Thursday, January 15, 2009

EDU 580 Yin's Ch. 5 Discussion Post

Explain the five analytic techniques and how you can "press for a high-quality analysis." Connect this to your own case study plan/idea too.

14 comments:

  1. The five analytic techniques:
    1. Pattern Matching:
    This compares an empirical (based on observation) pattern with a predicted one. If the patterns coincide then the results can strengthen the internal validity. This is the most desirable technique. This is divided into parts. The parts include:
    nonequivalent dependent variables as a pattern-The experiment may have more than one dependent variable.
    rival explanations as patterns-This is pattern matching for the independent variables. The researcher needs to focus on why the outcome occurred as it did. Only one explanation can be true.
    simpler patterns- There is minimal variety of dependent and independent variables.
    precision of pattern matching-There are no precise comparisons.
    2. Explanation Building
    This is a type of pattern matching that analyzes the case study data by building an explanation about the case. This is also divided in sections.
    elements of explanations-The hows and whys something happened.
    iterative nature of explanation building- I think this is when new perspectives are looked at after the case study is revisited.... (I think?)
    3. Time-Series Analysis
    simple time series-There can only be one dependent and independent variable. Starting and ending points may be unclear, which create complications. I think the book is trying to say that time-series design matches the observed trend with either the theoretically significant trend specified before the onset of the investigation or a rival trend specified earlier. (That's straight from the book, but I don't understand.)Please explain.
    complex time series-Just a more complex time series. More variables and more complex outcomes are present.
    chronologies-The events are sequenced. The events are not limited as far as variables.
    summary conditions for time-series analysis-Don't observe the trends alone... look at the hows and whys.
    4. Logic Models
    This model stipulates a complex chain of events over a long time. The events are staged and repeated (cause and effect repeating pattern). This will help define goals and actions. There are different models that would help in different situations.
    5. Cross-Case Synthesis
    This is used only with analysis of multiple cases.
    Wow, I was able to take notes from the book but after rereading my notes and book I am lost. I don't understand this stuff. I can’t explain how to apply this to my research.

    Dr. Hendrix~ Can we go over this when we meet or can you meet with me any Tuesday or Monday after 4:30? Call me if it is easier to set up a time!

    ReplyDelete
  2. Oh through all that I forgot to tell how a researcher can press for a high-quality analysis!

    Researchers should make sure they have attended to all evidence. Don't leave any loose ends. They should also address all major rival interpretations. Then researchers should address the most significant aspect of your case study. Lastly, a researcher should use their own knowledge.

    ReplyDelete
  3. Explain the five analytic techniques and how you can "press for a high-quality analysis." Connect this to your own case study plan/idea too.

    1) Pattern Matching- Compares an empirical (based on observation) pattern with a predicted one. If the patterns coincide, then the results can strengthen the “internal validity”. There are four kinds of pattern matching: A) Non-equivalent dependent variables as a pattern in which the experiment may have more than one dependent variable B)
    Rival explanations as patterns for matching the independent variables with the focus on why the outcome occurred as it did, but only one explanation is deemed to be true C)
    Simpler patterns where there is a minimal variety of dependent and independent variables D) Precision of pattern matching where essentially there are no precise comparisons

    2) Explanation Building- This is a special type of pattern matching, only more difficult, that analyzes the case study data by building an explanation about the case. A) Elements of explanations which explains the “how” or/and “why” something happened B) Iterative nature of explanation building is essentially when the case study evidence is examined, theoretical positions are revised, and examined again
    3) Time-Series Analysis- simple time series where there is only one dependent and independent variable. Starting and ending points may be unclear, which create complications. A) Complex time series is a more complex time series with more variables and more complex outcomes B) Chronologies involve sequencing the events C) Summary conditions for time-series analysis examine the relevant “how” and “why” questions about the relationship over time, not merely the trends alone

    4) Logic Models- Events are staged over a series of time. The events are staged and repeated in a “cause and effect” pattern. This helps define the goals and actions of thecase study. There are individual-level logic models, firm or organizational level logic models, and program-level logic models

    5) Cross-Case Synthesis- applies specifically to the analysis of multiple cases and how they relate to one another, ie. If there are any significant or relevant similarities
    You can press for a high-quality analysis by showing how you attended to all the evidence, your analysis should address all major rival interpretations, it should address the most significant aspect of your case, and it should use your own prior knowledge
    This chapter was by far the MOST confusing to me.
    Fact- Data analysis consists of examining, categorizing, tabulating, testing, or otherwise recombining evidence, to draw empirically based conclusions.
    Quote-“The best preparation for conducting case study analysis is to have a general analytic strategy”.
    Question- How far is everyone else on their IRB proposals? I’m scared!

    ReplyDelete
  4. Annie,

    Just meet me in my office on Tues. at 4:30 then. We will review this material. Everything will not apply to your own study here. In this part, Yin is showing you a variety of case studies. Which type does your work fall into in his book?

    I will be here Tues. Stop by and see me!!! I am here to help!!!!!

    ReplyDelete
  5. Erin,

    No worries! Come and see me about your IRB and/or informed consents!

    ReplyDelete
  6. Annie and Erin have done an amazing job detailing the points so as to not be too redundant I will try to only summarize.
    The five analytical techniques are:
    1. Pattern matching- it is the most desireable, it compares an empirically based pattern with a predicted one. If the patterns coincide it will help to strengthen the cases internal validity.
    2. Explanation building- is actually a special type of pattern matching however it is more difficult. The goal is to analyze the case study data by building an explanation about the case. It is mainly used in with explanatory case studies. The object is to "explain" a phenomenon- to stipulated presumed set of casual links about it, or "how" or "why" something happened. It is usually done in narrative form.
    3. Time-series analysis- can follow many intricate patterns and can lay a firm foundation for the conclusion of the case study. The researcher wants to find match between the observed trend and either a theoretically significant trends specified before the inset of the investigation, or some rival trend, also specified earlier.
    4. Logic models- useful with case study evaluations, deliberately stipulates complex chain of evens over an extended period of time. The events are staged in repeated cause-effect-cause-effect patterns, where a dependent variable at an earlier stage becomes the independent variable for the next stage. The use of logic models consists of matching empirically observed events to theoretically predicted events.
    5. Cross-case analysis- applies specifically to the analysis of multiple cases. Having more than 2 cases could strengthen your findings. The technique treats each case study as a seprate study. This one was too hard for me to try to summarize so good luck here.
    Pressing for higher quality analysis- 4 principles underlie all good social research:
    1. Show that you attended to all the evidence
    2. Address all major rival interpretations
    3. Address the most significant aspect of your case study
    4. Use your own prior expert knowledge, awareness of current thinking
    In order to connect these to my own case study plan I will need way more info. I think many of us will use either the pattern building or explanation building techniques. I think the pattern building could be tied to your research and lit review. This is the research already done and here are the patterns in my research, how are they alike and how are they different. With our interviews we can tie into the explanation building. What observations and comments are made to justify our explanations. The logic model I think of as a chain reaction, I didn't see this linked to the type of studies we have all discussed. The others were just over my head!
    Question-
    How will we decide what route to go on this, and when? Is it best to see which fits after we've completed more of the data collection?
    Quote
    "None of the analytic techniques should be considered easy to use, and all will need much practice to be used powerfully. Your objective should be to start modeslty, work throughly and introspectively, and build your own analytical repertoire over time."

    ReplyDelete
  7. Pattern Matching-
    compares an empirical pattern and a predicted one. If the 2 patterns parallel, then it increases your internal validity.
    Explanation Building-
    a type of pattern matching. it analyzes case study data and builds an explanation of the case. it explains how and why something happened.
    Time Series Analysis-
    simple- has only 1 depandent and independent variable.
    complex- has more variables and more complex outcomes. The essential logic behind the time-series design is the match between the observed trend and either a significant trend specified before the case study, or some rival trend also specified before the case study.
    Logic Model-
    deliberately stipulates a complex chain of events over an extended period of time. There is a cause/effect relationship between the events.
    **This one relates to my study- or seems to- because it said that the evaluators (ME) and the officials implementing a program (My adminstration) can work together to form a program's logic model. I think this may help to show them what they are lacking if they want the program to be successful, without me getting in trouble for pointing it out. That is my fear about this program implementation as a topic of my study.
    5. Cross-Case Synthesis-
    This is used only with analysis of multiple cases. It treats each individual case as a separate study.
    To insure you have a high quality analysis you need to:
    -be sure to address all of your data
    - be sure to address all rival interpretations of your data
    - be sure to address the most significant aspect of your case study.
    Question:
    How do you know what the most significant aspect of your case study is?

    Dr. Hendrix,
    I want to meet on this Tuesday too. I have questions about my study and research due to developments (or lack of progress) at my school.

    ReplyDelete
  8. The analytic strategies are used to develop internal validity (how and why event X lead to event Y) and external validity (generalization of results to some broader theory if research replicated 2-3 times with same results).
    1. Pattern matching – compares a pattern from experience or observation with a predicted one
    2. Explanation building – a more difficult type of pattern matching where the researcher looks for “how” or “why” something happened…these causal links may be more complex and difficult to measure in any precise manner
    3. Time-Series Analysis – to be used with more intricate and precise patterns to lay a firm foundation for the conclusions of a case study
    a. Simple time series - single variable
    b. Complex time series – mixed pattern, across time
    c. Chronologies – trace events over time
    4. Logic models – deliberately stipulates a chain of events over time (staged events); 4 types of logic models
    5. Cross-case Synthesis – specifically for the anaylsis of multiple cases; results will carry more strength than having only a single case
    As I am reading the text and your blogs, I am wondering which method to consider. However, I still haven’t written my research questions in the correct format yet and am quite overwhelmed. I think I may use pattern matching to discover the patterns for why or why not students enjoy reading. I think I may see a pattern in SRI scores and attitudes towards reading.
    Should I study my entire class or only those reading below grade level to see how reading is impacted once audiobooks are integrated?

    ReplyDelete
  9. Good questions class!
    I think I answered these in person here, but if not, please ask again on Tues. or in here. I will respond.

    See you Tues.,
    Dr. Hendrix

    ReplyDelete
  10. I somehow skipped this chapter and did chapter 6 first. Funny how I understood that one much better. At this point, I am completely overwhelmed with this class and my life in general, so I can’t wrap my mind around anything in this chapter.


    1. Pattern Matching- compares an empirically based pattern with a predicted one- results strengthen internal validity
    2. Explanation Building-type of pattern matching in which the goal is to build an explanation about the case, more complex form of pattern matching as it aims to answer the “how” and “why”
    3. Time-Series Analysis- I really am not clear on this one, but I think it looks at changes over time and compares the empirical trend to a theoretically significant trend or some rival trend. Yeah, I really don’t know what I just said.
    4. Logic Models- deliberately stipulates a complex chain of events over a period of time in repeated cause-effect pattern (like a domino effect)- involves matching empirical observed events to theoretically predicted events
    5. Cross-case synthesis- technique used in multiple case studies

    Pressing for a high-quality analysis
    1. Show that you attended to ALL the evidence
    2. Address all major rival interpretations
    3. Address the most significant aspect of your case study
    4. Use your own prior knowledge- show that you’re in expert on the subject

    ReplyDelete
  11. Explain the five analytic techniques and how you can "press for a high-quality analysis." Connect this to your own case study plan/idea too.

    Pattern matching: most desirable technique for case studies. Increases internal validity if the patterns coincide. Compares a empirically based pattern with a predicted one. Divided into three categories:
    Nonequivalent dependent variables as a pattern - experiment has multiple dependent variables
    Rival explanations as patterns - pattern matching for independent variables. Investigation focuses on the how and why a certain outcome occurs in each case.
    Simpler patterns - used in cases where there may only be two independent or dependent variables. Because there are fewer variables the more dramatic the different patterns will have to be.

    Explanation Building: the goal is to analyze the data by building an explanation about the case. Explains how or why something happened. Often done in a narrative form and requires much analytic insight.

    Time series analysis: Ok this one confuses me. I think it's a way to analyze data over a period of time. It is the match between the empirical (observed) trend and a rival trend or a theoretically significant trend specified before the investigation began. Divided into three types.
    Simple time series - only a single independent or dependent variable.
    Complex time series - multiple variables are present.
    Chronologies - compiling of chronological events. Allows you to trace events over time.

    Logic Models: Stipulates a chain of events over a period of time. Has sequential stages. Matching observed events to theoretically predicted events. Divided into four types: individual-level logic model, organizational-level logic model, alternative configuration for organizational-level, and program-level logic model. Each type is suited for a specific group (individual, organization, transforming an organization, and federal programs).

    Cross Case Synthesis:applies to the analysis of multiple cases. Treats each individual case study as a separate study.

    Pressing for high-quality analysis:
    1. Analysis shoudl show that you attended to all evidence.
    2. Should address all major rival interpretations
    3. Should address the most significant aspect of your case.
    4. Use you own expert knowledge

    Ok, I found this chapter to be completely overwhelming. I think that for the purposes of my case study I would be more likely to use pattern matching and explanation building. I think that I may find patterns as to how using CGI theory to teach math has increased my students mathematical understanding and confidence. Although I know explanation building can be difficult I am hoping that because of my knowledge and close proximity to the participants in my study I can build a plausible explanation.

    I am still struggling with my case study. I need to get my IRB sent in but I still have so many questions. I need to talk this over with you Dr. Hendrix either tomorrow night or during an office visit. Thanks!



    Explain the procedures in doing a case study report and "what makes an exemplary case study."

    In order to create good case studies one must either be good at composing or want to become good at it. It takes much practice and should not be seen as a burden but as an opportunity. You should not wait until the end of your study to begin composing. The earlier you begin the better. The bibliography and methodological sections can be written very early in your study. Most of the bibliography comes from your literature review which will be completed before you begin your study. This is why it is a good starting point when composing. As for the methodological section, your major procedures for data collection and data analysis will already have been determined.

    ReplyDelete
  12. The five analytical techniques are:
    1. Pattern-matching
    a. Compares an empirically based pattern with a predicted one
    2. Explanation Building
    a. Analyze the case by study by building an explanation about the case
    3. Time-Series Analysis
    a. Can only be one dependent or independent variable
    4. Logic Models
    a. Compares an empirically observed events with a predicted events
    b. Stipulates a complex chain of events over extended amount of time
    5. Cross-Case Synthesis
    a. treats each case as a separate study

    A cross-case synthesis would definitely work with my study because I could evaluate whether taking honors courses helped each particular student.

    ReplyDelete
  13. wedding photography Asian Luxury Asian Wedding Photographer based in Slough | London for Hindu, Sikh ... Further to your enquiry to Vivida Photography we will contact you within 24 ...

    ReplyDelete