
- December 2008 (1)
- November 2008 (2)
- July 2008 (3)
- February 2008 (2)
- December 2007 (1)
- November 2007 (4)
-
Categories
-
RSS
Entries RSS
Comments RSS
