Last but Not Least

This semester's course objectives:

1. Compare and contrast quantitative, qualitative and mixed-methods approaches to research.
2. Explain what experimental, quasi-experimental and non-experimental research designs entail and describe their application to different research questions.
3. Explain descriptive statistical techniques such as measures of central tendency, standard deviation and correlation.
4. Explain the ethical principles that pertain to research involving human subjects and research conducted in educational settings.
5. Select a research problem and formulate appropriate research hypotheses and/or questions.
6. Conduct a review of educational literature from texts, journals and computer library databases.
7. Write a coherent synthesis of such literature as it relates to the research problem.
8. Prepare a viable research proposal.

As the semester progressed I found that quantitative and qualitative methods can be used in research related to architecture. Buildings are solid masses we can quantify and measure but they only exist to serve in combination with occupants and context so the interaction between is often what is in question. That interaction is not so quantifiable but in combination I think the two methods can be successful in comprehensively approaching many of my specific questions. Research designs models were especially interesting because they provide frameworks. Developing a research question is daunting enough. Having models to provide the framework has given me more confidence in developing research questions, proposals and experiments. Research question development is similar to concept development - I like the back and forth process. Literature synthesis and review is still challenging to me. I could use more time learning how to review different types of documents and where to find sources. Through the process of developing my research proposal I've learned most specifically about idea/question development, literature searches and thorough development and documentation of a process. Our conversations about validity and accuracy helped explain the reason behind the structure for these proposals.

Blog Post #7: Zumbrunn's Exploration of Writing Feedback Perceptions, Self-Efficacy Beliefs, and Success in Middle School Students

Sharon Zumbrunn used 287 fourth grade students as her sample to better understand the relationship between student's perceptions of feedback toward their writing and their beliefs about their own writing. The experiment was designed to explore whether better understanding this relationship could become a predictive tool for student's writing success. She used 2 different surveys that asked students to rank their answers on a scale. This allowed her to not only gather answers to the survey questions but to also gather data toward understanding the frequency of the behavior or response in question or the student's commitment to the answer. The surveys were administered by the teachers in their classrooms and the findings showed that there is a relationship and that further research can inform instructional practice to increase student's writing success.


5 questions I have for the research team:

1. Why fourth graders and not high school students who may have a greater appreciation for self-reflection?
2. Were the surveys developed in collaboration with Sharon Zumbrunn?
3. If they were, was that because there were no pre-existing surveys?
4. Why were surveys and not interviews used?
5. Were teacher's feedback methods included in data gathering?

Ethnographic Observations

At 615pm on Monday, April 4th, I sat on a bench below Oliver Hall to observe the intersection of Harrison and Main Streets. The temperature was around 75 degrees with a few clouds and a light breeze. This intersection has heavy bicycle, pedestrian and auto traffic including buses, cars, motorcycles and scooters. There were several loud buses that emitted gray exhaust when starting forward after stopping at the stoplight. Approximately 4 cars, several SUVs, had their windows rolled down so you could hear ridiculously loud music. In between the auto traffic the intersection is surprisingly quiet. A vendor selling VCU Final Four shirts, hats and the like was set up in the 7-11 parking lot across Harrison Street from where I sat. He lured in several of the 7-11 customers. On the northeast corner of the intersection is Piccola Italy (restaurant) and on the southeast and southwest corners are VCU buildings. The northwest corner (7-11) had the most pedestrian and bicycle traffic.There are no bike lanes on Harrison or Main Streets so bicycle riders either go with the auto traffic or ride on sidewalks which line both streets. There were relatively few students going in and out of the VCU buildings (Oliver Hall) at the intersection. Pedestrians were casual in crossing the streets - sometimes going against the stoplight. During this observation there were no cars honking and no yelling. Main Street had more auto traffic than Harrison Street. More female than male pedestrians were observed but both genders were typically wearing sunglasses and had either a purse, backpack, messenger bag or a combination. The majority of pedestrians were wearing jeans - long or short- short sleeved shirts and appeared to be in their late 'teens, 20's or 30's.

Re: Blog Post 8 - Testing Historical Knowledge

Gabriel A. Reich's article, Testing Historical Knowledge: Standards, Multiple-Choice Questions and Student Reasoning, addresses the following research questions:



1.  How can the knowledge and reasoning used by relatively
successful and unsuccessful participants to answer a set of 
multiple-choice history questions be characterized?
2.  Did the items evoke the thinking called for in the
standards?

• qualitative study
• knowledge and reasoning = used in multiple choice test taking per the Regent's exams
• successful and unsuccessful participants = sampling from “Metropolitan School” which is a small, urban high school of fewer than 400 students - 15 students from a single 10th grade history class participated

The article explains the history, methodology, findings, discussion and conclusions for the above research questions. A small sample of 15 students in Mr. Heche's 10th grade history class at Metropolitan School (NY state) were given 15 multiple-choice questions, observed taking their "test" on historical topics and interviewed. The researchers used codes in their observations to assess students cognitive processes and used additional codes to express their inferences. The article provides many findings that point toward the tendency for teachers to focus on literacy skills and content knowledge in their teachings, vs. discipline-based skills, but that Mr. Heche did not do this. He did, however, abide by a curriculum that followed the test standards closely.

Questions:

1. What was the motivator for using such a small sample size? (15 students vs. that many students from several classes) Was it due to the selection criteria and if so why wasn't that criteria modified so that the results using a larger sample could be more generalizable?
2. If the research question is to explore how independent variables like knowledge and reasoning are to be classified - why was history, versus math, chosen as the test subject for the tests? 
3. Did the objectives stated in the New York State World History and Geography standards use the same variables (cognitive processes) to express knowledge in reasoning or were those imposed by researchers?
4. Would a rival hypotheses be to use an equal sample size with the same characteristics (class/test subject, SES, age, etc...) but with a teacher found not to be teaching to the test standards? If the results were similarly accurate could this add to the test's validity?
5. Were there any events (history) that affected the outcome of the research?
6. In the conclusion a generalization is made that teachers tend to focus on literacy skills and content knowledge in their teachings. Is this drawn from the literature research and not from the actual qualitative study?

What Causes to My Effect?

Revised research problem stated as a prediction study within a correlational design:
Predictor variables:

1. exposure to natural light which includes an outdoor view including the horizon, expressed in hours/day
2. reason for incarceration, expressed in type of non-violent offense 
3. length of time already served, expressed in years

Outcome variable:

1. violent behavior with other inmates and prison officers, expressed in number of reported violent incidences per month

Population: Non-violent offenders in low-security prisons

Sampling Frame: 3 specific prisons, yet to be identified

Sample: sample of prisoners representing different races and ages

I think it's a natural desire to search for the cause to an effect. Understanding why and how things happen provides an explanation and the ability to somewhat control the future. If we can understand what caused an outcome then we have a greater chance of manipulating that variable in a similar situation to our benefit.


In research, however, the determination of a causal relationship as one of the primary research goals can limit the observations made or data collected (problems). If my research question involved only two variables - say, the effect of an inmate's exposure to natural light has on the number of violent incidences in which that inmate is involved per month - then I would end the research with data regarding quantities and types of natural light as well as reports of violent incident frequencies and types. If the research showed a positive correlation between natural light and lower violent incidents I could claim, based on this research alone, that there is a positive correlation. However, I think this would be an example of an accurate but invalid conclusion. 


If the research topic has been previously explored and it has been determined there is a causal relationship then the new research can prove or disprove this. My research is the initial study and has a large number of variables. This seems fairly representative of social science research. 


By using a correlational design model that is predictive I can include several variables that provide additional context to the overall research and inferences made. A regression model can communicate the role in which the predictor variables affected the outcome variable based on surveys, reports and interviews.


Questions:
Would an Ex-Post Facto, quasi-experimental design model, be appropriate given the number of "pre-existing"  conditions participants may have?
Is a regression model the only communication tool, to express research outcomes, in prediction research as I've described above? 

Validity + Reliability

Original research problem:


The purpose of this study is to understand the effects of prison architectural design on inmate rehabilitation in 2011.


Revised research problem:


The purpose of this study is understand whether the architectural design elements employed in several prisons were perceived by prison staff and inmates as they were intended. 


The goal is to better understand how the architecture of juvenile detention facilities facilitate inmate rehabilitation.

• architecture - space planning, volumetric changes / associations, ventilation, visual and physical access to natural light and other inmates / staff / community, materiality
• facilitates - enables or hinders the activities of rehabilitation
• rehabilitation - improvement in emotional and mental well-being that enables an inmate to function both in prison and out of prison in a way that doesn't harm him/herself or society

Measures


Specific architectural elements can be evaluated to determine their level of success by comparing data collected through architectural team interviews to data collected through prison staff / inmate surveys. The instruments or measurements would be interviews (and therefore the interviewer as an observer) and surveys.


Prisoner rehabilitation can be evaluated through post-prison release supervision statistics and surveys administered to inmates having served a majority of their term in a specific facility. The instruments would be surveys and credible sources. Would these credible sources be a part of the literature review or introduced in the body of the research with other instruments of measure?


These instruments are being used to measure the following variables:

1. the architectural team's design intentions/strategies (opinions and documents)
2. the inmate's reactions to the design strategies (opinions and statistics)
3. the prison staff's reactions to the design strategies (opinions and statistics)

Validity + Reliability


Based on the definition of qualitative research:

"Qualitative  research uses a naturalistic approach that seeks to understand phenomena in
context-specific settings, such as "real world setting [where] the researcher does not attempt to
manipulate  the  phenomenon  of  interest"  (Patton,  2001,  p.  39) (The Qualitative Report, p. 600)


Reliability will be based on consistency in conducting interviews and administering surveys as well as the content of both measures. Specificity of the architectural elements in questions and effects being measured are important in providing data that can be compared between one or more sample groups from different prisons/different architectural teams.


Validity requires me (the researcher) to define the quality concepts being explored through the research. 
I'm still defining these but at this time I think the concept being explored is :
natural light as a tool to discourage violent behavior 
If I compare the amount of natural light (measured in square feet of glazing per cell / most consistently occupied space) accessed by inmates to the number of violent occurrences documented for those inmates - a possible correlation can be made. The challenge I am now facing is limiting variables like reason for incarceration (some inmates are more violent than others prior to incarceration), overcrowding, etc...


Resources:

The Qualitative Report Volume 8 Number 4 December 2003 597-607 
http://www.nova.edu/ssss/QR/QR8-4/golafshani.pdf

Research Ethics

In educational research, the most important ethical points to consider are those that concern the participants well-being, information regarding the participant's role in research and research goals. Full disclosure of the researcher's goals provides the participant with context surrounding his or her role.  Age can determine how and to whom the information should be disclosed. If the sample group contains students that are children or college-age, an Internal Revue Board may review the research project prior to allowing student participation. It is also important to consider participant's families and their opinions regarding the research topic, methods and goals. The need to disclose information to the research sample group, the group's families, and the organization through which participation is gained makes using minors as a sample group unique. 


Another important ethical issue, specific to educational research, is peer pressure. Whether the sampling frame is an entire school or just one class, a student and their family may be uncomfortable as the only subjects not participating in the research. The researcher needs to consider whether anonymity is required to prevent coercion or peer pressure. Anonymity can be extended to a participants status in the research as well as the outcome of their participation.


Unless considered necessary, incentives should be avoided. Although it is difficult to eliminate bias, not offering incentives allows potential participants to base their decisions on the merit of the research, not the incentive.