Posts Tagged ‘EDU 6975 Interpreting and Applying Educational Research II:’

Capstone-Standard 11 Meta-Reflection: Inquiry/Research

Standard 11 Meta-Reflection: Inquiry/Research

Competently consumes and produces where necessary empirical data to guide educational practice.

Initial reflection during C & I Orientation:

It is my responsibility to effectively use data at every level in my job including: selecting research-based curriculum, researching best practices for delivering services to address IEP goals, and collecting data on student performance. Equally important is to examine district records and track information to ensure that classroom goals are based on current IEPS which are in turn based on current evaluations.

Meta-Reflection following completion of EDU 6976 Interpreting and Applying Educational Research I and EDU 6975 Interpreting and Applying Educational Research II:

Please note: The requirement for EDU 6976 Interpreting and Applying Educational Research I was fulfilled via transfer credits through another university. I am grateful for SPU’s acceptance of credits for EDD/569 Introduction to Action Research and QNT/575 Measurement, Evaluation and Ethics in Research. Two of the artifacts included below are from these courses.

The need for data:

When our district’s new high school opened up in 2007, our principal wanted our team to implement a full-inclusion program….Our special education team felt strongly that “a continuum of services” needed to be offered, therefore, we took data. We carefully monitored every student’s progress in any way possible. We talked with teachers, met with students, checked and recorded online grades, and listened to concerned parents. Data became the magic key. Within three weeks, we were able to discern enough of a pattern and shared this data with our administrators–respectfully requesting that we be allowed to create a few class sections for SE  Math, English and Learning Strategies. Our request was granted and this model has since become our status quo. (Although now, the students with the greatest needs can be served in SE starting from the beginning of the year, but others are encouraged to spread their wings, knowing that we have a safety net with alternative approaches if needed). (excerpted from EDSP 6644 BLOG 2 Principles of HOPE, James, 2012)

EDD/569 Introduction to Action Research and QNT/575 Measurement, Evaluation and Ethics in Research.

The first two of the artifacts for Standard 11 provide examples of Artifact–Data Collection, Artifact–Descriptive Statistics. Please note: In the process of scanning and uploading hard copies, some distortion occurred within these documents.

Within these courses, a primary emphasis was the use of multiple data collection methods, regarding student performance in school settings at every grade level– leading to a more complete and composite picture of students’ strengths and weaknesses. A variety of data sources ranging from existing school electronic database, self-collected teacher observations, as well as responses to surveys collected directly from studentscan provide educators at every level with valuable baseline, target and progress information. It was noted that  among the most widely used methods for collecting more individualized responses are in-depth interviews, observations, surveys, questionnaires and document analysis.

EDU 6975 Interpreting and Applying Educational Research II:

In the context of this course, we wrote weekly reflections, although these were not required to be posted in WordPress. I am choosing to include these original reflections, in this meta-reflection, as this format allows me to retrace my steps in a way that is most beneficial for my learning within this C & I Capstone class.

Module 3 Reflection:

Within this module, I learned of the importance of looking at the differences between the percentages in studies versus the numbers, as well as the importance of always going back to the research question. Also, this week’s lecture helped to clarify for me that the sampling distribution is centered on “zero” as noted on the graph and that this represents the null hypothesis.  Additionally, I am learning of the importance of distinguishing between random assignment and random sampling; In analyzing the study regarding the “Murderous Nurse”, I found myself getting caught up with thinking about all the other potential variables instead of looking at the lack of randomization. The class notes indicate that “this study does not implement random sampling nor random assignment”. Lack of random sampling prevents generalization to the larger population. The lack of random assignment, prevents drawing a cause and effect conclusion.

This may be a stretch, but I created a visual in my own mind to help me differentiate the two terms. To associate sample with generalization, I visualize offering cookie samples to a larger group or population. The visual in my mind to associate assignment with cause and effect is; When I give an assignment to a student (cause), I anticipate that the student will complete it (effect).

In previous modules, we have discussed the fact that other variables can and do influence both the results and the interpretation of statistics. I am reminded that there can be correct statistical analysis, however, these results must also take into consideration the type of  study, number of trials, etc. so that appropriate inferences can be drawn. (ie Good data but incorrect procedures)

Module 4 Reflection:

In the midst of trying to grapple with significantly heavy material this week, I must admit I know I have SO MUCH more to absorb than I would have hoped by the end of this unit, however, I learned how to write up the results of a study. The example write-up (Sleep Deprivation Study) provided a tangible and very useful tool for guiding me through the critical steps in the process. Also, through the process of working with Group 1 on the Homework 2 assignment, a particular explanation from Laura Zylstra helped me to get a better understanding of how to calculate the p-value. (My understanding is still fuzzy and tentative—but at least her explanation made sense for the specific situation. Now if I can learn to apply the procedure…I’ll be moving in the right direction!) Her explanation (related to question #10 in the Latin American Study) is as follows:

”P-value is found by using the difference of the means of the educational achievement levels (5.92). This number (5.92) is then located on figure 2, the plot of the differences in means from the 500 simulated trials. All numbers at 5.92 and above are added together and divided by the number of trials. Since there is only one (1) dot at or above 5.92, and there are 500 trials, divide 1 by 500, and that is the p-value. Therefore the approximate p-value is 0.002.”

Module 5 Reflection:

While reading this week all about ANOVA and Tukey’s Post-Hoc tests (along with several others) I learned that “the F ratio is the resulting statistic” (Sprinthill, 2012, p. 367). A few basic concepts: High F ratio—high variability, low F ratio—low variability. One way ANOVA—one independent variable, effect size determined through using eta squares Factorial ANOVA—more than one independent variable, effect size determined through use of partial eta squares.

When considering the relationship between samples and populations, higher F-ratios suggest samples are from different populations whereas lower F-ratios suggest that the samples represent a single population.

Module 6 Reflection:

In the midst of studying the material for this week on Chi-square studies, I learned of the importance of this type of statistical analysis and its very practical application to many everyday situations. The textbook refers to this type of nominal data as ‘nose-counting data’ with ‘no shades of gray”. Perhaps this type of simplicity is relatively refreshing—in light of the complexity of the content of this course! In addition to the specific details of chi-square as a non-parametric procedure, I learned of the value of the added personal communication I experienced this week: 1) The honesty expressed by many of us when we are confused, 2). Dr. M’s compassion expressed in the discussion thread –especially when she posted the very applicable scripture to encourage us, 3) the follow-up email and subsequent phone appointment I made with her this week to discuss my mid-term results, 4) the blessing of working together in a small group on the homework, 5) the helpfulness of participating in the optional Tuesday afternoon “live-chat” session.

The specific details articulated throughout the lecture and discussion threads this week regarding chi square procedures have offered clarification of similarities and differences between these and previous tests we have been learning about such as the t-test and ANOVA. This type of contrast and comparison as well as spending additional time re-listening to lectures helps to bring more clarity to my thinking (although I still have much, much more to learn and understand). I am so grateful to God for His never-ending faithfulness and promise to be with me when I go through “deep waters”.

Module 7 Reflection:

Among the new concepts I learned about this week was that of the scatterplot. I learned that a pair of scores can be found on each point within the plot and that in correlational studies, the slope on a scatterplot indicates whether the correlation represented on the plot is positive or negative. (Lower left to upper right- positive. Upper left to lower right-negative). In previous units, we looked at experimental methods for determining cause and effect relationships (such as t-tests and ANOVA) whereas this unit focused on methods for effectively examining relationships between variables within the same group.

Module 8 Reflection:

During this week, I learned many new terms related to the concepts of correlation and prediction. Bivariate—two variable scatterplot, residual error—difference between actual values of Y and their predicted values, multicollinearity—when variables are “too highly correlated” as in the text’s example of income and unemployment, spurious correlation—a correlation that can be quite misleading, canonical correlation—two or more x variables correlated with two or more y variables (ex. several personality tests to predict several measures of leadership) Also, the standard error of the estimate relates to the interval where a true score might be located, and that the wider the interval—the higher the level of confidence. I wonder if this is like the larger a target is, the more likely you will hit it? I see the concepts surrounding correlation research as more applicable to research in educational settings than experimental research—and yet both can involve independent and dependent variables.

Module 9 Reflection:

Throughout this module I’ve finally learned to understand that one of the primary reasons for confusion between the terms standard error (SE) and standard deviation (SD) is that according to the article by Altman, “the standard error is a type of standard deviation”(2005, p. 903). A larger sample size decreases the standard error perhaps in the same way that a “closer look” at an object (making it appear larger) brings it into sharper focus and increases clarity and accuracy. If I want to obtain a clearer and more accurate “picture” of a situation (or sample mean), I need to ask more people. By contrast, regardless of the number of people I ask (size of the sample) the measure of variability or changes in what I am studying or looking at will not be apt to change.These concepts are important to understand when endeavoring to use inferential statistics to draw the most accurate-as-possible conclusions regarding what is “true” for a given population.

In conclusion and on a personal note:

In the midst of my measurement class, my mom became ill and passed away just two weeks before the end of the course. I can relate to Parker Palmer’s words in The Courage to Teach, as he describes the sudden loss of his father at a particularly stressful time in his teaching career; “I was devastated” (p. xi) Years earlier, just before taking my first statistics course, my dad had passed away. Thankfully, during the time I took EDU 6975 in the fall of 2012, I did not lose a family member, however, our family was in the midst of a major move and “in-between houses”—staying with family, and I became very sick and depleted. The reason I share these personal details in this meta-reflection, is that perhaps one of the greatest “life lessons” learned in the process of taking these research courses under particularly stressful situations is that of identifying even more with my students who struggle. As one who has always been blessed with high grades and success in school, I choose to “document” these vulnerabilities here in this reflection—that I might never lose sight of the importance of having empathy for my students who struggle. Palmer reminds us; ”Identity and integrity have as much to do with our shadows and limits, our wounds and fears, as with our strengths and potentials” (p. 13).

Artifacts for Standard 11:

UoP:  Artifact–Data Collection

UoP:  Artifact–Descriptive Statistics

SPU:   Homework 1:Homework 1 (second version)

SPU:   Homework 2:Sleep-Deprivation–Revised–Group Copy

SPU:   Homework 3:HW3 Group #1 graded

SPU:   Homework 4:Homework 4 Laurie James (graded)

SPU:   Homework 5:Homework5 Estimation Final

 

Resources:

Altman, D. (2005). Standard deviations and standard errors. British Medical Journal. (p. 903)

Palmer, P. (2007). The courage to teach: Exploring the inner landscape of a teacher’s life. San Francisco: Jossey-Bass.

Sprinthall, R. C. (2003). Basic statistical analysis (7th ed.). New York: Pearson Education, Inc.